VoxEU Column International Finance

Do capital controls deflect capital flows?

Capital controls may help countries limit large and volatile capital inflows, but they may also have spillover effects on other countries. This column discusses recent research showing that inflow restrictions have significant spillover effects as they deflect capital flows to countries with similar economic characteristics.

The size and volatility of capital flows to developing countries have increased significantly in recent years (Figure 1), leading many economists to argue that national policies and multilateral institutions are needed to govern these flows (Forbes and Klein 2013, Blanchard and Ostry 2012). The IMF itself has reviewed its position on the liberalisation and management of capital flows, while recognising that “much further work remains to be done to improve policy coordination in the financial sector” (IMF 2012, p. 28).

Figure 1. Capital flows to developing countries

The essence of the problem can be simply stated:

  • Controls on capital inflows may help countries deal with the negative effects caused by large and volatile capital inflows.
  • However, these controls may have spillover effects on other countries, possibly leading to a policy response and to coordination problems.

How are these policy spillovers transmitted across countries? Are they economically meaningful? A better understanding of the relevant international externalities associated with the use of capital controls is an essential building block to improving multilateral cooperation in the financial sector.

An example: Controls in Brazil and inflows to South Africa

Figure 2 shows South Africa’s gross capital inflows (as a share of GDP) and Brazil’s inflow restrictions during a period when controls in South Africa were essentially stable. The index used is a measure, developed by the IMF of policies specifically aimed at restricting capital flows – the so-called Schindler index.1 In 2009 Brazil imposed more stringent inflow controls, such as the introduction of a 2% tax on equity and debt inflows, as well as other restrictions that led to a substantial increase in the Schindler index.

Figure 2. Brazil’s capital controls and inflows to South Africa

Did this policy change in Brazil affect capital inflows to South Africa? The figure suggests that this might be the case, but there could be other explanations behind the apparent positive correlation between inflow controls in Brazil and capital inflows to South Africa, such as the existence of a common shock affecting both countries. Moreover, even if a causal relationship existed, how large was the spillover effect from Brazil’s policy change?

Capital flow deflection

In Giordani et al. (2014), we address these questions and use a simple model of optimal capital controls to guide the empirical analysis. We show that inflow restrictions imposed by a country distort international capital flows to other countries. Intuitively, increases in inflow controls lower the return that foreign investors receive from exporting capital to that country and lead them to reallocate their funds to other borrowers with similar characteristics.

We call this effect ‘capital flow deflection’ as it is similar to ‘trade deflection’ – the distortion of exports to third markets caused by import tariffs (Bown and Crowley 2007). An important question is whether this international externality identified by the theory is supported by the data. This is what we focus on next.

New evidence on spillover effects of inflow controls

Previous work on the spillover effects of capital controls such as Forbes et al. (2012) relied on an event study approach. Event studies often provide detailed information on flows and implemented policies. However, one always wonders if the results of event studies extend beyond the specific cases considered.

The empirical analysis in our work employs data on inflow restrictions and gross capital inflows for a large sample of developing countries between 1995 and 2009. To test the theory, we augment a standard model of the determinants of capital inflows (a push-pull factor model) with a variable that captures the spillover effect of inflow restrictions. We divide countries into groups of likely substitutes based on common characteristics, such as geographic location, export specialisation, return, and risk. Figure 3 reports the Schindler index of inflow controls by country groups.

Figure 3. Schindler index of capital inflow controls for different country groups

The estimation yields strong evidence that capital controls deflect capital flows to countries with a similar risk structure. Perhaps surprisingly, we find no significant spillover effects on countries in the same region. This finding is consistent with the view that investors are guided by the similarity of economic characteristics of countries, rather than by their geographic location. Capital flow deflection is also found to be economically relevant. For instance, Brazil’s imposition of restrictions in 2009, which is reported in Figure 1, is estimated to have increased capital flows to South Africa by 0.5–1.0% of GDP.

These results are robust to a number of tests. In particular, spillovers to countries with similar risk structure continue to be significant when we use different measures of capital controls and of risk, when we focus on episodes of rapid surges of capital inflows, and when we use an instrumental variable approach to address potential endogeneity problems.

An application: Spillovers from liberalisation in China

The findings of this line of work are also useful for addressing questions about the spillover effects from capital account liberalisation, namely regarding the country-composition of these effects. For example, our model estimates predict that a complete removal of inflow restrictions in China would reduce capital inflows to countries with similar risk characteristics (see Table 1).2 This is because, other things being equal, the removal of inflow controls would increase the returns foreign investors receive from exporting capital to China, leading them to reallocate funds away from countries that belong to the same risk group.

Table 1. Predicted effects of Chinese capital account liberalisation

 
Change in gross inflows as a share of GDP

 

Chile

Mexico

Malaysia

Russian

Poland

Lower bound

-1.96

-2.11

-1.97

-2.20

-2.01
Upper bound
-4.47
-4.82
-4.48
-5.01
-4.59

While this is a comparative statics exercise that ignores general equilibrium effects that might be relevant for large developing economies, it provides some useful information on which countries are likely to be most affected by the removal of inflow restrictions in China. This result complements the findings of studies such as Bayoumi and Ohnsorge (2013) that look at the impact of overall liberalisation (including the removal of controls on capital outflows) on flows to and from China, but abstract from the country composition of potential spillover effects.

Conclusion

The evidence discussed in this column informs the debate on the efficient design of the international financial system. While capital flow deflection may not always be a concern from a welfare perspective, theoretical work has shown that spillover effects of capital controls could lead to an inefficient equilibrium in environments in which governments have limited policy instruments (e.g. controls are costly) and/or markets are incomplete (Korinek 2014). In these situations, multilateral rules may be needed to help governments internalise the impact of inflow restrictions on the welfare of other countries through capital flow deflection and avoid a costly escalation of tit-for-tat measures. Going forward, the experience in regulating similar problems of trade deflection within the WTO may provide some useful lessons for the international financial system.

Disclaimer: The views expressed herein are those of the authors, and should not be attributed to the IMF, its Executive Board, or its management.

Footnotes

1. The Schindler index calculates inflow controls as the average of the restriction dummies on a series of international transactions. The index varies between 0 (i.e. no restrictions) and 1 (i.e. restrictions on all international transactions).

2. This analysis is a positive exercise of potential spillover effects of capital controls, not a normative statement or policy advice on capital account liberalisation in China.

References

Bayoumi, T and F Ohnsorge (2013), “Do Inflows or Outflows Dominate? Global Implications of Capital Account Liberalization in China”, IMF Working Paper WP/13/89.

Blanchard, O and J Ostry (2012), “The Multilateral Approach to Capital Controls”, VoxEU.org, 11 December.

Bown, C P and M A Crowley (2007), “Trade Deflection and Trade Depression”, Journal of International Economics, 72(1): 176–201.

Forbes, K and M W Klein (2013), “Policymaking in Crises: Pick Your Poison”, VoxEU.org, 24 December.

Forbes, K, M Fratzscher, T Kostka, and R Straub (2012), “Bubble Thy Neighbor: Direct and Spillover Effects of Capital Controls”, NBER Working Paper 18052.

Giordani, P, M Ruta, H Weisfeld, and L Zhu (2014), “Capital Flow Deflection”, mimeo, IMF.

IMF (2012), “The Liberalization and Management of Capital Flows: An Institutional View”, IMF Staff Paper, November.

Korinek, A (2014), “Capital Controls and Currency Wars”, mimeo, Johns Hopkins University.

24,672 Reads