Roundtable 2-12 on “Democracy and Victory”

In the following exchange Dan Reiter defends his argument that democratic states win most of the wars that they fight primarily because they choose which wars to engage in more carefully than authoritarian states do.[1] This is called the “selection effects” explanation because democracies are selecting which wars to fight and which to avoid. Here, Reiter is replying to previously published criticism by Michael C. Desch and Alexander Downes that detailed examinations of several historical cases that Reiter cites do not in fact support his arguments.[2] Desch and Downes respond and then Reiter has a rebuttal.  They primarily debate both how historical evidence should be interpreted and how their hypotheses should be evaluated in the 1920 Russo-Polish War, the 1956 Sinai War, the 1967 Six Day War, the 1982 Lebanon War, and the 1965 escalation of the Vietnam War.

H-Diplo | ISSF Roundtable, Volume II, No. 12 (2011)

Diane Labrosse and Thomas Maddux, H-Diplo/ISSF Editors
George Fujii, H-Diplo/ISSF Web and Production Editor
Commissioned for H-Diplo/ISSF by Christopher Ball

H-Diplo/ISSF Exchange on Democracy and Victory {a H-Diplo | ISSF Roundtable}

Published by H-Diplo/ISSF on 1 July 2011

Roundtable editors:  Christopher Ball and Diane Labrosse
Introduction by Christopher Ball

PDF-  http://issforum.org/ISSF/PDF/ISSF-Roundtable-2-12.pdf

Contents

Introduction by Christopher Ball, Loyola University Chicago

In the following exchange Dan Reiter defends his argument that democratic states win most of the wars that they fight primarily because they choose which wars to engage in more carefully than authoritarian states do.[1] This is called the “selection effects” explanation because democracies are selecting which wars to fight and which to avoid. Here, Reiter is replying to previously published criticism by Michael C. Desch and Alexander Downes that detailed examinations of several historical cases that Reiter cites do not in fact support his arguments.[2] Desch and Downes respond and then Reiter has a rebuttal.  They primarily debate both how historical evidence should be interpreted and how their hypotheses should be evaluated in the 1920 Russo-Polish War, the 1956 Sinai War, the 1967 Six Day War, the 1982 Lebanon War, and the 1965 escalation of the Vietnam War.

Beyond the specific cases, the debate over democratic victory – whether it exists and why it occurs – has broader implications. The 2011 “Arab Spring” has renewed questions about how domestic politics affect foreign policy. If Reiter is correct, the new democracies should be more prudent in their use of force than their authoritarian predecessors were. If Desch and Downes are correct, we should see little change in their predilection to initiate wars. One issue in this debate is how well we should expect democratic institutional constraints to work for recently established democracies (Poland in 1920 and Israel in 1956) versus older democracies (Israel and the United States).

As readers might expect, this exchange has not resolved the disagreements between the participants. It does, however, clarify the terms of the debate and suggest avenues for future research.

Participants:

Michael C. Desch is Professor and Chair of the Department of Political Science at the University of Notre Dame.  In addition to Power and Military Effectiveness:  The Fallacy of Democratic Triumphalism (Baltimore: Johns Hopkins University Press, 2008), he is the author of When the Third World Matters: Latin America and U.S. Grand Strategy (Baltimore: The Johns Hopkins University Press, 1993) and Civilian Control of the Military: The Changing Security Environment (Baltimore: The Johns Hopkins University Press, 1999).

Alexander B. Downes is Assistant Professor of Political Science at Duke University. He is the author of Targeting Civilians in War (Cornell University Press, 2008), which won the Joseph Lepgold Prize given by Georgetown University for best book in international relations published in 2008. Downes has published several articles and book chapters on the causes and effectiveness of civilian victimization in wartime, the military effectiveness of democracies, and the compatibility of covert intervention by democracies with theories of democratic peace. He is currently working on a book project on the domestic and international consequences of foreign-imposed regime change.

Dan Reiter is chair of the political science department at Emory University.  He received his Ph.D. in political science from the University of Michigan in 1994, and was an Olin post-doctoral fellow in Security Studies at Harvard University in 1994-1995.  He is the author of several scholarly articles, as well as Crucible of Beliefs: Learning, Alliances, and World Wars (Cornell, 1996), Democracies at War (Princeton, 2002; coauthored with Allan Stam) and How Wars End (Princeton, 2009).  He won the 2002 Karl Deutsch Award given by the International Studies Association to the leading international relations scholar under the age of 40 or within ten years of receiving his or her Ph.D.  How Wars End won the 2010 American Political Science Association Award for Best Book in Conflict Processes, and was recognized in January 2011 by Choice as an outstanding academic title of the year.

Dan Reiter, Emory University

“A Closer Look at Case Studies on Democracy, Selection Effects, and Victory”

Are democracies more likely to win the wars they fight?  This question has been of interest to historians and philosophers since Thucydides.  During the Enlightenment, the question was highly relevant to the great issues of the day, as thinkers such as Thomas Paine wondered how emerging republics like the United States and France would fare in war against monarchies.  It reemerged in the twentieth century, when some worried whether the Western democracies had the stuff to stand up to Nazi Germany and its fascist allies.  After World War II, Westerners fretted that an American Athens would ultimately fall short against a Soviet Sparta.

Allan Stam and I explored whether democracies win their wars in our 2002 book, Democracies at War.  In it, we examined the historical record back to 1815 and found strong empirical evidence that not only are democracies not disadvantaged in war, but actually they are more likely to win their wars than are other kinds of states.  We explored a variety of different explanations for why democracies win their wars, and ultimately found strong support for what we termed the selection effects explanation.  This theory posits that democracies are especially likely to win the wars they start, for two reasons.  First, elected leaders are likely to initiate wars only when they are very confident they will win.  If an elected leader starts a war the country goes on to lose, then that leader faces a high risk of losing office.  The prospect of losing office discourages an elected leader from initiating wars other than those she is highly confident she can win, meaning that when democracies initiate wars they are very likely to win.  Conversely, if an autocrat starts an ultimately unsuccessful war, he can use the tools of repression to stay in power even in the face of popular discontent.  Knowing that losing an initiated war does not present great risks of getting thrown from power, autocrats are more willing to initiate wars when they have a lower chance of winning, meaning that autocrats win their initiated wars less often than elected leaders win their initiated wars.

Second, elected leaders enjoy higher quality information about whether the country will win a war.  The marketplace of ideas provided by open societies enjoying freedom of the press and speech produces better public information about policy issues.  Also, the lesser politicization of civil-military relations in democracies means that military leaders provide better advice to the civilian leadership, both because military leaders in democracies are more likely to have been promoted on the basis of merit rather than political reliability, and because democratic military leaders are more confident that they can speak freely without risking personal punishment for having told the civilian leader something he or she does not want to hear.  Higher quality information means that democratic leaders can assess more accurately than autocratic leaders whether a possible war will go well or poorly, meaning that democracies will be better than autocracies at identifying winnable wars, and hence will be more likely to win the wars they initiate.

To our delight, our book has attracted substantial scholarly attention.  The book inspired debate over issues of theory, methodology, and quantitative empirical analysis.  It has also attracted attention over the application of the theory to individual case studies.  This paper addresses some of these case study critiques.  It avoids methodological debates about whether failure to predict individual cases can falsify a probabilistic theory.  Instead, it considers the specifics of the cases, assessing the critics’ claims that these cases provide evidence against selection effects theory.  Specifically, it assesses four of the cases Michael C. Desch examines in his 2008 book, Power and Military Effectiveness: the 1920 Russo-Polish War, the 1956 Sinai War, the 1967 Six Day War, and the 1982 Lebanon War.  It also evaluates the arguments regarding the Vietnam War made by Alexander Downes in his 2009 International Security article.[3]  These two scholars contest some of the independent and dependent variable codings in these cases, and also posit that some of these cases do not demonstrate the processes predicted by selection effects theory.  For reasons of space, the goal of this paper is not to present full blown case studies of all five cases, but rather to address the specific critiques within each case offered by Desch and Downes.

Closer examination of these cases reveals that they offer far greater support for selection effects theory than the critics have allowed.  Consistent with selection effects theory, in all cases the decisions to attack were popular and were made with the approval of the political oppositions.  In the Russo-Polish War, Sinai War, Six Day War, and Lebanon War, elected leaders initiated wars they went on to win.  The Vietnam War is perhaps an exception that proves the rule: it is a case in which a democracy (arguably) initiated a war it was only moderately confident it would win. Selection effects theory both explains the decision to initiate under these conditions as well as a number of other aspects of how the war was fought.

Though the cases generally provide support for selection effects theory, they do point to two aspects of the theory that need further scholarly attention.  First, though elected civilian leaders often enjoy higher quality advice from their military leadership, civilian decision-making can be hampered when civilian leaders are insufficiently informed about military affairs, and rely heavily and uncritically on the recommendations of military leaders.  The Lebanon War demonstrates this dynamic most clearly.  Second, secrecy poses a real dilemma for the theory.  Sometimes planning for war must be taken in secret to preserve military advantage.  However, planning in secret runs counter to one of the predictions of the theory, that in democracies open debates precede decisions for war, and these debates help better inform the leader’s decision-making, decreasing the chances that the democracy will initiate a foolish war.  The paper provides a more extensive discussion of the role of secrecy for selection effects theory, considers its presence in several of the cases, and offers suggestions for future research on this topic.

SEARCH FOR VICTORY: DEMOCRACY, INITIATION, AND WAR OUTCOMES.

Many have proposed that democracies do not fight each other because publics hate war, elected leaders must answer to publics, and such leaders are unlikely to initiate war and risk getting removed from office.  Tyrants need not answer to their publics, and hence are freer to start wars.[4]  Scholars built on this basic insight to understand how democratic political institutions might shape other aspects of conflict behavior.  One idea posited that if the public’s opposition to war was based on aversion to paying the costs of war, as opposed to a normative opposition to war, then public opposition might relax if war promised to bring benefits at acceptable costs.  A simple proposition from this insight is that an elected leader is less likely to suffer domestic political punishment when he or she initiates a war that the state goes on to win.  The testable proposition is that democratic leaders are likely only to start wars when the chances of winning are high, meaning that among the observed wars democracies initiate, democratic initiators are quite likely to win.  Autocratic leaders are less fearful of being thrown from power if they start wars and lose.  Hence, autocracies are more likely than democracies to start risky wars, so autocratic initiators are less likely to win their wars than are democratic initiators.  A robust marketplace of ideas means that democratic leaders are likely to make more accurate guesses about whether or not they can win wars they are considering launching.  As noted, Stam and I developed selection effects theory in our book, and the rigor of the logic has since been demonstrated formally.[5]

Selection effects theory has attracted a substantial body of quantitative empirical support.  We and others found that since 1816 democracies and democratic initiators in particular have been more likely to win their wars.[6]  Many other quantitative studies have offered support for the idea that democratic political institutions shape foreign policy behavior, and that institutions guide leaders to avoid costly, high risk military ventures.   Quantitative empirical studies have found that democracies fight significantly shorter wars,[7] and that democracies win because they start short wars.[8]  Relatedly, democratic initiators are especially likely to win interstate crises,[9] and crises involving democracies are significantly shorter in duration.[10]  Democracies become increasingly likely to initiate conflict as the balance of power becomes increasingly favorable.[11]  Public support for war declines as casualties accumulate and as the prospects for victory recede.[12]  Hein Goemans’ latest and most sophisticated research on the post-conflict political fates of leaders offers further support.  He finds that elected leaders who lose wars face significantly higher chances of losing office through extralegal means, thereby facing higher likelihoods of severe personal punishments such as death, prison, or exile.[13]   And of course, the general idea that political institutions constrain the conflict behavior of elected leaders underpins the widely verified empirical finding that democracies are unlikely to fight each other.  Notably, there have been some critiques of the quantitative findings that democratic initiators are especially likely to win their wars.  We in turn have published rebuttals to these critiques.[14]

CASE STUDIES AND SELECTION EFFECTS THEORY

The quantitative support for selection effects theory is impressive.  However, some scholars have presented case studies critiques of selection effects theory, arguing that in some instances the key variables of democracy, initiation, and/or victory were miscoded in the quantitative data sets, meaning that statistical analysis provided inaccurate correlative support for selection effects theory.  They also propose that process tracing reveals that the dynamics predicted by selection effects theory did not actually occur.  Desch (6) discussed the Russo-Polish War, the Sinai War, the Six Day War, and the Lebanon War, cases commonly thought to be of democratic initiators winning their wars and which “are instances of democratic states prevailing when the most likely alternative explanation—material power—would have predicted they would have lost.”[15]  Downes describes why he chose the Vietnam War as a case, though conceding that Vietnam might be an outlier: “In Vietnam, however, we observe a democratic leader entering a war even though he knew the odds of victory were slight.  …[I]n certain circumstances democracy may propel leaders to engage in wars with a low probability of success, this causing democracies to suffer the occasional draw or loss.  I do not claim these circumstances are common.”[16]

Before getting into the cases themselves, three methodological issues require discussion.  The first concerns the difficult task of coding the dependent variable, war outcomes.[17]  The conventional approach taken by many data sets, including the one used by Stam and I,[18] is to code which side better achieved its military operational goals by the end of the war, allowing for the possibility of a draw.  An alternative approach might be to code a state as winning a war if the war provided that state long term strategic benefits, operational success aside, and/or to frame operational success in the context of costs suffered.  Some have argued that some wars coded in the quantitative data sets as victories for democracies did not provide longer term benefits, and in line with the logic of the second, “long term effects” approach, ought to be recoded as draws or losses.

It is much more difficult to reach consensual codings using this latter approach than it is using the operational goals approach.  Coding whether or not an army is destroyed or which side conquered territory is much easier than coding long term strategic gains or losses.  Further, assessing whether a war provided long term geopolitical costs or gains can become an ideologically charged debate.  For example, leftists claim that the Six Day War was a disaster for Israel because it created the problems of the occupied territories, while rightists claim that it was a success because it boosted Israeli deterrence and territorial security.

A related issue to the ideological problem is that judging long term benefits may turn on one’s definition of “long term.” A war may be deemed net beneficial five years out, net costly ten years out, but net beneficial twenty years out.  In December 2003 following the capture of Saddam Hussein and before the flowering of the insurgency, the Iraq War seemed like a benefit to American interests.  In 2005, when no weapons of mass destruction were found and the country had slid into civil war, it seemed like an unmitigated disaster.  If Iraq emerges by 2014 or so as a reasonably democratic, reasonably stable, reasonably pro-American oil exporting country, then many may view the 2003 war as net beneficial to American interests.

Again, some critics of the democracy-victory thesis claim that the long term approach makes democratic military successes appear more modest, as, for example, one should code the Sinai and Lebanon Wars as draws for Israel rather than victories (see below).  However, embracing this standard would change many other war codings, including converting wars which are autocratic victories under the operations approach to autocratic defeats under the long term strategic approach.  Examples include: German victories over Belgium and Russia in World War I; German victories over Norway, France, Poland, Belgium, the Netherlands, Greece, and Yugoslavia in World War II; the 1990 Iraqi victory over Kuwait; the Soviet victory over Finland in the Winter War.  Hence, embracing this alternative approach to coding war outcomes would not necessarily overturn the systematic finding that democracies win the wars they initiate (or fight).

A second methodological issue concerns what kinds of evidence case studies examining selection effects theory should look for.  The heart of selection effects theory is that elected leaders seek to avoid launching unpopular wars because they want to remain in power.  The central hypothesis is that because losing wars will likely be unpopular, elected leaders are only likely to launch wars when they are confident that they will win.  The theory, as described by Stam and I, also allows that leaders are more likely to launch wars for popular causes, and relatedly that publics gripped by war fever may drag a reluctant leader into war.[19]  So, the kinds of processes envisioned are that elected leaders will launch wars when they are confident they will win and when they think the wars will be popular.  Leaders may vet the popularity of a potential war by examining public opinion data or by seeking approval for a war within the government, either by consulting with the cabinet or the elected legislature.  Approval by the opposing party of military action can be especially important in winning public support for the action, as it is a powerful signal.[20]

One important question is whether the theory predicts that an elected leader will make decisions for war in consultation with society and/or other components of government.[21]  The theory makes no prediction that elected leaders are driven by normative imperatives to engage in open, democratic debate of policy options.  Indeed, Stam and I discuss a number of instances in which elected leaders subverted democratic norms, for example by undermining elected governments in other countries.[22]  However, the theory does predict that elected leaders are motivated to include other actors in the decision-making process in order to provide themselves political cover, to gather evidence of and to demonstrate public support, and to minimize political fallout if things go poorly.  For example, the January 1991 Congressional votes in favor of Operation Desert Storm provided critical reassurance to President George H.W. Bush, who reported that, “I felt the heavy weight that I might be faced with impeachment lifted from my shoulders as I heard the [Congressional voting] results.”[23]

Stam and I address the issue of democracy and secrecy in our book.[24]  We discuss it in the context of covert action, in which elected leaders sometimes take steps to overthrow foreign governments secretly, in order to circumvent the constraints of public opinion.  We argue covert actions such as the Bay of Pigs invasion often fail because secrecy increases the risks of policy failure by cutting out the marketplace of ideas.  That is to say, we identified an internal tension between different components of our theory.  Though the theory declares that the constraints of public opinion and the marketplace of ideas increase the likelihood that democratic foreign policy initiatives are more likely to succeed, leaders in some cases may be motivated to circumvent public constraints by acting covertly.  However, acting covertly can have the effect of shortcircuiting the marketplace of ideas and thereby increasing the risk of policy failure.

Unlike covert action, secret deliberation over launching an overt war does not circumvent public constraints.   The war will be common knowledge once launched, and if it goes badly, the elected leader will still suffer adverse domestic political consequences.  Conversely, leaders may have more confidence that ‘plausible deniability’ will enable them to escape or mitigate the negative consequences of a failed covert plot.  Therefore, deliberating in secret should not make elected leaders more willing to launch riskier wars.

However, secret deliberation before overt war does prevent the marketplace of ideas from operating fully.  Though when deciding secretly on war elected leaders still benefit from higher quality advice from their military leadership than do autocratic leaders, they do not benefit from extensive public and government discussion of a proposed war, and hence are less likely to make an accurate guess about the outcome of the war.  That being said, in the context of making decisions for military attack, there may be important benefits of making the decision in secret, as secrecy may make victory more likely.  A plan for victory may rest on the tactical advantages of a surprise attack, an innovative military strategy, a new military technology, and/or secret diplomatic arrangements.  Revelation of any of these factors through public discussion or leaks may make victory less likely.  Keeping the decision-making process secret and contained to a small group decreases the chances of leaks, and may under these conditions increase the chances of victory.

In summary, examples of leaders deliberating in secret have complex implications for the theory.  First, selection effects does not forecast that elected leaders have a normative imperative to consult broadly.  Second, military or diplomatic conditions may push elected leaders to keep the decision-making group as small as possible to avoid leaks and increase the odds of victory.  The desire to maximize the chances for victory through secrecy is consistent with selection effects theory’s premise that elected leaders have a very strong incentive to avoid military defeat.  Third, even when elected leaders need to maintain secrecy, they may include at least some members of the political opposition in the decision-making process to maximize political support and minimize possible political fallout.  Fourth, though secrecy may mean gathering some military advantages, it also means increasing the risks of policy failure by cutting out much of the marketplace of ideas.

A third methodological issue concerns the pertinence of evidence about what material factors (such as the balance of power, allies, and military strategy) helped determine a war’s outcome.  Importantly, such factors do not directly speak to the empirical support for or against selection effects theory.  Selection effects theory predicts that an elected leader is more likely to attack when a war seems winnable, regardless of exactly why the war seems winnable.  So, observing that one side won because it had a larger army does not provide evidence against the selection effects theory, as it may just mean that the elected leader correctly observed before the war that his/her country’s numerical superiority would permit victory.  Note that in the quantitative studies, democratic initiators are still more likely to win even when controlling for material factors such as strategy, terrain, troop quality, military-industrial capabilities, and allies.[25]

THE RUSSO-POLISH WAR

In April 1920, Polish forces invaded Ukraine, ultimately defeating Soviet Russian forces and acquiring Russian territorial concessions in the 1921 Treaty of Riga.  The 1920 Russo-Polish War is, according to the quantitative data sets, an example of a democratic initiator, Poland, winning a war against an autocracy, Russia.  Desch makes a number of critiques of this interpretation.  They are addressed in turn.

Did Poland Initiate the Russo-Polish War?

We and others argued that Poland initiated the Russo-Polish War in April 1920.[26]  Desch argues that either Soviet Russia started the war before 1920, or it is not clear exactly which side started the war.

Low-level hostilities between Russia and Poland did commence prior to April 1920.  However, there are two problems with the claim that one should code Russia as initiating war before 1920: Poland initiated the pre-1920 violence, and the pre-1920 violence was relatively minor, far below the threshold of what is commonly thought of as “war” (armed conflict killing at least 1000).  Proving these points requires working through the specific events Desch describes in support of his critique.[27]  In arguing that war-scale violence predated April 1920, Desch refers to Russo-Polish clashes near Bialystock and Brest-Litovsk in December 1918.  However, these conflicts were very limited in scope, as the presence of German forces prevented any substantial violence at that time.[28]  Desch (74) refers to a conflict between Russian and Polish forces at Bereza Kartuska in February 1919.  However, Polish forces initiated military conflict by entering this small township.  Further, this was a very minor incident involving only 62 Polish soldiers capturing about 80 Soviet soldiers.[29]  The insignificance of this clash is demonstrated by the perseverance of political contacts between Poland and Russia.[30]  Desch refers to Poland “repulsing” Russian moves in the Baltics in March and April 1919 (p.11).  These events are clearly Poland initiating actions against Russia, not the reverse.  Russia had taken control of Wilno in sovereign Lithuania, and Poland reacted by dispatching forces to eject the Russian forces, accomplishing this mission in late April.[31]  Note that during this period it was Poland that took the initiative to seize territory in the turmoil of the Russian Civil War, capturing from February to October 1919 contested border areas with substantial numbers of ethnic Poles.  The Poles renewed their offensive when Russia demanded Polish evacuation of these areas.[32]  In his definitive history of the war, Norman Davies claims that in this first phase “the initiative lay with the Poles.”[33]  This is not surprising, as the Russians wanted to avoid conflict with Poland to focus on fending off foreign forces during its own civil war.[34]

Desch mentions Soviet Russia’s “Target Vistula” plan, declared in November 1918 (not 1919, as noted by Desch, 74).  Despite its grandiose vision of a march to Warsaw, it amounted to nothing more than revolutionary rhetoric, the formation of the Soviet Western Army, and a reconnaissance in depth as far as the River Bug.  Tellingly, there were no notable military clashes in this operation, as Soviet forces were entering territory just evacuated by German forces and did not seek to engage Polish forces.  Davies wrote: “It is problematical whether …’Target Vistula’ was intended to bring the Red army as conquering heroes into Warsaw.  Its name suggests so.  Yet the extremely tentative phrasing of its directives and the extremely parlous state of the Western Army suggest otherwise.  ‘Target Vistula’ was probably no more than a phrase inspired by revolutionary bravado.”[35]  Another scholar notes, “The Russian Western Army had never been intended to fight its way to the Vistula: it was far too weak.”[36]

Desch notes (74) that “a more reasonable argument” is that it is not clear which side started the war, as Poland and Russia entered the vacuum created by the withdrawal of German troops.  It is probably accurate that both Poland and Russia were opportunists, seeking to make geopolitical gains following Germany’s withdrawal.  However, the violence that did occur before April 1920 was initiated by Poland and far below the level of interstate war.  The action which clearly escalated the conflict to the ‘war’ level of intensity was Poland’s April 1920 invasion with some 52,000 soldiers.[37]

Was Poland Democratic?

Desch argues that Polish leader Jozef Pilsudski was a dictator, unconstrained by public opinion, though multiple cross-national data sets on democracy deem Poland to be democratic at this time.[38]  Closer inspection reveals that Poland was a democracy, and that Pilsudski was a constrained leader and not an autocrat.

Poland emerged as an independent state in late 1918.  Pilsudski was given authority over the armed forces by the Regency Council when he arrived in Warsaw that November.  The Council gave him only temporary power, to be handed back to the national government when the latter was formed.[39]  Critically, “Jozef Pilsudski supported elections from his first moments in Warsaw.  He believed elections to be the only way of creating a centre of state authority which would be generally recognized in Poland and abroad—a factor of the society’s consolidation and integration.”[40]   As early as November 1918 Pilsudski’s power was constrained, when he was informally the provisional head of state.  His first nominee for prime minister, Ignacy Daszynski, had to be withdrawn because of opposition from the Right.[41]  Pilsudski instead (on November 18) appointed Jedrzej Moraczewski as prime minister, commanding him that his appointment demanded that “in the course of one week you will produce an electoral law, just as if you had a trench to dig.”[42]

Moraczewski followed Pilsudski’s encouragement, and the first elections to the Sejm legislature were held in January 1919.  The election was competitive, providing legislative seats to several different parties.[43]  In February, Pilsudski honored his commitment made to the Regency Council and handed power to the elected Sejm.  The Sejm voluntarily handed power back to Pilsudski, a temporary measure until a permanent constitution could be formed.  Importantly, Pilsudski did not enjoy dictatorial powers.  At the time of Pilsudski’s reappointment, the Sejm made a short declaration on its own legislative powers which came to be known as the “Little Constitution.”  This declaration constrained Pilsudski’s power.  It stated that the Sejm itself “embodies the sovereign and legislative authority of the State of Poland.”  The declaration gave Pilsudski the power to appoint the prime and cabinet ministers, but such appointments required the consent of the Sejm, and the Sejm could dismiss at any time the prime and cabinet ministers.  Pilsudski accepted the Sejm’s appointment subject to the Little Constitution.[44]

In sum, Poland at this stage can be described as being institutionally democratic.  Poland held elections.  There was a constitution which empowered an elected legislature and constrained the executive.  A separate democracy-related question is, was the war against Soviet Russia popular?  The critical theoretical assumption underlying the selection effects argument is that democratic leaders follow public opinion because they desire to sustain their domestic political careers.  This means, critically, avoiding unpopular wars.

Pilsudski’s foreign policy choices which led to war commanded the assent of the Sejm.  Following Soviet peace overtures in early 1920, the Foreign Affairs committee of the Sejm approved the text of a peace feeler on March 27, which proposed a time and place for the commencement of peace negotiations.[45]  As the negotiations with the Soviets bogged down, the Polish government proceeded to make a military alliance with Ukraine, laying out plans for coordination of their two armies.  This alliance was approved by the Foreign Affairs committee.[46]  Even Sejm members opposed to the war recognized that the majority favored war over the renewal of peace negotiations.[47]  The Warsaw correspondent for the Times of London noted at the time:

“It is suggested that the Government in demanding the disannexation of the whole of the ancient Kingdom of Poland from Russia and helping the establishment of an independent Ukrainian State…was pursuing a policy unacceptable to the country as a whole.  These assumptions were entirely refuted at yesterday’s session, and the Government’s policy was approved by practically every member of the Committee except the National Democrats.”[48]

Poland’s successes during the war were recognized by the Polish public, and Pilsudski’s popularity surged with these victories.  Even the National Democrats quieted their opposition.[49]  When Pilsudski returned to Warsaw on May 18 following initial battlefield successes, he was given a hero’s welcome and thanked formally by the Sejm.  The political opposition halted their attacks.[50]  The May Soviet counteroffensive led to a string of Polish defeats, as the Red Army approached Warsaw.  Importantly, as Polish military fortunes worsened, Pilsudski did not cast aside the cloak of democracy and become a dictator.  Twice before the Council of National Defense, on July 13 and on July 19, Pilsudski submitted his resignation.  Both times the offer to resign was rejected.[51]  Desch reports (85) that in August 1920 Pilsudski was concerned about declining public support for continuing the war, a pattern predicted by the selection effects theory, that support for continuing the war might ebb in the face of military setbacks.  The reversal of Poland’s military fortunes did lead to a reduction in Pilsudski’s power.  During this period, the Sejm convened a Council of National Defense, which included Pilsudski, several Sejm representatives, and several cabinet members.  The Council had the power to make all decisions on war-related matters, and its creation marked a “considerable slimming down of Pilsudski’s authority.”[52]  In short, as selection effects theory forecasts, there were negative domestic political repercussions for the elected leader in the wake of disappointments on the battlefield.  Eventual Polish victory helped Pilsudski stay in power.

Was Poland Confident In Victory Before the War?

Desch (73) states that “Pilsudski reportedly launched the attack on the Ukraine in the spring of 1920 despite serious doubts that it would succeed.”  The evidence does not clearly support this conclusion.  The article Desch cites indicates that though some British emissaries had doubts about the operation, many also noted Pilsduski’s confidence.  It quotes one February 1920 report: “General Pilsudski is confident that he can carry on the war [with Soviet Russia] for several months…”[53]  Pilsudski attacked in April because delays in Russian military mobilization created a Polish advantage in the balance of forces.[54]

Desch (73) also states Pilsudski should have forecast that the invasion of the Ukraine would inflame nationalism in the Red Army, boosting its military effectiveness.  It is perhaps unreasonable to assume that the Pilsudski should have seen that the invasion would spark Soviet nationalism.  The pro-Soviet nationalism which emerged following the Polish invasion was bizarre and unique.  It was bizarre because it was a call to Russian nationalism to defend Ukraine, and this nationalism was then wedded to a Bolshevik call for class consciousness and unity.  It was unique because this was the first appearance ever of this odd mix of “Russian nationalism and Soviet internationalism.”[55]  Indeed, even the Bolsheviks themselves were “stunned by the success” of using Russian nationalism to mobilize the population to fight the Poles.  That is, it is unreasonable to criticize Pilsudski for failing to foresee a development as shocking as the emergence of pro-Soviet Russian nationalism. [56]

A related question is the quality of prewar strategic evaluation in Poland.  Desch (84) echoed the argument by Vladimir Lenin that the Poles might have been better off not going to war.  The Treaty of Riga, which ended the Russo-Polish War, provided Poland with less territory than it would have received if it had accepted the terms offered by the Soviets in January 1920.  This notion of the war as ultimately one of Polish loss was a common propagandistic theme made by Soviet leaders and later by members of the antiwar minority.[57]

It would be inaccurate to portray the episode as one in which Poland foolishly refused a generous settlement, only to receive its comeuppance when the war it chose did not work out as well as hoped.  The true promise of the January 29 offer is dubious, as its terms were for an armistice and demarcation line, not for a border and peace treaty.[58]  Further, it is doubtful that the Soviets truly desired peace.  The Bolsheviks were more interested in subverting Poland from within, hopefully inciting a Communist revolt, rather than peacefully settling a border dispute.  Indeed, soon after the January offer the Soviets declared, “Polish soldiers, turn the bayonets which you hold in your hands against your masters and government…[and] then peace will be concluded.”[59]  Pilsudski had good reason to doubt the veracity of Soviet declarations of peaceful intentions.  He had received intelligence reports that since the end of 1919 the Soviets were concentrating their forces in Belarus and Ukraine, quintupling their forces in the west from January to April 1920.[60]

Regardless, the Poles did not initially reject the Soviet offer, as on March 27 they suggested negotiations commence on April 10 in Borisov.  Wrangling over the specifics of the negotiations ensued, preventing actual talks from taking place.  Importantly, in the post-March 27 bargaining over negotiating terms, the Soviets indicated that the January 29 offer was no longer on the table.[61]

The Poles were reluctant to strike a deal with the Soviets.  Their concern was that accepting a peace treaty would leave the Soviet army intact, and since the critical phase of the Russian Civil War was passing, Soviet strength would grow in the future, eventually leading the Soviets to attack.[62]  Specifically, a peace treaty at that juncture would have permitted the Bolsheviks to crush the last of the opposition in the Ukraine as well as the forces under General Petr Wrangel.[63]  This is consistent with the more general proposition that war is one solution to a credible commitment problem created by a changing balance of power.[64]

ISRAEL’S INITIATED WARS

Desch presents a cluster of case studies of Israel’s post-independence wars.[65]  Desch claims that three wars Israel initiated, the 1956 Sinai War, the 1967 Six Day War, and the 1982 Lebanon War, do not support selection effects theory.  The remainder of this section examines these claims.  It then discusses the effects of war outcomes on leader tenure in the Middle East during this period.

1956 Sinai War

In 1955 and early 1956, Britain and France became increasingly worried about Egyptian leader Gamal Abdel Nasser’s nationalism, especially following Nasser’s seizure of the Suez Canal.  During this time, Israel was suffering from Egyptian and fedayeen attacks launched on southern Israel from bases in Gaza and northeastern Sinai.  Israel worried that a 1955 arms deal with Czechoslovakia would significantly expand Egyptian military power.  In 1956, Britain, France, and Israel agreed on a plan.  Israel would invade Sinai and Gaza, and the invasion would then provide a diplomatic pretext for Anglo-French forces to land in Egypt and seize control of the Suez Canal.  The three countries executed the plan later that year.  Israel captured Gaza and Sinai.

Was Israel Democratic?

Desch views Israeli wartime decision-making throughout its post-independence period as insulated and autocratic, though systematic data sets view Israel as democratic since independence.[66]  Israeli institutions provided for fair, regular, and competitive elections in a multiparty, parliamentary framework, safeguarded by the rule of law.  Across the 35 year period from 1948 to 1983, Israel experienced nine different leadership transitions, all peaceful and legal.  No single party dominated the array of ruling coalitions.

Desch’s depiction of Israeli decision-making before the Sinai War as autocratic is exaggerated.  Israeli Prime Minister David Ben-Gurion was more constrained than Desch allows.  This is demonstrated in part by Israel’s earlier decisions not to attack Egypt.  As Defense Minister, Ben-Gurion had suggested to Prime Minister Moshe Sharett in March 1955 that Israel should seize the entire Gaza Strip; Sharett refused.  When Nasser announced in September 1955 a massive Czech-Egyptian arms deal which promised to augment substantially Egyptian military power, Chief of Staff Moshe Dayan wanted to launch a preventive invasion immediately. Prime Minister, Ben-Gurion demurred on Dayan’s decision, but did support the use of force to open the Straits of Tiran.  However, no action was taken because the Israeli cabinet rejected the plan.  Ben-Gurion suggested to his cabinet in January 1956 the possibility of an Israeli preventive attack on Sinai in advance of the arrival of Czech arms, but again opposition within his cabinet killed the idea.[67]  This string of decisions not to invade also speaks to Desch’s (124-5) undocumented assertion that “many Israelis believed that Ben-Gurion’s Suez gambit was not carefully thought out,” as clearly the Israeli leadership had been considering multiple scenarios for offensive action against Egypt for more than a year.

