VoxEU Column Politics and economics

Financial liberalisation and democracy: The role of reform reversals

Will the current crisis reverse the past two decades of democratisation and financial liberalisation? This column documents the complex, non-linear relationship between political and financial reform. Financial liberalisation often reverses as countries move from autocracy to democracy, as “partial democracies” are less liberalised, and there are big differences between de jure and de facto liberalisation.

Since the early 1990s, both the process of democratisation and the process of financial liberalisation have accelerated globally. A relevant issue is whether the current global financial and economic crisis could provoke a slowdown or even a reversal in the process of financial liberalisation around the world, particularly in emerging economies (Calvo and Loo-Kung, 2009). Previous literature has found that recessions and banking crises have significant adverse impacts on financial reform. Further, financial liberalisation experienced a “great reversal” in the aftermath of the Great Depression of the 1930s (Rajan and Zingales 2003). Given the pressure on politicians to act, an important issue is how different political regimes will respond to the crisis and possibly reverse financial reforms.

In Campos and Coricelli (2009), we derive two results that may be relevant for the current debate on financial liberalisation and restrictions. First, in countries with weak or “partial” democracy, the state can be captured by powerful economic elites that will push for restrictions in the financial sector and reversals of liberalisation policies. Second, de jure measures may translate into sharply different de facto measures, depending on the political regime. Again, irrespective of de jure measures, regimes of so-called “partial democracy” are typically characterised by lower degrees of financial liberalisation, and even more important, by reversals in financial reform. By contrast, both democratic and autocratic regimes tend to favour financial liberalisation.

We could thus conjecture that those countries falling into the category of “partial democracy” are more at risk to undergo significant set-backs in financial liberalisations.1 The reason why partial democracy tends to display the worst outcome in terms of financial liberalisation has probably to do with the notion of “state capture”. In full-fledged democracies, the power of elites is checked by the “voice” of the population. In autocracies, political elites are strong and may withstand pressure from economic elites. In contrast, in partial democracies, governments tend to be weak and the hostage of economic elites. Hence, the relationship between democracy and financial reform may be non-linear, with the lowest level of financial liberalisation tending to occur in intermediate regimes of “partial democracy”. The non-linearity also indicates that during political reforms, as the system travels, say, from autocracy to partial democracy, financial reform is likely to go through important reversals.

When these non-linearities count, the effects of changes in the political regime on financial reforms depend on initial conditions. Cross-country analyses involving countries with highly heterogeneous starting points may be misleading. One way to tackle this problem is by choosing a set of economies that provide a natural experiment. The differences in the level and type of political competition across the 25 Eastern European and former Soviet Union “transition” countries in 1989 were minimal and the same can be said for their financial systems. Yet after the collapse of communism, these countries followed radically different economic and political trajectories (Campos and Coricelli, 2002).

Finance and democracy

What do the data on financial reform and political liberalisation reveal? Financial liberalisation is measured by two indicators – one reflects changes in the quantity of financial services while the other reflects the efficiency with which these are provided. The latter is the preferred measure. The first index is based on liquid liabilities, credit to the private sector, and commercial and central bank assets (all as a share of GDP). The second is based on the ratio of bank overhead costs to total assets and the net interest margin. Democracy is measured by indices of political rights, civil liberties, press freedom, presidential power, and a composite index of political reform. All are normalised to the unit interval and re-scaled so that larger values indicate more reform.

Figure 1 plots the predicted and actual values from regressions of financial reform on linear and quadratic terms of the index of political rights (from the between panel estimator). Clearly, the relationship between financial and political reform is U-shaped. The fit of the quadratic specification is better than that of the linear for all measures of financial and political reform (civil liberties, political rights, two democracy indexes, and press freedom).

Figure 1. Financial reform and democracy

The strength of this non-linearity implies that reform reversals are numerous, large, or both. Yet the paucity of empirical attention to reform reversals seems to suggest otherwise. Reform is defined as the change in levels of the indicators of political and financial liberalisation on a year-to-year basis. A reversal is defined as the situation in which the value of the difference is negative. Out of the 337 country-year cells for which data on the two reforms is available, we identify political reform reversals in 48% of the cases, financial reform reversals in 35% of the cells, and joint political and financial reform (“twin”) reversals in 17% of all possible cases (Figure 2).

