Democracy and economic growth

Does the type of regime affect how fast a country grows? One view holds that autocracies can suppress consumption, override vested interests and plan long-term, so they can grow faster; the growth of South Korea and Taiwan under military rule and of China since 1978 is often cited. Another holds that democracies protect property rights, invest in human capital, limit expropriation and correct errors, so they grow faster and more steadily. The question matters because claims that a regime 'succeeded because of authoritarianism' depend on it.

Robert Barro (1996) found at most a weak, non-linear relationship, positive at low levels of democracy and negative at higher levels. Przeworski, Alvarez, Cheibub and Limongi (2000), studying 135 countries from 1950 to 1990, found that political regimes had no impact on the growth of total national income, and that per capita incomes grew faster in democracies because population grew faster under dictatorships. Gerring and colleagues (2005) argued that the stock of democratic experience accumulated over time, rather than the current level, predicts growth.

Daron Acemoglu, Suresh Naidu, Pascual Restrepo and James Robinson (2019), in 'Democracy Does Cause Growth', used a panel of countries from 1960 to 2010, a new consolidated democracy indicator and dynamic models with country fixed effects. Their baseline estimate is that a country moving from non-democracy to democracy has about 20% higher GDP per capita in the following 25 years than one that remains a non-democracy. They found similar results using propensity-score reweighting and regional waves of democratisation as an instrument, effects that did not depend on initial income, and channels through greater investment in capital, schooling and health.

• South Korea and Taiwan grew rapidly under authoritarian governments and kept growing after democratising in the late 1980s.
• China and Vietnam have grown fast under one-party rule since their market reforms, while other long-lived autocracies, such as Mobutu's Zaire or North Korea, stagnated or declined.
• India and Botswana sustained growth over decades as electoral democracies.

Markus Eberhardt argued that the estimates are fragile when a fraction of observations is dropped and that the common-effect and cross-section independence assumptions do not hold; allowing for heterogeneity across countries he finds a positive long-run effect of around 10%, roughly half the original. Meta-analyses differ: Doucouliagos and Ulubaşoğlu (2008), pooling 483 estimates from 84 studies, found no direct effect but significant positive indirect effects through human capital, lower inflation, lower instability and greater economic freedom, while Colagrossi, Rossignoli and Maggioni (2020) analysed some 2,000 regressions. Others note that democratisation episodes are not random, that the effect may differ by region and period, and that average effects say little about individual autocracies that grew fast or democracies that stagnated.

The weight of recent evidence suggests that democracy does not harm growth and probably raises it modestly over the long run, while the size and mechanism of the effect remain disputed. There is no robust evidence for a general 'authoritarian advantage' in growth: fast-growing autocracies exist, but so do many stagnant ones, and autocracies show more variance in outcomes.