by JOHN M. OWEN IV
Foreign Affairs, November/December 2005
WHO SAYS DEMOCRACIES DON'T FIGHT?
Seldom if ever has the hostility between academics and the U.S. president been so pronounced. Of course, political scientists always seem to complain about the occupant of the White House, and Republicans fare worse than Democrats: Herbert Hoover was called callous, Dwight Eisenhower a dunce, Richard Nixon evil, Ronald Reagan dangerous, and George H.W. Bush out of touch. But professors have consigned George W. Bush to a special circle of their presidential hell. And the White House seems to return the sentiment.
According to the academics, Bush's chief transgressions have had to do with foreign policy, especially the Iraq war -- a mess that could have been avoided if only the president and his advisers had paid more attention to those who devote their lives to studying international relations.
The irony of this argument is that few other presidents -- certainly none since Woodrow Wilson, a former president of the American Political Science Association, scribbled away in the Oval Office -- have tied their foreign policies more explicitly to the work of social science. The defining act of Bush's presidency was grounded in a theory that the political scientist Jack Levy once declared was "as close as anything we have to an empirical law in international relations," namely, that democracies do not fight one another.
The theory, which originated in the work of the eighteenth-century philosopher Immanuel Kant and was refined in the 1970s and 1980s by several researchers working independently, has, since the 1990s, been one of the hottest research areas in international relations. Although some skeptics remain and no one agrees about why exactly it works, most academics now share the belief that democracies have indeed made a separate peace. What is more, much research suggests that they are also unusually likely to sign and honor international agreements and to become economically interdependent.
The administrations of Presidents George H.W. Bush and Bill Clinton made frequent appeals to the theory in public, and it seems to have informed their support for democratization in former communist lands and in Haiti. The current Bush administration, however, has gone much further in its faith in the idea, betting the farm that the theory holds and will help Washington achieve a peaceful, stable, and prosperous Muslim world as, over time, Iraq's neighbors, following Iraq's example, democratize. The United States' real motives for attacking Iraq may have been complex, but "regime change" -- the replacement of Saddam Hussein's gruesome tyranny with a democracy -- was central to Washington's rhetoric by the time it began bombing Baghdad in March 2003.
Why has a president who set his defining policy around one of political science's crown jewels come in for so much venom from the same academics who endorse the idea? After all, a host of peer-reviewed journal articles have implicitly supported the president's claim that a democratic Iraq would not threaten the United States or Israel, develop weapons of mass destruction, or sponsor terrorism. Are professors simply perpetual critics who refuse to take responsibility for the consequences of their ideas? Or does Bush hatred trump social science?
The Bush administration's desire to break with its predecessors and alter the authoritarian status quo in the Middle East was admirable. But the White House got its science wrong, or at least not completely right: the democratic peace theory does not dictate that the United States can or should remake Iraq into a democracy. In Electing to Fight: Why Emerging Democracies Go to War, the veteran political scientists Edward Mansfield and Jack Snyder make two critical points. Not only is turning authoritarian countries into democracies extremely difficult, much more so than the administration seems to have anticipated. The Middle East could also become a much more dangerous place if Washington and the rest of the world settle for a merely semidemocratic regime in Baghdad. Such an Iraq, Mansfield and Snyder imply, would be uncommonly likely to start wars -- a bull in the Middle Eastern china shop. Unfortunately, such an Iraq may also be just what we are likely to end up with.
At first glance, the realists' critique of the Iraq war is easier to understand than that of the democratic peace theorists. Indeed, realism -- which holds that a country's type of government has no systematic effects on its foreign policy -- is enjoying a revival in Washington these days, precisely because of the war. According to the realists, the best way to have dealt with Saddam would have been not to overthrow him but to use coercive bargaining: to have threatened him with annihilation, for example, if he ever used nuclear weapons.