Political opportunity presented itself later that year for an Israeli solution to the growth in Egyptian military power.  The French and British suggestion for a joint attack on Egypt came with the arrival of a large, secret shipment of French arms to Israel in July 1956.  On October 24, Dayan briefed Peres and Ben-Gurion on a plan for invasion of the Sinai.[68]  The cabinet, which included members of the political opposition, approved the plan soon after.  After the vote had been taken, more members of the political opposition were informed.  All opposition parties supported the operation, other than the Communists.[69]  Israel attacked on October 29.

Consistent with selection effects theory, the decision to attack was popular.  The rising Egyptian threat had been turning Israeli public opinion more belligerent since 1955, as indicated by the July 1955 elections which replaced the moderate Sharrett with the more hawkish Ben-Gurion, and increased the number of Knesset seats held by the nationalist Herut party.[70]  The September Czech arms deal further increased public support for war.  When the Knesset met in mid-October 1955 and discussed the Czech arms deal, almost all saw the pending growth in Egyptian power as a grave threat to Israeli security.  Many called for action, though at that point, before French arms had been secured, the calls for action frequently centered on the acquisition of more and better arms rather than for immediate preventive war.[71]  The ongoing fedayeen attacks on Israeli territory from Sinai and Gaza were escalating public insecurity, incubating public criticism against the Israeli government for its inaction.[72]  Public support for action was evident in Israeli newspapers and public statements by political leaders.  Editorials in Ma’ariv, the leading Israeli daily, warned of Egyptian aggression.  The head of the Centrist Liberal Party Peretz Bernstein argued for preventive war, as did Davar, the moderate voice of the Labor movement. [73]  Once the war began, there was “insignificant” dissent, coming only from the tiny Arab-Jewish communist party, Maki.[74] Consistent with selection effects theory, Ben-Gurion favored preventive war as a means of reducing Israeli casualties, as a preventive war would allow Israel to destroy enemy bombers on the ground before they could be used to attack Israeli civilians.[75]

Secrecy and the Decision for War

Desch (101) proposes that the secrecy surrounding the preparation for war provides two critiques of selection effect theory.  First, it exemplifies a non-democratic decision-making process.  Second, it is “evidence” that Ben-Gurion knew such a war would be unpopular.

Ben-Gurion kept war planning as secret as possible, because secrecy maximized the chances for victory.  As discussed above, selection effects theory does not necessarily predict that elected leaders will make decisions for war publicly, and that certain conditions may push elected leaders to decide in secret in order to maximize the chances for victory.  In 1956, the plan to attack Egypt hinged on maintaining secrecy.  The whole operation was predicated on the diplomatic ruse that Israel would attack Egypt, and then France and Britain by previous agreement with Egypt would deploy their forces to “protect” the canal.  Exposure of the prewar secret arrangement between Israel, Britain, and France would have jeopardized the entire operation.  Further, secrecy was critical for battlefield success, especially the commando operations undertaken at the beginning of the war.  General Moshe Dayan bluntly stated that, “Secrecy was imperative.”[76]  However, domestic politics did frame the decision-making.  As noted, Israeli society and politics had shifted in a more belligerent direction since 1955, and Ben-Gurion vetted the decision for war both with his cabinet and with members of the political opposition.  Notably, secrecy does not imply that Ben-Gurion thought the war would be unpopular.  The sources that Desch employs do not make this point.[77]

Did Israel Win the Sinai War?

A last critique Desch makes is that Israel did not win the war.  He declares (96) that Israel scored a tactical success, “but was denied a strategic victory when the United States forced it and its allies to withdraw from captured Egyptian territory.”  This point invokes methodological issues, discussed above, regarding whether a war outcome should be coded on the basis of operational or grand strategic success.

However, even using a longer term, grand strategy yardstick for determining who wins wars, the war yielded important gains for Israel, its withdrawal from the Sinai notwithstanding.  The war substantially improved Israeli relations with Britain and France, which had the specific benefit of providing Israel with a steady flow of French munitions.  U.S.-Israel relations improved in the wake of the war.  UN peacekeepers stayed in the Sinai after the Israeli withdrawal, providing improved security for southern Israel from Egyptian and fedayeen attacks.  Israeli military successes improved Israel’s military reputation, boosting its deterrent.  Israeli military successes dissuaded Nasser from getting involved in conflicts between Israel and Syria.  Israel was able to use Eilat as a port and acquired critical maritime access to the Gulf of Aqaba and thereby the Indian Ocean.[78]

The Six Day War

Israel launched the Six Day War on June 5, 1967 in reaction to Egypt’s decision to close the Straits of Tiran, and because of rising fear that an Arab attack loomed.  Israel defeated the militaries of Arab states including Egypt, Jordan, and Syria.  It captured Sinai, the Gaza Strip, the Golan Heights, the West Bank, and East Jerusalem.

Desch’s central critique is that Israeli decision-making was not democratic.[79]  He claims that as with the Sinai War Israeli decision-making was highly autocratic even within the inner circle.  He (100) specifically posits that Prime Minister Levi Eshkol was forced to take Dayan and Peres into his government, and that, “This government reshuffling was hardly the result of normal democratic procedure.”  He quotes Levi Eshkol’s wife Miriam Eshkol as declaring, “It was a real putsch.  Everyone was worried and nobody cared about democratic processes.”

There are some difficulties with this claim.  Miriam Eshkol’s comment was a reaction to the vehement criticism that Levi Eshkol received after he made a rambling and discouraging public speech on May 28, 1967, as the crisis reached a fever pitch.  Right after the speech, he met with military leaders in private.  The military leadership lambasted him, urging him in the strongest terms to take the country to war.  The Miriam Eshkol quote comes from Michael Oren’s book, from an interview she gave decades after the war.  It is not surprising that she would defend her embattled husband, and notably Oren does not agree with her assessment.  After presenting her quote, Oren notes that the military officers were expressing their opinion to the prime minister in private, and not attempting to seize power.  Importantly, the military was not alone in its opposition to Eshkol’s attempt at moderation.  Israeli newspapers attacked Eshkol for his weakness, some demanding that he step down.[80]

Desch’s proposition that Peres was forced into government as part of the rush to war is not accurate.  Peres was a member of the Rafi party, which belonged to the ruling coalition before the crisis emerged.  Representing Rafi, Peres sat on the Ministerial Defence Committee.[81]  Peres did not get promoted during the crisis.  Further, he made largely unheard suggestions for resolving the crisis without war which diverged from the opinions of those who demanded an immediate preemptive attack.  For example, he suggested on June 2 that Israel conduct a nuclear test as a means of restoring Israeli deterrence and averting war.[82]

Like Peres, Dayan represented the Rafi party on the Ministerial Defence Committee.  Dayan was promoted to Defense Minister through democratic means.  Golda Meir, a civilian and general secretary of the Mapai party (the largest member of the coalition), asked Dayan to serve as deputy prime minister on defense matters, though Dayan instead ended up with the defense portfolio.[83]  The appointment of Dayan, the war hero of 1956, to Defense Minister was widely and wildly popular.  The appointment was a great relief to many, and the announcement led to a cancellation of a planned massive protest in Jerusalem.[84]  Eshkol was forced to accept Dayan into his government not because of a feared coup, but because Dayan was so popular.  As Colonel Israel Lior noted, “Too many ministers, too many members of Knesset, too many generals, and the street, always the street, supported Dayan.”[85]

Desch claims (101-2) that Israel attacked not out of optimism for victory, but rather out of fear that the alternative was the annihilation of the state.  It is true that Israel’s decision to attack was greatly encouraged by a sense of overwhelming threat to its national security.  However, Israel did not view the attack on Egypt as a long shot launched in desperation, as Japan viewed the attack on Pearl Harbor in 1941.[86]  The Israeli military establishment was very confident in victory, and conveyed this confidence to the political leadership.  Desch himself concedes that “Rabin, Yariv, and their colleagues in the Israeli Defense Forces (IDF) were ‘spoiling for a fight’.”  An article cited by Desch agrees: “Israel’s military leaders, while certainly concerned by the situation in general, were also unequivocally convinced that they would be victorious in the forthcoming war, even if the Arabs did attack first.”[87]  More importantly, the political leadership became convinced that an attack could succeed, especially after Dayan became Defence Minister and approved a more aggressive plan of attack, and after the United States began to signal that its opposition to attack was waning.  When the final decision for war was made, the cabinet voted 12-2 in favor of attacking.[88]

Desch also claims (100) that Dayan as defense minister “ordered the assault on the Golan Heights without securing the prime minister’s approval.”  Desch uses this as evidence that Dayan eschewed democratic processes, perhaps evincing an image of a reckless defense minister unconstrained by an elected leader.  The reality is much more complex.  After the war started, Eshkol was more in favor of attacking Golan than was Dayan.  As late as June 8, Dayan vehemently opposed attacking Golan, even as Syrian forces were bombarding settlements in northern Israel.  It was at a meeting that day that Dayan convinced Eshkol and the cabinet ­not to attack the Golan.  Others besides Eshkol supported the attack on Golan, including Minister of Labor Yigal Allon and Minister of Education Zalman Aran.  The following day, Dayan changed his mind and ordered the invasion of Golan that Eshkol favored, albeit without consulting with Eshkol that day.  Eshkol was furious not because he opposed the operation, but rather because he feared that now Dayan would reap political credit for ordering the operation.[89]

Lebanon War

In 1982, Israel invaded Lebanon, in reaction to attacks of Palestinian Liberation Organization (PLO) forces based in Lebanon.  Stam and I coded this as an Israeli victory.  Desch (97) codes it as a draw, specifically that the invasion “again met with some tactical success but hardly constituted a clear strategic victory.”  While there has been debate in Israeli society since 1982 about the long-term advisability of the invasion, there is not debate that Israel achieved its operational goals, the successful seizure of swaths of Lebanese territory and the ejection of the PLO from Beirut.  By the operational success standard, Israel won the Lebanon War.

Desch’s central critique (100) is that the decision for war was not taken democratically.  According to Desch, Defense Minister Ariel Sharon “used the pretext of an assassination attempt against Israel’s ambassador to London, Shlomo Agrov, by the anti-PLO Abu Nidal faction to justify an invasion of Lebanon.  Publicly adhering to a ‘Little Plan’ designed only to drive the PLO forty kilometers away from Israel’s northern border, Sharon secretly implemented his ‘Big Plan’ in Lebanon to destroy the Palestinian Liberation Organization (PLO), weaken Syria, and fundamentally remake the Middle East, all without a formal cabinet vote.”  (Note that the correct names of the operations were “Little Pines” and “Big Pines” rather than “Little Plan” and “Big Plan.”)

Desch’s characterization of Sharon as successfully subverting democratic processes is overblown.  Menachem Begin of the hawkish Likud Party was elected to the office of Prime Minister in 1981.  Begin reshuffled the Israeli cabinet, replacing centrists with hawks, including Sharon as Defense Minister.  Desch (100n) cites Avi Shlaim’s book The Iron Wall, but Shlaim recounts a number of instances in which Sharon (and sometimes Begin) attempted to maneuver Israel into war, but was rebuffed by the cabinet.  In December 1981, cabinet opposition to Big Pines precluded a vote from being taken on it.  Sharon then proposed a more limited operation, bombing PLO targets in Lebanon, which he hoped might then draw Israel into a larger war, but the cabinet opposed this suggestion, as well.  In March 1982, Sharon suggested an invasion of Lebanon as a means of testing Egyptian intentions in the wake of Israeli withdrawal from eastern Sinai.  Begin rejected the idea.[90]

Begin remained open to the idea of war in Lebanon, but consistent with selection effects theory he would only act with broad support inside his government, as a means of providing him with domestic political protection once war started.  On May 10, Sharon and Begin secured an 11-7 cabinet vote in favor of a reduced form of Big Pines, but this slim majority was insufficient for Begin, and he declined to order the invasion.[91]  The debate on whether and how to attack was fundamentally changed on June 3 when Palestinian terrorists shot and severely wounded Agrov.  Desch implies that Sharon used this event to get his invasion approved, but the attack enraged Begin himself, who recommended launching an invasion to strike back at the Palestinians.  The cabinet voted 14-2 in favor of war, approving an operation intended to penetrate forty kilometers into Lebanon and avoid conflict with Syria.  The attack was launched on June 6, and enjoyed broad national support.  One June poll found that 77% of Israelis believed that the war was definitely justified, and a further 16% thought it was justified with reservations, though support declined as a quagmire developed following the end of the conventional phase of the war.[92]

A critic might reply that once the war started, Sharon expanded the operation without sufficient respect for democratic processes.   It would be inaccurate, however, to characterize Sharon as acting secretly or against the wishes of the cabinet majority and Begin.  Sharon sought cabinet support for a number of steps to escalate the war once it started.  On the second day of the war, Sharon asked the cabinet to authorize Israeli forces to outflank the Syrians and go as far as the Beirut-Damascus road, which lay beyond the 40 kilometer zone.  The cabinet approved, but took military and diplomatic measures to prevent a larger confrontation with Syria.[93] On June 8, the cabinet approved Sharon’s suggestion for an Israeli flanking move around Syrian forces, and the following day Sharon approved cabinet permission to attack Syrian surface to air missile batteries.[94]

The Lebanon War demonstrates selection effects theory.  Democratic Israel won a war that it initiated.  Begin declined to attack until there was sufficient political support for the invasion.  The decision to attack was popular.  The democratic political process constrained and shaped Begin’s decisions for war.  Once war began, individual decisions for escalation were approved by the civilian cabinet.

The Postwar Political Fates of Middle Eastern Leaders

Desch (103) argues that Israeli leaders were not punished for poor military performance, or rewarded for strong military performance.  He remarks that “the arguably mixed results of the Suez War…did not adversely affect Ben-Gurion’s political career,” though as noted that war provided important geopolitical benefits to Israel, and was certainly viewed by Israeli society as a great success.  Further, Desch concedes that Dayan and Meir resigned in disgrace for the series of errors committed in the prosecution of the 1973 War, that Sharon was forced out after the Lebanon War, and that Begin stepped down from power after the Lebanon War, albeit for personal reasons.  And of course, the War of Independence improved Ben-Gurion’s political standing.  More generally, as noted previously, the most sophisticated quantitative work on the relationship between war outcomes and leadership tenure supports selection effects theory.

Desch attributes a number of leadership transitions in the Arab world to outcomes of the wars with Israel.  However, in all of these cases it is not clear that it was defeat to Israel which caused the leader to lose power.  Husni Za’im of Syria was deposed and executed in 1949, though the coup against Za’im may have been more directly caused by Za’im’s political repression and policies of ethnic favoritism within the Syrian military.[95]  Desch mentions the assassination of King Abdullah in 1951, but that assassin was motivated by fear that Abdullah would make peace with the Israelis, not retribution for Jordan’s defeat in the War of Independence (the assassinations of King Faisel and Prime Minister Nuri al-Sa’id of Iraq in 1958 were also largely caused by fear that those leaders would make peace with Israel).  Desch mentions the 1952 overthrow of King Farouk of Egypt, but this was driven principally by anti-British sentiment and an upsurge in Egyptian nationalism.[96]  The other purges that Desch makes note of, such as those within the Egyptian military following the Six Day War, are not inconsistent with the selection effects theory, as a defeated authoritarian leader found scapegoats for a loss to deflect criticism from himself.  Desch notes that King Hussein of Jordan worried about public backlash following the defeat in 1967, but of course Hussein stayed in power, as did Nasser of Egypt and Assad of Syria.  Nasser also held power following the defeat in 1956, Sadat stayed in power following the failed 1969-1970 War of Attrition, Sadat and Assad stayed in power following the defeat in 1973, and Assad stayed in power following the 1982 defeat.

THE VIETNAM WAR

Downes argues that the 1965 American decisions to escalate its involvement in Vietnam provide evidence against selection effects theory.  He focuses on processes rather than coding of independent and dependent variables.  Importantly, the identity of the initiator is debatable (specifically, whether the February 1965 attacks on Pleiku or the American retaliatory airstrikes constituted the commencement of the North Vietnam-U.S. war), and American participation is often coded as ending in a draw in 1973.  So, this is not a clear case of a democracy initiating a war it goes on to lose.[97]

The American decision to escalate occurred in the context of a wave of Vietcong bombings against American troops in South Vietnam in February 1965 at Pleiku.  The United States launched a minor round of airstrikes against North Vietnam in retaliation for the bombings, which eventually grew into a major bombing campaign.  The United States also substantially escalated its ground commitment in Vietnam, to more than 500,000 American troops.  Downes proposed that the Johnson administration clearly understood in 1965 that the likelihood of winning in Vietnam was not high, and yet it initiated/escalated the war anyway.  Downes claimed that this is a counterexample to the selection effects proposition that democracies only initiate war when they are confident in victory.

As Downes argues, there were two reasons why Johnson initiated/escalated the Vietnam War, despite knowing that the chances for eventual American success were not high.  First, Johnson and others in the foreign policy elite believed that abandoning South Vietnam, an American ally, would have grave consequences for American reputation, and likely encourage aggression elsewhere.  Second, President Johnson feared that withdrawing from South Vietnam would expose him to vicious political attacks including from within his own party, which among other things would jeopardize his Great Society program.

The Vietnam War demonstrates many of the patterns predicted by selection effects theory.  If Downes’ critique is correct, then in one sense the marketplace of ideas seems to have worked reasonably well, as many components of government (including the White House) correctly understood that the Vietnam War was a risky bet in which victory was not guaranteed.  This is not to say that the Vietnam War is an unqualified success of the marketplace of ideas, as critics have long lambasted the Johnson administration for presenting inaccurate or incomplete information about Vietnam, most notably regarding the Gulf of Tonkin Incident.  But at least on one important issue, estimation of the likelihood of winning the war, the president was basically correct.

Consistent with selection effects theory, the decision to escalate was popular at the time.  Before the Pleiku attacks, there was broad support for either standing by South Vietnam or escalating.  A September 1964 poll found that 45% supported maintaining support for South Vietnam, 36% supported escalating to initiate attacks on North Vietnam, and only 19% favored withdrawal.[98]  A December 1964 poll found that 58% supported either standing by the Saigon government or bombing North Vietnam, as opposed to only 20% who supported leaving Vietnam and negotiating.[99]  This support held up after the Pleiku attacks.  A March 1965 poll found that 85% of the American public favored either holding the line to protect South Vietnam or invading North Vietnam proper, whereas only 15% supported withdrawal.[100]  In a June 1965 poll, 79% agreed with the Johnson administration’s argument that if the United States did not stand fast in Vietnam, then the Communists would take over the rest of Southeast Asia.[101]  More generally, opinion polls taken in the months after the escalation decisions were made indicated that about two thirds of the public rated Johnson’s Vietnam policy as excellent or very good, and only about one third rating it as only fair or poor.[102]

Other aspects of the Vietnam War are consistent with selection effects theory.  The escalation of casualties and increasingly grim prospects for victory steadily eroded American public support for the war.[103]  The poor course of the war had negative consequences for the elected leader who initiated/escalated it, as the setbacks forced Johnson to forgo running for reelection in 1968.  Last, even by Downes’ own argument Johnson elected to initiate/escalate the war because of fear of being attacked politically.  This domestic political motivation is fundamentally consistent with the assumption of selection effects theory that the foreign policy decisions of elected leaders are shaped by concerns about their domestic political fortunes.

Importantly, the American public shared the Johnson administration’s sober estimates of the likelihood of rapid victory.  That is, this is not an instance in which an elected leader duped the electorate into being falsely confident in the prospects for victory.  In a May 1965 poll, 44% thought that in five years’ time a neutralist or pro-Communist government would be in power in Vietnam and only 22% thought a pro-U.S. government would be in power.[104]  In an August 1965 poll, only 14% thought the war would end in U.S. victory, and 35% thought it would either end in a Korea-like stalemate or drag on interminably.[105]  In an October 1965 poll, only 29% indicated they thought the war would end in American victory, whereas 40% indicated it would either end in stalemate or drag on.[106]  A different poll found that 54% thought that Vietnam would be a long war, and only 24% though it would “settle soon.”[107]  In one December 1965 poll, 69% indicated that they thought the war would end in a compromise, whereas only 7% indicated that they thought the war would end in a clear-cut American victory.[108]  That is, the public agreed with the basic assessments of the Johnson administration: though the chances of victory were only moderate, the high stakes merited escalation regardless.

Is a case of a democracy initiating war when both public and leadership recognize the chances for victory to be moderate a challenge to selection effects theory?  Stam and I argue that publics are more likely to consent to war as the stakes involved get higher.[109]  Sometimes, as in 1965, the perceived stakes become high enough that the public is willing to support war initiation even when the chances of victory are only moderate.  The existence of such cases is in a narrow sense a challenge to the simple proposition that democracies only initiate wars when they are highly confident they will win.  However, the general pattern is consistent with the theory.  The theory forecasts that publics weigh costs and benefits, and their sensitivity to war costs means they will only support war when the expected utility of war are quite high.  When the stakes are high, it means that the expected utility of war is high even if the chances of victory are moderate, and the costs of inaction are high.

In sum, though the war was initiated despite the recognition by the leadership and the public that the chances of victory were only moderate, many of the internal dynamics within the Vietnam War are predicted by selection effects theory.  A meritocratic military provided the president with an accurate assessment of the likelihood of victory.  The public and president shared a reasonably accurate view of the likelihood of success.  The initial initiation/escalation of the war was popular.  The government’s decision for initiation/escalation was affected strongly by domestic political concerns.  Escalating casualties and declining military fortunes pushed down public support for the war.  The lack of success affected an elected leader’s political tenure.

CONCLUSIONS

Close examination of the Russo-Polish War, the Sinai War, the Six Day War, the Lebanon War, and the Vietnam War all reveal more support for selection effects theory than critics have allowed.  These wars generally describe the patterns predicted by selection effects theory.  Democracies tend to win the wars they initiate.  Domestic politics shape the war initiation decisions of elected leaders.  Democratic militaries usually provide accurate assessments of the likelihood of victory.  Elected leaders generally attack when they are confident they can win.  Elected leaders only initiate wars when they are confident that war initiation decisions will be popular.  Elected leaders vet war initiation decisions among the political opposition, and they do not attack if there is not adequate support.  If the war does not go well, it has negative domestic political consequences for the leadership.  These dynamics are all squarely at the heart of selection effects theory, and the broader proposition that domestic political institutions shape foreign policy decisions.

That being said, elements of the cases demonstrate two important aspects of selection effects theory that would benefit from future research.  The first concerns civil-military relations, especially civilian control of the military.  Though not extensively developed, an internal assumption of the selection effects theory is that democracies are often characterized by looser civilian control of the military.  Elected leaders are less fearful of coups d’état, and hence can afford to promote military officers on the basis of merit rather than political reliability.  The lower fear of coups d’état also means they can grant their military leaders more decision-making autonomy both before and during war.

Though looser control can provide benefits, it can also introduce risks.  Even the highest quality military officers sometimes give poor advice to civilian leaders or make bad decisions.  When civilians defer too much to military officers or to high level officials with military backgrounds, this can sometimes increase the chances that bad military advice gets translated into policy without being checked by civilian oversight.  This has been an issue of long standing concern in American history.  For example, the Union fought the Civil War badly for years because President Lincoln too easily deferred to the flawed judgments of military leaders like Generals John Pope and George McClellan.

The military advice in the cases examined here was generally good, notably military strategizing in the Sinai and Six Day Wars, and 1965 American military assessments of the likelihood of victory in the Vietnam War.  The dangers of excessive deferral to leaders with military backgrounds was evident in the 1982 Lebanon War.  Though Sharon achieved cabinet approval for each escalatory step taken by Israel, he was probably able to have a large influence in shaping the debate in the direction he wanted, broader action in Lebanon, because Begin and other members of the cabinet lacked military expertise.[110]  However, even in the Lebanon case the push to a broader war came from Sharon himself rather than from the military more broadly.  Sharon’s deputy, the former brigadier general Mordechai Zippori, told Begin and the cabinet that the June 1982 invasion plan would lead to a war with Syria, but he was ignored.[111]  Future research might explore whether decisions for war are of higher quality when many or most members of an elected leader’s inner circle enjoy military expertise, meaning that more high level decision-makers can evaluate military arguments more thoroughly and with greater confidence.

A second issues concerns secrecy and the marketplace of ideas.  The issue of secrecy introduces some puzzles for selection effects theory.  Though the theory does not forecast that elected leaders have a normative aversion to secrecy, it does allow that an elected leader might make decisions in secret in order to maximize the chances of victory, and hence minimize the chances that the war will go poorly and threaten her hold on office.  There are limits to how secret an elected leader will keep the decision-making process, as she is motivated to include at least the leadership of the political opposition in order to provide political cover.  Regardless, though keeping a decision secret may improve the chances of victory by concealing diplomacy or military strategy, it may increase the chances of policy failure by cutting out much of the marketplace of ideas.

Secrecy needs to be explored further as both a dependent variable and an independent variable.  As a dependent variable, we need to know what structural military conditions (such as new military strategies) might make secret war decision-making more likely.  We also need to know whether democracies are significantly less likely to engage in secret war planning and decision –making.  Though there are some instances of democracies deciding secretly, they are not the norm.  For example, the United States debated war decisions openly in 1898 (Spain), 1917 (Germany), 1941 (Japan), 1950 (North Korea), 1964/1965 (North Vietnam), 1990 (Iraq), 2001 (Afghanistan), and 2003 (Iraq).  One quantitative study found that excluding Israel, democracies are significantly less likely than other kinds of states to engage in secret mobilization moves during international crises, though Israel itself is more likely to engage in secret mobilization than other kinds of states.[112]

As an independent variable, we need to know whether secrecy makes military success more or less likely, perhaps exploring whether secrecy might make victory more likely but only under certain conditions.  Certainly, some of the bigger foreign policy decisions in American history had their roots in decisions or assumptions made in secret.  The Bay of Pigs invasion was not openly debated.  Johnson’s selective framing of the facts surrounding the 1964 Gulf of Tonkin Incident may have led the American public to exaggerate the extent of the North Vietnamese threat, in turn leading the public to accept war in 1965 knowing that the chances of victory were only moderate.  The 2003 invasion of Iraq was based on flawed and inadequate evidence about Iraq’s weapons of mass destruction and terrorist links that were not sufficiently exposed to open debate.  Future research on both the causes and effects of secrecy will help improve our understanding of the marketplace of ideas, and whether it plays a significant role in explaining why democracies win their wars.

Michael C. Desch, University of Notre Dame

Democracy and Victory: Very Selective Effects

I am pleased to have the opportunity to respond at length to Dan Reiter’s defense of the selection effects explanations for the apparent propensity of democracies to win a disproportionate share of their wars.  Following the old adage that the best defense is a good offense, my book Power and Military Effectiveness bears the brunt of his counterattack.[113]  In it, I examined a wide range of pre-war and wartime explanations for this intriguing statistical finding and could only devote a small portion of the case studies to examining the selection effects argument.  Now it seems to be emerging as the democratic triumphalists’ last redoubt in their campaign to argue that normal democratic processes operate even in the decisions to start and wage wars and this explains the propensity of democracies to win their wars more often than other types of regimes.  In this essay, I want to engage the big conceptual issues that divide us, taking the opportunity to respond to some of Reiter’s smaller points along the way.[114]

The core claim of the selection effects argument is that normal democratic politics continue to operate in wartime and that these very processes make it more likely that democratic states will choose winnable wars.  Two overarching mechanisms are purportedly at work here:  Some argue that because democratic leaders are primarily interested in retaining office, and losing a war is not propitious for one’s electoral fortunes, they are more careful about which wars they start, only choosing those they can win.  This is the electoral constraints argument.  Others posit that democratic leaders are more likely to select winnable wars because the unfettered marketplace of ideas in democratic societies provides them with better information with which to make their decisions about going to war.  Both versions of the selection effects argument assume that there is not much difference between peacetime and non-security democratic politics and the actual decision-making processes that lead to war.

Implicit in the selection effects arguments are three propositions about decision-making in democratic politics concerning national security decisions.  The first is that it is popularly elected leaders who make these decisions about war and peace.[115]  If in fact these leaders do not make the actual decisions about war, this is a problem for the electoral constraints version of the selection effects argument because it is only such leaders who should care about reelection.  Second, these leaders need to respond to, rather than shape, public opinion regarding whether to go to war or not.  This assumption is important for both the electoral constraints and the market-place of ideas versions of the selection effects argument because the public needs to be the independent variable in decisions about going to war.  But if leaders, democratic or otherwise, can shape public opinion, then the causal chain is reversed, undermining the selection effects argument.  Finally, for the marketplace of ideas version to hold, the public debates and rationales for war ought to reflect, indeed shape, the internal discussions and motives within the government.  If that is not the case, then there is no reason to think that the marketplace of ideas is really operating in the actual decisions to go to war.  Thus, public and internal governmental discussions and debates ought to proceed along similar lines.  If not, this calls into question the influence of the marketplace of ideas.

The claim that normal democratic politics operates even in the national security realm and during wartime stands in contrast to well-established scholarly findings.   In Power and Military Effectivess I recounted how the democratic triumphalist’ claim that democracy is an asset in the preparation for, and conduct of, war contradicts much classical political thought beginning with Thucydides and proceeding at least up through Alexis de Tocqueville.  Also, it stands opposed to modern Realist thought, including practioners like George Kennan and theorists like Kenneth Waltz.[116]   In addition, democratic triumphalism cuts against the extensive literature on war and state formation, which finds that states emerging from more intense security environments tended to be less democratic.  As Otto Hintze famously put it, “Power politics, mercantilism, and militarism are all related.”[117]  A related argument was made by Harold Lasswell in his well-known “Garrison State” thesis: “Only the iron heel of protracted military crisis can subdue civilian influences and pass ‘all power to the general.’”[118]  American historian Arthur Schlesinger, Jr. suggested that even in the democratic United States, war, or preparation for it, had the tendency to enhance executive over legislative power, resulting in what he lamented was an “Imperial Presidency.”[119] Finally, there is a robust literature in economics and political science that points out that the dynamics of national security affairs are very different from other aspects of domestic politics.[120]  If democratic triumphalism is correct, democracy is not a liability in war, war does not undermine normal democratic processes, and democracy in fact operates much the same in war as in peacetime.  Such bold claims impose a heavy burden upon their proponents.

I do not think Reiter and his democratic triumphalist colleagues have made this case.  First, the notion that normal democratic politics continue in wartime is an extreme position to defend.  We know war changes societies quite dramatically and even when democracy survives war it is largely in areas unrelated to its conduct.  Second, it is entirely possible to have democratizing (Poland) or reasonably democratic societies (Israel and the United States) in which decisions for war are taken in isolation from democratic politics. Finally, even in highly democratic societies (like Israel and the United States), leaders can use secrecy and other tools at their disposal to shape public opinion, turning it from an independent variable (as the selections effects argument holds) to a dependent variable.  Reiter concedes that secrecy is “a real dilemma” for the selection effects argument, but understates how central it is in allowing even otherwise democratic leaders to isolate decisions about war from normal democratic processes and procedures.[121]

In order to sustain these arguments, this essay proceeds in four sections:  First, I will make the case that, contrary to Reiter’s claim that the theoretical and statistical basis of the selection effects arguments are well established, the most direct test of these mechanisms is through in-depth process-tracing, in which they do not fare well.  The outcomes the selection effects arguments predict are often debatable or do not occur at all, and in any case the processes by which these occur do not follow those these theories posit.   Second, I expand upon the case of the Russo-Polish War of 1919-1921 as an example of how in a weakly institutionalized proto-democracy, a wartime leader can ignore democratic institutions and processes in decisions to start and wage wars.  Third, I then revisit three cases from modern Israel to show how even in relatively well-institutionalized democracies, leaders can by the use of secrecy and other tools of manufacturing consent, avoid public scrutiny or shape public opinion in a manner which reverses the selection effects casual chain.  Finally, I conclude by showing how these same dynamics explain important episodes in recent American history, particularly the George. W. Bush administration’s decision to start a war with Iraq in spring 2003.

WHY USE CASE STUDIES TO STUDY SELECTION EFFECTS?

Reiter touts what he sees as an emerging consensus in the quantitative literature on democracy in support of the selection effects argument.  In addition to his book and articles with Allen Stam, which employ a variety of statistical techniques on large data-sets to show an association between democracy, war initiation, and victory, there is other work in this research tradition that seeks to replicate this finding with democracy, initiation, and crisis outcomes and also scholarship that tries to tie the outcome of wars to the political fate of leaders.[122]  I do not have space here to examine all of this work, but did want to highlight that not all of it provides the strong support for the selection effects argument that Reiter claims.  For example, Hein Goemans points out that his “findings thus throw doubt on a straight forward application of the selection effects logic to war.”[123]

In their book Democracies and War, Reiter and Stam largely eschewed structured, focused, and in-depth case study tests of their various arguments relying largely on statistical tests and illustrative anecdotes. The problem with this approach, as Sebastian Rosato reminds us, is in addition to the “instrumentalist-empiricist” criteria of logical consistency and statistical association, a good theory also has to abide by the “scientific-realist” criteria of its casual mechanisms operating as stipulated in the real world.[124]  If these quantitative studies are in fact so definitive, one wonders why Reiter has devoted so much effort to critiquing my and others’ historical case studies?[125]

There are two explanations for Reiter’s eagerness to engage the historical case studies of the selection effects argument.  First, while all of this statistical research is certainly germane to the question of the relationship of democracy and the successful selection of wars, none really represents anything more than an indirect test of the selection effects arguments.   The actual causal mechanisms of the electoral constraints and marketplace of ideas are not really susceptible to statistical analysis, which simply measures the association of levels of democracy with various outcomes, as they involve decision-making that can only be directly analyzed through process-tracing.

Second, the statistical association between democracy, war initiation, and victory is brittle.  As I point out in Power and Military Effectiveness, out of 197 observations, there are only sixteen which involve democracies starting wars since 1815.  Of these, six are wrongly coded as democracies initiating the war, and of the remaining ten, process tracing does not support the selection effects argument in six more of them.   With just five of these cases removed from the data-set, the statistical association between democracy, war initiation, and victory is no longer significant.[126]

Finally, there is a growing consensus among all of the in-depth historical analyses that the selection effects posited mechanisms do not seem to explain the dynamics of lots of important cases.  Alexander Downes, for example, finds that Lyndon Johnson and his advisors decided to escalate in Vietnam despite knowing that the prospects for success were bleak.[127]  John Schuessler uses the case of Franklin D. Roosevelt’s actions in the run up to the Second World War to show that democratic leaders can “manufacture consent,” and thereby shape, rather than simply respond to, public opinion in their decisions to initiate war.[128]  I will talk at length below about the problems for the selection effects argument that the Russo-Polish War, the 1956 Sinai War, the 1967 Six Day War, and the 1982 Lebanon War present.

