The use of capital controls as a policy tool – especially as a stopgap to ward off financial crises – is controversial. For example, in 1998, Malaysia was castigated by policymakers and financial markets for imposing capital controls in response to the East Asian financial crisis. In 2010, however, the IMF revised its stand against capital controls, recognising that sudden capital surges can pose risks for some countries, and acknowledging that controls on capital inflows may be part of a toolkit that countries use to ward off financial crises (Ostry et al. 2010). This change in policy reversed the previous IMF position favouring the free movement of capital.
Since capital controls have been used in response to exchange-rate crises, understanding their macroeconomic effects relative to other policies is an important agenda for research (e.g. Eichengreen and Rose 2014). On the one hand, capital controls bottle up inflows, which could potentially drive new investment spending and fuel a recovery in the wake of a crisis. On the other hand, their imposition could provide central banks with room for manoeuver; in particular, central banks can maintain fixed exchange rates, but pursue expansionary monetary policy in the short run to stimulate output and return to long-run policy objectives. Research on the 1997–1998 East Asian financial crisis has suggested that restrictions on the movement of capital may have produced a faster economic recovery in comparison to countries that relied on help from the IMF (Kaplan and Rodrik 2002).
Capital controls during the Great Depression
Determining the relative benefits and costs of capital controls for economic recovery is ultimately an empirical question, and the Great Depression offers a fertile testing ground for shedding light on this issue. The Depression was the first financial crisis in the era of modern economic growth in which a large number of countries responded to balance-of-payments pressures by imposing restrictions on the movement of capital. Few, if any, financial crises since the Depression have rivalled its severity and global impact, and few have witnessed so many countries responding by imposing capital controls – perhaps in part because many subsequent crises have been regional in nature.
Deflation spread globally after 1929, and as production and incomes fell, countries found it increasingly difficult to maintain pegged exchange rates. By the mid-1930s, most had abandoned the gold-exchange standard and were seeking refuge in a variety of alternative exchange-rate arrangements, including capital controls. The abandonment of gold, however, was carried out in a haphazard manner, with some countries following England off gold in 1931 and others steadfastly staying on gold until after 1933. Some countries chose to re-peg at lower rates to particular currencies, such as the pound sterling; others floated their currencies. Many imposed exchange controls (the name for restrictions on the current and capital accounts during that era) in order to shield their economies from the effects of short-term capital flows (‘hot money’) and balance-of-payments pressures.
Imposing capital controls might have prevented short-run capital flight and, under the ‘policy trilemma’ framework, enabled policymakers more room to aid ailing banking systems using monetary policy. Moreover, imposing capital controls while maintaining a fixed exchange rate might have reduced the possibility that a dramatic decline in the value of the currency would further increase the probability of a banking crisis. On the other hand, if capital-control countries kept their exchange rates pegged (perhaps due to a ‘fear of floating’) or delayed an adjustment in their parity, then the scope to engage in a competitive devaluation to boost the domestic production of exports might be more limited relative to floaters.
In Mitchener and Wandschneider (2014), we analyse the effects of capital controls on economic recovery in the 1930s by assembling a large, new monthly data set of macroeconomic variables and information on exchange rates and capital controls, which spans 1925–1936 and contains almost all of the countries that were on the interwar gold standard. We use these data to examine both their immediate effects on capital flight as well as their medium-term effects on economic recovery. Our empirical analysis takes advantage of the variation in timing of going off gold and heterogeneous policy responses in order to estimate the causal effects of exchange controls on economic recovery from the Great Depression, and time-shifted, difference-in-differences estimators allow us to account for bias arising from the variation in timing of going off gold (i.e. when treatment began).
Capital controls as a short-run expedient with little macroeconomic benefit
As Figure 1 suggests, capital controls achieved the short-run policy objective of stemming capital outflows as gold cover ratios stabilised in the months following their imposition. We then show that capital controls did not accelerate recovery from the Great Depression relative to countries that went off gold and floated. In examining the impact of capital controls on industrial production, exports, and prices, we only find statistically significant effects on industrial production – even after controlling for additional policy variables such as movements in the discount rate and changes in trade barriers. However, as shown in Figure 2, the estimated coefficient on industrial production suggests that capital controls slightly reduced its rate of growth relative to floaters. Thus, while capital controls provided an immediate tool to combat capital flight, they appear have held no advantage over a free float, and likely even hindered recovery after they were imposed.
