2.1 Economic Structure of Country i

Country i is populated by a unit mass of identical consumers who earn $y_t^i$ at time t and value consumption $c_t^i$ according to the following utility function:

with u^ʹ > 0, u^ʺ < 0 and βⁱ < 1, and where $C_1^i$ and $Y_1^i$ denote, respectively, aggregate consumption and income in country i at time 1, which are both taken as exogenous by the individual. Country i is defined as borrower (lender) whenever $C_1^i-Y_1^i>0(C_1^i-Y_1^i<0)$. Function e(.) captures the fact that capital inflows may be associated with negative external effects, such as the risk that results from domestic households ignoring the consequences of their foreign borrowing decisions on the country’s financial stability (financial fragility). Specifically, this externality is defined as

where x(.) is a positive twice continuously differentiable function with x^ʹ(.),x^″(.) > 0.⁷ If a country is a lender $(C_1^i-Y_1^i\le 0)$, the externality is null. If instead a country is a borrower $(C_1^i-Y_1^i>0)$, the externality associated with capital inflows enters individual welfare function.

Consumers must satisfy the following intertemporal budget constraint:

where R ≡ 1 + r is the world gross real interest rate, and τⁱ is a tax/subsidy on individual borrowing at time 1. In the context of a representative agent economy, τⁱ can also be interpreted as the capital control policy at time 1. A positive τⁱ denotes a capital inflow tax when the country is a borrower, and a capital outflow subsidy when the country is a lender; and a negative τⁱ has the opposite interpretation. In the rest of the paper, we simply refer to τⁱ as capital controls. The tax revenue when τⁱ > 0 (or the cost of the subsidy when τⁱ < 0) is assumed to be lump-sum rebated to (or financed from) consumers, so that $T^i=(C_1^i-Y_1^i){\tau }^{i}R$

In each country i, the representative consumer maximizes her objective function (1) while satisfying the budget constraint (2), and taking the world interest rate R, the policy τⁱ as well as the aggregate variables as given. Maximization gives rise to the usual Euler Equation:

implying consumption demands $c_1^i(\overline{R},\overline{{\tau }^i})$ and $c_2^i(\stackrel{+}{R},\stackrel{+}{{\tau }^i})$. Assuming substitution effect dominates income effect, a higher interest rate implies higher savings $(c_1^i\downarrow ,c_2^i\uparrow )$, and hence a fall in the demand of capital inflows in borrowing countries and a surge in the supply of capital outflows in lending countries. On the other hand, a country i that raises its capital controls (τⁱ ↑) creates a wedge between the world interest rate and the domestic interest rate (Rⁱ = R(1 + τⁱ)), thus stimulating domestic savings $(c_1^i\downarrow ,c_2^i\uparrow )$ and, hence, reducing its demand of capital inflows.

2.2 Optimal Capital Controls

For a world economy made up of n countries we now characterize the optimal capital control policy of each individual country i and the resulting world market equilibrium. The mass of country i is denoted by mⁱ ϵ [0,1], with ${\mathrm{\Sigma }}_{n=1}^n{m}^i=1$. Hence, mⁱ =0 implies that country i is small with respect to the world economy. Whether country i is large or small, its national planner maximizes the following social welfare function:

subject to the budget constraint

The crucial difference between the individual’s and the social planner’s optimization problem is that the latter internalizes the aggregate impact of individual consumption plans on capital inflows to country i. The socially optimal consumption plans for country i are implicitly determined by the following Euler equation:

A comparison of the two Euler equations (3) and (6) shows that there are two reasons for policy intervention in this economy. A first reason results from the presence of a domestic externality. The government may restrict capital inflows as a means to manipulate the domestic interest rate (Rⁱ = R(1 + τⁱ)), thus affecting agents’ saving patterns and offsetting this distortion. We follow the literature on financial fragility and refer to this rationale as the prudential motive for capital controls. While not explicitly modeled here, other non-prudential motives commonly discussed in the literature (e.g. targeting the real exchange rate or dealing with the “trilemma”) can also fit into this category. As prudential controls, they aim at manipulating the domestic rather than international interest rate.

A second source of market failure arises whenever country i is large enough to affect the world interest rate (mⁱ > 0). It is well known in the literature that large countries attempt to exploit their market power by taking advantage of the effect that their policy has on the world interest rate (terms-of-trade effect). For instance, a borrowing country exploits its monopsonistic power in the capital market by taxing capital inflows so as to limit domestic demand and thus keep a “low” cost of capital. In solving the optimization problem defined above, the national planner of country i knows that R depends on its own domestic capital demand $m^i(c_1^i-y_1^i)$ through the following market-clearing condition for the world economy, stating that the sum of net supplies of capital for the world as a whole must be null:

In the rest of the paper, we refer to this second rationale for imposing capital controls as the terms-of-trade motive.⁸

We are now ready to formally characterize the optimal capital control for country i.

