Orkun Saka, Nauro Campos, Paul De Grauwe, Yuemei Ji, Angelo Martelli 25 June 2019
Severe economic and political turbulence occurs in the aftermath of financial crises. What starts as a panic in a single financial market or institution usually propagates rapidly to other agents of the economy, and might necessitate an urgent and decisive reaction from policymakers. However, it is difficult to predict, a priori, whether the reaction of policymakers to the crisis would be in the direction of further liberalisation.
On the one hand, as financial institutions and markets become dysfunctional in the midst of a crisis, governments may feel the urge to intervene, for instance by bailing out the failed banks or increasing efforts to better regulate the misbehaving institutions. This could be politically unavoidable especially when the cause of the crisis is commonly perceived to be the result of ‘unregulated capitalism’ and public sentiment turns against the financial industry as well as the bankers at its helm (Dagher 2018).
On the other hand, such periods of instability may act as a catalyst for pushing forward liberalisation agendas that might have been stuck due to private interests or a lack of political enthusiasm. In that case, financial crises could open a window of opportunity to make sharp changes in the policy space. This view is in line with the more general crises-beget-reforms hypothesis (Drazen and Grilli 1993, Drazen and Easterly 2001).
In a recent paper (Saka et al. 2019), we study how financial reforms evolve in the aftermath of financial crises by carefully merging the classic dataset on financial liberalisations by Abiad et al. (2010) with the more recent but narrow dataset by Denk and Gomes (2017) and the recently updated IMF dataset on financial crises (Laeven and Valencia 2018). This provides us with one of the most comprehensive set of observations that has been employed on the crises-reforms nexus, covering 94 countries over the period between 1973 and 2015. The dataset covers seven major areas of financial reform and includes five indices directly related to the domestic banking sector (credit controls, interest rate controls, entry barriers, privatisation, and supervision), one index on restrictions in international capital movements, and one on asset markets (security market regulation).
Using a quasi-difference-in-differences approach, we first find that the immediate impact of financial crises is towards de-liberalisation. This result generally holds for any individual reform area and crisis type (banking, currency, or sovereign debt crises, in order of increasing effect size). It seems that, at least in the specific context of ‘financial’ crises and ‘financial’ reforms, periods of turmoil do not spur liberalisation efforts – on the contrary, they end up reversing some of the previously liberal policies. This contrasts with the predictions of the earlier literature on the political economy of reforms (Drazen and Easterly 2001).
Instead, our results align with the view that governments may consider crises as market failures and respond by increasing their interventions in the hope of an urgent correction in the markets. Therefore, it is likely that governments may be using reform reversals as a form of self-help (Pepinsky 2012). This could be especially true for bank bailouts (privatisation reversals), as there has been an increasing demand since the 1970s for the protection of middle-class wealth during financial crises. As argued by Chwieroth and Walter (2019), this may even result in the punishment of incumbent politicians if they fail to provide such protection. If this view is true, government interventions should be temporary and hence disappear in the medium-term as the crisis wanes.
An alternative argument to explain our findings on the reversals of reforms could be that politics breaks down and governments become fractionalised following financial crises, with more veto players coming into play and affecting the government’s reform programme (Mian et al. 2012). In that case, one would expect reversals to be persistent over time and crises to have a longer-term negative effect on the liberalisation process. In order to investigate this hypothesis, we utilise local projections à la Jordà (2005) and plot the impulse response function of a continuous measure of banking crises on the average level of financial liberalisation.
Figure 1 The effects of a banking crisis on average level of financial liberalisation
Source: Saka et al. (2019).
Notes: The figure shows the estimated impulse responses via the local projections (LPs) method (see Equation 2 in our paper) using the average financial reform as the endogenous variable and banking crises as the exogenous shock with four lags included as controls on the right-hand side for each variable. The shaded area represents the 90% confidence intervals.
Figure 1 illustrates the time horizon of the relationship between financial crises and reversals, showing us clearly that the effects are concentrated on the very short-term. After the contemporaneous decrease in liberalisation levels, there is no additional divergence between crisis and non-crisis countries in the following years. Indeed after three years – which is approximately the average duration of a banking crisis in our sample – countries that had a crisis start catching up with the ones that did not. Overall in a period of five to six years there is almost no effect of the banking crisis on average liberalisation levels. Some countries may even end up being more liberalised than their counterparts a decade after experiencing a banking crisis.
These results are more in line with the argument that governments react to crises by using short-term policy reversals in order to intervene and correct market failures (Pepinsky 2012). Such urgent policy reactions might be in high demand especially with middle-class voters (Chwieroth and Walter 2019) and thus explain our finding that the negative effect is almost entirely driven by the contemporaneous relationship in Figure 1. As soon as the crisis pressures ease, countries go back to their liberalisation paths and catch up with others.
These findings also have important implications for the general literature on the crisis-reform nexus. As documented before, many studies on the political economy determinants of reforms produce contradicting and inconsistent results that make it hard to reach a consensus (Campos and Nugent 2019). Our findings imply that, not only the contemporaneous relationship but also the persistence of the effects might matter in this discussion and could reconcile some of the contrasting results in the literature on the effects of crises on reforms at varying horizons.
Finally, our results support the case-study evidence reported in Dagher (2018) arguing that financial regulation is inherently pro-cyclical – being more lenient during booms and stricter during recessions – and crises may act as turning points. Whether such regulatory cycles occur due to changing sentiments in the public and how they interact with the lobbying power of the financial sector and electoral incentives of the incumbent politicians constitute fruitful avenues for future research.
Abiad, A, E Detragiache, and T Tressel (2010), “A New database of financial reforms”, IMF Staff Papers, 57 (2), 281-302.
Campos, N, and J Nugent (2019), “Crises galore: New measures, better estimates”, USC, mimeo.
Chwieroth, J M, and A Walter (2019), “The wealth effect: The middle class and the changing politics of banking crises”, VoxEU.org, 3 June.
Dagher, J (2018), “Regulatory cycles: Revisiting the political economy of financial crises”, VoxEU.org, 22 March.
Denk, O, and G Gomes (2017), “Financial re-regulation since the global crisis? An index-based assessment”, OECD Economics Department Working Papers, no. 1396.
Drazen, A, and W Easterly (2001), “Do crises induce reform? Simple empirical tests of conventional wisdom”, Economics & Politics, 13 (2), 129-157.
Drazen, A, and V Grilli (1993), “The beneﬁt of crisis for economic reforms”, American Economic Review, 83 (3), 598-607.
Jordà, Ò (2005), “Estimation and inference of impulse responses by local projections”, American Economic Review, 95 (1), 161-182.
Laeven, L, and F Valencia (2018), “Systemic banking crises revisited”, IMF Working Paper, no. 206.
Mian, A R, A Sufi, and F Trebbi (2012), “Political constraints in the aftermath of financial crises”, VoxEU.org, 21 February.
Pepinsky, T B (2012), “Do currency crises cause capital account liberalization?” International Studies Quarterly, 56 (3), 544-559.
Saka, O, N Campos, P De Grauwe, Y Ji, and A Martelli (2019), “Financial crises and liberalization: Progress or reversals?”, CEPR Discussion Paper no. 13776.