Figure 2. Reversals across countries and over time: Political reform reversals, financial reform reversals, and twin reversals

The magnitude of the average reversal is 0.008 for political reform and 0.02 for financial reform (on the 0 to 1 scale), with standard deviations of 0.09 and 0.07, respectively. These figures suggest that reversals are much more common and severe than previously thought and that they play a crucial role in the relationship between financial and political reform.

Reversals, de jure and de facto financial liberalisation

Using the Abiad and Mody (2005) model, we show that banking crises hinder financial reform, while debt crises help it. Lower US interest rates boost domestic financial reform, while recessions and high-inflation show a systematic negative effect on financial reform. The results also suggest that the ideological orientation of the government (whether left- or right-wing) is less important vis-à-vis financial reform than the actual level of political liberalisation. Yet the most important result is that, within the Abiad-Mody model, there is robust evidence for a U-shaped relation between financial and political reform.

One of the central distinctions in the financial liberalisation literature is that between de jure and de facto measures. Kose et al. (2009) argue that many discrepancies on the impacts of financial liberalisation can be traced back to this distinction. De jure measures reflect changes in the legal framework, rules and regulations that affect the financial system, while de facto indices capture the size and actual workings of the financial system.

Beyond analysing the separate effects of de jure and de facto measures, which is standard in the literature, we propose within a production function framework that de jure are inputs for de facto financial reforms. More specifically, we posit that de facto financial liberalisation may be driven by two components – one is changes in laws and regulation (inputs) and the other is changes in the quality of the enforcement of these laws (“technology”). There is still a non-linear relationship between political and financial (de facto) liberalisations even accounting for the role of the legal inputs and the fact that most de jure measures are not good predictors of de facto reform (suggesting that enforcement dominates stroke-of-the-pen changes).

Figure 3. The relationship between de jure and de facto financial liberalisation

Policy implications

These results have at least two important policy implications.

  • First, the transition from autocracy to democracy may not lead to financial liberalisation if a country gets stuck in an intermediate political regime. Such intermediate political regimes are prone to reversals in financial reform as powerful elites can affect policy decisions. International institutions should then be careful when dealing with countries in such intermediate regimes.
  • Second, legal and de jure measures seem to be more ineffective in intermediate regimes. Therefore, the analysis of the link between political and economic reforms should perhaps pay more attention to the enforcement of policies, or the indeed to the link between de jure and de facto reforms.

1 See Epstein et al. (2004) for the theoretical and policy importance of “partial democracies”. Acemoglu and Robinson (2008) focus on “captured democracy”.


Abiad, A. and Mody, A. (2005), “Financial Reform: What Shakes It? What Shapes It?The American Economic Review 95(1): pp. 66-88

Acemoglu, D. and J. Robinson (2008), “Persistence of Power, Elites, and Institutions”, American Economic Review, 98:1, 267-293.

Calvo, Guillermo and Rudy Loo-Kung (2009), “Should we rush to further regulate financial institutions?” VoxEU.org, 29 June.

Campos, N. and F. Coricelli (2002) “Growth in Transition: What We Know, What We Don’t, and What We Should,” Journal of Economic Literature, XL (3): 793-836.

Campos, N. and F. Coricelli (2009), “Financial liberalization and democracy: The role of reform reversals,” CEPR DP 7393

Epstein, D., R. Bates, J. Goldstone, I. Kristensen, S. O’Halloran (2004), “Democratic transitions”, CID Working Paper n. 101, Harvard University.

Kose, A., E Prasad, K. Rogoff and S. Wei (2009), “Financial globalisation and economic policies,” CEPR DP 7117.

Rajan, Raghuram and Luigi Zingales (2003). “The Great Reversals: The Politics of Financial Development in the Twentieth Century.” Journal of Financial Economics, 69(1): 5–50.

630 Reads