These sets of cases pose a challenge for the selection effects argument because they demonstrate the compatibility of proto- or even highly-democratic polities, with largely unconstrained strategic decision-making by government leaders before and during wartime.  Leaders in these cases had three strategies available to them. First, leaders in weak and unconsolidated democracies can often simply ignore public opinion, as was the case in Poland.  Leaders in more consolidated democracies can employ secrecy not only to fool enemies, but also to dupe their publics, as Israeli leaders did in 1956 and 1982 and as the Bush Administration did in part in the run up to the second Iraq War.  Finally, leaders in consolidated democracies can also manipulate public opinion and “manufacture consent” as the Israeli military and its civilian allies did on the eve of the Six Day War and as the Bush administration did after September 11, 2001 as it prepared to topple Saddam Hussein.

In one way or another, all of these strategies change the causal relationship from the public acting as the independent variable shaping government policy to the opposite.  As I will show in the next two sections, Reiter has not done a thorough job of specifying and testing the precise casual mechanisms for the selection effects argument.  The root of the problem is that he has not adequately distinguished in his discussion of the cases between the mere existence of democratic institutions and the “going through the motions” of democracy on the one hand and a truly democratic decision-making process of the sort that the selection effects theory requires to sustain their claim that normal democratic procedures operate in decisions to go to war. These cases are even more problematic because they demonstrate the compatibility of democratic politics in general with decidedly undemocratic decision-making about issues of war and peace.

PIŁSUDSKI GOES TO WAR IGNORING POLAND’S PROTO-DEMOCRACY

I am not eager to refight with Reiter the battle over which side started the Russo-Polish War or when it started, but a few words in order simply to demonstrate that it is by no means as clear-cut as he makes it.  The Correlates of War data-set that Reiter employs has the war beginning with the Polish invasion of the Ukraine in the spring of 1920 and ending with the Polish victory enshrined in the Treaty of Riga in January 1921.  If you accept that story, the case looks pretty good for the selection effects argument, at least at the crude correlational level: Poland started the war and then won it.  The problem with that account is that it is too simple.

First, the majority of sources on the Russo-Polish War mark its beginning earlier, in early 1919.  As I discussed in Power and Military Effectiveness, the conflict really began with the evacuation of the German Army from the territory it occupied in the East, and the movement of Soviet and Polish forces in to replace them.  One source suggests that this was part of a larger Soviet plan hatched in late 1918 to invade Poland.[129] Given that the Soviets moved first into areas that had been part of historical Poland before 1772, it is plausible to argue that the Soviets, rather than the Poles, actually began the war.    Reiter quotes Norman Davies as saying the Poles had the “initiative” after their engagements with the Soviets but wrongly interprets this as meaning he is saying that they started the war.[130] Whether Warsaw itself was their ultimate objective or not, it was the Soviets who first occupied Wilno, an ethnically Polish city and the birthplace of Polish leader Jozef Piłsudski, in January 1919 with military forces. [131]  It is for this reason that Davies concludes that Poland “recaptured” territories occupied by the Soviets after the withdrawal of the Germans and John Posey refers to the Polish “counteroffensive” to liberate these areas. [132]

My view, as I argued in Power and Military Effectiveness, is simply that it is not clear who started the war, but it certainly began before the spring of 1920.  As Warren Lerner put it, “whether the original act of aggression that triggered the Russo-Polish War was perpetrated by the Soviet regime or Poland is a debatable point which can be adequately documented either way.”[133]  It is not as clear as Reiter insists who started this war and when, but he is forced to impose more certainty than the history will bear given how few cases of democracies starting wars there are.

Likewise, I think Reiter overstates the case on behalf of a Polish victory.  Again, let us listen to Norman Davies, whom Reiter acknowledges as the author of the “definitive history” of the conflict:  “In reality, it is difficult to award an outright victory to either side.  Although at the end of the War the Poles had been victorious on the battlefield they did not translate their advantage into political terms.” [134]  Reiter argues that we ought to define victory narrowly, primarily for reasons of convenience.[135]  But unless we are prepared to reject Clausewitz’s premise that war is a means to a political end, rather than an end in and of itself, we have to judge the war’s outcome by whether it achieved that end.[136]  It may be messier and more contested, but it is also more intellectually defensible.  By that criteria, the outcome of the Russo-Polish War seems at best a draw.

But the real problems for the selection effects argument are not so much who started the war or who won it, but rather the precise mechanisms through which it was initiated and then waged.  First, Reiter exaggerates the level and coherence of Polish democracy in the period from independence until the end of the war.   A contemporary observer Sidney Brooks explained that, “Poland had many men of high intentions with equally high energy for that creation.  Her trouble was that few of these men had the requisite training and experience, nor, what was more important, habits of organized association.”[137]  Or as American Ambassador Hugh Gibson colorfully put it in a private letter home in January of 1920: “The Poles are rent with internecine quarrels and all their old imperialism has revived.  And instead of taking to themselves trowel and mortar, and with prayer and fasting laboring at the foundations, they sit squabbling amid the ruins debating whether they shall ally themselves with the Persians or the Medes; whilst their trumpeters march in procession to all the seven cities of Philistia proclaiming loudly, ‘When our greatness is re-established, how great that greatness will be.’”[138]  But for the selection effects argument to work, there has to be a functioning and effective democratic political system.  Such was not the case in Poland before and during the war.  As Richard Watt calculates, “in the first three and half years of Polish independence there were seven separate cabinets, and few of them were capable of withstanding the most modest crisis.”[139]  There was even a coup attempt on January 4, 1919.[140]  Watt adds that “by the end of 1919 Poland did not even possess a prime minister of independent stature …”[141]  This is hardly the description of a well-functioning democratic political system.

To be sure, there were some of the forms of democracy in Poland during this period, including a parliament – the Sjem – and a constitution.  But Reiter does not look much beyond these formal institutions to assess whether they really had much concrete effect on day-to-day politics.  He makes much, for example, of Poland’s so-called “Little Constitution” which consisted of only 250 words and did little more than establish the bare bones of Polish democracy and was widely regarded as only a “stop-gap measure” that had little effect during the war.  It was not until three years after independence, well after the war was over, that Poland had a constitution that was actually “put into effect.”[142]  Given all this, it is not surprising that Davies characterizes Poland during its early years as a liberal democracy in only “the formal sense.”[143]

Piłsudski’s dominant role in Polish national security decision-making present’s two difficult choices for proponents of the selection effects argument. They could try to argue that somehow Piłsudski was not in fact the key figure shaping Polish foreign and defense policy.  But that case is impossible to make because most scholars agree with Davies, who calls Piłsudski “the architect and the organizer, of victory.”[144]  According to Michiel Shewchuk, Polish foreign policy “could be reduced to a single common denominator – Jozef Piłsudski.”[145]  Hardly anyone can name any of the prime ministers or presidents of this period of Polish history (save perhaps for Ignacy Paderewski, the famous concert pianist); the chronicle of this period is essentially the story of Piłsudski!

The other option is to admit Piłsudski was the key wartime decision-maker but then to try to make the case that he was somehow constrained by the Sjem and public opinion.  To make that case, one would have to show that Piłsudski 1) was animated by desire to stay in office via elections, and 2) paid attention to public opinion and the wishes of the civilian government.

The biggest problem for Reiter and the selection effects argument is that the most important actor in Poland at the time, especially in terms of defense and foreign policy was not an elected official at all.   The constant factor in terms of Polish leadership was Piłsudski, an ex-revolutionary activist and military man who assumed and maintained his power without recourse to free and fair elections.  He was originally appointed to the position of Chief of State by the puppet government of the German and Austrian occupation forces.[146]  And he did not have much interest in garnering the support of others to gain office.  As Watt recounts, “In the name of the Provisional Government in Lublin, [Prince] Lubomirski offered Piłsudski full military powers.  Piłsudski declined the offer.  He would not accept any office from any self-proclaimed Polish provisional government (particularly an exclusively socialist one) and thus be beholden to it.  He was determined that he would confer offices and positions on others – not they on him.”[147] It was that same attitude that led him to walk away from the presidency that he was all but guaranteed to win in 1921.[148]  And after his successful coup in 1926, Piłsudski again refused the presidency despite winning an election to ratify it.[149]

Piłsudski had little interest in democratic politics; reportedly he “imagined himself (in the romantic style of the Polish gentry) as the military dictator of Poland that was free, but under attack from external enemies.”[150]  After he was appointed Chief of State, he demonstrated very little interest in playing by the rules of normal democratic politics.  He reportedly dismissed one former comrade’s appeal thus: “’You talk to me about a ‘people’s government’ – I don’t care about your people at this moment.  I care only about a government which can give Poland what she requires.’”[151]  Foreign observes like British Minister to Poland Max Muller observed of Piłsudski that “’he is in no sense a democrat.’”  Viscount d’Abernon, another British diplomat, echoed this view: “The Polish Ministers who, theoretically, are supposed to advise [Piłsudski], possess in truth little real influence or authority.  Indeed, he definitely prefers to act in opposition to their counsel.”[152]  Piłsudski launched a coup in 1926 because of frustration with corruption and inefficiency of Polish proto-democratic politics.[153]  Given that, it is hard not to accept Davis’ characterization of the Polish political system during its early years as a “pseudo-parliamentary charade.”[154]

To be sure, immediately after independence and during the Russo-Polish War, Piłsudski did not try to run every aspect of the whole country, and did not initially rule out a democratic parliamentary system.  “His main interest was the army, and as commander in chief, he supervised every phase of its development,” Watt reports, “He had neither the time nor the interest for further service.”[155] But he demanded, and largely received, a free hand in foreign and defense policies.  Piłsudski would brook no interference with the military and ensured that it was never under parliamentary control.[156]

Piłsudski’s nonchalance about democratic constraints on his authority is evident in how he began the campaigns of 1919 to “repulse” Soviet incursions from Lithuania and Byelorussia.[157]  There was not overwhelming public support to begin with for Piłsudski‘s plan to liberate large tracts of pre-partition Poland and create a federation of puppet states as a buffer against the Soviet Union.  As Watt recounts, “A great many Poles, including most of the members of the foreign affairs committee of the Sjem, believed that Poland had gone far enough.  They favored stopping the Polish Army, annexing the territories in hand, and concluding a peace agreement with the Soviet government [in 1919]”[158]  But Piłsudski continued the advance, ignoring public opinion and the majority in the Sjem.  “So now, without any appreciable opposition, Piłsudski commanded Poland’s army and conducted Poland’s foreign affairs,” Watt concludes, and “He kept his own counsel and made his own plans, paying no great attention to the desires of the Sjem, the majority of whose members, both on the Left and on the Right, wanted peace with Soviet Russia.”[159]   The truth of the matter is that Piłsudski just did not pay much attention to democracy during course of the war:  According to Konstantin Symmons-Symonolewicz, “In politics … as every great statesman, [Piłsudski] was guided exclusively by the interests of his nation.”[160]

Two particular aspects of Piłsudski’s behavior directly challenge the selection effects arguments.  First, the marketplace of ideas could not have been functioning very efficiently during the war because, as Watt reports, “no one knew exactly what Piłsudski’s objective was.”[161]  Second, electoral constraints could not have been influencing Piłsudski’s conduct because public opinion reflected, rather than shaped, what Piłsudski did on the battlefield.  When he was winning he was popular, when he was not, his popularity fell, meaning that he, rather than the public or the civilian government, was the independent, rather than the dependent variable, in Polish politics.[162]

Further contradicting the electoral constraints argument, defeats had little substantive effect on his power, even if they did undermine public support for the Marshal.  During the dark days of the Russo-Polish war during the summer of 1920, Reiter regards the establishment of a War Council as the reassertion of democratic control over Piłsudski but in reality it represented the opposite: the virtual abdication of the Sjem.[163]  It ignored widespread calls to oust Piłsudski in July 1920 after the reverses of fortune in the Ukraine.  While Reiter sees the institutions of Polish democracy before and during the war as robust, the Polish civilian government itself hardly regarded Piłsudski as sufficiently constrained by either the “Little Constitution” or the wartime Sjem or its War Council.  Indeed, the drafters of the 1921 constitution actually imposed more restraints upon executive and military authority directed explicitly at containing the Marshal’s power in the future.[164]  Piłsudski’s “victory” apparently did little to enhance his prospects for elective office

In sum, the Russo-Polish War is hardly a strong case for democratic triumphalists to use to bolster their case for the selection effects theories of democratic victory.  Even if one concedes that Poland started the war and then won it, claims that are historically debatable, the processes by which that came about bore little resemblance to normal democratic politics in action central to the selection effects argument. If one takes a more charitable view of the state of Polish democracy during this period than I do, it is actually much worse for the democratic triumphalists because it demonstrates that democratic politics in some areas can coexist with a lack of democracy in decisions about war.

ISRAEL, THE HOLOCAUST, AND THE MANUFACTURING OF CONSENT FOR WAR IN 1956, 1967, AND 1982

Israel is, at first glance, a better case for the democratic triumphalists because it has been, in most respects, a vibrant democracy since its founding in 1948.  Indeed, the Israel cases are in many respects “most likely cases” for democratic triumphalism because their conditions favor the theory.  In other words, if there is any set of cases in which the selection effects mechanisms should operate, it is these.  Failure, therefore, to explain them is particular damaging for the selection effects theories.[165]   But upon closer examination, the Israeli cases in fact show that despite a robust democratic culture and strong democratic institutions, Israel was not, and in many respects remains, not that democratic in the area of national security decision-making.

There are two specific elements of these cases that call into question the selection effects argument.  First, rather than constraining leaders, as the electoral constraints argument requires, the public and many members of parliament were extremely deferential toward the Israel Defense Forces, the minister of defense, and the prime minister on issues of war and peace.[166]  Not surprisingly, as Avner Yaniv and Robert J. Lieber observe, “Israeli policymakers, like most other policymakers, take the public presentation of their cases very seriously.  But faced with a choice between the requirements of vindication and those of actual policy, the former is considered only an instrument of policy rather than its determinant.”[167] In other words, the democratic public is not the independent variable explaining Israeli defense policy.

Second, despite an otherwise unfettered press and wide-ranging discussion of most issues of government, defense is a dramatic exception. As Nadav Safran observes, “For the Jewish population of Israel the only important restriction on civil liberties on security grounds has been censorship of the press and of private correspondence on matters relating to defense.”[168]  Given those restrictions on discussion and debate of defense matters, it is not surprising that the marketplace of ideas did not play much of a role in Israel’s decisions about war and peace.

This deference to the military and the prime minister and reticence about full-throated public debate about security issues was exacerbated in Israel by virtue of the recent experience of the Holocaust, which almost wiped out Europe’s Jewish population and was one of the formative influences upon the establishment of the state of Israel.  The invocation of the Holocaust has served as a powerful tool for Israel’s leaders to mobilize the public on behalf of various policies. As historian Tom Segev recounts, “the most fateful decisions in Israeli history, other than the founding of the state itself – the mass immigration of the 1950s, the Six Day War, and Israel’s nuclear project – were all conceived in the shadow of the Holocaust.  Over the years, there were those who distorted the heritage, making it a bizarre cult of memory, death, and kitsch.  Others too have used it, toyed with it, traded on it, popularized it, and politicized it.”[169]  Writing in Foreign Affairs, Chief of Staff General Moshe Dayan directly linked the memory of the Holocaust to security discussions in Israel: “The state of Israel had come into existence in the shadow of imminent destruction, and the memories of escape from fearful dangers have attended the people of Israel from the very dawn of their independence.  The memories abide with us still, and go far to explain the depth of our preoccupation with security.”[170]  Elsewhere, Dayan approvingly quoted Israel’s first Prime Minister David Ben-Gurion’s mantra that “’Israel can win a hundred battles yet its problems will not be solved; but if the Arabs are victorious only once it will mean our end.’”[171]  The ever-present specter of a second Holocaust has been a ubiquitous part of every debate about whether and when Israel should go to war and has the effect of shaping the public discourse in favor of centralized, rapid, and secret decision-making when security issues are thought to be at stake.

Viewed through the lens of the Holocaust, the real threats from a number of hostile Arab states along its border were blown out of proportion.  Political entrepreneurs in Israel often manipulated this notion that there is no margin between defeat and destruction and thereby stopped normal democratic politics at the barracks door.   As Safran observes, “National security was … probably the most important countervailing force that helped modify the initial characteristics of the Israeli’s party system and their consequences.  The general concern with this problem not only prevented Israeli factionalism at its worst from tearing the whole political system asunder, but it also contributed more than any other factor to the development of a positive consensus among parties on some specific perceptions, policies, and procedures.  After an initial period of indulging their divergent views, all of the parties soon came to share the perception of Israel as engaged in an inescapable confrontation with its neighbors.  They all came to agree that building Israel’s deterrent power was the main, if not the only, assurance against destruction and constituted a sine qua non for any prospect for peace.”[172] As Holocaust historian Saul Friedlander explains, “since then, consciously or unconsciously, Israel considers herself as a community whose total physical destruction may be intended.  Such a feeling will of necessity inspire the utmost caution, even inflexibility, in regard to anything that might seem to weaken the country’s security in the slightest degree.  The common emotional heritage is engrained deeply enough to overcome any internal dissension in time of crisis.”[173]   Debate, discussion, and even political conflict are business as usual in democracies.  To be sure, one sees them surrounding most issues in Israel but not defense and national security.

The 1956 Suez War

We can see these factors in operation in the run up to the 1956 Suez War in which Israel secretly joined with France and Britain in a scheme to seize the Suez Canal and topple the Egyptian nationalist leader Gamal Nasser.  Israeli Prime Minister David Ben-Gurion hatched this plot with little input from his Cabinet colleagues, only presented it as a fait accompli to the Cabinet literally the night before it was launched and after it could not be stopped, and manufactured the consent of the Israeli public by not disclosing the war’s true objectives, instead playing upon the still-fresh memory of the Holocaust to portray it as a preemptive war of self-defense rather than a war of conquest and regime change.

The first problem with this case for the selection effects argument is that the key decision-maker in Israel – Ben-Gurion – was hardly a strongly committed democrat, particularly when it came to vital issues of state.  As his biographer Michael Bar-Zohar recounts, Ben-Gurion was not a big fan of American democracy and was in general quite autocratic in his view of politics.[174]  Israeli New Historian Benny Morris agrees: “In my book, Ben-Gurion emerges as a consummate statesman and practioner of realpolitik, a man, like Lenin (whom Ben-Gurion, incidently, greatly admired), who knows what must be done and how to get it done.  [Ben-Gurion biographer Shabtai] Teveth would have us believe that Ben-Gurion was some sort of anemic, pussyfooting liberal.  Far from it.  He was a ruthless, single-minded nation-builder.”[175]

Golda Meir recounted an anecdote in which another Israeli politician looked at Ben-Gurion “for a minute, smiled his charming smile and answered thoughtfully, ‘No, I wouldn’t say that [you conduct meetings democratically].  I would say rather that in the most democratic fashion possible, the party always decides the way you want it to.’”[176]  Oxford historian Avi Shlaim adds that “[Ben-Gurion’s] power was so great that his coalition partners used to joke that he submitted proposals to the cabinet only when he wanted them to be defeated.”[177]

Summarizing the process of Israeli national security decision-making under Ben-Gurion, Ha’aretz military correspondent Ze’ev Schiff concludes that: “For the first fifteen years of the state’s existence, from 1948 to 1963, Israel’s political and military leadership was embodied by one man: David Ben-Gurion. He exercised almost exclusive control over both spheres as Prime Minister and Defense Minister, and most important matters were resolved behind closed doors without the advice, consent, or even knowledge of the Knesset.  Often the Cabinet was uninformed, or made privy to only select information on essential defense matters; the decision to embark on the 1956 Sinai Campaign, for example, was reported to some ministers and parliamentary leaders after it was already a fait accompli.”[178]  This process hardly accords with either the electoral constraints or the marketplace of ideas versions of the selection effects arguments.

Indeed, Ben-Gurion set both institutional and procedural precedents that isolated defense and foreign policy decision-making from both parliamentary oversight and public discussion from the beginning of the Jewish state.  For example, it was Ben-Gurion who established the long-standing tradition of combining the prime minister and minister of defense portfolios in one man: himself.[179]  And we can see in a number of other instances in the 1950s where Ben-Gurion either ignored public opinion (the deeply emotional German reparations issue) [180] or made decisions unilaterally and in secret (the Israeli nuclear program).[181] This was reflected, as Amos Perlmutter points out, in the fact that Israel “had no formal legal-institutional structure for the making of national security policy.  Between 1947 and 1974, national security was conceived and implemented by a small, informal, unofficial body known first as Ben Gurion’s ‘inner circle’ and later as Golda Meir’s ‘kitchen cabinet.’”[182]

Given that pattern of behavior in general, it is hardly surprising that Ben-Gurion did not operate within the constraints of normal democratic politics when deciding whether Israel should collude with Britain and France to topple Nasser and regain control of the recently nationalized Suez Canal.  This was clear in the disconnect between Ben-Gurion’s public discussion of events in Egypt, about which he and other leaders repeatedly invoked the apocalyptic image of a second Holocaust in which he could not let the Jews “be slaughtered like cattle,” while privately were he and Dayan confident in a quick victory.[183]  The upshot, as Shlaim points out, was that “the popular perception of the 1956 war in Israel is that it was a defensive war, a just war, a brilliantly executed war, and a war that achieved nearly all of its objectives.  This version of the war was propagated not only by members of the Israel defense establishment but by a host of sympathetic historians, journalists, and commentators.  However deeply cherished, this version does not stand up to scrutiny in light of the evidence now available.  It is a striking example of the way history can be manipulated to serve nationalist ends.”[184]  The Suez case, therefore, poses two problems for the marketplace of ideas:  It highlights the disconnect between the public’s views and those of its leaders’  and it illustrates how the latter are able to manufacture the consent of the former in the national security realm.

Neither were the democratic institutions of the Israeli government much of an obstacle to Ben-Gurion’s plan.  Recall that the previous year the Israeli Cabinet rejected the idea of attacking Egypt.[185]  Rather than killing the plan for good, this opposition simply led Ben-Gurion to proceed in secret, not only to fool the Egyptians, but also to side-step domestic opposition to it in Israel.  Indeed, only ten civilians knew of Operation Kadesh prior to October 25th and Ben-Gurion was clearly the primary decision-maker in the operation’s implementation.[186]  True, he brought the plan to the cabinet, but only after it was too late to stop it on the night before the operation on October 28, 1956.[187]  The details of Ben-Gurion’s autocratic decision-making process have long been public knowledge, and this led political scientist Michael Brecher to conclude that “In making both [the tactical and strategic Suez decisions], Ben Gurion totally ignored Israel’s political structure.”[188]

Reiter concludes that the war nonetheless supports the electoral constraints version of the selection effects argument because no matter how the decisions were made, it was a victory in the end and so one could argue that Ben-Gurion’s confidence in the IDF’s prowess was consistent with his desire to avoid a losing war that would threaten his tenure in office.  But the notion that the Suez War was a ‘victory’ in anything other than a narrow tactical sense, is a hard case to make.  As historian Hugh Thomas concludes, the outcome of Suez was disaster for Britain and France–they did not gain control of the Canal or topple Nasser–and only a mixed benefit for Israel.[189]  Safran agrees, characterizing the victory as “illusory” given that Israel had to give up Gaza and the Sinai and only succeeded in setting back Egyptian armament by a year.[190]  The deployment of the U.N. force in the Sinai served not only to protect Israel but also Egypt. Free transit of Gulf of Aqaba, the putative casus belli, still depended primarily on Israeli force alone. To all that, Benny Morris adds that “If the destruction of Israel was not Arab policy before, after 1956 it most certainly was.  While border clashes and terrorist infiltration remained rare during 1957-62, the political will to belligerence had vastly increased in the Arab world as a result of Israel’s collusion with ex-imperialist powers and the onslaught against Egypt.”[191]   In sum, neither the decision-making process nor the outcome of the Suez War provide strong support for the selection effects argument.

The 1967 Six Day War

The Six Day War of June 1967 was hardly a model of normal democratic politics in action either.  Once again, there was a significant disconnect between what the public and most of the Israeli government thought was at stake in the war, and what the General Staff of the Israel Defense Forces and some of their political allies believed to be the case.  As with Suez eleven years earlier, “the Israeli public perceived the situation as ‘dire’ and ‘live or perish.’”[192] Israeli author Yitzhak Laor characterizes this as “orchestrated panic.”  “Anyone who knew the details, including the CIA in Tel Aviv and the general staff of Israel, knew that Israel was not facing a ‘second Holocaust,’ as the press, nourished by the military’s psychological operations warned….  A very small group of individuals floated an exaggerated, cataclysmic scenario for the benefit of ‘public opinion’ in the West and at home, a key part of Israel’s strategy to this day.”[193]  While the Israeli public was panicked at the thought that the Arabs were planning a second Holocaust, the military was calmly confident in victory.[194]  Roland Popp concludes that “it is quite obvious, then, that the claim of an impending Egyptian first strike was devised in order to convince both the civilian leadership in Israel as well as the main ally, the United States, of the necessity for a military solution to the crisis.”[195]  He reports that the General Staff “’was spoiling for a fight and willing to go to considerable lengths to provoke it,’” and used “alarmist” and even “dubious” information to pressure Eshkol into attacking.[196]   This hardly accords with the notion that an independent public opinion was constraining the Israeli military or the minority of hawks in the government.  Rather, it is a clear case of the latter manipulating public opinion.

Contrary to the marketplace of ideas version of the selection effects argument, Israeli public and civilian governmental support for war was not motivated by confidence in easy victory, but rather fear of a second Holocaust.[197]  This fear was hardly rational, as Segev recounts: “All over the country, one heard and read about the danger that the Arabs were about to ‘exterminate Israel.’  The phrase had no precise meaning, but everyone used it: no one said the Arab armies would ‘conquer’ Israel or that they would ‘destroy’ its cities, not even that they would kill its inhabitants. They said that the Arabs would ‘exterminate Israel’…. The threat of ‘extermination’ had not, then, been real.  But the fear of it had been real, and fear is what Eshkol’s opponents exploited.  More than any other factor, fear had prompted the war … ”[198] Michael Brecher also highlights the effect of the crisis atmosphere on large segments of the government: “The Holocaust psychology was deeply rooted in Israel’s national consciousness, especially among non-Sabras of European descent, and that included 16 of 21 Cabinet members of 4 June 1967.”[199]

Confounding the electoral constraints argument, Avi Shlaim argues that the “decision-making process of the Eshkol government during the war was complex, confused, and convoluted.  It did not bear the slightest resemblance to what political scientists like to call the ‘rational actor model.’”[200]  This is not surprising given that the Israeli government was divided politically and that the relationship between the General Staff of the Israel Defense forces and the civilian government had seriously deteriorated in the run up to the war.

Israeli Prime Minister Lev Eshkol was caught in the middle of intense political and civil-military conflict.  On the former, the relationship between Ben-Gurion and Eshkol had gotten so tense before the war that Segev reports that “Ben-Gurion even toyed with the idea of having Eshkol forcibly removed from office.”[201]  While that did not happen, it was a harbinger of the extra-democratic lengths to which his opponents would go to gain power.  During the crisis preceding the war, he documents how “anxiety, increasing from day to day, served [Eshkol’s] opponents well.  They stoked the fires; the papers called for ‘war now’ and demanded a new war cabinet led by a ‘strongman.’”[202]  Israel’s long-simmering political crisis weakened its democratic political system by dividing the civilian government.

On the latter, the relationship between the military and civilian leadership also reached a crisis during the run-up to the war.  The general climate was tense given that the General Staff, largely dominated by native Israeli Sabras, had thinly veiled contempt for the civilian leaders who were largely European refugees, referring to themselves as “Prussians” and Eshkol and company as “Jews.”[203] So when the General Staff and what Schiff calculates was “the majority in the Cabinet” disagreed regarding whether Israel should attack, “many IDF commanders permitted themselves certain liberties in the political area that had previously been regarded as unthinkable.  They began to intervene – directly and indirectly – in matters of political importance, ranging from the country’s desired borders and settlement in the occupied territories to the proper reaction to Soviet involvement in the Middle East talks with the United States on a settlement with Israel’s neighbors.”[204]  As Segev notes, “Countless letters from Israelis to friends overseas, as well as letters written by soldiers at the front, reflect a desire to avoid war.  But the military leaders insisted there was no way out of it, and they were the only people most Israelis still trusted.”[205]  This situation bears little resemblance to normal democratic politics in action.

The combination of the political crisis and deteriorating civil-military relations resulted in a situation in which Eshkol’s opponents in Ben-Gurion’s and Dayan’s splinter party Rafi and hardliners in the General Staff colluded to topple Eshkol as Minister of Defense and replace him with Moshe Dayan.  This was effected without new elections or other regular democratic procedures.  Once this was accomplished, they were free to start the war against Egypt, Jordan, and Syria that they had been clamoring for, not to avert an impending Arab attack aimed at perpetrating a second Holocaust as they convinced the public, but rather to change the regional military and territorial balance in Israel’s favor, their secret objective.

If the decision to start the war was hardly compatible with the normal democratic processes envisioned by the selection effects argument, its conduct was also scarcely affected by public opinion or even the preferences of most of the rest of the Israeli government.  Israel Lior, an IDF officer, complained of Defense Minister Dayan’s demand for complete freedom of action during the war: “’The proposal that military matters not be subjected to the decisions of the majority was the demand of a tyrant, not of a minister in a democratic government.’”[206]  Michael Oren quotes Dayan as telling other cabinet members, “‘In security matters there’s no democracy.’”[207] According to Brecher, many cabinet members abdicated their role in the decision-making process, voting “for Dayan’s proposals regularly because they accepted him as the authority in security matters.”[208]  In other words, he, rather than they or public opinion, was the independent variable in important wartime decisions.  All of this led Safran to conclude that Israel won in spite of, not because of, its democratic political system.[209]

Finally, as with the 1956 Suez War, it is not clear that the outcome of the Six Day War should be counted as a complete victory for Israel. To be sure, there were some very impressive tactical achievements on the Israeli side.  But when measured by the war’s political objectives, the outcome looks far less clear.  Israel’s most eminent military historian Martin van Creveld concludes that, “In retrospect, the smashing victory of 1967 was probably the worst thing that ever happened to Israel.  It turned ‘a small but brave’ people (Dayan’s words during his radio address on the morning of June 5), who with considerable justification believed itself fighting an overwhelmingly powerful coalition of enemies for dear life, into an occupying force, complete with the corrupting moral influences that this entails.”[210]

The 1982 Lebanon War

But the most damning case for the democratic triumphalist argument that normal democratic politics constrain and inform Israeli leaders is the saga of Israel’s ill-fated invasion of Lebanon in 1982.  Schiff concedes, “Israel is a thoroughly democratic country,” but then goes on to point out that

on the eve of the war and in its initial phases something happened that can only be described as a putsch – albeit a novel one in which control over the Israeli Army and its operations was arrogated by a single man who proceeded to flout governmental decision in matters of crucial importance.  It was not the army that engineered this sophisticated putsch but the Cabinet member in charge of the army on behalf of the civilian establishment – the Minister of Defense.  Sharon acted contrary to the government’s intentions and sent the IDF into action to achieve aims and serve purposes that lacked the Cabinet’s sanction – sometimes because the ministers were deliberately kept in ignorance.  In essence, he would lay out one plan before the Cabinet and then implement a very different one – on which his mind had long been set – using the IDF as his tool.  The result was that the Cabinet either had no idea what the General Staff was thinking or doing, or it received the pertinent information too late.  He fully succeeded in isolating the minister from their military.[211]

In short, there is almost nothing about the Lebanon War that supports either version of the selection effects argument.

As with the Suez and Six Day wars, there was a yawning chasm between what the Israeli public and most of the government thought was at stake and what Minister of Defense Ariel Sharon really had in mind.  For the public, the proposed military incursion into Lebanon in the summer of 1982 was to be limited to pushing Palestinian forces back 25 miles to end shelling of Israeli settlements near the border.  This is what Israeli commentators refer to as the “Little Plan.” [212]  Sharon, however, had a much more ambitious “Big Plan” in mind.  In addition to eliminating the threat to northern settlements from Palestinian shelling, Sharon also wanted to force Syria out of Lebanon; expel the PLO altogether from Lebanon; install a pro-Israel government in Beirut which would make peace with Israel; and solve the Palestinian “problem” in the West Bank and Gaza by encouraging the overthrow of King Hussein in Jordan and the establishment of a Palestinian state there.[213]

From the beginning, Sharon’s Big Plan was highly controversial among the Israeli public and within the government.  It is not even clear if Prime Minister Menachem Begin fully understood or supported it.[214]  Sharon presented parts of it to cabinet on December 20, 1981, when it was rejected.[215]   Yair Evron argues that this led Sharon to conclude that his “plan could not be revealed as it was the potential source of enormous domestic and external political opposition.”[216]

Given the deep divisions within the Knesset about invading Lebanon, Sharon had to find a pretext to implement his plan.[217]  That came with an assassination attempt against Shlomo Argov, Israel’s Ambassador to the United Kingdom, in early June 1982.  In fact, the attempt was masterminded by the dissident Abu Nidal faction, not Yassir Arafat’s P.L.O., but Begin and Sharon used the claim of P.L.O. sponsorship of Agorov assassination attempt to manufacture support for an invasion of Lebanon.[218]  Begin’s thinking about the Lebanon operation was different from Sharon’s, and clearly also influenced by the Holocaust analogy.  Begin’s fixation with the Shoah and events in Lebanon led Israeli Holocaust scholar Ze’ev Manowitz to lament that “‘Whatever its final outcome, the epitaph to be placed upon the war in Lebanon will read: Here lies the international stature and moral integrity of a wonderful people.  Died of a false analogy.’”[219]

Since there was scant public support for Sharon’s Big Plan, Sharon had to “misrepresent its true purpose and scope, i.e., to employ deception” to implement it.[220]  At a Cabinet meeting on June 5th, Sharon lied to Begin, the cabinet, the press, his colleagues in the Ministry of Defense, and even the public about his intention to implement his Big Plan.[221]  In their definitive account of the Lebanon War, Schiff and Ya’ari explain that “Sharon and the few others he privately convinced of his plan … behaved as if they existed in a vacuum, untouched by any political constraints and oblivious to the danger of causing a rift in the nation – an attitude far more pernicious than any real threat posed by the P.L.O.”[222]  Sharon’s conduct in pursuing one policy publicly and another behind the scenes reminded two other scholars of Ben-Gurion’s actions before the Suez War. [223]

Once the war began, Sharon continued to keep the cabinet and the public in the dark about his true aims.   According to Evron, “despite the deep cleavage within Israeli society about the war and the widespread bitterness among part of the population, and under continued criticism from abroad, the Israeli Government persisted in its objectives which rapidly became less and less realistic.”[224]  While some Israeli politicians, apparently thinking in terms that could be compatible with the electoral constraints version of the selection effects argument, opposed the expansion of the operation, they were easily circumvented by Sharon.[225]  In Schiff and Ya’ari’s view, “the Israeli Cabinet was reduced to a genteel debating society that received regular but carefully censored reports and was never advised of what was likely to result from the IDF’s anticipated moves.  It was a body whose decisions were often made after the fact, to rubber-stamp measures in progress or already completed in the field.”[226]

Contradicting the marketplace of ideas version of the selection effect argument, Schiff and Ya’ari maintain that “the PLO had far more accurate information about Sharon’s military plan than did the Israeli Cabinet.”[227]  If that was true of the government, it was undoubtedly the case that the public had even less information at their disposal, making it impossible for the marketplace of ideas to function in an effective manner in the decision-making about military operations in Lebanon.

