Statistical Evidence

This preview highlights the main points of the Statistical Evidence chapter from Pathways to Freedom: Political and Economic Lessons from Democratic Transitions, a publication of the Civil Society, Markets, and Democracy Initiative. For previews of other chapters, please see the Table of Contents.

Also see:
Statistical Evidence: Full Chapter Preview
Further Reading

Overview

Many of the factors thought to influence the trajectory of democratic transitions can be quantified. By looking across time at various factors and their statistical association with countries’ movements toward and away from democracy, it is possible to discover what makes democratic transitions more or less likely to begin, backslide, and endure. Statistical likelihoods are not perfect predictors—even a country with the "right" characteristics may well stumble on the road to democracy. However, these likelihoods reveal what often matters in democratic transitions and can usefully inform policy choices.

Studies on the determinants of democratization can be grouped around major questions such as:

■ Is democracy inevitable after countries reach a certain level of income, growth, or modernization, and how does this affect the chance of sustaining democratic advancements?

■ Are these structural issues irrelevant, and are strategic interactions among elites what really matter?

■ Is democracy forged from below, through a power struggle among social forces with competing economic interests?

Although much remains unknown, the statistical evidence reveals several major findings:

■ Increases in gross domestic product (GDP) per capita do not generally produce democratic improvements on their own. However, when a democratic transition has occurred, richer countries are more likely to sustain it.

■ In the short term, economic growth under autocracy impedes democratization while economic crisis tends to trigger democratic transitions.

■ Nonviolent mass mobilization against autocracies is a strong trigger of democratization. Armed rebellion, by contrast, typically does not lead to democratization.

■ Authoritarian regimes with partial political openness, especially through multiparty elections, are likeliest to grow more democratic.

■ The fate of new democracies does not appear to hinge on the choice of a presidential or parliamentary system.

■ Overall, quantitative studies do well at predicting long-term changes in the level of democracy. Eruptive short-term change, however, is less predictable.

Pathways to Freedom: Political and Economic Lessons From Democratic Transitions, a new book from the Council on Foreign Relations, examines the main quantitative findings in the literature around six areas of economic, political, and social development.

Socioeconomic Exclusion and Inclusion

In theory, transitions to democracy should be most likely in countries with intermediate levels of socioeconomic inequality. In highly unequal autocracies, average citizens have a strong incentive to revolt in order to gain power and redistribute wealth, but elites also have a strong interest in repressing uprisings. In more equal societies, elites have less fear of redistribution but average citizens have less incentive to revolt.

Although examples exist of transitions at intermediate levels of inequality, such as Indonesia in 1999, quantitative evidence has proved weak and contradictory. One possible explanation is poor data on income inequality. Another is that the relationship between inequality and democratization might differ between right-wing autocracies, such as South Africa’s apartheid-era regime, and socialist dictatorships, as in the former Soviet Union and Eastern Europe. In South Africa, income disparities were declining in the 1980s, consistent with the theory that rich white elites had less to lose from redistributing power to blacks. In Poland and Ukraine, however, transitions occurred while inequality was fairly low.

Studies of ethnolinguistic or religious fractionalization also show little effect on democratization. This could again result from low data quality. Data poorly reflect the extent to which cleavages are politicized, which most scholars argue is a crucial factor.

Economic Structure and Policies

The relationship between economic development and democratization is by far the most well-studied one in the quantitative literature. Studies have shown that growth in GDP per capita does not make an autocracy more likely to democratize, presumably because people enjoy their rising fortunes. Increased national income does, however, make democracy more likely to endure once it arises. Questions regarding what aspects of development are most critical to safeguarding democratic achievements—income, urbanization, economic structure, or others—remain largely unresolved.

Economic crises under authoritarianism make democratization more probable by subverting the regime’s hold on power. Some studies have also found a relationship between economic performance and democratic survival, but it appears that economic crisis hurts democracies less than autocracies.

To take one example, quantitative evidence indicates that China’s impressive growth is unlikely to lead to democracy. On the contrary, a serious contraction would threaten China’s one-party regime. However, the prospects that enduring democracy would result from such a crisis improve as China’s prosperity grows.

