Are Regime Changes Always Bad Economics? Evidence from Daily Financial Data

• In Event: Poster Session: Political Economy

Abstract

The image of a coup or popular uprising ousting an autocratic ruler and installing a democratic government is an evocative one. However, both recent events and prior research suggest that coups rarely lead to democracy. In the wake of the Arab Spring—in which rulers were deposed in four countries—only Tunisia ultimately transitioned to democracy. These results are in line with prior research that claims that in autocracies coups rarely lead to democratization, and in democracies they cause instability (Derpanopoulos, Frantz, Geddes, & Wright 2016; Powell & Thyne 2011; Thyne & Powell 2016; Varol 2011).

Of course, coups and other regime changes not only impact a country’s political environment, but its economy as well. For instance, investors, multinational firms, development agencies, and aid organizations rank political risk as a top consideration when making investment decisions in emerging markets. Common rationales are that unstable governments act myopically and adopt inefficient policies. Unstable governments have, for example, been argued to be less likely to invest in the legal system and the protection of property rights (Svensson 1998), and to be more likely to increases taxes (Devereux & Wen 1998).

These theories are supported by empirical research that suggests political instability—as measured by regime change frequency, political violence, or indices of uncertainty derived from news reports—is negatively correlated with investment, financial development, and GDP growth in cross country regressions (Aisen & Veiga 2013; Alesina, Ozler, Roubini, & Swagel 1996; Alesina & Perotti 1996; Baker, Bloom, & Davis 2016; Fosu 1992; Jong-A-Pin 2009; Roe & Siegel 2011) and that the variance of stock market returns increases in response to economic policy uncertainty (Jensen & Schmith 2005; Leblang & Mukherjee 2005; Liu & Zhang 2015).

In this study, we examine the impact of “irregular” regime changes and public protests on economic performance and remain agnostic about the direction of the effect. In contrast to the aforementioned cross-country studies, we conduct our analysis using daily financial data and provide separate estimates by source of instability. Market expectations in the form of stock market returns are used as an indicator of whether the market views different types of potentially destabilizing political events as “good” for economic growth. The advantage of this event study approach is that it allows us to use time-series variation in stock returns to estimate treatment effects and mitigate the endogeneity problems that are nearly ubiquitous in the cross-country regression approaches previously used to assess economic response to political instability.

We analyze 13 successful coups, 8 assassinations, 15 forced resignations, and 11 public protests. We recognize that this does not represent a full list of historical coups, assassinations, resignations, and public protests, but are restricted to the sample of events for which national-level daily financial data is available. We also examine the failed 2002 Venezuelan coup, as it provides a case that allows us to estimate the effects of both a seemingly successful pro-business coup, and the immediate reinstatement of a left-wing populist. Our study therefore not only uses an event-study approach to enhance the reliability of our estimates, but simultaneously examines a relatively large sample that enhances external validity and allows us to show that different types of political events have disparate economic effects.

We find that coups, assassinations, resignations, and public protests cause large increases in financial volatility. However, the magnitude and direction of the effects differ by source of instability. On average, abnormal returns following resignations are large and positive (4%) while abnormal returns following assassinations are negative and smaller in magnitude (2%). The average impact of coups is negative (2%); however, stock prices increased dramatically (10%) during the failed overthrow of Hugo Chavez’s socialist regime and decreased by approximately the same magnitude following his return to leadership. The effect of public protests on mean stock returns is uncertain, with some protests such as those during the Egyptian Revolution depressing them while others have the opposite effect.

Authors

• Devin Incerti, Princeton University

• Trevor Incerti, Yale University