Financial crises: the search for early warning indicators
Financial crises are a recurrent phenomenon in the history of emerging markets and advanced economies alike. To understand the common causes of these crises and to prevent future ones from developing, economists have a long tradition of studying early warning indicators. Two well-documented predictors of financial crises are credit booms and capital flow bonanzas. Indeed, many banking and current-account crises (sudden stops) have been preceded by unusual expansion of domestic and/or external credit (see, for example, Reinhart and Reinhart 2008, Forbes and Warncock 2012, Schularick and Taylor 2012, Mendoza and Terrones 2012). To the best of our knowledge, however, not much research has focused on early warning indicators outside the realm of economic variables. Generally, there is still limited knowledge on the political economy of financial crises (exceptions include the recent contributions by Fernandez-Villaverde et al. 2013, McCarthy et al. 2013, and Mian et al. 2010 and 2012).
This column, which is based on Herrera et al. (2014), studies the role of political frictions in the run-up to financial crises. We show that political variables can be helpful crisis predictors – in particular the level and change in government popularity pre-crisis. One innovation of our approach is that we propose a new cross-country measure of government popularity, which covers more than 100 countries as far back as 1984 and builds on the International Country Risk Guide (ICRG) database. Specifically, we focus on the ICRG sub-indicator of ‘government stability’, which measures “the government’s ability to carry out its declared program(s), and its ability to stay in office”. We show that this index commoves closely with actual public opinion data for those (few) countries and episodes for which actual government approval data is available (e.g. from Gallup).
Our finding: political booms predict crises
Our main finding is that ‘political booms’, defined as an increase in government popularity, precede financial crises in emerging markets. The public’s opinion of the current government improves significantly in the run-up to a country’s financial crash. Interestingly, however, political booms only predict financial crises in emerging markets, not in advanced economies.
Figure 1 shows the cumulative percentage change of our measure of government popularity in the five years prior to the start of a severe crisis, as defined by Reinhart and Rogoff (2009). A stark difference between advanced and emerging economies is apparent. Government popularity increases substantially prior to severe crises in emerging economies, including all countries that went through the Asian crisis, but also prior to the severe crises in Russia and Argentina. The opposite holds for crises in advanced economies, but to a lesser extent.
Figure 1. Cumulative change in government popularity (5 years pre-crisis)
Figure 2 shows for emerging economies that popularity increases by more than 50% in the five-year interval before severe crises. Moreover, the 90% confidence bands are rather narrow, indicating that the dynamics are similar for episodes across emerging economies. Figure 3 shows the opposite trend in advanced economies, but with less economic and statistical significance.
Figure 2. Emerging economies: Government popularity surrounding severe crises
Figure 3. Advanced economies: Government popularity surrounding severe crises
We confirm these results in a regression analysis covering a broader sample of banking crises and sudden-stop episodes back to the 1980s and controlling for economic fundamentals. This finding is not only statistically, but also economically significant – a one standard deviation increase in popularity in emerging markets roughly doubles the probability of a banking crisis.
Moreover, we show that political booms predict future financial crises above and beyond credit booms in emerging markets. The best model fit is achieved when including both credit booms and political booms. The two predictors jointly have much higher explanatory power than separately – an emerging market which experienced a political and credit boom at the same time is much more likely to suffer a crisis than a country experiencing only one of them.
Our explanation: regulation is politically costly
To rationalise these new stylised facts, we develop a model of reputation-concerned governments. The model tries to capture the idea that it may be politically costly to control a boom with regulation. We argue that governments have more and better information about the fundamentals in the economy than the public in general. In particular, governments know more about how sound an economic boom their country may be experiencing is. If the economic boom shows symptoms of exhaustion (for example because it is driven by speculation more than by productive investment opportunities, as highlighted by Gorton and Ordoñez 2014), then the ideal course of action for the government is to regulate the boom, applying corrective measures that prevent or at least reduce the chance of a future crash.
