The Dictator's Handbook

by Bruce Bueno de Mesquita

  Dictators seem to like to hold elections. Whether they do so to satisfy international pressure (and gain more foreign aid), to dispel domestic unrest, or to gain a misleading sense of legitimacy, their preference is to rig the vote count. Elections are nice, but winning is nicer. Still, sometimes the people seize the moment of an election to shock the incumbent, voting so overwhelmingly for someone else that it is hard to cover up the true outcome.

  Liberia’s Sergeant Doe was foolish enough to hold an election. In doing so, he provided the impetus for protest that he was lucky to survive. In 1985, Thomas Quiwonkpa challenged Samuel Doe after it took weeks for Liberia’s electoral commission to “count” the votes. Perhaps Quiwonkpa took the commission’s dalliance as a sign of popular support and equally a sign of the commission’s lack of support for him. As his insurgency approached the capital, Monrovia, the masses took to the streets against Doe’s government. Unfortunately for them, Doe’s essential supporters remained loyal. The costs of protest became very real. Doe’s soldiers killed hundreds in retribution.

  In post-Soviet Eastern Europe, “legitimizing” elections helped to promote citizen uprisings. Rather than sustaining the regimes in power, elections created the opportunity to replace them. In 2004, the incumbent Ukrainian leader, Leonid Kuchma, having served two terms, decided, perhaps to the surprise of his essential backers, to respect the two-term limit and retire. His chosen successor was Viktor Yanukovych. The runup to the election looked like it came straight out of a John Le Carre spy novel, with the leading opposition candidate, Viktor Yushchenko, allegedly poisoned with dioxin, which left him horribly disfigured.

  In the first round of the elections in October each of the leading candidates received about 39 percent of the vote. This necessitated a runoff election on November 21, in which the official results differed greatly from exit polls. Even before the second round presidential runoff was complete, Yushchenko called for the people to take to the streets. The electoral commission declared Yanukovych the winner. However, protests mounted and the security forces withdrew. Eventually the Supreme Court ruled that given the high level of fraud, another ballot was needed. Yushchenko then won the election handsomely.

  Coalition dynamics play a key role in explaining why the security forces allowed the people to take to the streets. The president was changing. Although the retiring incumbent, Kuchma, backed Yanukovych, he could not ensure core supporters within the security forces that they would be retained after the transition. As we saw with Louis XIV and many others, newly empowered leaders, even when they have been chosen by their predecessor, are wise to shake up their coalition, bring in their own loyalists, and dump many of their predecessor’s erstwhile backers. The security forces, being uncertain whether they would keep their long-run privileges, declined to attack the masses, hedging their bets about who would be more likely to reward them. Without force to control the masses on the street, Yanukovych’s supporters deserted. The people brought Yushchenko to power, but an essential factor in their willingness to take to the streets was the apparent lack of support for Yanukovych by the security forces.

  Sometimes the shocks that spark revolt come as a total surprise. Natural disasters, while bringing misery to the people, can also empower them. One frequent consequence of earthquakes, hurricanes, and droughts is that vast numbers of people are forced from their homes. If they are permitted to gather in refugee camps, then they have the opportunity to organize against the government. You see, refugee camps have the unintended consequence of facilitating free assembly. Earthquakes, storms, and volcanoes can concentrate large numbers of desperate people with little to lose. They also can substantially weaken the state’s capacity to control the people.

  On the morning of September 19, 1985, a large 8.1 magnitude earthquake occurred on the Michoacan fault in the Pacific Ocean about 350 kilometers from Mexico City. Mexico City is geologically vulnerable as it was built on the soft foundation of the remains of Lake Texcoco. The clay silts and sands that make up the lake bed plus the soil’s high water content led to liquefaction (wherein the ground behaves like a liquid) during the earthquake. The city was also built in the absence of democratic rule, so few building codes had been enforced. As a result, the distant quake caused enormous devastation throughout the city. The death toll is highly disputed, but is thought to be between 10,000 and 30,000 people. An additional 250,000 were made homeless. The government did virtually nothing. Left to rescue themselves, the people formed crews to dig for survivors and organized refugee camps.

