Income, democracy, and growth: Broadening the perspective

The long‑running debate on the income–democracy nexus contrasts two opposing views on the direction of causality. According to the modernisation view (Lipset 1959), higher levels of economic development beget democracy. In view of evidence that income increases often do not entail more democratic institutions once country and period effects are absorbed, the institutions view argues instead for causality running from democracy to income growth (Acemoglu et al. 2008). From a theoretical perspective, democratisation is driven less by income levels per se but rather by economic conditions that shape group interests, such as inequality, the importance of human capital, and the returns to broad‑based public policies. Correspondingly, the predicted effect of democratisation on growth is theoretically ambiguous: extending the franchise can heighten redistributive pressures but can also foster more inclusive, growth‑friendly environments. Recent empirical work, however, finds that democracy tends to raise income growth on average (Papaioannou and Siourounis 2008, Acemoglu et al. 2019).

A unified perspective on development and democratisation

In a new paper (Cervellati et al. 2025), we conjecture that both empirical findings can be rationalised by considering development dynamics that are not restricted to a balanced-growth path. Our theoretical analysis shows that the bi‑directional relationship between income and democracy is shaped by the phases of economic development, especially by the onset of sustained human capital accumulation and the associated fertility decline within the demographic transition. We extend a standard unified growth framework that accounts for the shift from stagnation to sustained growth (e.g. Galor and Weil 2000, Cervellati et al. 2023) to include institutional dynamics in the spirit of political economy models of democratisation (Acemoglu and Robinson 2006).

The core prediction, overlooked so far, is that the start of human capital accumulation affects both the likelihood of democratisation and the growth impact of democracy. Intuitively, as the importance of human capital rises and returns to natural resources decline, conflicts of interest across groups in society diminish. Elites have weaker incentives to resist enfranchisement, including by repression, and their interests shift towards democracy. This strengthens the link from development to democracy.

On the growth side, once economies rely less on natural resources and more on human capital, the benefits from inclusive institutions reflected in less repression and more democracy, better public good provision (notably schooling), and wider policy accountability are amplified. While the theoretical growth effect of democracy is ambiguous in general, it becomes systematically more positive after the transition to sustained human capital-led growth.

Two empirical checks

Based on these theoretical predictions, we revisit two canonical empirical exercises.

First, we replicate the result that income per capita, on its own, does not predict democratisation in panel regressions with country and period fixed effects using five-year panels from 1960-2010 (Acemoglu et al. 2008). We then show that an indicator for the demographic transition, proxied by the onset of a sustained fertility decline (Reher, 2004), is a significant positive predictor of democratisation (see Figure 1). Moreover, the partial effect of income turns positive only after the demographic transition. Development appears to matter for democracy not because higher income directly yields better institutions, but because the economic structure changes along the development path in ways that raise the cost of autocracy for elites.

Second, we replicate the finding that democracy is associated with higher subsequent income growth on average using annual panel data (Acemoglu et al. 2019). Extending this model to incorporate the broader development perspective, we find that the growth dividend from democracy is larger once countries have passed the demographic transition, precisely when ending repression and improving broad-based policies are most productive. This helps explain why democratisation does not always deliver the same growth payoff: timing along the development path matters.

Figure 1 The demographic transition and democracy

Note: Local projections.
Source: Cervellati et al. (2025).

Reframing stylised facts

Our results reframe several stylised facts. The null effect of income on democracy in within-country fixed-effects models is not a contradiction to the modernisation hypothesis; it points to an incomplete conditioning set in which the transition from stagnation to growth is the relevant state variable. Likewise, uneven growth payoffs from democracy across countries and decades may reflect whether democratisation occurred before or after human capital expansion. This view fits the historical experience of many early democracies in Western Europe, which grew slowly until the late 19th century when education expanded, and it also aligns with later accelerations in countries that democratised after sustained fertility declines and rapid schooling gains.

Policy implications

From a policy perspective, the evidence suggests that coupling institutional change with measures that speed up the demographic transition and the structural transformation of the economy can improve both the odds of democratisation and its economic payoff. Investments that broaden access to quality schooling, basic health, and family planning raise the cost of repression for elites and increase the social return to inclusive institutions. Where democratisation precedes the transition, complementary policies that accelerate human capital accumulation are likely to enhance the growth dividend of democracy.

Source: cepr.org