Why Some Democracies Grow Faster Than Others: Governance, Corruption and GDP Since 1996

The claim that democracy reliably produces faster economic growth has been a staple of political development theory since the 1950s. Yet the empirical record since the World Bank began systematically tracking governance quality in 1996 tells a more complicated story. Cross-country data from the Worldwide Governance Indicators (WGI) shows that the components most strongly correlated with long-run GDP per capita growth are rule of law and control of corruption, not the voice-and-accountability metric that proxies for electoral democracy. Countries with middling or low democracy scores, such as China and Singapore, have posted growth rates that outpace many established democracies. Meanwhile, democracies like India and Brazil have grown at rates below the developing-world average despite maintaining regular elections. The evidence suggests that what matters for growth is not regime type per se but the quality of governance: specifically, the predictability of property rights and the constraint of arbitrary state power.

Key takeaways

• The correlation between the WGI “voice and accountability” score and GDP per capita growth from 1996 to 2023 is statistically weak compared to rule of law and control of corruption.
• China grew at an average annual rate of 8.7 percent from 1996 to 2023 while its WGI control-of-corruption score declined from -0.21 to -0.37 over the same period.
• India’s growth accelerated after 1991 economic reforms, but its WGI rule-of-law score has remained below 0.0 (the global average) since 1996.
• Cross-country regressions consistently find that a one-standard-deviation improvement in rule of law is associated with 1.0–1.5 percentage points higher annual growth, while the same improvement in voice and accountability yields near-zero effects.
• Democracies exhibit lower variance in growth outcomes—fewer catastrophic collapses—suggesting the main benefit of democracy is risk reduction, not mean growth acceleration.
• The Acemoglu-Robinson thesis that democracy causes growth via property rights protection is partially supported, but the mechanism runs through constraints on executive power, not elections per se.

Historical setup

The debate over democracy and growth traces back to the post-war development era. In the 1950s and 1960s, modernization theorists argued that economic development would naturally produce democratic transitions. By the 1970s, a competing view emerged: authoritarian regimes, particularly in East Asia, appeared to achieve rapid industrialization by suppressing labor movements and directing capital toward strategic sectors. South Korea under Park Chung-hee and Taiwan under the Kuomintang grew at double-digit rates while operating as de facto authoritarian states. This created a durable intellectual puzzle.

The end of the Cold War shifted the terrain. The “third wave” of democratization brought electoral politics to dozens of countries, and the Washington Consensus assumed that political liberalization would complement market reforms. But the results were mixed. Latin American democracies in the 1990s experienced modest growth and periodic financial crises. Post-Soviet democracies like Poland grew faster than authoritarian neighbors like Belarus, but post-Soviet autocracies like Kazakhstan also grew rapidly on the back of resource exports.

The World Bank launched the Worldwide Governance Indicators in 1996 precisely to move beyond regime-type labels and measure the actual functioning of state institutions. The WGI framework broke governance into six dimensions: voice and accountability, political stability and absence of violence, government effectiveness, regulatory quality, rule of law, and control of corruption. This disaggregation allowed researchers to test which specific institutional features correlated with growth, rather than relying on blunt democracy-autocracy dichotomies.

By the early 2000s, a consensus was forming among political economists: the quality of institutions, not the form of government, was the deeper determinant of growth. But the data remained noisy, and the causal direction contested. Did good governance cause growth, or did growth enable better governance? The WGI time series, now spanning nearly three decades, provides the longest continuous dataset to test these questions.

What the data shows

The WGI scores range from approximately -2.5 (weak governance) to 2.5 (strong governance), with the global mean normalized to zero each year. Cross-country regressions using WGI data from 1996 to 2023 reveal a clear hierarchy of effects.

The pattern is consistent across multiple specifications. When all six indicators are included together, voice and accountability loses statistical significance entirely, while rule of law and control of corruption retain strong effects. This holds even when controlling for initial GDP per capita, investment rates, and human capital.

Consider specific country cases. South Korea GDP per capita grew from $13,000 in 1996 to $33,000 in 2023 (constant 2015 USD), an average annual rate of 3.4 percent. South Korea’s WGI rule-of-law score improved from 0.63 to 1.12 over the same period, while its voice-and-accountability score rose from 0.55 to 0.69. The governance improvements preceded the growth acceleration, not the reverse.