  Born of necessity, these camps became the foundation for an important political force in Mexico City. Instead of separate individuals unhappy with their government, the earthquake formed a concentrated mass of desperate people. Forced together into crowded camps, they shared their disillusionment with the government. Organizing a protest rally was suddenly relatively easy. Ready and willing participants were on hand and had little to lose. With the government largely absent, these social groups became important political forces that rapidly deployed as large antigovernment demonstrations. Unable to oppose these groups, the government sought to accommodate them. It is widely believed they played a key role in Mexico’s democratization.4

  The story of Anastasio Somoza’s deposition in the Nicaraguan Revolution in 1979 is broadly similar. In 1972, a 6.2 magnitude earthquake struck the capital of Managua, killing around 5,000 people and forcing about 250,000 homeless people into camps. Somoza and his cronies profited from disaster relief but did nothing to resettle the enormous number of homeless people who had gathered in refugee camps in the capital. These camps became organizing grounds for the activists who eventually ended Somoza’s reign.

  Not all autocrats make the mistake of ignoring disasters or ignoring the creation of refugee camps. Consider the case of Myanmar in 2008. Than Shwe is the military leader of Burma (officially known as Myanmar). Although he has been described as an unremarkable man, he understands the essentials of staying in power.5 On May 2, 2008, a massive cyclone, named Nargis, swept across the Irrawaddy Delta in southern Burma causing havoc. The delta’s residents, mainly poor fishermen and farmers, received no warning of the coming storm. The storm destroyed entire towns and villages. The official death toll is 138,000, though other estimates suggest it might be as high as 500,000.

  No one can blame Than Shwe for the storm or for the low-lying villages’ vulnerability to storm surge. However, Burma’s military regime provided no warning and did nothing to help the survivors, and for that they can be blamed. Indeed they did worse than nothing: they actively prevented help from being delivered. Many people in Rangoon, the major city in southern Burma that was itself heavily damaged by the storm, attempted to help those in the delta. They were not allowed. Small businessmen and traders were reduced to smuggling small amounts of food into what remained of towns and villages.

  The international community rallied to offer assistance. As tens, or possibly hundreds, of thousands of people died of hunger and thirst in the aftermath of the storm, ships full of disaster relief supplies sat off the coast. The military junta refused to allow relief workers in. Visas were almost impossible to obtain. Information was extremely scarce. The government requested aid, but asked that it be in the form of bilateral government-to-government assistance. Effectively, Than Shwe was saying, “send cash, but you can’t come in.”

  About a week after the disaster, the army started entering the larger towns and villages of the delta. They were not there to help. They were there to disperse survivors who had congregated in schools and temples. Even though their numbers rarely exceeded a few hundred, survivors were expelled from their shelters and told to return home. It mattered little that, in most cases, their entire village had been destroyed and they had no food, water, clothing, or shelter to return to. Indeed, one report observed,Survivors were loaded onto boats and ferried back to the destroyed villages they had recently escaped from. In some areas the clearances happened quickly; as the emergency phase was now
officially over, the authorities wanted people back in their villages by June 2, when the next school term was scheduled to begin. But survivors had no idea what they were returning to; was there even anything left at places they had once called home? And how would they get food and water there?6

  The government did not even attempt to answer these questions.

  In the PBS documentary, Eyes of the Storm, a senior Burmese general is seen addressing a group of survivors.7 Starving and destitute, they ask for a handful of rice. The general tells them that he is here now (but still he makes no offer of assistance) and that they must go back to their village and “work hard.” While the army seized (and sold on the black market) the few relief supplies allowed in, the people were told they could eat frogs. Effectively the government told these survivors to go away and die quietly: inhumane in the extreme, but good small-coalition politics. Dead people cannot protest.

  Are Disasters Always Disasters for Government Survival?