The truth of the matter is that Sharon brilliantly used deception and secrecy to shape government and public opinion from its onset throughout much of the rest of the war.  He, rather than the institutions of government and public opinion, was the independent variable here.  He achieved this not by overthrowing Israel’s democratic political system, but rather by exploiting weaknesses inherent in it.  As Schiff and Ya’ari conclude, “Sharon devised a formula for bypassing the decision-making process and evading the supervisory prerogatives of the country’s parliamentary system.  Through chinks in that system, he gained the freedom of maneuver necessary to implement his plan.”[228]  In doing so, he not only did irreparable damage to Israel’s security but he also fatally undermined both versions of the democratic triumphalists’ selection effects argument.

One can at least make a plausible argument that Israel won short-term, tactical victories in 1956 and 1967, even if the larger grand strategic results were more mixed.  One can also point to some impressive tactical victories in the Lebanon War too, such as the lop-sided outcome of the Israeli-Syrian air battle over the Bekka Valley.  But as respected American military analyst Anthony Cordesman concludes, despite some tactical success, “Israel suffered its first major grand strategic defeat in the sense that it emerged from the war far weaker in international and domestic terms than if it had not fought at all, and sacrificed many of the technical and tactical secrets it had developed to suppress Syrian air defenses, without any meaningful military result.”[229]  Shai Feldman and Heda Rechnitz-Kijner reflect the near universal consensus that “Even a superficial review of Israel’s invasion of Lebanon, its goals, and the extent to which they have been achieved, demonstrates that the results of the war have not justified the costs.”[230]  Given that, it is only by the narrowest definition of success, that one could, as Reiter does, treat the Lebanon War as a victory for Israel.  If democratic politics as usual should operate in any set of cases, it should have been these.  Its failure in the Israeli cases is therefore a serious blow to the selection effects theory.

MISSION ACCOMPLISHED: THE UNITED STATES SELECTS A DEBACLE IN IRAQ

The debate between Reiter and myself clearly matters in terms of the history of these and other cases, and also has important theoretical implications for political scientists interested in whether democracies, by virtue of their domestic political systems, behave differently than other types of regimes in international relations.[231]  In addition to these two stakes, our debate also speaks to the dynamics of contemporary American foreign policy.  If I am correct in my critique of democratic triumphalism, we should not rest on its democratic laurels in the belief that our political system will ensure that we make the right strategic decisions.  The flawed decision-making process that led to the United States invasion of Iraq in March 2003 is a cautionary tale about the dangers of overconfidence in democratic exceptionalism.

In an important analysis of the decision-making process that led up to the Iraq War, political scientist Chaim Kaufmann highlights the differences between how the marketplace of ideas operates effectively in most domestic political issues as opposed to in foreign policy and national security.[232]  Even in otherwise vibrant democracies like our own, presidents can manipulate public opinion by controlling how issues are framed and thereby shape, rather than respond to, public opinion.   Because presidents control access to intelligence information and have greater weight in public debates, they can often overwhelm the countervailing institutions like Congress, the press, and independent experts that are so central to the functioning of the marketplace of ideas version of the selection effects argument.  These normal advantages of the presidency in war and foreign policy, as Alexander Downes and John Schuessler have shown, allowed previous presidents like F.D.R. and L.B.J. avoid the constraints of democratic politics in their decisions for war.

With these executive advantages in national security decision-making, combined with an extraordinary event like the attacks on the World Trade Center and the Pentagon on September 11, 2001, which have added to the American public’s collective memory an historical analogy nearly as potent as the Holocaust is in Israel, it is easy to see how even in an otherwise robust democracy, decisions for war can be made in a fashion quite inconsistent with the selections effects argument.  As an aside, and telegraphing the connection between 9/11 and the Holocaust, President Bush met at the White House with Holocaust survivor Elie Wiesel in February of 2003.  During that meeting, Wiesel compared Saddam to Hitler and urged Bush not to repeat the West’s failure to confront Holocaust evil early on, an appeal that resonated with important elements of elite opinion such as the so-called Liberal Hawks and thereby helped create bi-partisan support for the Iraq War.[233]

Advocates of the war to topple Hussein in the Bush administration used all of these mechanisms to manipulate public opinion in support of the war.[234]  As is now widely recognized, the two public rationales for this war – Iraq’s alleged pursuit of nuclear weapons and the Iraqi government’s putative connection with the 9/11 attacks – were willfully exaggerated, and in some cases made up out of whole cloth, to make the case for war. The evidence for this is threefold. First, neither claim turned out to be true once the United States had free rein to search the country, ransack the former regime’s files, and interrogate its members.  Second, we now know that there was a conscious effort by members of the Bush administration to twist intelligence to make the case for war.[235]  Finally, high-level officials have admitted that the public case for the war was not the same as the Bush administration’s real rationale, which was instead focused on how Saddam’s ouster might, like Ariel Sharon’s Big Plan, start a process of regional transformation that would dramatically alter the security environment in the Middle East.[236]

As political scientist Jon Western explains, “a cohesive administration intent on framing and selling the need for war can frequently influence and mobilize public support by controlling, managing and even distorting information.  When information is controlled and manipulated, and when counterarguments are suppressed, co-opted, or delegitimized, even rational publics may end up endorsing policies that later seem irrational.”[237]   In other words, even in otherwise robust democracies, public opinion can be turned into a dependent variable and consent for war can be manufactured.  Acknowledging this was his objective, President Bush reportedly told Italian Prime Minister Silvio Berlesconi in January 2003, “’You watch, public opinion will change. We lead our publics.  We cannot follow our publics.’”[238]

The combination of the distinct nature of national security issues from other elements of domestic politics and the institutional advantages of the presidency in dominating them, explains why normal democratic politics cease when the guns begin to fire.  Central to this is the role of secrecy, which Reiter downplays as a serious challenge to the selection effects arguments.  In fact, the ability of presidents to conduct large parts of their wars in secret, not only to prevent enemies from divining U.S. strategy, but also, and often just as importantly, to derail and short-circuit domestic opposition, presents fundamental challenges to both the electoral constraints and the market place of ideas versions of the selection effects argument.  Secrecy makes it possible for presidents to initiate wars that the public and other parts of government might not otherwise support, thereby thwarting democratic checks and balances.  And secrecy also makes it possible for presidents to bankrupt the marketplace of ideas by either withholding vital intelligence information or by manufacturing consent by introducing false information into the public debate or intergovernmental deliberations.  Both elements of this strategy were evident in the run-up to the Iraq War.

The profoundly disturbing implication of both the Israeli and U.S. cases is that mature democracy and undemocratic war-making are fully compatible.  They demonstrate clearly that there are ways, even in the context of otherwise robust and vibrant democratic politics, to shield strategic decision-making from the rest of politics and reverse the causal arrow from the democratic public being the ultimate decider of key political decisions to placing them in the hands of a small number of leaders who may in fact make bad decisions about war and peace.

Reiter is, of course, correct that we do not yet know for sure how Iraq will turn out (thought it does not appear it will have the happy ending envisioned by its architects, who privately conjured pipe dreams of a democratic Iraq serving as the first of a line of democratic dominoes falling to produce a complete regional transformation), but there is no doubt that the process by which we got into it bears little resemblance to either version of the democratic triumphalists’ selection effects arguments.  The danger of democratic triumphalism, then, is that it blinds us to the sad fact that democracies are often no better than other regimes in how they select their wars or otherwise behave in international relations.

Alexander B. Downes, Duke University

The Myth of Choosy Democracies: Examining the Selection Effects Theory of Democratic Victory in War

The debate over what some have labeled “democratic efficacy” or “democratic triumphalism”—the view that “democracies systematically outperform non-democracies in the hurly-burly of international relations”—continues to rage among scholars of international politics.[239] One specific manifestation of this debate concerns the question of whether democracies are more likely than non-democracies to prevail in war. Dan Reiter and Allan Stam have argued that democracies tend to win wars because democratic leaders have strong incentives to start only those wars that they are confident they will win (known as the selection effects argument), and because democratic soldiers fight harder in battle, displaying better leadership and initiative (known as the warfighting argument). Their book Democracies at War advances these explanations for democratic effectiveness and counters others, such as the contention that democracies “win wars on factory floors,” overpowering their autocratic foes with material power, or that “birds of a feather flock together,” whereby democracies come to each others’ aid in wartime and form overwhelming alliances.[240]

A number of scholars have criticized the arguments and evidence put forward in Democracies at War on both quantitative and qualitative grounds.[241] Reiter and Stam have responded by vigorously defending their position, which has prompted several exchanges in the correspondence section of International Security.[242] In the latest iteration of this debate, Dan Reiter offers a spirited brief for the selection effects explanation for democratic victory in his essay “A Closer Look at Case Studies on Democracy, Selection Effects, and Victory.”[243] The selection effects argument pertains to wars initiated by democracies, and consists of two mechanisms that purport to explain why democracies prevail in a disproportionate number of these conflicts. First, because democratic leaders are relatively easy to remove from power via regular elections, they attempt to avoid major policy failures—like losing wars—and thus will start only those wars they believe they have a very good chance of winning. Second, democratic leaders benefit from a robust marketplace of ideas—consisting of vigorous public debate as well as good advice from the military—that helps them better estimate the likelihood of victory. In his essay, Reiter revisits several cases of democratic war initiation originally explored by Michael Desch in his book Power and Military Effectiveness, and in my article “How Smart and Tough Are Democracies?” in International Security.[244] Reiter argues that upon close examination, these cases—the Russo-Polish War, the Suez, Six-Day, and Lebanon wars involving Israel, and the Vietnam War—actually reveal substantial support for selection effects.[245]

I am pleased to have the opportunity to respond to Reiter’s essay and further extend the debate. I divide my contribution into two parts, the first mainly theoretical and the second largely empirical. In the first part of the essay, I make four arguments about the logic of selection effects theory and the evidence offered to support it. First, I argue that Reiter and Stam’s electoral accountability argument contains internal contradictions that neither its proponents nor its critics have previously acknowledged. Reiter and Stam assume that leaders seek contemporaneous consent for their decisions rather than retrospective approval, and that leaders follow public opinion but cannot lead it. These assumptions create two problems: the various mechanisms Reiter claims are consistent with the electoral accountability argument for how democracies choose which wars to enter—the leader expects to win, the leader expects the war to be popular, or a hawkish public drags a leader into war—can be mutually contradictory, and war outcome is not the correct dependent variable to test the theory.

Second, Reiter offers no evidence that a key aspect of the marketplace of ideas—public debate and deliberation—leads to better decisions for war. Because this part of the marketplace contributes nothing to decision-making, I contend that Reiter and Stam’s marketplace of ideas mechanism reduces to an argument that democracies are characterized by better civil-military relations than non-democracies. Other scholars have written on this subject and found that only some non-democracies have toxic relations between political and military leaders, while civil-military relations in democracies are not uniformly positive.[246] The implication is that scholars should redirect their attention away from the marketplace of ideas and toward civil-military relations as an explanatory factor of military effectiveness.

Third, the issue of secrecy in democratic preparations for war is a major conundrum for selection effects. Reiter ends up arguing that secrecy can both increase the chances of victory and the chances of disaster, but provides no guidance for when it will have which effect. The effect of secrecy on the odds of democratic victory is thus indeterminate.

Finally, I argue that the quantitative evidence for democracy and victory is less clear than Reiter would have us believe. Reiter dismisses quantitative critiques of selection effects in a single sentence when these critiques actually pose a major challenge to the correlation he and Stam identified in their earlier work.

In the second section of the essay, I argue that contrary to Reiter’s claims, the details of the Vietnam case do not comport with selection effects theory and exemplify some of the problems with the theory mentioned above. In my 2009 article, I argued that President Lyndon B. Johnson and his closest advisors initiated war in Vietnam despite knowing that victory was unlikely and the war would be costly and protracted. Reiter does not contest this argument, but he suggests several other ways in which the case is supposedly consistent with selection effects: the United States did not initiate or lose the war; the marketplace of ideas operated effectively; going to war was popular with the American people; Johnson’s domestic political reasons for escalating are explained by the theory; and the public believed the stakes were so high in Vietnam that it approved of fighting even though the probability of winning was low. I show, to the contrary, that the United States should be coded as the initiator of the war against North Vietnam and that the war ended in a draw (which is nevertheless inconsistent with selection effects because the United States failed to win); the marketplace of ideas was an utter failure; the American public was remarkably ambivalent about fighting in Southeast Asia; Johnson’s reasons for escalation—although rooted in domestic politics—are not explained by selection effects theory; and that the public did not view the stakes in Vietnam as so critically important that it offered its consent for war even though the outlook was not sanguine.

I conclude the essay by offering a few thoughts on the prospects for future development of selection effects theory and on the study of military effectiveness more broadly.

THEORY, EVIDENCE, AND SELECTION EFFECTS

The Contradictions of Selection Effects Theory

The selection effects theory as described by Reiter and Stam is disarmingly simple. Democracies prevail more often in wars they initiate because leaders in democracies start only those wars they have a high degree of confidence they will win. Why? Because leaders understand that starting a losing war is a good way to get voted out of office, and the lively marketplace of ideas that characterizes democracies allows them to differentiate wars that promise to be easy victories from those that will be costly slugfests.

Underneath the surface of this seemingly simple theory, however, things quickly become complicated. In fact, a fundamental contradiction lies at the heart of the theory. Reiter and Stam’s electoral accountability argument appears at first glance to be a straight-forward application of the retrospective model of voting. In this model, voters “reward success and punish failure” focusing on the incumbent’s past record.[247] In the selection effects theory of war initiation, the key question voters would ask when they went to the polls would be: did the leader win the war he or she started during the last term? If yes, then reward him or her with another term; if no, punish the leader by voting for another candidate.[248] Because elections are not held every year, and rarely coincide with the beginnings or endings of wars, the public can only register its approval or disapproval of a leader’s policies after the fact. Leaders seek to achieve policy success with the hope of being retained in office the next time the country votes.

According to Reiter and Stam, however, this is not the model of voting that underpins their argument. Rather, they prefer a model termed “contemporaneous consent,” which they describe as follows: “Leaders in liberal democracies seek out contemporaneous approval for political choices. Voters then punish leaders not so much for particular failure or success, but instead for failing to heed the more popular sentiments at the time the leaders settle on a particular policy.”[249] What matters most, in other words, is what public opinion supports at the time a decision is made, rather than how the public will judge the success or failure of leaders’ policies come election time. Indeed, although the bumper sticker version of the selection effects argument is that leaders start only those wars they believe they have a high chance of winning, as Reiter points out there are at least two other mechanisms that are consistent with the argument: (1) when the war is likely to be popular with the public or is for a popular cause; and (2) when “publics gripped by war fever…drag a reluctant leader into war,” the leader following presumably because the public’s war fever guarantees that the war will be popular.[250] These two mechanisms flow from the contemporaneous consent model of leader behavior that lies at the heart of the electoral accountability mechanism of selection effects theory.

Reiter and Stam’s understanding of leaders as responding to public consent at the time policy is made, however, does not square with their description of the consequences of failure in war, which is clearly retrospective. When voters go to the polls, do they cast their ballots based on whether the leader obeyed public opinion when he originally made a policy decision, or based on the success or failure of the war? The way Reiter and Stam describe their argument throughout the book implies the latter: “Because democratic executives know they risk ouster if they lead their state to defeat, they will be especially unwilling to launch risky military ventures.” As the first President Bush put it, “I’ll prevail or I’ll be impeached.”[251] A theory based on contemporaneous consent, however, predicts that a leader who won a war but chose to start it when public opinion did not support war should be removed from office. Similarly, a leader who lost a war that the public supported when it was declared should not be punished for losing. In short, according to Reiter and Stam’s own description of their theory, war outcomes are not the proper dependent variable to test the theory.

Furthermore, as articulated by Reiter and Stam, selection effects theory is internally contradictory. One mechanism maintains that leaders choose wars they think they can win, while two others contend that leaders start wars they believe will be popular. But what is popular is not always what is strategically sound, and although the theory seems to assume that all good things go together, sometimes they do not. What if the leader expects a quick and decisive victory but public opinion opposes going to war? What if initiating war is a popular option but the leader doubts victory is possible? What if a public gripped by war fever compels a leader to start a war he or she believes stands little chance of yielding a victory? What if the leader gets sound advice from the military that victory is unlikely but goes to war anyway? What if the public debate on the war is weak and ineffective, but the war is still won? Are cases like these consistent with selection effects, or do they contradict it?[252] As the theory is currently articulated, it is impossible to determine, and proponents of the theory can always fall back on a different mechanism in the event that the historical evidence contradicts one of them.[253]

If war outcomes are not the correct measure to test an electoral accountability mechanism based on contemporaneous consent, how can the argument be tested? The correct metric for testing the theory is whether leaders followed the will of the public when they initiated wars. The evidence to look for in case studies of democratic war initiation would thus be measures of public approval for going to war. Unfortunately, there are two serious problems with this approach. First, the key assumption of the contemporaneous consent model is that leaders are followers of public opinion, not shapers of the public’s views. Leaders are constantly trying to determine which way the wind is blowing, and do their best to do what is popular at the time. Not only is this description inconsistent with how Reiter and Stam test the theory, as shown above, it also neglects the possibility that politicians can shape and lead public opinion. In Democracies at War, Reiter and Stam make repeated references to leaders being able to “generate” consent, i.e., persuade the public to support an action of which they previously disapproved.[254] John Schuessler has shown how President Franklin D. Roosevelt was able to deploy deception to swing the American public behind a policy of intervening in Europe before U.S. entry into World War II.[255] Reiter and Stam also acknowledge that “there is great variability among democracies in different eras as to the conditions that are sufficient to generate public consent for the use of force against some other state.”[256] These kinds of statements suggest—in contradiction to the view that political elites merely follow some preexisting public consensus—that leaders in fact have a fair amount of leeway in persuading publics that something is in “the national interest.” If leaders in democracies have formidable resources at their disposal to shape and mould the public’s views on the use of force, then it is hard to sustain the argument that public opinion acts as much of a constraint on leaders’ decisions to start wars.

Second, public opinion is notoriously difficult to measure. The answer one gets often depends heavily on the way one asks the question. As I demonstrate below, arriving at a coherent view of “public opinion” on the course the United States should have followed in Vietnam in the months before Johnson launched the air campaign against North Vietnam in February 1965 is confounded by the baffling array of polls that show support for options ranging from escalation to holding the line to negotiations and withdrawal. A related difficulty concerns the type of evidence that can be deployed to support arguments about public opinion for war. Because of the possibility of rally effects, it is not valid to use measurements of public support for military action taken after the war has begun. To test whether a leader is following public opinion when choosing to go to war, evidence of the public’s views must be obtained from the period before force has been used on either side.[257]

In sum, Reiter and Stam’s version of selection effects theory posits multiple causal mechanisms for how democratic leaders choose to go to war that are mutually contradictory. The dependent variable they have used to evaluate the theory—war outcome—is also incorrect. To be consistent with the logic of contemporaneous consent, they should examine pre-war public opinion to determine if leaders are following the policy that is popular with the majority. Pursuing this investigation is fraught with difficulty because leaders have the ability to change the public’s views over time and it is hard to get an unambiguous reading of public opinion.

Laying the Marketplace of Ideas to Rest

The second half of selection effects theory consists of the marketplace of ideas, which is composed of two separate mechanisms that combine to allow democratic leaders to make informed decisions for war. One mechanism is the venerable idea of discussion: the more voices that participate in debate on a given subject, the more likely that the decision arrived at will be correct.[258] Freedom of expression and freedom of the press in democracies foster a vibrant public sphere and stimulate productive debate on important decisions like whether to use force abroad. The second mechanism focuses on the military: because democracies tend to have amicable civil-military relations, officers are promoted based on merit and are able to provide high quality military advice to civilian leaders.[259]

In his essay, Reiter neither provides evidence for the first plank of the marketplace of ideas, nor is any such evidence offered in Democracies at War. Indeed, to date there is little evidence in the literature that this mechanism operates to produce better decisions on national security in democracies, but a fair amount of evidence that the marketplace is vulnerable to subversion by elites bent on persuading the public of the wisdom of a particular military venture.[260] A multitude of critics have argued quite persuasively, for example, that the George W. Bush administration was able to manipulate public debate to obtain support for the invasion of Iraq in 2003.[261] In the face of these critiques, proponents of the public discussion aspect of the marketplace of ideas need to provide compelling evidence of debate in the public sphere working to improve the quality of democratic decisions for war or risk seeing this concept consigned to the theoretical dustbin.

Given the dearth of evidence for public discussion, the marketplace of ideas part of selection effects theory reduces to the argument about meritocratic militaries. The argument maintains that because civilian control of the military is well-established in democracies, civilian elites need not fear military coups and thus are able to promote officers solely on the basis of merit. A thoroughly professionalized officer corps then provides expert military advice to political leaders, which aids them in formulating strategy and choosing wars that will turn out victoriously.

There is some support for the view that certain types of non-democracies are handicapped by civil-military relations plagued by mistrust, and that this problem undercuts their military effectiveness. Stephen Biddle and Robert Zirkle, for example, chronicled how Saddam Hussein’s fear of overthrow at the hands of his officer corps led him to purge it frequently of any contenders for power, thereby crippling the army’s ability to fight external foes.[262] Similarly, Risa Brooks has shown how the parlous state of relations and unclear lines of authority between President Gamal Nasser and his chief military advisors contributed to the disaster suffered by Egypt in the Six-Day War.[263] However, these and other scholars also identify non-democracies characterized by much healthier civil-military relations and link this variable directly to improved war performance. This literature contends that authoritarian states are not uniformly characterized by poor civil-military relations or disastrous levels of military effectiveness.[264] Nor is it the case that civil-military relations in democracies are uniformly harmonious. Some scholars have linked cases of ineffective military performance by democracies with unhealthy relations between officers and civilian leaders.[265] To preview my discussion of Vietnam later in the essay, President Johnson and his civilian advisors in 1964-1965 went to great lengths to exclude the Joint Chiefs of Staff (JCS) from crucial deliberations on strategy, misrepresented the chiefs’ views to others in government, misled them on numerous occasions, and politicized relations with the JCS to shape the content of military advice the chiefs would provide. It is thus not clear that there is a necessary connection between democracy, positive civil-military relations, and victory in war. This is a potentially fruitful avenue for further research.

Selection Effects and Secrecy

Reiter discusses at some length how secrecy in democratic war preparations fits with selection effects theory. An earlier generation of democratic peace theorists maintained that one reason democracies did not fight each other was because the public nature of war preparations in democracies precluded them from launching surprise attacks.[266] This feature of democratic decision-making meant that inter-democratic crises could be resolved peacefully since neither side had to be concerned about a bolt from the blue. By contrast, although Reiter concedes that “examples of leaders deliberating in secret have complex implications for the theory,” he ultimately concludes that keeping these deliberations out of the public eye does not contradict selection effects theory.[267]

Reiter’s discussion of secrecy demonstrates that the effect of secrecy on the likelihood of democratic victory is indeterminate. On the one hand, he argues that secrecy sometimes contributes to victory by facilitating the success of surprise attacks. On the other hand, secrecy—because it bypasses the marketplace of ideas and thus sacrifices the benefits of full and frank discussion of the merits of going to war and the modalities of attacking—also increases the risk of failure. Reiter does not give any hint as to which of these effects may be greater, and thus theoretically the effect of secrecy on the likelihood of democratic victory is indeterminate. It is not possible to declare, as Reiter does, that the “desire to maximize the chances for victory through secrecy is consistent with selection effects theory’s premise that elected leaders have a very strong incentive to avoid military defeat” because—as he admits—secrecy also raises the likelihood of defeat, and is thus inconsistent with the incentive to avoid military defeat.[268] It is also worth noting that it is difficult for contemporaneous consent to operate if an attack is planned secretly because leaders cannot be sure whether the public supports or opposes such an attack beforehand.

Reiter’s discussion of secrecy in the context of covert action merely inserts another source of indeterminacy into the equation. Reiter argues that “leaders in some cases may be motivated to circumvent public constraints by acting covertly,” and that “covert actions such as the Bay of Pigs invasion often fail because secrecy increases the risk of policy failure by cutting out the marketplace of ideas,” causing democratic decision-making to approximate autocratic decision-making.[269] But how constraining are democratic institutions and public opinion if leaders can simply sidestep them by employing covert measures? One recent study identified literally hundreds of instances of covert action by the Eisenhower and Kennedy administrations during the Cold War.[270] This seems to be an admission that selection effects theory applies to a rather small proportion of democratic foreign policy decisions.

Reiter’s argument about secrecy and covert action also states quite clearly that secrecy increases the likelihood of failure, which contradicts his earlier argument that it is consistent with selection effects because it increases the likelihood of victory. Reiter tries to distinguish between secrecy in overt wars versus covert action by saying that wars obviously become public knowledge whereas “leaders may have more confidence that ‘plausible deniability’ will enable them to escape or mitigate the negative consequences of a failed covert plot.”[271] On the contrary, the latest research on democracy and covert action has found that democratic leaders often proceed despite having little confidence that covert plots will succeed and little confidence that failed plots will remain secret.[272] Some covert plots—such as the Bay of Pigs or the Iran-Contra affair—have failed publicly and dramatically, resulting in enormous embarrassment at a minimum to criminal indictments and convictions at a maximum. Democratic leaders, in short, are typically not confident that covert failures will remain covert, yet they order them nonetheless, in contradiction to the dictates of selection effects.

Quantitative Critiques of Democracy and Victory

My final point in this section concerns the quantitative evidence for democracy and victory. In their 1998 article in the American Political Science Review and four years later in their book Democracies at War, Reiter and Stam presented a statistical analysis of democracy and victory in interstate wars.[273] They found that democracies that initiated wars and democracies that were targeted by other states were each more likely to prevail than non-democratic initiators and targets. In my 2009 article I reanalyzed their data and came to a different conclusion.[274] Reiter and Stam restricted their analysis to wars that ended in victory or defeat, excluding draws and analyzing them separately. The logic of selection effects, however, clearly states that democratic leaders only start those wars they think they can win, and avoid wars that are likely to end in defeat or stalemate. According to that logic, all war outcomes—including draws—should be analyzed. Reiter and Stam also classified all belligerents as either initiators or targets, thereby omitting an important class of combatants: states that join ongoing wars after they have started. Using these tripartite classifications of war outcomes and belligerents, I found that although democratic initiators, targets, and joiners were somewhat more likely to win than their autocratic counterparts, in no case were the differences between them statistically significant. In other words, statistically speaking, we cannot be confident that there is actually no difference between the rates at which democracies and autocracies prevail in war.

Reiter and Stam penned a response to my article that was published (along with my rejoinder) in the correspondence section of an ensuing issue of International Security.[275] Curiously, they did not dispute my findings directly; instead, they argued that I had incorrectly tested their theory, which they maintained divided states into three types—democracies, autocracies (states at the lowest end of the democracy-autocracy spectrum), and oligarchies (states falling between autocracy and democracy, sometimes also called anocracies)—whereas I had lumped all non-democracies together and tested their effectiveness against that of democracies. Although Reiter and Stam had performed exactly this same test in their work,[276] they contended that my failure to differentiate among three regime types rather than two was the source of my insignificant findings. They used fractional polynomials to argue that the relationship between regime type and victory is actually curvilinear: democracies are best, oligarchies are worst, and autocracies fall in between.

In my rejoinder I tested this argument and found it to be incorrect. There is no significant curvilinear relationship between regime type and victory. Reiter and Stam claimed that merely showing that the fractional polynomial terms in their model were statistically significant sufficed to show evidence of a curvilinear relationship.[277] But this is mistaken: the way to test for curvilinearity is to compare the explanatory power of the model including the fractional polynomial terms with that of the model that includes the linear terms. If the former is significantly larger than the latter, then the curvilinear version is superior.[278] Comparing these figures showed them to be exactly the same, even when using Reiter and Stam’s original data (excluding draws and war joiners).[279] There is also no evidence of curvilinearity when draws and joiners are added. Thus, whether one posits a linear or a curvilinear relationship between regime type and victory, when all war outcomes are examined, democracies are no more likely to prevail than other types of polities.[280]

Reiter argues that his findings with Stam regarding the efficacy of democracies in conflict have been widely corroborated by other studies.[281] He neglects to mention a growing body of work that reaches different conclusions. Reiter, for example, claims that a recent article by H. E. Goemans supports selection effects because Goemans found that “elected leaders who lose wars face significantly higher chances of losing office through extralegal means, thereby facing higher likelihoods of severe personal punishments such as death, prison, or exile.”[282] This claim is not only disingenuous, but if true actually undermines the argument. The logic of selection effects theory is that democratic leaders are cautious to start only those wars they think they can win because defeat increases the risk of removal from office via elections (or other regular procedures, such as impeachment). The central insight of the theory is that democratic leaders are constrained because there are regular institutionalized opportunities to remove them, whereas authoritarian leaders—who must be ousted violently—are not so constrained. The theory says nothing about the possibility of violent overthrow of democratic leaders, and thus evidence that democratic leaders who lose wars are at increased risk of falling victim to a coup cannot be used as support for selection effects. Indeed, the fact that democratic losers are more likely to be removed by irregular means undermines the argument that democratic procedures are sufficient to punish (and hence deter) errant leaders. Goemans further shows that defeat in war does not increase the likelihood that leaders are removed from office via regular procedures.[283] In an earlier article, Goemans and a co-author similarly found that neither victory nor defeat had any systematic effect on the likelihood of democratic leaders’ removal from office.[284] These sophisticated empirical studies on the effect of conflict involvement on leader survival thus contradict selection effects theory.

Other recent work has begun to chip away at democratic exceptionalism in war. Stephen Biddle and Stephen Long, for example, have shown that the effect of democracy on the casualty ratio between attacker and defender depends crucially on how one interprets the contribution of human capital.[285] When an indicator of human capital is included along with democracy in a regression, regime type becomes insignificant.[286] If it could be shown that democracy causes higher levels of human capital, this result would support democratic triumphalism, but if better human capital leads to democracy, then democracy is at best epiphenomenal and perhaps even harmful to military effectiveness. Other studies have found that democracies are at best no better—and possibly worse—than non-democracies at fighting irregular, counterinsurgency wars.[287] Still others maintain that democracies with high levels of income inequality are systematically biased toward building capital-intensive militaries and fighting small wars with inappropriate firepower-intensive strategies because this way of fighting lowers the economic burden on the median voter even as it makes democracies likely to lose.[288]

In the broader literature on regime type and conflict, a number of studies have begun to challenge or qualify democratic efficacy. Jessica Weeks, for example, argues that leaders in several types of non-democratic regimes are subject to institutional constraints on their power that cause them to exercise caution in the types of crises and wars they undertake. Weeks finds that democracies are not uniquely advantaged in being able to make credible threats: most non-democracies have their threats reciprocated at rates that are not statistically different from the rates that democracies’ threats are reciprocated.[289] Douglas Gibler and Marc Hutchison argue that there is no general democratic audience costs advantage that makes democratic threats more effective; audience costs are present only in those rare cases when the issue is highly salient to the public, such as during a dispute with an enduring rival.[290] In a major reassessment of the democratic audience costs thesis, I argue with a co-author that the evidentiary basis for the argument that democracies make more credible threats in international crises is deeply problematic.[291] The reason is that the principal datasets used to test the argument—Militarized Interstate Disputes (MID) and International Crisis Behavior (ICB)—are nearly devoid of actual threats. We reanalyze prominent empirical studies that use these datasets and find that the results are driven almost entirely by cases that do not contain threats. Analysis of an original dataset of compellent threats shows that democracies are no more likely than autocracies to obtain their demands.

In short, the quantitative debate about democratic efficacy in international relations is far from over. A large body of research questions the view that democracies are both more likely to prevail in wars and in international crises.

Summary

To sum up, I argue that selection effects theory as articulated by Reiter and Stam, relies on an understanding of leaders as seeking contemporaneous consent for their policies, thereby making mutually contradictory predictions about when democratic leaders will initiate wars. War outcome is also not the correct dependent variable to test the theory; instead, public support for war should be examined prior to the onset of fighting. Second, empirical support for the marketplace of ideas is surprisingly sparse; I argue that the public discussion aspect of the marketplace is moribund and thus scholars should focus on the claim that superior civil-military relations facilitate democratic success. Third, secrecy cannot be reconciled with selection effects because it can both increase and decrease the likelihood of victory; its effect on the likelihood of victory is thus indeterminate. Lastly, the quantitative evidence for democracy and victory—as well as for a variety of other positive conflict outcomes—is much more contested than Reiter admits.

VIETNAM

I now turn to Reiter’s reaction to my analysis of the U.S. decision to engage in war in Vietnam, which exemplifies some of the issues discussed above. In my International Security article I argued that President Johnson chose to initiate war against North Vietnam despite deep pessimism among top U.S. political and military officials regarding the ability of the United States to win. I contended that this decision contradicted the selection effects theory’s prediction that democratic leaders start only those wars they believe they can win. Reiter does not dispute my core claim, but he does raise five other objections to my interpretation of the case.[292] First, he argues it is not clear that the United States initiated or lost the war. Second, Reiter claims that the marketplace of ideas actually functioned quite well in the decision-making that led to Vietnam. Third, he asserts that Johnson’s decision to fight was popular at the time, and thus was consistent with the selection effects mechanism that leaders start wars they believe will be popular. Fourth, Reiter contends that because I argue that Johnson was motivated to fight in Vietnam by domestic politics, this is consistent with selection effects. Finally, Reiter claims that even though the chances of victory in Vietnam were relatively low, the public perceived the stakes involved to be so high that fighting was worthwhile. Overall, Reiter concludes that Vietnam “demonstrates many of the patterns predicted by selection effects theory.”[293] I address each of these points in turn.