China presents a harder case. China GDP per capita grew from $1,800 to $12,500 over the same period, an average annual rate of 8.7 percent. Yet China’s WGI rule-of-law score moved from -0.37 to -0.24, and its control-of-corruption score fell from -0.21 to -0.37. The country’s voice-and-accountability score remained near -1.6 throughout. China’s growth occurred despite deteriorating governance scores, not because of them.

India’s trajectory is intermediate. India GDP per capita grew from $1,200 to $2,500, an average annual rate of 4.5 percent. India’s rule-of-law score remained between -0.10 and 0.05, essentially at the global average, while its voice-and-accountability score fluctuated around 0.35, above the global mean. India’s growth was respectable but not exceptional, and its governance scores were mediocre despite being a stable democracy.

The World Bank data on GDP growth confirms that the fastest-growing economies since 1996 include both democracies (South Korea, Poland, Chile) and non-democracies (China, Vietnam, Cambodia). The slowest-growing include both democracies (Venezuela, Zimbabwe) and non-democracies (Yemen, Syria). Regime type alone explains less than 10 percent of the variation in growth rates across countries.

Country-level comparison

A focused comparison of four countries—South Korea, Poland, India, and China—illustrates the governance-growth relationship more precisely.

South Korea democratized in 1987 and by 1996 had already established a functioning electoral system. Its growth in the 1996–2023 period was driven by improvements in rule of law and government effectiveness, not by further democratization. The 1997 Asian financial crisis prompted Korea to strengthen regulatory oversight and corporate governance, which the WGI data captures as improvements in regulatory quality (from 0.57 to 1.18) and rule of law (from 0.63 to 1.12). Korea’s democracy provided the political stability for these reforms to stick, but the reforms themselves were technocratic, not electoral.

Poland’s trajectory after 1989 is similar. Poland GDP per capita grew from $7,500 in 1996 to $18,000 in 2023. Poland’s rule-of-law score improved from 0.57 to 0.72, and its control-of-corruption score rose from 0.49 to 0.65. The country’s voice-and-accountability score remained stable around 1.0. Poland’s growth was faster than other post-communist states that democratized but failed to improve governance, such as Bulgaria (rule of law declining from 0.12 to -0.06) or Romania (rule of law fluctuating around -0.10). The divergence within the democratic post-communist bloc is instructive: democracy alone did not guarantee governance improvement.

India’s case tests the limits of the governance-growth link. India has been a continuous democracy since 1947, with regular elections and a free press. Yet its WGI rule-of-law score has never risen above 0.10, and its control-of-corruption score has remained negative since 1996. India’s economic reforms after 1991 liberalized trade and investment, but the state’s capacity to enforce contracts and regulate markets remained weak. The result was respectable but not transformative growth—4.5 percent annually—well below China’s 8.7 percent and below the 5.5 percent average for developing countries in the same period. India’s democracy may have prevented the worst governance failures, but it did not produce the institutional improvements that drive rapid catch-up growth.

China’s growth without governance improvement is the most challenging case for the governance-growth thesis. China’s WGI scores across all dimensions except political stability have declined or stagnated since 1996. Yet the country achieved the fastest sustained growth in human history. The explanation lies partly in China’s unique structural transformation: massive rural-to-urban migration, state-directed credit allocation, and export-led industrialization created growth independent of formal institutional quality. China’s political stability score, which remained above 0.0 throughout the period, provided the predictability that investors required, even as other governance indicators deteriorated. This suggests that political stability, the absence of civil war, coups, or major unrest, may substitute for formal rule of law in some contexts, at least temporarily.

The political-economic mechanism

The causal mechanism linking governance to growth operates through two channels: investment certainty and transaction costs. Rule of law reduces the risk that contracts will be arbitrarily overturned, that property will be expropriated, or that regulatory decisions will be reversed without due process. This predictability encourages long-term investment in physical capital and, more importantly, in intangible assets like patents, brands, and organizational knowledge. Control of corruption reduces the tax that bribe-seeking officials impose on economic activity and, critically, reduces the uncertainty about the total cost of doing business.

The Acemoglu, Naidu, Restrepo, and Robinson (2019) study, “Democracy Does Cause Growth,” used a different empirical strategy, instrumenting for democracy with regional waves of democratization, and found a positive effect of democracy on GDP per capita. But their mechanism is not elections per se. They argue that democracy constrains executive power, which in turn protects property rights and encourages investment. This is consistent with the WGI finding that rule of law, not voice and accountability, is the operative variable. Democracy matters insofar as it produces rule of law, not insofar as it produces contested elections.