  Earthquakes and other disasters shake up political systems. However, the nature of the shakeup is very different under different institutions. Democratic leaders are very sensitive to disaster-related casualties. Allowing people to die reveals serious policy failure. Democrats need to deliver good public policy to reward their large number of backers. When they fail to do so, they are liable to be removed. Disaster-related deaths result in protest and in the removal of leaders in democracies.

  To illustrate the difference in political responses to poor disaster relief in a non-democratic and democratic setting, we contrast Cyclone Nargis with Hurricane Katrina. Katrina struck the US Gulf Coast in August 2005. This was the most costly natural disaster in US history, with damages estimated at $81 billion. The death toll was 1,836.

  The government, from President George W. Bush down to New Orleans’s mayor, Ray Nagin, stood accused of mismanagement and lack of leadership. Nagin delayed the evacuation order for the city until nineteen hours before the storm struck. As a result many people became trapped. Then, once the New Orleans Superdome football stadium was set up as an emergency center, it became overwhelmed when 30,000 rather than the anticipated 800 people showed up. Federal disaster relief was slow in arriving. Many of the casualties were the sick and elderly who were overcome by heat and dehydration.

  The tenure of US leaders was seriously jeopardized by the disaster. Many observers think Katrina contributed significantly to the Republican Party’s midterm electoral losses in 2006 and their significant losses, including the presidency, in 2008. Yet, while it is clear that the situation could have been handled much better, it bears no resemblance to Cyclone Nargis. In contrast, despite having allowed at least 138,000 people to die, Than Shwe felt sufficiently well entrenched to allow a farcical election in 2010, which the government-backed parties won easily (at least according to official sources).

  As seen in the cases of Mexico and Nicaragua, disasters can serve as rallying points in autocracies. Disasters can concentrate opponents of the regime, making it easier for them to coordinate. Yet the death toll from disasters has relatively little effect on a dictator’s chance of staying in power. Indeed, if anything, large numbers of people dying in disasters actually enhance the political survival of autocratic leaders.

  As we know, autocrats don’t buy political support with efficient public policy. Resources spent saving the lives of the people cannot be spent on cronies. In addition, as we have seen, autocrats are skilled at exploiting the international community. By letting more people die they may in fact be able to extract more relief assistance. The implications of these results are frightening. Small wonder, then, that far more people die in natural disasters in autocracies than in democracies.

  Letting people die is good governance in autocracy, but it is disastrous for the tenure of democrats. Although a detailed statistical analysis of the relationship between disasters,8 deaths and leader tenure is complex, we compared what happens in a country when 200 or more people die in a magnitude 5+ earthquake, to what happens in the same size earthquake if fewer people die. In particular, we looked at the effect of such circumstances on the odds of a country’s leader being removed from office within two years following the earthquake.

  An earthquake alone does not threaten the survival of democrats. However, if there are more than 200 people killed by the quake then a democratic leader is almost certain to be removed from office. Under normal circumstances, any democrat has a 40 percent chance of being ousted from office in any two-year period. But for a democrat whose country suffered 200 or more deaths in an earthquake, those odds rise to 91 percent. We believe this is the case because democratic leaders are supposed to deliver effective public policies, and those effective policies include ensuring good building codes are enforced and excellent rescue and recovery is implemented following a natural disaster. The death of many in such a disaster is a signal to everyone else that the leadership has not done an adequate job of protecting the people and so out go the leaders.