Proposition 1 (Unilaterally Optimal Capital Controls) The unilateral optimal policy (τ^i*) can be decomposed into a prudential motive $\left(\widehat{{\tau }^i}\right)$ and a terms-of-trade motive $\left(\widetilde{{\tau }^i}\right)$, such that

with

and

where ε⁻ⁱ is the inverse elasticity of global savings faced by country i, and it is given by ${\epsilon }^{-1}=[dR/d(Y_1^{-i}-C_1^{-i})]\cdot (Y_1^{-i}-C_1^{-i})/R$ with $Y_1^{-i}\equiv {\mathrm{\Sigma }}_{j\ne i}{m}^j{Y}_1^j$ and $C_1^{-i}\equiv {\mathrm{\Sigma }}_{j\ne i}{m}^j{C}_1^j$.

Capital controls can be imposed for prudential or terms-of-trade reasons. Expression (9) captures country i’s optimal prudential controls. They are null whenever i is a lender. In fact, since the externality is null, the solution to the welfare maximization for a rational forward-looking national planner coincides with the utility maximization of the representative consumer, and the optimal prudential policy is no-intervention: $\widehat{{\tau }^i}$. Capital controls are instead strictly positive when country i is a borrower. In particular, the expression $x^{\prime }(c_1^i-y_1^i)/[u^{\prime }\left(c_1^i\right)-x^{\prime }(c_1^i-y_1^i\left)\right]$ tells us that (i) low levels of current consumption -and thus a high marginal utility of it- imply weak optimal capital controls; (ii) a strong marginal effect of the negative externality implies strong capital controls.

Expression (10) is the well known formula for the unilaterally optimal tax of a country that exploits its market power. A borrowing country faces a positive elasticity (ε⁻ⁱ > 0), and hence taxes capital inflows $(\widetilde{{\tau }^i}>0)$. In particular, a more rigid supply of capital from the rest the world (that is, a higher ε⁻ⁱ) commands stronger controls on capital inflows (that is, a higher $\widetilde{{\tau }^i}$). On the other hand, a lending country faces a negative elasticity (ε⁻ⁱ < 0), and hence taxes capital outflows $(\widetilde{{\tau }^i}<0)$. Moreover, a more rigid demand of capital from the rest of the world (that is, a lower ε⁻ⁱ) implies stronger controls on capital outflows (that is, a lower $\widetilde{{\tau }^i}$).

We conclude this section with the definition of equilibrium for this economy. The equilibrium is a configuration in which agents maximize their utility, each national planner implements the unilaterally optimal capital control policy, and the international market for capital clears. Specifically,

Definition. (World Market Equilibrium) A world market equilibrium is defined as the gross real world interest rate R* and each country’s consumption plan and capital controls ${\{c_1^{i*},c_2^{i*}{\tau }^{i*}\}}_{i=1}^n$ that satisfy (3), (5), (7), and (8).

2.3 Capital Flow Deflection

In this section and the next we focus on countries that are borrowers, and we analyze the cross-border spillovers associated with capital controls as well as the policy reactions to these spillovers. In particular, we carry out a comparative statics exercise in this section to analyze the effects of a change in capital controls in a set of countries on the world interest rate, as well as on the amount and the distribution of capital flows to the borrowing countries. Denote by Ω the set of borrowing countries, and by S any subset of it, S ⊆ Ω, with country i belonging to S.⁹ We start with the following statement.

Lemma 1. A rise in the capital controls in the set of borrowing countries S⁻ⁱ (with S⁻ⁱ ≡ S\i, (τ^S−i ↑)

• (i) lowers the equilibrium world interest rate (dR*/dτ^S−i < 0) and

• (ii) lowers the total amount of world savings $(d\left[{\mathrm{\Sigma }}_{\omega }{m}^{\omega }[Y_1^{\omega }-C_1^{\omega }(R,{\tau }^{\omega })]\right]/d{\tau }^{S^{-i}}<0\forall \omega :Y_1^{\omega }-C_1^{\omega }>0)$.