Initiator and Outcome of the Vietnam War

Reiter’s first counterargument is that the United States did not clearly initiate the Vietnam War and also did not lose it. On the war’s outcome, I never claimed the United States lost the war,[294] although this is a common interpretation. The U.S. war aim in the conflict was to preserve an independent, noncommunist South Vietnam. The peace settlement that ended U.S. participation in the war in January 1973 partially achieved this outcome: the Republic of Vietnam continued to exist, but roughly 200,000 North Vietnamese soldiers were permitted to remain south of the border. The settlement in reality purchased a “decent interval” to separate the U.S. withdrawal from South Vietnam’s ultimate defeat, which did not occur until April 1975. For their part, the North Vietnamese failed to conquer South Vietnam but they did obtain the U.S.’s exit from the conflict. Many analysts code the U.S.’s inability to impose its will on Hanoi as a defeat.[295] I think it is more accurate to code the war as a draw because each side achieved some but not all of its objectives. Because risk-averse, forward-looking democratic leaders want to avoid bloody eight-year battlefield stalemates, however, I argue that selection effects theory predicts that Johnson should have avoided war in Vietnam.

Reiter does not contest my decision to code the United States as the initiator of the war against North Vietnam. After repeating my argument (drawn from a source not at all sympathetic to the communist side) that the Pleiku bombing by the Viet Cong in February 1965 was not ordered by Hanoi, Reiter suggests that U.S. leaders perceived the attack to have been instigated by North Vietnam, and thus “the decision to escalate was seen more as a reaction to attack rather than a war of choice.”[296] Reiter neglects to mention that LBJ had already decided to initiate a campaign of sustained bombing of North Vietnam on December 1, 1964, and sought only a propitious provocation to justify his attack. Johnson’s National Security Advisor McGeorge Bundy said at the time that “‘Pleikus are streetcars’—meaning that the United States had been waiting for such an incident and planned to make use of it.”[297] The United States was the first state actor to use military force against another state, which is the criterion used by the Correlates of War project to identify war initiators, and the same criterion endorsed by Reiter and Stam in Democracies at War.[298] Recoding war initiation along the lines suggested by Reiter—whether a state’s leaders perceived themselves to be acting defensively or offensively—would be an arduous task and could result in cases where both sides or neither are coded as the initiator.

The Marketplace of Ideas and Vietnam

Reiter’s second counter-claim is that “the marketplace of ideas seems to have worked reasonably well, as many components of government (including the White House) correctly understood that the Vietnam War was a risky bet in which victory was not guaranteed.”[299] Reiter refers here to the assessments of the U.S.’s military prospects in the war offered by both civilian analysts and military officers, which forecast that victory was not likely even if the United States introduced hundreds of thousands of ground troops into South Vietnam. He does not address a large body of evidence that the discussion aspect of the marketplace was subverted and rendered ineffective by the Johnson administration. Here I first provide evidence on how Johnson avoided a full public debate on the merits of escalation in Vietnam and the level of deception the president practiced on Congress, the public, and even members of his own administration. Second, I demonstrate that although civilian and military officials correctly predicted that fighting in Vietnam would be a tough, costly, and protracted slog, these pessimistic forecasts had little impact on decisions to escalate. Moreover, civil-military relations under Johnson bore little resemblance to Reiter and Stam’s ideal: LBJ largely excluded military officers from planning and decision-making on Vietnam, relied almost exclusively on civilians for military advice, and co-opted members of the JCS to prevent them from publicly airing their disagreements with administration strategy. Johnson wanted the credibility that their uniforms lent to his preferred strategy of gradual escalation, not their professional military advice.

Administration Subversion of Public Discussion.  The discussion logic implies that a multiplicity of voices should participate in a free-wheeling debate on going to war that improves the quality of information available, helping to produce better decisions. One is struck, therefore, by the narrow range of the debate on Vietnam within the Johnson administration, as well as the extent to which the president restricted debate to a few individuals and concealed his decisions for war not only from the public and Congress, but also from second-tier officials within the government who disagreed with escalation.[300]

One of the remarkable facts about the decisions for war in 1964 and 1965 is the narrowness of the debate both inside and outside the administration. Certain options were simply out of bounds. Neither the Kennedy nor Johnson administrations, for example, ever seriously pursued negotiations as an alternative to standing firm or escalating in Vietnam. Kennedy administration officials opposed any diplomatic settlement until the insurgency in South Vietnam had been defeated, and Johnson repeatedly rebuffed feelers for talks well into 1965, including two orchestrated by U.N. Secretary General U Thant.[301] Just as negotiations were beyond the pale, any discussion of neutralizing South Vietnam was similarly anathema. When Senate Majority Leader Mike Mansfield presented the president with a memorandum in January 1964 suggesting neutralization as a possible way out of Vietnam, Johnson’s top aides blasted it. Bundy, Secretary of State Robert McNamara, Secretary of State Dean Rusk, and others pointed to the apocalyptic international and domestic consequences of accepting a neutral South Vietnam (since they equated neutralization with communization), including falling dominoes in Southeast Asia, a decline in U.S. prestige, doubts about the credibility of U.S. commitments to defend other countries, and the possible evisceration of the administration’s domestic effectiveness because it “lost” Vietnam.[302]

Suggestions that the United States should withdraw its forces from South Vietnam, finally, were enough to get officials—even the vice-president—banned from participation in further deliberations on Vietnam. Vice-President Hubert Humphrey gave Johnson a memo on February 17, 1965, in which he advised the president to cut rather than enlarge the U.S. presence in Vietnam. Arguing that “American wars have to be politically understandable by the American public” to “enjoy sustained public support,” Humphrey warned that in Vietnam, the public simply “can’t understand why we would run grave risks to support a country which is totally unable to put its house in order.” Citing the unpopularity of the stalemated Korean War, Humphrey presciently cautioned that should the United States find itself “embroiled deeper in fighting in Vietnam over the next few months, political opposition will steadily mount….with serious and direct effects for all the Democratic internationalist programs to which the Johnson Administration remains committed: AID, United Nations, arms control, and socially humane and constructive policies generally.” Humphrey’s dissent merely infuriated Johnson and resulted in the vice-president’s banishment from policymaking on Vietnam.[303]

Much has been made of Undersecretary of State George Ball’s October 1964 memorandum as an example of the consideration and acceptance of dissenting views within the Johnson administration. One problem with this view is that Ball’s paper received limited distribution within the government; the president, for example, did not read it until late February 1965, after he had decided to initiate sustained bombing of North Vietnam. Moreover, it is not clear that Ball’s views were taken seriously in the administration since some now believe that he had accepted the role of “house dove” assigned by the president “to shoot holes” in the plans to fight in Vietnam.[304] Ball denied this in his memoirs, but others in the administration seem to have understood this to be his role. Regardless of whether Ball was playing a role or not, he was virtually alone among the upper echelon of officials in suggesting options other than escalation for Vietnam. Ball did not even do this consistently, as he often kept silent or supported the administration’s policies in internal debates or discussions with Congressmen. Ultimately, his loyalty to the president kept him on board and prevented him from breaking with the path Johnson had chosen.[305]

The Johnson administration also practiced an exceptional level of deception on Congress and the public. The Tonkin Gulf episode in August 1964 is a good example.[306] On August 2, three North Vietnamese patrol boats approached the U.S.S. Maddox, a destroyer on an intelligence-gathering patrol near the North Vietnamese coast. Perceiving that the vessels were intent on attacking, the Maddox fired on and damaged one of the ships. The North Vietnamese were retaliating for a covert attack on their coast under OPLAN 34-A that had taken place in the area on the night of July 30-31, but the actual attack on the Maddox was unauthorized; U.S. intelligence intercepted a message to the three boats canceling the mistaken order to attack. The president and Bundy both understood that the OPLAN 34-A operation had caused the attack on the Maddox, but the White House ordered a continuation of the destroyer patrols (known as DeSoto patrols), and specifically ordered another 34-A mission for the next day. On August 4, the Maddox reported it was under attack again. Administration officials, who had already discussed retaliatory options if a second attack occurred, decided to strike back. Later in the day, Admiral Sharp, commander of U.S. forces in the Pacific, and Captain Herrick of the Maddox contacted Washington and raised doubts about the attack: Herrick reported that it now appeared that the sonar contacts and torpedo firings were doubtful, and carrier pilots on the U.S.S. Ticonderoga were unable to find any enemy vessels in the area. Later scholarly analysis concluded that a second North Vietnamese attack almost certainly did not occur.[307] Nonetheless, Johnson and his top aides decided to launch a retaliatory strike against North Vietnam, and quickly followed that up by sending the Gulf of Tonkin resolution to Congress, which permitted the president to use armed force to prevent further attacks on U.S. forces and to defend any ally in Southeast Asia from aggression.

In order to obtain Congressional authorization for the resolution, however, administration officials departed significantly from the truth in their version of events in the Tonkin Gulf. Secretary McNamara, for example, told the Senate Foreign Relations Committee that the Maddox had been on a “routine patrol” rather than a classified mission to obtain electronic intelligence.[308] This was obviously false. He also said that the Navy knew nothing about any South Vietnamese operations against North Vietnam and denied any connection between such operations and the destroyers’ presence in the Gulf. McNamara of course knew the two were connected, and knew that was also the North Vietnamese understanding. As Director of Central Intelligence John McCone put it in an NSC meeting on August 3, “The North Vietnamese are reacting defensively to our [Oplan 34-A] attacks on their off-shore islands.”[309] The administration also neglected to mention that it had ordered another covert attack on North Vietnam and a continuation of the DeSoto patrols after the first incident, hoping to provoke a reaction from Hanoi that would justify retaliation and a Congressional resolution.[310]

It is also startling the lengths to which Johnson and his advisors went to stifle debate on the war and to conceal their decision to wage war against North Vietnam. According to George Herring:

Johnson thus took the nation into war in Vietnam by indirection and dissimulation. The bombing was publicly justified as a response to the Pleiku attack and the broader pattern of North Vietnamese ‘aggression,’ rather than as a desperate attempt to halt the military and political deterioration in South Vietnam. The administration never publicly acknowledged the shift from reprisals to ‘sustained pressures.’ The dispatch of ground troops was explained solely in terms of the need to protect U.S. military installations, and not until June, when it crept out by accident in a press release, did administration spokesmen concede that American troops were authorized to undertake offensive operations.[311]

Similarly, Logevall writes that “Johnson in the first half of 1965 escalated the Vietnam War by stealth. More than that, he set out to quash debate on the war, to manufacture consensus….to keep the American people in the dark about Vietnam, to foster apathy, to conceal for as long as possible the Americanization of the war.”[312] The president, for example, repeatedly stymied Senate debates on the way forward in Vietnam proposed by Senator Mansfield by deploying his aides to conduct personal discussions with senators and dissuade them from airing their differences with the administration in public. In a meeting with key legislators on January 21, 1965, for example, Johnson asked for bipartisan support for his policies and swore participants to secrecy. Briefings at this meeting by Rusk and McNamara, however, significantly misrepresented those policies, claiming that the United States was open to negotiations and a multi-power conference on Vietnam (it was not) and that significant military progress was being made by the South Vietnamese (also not true).[313] But the biggest whopper came from Johnson himself: the president, who had cabled Maxwell Taylor in Saigon three weeks earlier to indicate his interest in the deployment of U.S. ground troops, told the assembled legislators that “[t]he war…must be fought by the South Vietnamese. We cannot control everything that they do and we have to count on their fighting their war.”[314]

A few weeks later, when Johnson endorsed Bundy’s policy of sustained reprisal, he refused Bundy’s recommendation to inform the public, preferring to pretend that no policy shift had taken place. Indeed, as the bombing got underway, some commentators wondered why the president had made no public statements regarding U.S. policy even as American involvement in the war deepened.[315] “In Washington,” writes Robert Mann, “there was no public announcement that the United States was finally at war with North Vietnam.”[316] The president at this time (late February-early March) was busy suppressing dissent in the Senate, and succeeded in keeping the most important Senate Democrats who opposed enlarging the war—Mansfield, Chairman of the Senate Foreign Relations Committee J. William Fulbright, and Johnson’s mentor Richard Russell—from voicing their dissent publicly. On April 1, the president in a press conference denied knowing of any plans in the U.S. government for expanding the war, and the next day Bundy told participants at an NSC meeting: “Under no circumstances should there be any reference to the movement of U.S. forces or other courses of action.”[317] This campaign of deception continued well into June, when Johnson told Fulbright that he had resisted the military’s attempts to escalate and had worked hard to open discussions with the North.[318] Finally, when Johnson announced to the country on July 28 that he was sending additional troops to Vietnam, he first misrepresented the true number he had approved (50,000 instead of 100,000), and then denied that these deployments constituted any change in policy.[319]

Perhaps the best illustration of Johnson’s systematic “policy of minimum candor” was the flap that arose in response to the statement by a U.S. State Department spokesman on June 8, 1965, regarding the mission of U.S. troops in Vietnam.[320] Two months earlier, when LBJ approved the deployment of two additional Marine battalions and 18,000 to 20,000 support troops, the president had also signed off on a change of mission for the Marines from static base security to a more active patrolling and combat role. Formalized in NSAM-328, Johnson concealed this decision to allow American troops to engage in combat from all but his top three advisors. On June 8, however, the spokesman told the press that U.S. troops were “available for combat support together with the Vietnamese forces as and when necessary.” Although the White House quickly issued a denial, the announcement caused widespread comment and provoked the following editorial by the New York Times: “The American people were told by a minor State Department official yesterday, that, in effect, they were in a land war on the continent of Asia. This is only one of the extraordinary aspects of the first formal announcement that a decision has been made to commit American ground forces to open combat in South Vietnam: The nation is informed about it not by the President, not by a Cabinet member, not even by a sub-Cabinet official, but by a public relations officer.”[321]

The president thus systematically concealed his decisions to escalate the conflict from the public, worked to suppress debate in Congress, and lied to senators (and the American people) about steps he had taken. Rather than encourage debate, the administration squelched it. Officials in the executive and legislative branches who dissented from the steadily escalating commitment were excluded from the policymaking process or misled about decisions that had been made.

Quality of Strategic Assessment.  What about Reiter’s claim that the strategic assessment plank of the marketplace of ideas worked effectively because Johnson had good information about what he was up against in Vietnam? As I demonstrated in my article, the bulk of the evidence on this question is that after 1963, civilian policymakers had access to accurate information about the nature of the political and military situation in South Vietnam, and the likelihood that using force against North Vietnam would help the South defeat the Viet Cong insurgency.[322] That information was overwhelmingly pessimistic: the Viet Cong gained strength and ground throughout 1964 while coup followed coup in Saigon. Moreover, few believed that deploying U.S. air power against the North would do much to improve things in the South. Even with the infusion of massive American ground power, all signs pointed to a protracted war of attrition. Yet with all of this accurate information in hand, Johnson and his men continued to escalate U.S. involvement in Vietnam. Officials were getting the bad news by 1964, but it did not have much of an impact on their willingness to escalate. As Logevall has put it, “To independent observers, this reality foretold gigantic, perhaps insurmountable, problems in any defense of South Vietnam…to the top figures in the Johnson administration, it necessitated merely a redoubled effort.”[323]

Indeed, administration officials seemed to be able to convince themselves that the course of action they wanted to take would have positive effects that previous analysis and neutral observers did not find credible. A good example is Bundy’s recommendation for sustained reprisals of February 8, 1965. First of all, Bundy’s report—contradicting all evidence—claimed that morale in South Vietnam was not nearly as bad as previously thought and that the South Vietnamese Army’s ability to engage in combat was improving. Bundy went on to argue that bombing the North would improve morale in the South, get contending political factions to work together, weaken support for the Viet Cong, and perhaps persuade Hanoi to halt its support for the insurgency. As Logevall notes, however, “Each of these claims about the benefits of bombing North Vietnam went against the bulk of U.S. intelligence analysis dating back several months.”[324] Second tier State Department officials who saw Bundy’s report were stunned to see these assertions—which they believed had been thoroughly discredited—recycled.

The historical record is replete with incongruous episodes of this type, many of which I reported in my article. In September 1964, for example, LBJ declared that he would not begin bombing North Vietnam until the political and military situation stabilized in Saigon. When improved conditions failed to materialize, however, Johnson and his advisors argued that bombing was necessary to stabilize the situation.[325] Perhaps most striking in this regard is McNamara’s paper from July 20, 1965, that formed the basis for the crucial discussions in Washington on the escalation to ground troops. McNamara’s memo, after recounting a rapidly deteriorating military situation, recommended an increase to 175,000 troops by the end of 1965, a further increase of 100,000 in 1966, and possibly more to follow, all while conceding that the Viet Cong might simply revert to a strategy of protracted war. McNamara, in complete contradiction to his own arguments, then concluded that this plan offered “a good chance of achieving an acceptable outcome within a reasonable time in Vietnam.”[326]

In short, the administration substituted hope for analysis and chose to continue escalating in Vietnam despite having access to highly prescient assessments of the consequences of their actions. The question, therefore, is whether one can claim that strategic assessment was effective if policymakers—fully aware of the perils of their preferred path—simply ignored the advice and continued to advocate policies they knew would lead to an unhappy outcome.

An even more damning critique of strategic assessment on Vietnam—especially given the weakness of the public discussion plank of the marketplace of ideas detailed above—is the malignant state of relations between the president’s legally designated military advisors—the JCS—and the civilian leadership. Reiter and Stam argue that because civilian control of the military is robust in democracies and the military does not represent a threat to political leaders’ tenure, officers are promoted solely on merit and thus provide leaders with expert military advice that allows civilian elites to choose wisely which wars to enter and which to avoid. In his essay, Reiter provides no direct evidence from the Vietnam case to support this argument; instead, he cites the fact that decision-makers understood that the odds of success were poor as proof that the system of military advice must have worked as posited. Again, as with the public discussion argument, the available evidence tells a very different story. In fact, civil-military relations under Johnson diverged in several important ways from the Reiter and Stam model that had deleterious consequences. Johnson and McNamara largely excluded the JCS from the planning process, and misrepresented or even suppressed the military’s views on strategy in Vietnam. Moreover, relations between soldiers and civilians were highly politicized, which allowed Johnson and McNamara to shape the advice they received from the chiefs.

The fundamental difficulty that Johnson faced in his relationship with the Joint Chiefs was that the chiefs in 1964 favored a much more aggressive approach to containing communism in Southeast Asia than did the president.[327] Johnson, influenced heavily by his civilian advisors, preferred a course of gradual escalation and the use of force primarily to send signals of American resolve to stand firm in Vietnam rather than to achieve outright victory. Gradualism also fit well with Johnson’s domestic political priorities, which were to win the November 1964 election and achieve passage of his Great Society programs before taking any serious action one way or the other on Vietnam. Finally, Johnson feared repeating the mistakes of U.S. intervention in Korea, where the U.S.-led invasion of North Korea triggered Chinese entry into the war, leading first to a disastrous U.S. retreat and then to a costly and protracted battlefield stalemate. The JCS, by contrast, wanted from the president a clear commitment to winning in Vietnam and a rapid escalation to forceful measures against Hanoi necessary to achieve victory. This divergence between the JCS and the president was highlighted soon after Johnson took office by a memorandum the chiefs sent to McNamara in January 1964, which rejected limited force and called for more aggressive actions against North Vietnam, including air attacks and sea mines.[328] Johnson, faced with military advice that contradicted his own objectives and strategy, neither changed his approach to the war in response to the chiefs’ views nor openly rejected their recommendations. Rather, Johnson—aided by McNamara—moved to sideline the JCS from the planning process but keep them on board so that he could benefit from the credibility their uniforms lent to his administration’s policies.

The leading study of U.S. civil-military relations in the period of U.S. escalation in Vietnam—H. R. McMaster’s Dereliction of Duty—provides an abundance of evidence regarding the ways in which the president and his secretary of defense marginalized and manipulated the JCS. First, McMaster argues that Johnson and McNamara worked to exclude the JCS from the planning and decision-making process on Vietnam. Before his trip to South Vietnam in early March 1964, for example, McNamara promised the chiefs that they would be able to review his report before the president acted on it. Little did they know that William Bundy had already written two drafts of the report before McNamara even departed for Saigon and that Johnson had read the document twice before the chiefs even saw it.[329] Beginning in May, McNamara ended the JCS’s ability to speak directly to the president by requiring that all communications from the chiefs go through his office rather than straight to Johnson.[330] The role of the Joint Chiefs was further diminished that summer when Johnson appointed former JCS chairman Taylor as ambassador to Saigon and granted him extraordinary powers to coordinate the entire military effort in South Vietnam.[331] Johnson and McNamara also shaped the recommendations the chiefs could make by issuing narrow guidelines that essentially forced the JCS to produce the type of gradualist advice the administration sought.[332] Military participation in important reviews of administration policy on Vietnam—such as the Vietnam Working Group in November 1964 and the decision to fulfill MACV commander Gen. William Westmoreland’s thirty-four battalion request in July 1965—was limited; the chiefs often found themselves reacting to positions developed without their input by civilians in the Defense Department.[333]

The corollary to Johnson’s and McNamara’s exclusion of the JCS from deliberations on Vietnam was their reliance on civilians with little military experience to formulate military strategy and plans. While Johnson distrusted his military advisors, McNamara simply thought that “military experience mattered little” and that “JCS advice” in particular “was inapplicable.”[334] Johnson’s inner circle consisted of McNamara, Rusk, and Bundy. This group met for lunch most Tuesdays starting in early February 1964. Although Vietnam was often on the agenda, no military representative was invited until the war was well underway in 1966. Furthermore, none of the second-tier officials who helped draft plans for military action in Vietnam—men like Paul Nitze, John McNaughton, William Bundy, and Roswell Gilpatric—had any combat experience. As McMaster puts it, “William Bundy of the State Department and McNaughton of the Defense Department became the principal planners for Vietnam.”[335]

Second, McMaster argues that Johnson and McNamara politicized relations with the JCS by cultivating close relationships with members of that body to shape the advice they dispensed.[336] In February 1964, for example, after the chiefs voted to recommend an Air Force general for the job of commander-in-chief of Pacific Command, a post traditionally held by a Navy officer, the Chief of Naval Operations, Admiral David McDonald, protested to McNamara. The defense secretary intervened and gave the job to Admiral Ulysses Sharp. According to McMaster, however, “McNamara’s intervention on behalf of the Navy had a price—McDonald, who had hitherto opposed McNamara’s plans [for Vietnam], no longer objected to them outside JCS meetings.”[337]

Johnson and McNamara also capitalized on their relationships with JCS chairmen Gens. Taylor and Earle Wheeler. McNamara, for example, exploited his close relations with Taylor when he was chair of the Joint Chiefs to quell dissent within that body. According to McMaster, Taylor “assisted McNamara in suppressing JCS objections to the concept of graduated pressure” and protected Johnson from “the views of his less politically sensitive colleagues while telling the chiefs that their recommendations had been given full consideration.”[338] In one episode in May 1964, Taylor withdrew a memo to McNamara written by the other chiefs without his input that called for sharp escalation against North Vietnam; when his colleagues redrafted the paper and asked him to submit it, he refused.[339] As McMaster puts it, Taylor had “foiled another attempt to question the validity of the assumptions on which the concept of graduated pressure was based.”[340] Relations with Wheeler followed a similar pattern. Johnson and Wheeler became good friends, and this was no accident. As Taylor’s former assistant Gen. Andrew Goodpaster once put it, “Johnson had a way of ‘befriending’ people and then using that friendship to extract acquiescence on controversial issues.”[341] According to Robert Buzzanco, Wheeler “enforced conformity within the JCS” and “pressured the Army and Navy into supporting the air war in order to send unified proposals to McNamara.”[342] Wheeler could also be counted on to keep quiet when civilian officials lied and to mislead his colleagues and members of Congress.[343]

Third, Johnson and McNamara employed a significant amount of deception in their dealings with the JCS. McNamara sometimes misrepresented—and even lied about—the chiefs’ views in high-level administration meetings. The chiefs, for example, lodged serious objections to McNamara’s report advocating gradual escalation in March 1964. Taylor first watered down the chiefs’ language in a memo to McNamara; the Secretary of Defense then quashed even this milder memo and proceeded to tell the National Security Council that the JCS “supported the McNamara report.”[344] Similarly, a year later McNamara reported to the president that the military agreed with the pace of bombing in North Vietnam when in fact Wheeler and Admiral Sharp both recommended it be augmented.[345]

Johnson, for his part, repeatedly misled the chiefs, telling them that although he could not presently implement their recommendations for more aggressive military measures, he would follow them in the future. The president used this ploy to prevent the JCS members from becoming disaffected and to keep them on the team. McMaster identifies a typical interaction between the president and the JCS after the Viet Cong attack on Bien Hoa airfield on November 1, 1964: “The Chiefs tried to use the attack to gain approval for additional actions on the list they had developed less than a week earlier. To keep the Chiefs ‘on board,’ President Johnson and his closest advisors appeared sympathetic to the JCS recommendation and held out the promise of future action” while denying permission to strike now.[346] Five months later, Johnson similarly “promised more determined action in the future” while rejecting a JCS recommendation to deploy three divisions to South Vietnam.[347] McNamara also took this tack with the chiefs to tamp down their frustration over the administration’s refusal to heed JCS advice.[348]

Finally, it is worth noting that although Reiter and Stam’s model assumes that “the military” speaks with one voice in providing advice to political leaders, in fact the armed forces are comprised of several services, each with its own interests and culture, which often disagree over strategy. Indeed, the JCS in the 1960s was riven with interservice rivalries that hobbled its ability to offer unanimous advice on Vietnam.[349] In July 1964, for example, the chiefs were unable to agree on a request from Westmoreland for 4,200 additional troops and failed to submit a recommendation until after Johnson had already authorized the increase. In early August, a debate among the chiefs over Westmoreland’s request for Army aviation assets devolved into an ugly shouting match between Gens. Curtis LeMay of the Air Force and Harold Johnson of the Army. According to McMaster, the Air Force and Marines “questioned the need for more Army aircraft in South Vietnam because they believed that Westmoreland’s request represented a surreptitious attempt to expand the Army’s air arm.”[350]

Disagreements among the chiefs were particularly salient on the question of bombing North Vietnam. Throughout 1964 and 1965, the chiefs feuded over the role of air power. Underlying the dispute were fundamental disagreements on the nature of the war in Vietnam and how to fight it that were never resolved. LeMay and Marine Corps Commandant Gen. Wallace Greene believed strongly that air attacks on North Vietnam would reduce Hanoi’s capability and will to support the insurgency.[351] Harold Johnson of the Army disagreed, arguing that the insurgency was largely indigenous and could not be defeated by the application of air power against the North. These internal disputes erupted periodically, as in September 1964 and March 1965.[352] The chiefs tended to paper over their disagreements in an attempt to present a united front to the civilian leadership, but these compromises and logrolls rendered it difficult for the JCS to formulate coherent strategy.

In sum, civil-military relations during the period when the key Vietnam decisions were made bear little resemblance to Reiter and Stam’s democratic ideal of a meritocratic military providing expert, nonpolitical strategic advice to policymakers, thereby enabling them to select wars that can be won quickly and decisively. Instead, Johnson and his right-hand man McNamara worked assiduously to prevent meaningful military participation in decision-making, politicized relations with the JCS to manage dissent and solicit advice that supported their preferred policies, and deceived the chiefs about the possibility of future escalation to keep them from going public with their frustrations. The chiefs eventually decided to work within the administration’s constraints of graduated force, supporting measures they believed were ineffective in the hope that more aggressive moves would be forthcoming.[353] Lastly, intra-military relations also diverge from Reiter and Stam’s model because rivalries among the services prevented the JCS from providing unanimous recommendations.

A Popular War?

Reiter’s third counterargument is that going to war in Vietnam was popular with the public, and thus the decision to fight is consistent with the selection effects argument that leaders will initiate war if they perceive that public opinion approves. Reiter offers evidence from several Harris surveys that he contends show that the American public supported escalation in Vietnam. We can investigate this claim by scrutinizing public and Congressional attitudes before President Johnson chose to attack North Vietnam. It is sometimes noted that support for taking the war to the North was high after the bombing began and that by inference this means there was solid support for going to war before-hand. This is incorrect. As noted above, research has shown that leaders may receive a bounce in approval when the country becomes involved in external hostilities as people rally behind the flag. In the case of Vietnam, for example, public approval for Johnson’s handling of the situation in Vietnam jumped almost twenty points from before the United States undertook the February reprisal attacks (41 percent approve) to after (60 percent approve).[354] One cannot, therefore, infer pre-existing support for war by observing approval ratings after hostilities commence.[355] This disqualifies most of the polling numbers that Reiter cites as evidence supporting his argument that the war was popular.[356]

In fact, the evidence shows that there was little public pressure for escalating American involvement in Vietnam. The public was not strongly opposed to expanding U.S. participation, but nor was it strongly in favor. Polls in the first half of 1964 showed that most Americans knew little about Vietnam and the war. A Gallup poll from April 1964, for example, found that slightly more than one-third of Americans said that they paid attention to events in Vietnam.[357] Of the few who knew what was going on, opinions were equally split among those who wanted to get out, maintain the current policy, get tougher without using force, fight a war, and the difficult to classify all-or-nothing category of “go all the way or pull out.”[358] A Harris poll from the previous month—which asked more directly about approval for putting “military pressure on the Communists in North Vietnam” at the risk of China entering the war—showed that nearly half of those queried opposed such a move while only a quarter favored it.[359]

Table 1 summarizes public opinion on Vietnam in mid-1964, showing the diversity of views on the question. The Gulf of Tonkin incident in early August caused a spike in the percentage of those polled who preferred to hold the line in Vietnam and a decline in the proportion of the public who wanted the United States to extricate itself from the country. The Gulf of Tonkin—at least in these polls, however—did not lead to an increase in the numbers in favor of going to war, which actually dropped from 12 percent in April to 9 percent in early August and further to 4 percent in late August.[360]

Table 1

Results of Gallup Polls:

“In your opinion, what would you like to see the United States do next about Vietnam?”

April 1964 Early August 1964 Late August 1964
Withdraw completely 11 4 7
Sit down and talk it over NA 10 4
Continue present policy 7 27 10
Step-up present efforts, but no war 12 12 13
Go all-out, declare war 12 9 4
Go all the way or pull out 11 3 7
Don’t know/No opinion 27 30 49

Note: Figures are percentages. Gallup Polls USAIPO1964-0689, 0696, and 0697.

Public awareness of events in Southeast Asia grew as the year went by (spiking around the Gulf of Tonkin episode), but public views about appropriate U.S. policy remained divided. In a Council on Foreign Relations poll taken in December, the proportion of those surveyed who were unaware there was fighting in South Vietnam dropped to one quarter. Of the remaining three quarters, nearly 50 percent were in favor of pulling out, while only about half that many preferred committing U.S. ground troops to defeat the communists. According to a University of Michigan study also from December, when asked if the United States should withdraw from Vietnam right away, twice as many Americans were strongly opposed as were in favor (37 versus 18 percent). But these figures reversed when the question was should the United States commit its own troops to fight in Vietnam: 32 percent opposed such a move, with only 24 percent in favor. According to a Harris poll from the same time, 18 percent backed the idea of bombing North Vietnam, while 20 percent said they preferred to withdraw.[361] Another Harris poll taken at the end of January 1965 showed that a plurality of Americans (40 percent) wanted to continue the current policy of holding the line, but when it came to other options, the public was about evenly split between negotiating and getting out (23 percent) versus carrying the fight to North Vietnam (17 percent). As the figures in Table 2 show, the largest group consistently favored maintaining the status quo, but those who desired negotiations and a quick exit from Vietnam consistently outnumbered advocates of escalation. The gap between the two grew wider in spring 1965 until nearly three times as many Americans preferred negotiations to escalation, but by summer the difference between the two positions was negligible.[362]

Table 2

Results of Harris Polls:

“Which of these three courses do you favor for the United States in Vietnam: Carry the war into North Vietnam, at the risk of bringing Red China into the war; negotiate a settlement with the Communists and get out now; or continue to hold the line there to prevent the Communists for taking over South Vietnam?”

Nov.

1964

Jan.

1965

Feb.

1965

March

1965

Early April

1965

Late April

1965

Late June

1965

Hold Line 40 40 40 46 48 43 36
Negotiate 20 23 23 35 31 28 24
Carry war into North Vietnam 20 17 13 12 17 20 20
Not sure 20 20 24 7 4 9 20

Note: Figures are percentages. Compiled from William Conrad Gibbons, The U.S. Government and the Vietnam War: Executive and Legislative Roles and Relationships, Part 3: January-July 1965 (Princeton, N.J.: Princeton University Press, 1989), pp. 42, 145, 353.

Even after the United States began bombing the North, and strong majorities expressed their approval of the policy, large numbers also hoped the war could be ended by negotiations.[363] A Harris poll in mid-February 1965 showed that while 83 percent supported bombing, 75 percent also favored trying to obtain a negotiated settlement.[364] A Gallup poll from March showed similar numbers, with 81 percent supporting a conference to negotiate a peace agreement.[365] In a head-to-head comparison of escalation versus negotiations in March, Gallup found equal support for each alternative (42 to 41 percent).[366] The diversity of opinion displayed in the Gallup polls in Table 3 suggests that although a majority of Americans favored both bombing and negotiations, when asked to get more specific there was little unanimity on the best course of action. Support for dovish (withdrawal, negotiations) and hawkish (escalation, going all-out) options was fairly similar, with maintaining the status quo also drawing double-digit support.[367]

Table 3

Results of Gallup Poll:

“In your opinion, what would you like to see the United States do next about Vietnam?”

April 23 May 16 June 9 July 2
Withdraw completely 17 13 12 13
Start negotiating, stop fighting 12 12 16 11
Continue present policy 14 13 20 16
Step-up present efforts 12 8 4 6
Go all-out, declare war 19 15 17 17
No opinion 33 39 33 37

Note: Figures are percentages. See William Conrad Gibbons, The U.S. Government and the Vietnam War: Executive and Legislative Roles and Relationships, Part 3: January-July 1965 (Princeton, N.J.: Princeton University Press, 1989), p. 353.