  Autocrats are less vulnerable to removal than democrats and earthquake related deaths have little effect on their hold on power. Over a typical two-year period, 22 percent of autocrats lose power. If their country suffers a magnitude 5 or greater earthquake in the first year of this two-year window, the dictator’s risk of being removed goes up to 30 percent. However, the autocrat’s risk of removal is reduced to 24 percent if the earthquake killed more than 200 people. Earthquakes pose a threat to autocratic leaders when people are forced into refugee camps and can organize against the regime. People dying from an earthquake can’t organize and so they do not endanger a dictator’s survival in office. As might be expected, given these facts and the incentives they suggest, instances of 200 or more people dying in earthquakes is much more common in autocracies than democracies.9

  Not all disasters are equal in the eyes of autocrats. Dictators are particularly wary of natural disasters when they occur in politically and economically important centers. Disaster management in China emphasizes this point. When an earthquake struck the remote province of Qinghai in 2010, the Chinese government’s response was, at best, halfhearted. In contrast, its handling of disaster relief in the wake of a 2008 earthquake in Sichuan won the approval of much of the international community. The differences are stark and driven by politics. The Sichuan quake occurred in an economically and politically important center where a massed protest could potentially threaten the government. Qinghai is remote and of little political importance. Protest there would do little to threaten the government. The government did much less to assist people who could not threaten them.

  Responding to Revolution or Its Threat

  Whether because of an unforeseen earthquake, a succession crisis, or a financial meltdown, the threat of rebellion can rise, striking a leader like a lightning bolt. What then is the right response to such a threat? History teaches us that some crack down hard on rebels; some succumb to them; and some reform on their own. The rules governing politics help us understand how different circumstances lead to different choices among these options.

  Successful rebellions, mass movements, and revolutions are not commonplace, but neither are they extremely rare. Successful rebellions that turn into democracy are pretty rare but they do happen. What characterizes revolutions or revolutionaries who actually do what they promise: create a democracy to try to better the lives of the people? And what characterizes revolutions that don’t take off or revolutionaries who don’t democratize! We start with our old friend, General Than Shwe of Burma.

  The Than Schwe government makes sure that the people of Burma are kept poor, isolated, and ignorant. There is no free press. The people are not allowed to congregate. Few foreigners are allowed in, and those that are, are constantly watched by the police. All these actions are designed to make it hard for the people to coordinate and organize against the government. The people are desperate for change, but the government makes it virtually impossible for them to achieve it. In a telling 2005 account of how unhapp
y the people are, a journalist for the Economist magazine recalls how they were continually asking him how the United States could be prevailed upon to invade: “the prospect of a foreign invasion is a fond hope, not a fear.”10 The people of Burma want to be the next Iraq! With such demand for change, it is little wonder that Shwe is terrified of protest and that he focuses his attention on preventing it.

  Than Shwe, like many others, takes the autocrat’s preferred path to eliminating the threat from mass political movements. He suppresses the people. He doesn’t need to buy them off because Burma is blessed, or cursed, depending upon your point of view, with natural resources. Burma is a huge exporter of natural gas, hardwood, gems, gold, copper, and iron.11 For instance, it is thought to earn about $345 million through the annual export of 1.4–1.6 million cubic meters of hardwood, much of it extremely valuable teak. We use the term “thought to” because it is hard to know the figures for sure. For instance, in 2001, China reported that it imported 514,000 cubic meters of wood from Burma, but Burma only records exports of 3,240 cubic meters. Presumably the income from the unaccounted-for hardwood lines the pockets of the generals, rather than funding the welfare of the people. It certainly does not fund infrastructure. Indeed, the timber industry’s attempts to process its products before export have been nearly completely stymied by the absence of infrastructure. Of course the absence of roads makes it even more difficult for the people to assemble and threaten the government. This became particularly true after 2005, when the government moved the capital to a remote mountain location where few citizens are allowed to visit.

  Burma is also the world’s major producer of jade and rubies. Gem auctions in 2007 are thought to have earned the nation $370 million. Yet Burma’s biggest export earner is natural gas. Currently the offshore natural gas fields generate between $1–1.5 billion. These earnings are likely to increase over the next few years with the development of additional fields and the opening of a pipeline to ship gas directly to China. Little of this money makes its way into the government’s economic accounts. The official exchange rate is 6 kyaks to the dollar. However, the real rate is around two hundred times higher. This means the regime can deposit all gas export earnings in government accounts at the official exchange rate and still keep 99.5 percent of the money for themselves.