Figure 1. Average cover ratios for exchange control countries
Note: The cover ratio is calculated as the ratio of bank gold reserves and foreign assets to domestic liabilities. Source: Statistisches Handbuch der Weltwirtschaft (1934, 1936/7).
Figure 2. Average growth in industrial production by regime and group
The fact that exchange-control countries broke from the gold standard earlier than gold bloc countries meant that their recoveries began sooner; however, once the latter group also finally abandoned gold, our analysis shows that countries imposing capital controls did not experience faster growth in industrial production and export growth relative to the gold bloc. Exchange controls allowed countries that perhaps ‘feared floating’ to maintain a fixed parity, but it offered scant improvement in terms of macroeconomic recovery.
Time-series analysis suggests that countries imposing capital controls did not actively pursue expansionary monetary policy after abandoning gold. An examination of discount-rate policy of capital-control countries suggest that, while capital-control countries did not follow France’s lead and continue to raise rates after imposing controls, they also did not pursue a discount-rate strategy similar to the US – a country which floated and then aggressively pursued expansionary monetary policy. Figure 3 shows that the average growth rate in the money supply of capital-control countries turned positive after their imposition, but it was slower than the growth rates of either floaters or gold bloc countries once they finally abandoned. For the countries that imposed capital controls in 1931 and 1932, these findings are consistent with a large body of research suggesting that monetary policies were far too tight in the early 1930s (Eichengreen 1992, Friedman and Schwartz 1962, Temin 1989) – promoting deflation and, in some cases, contributing to the collapse of banking systems. In contrast, countries that moved to floating rates pursued expansionary monetary policies and appear to have halted further deflation and declining incomes and production.
Figure 3. Average money growth rates by regime and group
Note: GB = gold bloc countries, FL= countries that floated, and ExC = exchange control countries.
Capital controls appear not to have been successfully used as tools for rescuing banking systems, stimulating domestic output, or for raising prices. Rather they appear to have been maintained as a means for restricting trade (working alongside or in lieu of restrictions on imports) and repayment of foreign debts. While our analysis suggests capital controls provided little macroeconomic benefit relative to other policies that were implemented in the 1930s, it would be difficult to conclude that they would have no ameliorative effects in other crises if employed with that purpose in mind. On the other hand, the experience of the 1930s suggests capital controls are often implemented with very short-run objectives in mind – to prevent capital flight. If kept in place, however, macroeconomic objectives can end up sharing the stage with other goals of policymakers.
Eichengreen, Barry (1992), Golden Fetters: The Gold Standard and the Great Depression, New York: Oxford University Press.
Eichengreen, Barry and Andrew K Rose (2014), “Capital controls in the 21st century”, VoxEU.org, 5 June.
Friedman, Milton and Anna Schwartz (1963), A Monetary History of the United States, 1867–1960, Princeton, NJ: Princeton University Press.
Kaplan, Ethan and Dani Rodrik (2002), “Did the Malaysian Capital Controls Work?”, in Sebastian Edwards and Jeffry Frankel (eds.), Preventing Currency Crises in Emerging Markets, Chicago: University of Chicago Press and NBER.
Mitchener, Kris James and Kirsten Wandschneider, (2014), “Capital Controls and Recovery from the Financial Crisis of the 1930s”, CEPR Discussion Paper 10019.
Ostry, Jonathan, Atish Ghosh, Karl Habermeier, Marcos Chamon, Mahvash Qureshi, and Dennis Reinhardt (2010), “Capital Inflows: The Role of Controls”, IMF Staff Position Paper 10/04.
Temin, Peter (1989), Lessons from the Great Depression, Cambridge, MA: MIT Press.