As Harris concluded in a February 22 report for the Johnson administration, “The clear mainstream of America [sic] opinion is this: We should shore up the efforts of the South Vietnamese to resist further Communist advances, use retaliatory airstrikes only when extreme guerrilla activity warrants it, and when we have made enough show of power so that the Communists can see that we will not yield, then finally negotiate a settlement.”[368]

Just as the public was not crying out for escalation in Vietnam, nor was Congress. Almost all of the leading skeptics of increasing American involvement in Southeast Asia were from the president’s own party, and included several of the Senate’s most influential members: Majority Leader Mansfield, Foreign Relations Committee Chairman Fulbright, and Russell. As discussed above, Johnson had to work hard to prevent these men from voicing their views publicly. Although the Gulf of Tonkin Resolution passed unanimously in the House and with only two negative votes in the Senate, this should not be interpreted as a sign that legislators were enthusiastic about the prospect of war. Many legislators made clear that they were supporting retaliation for specific attacks, not advocating a wider war.[369]

In short, before the bombing of North Vietnam began, opinion in the United States was divided over the wisdom of taking military action against Hanoi and sending U.S. troops to South Vietnam. After the bombing started, large majorities supported it but at the same time supported attempts to negotiate a way out of Vietnam. The evidence indicates that rather than following a public mandate for escalation, LBJ led a public that was uncertain regarding the correct course in Vietnam but was willing to follow its commander-in-chief. But follow him how far? The president’s old friend Mike Mansfield warned Johnson that the public did not support his policies because they understood or were strongly committed to them; rather, their support was based largely on his status as president. If the war began to go badly, this support would evaporate.[370] Johnson himself, according to Moyar, “remained convinced that domestic opinion did not provide compelling grounds for intervention in Vietnam,” noting that although the public appeared to support his policies now, “if you make a commitment to jump off a building, and you find out how high it is, you may withdraw the commitment.”[371]

Domestic Politics Does Not Equal Selection Effects

A fourth critique leveled by Reiter is that because I argue that Johnson considered domestic politics when he chose to go to war in Vietnam, my argument is “consistent” with selection effects. This statement betrays a misconception that any argument that relies on domestic politics is necessarily consistent with selection effects. Reiter, as noted, lays out three mechanisms that selection effects theory posits can lead to war: the leader believes victory is likely, the war is likely to be popular, or the leader is dragged unwillingly into war by a fire-eating public. I argued that LBJ felt compelled to fight in Vietnam because if he did not, the Republican opposition would torpedo his cherished Great Society legislative program. This motivation does not fall into any of Reiter’s three categories.[372]

Not a High Stakes War

The final argument Reiter makes is that although the American people understood that victory was not likely in Vietnam, “the high stakes merited escalation regardless.”[373] Reiter offers no evidence in support of this assertion. There is plentiful evidence to the contrary, however. Table 4 summarizes several Gallup polls from 1964 that asked people what they thought was the most important issue facing the country. The percentage of respondents who said Vietnam is compared to the proportion who indicated racial discrimination or civil rights, which during this period was consistently the first or second most important issue in the public consciousness. As the figures in the table clearly show, Vietnam was not foremost in the minds of the vast majority of Americans prior to the Johnson administration’s decision to bomb North Vietnam in early February 1965. Even immediately after the Gulf of Tonkin incident, a mere 14 percent of those polled thought Vietnam was the most important issue confronting the nation. This figure quickly receded to 6 percent by the end of the same month. Vietnam began to eclipse civil rights only after the war officially began, but still less than 30 percent of the public named it as the preeminent problem facing the United States. It is hard to imagine that if the stakes involved in Vietnam were as high as Reiter claims they were that less than one-third of the population would rank the conflict as the country’s most important problem even after it had started. Of course, only a very few felt this way before the war began. This evidence undercuts his claim that the magnitude of the stakes in the Vietnam War drove people to support fighting it even though the prospects for victory were poor.

Table 4

Results of Gallup Poll:

“What do you think is the most important problem facing this country today?”

April 1964 July 1964 Early Aug. 1964 Late Aug. 1964 Feb. 1965 May 1965 July 1965 Sept. 1965
Vietnam

 

2 3 14 6 29 23 37 19
Racial Discrimination or Civil Rights NA 58 36 46 24 23 20 25

Note: Figures are percentages. Gallup Polls USAIPO1964-0689, 0695, 0696, 0697, and USAIPO1965-0706, 0711, 0714, and 0717.

CONCLUSION: TAKING STOCK OF THE DEBATE

In this essay I have argued that selection effects theory is flawed and that the empirical evidence—both quantitative and from the important case of Vietnam—does not support it. The most important problem for the theory is the conflict between the idea that leaders seek contemporaneous consent for their policies and the way in which the theory is tested, which relies on retrospective judgments of policy success. A theory based on contemporaneous consent predicts that publics reward or punish leaders based on whether they did what was popular at the time a decision was made, not whether the policy ultimately succeeded or failed. Testing such a theory by looking at war outcomes thus does not fit well with the theory’s logic, but even if it did, proponents would have to confront quantitative evidence that democracies do not prevail more frequently than non-democracies. The proper way to test the theory would be to look at public support for war as a policy option at the time a leader chose to fight rather than whether the war was won or lost. Performing such an exercise is perilous, however, because democratic leaders—contrary to the theory’s depiction of them as mere followers of public opinion—have powerful tools at their disposal to manufacture consent for war. This observation reveals the second problem with selection effects theory: the marketplace of ideas does not function as predicted to inform decisions for war and peace or constrain leaders from launching wars for which there was little public impetus. Leaders are able to shape public opinion via deception, shifting the blame for war onto the adversary, exaggerating threats, and downplaying or concealing the probable costs of fighting. Military advisors who oppose a leader’s strategy can be excluded from the debate, deceived into believing that their views are being taken seriously when they are not, or co-opted into providing the advice a leader wants to hear.

These problems manifested on the Johnson administration’s road to war in Vietnam. There was no consensus among the American people that attacking North Vietnam or fighting a major ground war in South Vietnam was a good idea. Johnson sought to escalate the war by stealth, painting Hanoi as the aggressor against the United States and capitalizing on events like the Gulf of Tonkin, Pleiku, and others to rally support for air strikes and eventually a ground commitment. The administration deceived the public and its elected representatives on numerous occasions and the president used his influence to avoid substantive debate in Congress on the wisdom of fighting in Vietnam. The JCS—the president’s designated military advisors—were largely unable to influence the administration’s decision-making because they were excluded from Johnson’s inner circle. Johnson and McNamara kept them on board, however, by leveraging their personal relationships with some of the chiefs (including chairmen Taylor and Wheeler) and by promising that more aggressive policies would be forthcoming in the future. In this way, Johnson muzzled military dissent and benefited from the credibility that the chiefs’ uniforms lent to his strategies.

I see two possible ways forward for proponents of selection effects theory. First, they could try to repair the logical inconsistencies inside the theory by abandoning contemporaneous consent and adopting a retrospective model of voting. This would make the theory internally consistent, but would still face the challenges posed by evidence that democracies are not more likely to prevail in war, that democratic leaders’ likelihood of regular removal from office is unaffected by victory or defeat in crises or wars, that the audience costs argument that heavily informs the theory is empirically questionable, and that audience costs and selection effects are not unique to democratic institutions. Second, proponents could retain contemporaneous consent and drop war outcomes as the dependent variable, instead investigating the prediction that leaders initiate war when the public supports or demands it. Going this route would entail grappling with the point raised by Desch, Schuessler, and myself that public support is not an independent variable. Acknowledging this fact would force selection effects theorists to confront tough questions about the conditions under which leaders choose to initiate or join wars beyond the facile assumption that they somehow know they will win.

More broadly, the study of military effectiveness would benefit by refocusing on explanatory factors that cross-cut regime type. Institutional constraints, for example, are not unique to democracies; all governments have institutions that may improve or detract from their ability to choose which wars to enter and how they fight those wars. Similarly, almost all modern states have professional militaries, and the relations between civilian and military elites can explain variation in military performance that cuts across regime type. The same can be said for cultural or identity-based factors.[374] One can imagine theories of institutional constraints, civil-military relations, and identity that explain military effectiveness regardless of regime type.[375] In my view, this is the research challenge for scholars of military power and military effectiveness.

Essay by Dan Reiter, Emory University

“Democracy and Victory: Final Thoughts”

I thank Michael Desch and Alexander Downes for their continued engagement with this topic. This concluding essay is too short to address all of their points, or even to treat adequately the handful of themes I touch on here. In this essay I will mostly focus on issues which emerged in their papers which were not addressed in my “Closer Look” essay.

A BIT OF HOUSEKEEPING: WHAT SELECTION EFFECTS THEORY DOES AND DOES NOT SAY

Selection effects theory proposes that democracies win wars for two reasons. First, because elected leaders are especially averse to initiating unpopular foreign policies, they are more likely than autocrats to avoid starting wars they will go on to lose. Second, the open marketplace of ideas provides more information to an elected leader, making it more likely (as compared to an autocrat) that she will avoid misestimating her chances of success, and initiating a war the country is doomed to lose.

Examining selection effects theory, Downes explores the theoretical distinction between contemporaneous consent and electoral punishment (retrospective voting). He asserts that my coauthor Allan Stam and I “prefer a model termed ‘contemporaneous consent’” (p.5). This is an incorrect reading of our 2002 book, which makes more nuanced claims.  We stated in our conclusion: “Our view is that the true nature of the relationship between public consent and democratic foreign policy-making contains elements of both [contemporaneous consent and electoral punishment] models and likely evolves over time.”[376]

I disagree with Downes’ assertion that a model which allows for both contemporaneous consent and electoral punishment suffers from a “fundamental contradiction” (p.5). There is no logical inconsistency between the theoretical assumptions of the two strands. A leader cares about contemporaneous consent because adopting policy which contravenes contemporaneous public sentiment can threaten that leader’s hold on power, ergo the 1962 and 1990 comments of John F. Kennedy and George H. W. Bush that each would be impeached if he appeared weak or failed in the Cuban Missile Crisis or Gulf War, respectively. Aside from removal from office, adopting unpopular policies can diminish an elected leader’s power in other ways, such as hurting his or her party in midterm legislative elections. Further, it is conceivable that contemporaneous consent and electoral punishment could make different predictions (e.g., when the public demands war but the leader thinks war would go poorly), but this is not a contradiction. The two assumptions can logically coexist. An empirical approach is more productive: identifying cases in which the two logical assumptions make predictions in different directions, and then collecting data to see which is more often correct. Importantly, this theoretical distinction has no real empirical consequence. Downes points to no examples of contemporaneous consent making opposite predictions from electoral punishment.

Downes also states “war outcomes are not the proper dependent variable to test the theory” (p.6), as contemporaneous consent predicts that leaders do what is popular with the majority at the outset of war. A more constructive approach is to consider multiple dependent variables, rather than a “right” or “wrong” variable. Retrospective voting predicts that democracies are much more likely to initiate wars which promise to be lower cost and successful. There is strong evidence demonstrating these patterns. Contemporaneous consent predicts that democracies only launch wars when there is popular support. There is support for this claim, as well. One study, cited in our book, found that since World War I there has been only one instance in which a democracy participated in a war and faced at the outset public opposition sufficient to threaten the stability of the government.[377] For the United States, there has been public support at the beginning of all its major wars since the Civil War, including the Spanish-American War, World War I, World War II, Korea, Vietnam, the Gulf War, Afghanistan, and Iraq.

Downes disagrees with my assessment that there was contemporaneous consent for the escalation of the Vietnam War, despite polls showing support for escalation. He takes the extreme view that post-Gulf of Tonkin and post-Pleiku polls mean little because they are contaminated by rally ‘round the flag effects. His view assigns too much power to rally effects, which are only occasional and are always short-lived.[378] The questionable validity of his argument aside, polls revealed American public support for the defense of South Vietnam even before the August 1964 Gulf of Tonkin Incident. For example, a July 1964 poll revealed a plurality of Americans (35%) supporting all-out victory in South Vietnam, with an additional 25% supporting holding the line.[379] Downes also argues that there may have been support for negotiations as well as war, but the domestic political viability of a negotiations option does not disprove the claim that there was support also for the escalation option.

Another issue is whether the public saw important issues at stake in Vietnam. As I noted, 79% of respondents in a June 1965 poll agreed that if the United States did not stand fast in Vietnam, then the Communists would take over the rest of Southeast Asia. More generally, Americans believed that the stakes were sufficiently important to warrant running very serious risks. In a February 1965 poll, a plurality (47%) indicated that the United States should continue its present course in Vietnam even at the risk of nuclear war.[380]

Downes proposes that leaders can generate public consent, asserting that Stam and I concede that leaders do so. This is an inaccurate assessment of our theory. We stated in our book: “The open marketplace of ideas in democracies helps undercut fallacious claims made by leaderships about foreign threats, however, reducing the ability of leaders to shape public opinion. In contrast, authoritarian leaders have an easier time controlling the flow of information to create an unchallenged image of the enemy. Though democratic leaders can sometimes shift public opinion at the margins, in general it is more accurate to think of the public as controlling the flow of information to create an unchallenged image of the enemy.”[381] Downes also mis-frames one quote from our book. Our quote is “there is great variability among democracies in different eras as to the conditions that are sufficient to generate public consent for the use of force against some other state,” implying, according to Downes, that “leaders in fact have a fair amount of leeway in persuading publics that something is in ‘the national interest’” (p.7).This was not our point. Rather, we proposed simply that what the public considers a casus belli varies across historical eras. In the next sentence from our book, we stated that “what is a reasonable cause for war for one public may not be reasonable for another.”[382]

Downes’ summary aside, is our assumption empirically valid? Here, consider two different categories of consent manipulation: deceiving the population about the likelihood of winning, and deceiving the population about the stakes. Desch and Downes provide no examples of a leader manipulating a population’s belief about the probability of winning, that is, knowingly persuading the population to have an inappropriately high belief in the likelihood of victory. The absence of such cases is not surprising. Leaders avoid exaggerating the likelihood of victory, because, if the war goes worse than predicted, leaders are more likely to suffer adverse domestic political consequences. The closest case might be the Vietnam War, but any attempts to inflate public optimism about victory failed. As I demonstrated in “A Closer Look,” the public had in 1964 and 1965 an appropriately sober view of the likelihood of eventual victory.

Desch and Downes describe three cases of alleged manipulation of beliefs about stakes. The first is the Iraq War, though we will not know until the complete disclosure of classified materials, some decades from now, whether the Bush leadership secretly believed that Saddam Hussein did not have weapons of mass destruction or terrorism links, or instead acted in good faith but got it wrong, as Bush claimed in his 2010 memoir. Desch quotes Bush as boasting to the Italian prime minister that the U.S. government will “lead” rather than “follow” public opinion (p.27), but Bush likely made this argument because he needed to bring the Italians on board, not because he believed it. Note that Bush was willing to make the opposite argument, that he was constrained by public opinion, when circumstances required it. He told the leaders of Britain, Spain, and Portugal in March 2003 that he opposed delaying the start of war because, “Public opinion won’t get better and it will get worse in some countries like America.”[383] The second is Lyndon Johnson’s misrepresentation of the 1964 Gulf of Tonkin Incident, probably the best case for the manipulation of consent argument. But even here the misrepresentation was limited, as though Johnson misrepresented the degree of North Vietnamese belligerence in the Incident itself, he was right about North Vietnam’s core goal of destabilizing South Vietnam.

The third case is America’s entry into World War II. Downes and Desch refer to John Schuessler’s argument that Franklin Roosevelt manufactured public support for American entry into World War II, by secretly embargoing oil exports to Japan to provoke a Japanese attack, and secretly provoking Germany in the Atlantic to manufacture a naval incident. I have replied to Schuessler’s argument elsewhere, demonstrating that the facts do not support this claim.[384] In developing American naval policy in the Atlantic, Roosevelt was careful not to get ahead of public opinion. In April 1941, Roosevelt initially allowed the American Navy to escort shipping convoys in the Atlantic but then backed off from this decision, primarily because, according to Henry Morgenthau, “public opinion was not yet ready for the United States to convoy ships.”[385] When a German submarine sank the Robin Moore in June, Roosevelt rejected suggestions from within his administration that he use the policy to justify a more aggressive American naval stance, because, “No doubt Mr. Roosevelt’s circumspection derived in part from realization that the country was not yet willing to approve a more positive policy in the Atlantic.”[386] When the USS Greer was attacked in early September, Roosevelt did shape the presentation of the facts of the incident to amplify the appearance of Germany’s responsibility. However, when the complete account came out in the following weeks, they did not dent public or congressional support for the “shoot-on-sight” policy Roosevelt announced after the incident.[387] That is, consent was not manipulated, because support did not change even when Roosevelt’s ruse was exposed. The allegation that Roosevelt manipulated consent regarding war with Japan is even more dubious, given that the public perceived the late July economic sanctions against Japan to be a total embargo, and that the public supported a hard line on Japan even if it risked war.[388] In sum, as Bear Braumoeller recently summarized, “American involvement was neither secret nor unpopular.”[389]

QUANTITATIVE EVIDENCE

My intention in “A Closer Look” was not to focus on the quantitative empirical dimension of this debate. However, Downes reasserts his 2009 claim that the quantitative evidence favoring selection effects is weak. Downes pulls out parts of the 2008 H. E. Goemans’ article to critique selection effects theory, but in the conclusion of his sophisticated study Goemans states that his analysis “yields results that resolve some questions and now offers empirical support for some of these theories (Bueno de Mesquita et al. 1999, 2003; Reiter and Stam 2002).”[390] Goemans’ earlier quantitative work on the war behavior of mixed regimes also supports selection effects theory. His 2000 findings that democracies fight shorter wars and mixed regimes fight longer wars “would fit the findings of Reiter and Stam (1998) that mixed regimes appear quite risk acceptant and lose more often than not when they start wars.”[391] Much of the other quantitative work Downes cites is not relevant to the selection effects claim, including the Biddle and Long study on soldier effectiveness. The empirical scholarship casting doubts on the audience costs hypothesis does not disprove selection effects theory, as selection effects theory does not predict that democracies make more effective threats. Indeed, we observe in our book that democracies are often perceived as having low resolve exactly because of the institutional constraints described in selection effects theory.

Downes recapitulates our 2009 exchange in International Security. Due to space constraints, I can make only a few brief observations here. First, Downes’ critique of our work aside, other scholars have found that democracies are especially likely to win the wars they initiate.[392] There are additional areas of empirical support for selection effects theory. We find in our book that democracies fight shorter wars and that as wars endure democracies tend to seek draws. Goemans finds that democracies fight shorter wars, and Branislav Slantchev finds that democratic initiators fight shorter wars. Benjamin Valentino and his colleagues find that when democracies fight wars they suffer fewer friendly casualties. Two studies find that democracies are more likely to win the crises they initiate. We and another study find that democracies become increasingly likely to initiate conflict as their likelihood of winning increases.[393]

Second, Downes’ “curvilinear” critique is misplaced. In our book, we lay out two variants of our theory, one predicting a linear interactive effect between democracy, initiation, and victory, and one predicting a curvilinear effect. In the empirical tests we found that democratic initiators are more likely to win, whether you use a linear or curvilinear functional form (as noted, Goemans’ work on war duration supports the curvilinear hypothesis). In our 2009 reply to Downes, we did not concede his critiques of our research design and data (as he implies), but we did point out that even if one accepts nearly all of his alterations to our research design and data, then democratic initiation is still significant using the curvilinear form. Downes makes a very specific critique that statistical tests speak against our use of a curvilinear (fractional polynomial) functional form. I disagree with his application of these tests. Stam and I employ a curvilinear measurement of regime type in our models because that is the functional form that variant of our theory suggests. The sources cited by Downes are by statisticians working through whether or not, when you are theoretically indifferent, you should use a curvilinear functional form. His critique is therefore not applicable since, if we did not use a curvilinear form, it would not have tested our theory. We use that functional form, which shows our results are significant, because it is consistent with our theory, a decision supported by the statisticians Downes cites. These statisticians describe using the test to explain variance, not (necessarily) to test theory. Referring to the section cited by Downes, the statisticians comment that in that section they “discuss how to derive an FP [fractional polynomial] model which is intended to fit the data ‘best.’ The criteria used to judge what is ‘best’ depends on the aims of the study.”[394]

SECRECY AND THE MARKETPLACE OF IDEAS

The issues of secrecy and the marketplace of ideas are linked. Selection effects proposes that democracies are better at choosing winnable wars because a thriving marketplace of ideas helps leaders assess accurately their chances of winning. However, democratic leaders occasionally engage in secret foreign policy initiatives, such as covert actions against foreign governments, to circumvent public opinion constraints (knowing that failed covert actions pose few domestic political costs, because there is plausible deniability of involvement, as well as fewer prospective financial and human costs). Such actions also circumvent the marketplace of ideas, and as a result have a lower rate of success.

Downes critiques this outlook on secrecy and the marketplace of ideas. He points to my observation that the effects of secrecy on war and foreign policy outcomes are complex, summarizing my view as that sometimes secrecy makes success more likely and sometimes less likely, drawing the conclusion that the effects of secrecy on policy outcomes are indeterminate. The connection between secrecy and policy outcomes has become even more salient in the wake of the Wikileaks disclosures, and deserves more study. Let me offer a brief observation about when secrecy is likely to be harmful and when it is likely to be helpful. In general, more discussion and less secrecy will improve the chances that a decision-maker will be better informed about the chances that its side will win. However, there are certain conditions in which less secrecy would directly decrease a country’s chance of winning by revealing information and undermining its military power. Japan’s 1941 attack on Hawaii would have been less successful if the United States had known about its possession of torpedoes that could operate in the shallow waters of Pearl Harbor. Israel’s 1967 attack would have been less successful if Egypt had known about Israel’s operational plan for destroying the bulk of Egypt’s air force on the ground in the first hours of war.

On balance democracies often (though not always) enjoy the best of both worlds, experiencing the benefits of open debate but being able to keep secret the particulars of military operations. Examples of democracies maintaining military operational secrecy and enjoying success include the 1944 D-Day landing, the 1950 Inchon landing, and the 1991 plan not to land Marines in Kuwait but rather to launch a left hook around Kuwait. We have some confidence, then, that democracies are in general likely to enjoy the benefits of secrecy when it is needed as well as the benefits of open debate. But this area needs more study.

Downes claims that the willingness of democracies to engage in covert action knowing that such actions are likely to fail and be exposed undermines selection effects theory. The one citation he provides on this point (to his own 2010 Security Studies article) refers to a single case, Chile in 1970. However, the United States did not support a coup attempt knowing that such an effort had a very low chance of succeeding and a high chance of being exposed. On October 15, 1970, having (accurately) concluded that the chances of a coup launched by Roberto Viaux alone had a low chance of success, U.S. national security advisers Henry Kissinger and Alexander Haig decided “that the [Central Intelligence] Agency must get a message to Viaux warning him against any precipitate action.”[395] Downes implies that after this meeting U.S. policy switched towards actively supporting a coup attempt despite persisting pessimism, but the historical record does not support this view. A more accurate picture is that the Chilean plotters decided on their own on October 16 to launch a coup, and C.I.A. officers in Chile, also pessimistic about the chances of success, attempted to convince the plotters not to undertake their scheme.[396] Specifically, on October 19, the C.I.A. station in Chile “attempted to dissuade the Viaux group from undertaking a coup,” proposing that “any coup would be unsuccessful.”[397] In sum, the Chile 1970 case fits with selection effects theory, as the United States tried to prevent the implementation of a foreign policy initiative that it perceived had a low probability of success.

Downes suggests that the marketplace of ideas remains a largely untested idea. Though the role policy debate in and outside of government plays in decisions for war deserves more study, there is supporting evidence that democracies avoid misguided wars because of better marketplaces of ideas. Jack Snyder studied several great powers with varying regime types, finding that non-democracies like Wilhelmine Germany and Imperial Japan launched foolish wars that democracies like the United States and Britain avoided, in part because non-democratic political institutions severely biased policy debates about decisions for war.[398] There are other examples in which active policy discussions helped improve policy-making in democracies. As mentioned, in 1941 Roosevelt made one attempt to shape public opinion, following the Greer incident. However, the gaps in Roosevelt’s account were publicly exposed within weeks through military cooperation with a congressional investigation. Columnist Arthur Krock commented at the time that “the value of a legislative check on the executive was demonstrated” by the episode.[399] Or consider Downes’ own example in the Vietnam War, when in June 1965 the media exposed the White House’s attempt to downplay the State Department’s announcement of major military escalation in Vietnam.

Though this last instance, described by Downes, is an example of the marketplace of ideas prediction that the media can undermine a government attempt to skew policy debate, Downes argues that the marketplace of ideas failed in Vietnam. I want to make three points on this issue. First, the failure of the marketplace of ideas is exaggerated. The purpose of the marketplace of ideas is to expose the leadership to a range of different opinions and information. And in 1964 and 1965, Johnson was exposed to a range of different recommendations and data. For example, George Ball and Hubert Humphrey argued for withdrawal, and on the other side, members of the the Joint Chiefs of Staff (JCS), in face to face meetings with Johnson, argued for more aggressive military escalation. Downes recognizes that a variety of arguments were made and policy options were presented, but he claims that the marketplace failed because the debate was “narrow.” If one can argue that a debate was skewed even when there is widespread discussion in and out of government and a variety of views are presented and discussed, then we must ask the methodological question of how do we know if a debate is narrow or robust. There may be falsifiability concerns, as we are always after the fact tempted to declare a debate to be narrow if we know the policy did not accomplish its goals.

Second, the quality of the debate aside, the Johnson administration did get two assessments right. It correctly concluded that the likelihood of victory was not high (this was Downes’ central point in his 2009 article). It also correctly predicted that the escalation policies adopted in 1965 did not present an excessive risk of provoking Soviet or Chinese involvement. Some excoriate Johnson for ignoring the JCS proposals for heavier escalation, but Johnson rejected those suggestions, fearing that doing so risked dragging in the Soviet Union or China. And though we do not know if more extensive American military involvement would have brought in the Soviet Union or China, we do know that Johnson’s more moderate escalation accomplished the important goal of keeping them out.

Third, American decision-making in Vietnam is a single case, and must be viewed comparatively. If one accepts the entirety of Downes’ position on Vietnam, which I do not, this constitutes a single case mispredicted by a probabilistic theory, a data point which by itself is insufficient to reject the theory, and which must be compared with other cases. As bad as the Vietnam War escalation debates may seem in hindsight, they do not seem nearly as misguided when contrasted with the warped policy-making debates that often occur in authoritarian states. Iraq’s disastrous invasion of Iran in 1980 was informed by the delusion of its politicized intelligence agencies that the Iranian military would surrender once war began.[400] Hitler approved the calamitous invasion of the Soviet Union under the false expectation that the Communist regime would quickly collapse.[401] As noted, Wilhelmine Germany and Imperial Japan each made doomed attacks because of biased estimates of the likelihood of victory. The Germans embraced a strategy (the Schlieffen Plan) that faced insurmountable logistical hurdles, and the Japanese were (mis)guided by cultural blinders that convinced them that Americans had no stomach for war. Conversely, democracies often attack and win because they develop reasonably accurate plans for victory. Think of the U.S. attack on Spain in 1898, the Israeli attacks on its Arab neighbors in 1956 and 1967, the Indian attack on Pakistan in 1971, and the U.S. attack on Iraq in 1991, among others.

A more direct comparison to the Vietnam War might be the Soviet invasion of Afghanistan. In 1979, Moscow underestimated the ease with which a Communist government could rule Afghanistan, and then strangely came to believe that the Communist Afghan leader, Hafizullah Amin, was close to switching sides to the West. To secure support for an invasion, Yuri Andropov exaggerated to Leonid Brezhnev the ease of an invasion, downplaying the size of the needed force. This was standard operating procedure in the Soviet Union; Andropov similarly sent the Politburo exaggerated reports in 1968 to boost support for the invasion of Czechoslovakia. The final Afghanistan decision was made by five people in the Politburo, kept so secret that the hard copy approval order made no reference to Afghanistan, and was handwritten to prevent possible leaks through the typing pool. A report submitted to the Politburo a few days after the invasion began “was so disingenuous that it effectively amounted to an active measure designed to mislead the rest of the Soviet leadership about the harsh reality of the Afghan situation. Probably composed chiefly for Brezhnev’s benefit, the report maintained the fiction that the assassination of Amin had been chiefly the work of the Afghans themselves rather than K.G.B. special forces.”[402] This dynamic represents a marketplace of ideas failure of a different order than in Vietnam. At worst, Johnson paid insufficient attention to parts of the bureaucracy, and concealed information from the public. In authoritarian systems, in contrast, the bureaucracy provides systematically biased estimates of victory to the leadership, introducing a far greater risk of a leadership unknowingly stumbling into a foreign policy venture doomed to failure. As Winston Churchill remarked, “democracy is the worst form of government except all the others that have been tried.”

DEMOCRACY AND FOREIGN POLICY

A final thought on policy. Though the three of us disagree about many theoretical and empirical particulars, we end up with (I think) fundamentally similar policy recommendations: more democracy can mean better foreign policy. For Desch and Downes, some of the great foreign policy disasters committed by democracies (Poland’s invasion of the Soviet Union, the Israeli attack on Arab states in 1967 and Lebanon in 1982, the Vietnam War, and the Iraq War) are attributable to poor prewar decision-making, in turn caused in part by anti-democratic patterns such as government manipulation of public opinion and government distortion of prewar policy debates. Their critique, as I take it, is not that a properly functioning marketplace of ideas does not improve policy outcomes, but rather that democratic institutions do not prevent the manipulation of public opinion or help ensure the existence and operation of a functioning marketplace of ideas. The view that democracy aids foreign policy is not a trivial position. Older realists such as George Kennan and Walter Lippmann took a different view, that foreign policy ought to be left to the experts, excluding Congressional representatives, the public, and the media. Though the point can be pushed too far, as demonstrated by the recent Wikileaks disclosures, we should recognize that open debate and democracy more generally can improve foreign policy effectiveness rather than undermine it. Indeed, for those of us in the academy who care about policy outcomes, this is the core assumption behind what we do: open and vigorous debate about foreign policy has at least the potential of guiding our elected leaders towards making better decisions.

© Copyright 2011-2015 The Authors.
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 United States License.

 

Notes

[1] Dan Reiter, How Wars End (Princeton, N.J: Princeton University Press, 2009).

[2] Michael C. Desch, Power and Military Effectiveness: The Fallacy of Democratic Triumphalism (Baltimore: Johns Hopkins University Press, 2008). Michael C. Desch, Power and Military Effectiveness: The Fallacy of Democratic Triumphalism (Baltimore: Johns Hopkins University Press, 2008), and Alexander B. Downes, “How Smart and Tough are Democracies? Reassessing Theories of Democratic Victory in War,” International Security 33:4 (Spring 2009): 9-51.

[3] Michael C. Desch, Power and Military Effectiveness: The Fallacy of Democratic Triumphalism (Baltimore: Johns Hopkins University Press, 2008), and Alexander B. Downes, “How Smart and Tough are Democracies? Reassessing Theories of Democratic Victory in War,” International Security 33 (Spring 2009): 9-51.  Selection effects theory critics also discuss the 2003 Iraq War.  Risa Brooks, Shaping Strategy: The Civil-Military Politics of Strategic Assessment (Princeton: Princeton University Press, 2008); Chaim Kaufmann, “Threat Inflation and the Failure of the Marketplace of Ideas: The Selling of the Iraq War,” International Security 29 (Summer 2004), 5-48.   Space precludes examination of this case, and the public historical record is as yet incomplete.  Gil Merom examined why democracies lose small wars, looking at Vietnam, Lebanon, and the French war in Algeria, but did not test selection effects theory.  How Democracies Lose Small Wars: State, Society, and the Failure of France in Algeria, Israel in Lebanon, and the United States in Vietnam (Cambridge: Cambridge University Press, 2003).  John M. Schuessler used Franklin Roosevelt’s actions in 1941 as a critique of selection effects theory.  “The Deception Dividend: FDR’s Undeclared War,” International Security 34 (Spring 2010), 133-165.  For a rebuttal to Schuessler’s argument, see Dan Reiter, “Correspondence,” International Security (forthcoming).  For reasons of space, this paper is also unable to examine case study critiques of other explanations of democratic victory, including that democracies win because of economic advantages, that democracies enjoy larger and more effective alliances, and that democratic soldiers fight with higher battlefield effectiveness.  See Lake, “Powerful Pacifists”; Reiter and Stam, Democracies at War; Choi, “Democratic Synergy” and “The Power of Democratic Cooperation”; Bruce Bueno de Mesquita, Alastair Smith, Randolph M. Siverson, and James D. Morrow,  The Logic of Political Survival (Cambridge: MIT Press, 2003); Benjamin E. Goldsmith, “Defense Effort and Institutional Theories of Democratic Peace and Victory: Why Try Harder?” Security Studies 16 (April-June 2007): 189-222.

[4] Bruce Russett and John Oneal, Triangulating Peace: Democracy, Interdependence, and International Organizations (New York: Norton, 2000).

[5] Bueno de Mesquita et al, Logic of Political Survival; Darren Filson and Suzanne Werner, “Bargaining and Fighting: The Impact of Regime Type on War Onset, Duration, and Outcomes,” American Journal of Political Science 48 (April 2004): 296-313.

[6] Reiter and Stam, Democracies at War.  See also Dan Reiter and Allan C. Stam, “Democracy and War Outcomes: Extending the Debate,” presented at the annual meeting of the International Studies Association, Chicago, IL, March 2, 2007; David A. Lake, “Powerful Pacifists: Democratic States and War,” American Political Science Review 86 (March 1992): 24-37; David H. Clark and William Reed, “A Unified Model of War Onset and Outcome,” Journal of Politics 65 (February 2003): 69-91.

[7] D. Scott Bennett and Allan C. Stam III, “The Declining Advantages of Democracy: A Combined Model of War Outcomes and Duration,” Journal of Conflict Resolution 42 (June 1998): 344-366; Reiter and Stam, Democracies at War, chapter 7.

[8] Branislav Slantchev, “How Initiators End Their Wars: The Duration of Warfare and the Terms of Peace,” American Journal of Political Science 48 (October 2004): 813-29.

[9] Jean-Sebastien Rioux, “A Crisis-Based Evaluation of the Democratic Peace Proposition,” Canadian Journal of Political Science 31(June 1998): 263-283; Christopher F. Gelpi and Michael Griesdorf, “Winners or Losers?  Democracies in International Crisis, 1918-94,” American Political Science Review 95 (September 2001): 633-647.

[10] Bruce Bueno de Mesquita, Michael T. Koch, and Randolph M. Siverson, “Testing Competing Institutional Explanations of the Democratic Peace: The Case of Dispute Duration,” Conflict Management and Peace Science 21 (Winter 2004), 255-267.

[11] Reiter and Stam, Democracies at War, chapter 2; Bruce Bueno de Mesquita, James D. Morrow, Randolph M. Siverson, and Alastair Smith, “Testing Novel Implications of the Selectorate Theory of War,” World Politics 56 (April 2004): 363-88.

[12] John E. Mueller, War, Presidents, and Public Opinion (Lanham, MD: University Press of America, 1985); Gartner, Scott Sigmund and Gary Segura, “War, Casualties, and Public Opinion,” Journal of Conflict Resolution 42 (June 1998): 278-300; Christopher Gelpi, Peter D. Feaver, and Jason Reifler, Paying the Human Costs of War: American Public Opinion and Casualties in Military Conflicts (Princeton: Princeton University Press, 2009); Scott Sigmund Gartner, “The Multiple Effects of Casualties on Public Support for War: An Experimental Approach,” American Political Science Review 102 (February 2008): 95-106.

