Liberal democratic peace theory (LDP), formulated largely in the 1960s-1980s, is viewed as one of the great wonders of liberalism and a major point cited in favor of evangelizing liberalism to the rest of the world.
The basic thesis is that liberal democratic states tend not to go to war with one another, because of the structural and/or normative features of liberal democratic states. That is, liberalism as such is directly responsible for the miraculous track record of peace that we’ve seen post-World War II, especially in Europe and North America.
In the post-World War II environment and doubly so in the post-Cold War environment, more and more of the world was moving towards liberal democracy. This was cause for excitement; world peace was on the horizon. The Soviet Union had fallen, dictatorships were collapsing, and the number of democracies was skyrocketing. Noted historian Spencer Weart boldly claimed that a “preponderance of democracies will transform the entire system of international relations.”
The arc of history was bending in the direction of liberalism and towards the democratic peace.
But something happened in the early 2000s. The unfolding of history as liberalism had been disrupted. The attacks of September 11, the subsequent bungling of intelligence regarding WMDs, the United States’ decision to drag its NATO allies into a series of embarrassing and ultimately ineffectual wars in the Middle East, numerous botched attempts at regime change and nation-building, and aggressive ideological demands as pre-conditions for diplomacy, have devastated the credibility of the U.S.-driven, Western liberal order. In some areas, history is now forking from the liberal path with a surprising confidence.
Axios declared 2018 the year of the strongman, but it can also be called the year of liberalism in crisis. Russia, Egypt, China, Poland, Hungary, the Philippines, and Turkey—among others—have gained a renewed sense of self and conviction in their opposition to liberalism, as opposed to merely carving out for themselves an unprincipled exception to liberalism and ceding it the moral high ground. Even in Europe, where liberal democracy claims some of its greatest accomplishments, liberalism is beginning to unravel on the periphery in countries like Poland, Hungary, and Italy. Jair Bolsonaro, who notably expressed sympathy for a military dictatorship, has done tremendously well so far in the Brazilian presidential election.
Neither soft, nor hard evangelism of liberalism is working the way it used to.
For liberal theorists, the prospect of a liberal democratic world peace is now questionable. The Economist has put out a call to arms to defend liberalism as the most successful idea of the last 400 years. Whether this call results in anything substantial remains to be seen. But there is a palpable sense that liberalism is facing a serious new opposition.
Despite this trend, liberalism’s theorists have a hope encapsulated by Daniel Deudney and G. John Ikenberry’s 2009 pronouncement that “liberal states should not assume that history has ended, but they can still be certain that it is on their side.”
And yet, in the nine years following this bold pronouncement, history has not been especially kind to liberalism.
Texas A&M political science professor Christopher Layne, writing in early 2018, has summed up the current situation aptly, namely that China as an increasingly self-aware superpower is already reshaping the world order away from Pax Americana:
The US foreign policy establishment does not grasp this, and, instead, has invested the idea of a ‘rules-based, institutionalized’ international order with a talismanic quality. It claims that rules and institutions are politically neutral, and, ipso facto, beneficial for all. However, in international politics, who rules makes the rules. Rules and institutions reflect the distribution of power in the international system. A power transition is taking place in the early twenty-first century: US power is in relative decline and China is rising quickly. No international order—not even the Pax Americana—lasts forever. The liberal world order cannot survive the erosion of US hegemonic power. It is this structural change, not Donald Trump, that threatens the post-Second World War international order’s survival. It requires a huge leap of faith to believe that a risen China will continue to subordinate itself to the Pax Americana.
Which brings us to the question: given that the prospect of an expanded democratic peace is starting to hit some hard limits, how well does liberal democratic peace theory actually explain the peace that has existed among liberal democratic states?
Although LDP theory is taken as an empirical law in international relations, it has picked up a number of qualifications, making the theory far less interesting than first presented. Here are various renditions of the theory since its modern inception in the 1960s:
Claim 1: liberal democracies go to war less than non-liberal democracies.
The most significant abandonment in the field has been of the monadic hypothesis: that liberal democratic states are less likely to go to war compared to non-liberal democratic states. Henry Farber and Joanne Gowa have convincingly shown that, statistically, democracies on average have the same probability of going to war as non-democracies. At this point, no one holds the monadic hypothesis.
