Moreover, there is no reason to believe that democracies invest less than dictatorships, even in poor countries. Hence, the very claim that in poor countries democratic competition leads governments to opt for higher rates of consumption finds no support in empirical evidence. Investment rates are simply low in poor countries. Poverty is so tightly constraining that there is no room for political regimes to make a difference.

Very surprisingly, however, political regimes affect the population’s growth rate, which is higher in dictatorships at every income level. This difference is not due to mortality rates. In fact, mortality rates are higher and life expectancies lower under authoritarian regimes. Neither is it due to the age structures of the population. The main difference is that fertility is higher under dictatorships. Very poor countries have fertility rates of about six births per woman regardless of their regime, while very wealthy countries converge to replacement fertility rates under both regimes. But within the entire intermediate range, from per capita income of US$1,000 to $12,000, fertility is higher under dictatorships. When regimes are matched for social and economic conditions, statistics reveal that an average woman has one-half of a child more under dictatorship than under democracy.

Why this is true remains unclear, but the consequence is that per capita, as distinct from total, income grows faster under democracies. The observed rate of annual growth of per capita income was 1.84 percent under dictatorships and 2.26 percent under democracies. Correcting for exogenous conditions, we still conclude that per capita incomes grew at the annual rate of 1.93 percent under dictatorships and at the rate of 2.11 percent under democracies. This is not a large difference, but nonetheless favors democracy.

Looking at the longer run, data generates the same conclusions. Poverty can trap societies in its grip. Of the 100 countries with per capita incomes of less than US$2,000 in 1950 or when they became independent, 56 remained equally poor or even poorer decades later. Yet the bonds of poverty are not inexorable. Some countries, notably Taiwan, South Korea, Japan, Singapore, Portugal, Malta, Ireland, Spain, Thailand, Greece, Malaysia, and Botswana, developed spectacularly, at least quintupling their per capita incomes. Of this list, Japan, Ireland, and Malta were parliamentary democracies during the entire period. Greece was a parliamentary democracy before and after a period of military rule. Botswana had more than one party with reasonably free elections in which the same party always won an overwhelming majority. Singapore and Malaysia had authoritarian regimes during the entire period. Portugal, Spain, South Korea, and Taiwan proceeded from various forms of dictatorship to different forms of democracy. Finally, Thailand has been so politically unstable that its history cannot be summarized. Looking at this list, it is clear that while such spectacular successes are rare, there is nothing to indicate that it takes one regime or the other to generate them. Indeed, contrary to conventional wisdom lauding the success of quasi-authoritarian East Asian states, the Asian “tigers” have been democratic as well as authoritarian, politically stable as well as highly unstable.

The list of economic disasters, by contrast, is much longer. A total of 19 countries that were independent before 1990 had lower incomes at the end than at the beginning of the period. Among them, Kiribati and Papua New Guinea were parliamentary democracies throughout. Five countries remained under different dictatorships during the entire period. Seven countries started as authoritarian and ended as democracies, but the transition typically occurred too late to impact observed growth patterns. Somalia began as a democracy and disintegrated under military rule. Finally, Suriname had a convoluted political history. Hence, patterns are again hard to find. But almost all these countries were ruled by dictatorship during most of their histories, having often experienced periods of civil strife and a lack of political institutions.

While these findings are not very satisfactory from an intellectual point of view—we still do not know which political institutions matter for economic development—they put to rest any notion of a trade-off between democracy and development. While some of the successful economic tigers have been dictatorships, dictatorships are no tigers. When one looks at the average performance of the two types of regimes, it is clear that, if anything, democracies generate a somewhat higher rate of growth of per capita income, lower mortality, and lower fertility.

Altogether, these findings add up to a bleak picture of dictatorships. While democracies are far from perfect, lives under dictatorships are grim and short. Because they rule by force, dictatorships are highly vulnerable to any visible signs of dissent. Economic growth under dictatorships is highly sensitive to any visible signs of political mobilization and to turnover of heads of government, while under democracy the same phenomena are an expected part of life and have no effect on growth rates. Since policies in dictatorships depend on the will or whim of a ruler, they exhibit much higher variance in economic performance: some generate miracles, some disasters, and many generate both. In the end, then, democracy does make a difference, not only for political liberty but also for material well-being.

Explaining the Correlation

There is no doubt that democratic regimes are more frequent in the more developed countries, while dictatorships predominate in poor ones; Lipset’s original observation holds 40 years later. But the inference that development makes it more likely that a country will become democratic is fallacious.

There are two distinct reasons that a relation between per capita income and the frequency of democracies may be observed. One is that countries with higher incomes are more likely to become democratic; this is the hypothesis of the modernization theory. But the second possibility is that if democracy emerges—for whatever reason—in a wealthy country, it is more likely to survive than if it appears in a poor country. In fact, Lipset thought that it was the second mechanism that explained the observed patterns. His central hypothesis was, as he wrote in his famous 1959 essay “Some Social Requisites of Democracy,” the “more well-to-do a nation, the greater the chance it will sustain democracy.”