[13] “Which Way Out? The Manner and Consequences of Losing Office,” Journal of Conflict Resolution 52 (December 2008): 790.

[14] Desch’s first published critique was “Democracy and Victory: Why Regime Type Hardly Matters,” International Security, 27 (Fall 2002), pp. 5-47, and was rebutted by Dan Reiter and Allan C. Stam, “Understanding Victory: Why Political Institutions Matter,” International Security, 28 (Summer 2003), 168-179.  Downes’ 2009 critique was rebutted by Dan Reiter and Allan C. Stam, “Another Skirmish in the Battle over Democracies and War,” International Security, 34 (Fall 2009), 194-204.

[15] In the remainder of this paper, references to page numbers in the text and notes are to page numbers in Desch’s book, Power and Military Effectiveness.

[16] Downes, “How Smart,” 31.

[17] See Dominic D. P Johnson and Dominic Tierney, Failing to Win: Perceptions of Victory and Defeat in International Politics (Cambridge: Harvard University Press, 2006).

[18] See Correlates of War (COW); Reiter and Stam, Democracies at War; Slantchev, “How Initiators End Their Wars.”

[19] Reiter and Stam, Democracies at War, chapter 6.

[20]Adam Berinsky, In Time of War: Understanding American Public Opinion From World War II to Iraq (Chicago: University of Chicago Press, 2009).

[21] On covert action and the democratic peace, see Alexander B. Downes and Mary Lauren Lilly, “Overt Peace, Covert War?: Covert Action and the Democratic Peace,” Security Studies 19 (2010), 266-306.

[22] Reiter and Stam, Democracies at War, chapter 6.

[23] George Bush and Brent Scowcroft, A World Transformed (New York: Vintage Books, 1998), 446.

[24] Reiter and Stam, Democracies at War,  esp. 159-162.

[25] Reiter and Stam, “Understanding Victory.”

[26] Stam, Win, Lose, or Draw; Reiter and Stam, Democracies at War; Slantchev, “How Initiators End Their Wars”; M. B. Biskupski, “Pilsudski, Poland, and the War With Russia, 1919-1920,” Polish Heritage 35 (Winter 1984): 5; Micheal Clodfelter, Warfare and Armed Conflicts: A Statistical Reference to Casualty and Other Figures, 1500-2000 (Jefferson, NC: McFarland, 2002), 387.

[27] Desch (74) includes a quote from Leon Trotsky, but that quote does not appear on the page or in the source Desch cites, Viscount d’Abernon, The Eigthteenth Decisive Battle of the World: Warsaw, 1920 (London: Hodder and Stoughton, 1931, esp. 14.

[28] Richard Debo, Survival and Consolidation: The Foreign Policy of Soviet Russia, 1918-1921 (Montreal: McGill-Queen’s University Press, 1992), 57.

[29] Norman Davies, White Eagle, Red Star: The Polish-Soviet War, 1919-1920 (London: Orbis, 1983), 27.

[30] Andrzej Garlicki, Jozef Pilsudski, 1867-1935, John Coutouvidis, ed. and trans. (Aldershot, UK: Scolar, 1995), 96.

[31] Adam Zamoyski, The Battle for the Marchlands (New York: East European Monographs, 1981), 7; Davies, White Eagle, 25; Garlicki, Pilsudski, 93.  It is difficult to know exactly what military operations Desch is referring to.  Desch cites Sidney Brooks, “America and Poland 1915-1925,” American Relief Administration Bulletin, ser. 2 no. 44 (April 1925), 25, who says merely, “The repulse of Bolshevik troops by the Poles during the months of March and April 1919 opened up a large territory needing immediate relief.”

[32] Clodfelter, Warfare and Armed Conflicts, 387.  See also Orlando Figues, A People’s Tragedy: The Russian Revolution, 1891-1924 (London: Jonathan Cape, 1996), 697.

[33] White Eagle, 396.  See also Clodfelter, Warfare, 387.

[34] Debo, Survival, 58; Zamoyski, Marchlands, 8.

[35] White Eagle, 26-27.

[36] Zamoyski, Marchlands, 7.

[37] Clodfelter, Warfare and Armed Conflicts, 387.

[38] POLITY (see http://www.systemicpeace.org/polity/polity4.htm <downloaded June 9, 2010>; Michael W. Doyle, “Liberalism and World Politics,” American Political Science Review 80 (December 1986): 1151-1169.

[39] Richard M. Watt, Bitter Glory: Poland and Its Fate 1918 to 1939 (New York: Simon and Schuster, 1979), 87.

[40] Andrzej Ajnenkiel, “The Establishment of a National Government in Poland, 1918,” in Paul Latawski, ed., The Reconstruction of Poland, 1914-1923 (London: MacMillanm 1992), 141.

[41] Watt, Bitter Glory, 82.

[42] Quoted in Watt, Bitter Glory, 83.

[43] Roszkowski, 164.

[44] Watt, Bitter Glory, 87-8; Jedrzejewicz, Pilsudski, 81; Garlicki, Pilsudski, 91.

[45] Piotr S. Wandycz, Soviet-Polish Relations, 1917-1921 (Cambridge, MA: Harvard University Press, 1969), 172.

[46] Wandycz, Soviet-Polish Relations, 193.

[47] Stanislaw Grabski,  The Polish-Soviet Frontier (New York: Polish Information Center, 1944), 21.

[48] Quoted in Titus Komarnicki, Rebirth of the Polish Republic: A Study in the Diplomatic History of Europe, 1914-1920 (Melbourne: William Heinemann Ltd, 1957), 573-574.  Desch provides (73) a skeptical quote from a Polish Foreign Ministry Official about the goals of the war, though the quote is from 1919, long before specific consideration of the April 1920 operation.

[49] Jedrezejewicz, 103.

[50] Watt, Bitter Glory, 120; Davies, White Eagle, 113; Robert Szymczak, “Bolshevik Wave Breaks at Warsaw,” Military History 11 (February 1995): 58.

[51] Jedrzejewicz, Pilsudski, 112-113.

[52] Garlicki, Pilsudski, 101.

[53] Zygmunt J. Gasiorowski, “Joseph Pilsudski in the Light of British Reports,” Slavonic and East European Review 50 (October 1972): 558.  Brackets in Gasiorowski.

[54] Davies, White Eagle, 99-100.

[55] Davies, White Eagle, 101-115.

[56] Figues, People’s Tragedy, 699.

[57] Eg, Grabski, Polish-Soviet Frontier.

[58] Piotr Wandycz, Polish Diplomacy 1914-1945: Aims and Achievements (London: Orbis Books, 1988), 15.

[59] Italics in original.  Quoted in Serge Michiel Shewchuk, “The Russo-Polish War of 1920,” Ph.D. thesis, University of Maryland, 1966, 127; see also 133.

[60] Michael Palij, The Ukrainian-Polish Defensive Alliance, 1919-1921 (Edmonton: Canadian Institute of Ukrainian Studies Press, 1995), 90.

[61] Davies, White Eagle, 99; S. Kutrzeba, “The Struggle for the Frontiers, 1919-1923,” in W. F. Reddaway, J. H. Penson, O. Halecki, and R. Dyboski, eds., The Cambridge History of Poland From Augustus II to Pilsudski (1697-1935) (Cambridge: At The University Press, 1941), 525.

[62] Watt, Bitter Glory, 107-108; Davies, White Eagle, 96-99; Wandycz, Soviet-Polish Relations, 156-7.

[63] Wandycz, Soviet-Polish Relations, 15.

[64] Dan Reiter, How Wars End (Princeton: Princeton University Press, 2009).

[65] Desch coded the Egypt-initiated 1969-1970 War of Attrition a draw, though others, including Stam and I, code it as an Israeli victory because Israel better accomplished its operational goals than did Egypt.  Desch’s only discussion of this coding comes in his 2002 (14) International Security article, in which he relates the following Ezer Weizman quote from Martin L. van Creveld’s book, The Sword and the Olive: A Critical History of the Israeli Defense Force (New York: Public Affairs, 1998), 215: “It is no more than foolishness to claim that we won the War of Attrition.  On the contrary, for all their casualties it was the Egyptians who got the best of it.”  However, the context is that Weizman, an Israeli military commander during the War of Attrition, was not making judgment on the lack of Israeli operational success.  Rather, he was sorely disappointed that Israel did not press its operational advantage and seize the west bank of the Suez Canal with ground forces.  Ezer Weizman, On Eagles’ Wings: The Personal Story of the Leading Commander of the Israeli Air Force (New York: Macmillan, 1976), esp. 280-281.  Egypt failed to compel Israel to abandon the Sinai or its forward military positions.  Also, Egypt suffered far, far greater losses than did Israel.  Israel lost 260 soldiers killed in action, and suffered an additional 687 wounded.  Egyptian losses are not precisely known, but reached into the thousands, as the death rate probably ranged from 100 per week to 300 per day, across the 18 or so months of war.  Casualties aside, “Egyptian towns along the canal were almost totally destroyed and some 750,000 residents were evacuated.  Important industrial plants were wrecked and with them the refineries and oil port of Suez.”  Ze’ev Schiff, A History of the Israeli Army: 1874 to the Present (New York: MacMillan Publishing, 1985), 183.  In short, the War of Attrition is consistent with the selection effect theory: an autocracy initiated a war and lost.

[66] POLITY and Doyle, “Liberalism.”

[67] Dan Kurzman, Ben-Gurion: Prophet of Fire. New York: Simon and Schuster, 1983), 382-4.

[68] Kurzman, Ben-Gurion, 386-9.

[69] Kurzman, Ben-Gurion, 390-1, Michael Bar-Zohar, Ben-Gurion: A Biography, Peretz Kidron, trans. (New York: Adams Books, 1978), 244; Schiff, History of the Israeli Army, 93.

[70] Jack Levy and Joseph R. Gochal, “Democracy and Preventive War: Israel and the 1956 Sinai Campaign” Security Studies 11 (Winter 2001/2002), 22.

[71] Bar-On, Gates of Gaza, 6-9.

[72] Michael Brecher, Decisions in Israel’s Foreign Policy (New Haven: Yale University Press, 1975), 261.

[73] Bar-On, Gates of Gaza, 6-9.

[74] Gad Barzilai, “War, Democracy, and Internal Conflict: Israel in a Comparative Perspective,” Comparative Politics 31 (April 1999), 323.

[75] Mordechai Bar-On, The Gates of Gaza: Israel’s Road to Suez and Back, 1955-1957 (New York: St. Martin’s Press, 1994), 39-40.

[76] Quoted in Bar-On 1994, Gates of Gaza, 250.

[77] See Desch (2008, n33 on p.209).  Avner Yaniv and Robert J. Lieber “Personal Whim or Strategic Imperative? The Israeli Invasion of Lebanon,” International Security 8 (Fall 1983), 140 say merely and without citations, “In 1956, a similar deception did not generate much criticism because the operation was quick and decisive.”  Thomas (1966, 16) doesn’t make this argument either; on the cited page, he discusses deliberations in 1955, when, as noted, the Israeli cabinet rejected Ben-Gurion’s suggestion that Israel capture the Straits of Tiran.  Brecher, Decisions, 65 discusses Ben-Gurion’s ideas about Germany and the Holocaust on the cited page.

[78] Bar-On, Gates of Gaza, 317-323.

[79] Desch (101) also summarizes Michael Brecher’s 1975 study of Israeli decision-making by claiming that Brecher found that “domestic political considerations played only a ‘marginal’ role in 75 percent of its major foreign policy decisions.”  Desch appears to be referring to Brecher’s hypothesis 19, “Decision-makers of middle power will be marginally influenced in their decisional choice by their own political structures.”  However, Brecher notes that domestic political considerations were critical in the approach to the Six Day War: “The political structure, in the form of a split coalition Cabinet, was important in the decision to delay and to form a National Unity Government.  It became a decisive input into the strategic decision.  The growing demand for a wall-to-wall Coalition, including Rafi and Gahal, from 24 May onward was a focus for the nation at large—over the correct form of response to Nasser’s threat.  The victory of the ‘pro-changers’ and the consequent formation of a National Unity Government on 1 June was a manifestation of the influence of the political structure in the 1967 decision-making process as a whole.”  Brecher, Decisions, 544-546.

[80] Michael B. Oren, Six Days of War: June 1967 and the Making of the Modern Middle East (Oxford: Oxford University Press, 2002), 134.

[81] Oren, Six Days, 87-8.

[82] Tom Segev, 1967: Israel, the War, and the Year That Transformed the Middle East, Jessica Cohen, trans. (New York: Metropolitan Books, 2007), 327.

[83] Oren, Six Days, 135.

[84] Segev, 1967, 328.

[85] Oren, Six Days, 138.

[86] Reiter and Stam, Democracies at War, 35-37.

[87] Roland Popp, “Stumbling Decidedly Into the Six Day War,” Middle East Journal 60 (Spring 2006), 297.

[88] John J. Mearsheimer, Conventional Deterrence (Ithaca: Cornell University Press, 1983), 150-153; Oren, Six Days of War, 156-7.

[89] Segev, 1967, 386-391.

[90] Avi Shlaim, The Iron Wall: Israel and the Arab World (New York: Norton, 2000, 397-9).

[91] Avner Yaniv, Dilemmas of Security: Politics, Strategy, and the Israeli Experience in Lebanon (New York: Oxford, 1987, 109).

[92] Shlaim, Iron Wall, 403-7; Yaniv, Dilemmas of Security, 128.

[93] Yaniv, Dilemmas of Security, 113.

[94] Shlaim, Iron Wall, 409.

[95] Tabitha Petran, Syria (New York: Praeger, 1972), 98.

[96] See Derek Hopwood, Egypt: Politics and Society 1945-1984 (Boston: Allen & Unwin, 1985), 29-33.

[97] The quantitative data sets are mixed as to whether or not the United States initiated the war.  Downes argues that the United States should be coded as the initiator, because the latest evidence indicates that the first interstate violence, airstrikes following the February 1965 Pleiku bombings, was initiated by the United States (the Pleiku bombings themselves were a Vietcong operation, and Downes argues that the bombings were approved by a local VC commander, and not the North Vietnamese commander).  At the time, however, the United States clearly perceived itself as a target, and the attack as being ordered by North Vietnam.  Hence, from the U.S. perspective the decision to escalate was seen more as a reaction to attack rather than a war of choice.  Lyndon Baines Johnson, The Vantage Point: Perspectives of the Presidency 1963-1969 (New York: Holt, Rinehart, and Winston, 1971), 124.

[98] Gallup Survey, [USGALLUP.633POS.Q24].

[99] Harris Survey, [USHARRIS.122164.R2].

[100] Harris Survey, [USHARRIS.65MAR1.R1].  All polling data are from iPoll, available from the Roper Center (http://www.ropercenter.uconn.edu/).  The abbreviated references used here are the same as those used by iPoll.

[101] [USHARRIS.062865.R9], from iPoll.

[102] Eg Harris Survey, [USHARRIS.071965.R2], Harris Survey [USHARRIS.100465.R2I], and Harris Survey [USHARRIS.120665.R4].

[103] Mueller, War; Gartner and Segura, “War.”

[104] The question refers to “Vietnam,” not “South Vietnam” or “North Vietnam.”  [USGALLUP.711.Q004D], from iPoll.

[105] [USGALLUP.715.Q005], from iPoll.

[106] [USGALLUP.719.Q005], from iPoll.

[107] [USHARRIS.65OCT.R1], from iPoll.

[108] [USGALLUP.65-722.R004], from iPoll.

[109] Reiter and Stam, Democracies at War, 148-149.

[110] Yaniv, Dilemmas of Security, 113.

[111] Shlaim, Iron Wall, 405; Yaniv, Dilemmas of Security, 111.

[112] Brian Lai, “Military Mobilization and the Outcome of International Crises,” Ph.D. diss. (Emory University, Atlanta, Georgia, 2001), 95.

[113] Power and Military Effectiveness: The Fallacy of Democratic Triumphalism (Baltimore: Johns Hopkins University Press, 2008).

[114] Two corrections I would like to make:  First, Reiter is correct that “Target Vistula” was drafted in November 1918, not 1919, as I said in the book.  Second, the Trotsky quote was wrongly attributed in 74, fn. 26.  The correct quote should be Leon Trotsky, “On Guard for the World Revolution,” Report read at the joint session of the Voronezh Soviet of Workers’, Peasants’ and Red Army Men’s Deputies, November 18, 1918 at https://www.marxists.org/archive/trotsky/1918/military/ch37.htm#onguard.  I regret any confusion these caused.

[115] Reiter, “A Closer Look at Case Studies on Democracy, Selection Effects, and Victory,” H-Diplo/ISSF Essays, No. 2 (August 2010): 3 makes this explicit but then ignores it in his discussion of the Russo-Polish case.

[116] George F. Kennan, American Diplomacy: 1900-1950 (Chicago: University Chicago Press, 1951 46 and 53 links what he sees as problems in U.S. foreign policy to democracy.  In contrast, Kenneth N. Waltz, Theory of International Politics (Reading: Addison-Wesley, 1979), 60-78 simply argues that regime-type, like other domestic factors, is not that important in international relations.  For Waltz’s view of this debate, see his “Correspondence: Fair fights or Pointless Wars? International Security Vol.  28, No. 3 (Winter 2003/4): 181.

[117] Otto Hintze, “Military Organization and State Organization” in Felix Gilbert, ed., The Historical Essays of Otto Hintze (New York: Oxford University Press, 1975), 201.  For a more nuanced view see Bruce D. Porter, War and the rise of the State: The Military Foundations of Modern Politics (New York: The Free Press, 1994).

[118] Harold D. Lasswell, “Sino-Chinese Crisis: The Garrison State Versus the Civilian State,” China Quarterly Special No. (Fall 1937): 649.  Also see Lasswell, “The Garrison state,” American Journal of Sociology Vol. 46, No. 4 (1941): 455-68.  For a Democratic Triumphalist challenge, see Aaron L. Friedberg, In the Shadow of the Garrison State (Princeton: Princeton University Press, 2000).

[119] The classic statement of how war leads to centralization of power in the executive is Arthur Schlesinger, Jr. The Imperial Presidency (Boston: Houghton-Mifflin, 1973).  Schlesinger is a normative democratic triumphalist inasmuch as he would like to see war more subject to democratic control but as a historian he cannot help but be struck by how often it works against democracy by strengthening the executive at the expense of the legislature.  For a more recent articulation see Garry Wills, Bomb Power: The Modern Presidency and the National Security State (New York: The Penguin Press, 2010).

[120] For a compelling discussion of how security and other political issues differ, see Jacob Viner, “Power and Plenty as Objectives of Foreign Policy in the Seventeenth and Eighteenth Centuries,” World Politics, Vol. 1, no. 1 (October 1948): 1-29 and Charles Lipson, “International Cooperation in Economic and Security Affairs,” World Politics, 37 (October 1984), pp. 1-23.

[121] Reiter, “A Closer Look,” 5.

[122] See the works cited by Reiter, “A Closer Look,” 4-6.

[123] Hein Goemans, “Which Way Out?: The Manner and consequences of Losing Office,” The Journal of Conflict Resolution Vol. 52, No. 6 (December 2008): 784

[124] Sebastian Rosato, “On the Democratic Peace” in in Christopher J. Coyne, ed., The Handbook on the Political Economy of War (New York: Edward Elgar Publishing, 2011), forthcoming.

[125] In addition to the case studies in my book Power and Military Effectiveness: The Fallacy of Democratic Triumphalism (Baltimore: Johns Hopkins University Press, 2008), two other case study tests of central elements of the selection effects argument include Alexander B. Downes, “How Smart and Tough Are Democracies? Reassessing Theories of Democratic Victory in War,” International Security Vol. 33, No. 4 (Spring 2009): 9-51 and John M. Schuessler, “The Deception Dividend: FDR’s Undeclared War,” International Security Vol. 34, No. 4 (Spring 2010): 133-65.  Reiter and Stam responded to Downes’ quantitative analysis in “Another Skirmish in the Battle Over Democracies and War,” International Security Vol. 34, No. 2 (Fall 2009): 194-200 and Reiter critiques Downes’ case study in “A Closer Look,” 25-28.  Reiter also has a response to Schuessler forthcoming in International Security.

[126] Desch, Power and Military Effectivness, 38-51.

[127] Downes, “How Smart and Tough Are Democracies?” 11.

[128] Schuessler, “The Deception Dividend,” 134. Schuessler’s piece sparked a very extensive discussion on H-Diplo after a favorable review of it by historian Marc Trachtenberg.

[129] Komarnicki, The Rebirth of the Polish Republic, 435.

[130] Reiter, 11, fn. 31 wrongly attributes this quote to Davies, White Eagle, 396.  The correct quote is Norman Davies, God’s Playground: A History of Poland [Vol. II] 1795 to the Present (New York: Columbia University Press, 1982) 396.  The important point here is not the mis-citation, but rather the misinterpretation of “initiative.”  Ironically,  Reiter and Stam made the exact opposite mistake in their use of the HERO data-set “initiative” variable, which there referred to who went first, rather than, as they claimed, who acted “more aggressively and better exploited opportunities” in combat.  On this, see Risa Brooks, “Making Military Might: Why Do States Fail and Succeed? A Review Essay” International Security, Vol. 28, No. 2 (Fall 2003): 181-82.

[131] Richard M. Watt, Bitter Glory: Poland and Its Fate, 1918-1939 (New York: Hippocrene Books, 1982), 97 and 168-69; Michael Palijj, The Ukrainian-Polish Defensive Alliance, 1919-1921: An Aspect of the Ukrainian Revolution (Edmonton: Canadian Institute of Ukrainian Studies Press, 1995), 99; Sidney Brooks, “America and Poland: 1915-1925,” American Relief Administration Bulletin, (Unpublished Report, April 1925), 25 and 40; Titus Komarnicki, The Rebirth of the Polish Republic (London: Wm. Heinemann Ltd., 1957), 428-33.

[132] God’s Playground, 395-96 and John P. Posey, “Soviet Propaganda as a Diplomatic Prelude to the Russo-Polish War, 1920,” The Southern Quarterly, vol. 7 (1968): 141.  Also see Michael J. Fibich, “On the Polish-Bolshevik Front in 1919 and 1920,” Field Artillery Journal, Vo.l 13 (1923): 272-73.  On the Russo-Polish War Beginning in February 1919, see Davies, “The Missing Revolutionary War: The Polish Campaigns and the Retreat From revolution in soviet Russia, 1919-21,” Soviet Studies, Vol. 27, No. 2 (April 1975): 181.  Finally, for a very useful timeline of this complex period, see Serge Michiel Shewchuk, “The Russo-Polish War of 1920,” (Unpublished Dissertation, University of Maryland, 1966), 359-60.

[133] Warren Lerner, “Attempting a Revolution From Without: Poland in 1920,” Studies on the Soviet Union, Vol. 11, No. 4 (1971)96.  Also see Mariuz Mazurek, “Great Britain, the United States and the Polish-Soviet Peace Treaty of Riga, 1920-1921,” Niepodlegiosc, vol 21 (no date): 55; John Erickson, The Soviet High Command  (New York: St. Martins Press, 1962), 85; Stanislaw Kutrzeba,”Chapter 22: The Struggle for the Frontiers, 1919-23” in W. F. Reddaway, et al, eds., The Cambridge History of Poland (Cambridge: Cambridge University Press, 1941), 521; M.K. Dziewanowski, “Polish-Soviet War of 1919-1921” in Joseph L. Wieczynski, ed., The Modern Encyclopedia of Russian and Societ History, [Vol. 28] (Gulf Breeze, FL: Academic International Press, 1982), 229; “The red Army’s Only Defeat,” Kosiuszko Foundation Monthly Newsletter (December 1974): 3; M.B. Biskupski, “Piłsudski, Poland, and the War With Russia, 1919-1920,” Polish Heritage, Vol 35 (Winter 1984): 5; Adam Zamoyski, The Battle for the Marchlands (New York: Columbia University Press, 1981), 2.

[134] Norman Davies, “The Genesis of the Polish-Soviet War, 1919-20,” European Studies Review vol. 5, No. 1 (January 1975): 54.  For a similar verdict, see Thomas C. Fiddick, Russia’s Retreat From Poland, 1920: From Permanent Revolution to Peaceful Coexistence (New York: St. Martin’s Press, ?), 24 and 254; John James Flynn, “British Diplomacy and the Polish-Soviet War of 1920,” (Unpublished M.A. Thesis, Corcoran Department of History, University of Virgina, May 1983), 49; Zamoyski, The Battle for the Marchlands, 189; Palijj, The Ukrainian-Polish Defensive Alliance, 170.; and Davies, White Eagle, Red Star: The Polish-Soviet War, 1919-1920 (New York: St. Martins, 1972), 263.

[135] Reiter, “A Closer Look,” 6-7.

[136] Carl von Clausewitz, On War Michael Howard and Peter Paret, eds. and trans. (Princeton, NJ: Princeton University Press, 1989), 87.

[137] Brooks, “America and Poland,” 26.

[138] Quoted in Serge Michiel Shewchuk, “The Russo-Polish War of 1920,” (Unpublished Dissertation, University of Maryland, 1966), 135.

[139] Watt, Bitter Glory, 185.  Also see 183.

[140] Palijj, The Ukrainian-Polish Defensive Alliance, 35.

[141] Watt, Bitter Glory, 103.  Also see Zamoyski, The Battle for the Marchlands, 6.

[142] Watt, Bitter Glory , 86-88.  Also see 186-87

[143] Davies, God’s Playground, 402.

[144] Norman Davies, “The Soviet command and the Battle of Warsaw,” Soviet Studies Vol. 23 91972): 585.

[145] Shewchuk, “The Russo-Polish War of 1920,” 146.

[146] Davies, God’s Playground, 384 and 391-92.

[147] Watt, Bitter Glory, 60.  Also see 213.

[148] Watt, Bitter Glory, 190-91.

[149] Watt, Bitter Glory, 237.

[150] Watt, Bitter Glory, 35.

[151] Watt, Bitter Glory, 85.

[152] Quoted in Zygmunt J. Gasiorowski, “Joseph Piłsudski in Light of British Reports,” Slavonic and East European Review, vol. 50 (1972): 566 and Viscount d’Abernon, The Eighteenth Decisive Battle of the World: Warsaw, 1920 (London: Hodder and Stoughton, 1931), 39.

[153] Watt, Bitter Glory, 217-35.

[154] Davies, God’s Playground, 422.

[155] Watt, Bitter Glory, 81.  Also see 113.

[156] Davies, God’s Playground, 419.

[157] Watt, Bitter Glory,  94 and 99; Anonymous, “Introduction,” to Jozef Piłsudski, Year 1920 and Its Climax Battle of Warsaw During the Polish-soviet War, 1919-1920 (London: Piłsudski Institute, 1972), 30-31.

[158] Watt, Bitter Glory, 99.

[159] Watt, Bitter Glory, 105.

[160] Konstantin Symmons-Symonolewicz, “Polish Political Thought and the Problem of the Eastern Borderlands of Poland (1918-1939),” Polish Review, Vol. 4 (winter-Spring 1959): 68.

[161] Watt, Bitter Glory, 103.

[162] Watt, Bitter Glory, 120 and 125; Palijj, The Ukrainian-Polish Defensive Alliance, 105; .

[163] Watt, Bitter Glory, 125.

[164] Watt, Bitter Glory, 188.

[165] For further discussion see Harry Eckstein, “Case Study and Theory Development in Political Science” in F. Greenstein and N. Polsby, eds, Handbook of Political Science (Reading, MA: Addison-Wesley, 1975), 94-137.

[166] Nadan Safran, Israel: The Embattled Ally (Cambridge: The Belknap Press of Harvard University Press, 1981), 327.

[167] Avner Yaniv and Robert J. Lieber, “Personal Whim or Strategic Imperative?: The Israeli Invasion of Lebanon,” International Security, Vol. 8, No. 2 (Autumn 1983): 117-42.

[168] Safran, Israel, 319;

[169] Tom Segev, The Seventh Million: The Israeli and the Holocaust (New York: Hill and Wang, 1993), 11.  Also see Michael C. Desch, “The Myth of Abandonment: The Use and Abuse of the Holocaust Analogy,” Security Studies Vol. 15, No. 2 (January-March 2006): 106-45.

[170] Major General Moshe Dayan, “Israel’s Border and Security Problems,” Foreign Affairs, Vol. 33 (January 1955): 252-53.

[171] Quote in Moshe Dayan, Moshe Dayan: Story of My Life (New York: William and Morrow, Inc., 1976), 441.

[172] Safran, Israel, 165.  Also see

[173] Saul Friedlander, “Policy Choices Before Israel” in Paul Y. Hammond and Sidney S. Alexander, eds., Political Dynamics in the Middle East (New York: American Elsvier Co., 1972), 116.

[174] Michael Bar-Zohar, Ben-Gurion: The Armed Prophet, Len Ortzen, trans., (Englewood Cliffs, NJ: Prentice Hall, 1967), 28 and 36.

[175] Benny Morris, “The Eel and History: A Reply to Shabtai Teveth,” Tikkun, Vol. 5, No. 1 (January/February 1991): 82.

[176] Quoted in Golda Meir, My Life (New York: G.P. Putnam’s Sons, 1975), 154.

[177] Avi Shlaim, The Iron Wall: Israel and the Arab World (New York: W.W. Norton & Co., 2001), 187.

[178] Ze’ev Schiff, A History of the Israeli Army: 1874 to the Present (New York: MacMillian Publishing Co., 1985), 232

[179] Safran, Israel, 321.

[180] Michael Brecher, Decision in Israel’s Foreign Policy (New Haven: Yale University Press, 1975), 61;

[181] Schiff, A History of the Israeli Army, 233.

[182] Amos Perlmutter, “Israel’s Fourth War, October 1973: Political and Military Misperceptions,” Orbis vol. 19 (summer 1975): 435.

[183] Hugh Thomas, Suez (New York: Harper and Row, 1966), 16.

[184] Shlaim, The Iron Wall, 185.

[185] Thomas, Suez, 16.

[186] Brecher, Decision in Israel’s Foreign Policy, 232-34;

[187] Schiff, A History of the Israeli Army 93; Thomas, Suez, 119.

[188] Brecher, Decision in Israel’s Foreign Policy, 545.

[189] Baylis Thomas, How Israel Was Won: A Concise History of the Arab-Israeli Conflict (Lanham, MD: Lexington Books, 1999), 125.

[190] Safran, Israel, 368.

[191] Benny Morris, Righteous Victims: A History of the Zionist-Arab Conflict, 1881-1999 (New York: Alfred A. Knopf, 1999), 301.  Also see Shlaim, The Iron Wall, 187;

[192] Thomas, How Israel Was Won, 166.  Also see Tom Segev, 1967: Israel, the War, and the Year That Transformed the Middle East (New York: Henry Holt and Co., 2007), 14;

[193] Yitzhak Laor, “Orchestrated Panic,” London Review of Books, (1 November 2007):  25.  Also see Shlaim, The Iron Wall, 239;

[194] Segev, 1967, 278.

[195] Roland Popp, “Stumbling Decidedly Into the Six Day War,” Middle East Journal, Vol. 60, No. 2 (Spring 2006): 297.

[196] Popp, “Stumbling Decidedly Into the Six Day War,” 298.

[197] Martin Gilbert, Israel: A History (New York: William Morrow and Co., 1998), 375.  Also see Howard M. Sacher, A History of Israel: From the Rise of Zionism to Our Time [2nd Ed.] (New York: Knopf, 1996), 660.

[198] Segev, The Seventh Million, 391-92.

[199] Brecher, Decision in Israel’s Foreign Policy, 335.

[200] Shlaim, The Iron Wall, 250.

[201] Segev, 1967, 93.  Later Eshkol himself became concerned about a military coup in the run-up to the war.  See 306-07.

[202] Segev, The Seventh Million, 390.

[203] Segev, 1967, 155.

[204] Schiff, A History of the Israeli Army, 233.

[205] Segev, 1967, 287.

[206] Quoted in Segev, 1967, 390.

[207] Michael B. Oren, Six Days of War: June 1967 and the Making of the Modern Middle East (New York: Oxford University Press, 2002), 229.  Reiter makes much of the fact that Oren, currently Israel’s ambassador to the United States, downplays this comment in his subsequent analysis.  I do not find that convincing inasmuch as Oren is basically trying to tell the story of the Six Day War in the most favorable light for Israel and admitting that the decisions for war were not taken in a very democratic fashion is to say the least embarrassing.  For a devastating review making this case, see Norman Finkelstein, “Abba Eban With Footnotes,” Journal of Palestine Studies, vol. 32, No. 3 (Spring 2003): 74-89.  Oren is hardly an unproblematic source for Reiter in asmuch as he both strongly implies that it was the Arabs who started the war and then argues that it is impossible to pin down historical questions like who started the war with any certainty.  Neither point helps Reiter’s argument.  See Oren, Six Days of War, 1-2.

[208] Brecher, Decision in Israel’s Foreign Policy, 466.

[209] Safran, Israel, 405.

[210] Martin Van Creveld, The Sword and the Olive: A Critical History of the Israeli Defense Force (New York: Public Affairs, 1998), 299.

[211] Schiff, A History of the Israeli Army, 237.  Also see Zee’ev Schiff and Ehud Ya’ari, Israel’s Lebanon War (New Uork: Simon Schuster, 1984), 38-39; Trevor N. Dupuy and Paul Martell, Flawed Victory: The Arab-Israeli Conflict and the 1982 War in Lebanon (Fairfax, VA: HERO, 1986), 142

[212] Reiter confuses the Big Plan/Little Plan with Big Pines/Little Pines: the former were the competing grand strategic concepts while the latter were the actual operational plans.  They are obviously related, but also distinct.  See, for example, Feldman and Rechnitz-Kijner, Deception, Consensus and War, 21-24; and Shlaim, The Iron Wall, 395 and 406.

[213] Feldman and Rechnitz-Kijner, Deception, Consensus and War, 10-19.

[214] Feldman and Rechnitz-Kijner, Deception, Consensus and War, 72.  Also see Dupuy and Paul Martell, Flawed Victory, 96-97; Schiff and Ya’ari, Israel’s Lebanon War, 188;

[215] Feldman and Rechnitz-Kijner, Deception, Consensus and War, 25.

[216] Yair Evron, War and Intervention in Lebanon: The Israeli-Syrian Deterrence Dialogue (Baltimore: Johns Hopkins University Press, 1987), 30.

[217] Feldman and Rechnitz-Kijner, Deception, Consensus and War, 44.