Claim 2: no liberal democracy has fought another liberal democracy since 1816 (notably after 1812)
This claim from Michael Doyle has increasingly fallen out of favor. There weren’t many democracies prior to 1945, and so the number of dyadic pairs of liberal democratic countries that could have gone to war was extremely limited. For example, Doyle excludes Great Britain from counting as a liberal democracy until 1832 and denies that the Spanish-American War of 1898 should count, even though the Polity II data set lists Spain as a democracy in 1898. Others have argued that the Second Philippines War of 1899 shouldn’t count either, as at the time the Philippines hadn’t held an election yet. This strong claim has largely been dropped because of a) exception cases that require new qualifications to explain, and b) the fact that there weren’t enough existing liberal democracies to achieve significance. All totaled, there are around 50 exception cases suggested in the literature.
Claim 3: mature liberal democracies tend not to go to war/don’t go to war with one another
Although Peru and Ecuador fought while both were liberal democracies, some maintain that the war occurred before the “pacifying effects” of liberal democracy had enough time to work, meaning maturity was lacking. Yet another qualification. Interestingly, the U.S. intervened militarily to topple Salvador Allende after he was democratically elected in Chile. Israel invaded Lebanon in 1981, but at the time, Lebanon’s government was in disarray. R.J. Rummel states that the Israeli attack on the United States’ USS Liberty in 1967 counts as conflict below the level of war, but argued that Israel at the time was “only partially free.” That Finland joined up with the fascists from 1941 to 1944 apparently shouldn’t count, as no Finnish troops attacked troops from a liberal democratic country, though Great Britain bombed Finland in 1941.
Claim 4: mature liberal democracies in the post-1945 era don’t go to war with one another
The addition of the ‘post-1945 era’ qualification came later in the development of LDP theory, and while it may be a disappointing qualification, it may be necessary, given the small-N universe of liberal democracies before 1945. Bruce Russett admits that “the absence of murderous quarrels between democracies was not too surprising, and may need—at least for the pre-1945 era—little further explanation.”
Claim 5: mature liberal democracies in the post-1945 era don’t go to war with one another, but for normative reasons, not structural reasons
It’s claimed that the structural elements of liberal democratic states, namely separation of powers, checks and balances, elections, the free press, etc. should result in less war in general, but this is essentially equivalent to the monadic hypothesis, which has been largely abandoned. So, it might be possible to explain the absence of war through the normative mechanisms of liberal democratic states, as opposed to structural elements. Zeev Maoz and Bruce Russett argue that while there is some support for the structural model, the normative model has more support and better consistency.
The above examples illustrate continuing iterations of the theory. This trajectory is by no means perfectly linear, nor is there consensus on how many qualifications are required, but the fact that these qualifications have continued to pile on is evidence that LDP theory is now much closer to triviality than when first introduced.
But on some level, LDP seems intuitive, first because we’re naturally more amenable to arguments that uphold our traditions of liberalism, and second because both North America and Europe are now overwhelmingly comprised of liberal democratic states and have enjoyed a near-unparalleled level of peace and prosperity over the last 70 years.
While it is true that there is an empirical track record of peace and prosperity between liberal democratic states in Europe and North America, the absence of war does not automatically vindicate the theory. Zero is not necessarily statistically significant, since that result might simply be predicted by random chance, as is the case with lightning strikes. People are struck by lightning all the time, but none of my friends have ever been. In general, the statistical approach adopted by much of the literature is somewhat methodologically wrong-headed; it assumes statistical properties which don’t necessarily hold in fat-tailed structural processes like war.
Aside from issues of statistical significance, the objective is to scrutinize the causal mechanisms supporting LDP theory—a fundamental step of theory-building. While there are a host of mechanisms adopted to explain democratic peace, astonishingly little research has been conducted into whether these mechanisms of liberalism actually work to guarantee the peace.
Michael Doyle’s recent expression of LDP theory in the article “Three Pillars of the Liberal Peace” stands out as a careful explication of the causal logic underlying LDP theory.
For Doyle, there are three mechanisms at play, all of which must be in operation simultaneously to generate the peace. The first mechanism is that LD states are accountable to voters, especially the median voter. As such, these states “preclude monarchs or dictators turning their potentially aggressive interests into public policy while assuming that the costs will be borne by a subordinate public.”
The second mechanism is that liberal principles or norms entail respect for legitimate individual rights, which are externalized in foreign policy. As Doyle puts it, “Domestically just republics…presume foreign republics to be also consensual, just, and therefore deserving of the accommodation that the individuals that compose them deserve.” The free press in liberal democratic states also helps to hold officials accountable to liberal principles.
The third mechanism is that commerce between liberal democratic states binds them together in a way that strongly discourages war.
The first mechanism of accountability assumes that voters pay close attention to war, don’t like war, and can’t be manipulated into bearing the costs of war. These assumptions seem implausible when examining America’s track record over just the last two decades. Yemen, as just one example out of countless, probably does not even register as a country to the median voter. The U.S. is currently at war in at least seven different countries.