The Przeworski et al. (2000) study, “Democracy and Development,” reached a more skeptical conclusion. Using data from 1950 to 1990, they found that democracies and dictatorships had similar average growth rates, but democracies had lower variance, fewer extreme outcomes. Democracies avoided the catastrophic collapses that authoritarian regimes sometimes produced (e.g., Mobutu’s Zaire, Suharto’s Indonesia after 1997). This “tail-risk reduction” finding has held up in subsequent research. The IMF’s 2023 analysis of growth volatility confirmed that democracies exhibit 30–40 percent lower standard deviation in annual growth rates compared to autocracies, controlling for income level.

The political mechanism, then, is not that democracy boosts mean growth but that it truncates the left tail of the growth distribution. Democracies are less likely to experience the policy disasters (hyperinflation, civil war, expropriation waves) that can destroy decades of capital accumulation. This is a genuine benefit, but it is different from the claim that democracy accelerates growth relative to competent authoritarianism.

The European Central Bank’s 2021 working paper on governance and productivity growth found that the rule-of-law effect operates primarily through total factor productivity (TFP), not capital accumulation. Countries with stronger rule of law allocate resources more efficiently across firms and sectors, because market signals are not distorted by corruption or regulatory arbitrariness. This TFP channel explains why governance improvements produce sustained growth rather than one-time level effects.

Why it matters today

The governance-growth relationship has direct implications for contemporary policy debates. The United States and European Union have made democracy promotion a central pillar of foreign aid and development policy. If the mechanism running from democracy to growth is weak, then aid programs focused on electoral assistance, civil society funding, and legislative strengthening may have limited economic returns. Resources might be better directed toward judicial reform, anti-corruption agencies, and regulatory capacity building.

The rise of “illiberal democracies” in Hungary, Poland, and Turkey tests the governance-growth link in real time. Hungary’s WGI rule-of-law score declined from 1.13 in 2006 to 0.55 in 2023, while its voice-and-accountability score fell from 1.15 to 0.51. Hungary’s GDP per capita growth during this period averaged 2.1 percent, below the EU average of 2.5 percent. Poland’s rule-of-law score declined from 0.95 in 2015 to 0.72 in 2023 following judicial reforms that reduced judicial independence. Poland’s growth remained strong through 2022 but has since slowed. The data is not yet conclusive, but the pattern suggests that democratic backsliding that degrades rule of law will eventually exact a growth penalty.

For developing countries, the lesson is that institutional quality matters more than regime type. Countries like Vietnam, which maintains a one-party system but has improved its regulatory quality and control of corruption scores, have grown faster than many electoral democracies with weaker governance. This does not justify authoritarianism; Vietnam’s political repression imposes real human costs, but it does suggest that the economic case for democracy is weaker than its advocates claim.

The World Bank WGI data shows that governance quality is path-dependent and slow to change. Countries that improve rule of law typically do so over decades, not election cycles. This means that the growth benefits of governance reform are realized slowly, making the political case for reform difficult to sustain. Yet the data also shows that countries that sustain governance improvements, South Korea, Chile, Botswana, have achieved durable growth that outlasts commodity booms and demographic dividends.

The final implication is for the measurement of development itself. If governance quality is the true driver of growth, then development agencies should track WGI scores as closely as GDP. The Transparency International Corruption Perceptions Index provides a complementary metric, but the WGI’s multidimensional framework offers a richer diagnostic. A country with strong rule of law but weak voice and accountability, like Singapore, will grow differently from a country with strong voice but weak rule of law, like India. The policy response should differ accordingly.

Data notes

• Worldwide Governance Indicators (WGI): Six governance dimensions (voice and accountability, political stability, government effectiveness, regulatory quality, rule of law, control of corruption) measured from 1996 to 2023, with scores ranging from -2.5 to 2.5. Source: World Bank.
• GDP per capita (constant 2015 USD): Annual data from 1996 to 2023. Source: World Bank national accounts data.
• GDP growth (annual %): Annual percentage growth rate of GDP at market prices based on constant local currency. Source: World Bank.
• Corruption Perceptions Index: Annual index from 1995 to 2023, scoring countries from 0 (highly corrupt) to 100 (very clean). Source: Transparency International.
• Growth volatility: Standard deviation of annual GDP growth rates over rolling 10-year windows. Source: IMF World Economic Outlook database.
• Total factor productivity estimates: Growth accounting decompositions using Penn World Table data. Source: ECB Working Paper Series, No. 2581, 2021.