[218] Yaniv and Lieber, “Personal Whim or Strategic Imperative?,” 136-37.

[219] Quoted in Sacher, A History of Israel, 913.  Also see Schiff and Ya’ari, Israel’s Lebanon War, 39.

[220] Shai Feldman and Heda Rechnitz-Kijner, Deception, Consensus and War: Israel in Lebanon (Boulder, Co: Westview, 1984), 5 and 61-63.

[221] Schiff and Ya’ari, Israel’s Lebanon War, 40-41.  Also see Feldman and Rechnitz-Kijner, Deception, Consensus and War, 29; Richard Gabriel, Operation Peace for Galilee: The Israeli-PLO War in Lebanon (New York: Hill and Wang, 1984), 62; Dupuy and Paul Martell, Flawed Victory, 151; and Shlaim, The Iron Wall, 406.

[222] Schiff and Ya’ari, Israel’s Lebanon War, 43.

[223] Yaniv and Lieber, “Personal Whim or Strategic Imperative?,” 140.

[224] Evron, War and Intervention in Lebanon, 161

[225] Dupuy and Martell, Flawed Victory,  

[226] Schiff and Ya’ari, Israel’s Lebanon War, 58.

[227] Schiff and Ya’ari, Israel’s Lebanon War, 94.

[228] Schiff and Ya’ari, Israel’s Lebanon War, 302.

[229] Anthony Cordesman, The Arab-Israeli Military Balance and the Art of Operations (Washington, DC: AEI, 1987), 54-55.  For similar assessments, see Gabriel, Operation Peace for Galilee, 126

[230] Feldman and Rechnitz-Kijner, Deception, Consensus and War, 3.  Also see

[231] This debate is inextricably linked to that about whether common democracy makes war between states highly unlikely.  The literature on the democratic peace is voluminous, but a good overview of it from proponents and critics can be found in Michael E. Brown, Sean M. Lynn-Jones, and Steven E. Miller, eds., Debating the Democratic Peace: An International Security Reader (Cambridge: MIT Press, 1996).

[232] Chaim Kaufmann, “Threat Inflation and the Marketplace of Ideas: The Selling of the Iraq War,” International Security Vol 29, No. 1 (Summer 2004): 36.

[233] Woodward, Plan of Attack, 320-21 and 428-29.  Also see Desch, “The Myth of Abandonment,” 137-38.

[234] Kaufmann, “Threat Inflation and the Marketplace of Ideas,” 7-9.

[235] For evidence that WMD and links to al Qaeda were “being fixed around the policy” of regime change in Iraq apparently decided on other grounds, see the so-called Downing Street Memorandum, quoted in Mark Danner, “The Secret Way of War,” New York Review of Books Vol. 52, No. 10 (June 9, 2005), www.nybooks.com/articles/18034.  This is also the theme of much of the testimony before the Chilcot Panel in Britain.  See Sarah Lyall, “Ex-Official Says Afghan and Iraq Wars Increased Threats to Britain,” The New York Times, July 20, 2010 at http://www.nytimes.com/2010/07/21/world/europe/21london.html?_r=1&scp=3&sq=britain%20iraq%20war&st=cse and John F. Burns, “Diplomat Harsh on Leaders in Testimony for Iraq Inquiry,” New York Times, July 27, 2010 at http://www.nytimes.com/2010/07/28/world/europe/28blix.html?scp=1&sq=britain%20iraq%20war&st=cse.

[236] United States Department of Defense, “Deputy Secretary Wolfowitz Interview with Sam Tannenhaus, Vanity Fair,” news transcript, Friday May 9, 2003,  http://www.defenselink.mil/transcripts/2003/tr20030509-depsecdef0223.html and Bob Woodward, Bush at War (New York: Simon and Schuster, 2002), 282, and 339-40.

[237] Jon Western, “The War Over Iraq: Selling the War to the American Public,” Security Studies vol. 14, No. 1 (January-March 2005) : 102.  Also see Kaufmann, “Threat Inflation and the Marketplace of Ideas,” 30-31.

[238] Quoted in Bob Woodward, Plan of Attack (New York: Simon and Schuster, 2004), 296.   Also see 377 regarding more evidence that the Bush Administration regarded public opinion as a dependent variable that could be manipulated.

The author would like to thank Michael Desch, Matthew Fuhrmann, and John Schuessler for their helpful comments and suggestions. Any errors are the sole responsibility of the author.

[239] Peter D. Feaver, “Correspondence: Brother Can You Spare a Paradigm? (Or Was Anybody Ever a Realist?),” International Security, Vol. 25, No. 1 (Summer 2000), p. 168. See also Michael C. Desch, Power and Military Effectiveness: The Fallacy of Democratic Triumphalism (Baltimore, Md.: Johns Hopkins University Press, 2008).

[240] Dan Reiter and Allan C. Stam, Democracies at War (Princeton, N.J.: Princeton University Press, 2002). See also Dan Reiter and Allan C. Stam, III, “Democracy, War Initiation, and Victory,” American Political Science Review, Vol. 92, No. 1 (June 1998), pp. 377-389; and Reiter and Stam, “Democracy and Battlefield Military Effectiveness,” Journal of Conflict Resolution, Vol. 42, No. 3 (June 1998), pp. 259-277.

[241] For a sampling of these critiques, see Michael C. Desch, “Democracy and Victory: Why Regime Type Hardly Matters,” International Security, Vol. 27, No. 2 (Fall 2002), pp. 5-47; Desch, Power and Military Effectiveness; Risa A. Brooks, “Making Military Might: Why Do States Fail and Succeed? A Review Essay,” International Security, Vol. 28, No. 2 (Fall 2003), pp. 149-191; Stephen Biddle and Stephen Long, “Democracy and Military Effectiveness: A Deeper Look,” Journal of Conflict Resolution, Vol. 48, No. 4 (August 2004), pp. 525-546; Alexander B. Downes, “How Smart and Tough Are Democracies? Reassessing Theories of Democratic Victory in War,” International Security, Vol. 33, No. 4 (Spring 2009), pp. 9-51; and John M. Schuessler, “The Deception Dividend: FDR’s Undeclared War,” International Security, Vol. 34, No. 4 (Spring 2010), pp. 133-165.

[242] See Dan Reiter and Allan C. Stam, “Understanding Victory: Why Political Institutions Matter,” International Security, Vol. 28, No. 1 (Summer 2003), pp. 168-179; Michael C. Desch, “Democracy and Victory: Fair Fights or Food Fights?” International Security, Vol. 28, No. 1 (Summer 2003), pp. 180-194; Dan Reiter and Allan C. Stam, and Alexander B. Downes, “Correspondence: Another Skirmish in the Battle over Democracies and War,” International Security, Vol. 34, No. 2 (Fall 2009), pp. 194-204. An exchange between Reiter and Schuessler on democracy and deception is forthcoming.

[243] Dan Reiter, “A Closer Look at Case Studies on Democracy, Selection Effects, and Victory,” H-Diplo | ISSF Essays, No. 2.

[244] Desch, Power and Military Effectiveness; and Downes, “How Smart and Tough Are Democracies?”

[245] The first four of these cases are examined in Desch, Power and Military Effectiveness, while the last is explored in Downes, “How Smart and Tough Are Democracies?”

[246] See, for example, Risa Brooks, Shaping Strateg: The Civil-Military Politics of Strategic Assessment (Princeton, N.J.: Princeton University Press, 2008).

[247] Reiter and Stam, Democracies at War, p. 6.

[248] Technically, voters compare the payoff they received under the incumbent with what they expect to receive under a prospective challenger. Failing to win a war reduces voters’ payoffs and makes choosing the challenger more attractive. Bruce Bueno de Mesquita, Alastair Smith, Randolph M. Siverson, and James D. Morrow, The Logic of Political Survival (Cambridge, Mass.: MIT Press, 2003), p. 228.

[249] Reiter and Stam, Democracies at War, p. 6.

[250] Reiter, “Closer Look,” p. 7. Reiter does not explain how a leader being forced to start a war he otherwise would not by a bloodthirsty public is consistent with the idea of a selection effect.

[251] Reiter and Stam, Democracies at War, p. 20.

[252] At a broader level, it is unclear if the electoral accountability and marketplace of ideas parts of the theory are individually or jointly sufficient. Is the vibrant public debate and expert advice from a professional and meritocratic military the source of leaders’ beliefs about the winnability of wars, or are those views formed separately?

[253] One way to try to square this circle would be to argue that public consent is only forthcoming for wars that are likely to be relatively easy victories. Reiter does not make this argument, however, and it is not clear the historical record supports it given that democracies have fought many long wars. One would also have to assume that members of the public base their decisions on whether or not to support a war on the same kinds of short-term factors that drive democratic leaders and not on calculations of what is in the national interest, and that leaders are unable to persuade constituents (using truthful or deceptive arguments) that a conflict is winnable. Reiter and Stam argue that publics offer support for wars that they believe are in the national interest rather than those they perceive will be won easily; see Democracies at War, pp. 148-149. I address the point about the persuasive abilities of leaders below.

[254] Reiter and Stam, Democracies at War, pp. 148, 200.

[255] Schuessler, “Deception Dividend.”

[256] Reiter and Stam, Democracies at War, p. 148.

[257] Rally effects do not always occur when a democratic leader puts troops in harm’s way, but scholars have identified several cases (including after U.S. entry into the Korean and Vietnam wars), and the possibility of such effects taints ex post evidence as potentially unreliable. See John E. Mueller, War, Presidents, and Public Opinion (New York: Wiley, 1973); Matthew A. Baum, “The Constituent Foundations of the Rally-Around-the-Flag Phenomenon,” International Studies Quarterly, Vol. 46, No. 2 (June 2002), pp. 263-298; Marc J. Hetherington and Michael Nelson, “Anatomy of a Rally Effect: George W. Bush and the War on Terrorism,” PS: Political Science and Politics, Vol. 36, No. 1 (January 2003), pp. 37-42; and Michael Colaresi, “The Benefit of the Doubt: Testing an Informational Theory of the Rally Effect,” International Organization, Vol. 61, No. 1 (Winter 2007), pp. 99-143.

[258] This is also known as the Condorcet jury theorem, which holds that if each member of a jury is more likely to be right than wrong, increasing the size of the jury increases the probability that it will arrive at the correct decision. See Christian List and Robert E. Goodin, “Epistemic Democracy: Generalizing the Condorcet Jury Theorem,” Journal of Political Philosophy, Vol. 9, No. 3 (September 2001), pp. 277-306.

[259] Reiter and Stam, Democracies at War, pp. 23-24.

[260] Chaim Kaufmann, “Threat Inflation and the Failure of the Marketplace of Ideas: The Selling of the Iraq War,” International Security, Vol. 29, No. 1 (Summer 2004), pp. 5-48; and Schuessler, “Deception Dividend.”

[261] Kaufmann, “Threat Inflation and the Failure of the Marketplace of Ideas”; Jon Western, “The War over Iraq: Selling War to the American Public,” Security Studies, Vol. 14, No. 1 (January 2005), pp. 106-139; and Jane Kellett Cramer, “Militarized Patriotism: Why the U.S. Marketplace of Ideas Failed before the Iraq War,” Security Studies, Vol. 16, No. 3 (July 2007), pp. 489-524. These and other pieces on this question are collected in A. Trevor Thrall and Jane K. Cramer, eds., American Foreign Policy and the Politics of Fear: Threat Inflation since 9/11 (London: Routledge, 2009).

[262] Stephen Biddle and Robert Zirkle, “Technology, Civil-Military Relations, and Warfare in the Developing World,” Journal of Strategic Studies, Vol. 19, No. 2 (June 1996), pp. 171-212. See also James T. Quinlivan, “Coup-proofing: Its Practice and Consequences in the Middle East,” International Security, Vol. 24, No. 2 (Fall 1999), pp. 131-165.

[263] Brooks, Shaping Strategy.

[264] See, for example, Biddle and Zirkle’s discussion of North Vietnam during the Vietnam War, and Brooks’s discussion of Egypt in the Yom Kippur War. See also Jessica L. Weeks, “Leaders, Accountability, and Foreign Policy in Non-Democracies,” Ph.D. dissertation, Stanford University, 2008; and Caitlin Talmadge, “Explaining Military Effectiveness: Political Control and Battlefield Performance,” Ph.D. dissertation, Massachusetts Institute of Technology, 2010.

[265] On France and Britain before World War II, see Elizabeth Kier, Imagining War: French and British Military Doctrine between the Wars (Princeton, N.J.: Princeton University Press, 1997). On the United States in Vietnam, see H. R. McMaster, Dereliction of Duty: Lyndon Johnson, Robert McNamara, the Joint Chiefs of Staff, and the Lies That Led to Vietnam (New York: HarperPerrenial, 1997).

[266] Bruce Russett, Grasping the Democratic Peace: Principles for a Post-Cold War World (Princeton, N.J.: Princeton University Press, 1993), pp. 38-40.

[267] Reiter, “Closer Look,” p. 9. Secrecy in war preparations, however, cannot be a reason that democracies prevail at a higher rate than non-democracies unless one makes the argument that democracies are better than non-democratic states at cloaking their preparations for war in secrecy. I have not encountered this argument in the literature (indeed, democratic peace theorists, as noted above, have made the opposite argument), and Reiter does not make it in his essay. He simply argues that secrecy is not inconsistent with selection effects.

[268] Reiter, “Closer Look,” p. 9.

[269] Reiter, “Closer Look,” p. 8.

[270] Tim Weiner, Legacy of Ashes: The History of the CIA (New York: Doubleday, 2007), p. 180.

[271] Reiter, “Closer Look,” p. 9.

[272] Alexander B. Downes and Mary Lauren Lilley, “Overt Peace, Covert War?: Covert Intervention and the Democratic Peace,” Security Studies, Vol. 19, No. 2 (Summer 2010), pp. 297-299.

[273] Reiter and Stam, “Democracy, War Initiation, and Victory”; and Reiter and Stam, Democracies at War.

[274] Downes, “How Smart and Tough Are Democracies?” pp. 15-31.

[275] Reiter and Stam, and Downes, “Another Skirmish in the Battle over Democracies and War.”

[276] Reiter and Stam, Democracies at War, p. 45, Table 2.2, models 1 and 4.

[277] Reiter and Stam, “Another Skirmish in the Battle over Democracies and War,” p. 199.

[278] Patrick Royston and Douglas G. Altman, “Regression using Fractional Polynomials of Continuous Covariates: Parsimonious Parametric Modeling,” Applied Statistics, Vol. 43, No. 3 (1994), pp. 429-467; Patrick Royston and Douglas G. Altman, “Using Fractional Polynomials to Model Curved Regression Relationships,” Stata Technical Bulletin, Vol. 21 (September 1994), pp. 11-23; Patrick Royston and Willi Sauerbrei, “A New Approach to Modelling Interactions between Treatment and Continuous Covariates in Clinical Trials by Using Fractional Polynomials,” Statistics in Medicine, Vol. 23, No. 16 (July 2004): 2509-2525; and Patrick Royston and Willi Sauerbrei, Multivariable Model-Building: A Pragmatic Approach to Regression Analysis Based on Fractional Polynomials for Modelling Continuous Covariates (Chichester, U.K.: John Wiley, 2008), pp. 82-83.

[279] Reiter and Stam, Democracies at War, p. 45, models 4 and 5.

[280] In addition to my contribution to “Another Skirmish in the Battle over Democracies and War,” see Alexander B. Downes, “Comparing the Effect of Linear versus Curvilinear Specifications of Democracy on War Outcomes,” and “A Note on Curvilinearity, Democracy, and Victory,” both available on my website, http://www.duke.edu/~downes/publications.htm.

[281] Reiter, “Closer Look,” pp. 4-5. See also Reiter and Stam, “Correspondence,” pp. 194-195.

[282] Reiter, “Closer Look,” p. 5. The article in question is H. E. Goemans, “Which Way Out? The Manner and Consequences of Losing Office,” Journal of Conflict Resolution, Vol. 52, No. 6 (December 2008): 771-794.

[283] Goemans, “Which Way Out?” p. 784.

[284] Giacomo Chiozza and H. E. Goemans, “International Conflict and the Tenure of Leaders: Is War Still Ex Post Inefficient?” American Journal of Political Science, Vol. 48, No. 3 (July 2004), pp. 604-619.

[285] Biddle and Long, “Democracy and Military Effectiveness.”

[286] Worse, when variables accounting for different combinations of belligerents’ cultures are included, attacker democracy predicts a significantly worse loss exchange ratio for the attacker. Ibid., pp. 537-538.

[287] Jason Lyall, “Do Democracies Make Inferior Counterinsurgents? Reassessing Democracy’s Impact on War Outcomes and Duration,” International Organization, Vol. 64, No. 1 (Winter 2010), pp. 167-192; Jason Lyall and Isaiah Wilson III, “Rage against the Machines: Explaining Outcomes in Counterinsurgency,” International Organization, Vol. 63, No. 1 (Winter 2009), pp. 67-106; and Gil Merom, How Democracies Lose Small Wars: State, Society, and the Failure of France in Algeria, Israel in Lebanon, and the United States in Vietnam (Cambridge: Cambridge University Press, 2003).

[288] Jonathan D. Caverley, “The Myth of Military Myopia: Democracy, Small Wars, and Vietnam,” International Security, Vol. 34, No. 3 (Winter 2009/10), pp. 119-157. See also Jonathan D. Caverley, “The Political Economy of Blood and Treasure,” unpublished manuscript, Northwestern University, July 2009.

[289] Jessica L. Weeks, “Autocratic Audience Costs: Regime Type and Signaling Resolve,” International Organization, Vol. 62, No. 1 (Winter 2008), pp. 35-64. Also making the case that autocracies can generate audience costs is Jessica Chen Weiss, “Playing with Fire: Nationalist Protest and Authoritarian Diplomacy,” paper presented at the Annual Meeting of the American Political Science Association, Washington, D.C., September 2010.

[290] Douglas M. Gibler and Marc L. Hutchison, “Territorial Issues, Audience Costs, and the Democratic Peace: The Importance of Issue Salience,” paper presented at the Annual Meeting of the American Political Science Association, Washington, D.C., September 2010.

[291] Alexander B. Downes and Todd S. Sechser, “Are You Threatening Me? A Reassessment of the Credibility of Democratic Threats in International Relations,” paper presented at the Annual Meeting of the American Political Science Association, Washington, D.C., September 2010.

[292] Although Reiter does not contest my fundamental claim, he downplays it by stating repeatedly that members of the Johnson administration and the U.S. public thought the chances of victory in Vietnam were “only moderate” rather than poor. Reiter, “Closer Look,” pp. 3, 27, 28. I am satisfied that the large number of documents I cited in my article supports my argument that Johnson and his advisors were more pessimistic than Reiter allows. My view also echoes the conclusion of most historians. See Downes, “How Smart and Tough Are Democracies?” pp. 34-43, and the works cited therein. See also Robert Buzzanco, Masters of War: Military Dissent and Politics in the Vietnam Era (Cambridge: Cambridge University Press, 1996).

[293] Reiter, “Closer Look,” p. 26.

[294] Downes, “How Smart and Tough Are Democracies?” p. 19, n. 31.

[295] See, for example, Jeffrey Record: The Wrong War: Why We Lost in Vietnam (Annapolis, Md.: U.S. Naval Institute Press, 1998); Merom, How Democracies Lose Small Wars; Charles E. Neu, America’s Lost War: Vietnam: 1945-1975 (Arlington Heights, Ill.: Harlan Davidson, 2005); and Gary R. Hess, Vietnam: Explaining America’s Lost War (Malden, Mass.: Blackwell, 2008).

[296] Reiter, “Closer Look,” p. 25, n. 95. My information is from Mark Moyar, Triumph Forsaken: The Vietnam War, 1954-1965 (Cambridge: Cambridge University Press, 2006), p. 486, n. 14.

[297] Quoted in Robert L. Galluci, Neither Peace nor Honor: The Politics of American Military Policy in Viet-Nam (Baltimore, Md.: Johns Hopkins University Press, 1975), p. 46.

[298] Reiter and Stam, Democracies at War, p. 40.

[299] Reiter, “Closer Look,” p. 26.

[300] The policymaking predilections of Lyndon Johnson are well-known: the president preferred unanimity among his advisors, disliked free-wheeling debate, and for those reasons confined important deliberations to a small circle of trusted aides.

[301] Fredrik Logevall, Choosing War: The Lost Chance for Peace and the Escalation of War in Vietnam (Berkeley: University of California Press, 1999), p. 212; George McT. Kahin, Intervention: How America Became Involved in Vietnam (Garden City, N.Y.: Anchor, 1987), pp. 288-289; and William Conrad Gibbons, The U.S. Government and the Vietnam War: Executive and Legislative Roles and Relationships, Part 3: January-July 1965 (Princeton, N.J.: Princeton University Press, 1989), pp. 97-103.

[302] Mansfield’s memo was originally given to President Kennedy in August 1963. See Memorandum, Mansfield to Kennedy, in Foreign Relations of the United States (hereafter FRUS), 1961-1963, Vol. 3, doc. 258. The responses of Johnson’s advisors are in FRUS, 1964-1968, 1: 8.

[303] Quoted in Gibbons, U.S. Government and the Vietnam War, 3: pp. 94, 95. A similar fate befell a lower-ranking official—Paul Kattenburg, a widely-respected Vietnam specialist in the State Department—late in the Kennedy administration for advocating withdrawal from Vietnam. See Logevall, Choosing War, p. 51; and David Kaiser, American Tragedy: Kennedy, Johnson, and the Origins of the Vietnam War (Cambridge, Mass.: Belknap Press, 2000), pp. 244-245.

[304] Logevall, Choosing War, pp. 248-250.

[305] Ibid., p. 249.

[306] My narrative of this incident relies on Kaiser, American Tragedy, pp. 331-336.

[307] Edwin E. Moïse, Tonkin Gulf and the Escalation of the Vietnam War (Chapel Hill: University of North Carolina Press, 1996).

[308] Kaiser, American Tragedy, p. 336.

[309] Quoted in Logevall, Choosing War, p. 201.

[310] This episode fits well with Schuessler’s argument that democratic leaders use deception to shift the blame for war onto the opponent, thereby generating public support for the use of force. Schuessler, “Deception Dividend.”

[311] George C. Herring, America’s Longest War: The United States and Vietnam, 1950-1975, 2d ed. (Philadelphia, Pa.: Temple University Press, 1986), p. 133.

[312] Logevall, Choosing War, pp. 304-305. For similar assessments, see Stanley Karnow, Vietnam: A History (New York: Viking, 1983), pp. 430, 433; Moyar, Triumph Forsaken, pp. 355, 369, 414-416; Brian VanDeMark, Into the Quagmire: Lyndon Johnson and the Escalation of the Vietnam War (New York: Oxford University Press, 1991), pp. 39, 54, 68, 70-71, 110, 130-131, 211; and Gibbons, U.S. Government and the Vietnam War, 3: p. 78.

[313] On this meeting, see Logevall, Choosing War, pp. 314-315.

[314] Quoted in Gibbons, U.S. Government and the Vietnam War, 3: p. 40.

[315] Ibid., p. 86, n. 116.

[316] Robert Mann, A Grand Delusion: America’s Descent into Vietnam (New York: Basic Books, 2001), p. 408.

[317] Quoted in Kaiser, American Tragedy, p. 419.

[318] Ibid., p. 449. Perhaps Johnson was referring to his speech at Johns Hopkins University in early April where he stated he was open to “unconditional discussions” with Hanoi. This speech was an empty gesture because it was designed primarily to quiet the administration’s critics and was not an offer to negotiate. Indeed, there was no change in policy at all. Logevall, Choosing War, pp. 371-372.

[319] VanDeMark, Into the Quagmire, p. 211.

[320] This expression is from Karnow, Vietnam, p. 430.

[321] Gibbons, U.S. Government and the Vietnam War, 3: p. 278. See also VanDeMark, Into the Quagmire, pp. 109-111, 155-156.

[322] Evidence indicates that in late 1963 the military presented an overly optimistic picture of progress in South Vietnam by omitting data on increased Viet Cong attacks from its reports. See Kaiser, American Tragedy, p. 268.

[323] Logevall, Choosing War, p. 216.

[324] Ibid., p. 330.

[325] Downes, “How Smart and Tough Are Democracies?” p. 38.

[326] On this episode, see ibid., pp. 42-43. Quote is from McNamara to Johnson, Memorandum, July 20, 1965, in FRUS, 1964-1968, 3: 67.

[327] Johnson’s predecessor, John F. Kennedy, faced the same problem, first in Laos and then in South Vietnam. In Laos, for example, where the president favored negotiations and neutralization, the JCS on March 20, 1961, recommended inserting up to 60,000 U.S. troops backed by air power and nuclear weapons. Kaiser, American Tragedy, p. 43. Similarly, the chiefs favored sending troops to South Vietnam in October 1961. Logevall, Choosing War, p. 26.

[328] McMaster, Dereliction of Duty, pp. 64-65.

[329] Ibid., p. 72. McMaster notes that sending “draft memorandums to the president before the Joint Chiefs had an opportunity to comment” was common practice for McNamara. Ibid., pp. 74-75.

[330] Ibid., pp. 90-91.

[331] Ibid., p. 111.

[332] Ibid., pp. 115-116, 222.

[333] Ibid., pp. 181, 301-303. See also pp. 208-209.

[334] Ibid., pp. 50, 72-73.

[335] Ibid., p. 160.

[336] This trend had begun under Kennedy, who—in an attempt to dilute the chiefs’ influence after the Bay of Pigs fiasco—appointed retired Gen. Maxwell Taylor as “Military Representative of the President,” essentially his personal advisor on military matters.

[337] McMaster, Dereliction of Duty, p. 83.

[338] Ibid., p. 106.

[339] Ibid., pp. 100-101.

[340] Ibid., p. 102.

[341] Ibid., p. 110.

[342] Buzzanco, Masters of War, p. 194.

[343] McMaster, Dereliction of Duty, pp. 135, 221, 320. Other members of the JCS also occasionally were not fully truthful with Congressmen. Ibid., pp. 309-312.

[344] Ibid., p. 77.

[345] Ibid., p. 280. For other examples, see ibid., p. 301

[346] Ibid., p. 175.

[347] Johnson approved an increase of only one brigade. Ibid., pp. 270-271.

[348] Ibid., pp. 175-176, 317.

[349] Ibid., pp. 82-84.

[350] Ibid., pp. 114-115. Each service tended to view the conflict from its own perspective and proposed solutions in which it would play the leading role. Ibid., pp. 142-144.

[351] Greene also favored inserting Marines into coastal enclaves in South Vietnam. Ibid., pp. 143-44.

[352] On the September 1964 episode, see McMaster, Dereliction of Duty, pp. 148-149; and Buzzanco, Masters of War, p. 171. On the March 1965 dispute, see McMaster, Dereliction of Duty, pp. 249-250.

[353] Buzzanco, Masters of War, pp. 186, 188, 227; and McMaster, Dereliction of Duty, pp. 146-147, 169, 209.

[354] VanDeMark, Into the Quagmire, p. 76; Gibbons, U.S. Government and the Vietnam War, 3: pp. 75-76.

[355] Logevall, Choosing War, pp. 281-282.

[356] Only the first two polls he cites—in September and November 1964—preceded the beginning of the air war. Reiter, “Closer Look,” p. 26. The first of these polls, which found that 45 percent favored continuing to support South Vietnam, 36 percent supported attacking North Vietnam, and 19 wanted to withdraw, was taken shortly after the Gulf of Tonkin episode and reflects the surge in public bellicosity in reaction to that skirmish.

[357] Gallup Poll USAIPO1964-0689, April 24-29, 1964, Q7. Gallup and Harris polls referenced here are available from the Roper Center Public Opinion Archives, http://www.ropercenter.uconn.edu/.

[358] Gallup Poll USAIPO1964-0689, April 24-29, 1964, Q9.

[359] Harris Survey, USHARRIS.033064.R2, March 1964. It also demonstrates the importance of question wording: the Harris question, by cueing the possibility of Chinese intervention, elicited much more opposition to escalation than questions that omitted this possibility.

[360] Members of Congress were no more enthusiastic, exhibiting “low understanding or care about the problem.” Of the handful of legislators who cared about Vietnam, most opposed any escalation there. Logevall, Choosing War, p. 148.

[361] Ibid., p. 282.

[362] Elite opinion was similarly divided, according to another CFR survey: 20 percent favored military disengagement as opposed to 24 percent who wanted to expand the war. Gibbons, U.S. Government and the Vietnam War, 3: pp. 41-42.

[363] Kahin, Intervention, p. 324.

[364] Logevall, Choosing War, p. 360.

[365] Gibbons, U.S. Government and the Vietnam War, 3: p. 74, n. 85.

[366] VanDeMark, Into the Quagmire, p. 120.

[367] As James Reston commented about public views of LBJ’s policies in the New York Times on March 7, “one has the impression that the mood would be about the same if he stepped up the bombing, or stopped it, or followed any other policy except ordering the American Army into battle on the ground.” Quoted in Logevall, Choosing War, p. 361.

[368] Quoted in Gibbons, U.S. Government and the Vietnam War, 3: p. 75. The editorial pages of the nation’s leading newspapers at the beginning of 1965 reflected the public’s uneasiness about escalation, with many staff editorials expressing opposition to escalation. Even after the bombing began, about half of the editorial boards opposed more energetic involvement and favored a negotiated exit. Logevall, Choosing War, pp. 283, 361.

[369] Kaiser, American Tragedy, p. 337.

[370] Herring, America’s Longest War, p. 143.

[371] Moyar, Triumph Forsaken, pp. 413 and 501, n. 59.

[372] Nor does Johnson’s desire to postpone escalation in (or withdrawal from) Vietnam until after the 1964 elections support selection effects. I address this issue in “How Smart and Tough Are Democracies?” p. 37, n. 77.

[373] Reiter, “Closer Look,” p. 27.

[374] See, for example, Jasen Castillo, “The Will to Fight: National Cohesion and Military Staying Power,” book manuscript in progress, Bush School of Government and Public Service, Texas A&M University; and Jason Lyall, “Paths of Ruin: How Identity Shapes Military Effectiveness in Modern War,” book manuscript in progress, Yale University.

[375] Indeed, one can more than imagine it; I have cited many examples of such work throughout this essay.

[376] Dan Reiter and Allan C. Stam, Democracies at War (Princeton: Princeton University Press, 2002), 201.

[377] Reiter and Stam, Democracies at War, 21.

[378] Matthew A. Baum and Phillip B. K. Potter, “The Relationships Between Mass Media, Public Opinion, and Foreign Policy: Toward a Theoretical Synthesis,” Annual Review of Political Science 11 (2008): 39-65.

[379] [USORC.64SEP.R14], from iPoll available from the Roper Center (http://www.ropercenter.uconn.edu/).  The abbreviated references used here are the same as those used by iPoll.

[380] [USGALLUP.706.Q018D], from iPoll.

[381] Reiter and Stam, Democracies at War, 146.

[382] Reiter and Stam, Democracies at War, 148.

[383] Bob Woodward, Plan of Attack (New York: Simon & Schuster, 2004), 357.

[384] Dan Reiter, “Correspondence: FDR, U.S. Entry into World War II, and Selection Effects Theory,” International Security 35 (Fall 2010): 176-181.

[385] Quoted in Robert Dallek, Franklin D. Roosevelt and American Foreign Policy, 1932-1945 (Oxford: Oxford University Press, 1995), 261.

[386] William L. Langer and S. Everett Gleason, The Undeclared War, 1940-1941 (New York: Harper and Brothers, 1953), 520. See also Dallek, Roosevelt, 267-278.

[387] Dallek, Roosevelt, 288.

[388] Reiter, “Correspondence”; Irvine H. Anderson, Jr., The Standard-Vacuum Oil Company and United States East Asian Policy, 1933-1941 (Princeton: Princeton University Press), 176n.

[389] Bear F. Braumoeller, “The Myth of American Isolationism,” Foreign Policy Analysis 6 (October 2010): 366.

[390] H. E. Goemans, “Which Way Out? The Manner and Consequences of Losing Office,” Journal of Conflict Resolution 52 (December 2008), 790.

[391] H. E. Goemans, “Fighting for Survival: The Fate of Leaders and the Duration of War,” Journal of Conflict Resolution 44 (October): 579.

[392] David H. Clark and William Reed, “A Unified Model of War Onset and Outcome,” Journal of Politics 65 (February 2003): 69-91; Branislav L. Slantchev, “How Initiators End Their Wars: The Duration of Warfare and the Terms of Peace,” American Journal of Political Science 48 (October 2004): 813-829.

[393] For these citations, see Dan Reiter and Allan C. Stam, “Correspondence: Another Skirmish in the Battle over Democracies and War,” International Security 34 (Fall 2009): 194-200; Goemans, “Fighting for Survival”; Benjamin A. Valentino, Paul K. Huth, and Sarah E. Croco, “Bear Any Burden? How Democracies Minimize the Costs of War,” Journal of Politics 72 (April 2010): 528-544.

[394] Patrick Royston and Willi Sauerbrei, Multivariable Model-building: A Pragmatic Approach to Regression Analysis Based on Fractional Polynomials for Modeling Continuous Variables (Chichester, UK: Wiley, 2008), 80.

[395] Memorandum of Conversation, Kissinger, Karamessines, Haig, October 15, 1970, available at http://www.gwu.edu/~nsarchiv/NSAEBB/NSAEBB8/ch06-01.htm <downloaded November 15, 2010>. Note the reference to Viaux “alone” in the October 16 CIA Operating Guidance Cable. Available at http://www.gwu.edu/~nsarchiv/NSAEBB/NSAEBB8/ch05-01.htm <downloaded November 15, 2010>.

[396] CIA Cable, October 17, 2010. Available at http://foia.state.gov/documents/Pcia3/000091BC.pdf <downloaded November 15, 2010>.

[397] CIA Cable, October 19, 2010. Available at http://foia.state.gov/documents/Pcia3/00009105.pdf <downloaded November 15, 2010>.

[398] Jack Snyder, Myths of Empire: Domestic Politics and International Ambition (Ithaca: Cornell University Press, 1991).

[399] New York Times, October 16, 1941, 20.

[400] Kenneth M. Pollack, Arabs at War: Military Effectiveness, 1948-1991 (Lincoln, NE: University of Nebraska Press, 2002, 183).

[401] David M. Glantz and Jonathan House, When Titans Clashed: How the Red Army Stopped Hitler (Lawrence, KS: University Press of Kansas, 1995), 31.

[402] Quoted in Christopher Andrew and Vasili Mitrokhin, The World Was Going Our Way: The KGB and the Battle for the Third World (New York: Basic Books, 2005), 403. See also 399-402.