The first mechanism also assumes voters who are able to accurately assess the success or failure of policies, accurately assign responsibility for success or failure, and then electorally reward or punish elites. This concept is known as retrospective voting. In retrospective voting, voters cast ballots based on results, not policies. Outcomes, so it goes, function as a heuristic for voters to rely on to overcome policy ignorance and compel liberal to act beneficially.
But in a 2013 study, Bryan Caplan et al. determined that voters systematically overestimate the influence of politicians on the economy and systematically underestimate their influence on education and the federal budget. The prognosis for retrospective voting is grim: “As long as ignorant and irrational voters know enough to properly reward success and punish failure, democracy can still work well. Unfortunately, retrospective voting requires a largely undefended assumption: Voters’ beliefs about political influence are unbiased.” That is, voters overestimate, underestimate, and misallocate influence.
Numerous other studies have examined voters’ strange propensity to vote out incumbents based on bad circumstances out of incumbents’ control. One such study found that voters punished U.S. governors in oil-producing states based on fluctuating world oil prices. Another determined that the success or failure of incumbent politicians in Latin America was based on U.S. interest rates, in addition to international commodity prices. Voters also consistently punish incumbents for droughts and floods.
While the above work undermines retrospective voting in the abstract, Sebastian Rosato looked specifically at voter behavior in the context of lost wars and found that both liberal democratic and autocratic leaders have similar chances of being removed from power for losing wars, but not a single one of the liberal democratic leaders was actually punished. In contrast, 29 percent of autocratic leaders were either killed, imprisoned, or exiled. For wars that are costly, but not lost, that is, wars where there is one fatality per 2,000 population, autocrats were removed from power 35 percent of the time. They were also punished in 27 percent of cases. Democratic leaders, on the other hand, get off a lot easier, and were only removed 27 percent of the time and punished in just 7 percent of cases.
Such findings cast doubt on the first pillar for explaining the peace.
The second pillar focuses on the externalization of liberal democratic norms and principles to the international arena via media publicity and other avenues, which causes liberal democratic states to treat other liberal democratic states with trust and respect, even in cases where there are serious clashes of interests.
But norm externalization has not applied in crucial cases. In a 1994 paper, Christopher Layne used process-tracing to determine that liberal norms and principles did not explain any part of why near-miss wars did not occur between the United States and Britain in 1861, France and Great Britain in 1898, and France and Germany in 1923, as public opinion was not pacific and elites did not refrain from making threats, among other indicators.
Additionally, if liberal democratic norms operate internationally to prevent conflict, a fortiori they should operate domestically. But according to Håvard Hegre, once GDP per capita is controlled for, it turns out that democratic states are no less likely to suffer from internal conflict, compared to autocracies. The historical record also indicates that the U.S. aggressively intervened in democracies and subsequently installed autocratic regimes. Several recent and notable interventions were: Iran in 1953, Guatemala in 1954, Indonesia in 1957, British Guyana in 1961, Brazil in 1961 and 1964, Chile in 1973, and Nicaragua in 1984.
Doyle’s response is two-fold: first, the democracies weren’t democratic enough or weren’t fully liberal democratic states, and second, these U.S. interventions were covert, meaning that elites subverted a crucial liberal democratic mechanism.
The first response could arguably apply to Iran, Indonesia, and Nicaragua, which at the time were quite immature democracies, but certainly the regimes at the time were more democratic than previous regimes. The other states, however, were much more democratic.
The second response about covert action misses the point. If the mechanisms to explain such an outcome do not work as described in even a small number of cases, so much the worse for the mechanisms. Appeal to covert action to save the mechanism is simply an admission of its limited explanatory power.
The third and final pillar concerns the claim of conflict-reducing interdependencies between nations caused by trade, but even if this is true—and the literature is mixed on this point—trade is not exclusive to liberalism. And the argument that free trade specifically is necessary for this mechanism to work is 1) putting an enormous amount of stock in the difference between “free” and “unfree” trade, and 2) missing the point that free trade may be best understood here as a proxy either for liberal democratic states in the same military alliance or for the relationship between a liberal democratic state and a small vassal state, in which no military conflict is necessary to achieve desired economic domination.
For Doyle, the underlying mechanisms must always be in play for LDP theory to work. But this position is dangerously close to a tautology, as the second mechanism holds that liberal democratic states’ liberal principles of respect, trust, and the recognition of legitimate individual rights are externalized in its foreign policy. What this effectively translates to is that it necessarily has to be the case that liberal democratic states treat other liberal democratic states with respect and trust for there to be a low likelihood of war, which is uninformative.