Even the democratic peace theory, however, does not necessarily prescribe the use of force to transform despotisms such as Iraq into democracies. Indeed, by itself, the argument that democracies do not fight one another does not have any practical implications for the foreign policymaker. It needs an additional or minor premise, such as "the United States can make Iraq into a democracy at an acceptable cost." And it is precisely this minor premise about which the academy has been skeptical. No scholarly consensus exists on how countries become democratic, and the literature is equally murky on the costs to the United States of trying to force them to be free.
This last part of the puzzle is even more complicated than it first appears. Enter Mansfield and Snyder, who have been contributing to the democratic peace debate for a decade. Their thesis, first published in 1995, is that although mature democracies do not fight one another, democratizing states -- those in transition from authoritarianism to democracy -- do, and are even more prone to war than authoritarian regimes. Now, in Electing to Fight, the authors have refined their argument. As they outline in the book, not only are "incomplete democratizing" states -- those that develop democratic institutions in the wrong order -- unlikely ever to complete the transition to democracy; they are also especially bellicose.
According to Mansfield and Snyder, in countries that have recently started to hold free elections but that lack the proper mechanisms for accountability (institutions such as an independent judiciary, civilian control of the military, and protections for opposition parties and the press), politicians have incentives to pursue policies that make it more likely that their countries will start wars. In such places, politicians know they can mobilize support by demanding territory or other spoils from foreign countries and by nurturing grievances against outsiders. As a result, they push for extraordinarily belligerent policies. Even states that develop democratic institutions in the right order -- adopting the rule of law before holding elections -- are very aggressive in the early years of their transitions, although they are less so than the first group and more likely to eventually turn into full democracies.
Of course, politicians in mature democracies are also often tempted to use nationalism and xenophobic rhetoric to buttress their domestic power. In such cases, however, they are usually restrained by institutionalized mechanisms of accountability. Knowing that if they lead the country into a military defeat or quagmire they may be punished at the next election, politicians in such states are less likely to advocate a risky war. In democratizing states, by contrast, politicians know that they are insulated from the impact of bad policies: if a war goes badly, for example, they can declare a state of emergency, suspend elections, censor the press, and so on. Politicians in such states also tend to fear their militaries, which often crave foreign enemies and will overthrow civilian governments that do not share their goals. Combined, these factors can make the temptation to attack another state irresistible.
Mansfield and Snyder present both quantitative and case-study support for their theory. Using rigorous statistical methods, the authors show that since 1815, democratizing states have indeed been more prone to start wars than either democracies or authoritarian regimes. Categorizing transitions according to whether they ended in full democracies (as in the U.S. case) or in partial ones (as in Germany in 1871-1918 or Pakistan throughout its history), the authors find that in the early years of democratic transitions, partial democracies -- especially those that get their institutions in the wrong order -- are indeed significantly more likely to initiate wars. Mansfield and Snyder then provide several succinct stories of democratizing states that did in fact go to war, such as the France of Napoleon III (1852-70), Serbia between 1877 and 1914, Ethiopia and Eritrea between 1998 and 2000, and Pakistan from 1947 to the present. In most of these cases, the authors find what they expect: in these democratizing states, domestic political competition was intense. Politicians, vying for power, appeased domestic hard-liners by resorting to nationalistic appeals that vilified foreigners, and these policies often led to wars that were not in the countries' strategic interests.
Although their argument would have been strengthened by a few comparative studies of democratizing states avoiding war and of full democracies and authoritarian states starting wars, Mansfield and Snyder are persuasive. In part this is because they carefully circumscribe their claims. They acknowledge that some cases are "false positives," that is, wars started by states that have wrongly been classified as democratizing, such as the Iran-Iraq War, started by Iraq in 1980. They also answer the most likely objections to their argument. Some skeptics, for example, might counter that Mansfield and Snyder get the causality reversed: it is war or the threat of it that prevents states from becoming mature democracies. Others might argue that democratizing states become involved in more wars simply because their internal instability tempts foreign states to attack them -- in other words, that democratizers are more sinned against than sinning. Analyzing data from 1816 through 1992, Mansfield and Snyder put paid to these alternative explanations. Bad domestic institutions usually precede wars, rather than vice versa, and democratizing states usually do the attacking.
