Cross-country analysis often shows only a weakly positive correlation between democracy and economic growth. The correlation is weak because there are some authoritarian countries that have strong growth figures. Most notably China of course. The impressive growth rates of a few oppressive regimes has successfully undermined the once popular theory about democracy’s positive effect on growth, and has even fostered the opposite belief: that authoritarian government is necessary for growth. (The story goes somewhat like this: authoritarian rule means longterm planning, discipline in production and consumption, national harmony and popular respect for often difficult decisions, which in turn means efficiency and productivity, and hence growth).
However, those non-democratic countries that do indeed show high growth rates shouldn’t be viewed as typical: there are just as many authoritarian countries with very weak growth figures. It’s a bit silly therefore to derive a general law about authoritarian economic success when that supposed law can be so easily falsified.
The success of China and a few other authoritarian countries doesn’t warrant a general conclusion about the beneficial effects of autocracy on growth. Bill Easterly in his “Tyranny of Experts” has argued that the prosperity of successful autocracies may not be due to a lack of freedom. Most of those countries experienced a recent move towards relatively more freedom and democracy. It was only after China started to soften its horrific totalitarian rule that prosperity began to rise. It’s not crazy, therefore, to assume that a more rapid liberalization would have resulted in even higher growth rates. Furthermore, most autocracies start from nowhere. It’s relatively easy to produce good growth figures when baseline prosperity is very low, as was China’s some decades ago (not in the least because of authoritarian rule). It’s relatively easy, even – one is tempted to say – for fools and autocrats.
The low baseline from which most autocracies start shows up when we compare not the growth rates but the level of GDP between countries. The correlation between democracy and GDP is stronger when we look at the level rather than the growth of GDP. Richer countries (with the exception of most wealthy Muslim countries) tend to be or become democracies. The data linked to in the previous sentence plot income in 1971 against democracy scores in the following decades, you can see that the causation seems to go from income to democracy. A high level of GDP predicts a flourishing (or at least continuation) of democracy. However, this could again be used by the authoritarian growth crowd. They can use this to argue that poor countries need autocracy in order to kick-start growth, because democracy can only come when the level of GDP is sufficiently high (the “democracy as luxury” argument). It’s probably true that prosperity fosters democracy (for the obvious reasons: democracy requires money, leisure, education etc.). But it’s good to see some evidence of causation in the opposite direction, from democracy to growth – if only to undermine self-congratulatory autocrats. For example, here‘s a study plotting current income against older democracy scores, suggesting that democracy also promotes growth. And here are some more correlations between the levels of GDP and of democracy. (More here).
If we accept that there is indeed a causal effect of democracy on the level of GDP, how exactly does that effect occur? Perhaps transparency, the rule of law, accountability, property rights and other characteristics of democracy are good for growth.
If we want further evidence of a causal effect of democracy on growth, we can do an in-country analysis. This paper examines the effect of democratic transitions on economic growth. The encouraging conclusion is that countries which have experienced a transition to democracy experience higher average growth after the transition.
The paper plots the evolution of real per capita GDP growth in the years surrounding a successful democratization (the year of the democratization being T), compared to the global growth rates in each year. It shows that the transition itself may imply economic costs, but in the longer term democracy pays off.
I should also mention a recent paper by Acemoglu et al that points in the same direction.