As the Watergate crisis escalated in late 1973, rumors traveled around Washington that President Richard Nixon might call on the military for support. Years later, U.S. Admiral Elmo Zumwalt, who had been chief of naval operations in those days, recalled a meeting between Nixon and the Joint Chiefs of Staff in December 1973. According to Zumwalt, Nixon launched into a rambling speech about how “the Eastern liberal establishment was out to do us all in.” Nixon’s next suggestion was shocking. “We gentlemen here,” declared the president, “are the last hope, the last chance to resist.”

“I got the impression,” Zumwalt recalled, that “he was sort of testing the water with us, to see whether there would be support -- any nodding of heads -- at some of these things. One could well have come to the conclusion that here was the Commander-in-Chief trying to see what the reaction of the Chiefs might be if he did something unconstitutional. . . . He was trying to find out whether in a crunch there was support to keep him in power.”

None of the military commanders responded. After the meeting, Zumwalt conferred with Army Chief of Staff Creighton Abrams, who said he would simply pretend the episode never occurred. If Zumwalt’s depiction is accurate, Nixon floated a trial balloon to see whether the military brass would support him in a coup d’état against Congress. But the balloon popped instantly, and Nixon faced the humiliating trial with stoicism, not Machiavellian machination.

This was not the first time in U.S. history in which a leading official or military commander considered staging a coup. During the U.S. Civil War, George McClellan, a top Northern general, apparently contemplated deposing President Abraham Lincoln over his stance on slaves. Shortly after Lincoln issued the Emancipation Proclamation, which freed all slaves living in Confederate territory, McClellan invited three fellow Union generals to dine with him. After dinner, he told them that his admirers were urging him to take a public stand against the proclamation. The generals pleaded with him to avoid a confrontation with the president and assured him he would be without support. McClellan backed off.

Today, Americans take for granted that their leaders ascend to power through constitutional procedures and step down when legally required -- that no one, not even a president or a major general, can overturn the country’s legal order. They have never suffered a coup or even seen a serious attempt at one. Likewise, almost all of the world’s developed states have gone decades, if not centuries, without a major upheaval, and it is difficult to imagine tanks rolling through modern-day London, Stockholm, or Tokyo. Yet coups happen regularly around the world, in such places as Caracas, Kathmandu, Kinshasa, and Quito. 

For a long time, political scientists have argued that the chances of a coup are a function of military professionalism. The armies of coup-free states, the reasoning goes, are deeply committed to protecting constitutional norms. But a closer look at the historical record leads to a different conclusion: stability is a product not of civil virtues, but of the rule of law. Impartial judicial systems weaken personal loyalties, making it far more difficult to plan grand criminal enterprises. No one can execute a coup without trustworthy followers, and as McClellan and Nixon discovered, in modern democracies, that sort of loyalty is in short supply.


To pinpoint the factors that prevent coups, one must first specify which countries have joined what could be termed “the coup-free zone” -- by avoiding violent overthrows for at least 50 years.

Between 1901 and 1960, nearly every independent country experienced at least one coup. (Sweden, Switzerland, the United Kingdom, and the United States were the only outliers.) The coups came in the form of military takeovers, palace intrigues, and street revolutions. Some posed serious threats to the existing order but ultimately failed. Others were dramatic expansions of executive power, whereby a constitutional leader forcefully extended his or her reach, as Hitler did in Germany in 1933.

Between 1961 and 2010, the coup-free zone expanded by an additional eighteen countries. And the club will soon grow larger. France, the most recent member, passed its 50-year milestone in 2011, half a century after the 1961 putsch by French generals in Algeria. Among the states poised to enter the zone are Singapore, in 2015; Botswana, in 2016; and Italy, in 2020.