Where does Electing to Fight leave realism, the dominant theory of international conflict? The quantitative data support the realist claims that major powers are more likely to go to war than minor ones and that the more equal are the great powers, the more likely are wars among them. But democratization makes war more likely even after one takes these factors into account. Furthermore, the case studies suggest that democratizing states very often lose more than they gain from the wars they begin, which implies that they do not respond to international incentives as rationally as realism would expect. That said, notwithstanding its preference for viewing states from the inside, the Mansfield-Snyder theory is still "realist" in the general sense that it assumes that politicians and other actors are rationally self-interested. Their self-interest simply involves building and maintaining domestic power as well as external security -- and sometimes trading some of the latter in order to gain the former.
The authors' conclusions for foreign policy are straightforward. The United States and other international actors should continue to promote democracy, but they must strive to help democratizing states implement reforms in the correct order. In particular, popular elections ought not to precede the building of institutions that will check the baleful incentives for politicians to call for war. Mansfield and Snyder are unsparing toward well-intentioned organizations that have pressured authoritarian governments to rush to elections in the past -- often with disastrous consequences. As the authors show, for example, it was organizations such as the World Bank and the National Democratic Institute that pushed Burundi and Rwanda to increase popular sovereignty in the early 1990s -- pressure that, as Mansfield and Snyder argue, helped set off a chain of events that led to genocide. Acknowledging their intellectual debt to writers such as Samuel Huntington (particularly his 1968 book Political Order in Changing Societies) and Fareed Zakaria, Mansfield and Snyder have written a deeply conservative book. Sounding like Edmund Burke on the French Revolution but substituting statistics and measured prose for rhetorical power, the authors counsel against abruptly empowering people, since premature elections may well usher in domestic upheavals that thrust the state outward against its neighbors.
BACK IN BAGHDAD
This brings the conversation back to Iraq, and in particular the notion that the United States can turn it into a democracy at an acceptable cost. In effect, Mansfield and Snyder have raised the estimate of these costs by pointing out one other reason this effort may fail -- a reason that few seem to have thought of. Forget for a moment the harrowing possibility of a Sunni-Shiite-Kurdish civil war in Iraq. Set aside the prospect of a Shiite-dominated state aligning itself with Iran, Syria, and Lebanon's Hezbollah. What if, following the departure of U.S. troops, Iraq holds together but as an incomplete democratizer, with broad suffrage but anemic state institutions? Such an Iraq might well treat its own citizens better than the Baathist regime did. Its treatment of its neighbors, however, might be just as bad.
Although Saddam was an unusually bellicose and reckless tyrant, attacking Iran in 1980 and Kuwait in 1990 and engaging in foolish brinkmanship with the United States, as Mansfield and Snyder imply, a democratic Iraq may be no less bellicose and reckless. In the near future, intensely competitive elites there -- secularists, leftists, moderates, and both Shiite and Sunni Islamists -- could compete for popularity by stirring up nationalism against one or more of Iraq's neighbors. And Iraq lives in a dangerous neighborhood. Already, Iraqi Shiite parties have been critical of Sunni-dominated Jordan; Iraqi Sunni parties, of Shiite-dominated Iran; and Iraqi Kurdish parties, of Turkey.
One hopes that the White House contemplated this scenario prior to March 2003. Whether it did or not, the possibility must be considered now, by U.S. civilian and military leaders, academics, and U.S. allies who agree with those academics. If Mansfield and Snyder are correct about the bellicose tendencies of young, incompletely democratized states, the stakes of Iraq's transition are higher than most have supposed. They are high enough, in fact, that those who called so loudly in the 1990s for an end to UN sanctions because Iraqis were dying but who are silent about the Iraqis who are dying now ought to reconsider their proud aloofness from the war. An aggressive Iraq, prone to attack Kuwait, Iran, Saudi Arabia, Syria, or Israel, is in no one's interest. The odds may be long that Iraq will ever turn into a mature democracy of the sort envisaged by the Bush administration. But those odds are lengthened by the refusal of those states in Europe and the Middle East that could make a difference actually to do so.