The findings of Lemma 1 are instrumental to prove our next propositions on the multilateral effects of capital controls. The first of these effects is capital flow deflection, which is the spillover effect of more stringent controls in one (or more) borrowing countries on the amount of capital inflows accruing to other borrowing countries. This effect is studied in the following

Proposition 2 (Capital Flow Deflection) A rise in the capital controls in the set of borrowing countries S⁻ⁱ (τ^S−i ↑) causes an increase in capital inflows to country i, that is:

Proposition 2 simply states that, when a borrowing country raises its barriers to capital inflows, it deflects part of these flows to other borrowing countries. The existence of this cross-country spillover effect of capital controls will be verified empirically in the second part of this paper.

The intuition behind capital flow deflection is illustrated graphically in Figure 3. The graph is divided in three parts. Part (i) plots the world demand (CFD) and supply (CFS) of capital. Parts (ii) and (iii) instead represent the demand of capital for respectively the borrowing country i (CFD_i) and the rest of the borrowing countries S⁻ⁱ (CFD^S-i), where for simplicity we assume S = Ω. The initial world market equilibrium is such that R* equalizes the supply of capital (measured by the distance AB) to the demand of capital (measured by the distance CD + EF). A rise in the capital controls of the borrowing countries S⁻ⁱ (τ^S−i ↑) shifts the demand of capital to the left (from CFD to CFD^′), thus generating lower world savings and a lower world interest rate R^*′ (Lemma 1). In particular, R^*′ is now such that the new lower capital supply (AB^′) equalizes the new lower capital demand (CD^′ + EF^′). The spillover effect arises because, as a consequence of a lower interest rate, capital inflows accruing to the borrowing country i increase (by the measure DD^′). In other words, while less capital flows to borrowing countries as a whole (capital flow “depression”), part of the capital flows previously directed towards countries S⁻ⁱ are now deflected to the borrowing country i.

Capital Flow Deflection

Citation: IMF Working Papers 2014, 145; 10.5089/9781498383202.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Capital Flow Deflection

Citation: IMF Working Papers 2014, 145; 10.5089/9781498383202.001.A001

• Download Figure
• Download figure as PowerPoint slide

Capital Flow Deflection

Citation: IMF Working Papers 2014, 145; 10.5089/9781498383202.001.A001

• Download Figure
• Download figure as PowerPoint slide

2.4 Capital Controls’ Response

We now introduce the policy response of borrowing countries to capital controls imposed by other borrowers. We show that the nature of this policy response depends on the motive (prudential or terms-of-trade) for capital controls. Our findings are summarized in the following

Proposition 3 (The Policy Response) A rise in the capital controls in the set of borrowing countries S⁻ⁱ (τ^S−i ↑) causes a policy response by country i which can be decomposed into

where

We have isolated two, possibly opposing, forces that govern the relationship of capital control policies among borrowing countries. The presence of a capital inflows’ negative externality makes the policies complementary, while the exploitation of market power by large borrowing countries may or may not act in the same direction depending on the impact of capital controls on the elasticity of global savings.

This result can be rationalized as follows. A rise in capital controls in a set of countries (or just in a single large country) lowers the equilibrium interest rate (dR*/dτ^S−i < 0) and raises capital inflows to country i (Proposition 2). This exacerbates the negative domestic externality and thus leads country i to “defend itself” from an excessive capital inflow by raising its own capital controls for prudential reasons $(\widehat{d{\tau }^i}/dR<0)$. If country i is small, its reaction to an increase of capital controls in other borrowing countries is unambiguous: it responds by raising its own barriers (part (i)).

If country i is large, instead, the nature of the relationship among capital controls -that is, whether capital controls are complements or substitutes - also depends on the sign of $\widetilde{d{\tau }^i}/d{\tau }^{S^{-i}}$ which in turn depends on how the (inverse of the) elasticity of global savings faced by borrowing country i (ε⁻ⁱ) responds to changes in the capital controls of the countries involved (part (ii)). In fact, in deciding its optimal policy response, a large country i takes into account, not only the effect of countries S⁻ⁱ’s policies but also the one of its own policy on elasticity ε⁻ⁱ. In the proof of Proposition 3, we show that both effects are ambiguous and ultimately rest on the preference fundamentals of consumers. In particular, we show that the complementarity or substitutability of the terms-of-trade driven capital controls depend on whether these controls become more or less effective in affecting the world interest rate as they increase.¹⁰