Corrective policies can be politically costly for a variety of reasons. For instance, regulation may reveal to the public that the economic boom they are experiencing, and which the government is taking credit for, is not sound. More specifically, we refer to ‘bad booms’ as booms that are supported by unsound fundamentals and hence, having a high enough chance to end in a crash, ideally should be regulated. If governments of poor ability/competence are more likely to generate bad booms, then enacting regulatory actions is politically costly, as it becomes the tell-tale sign of poor quality of governance. This political reputation concern is a force that may prevent governments from enacting optimal regulatory measures and avoiding crises. A government concerned about its popularity might thus prefer to ride economic booms, even if that means a larger risk of a subsequent crash.
In line with this, we document a negative correlation between financial regulation and government popularity. Especially in emerging markets, regulatory tightening is associated with a decline in government popularity. Moreover, we show that, in emerging markets, most crises were preceded by regulatory loosening (not tightening), suggesting that loose regulation was one reason behind past crashes in these countries.
Emerging markets vs. advanced economies
Why do political booms precede financial crises only in emerging markets? According to our model, governments with lower initial levels of popularity have more incentives to ride political booms simply because there is more margin for improvement. Moreover, if there is high uncertainty about the government’s quality to begin with, then riding a boom also has more potential to change public opinion. We argue that both of these preconditions are more likely to be observed in young democracies (including many emerging markets) rather than established ones. We then provide evidence in support of this reputational channel.
The data show that government popularity in emerging markets is indeed significantly lower than in advanced economies. The data also show that popularity is more volatile in emerging countries, meaning that the public is more uncertain about the quality of its politicians. According to the model, this implies that popularity is more responsive to the perceived economic environment and to governments’ actions and policies. Indeed, we show that governments with lower initial popularity levels are more likely to experience financial crises in the future. This result holds even in the subsample of emerging markets.
In sum, emerging markets may be more prone to crises than advanced economies because their governments have on average lower reputation, and because in emerging markets there is higher uncertainty about the quality of political leaders. Governments in emerging markets have more to gain by riding the popularity benefits of a credit boom.
It is a common belief that policymakers’ concerns about their popularity and reputation have positive effects on economic outcomes. In contrast to this view, we show that popularity concerns may also have negative effects, increasing the likelihood of financial crises. This reputational mechanism may also be of relevance in other policy fields such as redistributive policies, privatisations, fiscal stimulus, or taxation decisions.
Fernandez-Villaverde, Jesús, Luis Garicano, and Tano Santos (2013), “Political Credit Cycles: The Case of the Eurozone”, Journal of Economic Perspectives, 27(3): 145–166.
Forbes, Kristin and Frank Warnock (2012), “Capital Flow Waves: Surges, Stops, Flight and Retrenchment”, Journal of International Economics, 88(2): 235–251.
Gorton, Gary and Guillermo Ordoñez (2014), “Crises and Productivity in Good Booms and in Bad Booms”, mimeo, University of Pennsylvania.
Herrera, Helios, Guillermo Ordoñez, and Christoph Trebesch (2014), “Political Booms, Financial Crises”, NBER Working Paper 20346.
McCarty, Nolan, Keith Poole, and Howard Rosenthal (2013), Political Bubbles: Financial Crises and the Failure of American Democracy, Princeton University Press.
Mendoza, Enrique and Marco Terrones (2012), “An Anatomy of Credit Booms and their Demise”, NBER Working Paper 18379.
Mian, Atif, Amir Sufi, and Francesco Trebbi (2010), “The Political Economy of the US Mortgage Default Crisis”, American Economic Review, 100(5): 1967–1998.
Mian, Atif, Amir Sufi, and Francesco Trebbi (2014), “Resolving Debt Overhang: Political Constraints in the Aftermath of Financial Crises”, American Economic Journal: Macroeconomics, 6(2): 1–28.
Reinhart, Carmen and Vincent Reinhart (2008), “Capital Flow Bonanzas: An Encompassing View of the Past and Present”, NBER Working Paper 14321.
Reinhart, Carmen and Kenneth Rogoff (2009), “The Aftermath of Financial Crises”, American Economic Review, 99(2): 466–472.
Schularick, Moritz and Alan Taylor (2012), “Credit Booms Gone Bust: Monetary Policy, Leverage Cycles, and Financial Crises, 1870–2008”, American Economic Review, 102(2): 1029–1061.