What explains the shift toward orderly politics? Most political scientists have adopted a standard theory: that in coup-free countries, professionalized military officers have internalized the values of civilian control and constitutionalism. “The major factor in civil control,” wrote Louis Smith, a military historian, “lies in the fact that the military have never manifested any ambition to usurp first power in America and to overwhelm for all our citizens the great values of freedom under the law. In entering the armed forces, the American does not put off the citizen in becoming the soldier. The habits of obedience to authority and respect for law persist.” Other prominent scholars, including Samuel Huntington, Samuel Finer, and Eric Nordlinger, have offered a similar rationale. As Huntington wrote in The Soldier and the State, “Professional attitudes and behavior among the members of the officer corps” assure civilian control. “For the military to remain subject to civilian control,” concluded the historian Peter Karsten, commanders must “believe that it should remain subject to civilian control.” 

Thus far, the only serious challenge to this virtue-based theory has come from Bruce Farcau, a U.S. Foreign Service officer who was stationed in Bolivia. While there, he observed two coups firsthand. Farcau has argued that the scale and complexity of modern states are what make coups difficult. In the United States, he wrote, “even a modicum of control over the country could not be established without seizing dozens of locations in half a dozen massive cities, to say nothing of hundreds of transportation chokepoints, airports, television and radio stations, microwave transmission centers, and many major military bases scattered over tens of thousands of square miles of territory.” In this view, the prospects for success depend on “the number, diversity, and dispersion of targets.” 

The historical record, however, suggests otherwise. The Republic of Venice, which stayed coup-free for 486 years, from 1310 to 1797, was a highly centralized imperial state, with just a few critical targets inside the city. Other tiny countries, such as Luxembourg and Iceland, have also gone long periods without experiencing a coup, whereas historical giants, such as the Byzantine, Chinese, Ottoman, and Roman empires, experienced one takeover after another.

Still, the story many U.S. military leaders tell -- that their own loyalty to the Constitution has prevented coups -- fails to account for centuries of orderly succession. Their explanation has two main problems. First, all political leaders seek to instill obedience and loyalty within their security forces. Civic virtue, in other words, does not in itself explain how professional values take root within certain states but not in others. Second, inculcating these values is extremely difficult in such large and complex organizations as the U.S. military. Education cannot nullify basic self-interest over the long term. Virtue simply doesn’t explain why more politicians or military commanders have not tried to mimic Caesar, Napoleon, or Mussolini.


Countries enter the coup-free zone not by changing their militaries, but by altering incentives. The networks that typically support coups are usually based on bonds of family, clan, and patronage -- ties that transcend loyalty to a weak state. When loyalties are strong, members of the same group can safely suggest, discuss, and plan a new criminal enterprise. Even if a majority within the group reject a proposed coup, nobody gets turned in; members trust one another more than they trust outsiders. And if the group does decide to move ahead, it might need to make contact with representatives of only a few other similar groups.

But in the absence of tight-knit communities, crossing the threshold into a criminal conspiracy becomes fraught with peril. As the experiences of Nixon and McClellan suggest, senior officials subject to the rule of law cannot even insinuate a coup without maintaining plausible deniability. Merely mentioning such a plot is far too risky, so the notion effectively disappears from national political life.

Of course, all this hinges on the existence of an independent judicial system that systematically subjects the state to legal scrutiny -- correcting abuses, minimizing corruption, and resolving disputes in an objective manner. When laws mediate relationships among individuals, people have less need for personal allegiances and mutual assistance. When laws don’t, members of particular groups can assure their rights only through their personal networks; and in return, they are expected to favor their own people whenever possible.

Historically, the rule of law has always gone hand in hand with long-term stability. In Renaissance Venice, all citizens, regardless of rank, were subject to a uniform, dependable, and impartial legal code. And in England, the British legal system became exceptionally fair by the middle of the eighteenth century, after which coups no longer took place.

In both instances, feudal and familial ties faded away as the law took center stage. Early medieval Venice seethed with blood feuding and factional rancor, but from the Renaissance onward, the city became comparatively tranquil. John Julius Norwich, a historian, describes a “unique spirit of cohesion and cooperation” and a “mutual trust of a kind that in other cities seldom extended far outside the family circle.” England similarly transitioned from a medieval society marked by blood feuds, through a period of lordship and vassalage in the High Middle Ages, to a law-based society of individuals by the eighteenth century.