Another important economic attribute is natural resource wealth. Regimes can use such wealth to deploy both carrots (tax cuts and patronage) and sticks (repression) to hold democratic movements at bay. For example, oil has hindered Nigeria’s democratic prospects. Many quantitative studies have shown oil wealth to decrease the chances of democratization, but the evidence is not overwhelming and some believe flawed statistical techniques might explain the findings.

On the question of state involvement in the economy, the expectation is that less state intervention leads to more democracy. According to this logic, if economic actors depend on state subsidies, then they will be less inclined to oppose an autocratic regime. Some studies claim to show a positive effect of economic liberalization on democratization, but these findings have been difficult to replicate. Finally, on trade, most studies show that more dependence on trade makes countries less likely to democratize. However, the reasons for this are unclear.

Civil Society and Media

Studies find that peaceful demonstrations are powerful drivers of democratization, whereas violent rebellions are not. Violent opposition is often narrowly based, which helps autocracies rally the broader public and legitimizes the use of repression. When a regime uses force to confront peaceful protest, by contrast, moral outrage spurs further opposition. Peaceful protests also spread across borders, as was the case with the "colored revolutions" in the former Soviet Union and the Arab uprisings that began in 2011. To be sure, though, mass protests do not necessarily reflect durable civil societies. And as Syria and Libya show, the possibility of a peaceful uprising hinges on at least a modicum of political openness.

Evidence is more speculative on the media’s impact on democratization, though a large recent study finds that the proliferation of media outlets robustly safeguards democratic advances.

Legal System and Rule of Law

Few quantitative studies examine the relationship between the rule of law, or its absence, and democratization. Building an effective legal framework and functioning court system is a direct way of strengthening democracy, since these institutions can uphold rights and liberties critical to the democratic process. Beyond this, though, the statistical evidence is unclear on whether boosting administrative capacity, combating corruption, bolstering property rights, and improving courts exert an independent influence on democracy.

Government Structure

Recent studies show that authoritarian regimes with the trappings of a democratic system—especially multiparty elections—are the likeliest to democratize, even if elections are flawed or fraudulent. This is because elections promote both divisions within the ruling party and the unification of opposition forces. These processes reinforce each other and make democratic breakthroughs more probable.

Mexico, whose authoritarian regime maintained multiparty competition, illustrates these patterns. In a tug-of-war with the government, the opposition secured incremental reforms to the framework governing elections, eventually enabling it to oust the ruling party in the 2000 presidential vote.

As for institutional design, the conventional argument is that presidential democracies are less stable than parliamentary ones because the executive rests on a narrow support base and tends to be caught in deadlock with the legislature, producing stalemate and crisis. Empirical scrutiny, however, proves this inaccurate. Presidential democracies are more prone to breakdown, but this is because of the legacy of military dictatorships that often precede presidential systems. Absent a recent history of military interventions, presidential democracies are no weaker than the parliamentary variety.

Education and Demography

Some studies argue that higher societal levels of education positively influence democratization, but these studies generally ignore the fact that rising education is part of a modernization process that includes greater income, urbanization, and other elements. When these are controlled for, education’s effect weakens considerably.

Many studies have also found that a predominantly Muslim population impedes democratization. This so-called Muslim gap is, however, mostly an Arab gap, suggesting that any apparent religious correlations with democracy are spurious. Indeed, Muslims around the world express no weaker democratic sentiments than people of other religions. Religious obstacles to democratization have been cited before, only to be disproved by history.

Conclusion

Studies show that many factors affect the likelihood that countries will become and remain democracies. Quantitative evidence has been relatively successful in explaining long-run democratic trajectories, but it cannot predict revolutions or coups in the short term, and it is impossible to know when or how particular reforms might happen.

Three priorities for policy emerge from quantitative studies. The first is increasing the capacity for nonviolent protests against autocratic regimes. The second is support for multiparty elections. Finally, maximizing access to mass media is a sound strategy to safeguard democratic progress already achieved. What does not seem to promote democracy is economic development in authoritarian states. This suggests that, at least in the short run, promoting democracy and prosperity may be incompatible.

Additional Material

Statistical Evidence: Full Chapter Preview
Further Reading