If the causal logic of these mechanisms is broken, or if there are numerous historical case studies in which these mechanisms actually do not prevent war or explain near-misses between liberal democratic states, then other candidate mechanisms should be explored to explain the democratic peace.
One way of trying to answer the question of what best explains the lack of war in Europe and North America is to imagine hypothetical scenarios from the perspective of elite decision-makers in a dyad and come up with plausible conditions under which war might emerge. What conditions would have to obtain in the current geopolitical environment for Germany to go to war with France, or England to go to war with France, or Germany to go to war with Greece, etc.?
Let’s start with the last hypothetical, although it will sound odd because war in Western Europe is almost inconceivable in the post-World War II environment.
But imagine we broaden the circle and German decision-makers become so incensed with Greece’s lack of fiscal continence that Germany decides to invade Greece.
First, Germany’s military is currently not capable of launching and sustaining a land or air war in Europe.
Second, both Germany and Greece belong to the same U.S.-engineered military alliance, NATO, and so other NATO member states would be obliged to defend Greece.
Third, in order to get to the point where the proposal of war in this instance would even be on the table, the German state would have to be filled with—and German voters would have to vote for—an entirely new elite dominated by an entirely new and aggressive way of thinking. The United States engineered the post-World War II environment specifically to ensure this does not occur, and continually maintains this environment with both hard and soft power—the former via its enduring military presence on the continent and the latter via diplomatic censure and its many U.S.-funded or coordinated non-governmental organizations.
Replace Greece in this example with France, and an additional point of consideration appears: France has nuclear weapons.
For any other conceivable dyad, many or most of these considerations will also apply, and none have any strong relationship with liberal principles or mechanisms, with the exception in this case of the liberal character of the German elite. However, this of course is due to the fact that a large part of Germany was a U.S. vassal state immediately after 1945 and still maintains much of that character as a militarily and diplomatically-occupied country.
It would be a mistake to think of foreign policy decision-makers as purely motivated by power, just like it would be a mistake to conceive of them as purely utility maximizers. Humans in general are motivated by a multiplicity of goods and possess a certain baseline level of sociability. But the claim is that the motivations of sociability for peace are downstream of identity relations, and identity relations in the post-1945 era emphatically were constructed as part of Pax Americana, the new, U.S.-led order.
Moreover, due to the way sovereignty is de facto distributed in the post-World War II environment, Germany is not quite the right unit for war and is part of an arguably post-national order. While Germany is technically considered to be a sovereign state, the idea of it practically exercising this sovereignty to invade Greece and get away with it is not very plausible. The long arm of the U.S., NATO, the European Union, and the United Nations would put a stop to such a move.
The more appropriate unit for war is more like NATO versus a rogue liberal democratic state or a rogue NATO member state—perhaps Turkey.
If NATO did in fact go to war against its member state Turkey, democratic peace theorists might claim that this is no way counts as a point against their thesis. After all, Turkey isn’t exactly a shining example of liberal democracy.
But what they’re missing is that regime character is downstream of power relations and historical processes. In the West, liberal democracy is the only game in town, and the extent to which other countries have adopted liberal democracy is a reliable indicator of the extent to which they have become integrated into the U.S.-dominated geopolitical order. The process of “becoming liberal,” in other words, is virtually indistinguishable historically from “integration into the U.S.-dominated geopolitical order.”
And integration has not been a passive process. Non-democracies, as Christopher Hobson has noted, have historically been labeled threatening, both behaviorally and ontologically by the West. In what almost follows naturally from this labeling process, the West has taken upon itself to employ coercive democratization or sanctions to subordinate recalcitrant polities, often resulting in worse developmental indicators in these countries, which again reinforces the superiority of liberal democracies over other competing systems. We can see this process currently at play in Iraq, Iran, Russia, and North Korea, among many others.
Unfortunately, we don’t have the luxury of examining a parallel universe identical to ours with the exception that geopolitical historical processes have no impact on regime character, such that it can be observed that states adopt liberal democracy in a vacuum and then exhibit the democratic peace among one another. In other words, we have a problem of collinearity. Still, that doesn’t mean a case for LDP can’t be made in the actual world. However, there are few convincing case studies of countries adopting liberal democracy in the absence of other pacifying factors, such as power relations during the Cold War, in which countries either sorted themselves or were sorted into blocs based on alliance with the Western order or the Soviet order.
The world order is changing. Liberal democratic peace theory is slowly losing relevance. Liberal evangelism feels passé, given recent regime change failures in the Middle East and North Africa, and backsliding in Eastern Europe. Theorists may soon have to confront the possibility that liberal democratic regimes, rather than demonstrating intrinsic benefits of liberalism, are little more than the preferred regime-type of Pax Americana.