Once states have credible legal systems in place, corruption plummets as well. Transparency International, a Berlin-based think tank, annually ranks countries according to their levels of public sector corruption, as perceived by analysts and businesspeople. The following chart presents the organization’s 2008 results. The twenty-two coup-free states -- all of which experienced no coups from 1961 through 2010 -- are listed in bold.

Ivan Perkins

Admittedly, the correlation is not perfect. Four coup-free states have relatively high corruption scores: Costa Rica, Israel, Mexico, and South Africa. But that is because they are recent entrants to the coup-free zone: South Africa gained independence from the United Kingdom in 1934; Mexico saw a final coup attempt in 1938; Israel has remained stable since 1948; and Costa Rica’s last coup attempt came in 1955. Also, Costa Rica and Mexico are far from stable. Over the past 50 years, rumors of plots in both countries have been rampant. By and large, then, the connection between corruption and coups remains strong.


Just as the rule of law discourages military coups, it also discourages revolutionary ones. Like any other coup plotters, street revolutionaries must also attract followers. But they appear heroic -- and galvanize imitators -- only when they sacrifice their own safety to oppose a truly malevolent regime. After weathering decades of grim surveillance, for example, eastern Europeans poured onto city streets in 1989. Sheer numbers inspired hope and fed momentum. When government security services refused to crack down, though, the regimes stepped aside. A similar process unfolded in 2011, as Tunisians and Egyptians gathered to end decades of dictatorship.

Governments subject to the rule of law are usually incapable of kindling such outrage in the first place. Instead, they make would-be revolutionaries seem dangerous. In established democracies, the closest parallels to insurrectionary violence are urban riots, such as those in Los Angeles in 1992 and Paris in 2005. Many of these episodes involve relatively poor members of racial minorities who feel aggrieved by systemic biases in the justice system. And although the rioters make cities momentarily ungovernable, they don’t have a chance at toppling governments, partly because the general public and state security services do not share their grievances.

Mass revolutions are impossible in rule-of-law societies for another reason: no civilians -- no matter how angry or militant -- can effectively challenge the immense power of a united political-military establishment. As Lenin famously put it, “No revolution of the masses can triumph without the help of a portion of the armed forces.” As long as the state is united, it can almost always contain civil unrest.


Flawed theories about coup-free states have come at a high cost. In Afghanistan and Iraq, the United States spent valuable resources attempting to build professional militaries. But as the results have made clear, simply training soldiers to observe professional norms does not work in the absence of strong laws. And legal systems take many years to mature.

Take Haiti, for instance. Following the U.S.-led intervention in 1994, the UN launched a major effort to create a national police force. It recruited, trained, and deployed some 5,000 new officers. Selection standards were high, and recruits went through an intensive four-month training course conducted by U.S., Canadian, and French instructors. Initially, things appeared to be working. Officers who committed abuses were investigated and punished -- a radical departure from the past. But persistent corruption in the courts and the political system ultimately troubled the ranks. Many officers quit, and others, imitating judges and politicians, went on the take.

As Washington prepares to leave Afghanistan, U.S. leaders should keep this cautionary tale in mind. They should not fool themselves into believing that Afghan soldiers will adhere to civilian control. Instead, Afghanistan’s profoundly corrupt, tribal, and warlord-riven state portends a violent future. Similarly, Iraq is now experiencing the travails that accompany an absence of the rule of law -- despite years of U.S.-led training of the military.

But in the long run, the prospects for stability are not all that bleak. The coup-free zone is continuing to expand, and as rule-of-law systems have grown stronger, the global coup rate is dropping. The world now sees about six coup attempts per year, less than half the average rate during the 1970s. There is reason, then, to believe that coups might one day become rarer — even in Afghanistan.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now