The proposition shows how prudential controls (and, more broadly, any inflow restrictions aiming at manipulating the domestic interest rate) respond to exogenous shocks. Taken together, Lemma 1 and Proposition 3 delineate a complementary relationship between the world interest rate and prudential capital control policies, whereby a lower interest rate leads to more restrictions on capital inflows, and vice versa. Such complementarity may amplify the effects of exogenous shocks to the world economy or to any subset S of it, thus giving rise to what is usually called a multiplier effect. More formally, denote by $\widehat{{\tau }^i}(R*,\rho )$ the optimal prudential policy function for country i ϵ S, and where parameter ρ captures any feature that affects $\widehat{{\tau }^i}$ other than changes in R* (such as the endowments $y_t^i$, or the discount factor βⁱ). We are now ready to formulate the following:

Corollary 1. (The Multiplier Effect) A shock to any subset of borrowing countries S causes a prudential policy response of each borrowing country i ϵ S which can be decomposed into

A multiplier effect characterizes the prudential capital control policy of each country i, as the total equilibrium response is higher than the partial equilibrium response because $(d\widehat{{\tau }^i}/dR)(d{R}^{*}/d{\tau }^{S^{-i}})>0$.

The complementary relationship between $\widehat{{\tau }^i}$ and R* that gives rise to the multiplier effect is illustrated graphically in Figure 4. To gain an intuition of this corollary, suppose a non-negligible fraction of borrowing countries are hit by an exogenous shock (ρ ↑) that induces them to raise their restrictions on capital inflows, say from point E to E^′ in Figure 4. This rise in capital controls lowers the world interest rate (as proven in point 1 of Lemma 1), which moves from E^′ to E^″. Each country, however, reacts to a lower R by further increasing its optimal capital controls from E^″ to E^‴, which in turn triggers a further decrease in the world interest rate, and so on and so forth. As a result of this chain reaction, the aggregate policy response to the shock, as measured by the horizontal distance between point E and point F, exceeds the initial capital controls imposed by each individual borrowing country, as measured by the length EE^′.

The Multiplier Effect in Capital Controls

Citation: IMF Working Papers 2014, 145; 10.5089/9781498383202.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

The Multiplier Effect in Capital Controls

Citation: IMF Working Papers 2014, 145; 10.5089/9781498383202.001.A001

• Download Figure
• Download figure as PowerPoint slide

The Multiplier Effect in Capital Controls

Citation: IMF Working Papers 2014, 145; 10.5089/9781498383202.001.A001

• Download Figure
• Download figure as PowerPoint slide

2.5 Welfare

Let us close this section with a remark on the welfare implications of the theoretical model. It is well known that, from a global welfare perspective, prudential and terms-of-trade-driven controls have radically different consequences. Under the conditions stated in Korinek (2014) that the set of policy instruments and markets are complete, prudential controls are efficient both from a unilateral and from a multilateral point of view. Policy intervention aimed at exploiting the terms of trade effect is instead sub-optimal.¹¹ As a result, while unilateral capital controls driven by terms of trade motives call for a role of multilateral institutions to improve international policy coordination, no need for such institutions arises when policy measures are set for prudential reasons only. However, if the government has imperfect policy instruments (e.g. costly capital controls) or if markets are incomplete, the complementarity highlighted in the multiplier effect may lead to an inefficient world equilibrium characterized by excessively high restrictions and thus call for international policy coordination.

3.1 Data

Our sample consists of 78 less industrialized countries and emerging markets (see Table 1 for the full list). We focus on developing countries as they are more likely to employ capital controls than more advanced countries which are blessed with a wider set of policy instruments to address domestic distortions.¹³ The sample period spans 15 years from 1995-2009 and the data are in annual frequency. A country is included in our sample only if it has at least ten years of observations for all variables used in the regression models. While the sample period, data frequency and the selection of countries are inevitably constrained by data availability, the panel covers several interesting episodes including the crises in Asia in the 1990s and in South America in the early 2000s and the first two years of the Great Recession, 2007-2009.

Table 1.

Country List by Risk

Note: Time-invariant groups by composite risk. The countries in bold are used in the regression analysis. All the countries listed are used in computing τ^S−i.

Table 1.

Country List by Risk

Hish Risk	Hish-Moderate Risk	Low-Moderate Risk	Low Risk
Angola	Albania	Argentina	Bahamas, The
Burkina Faso	Algeria	Azerbaijan	Bahrain
Congo, Republic of	Armenia	Bolivia	Botswana
Côte d’Ivoire	Bangladesh	Brazil	Chile
Dem. Rep. of the Congo	Belarus	Bulgaria	China
Ethiopia	Cameroon	Dominican Republic	Costa Rica
Guinea	Colombia	Egypt	Croatia
Guinea-Bissau	Ecuador	El Salvador	Hungary
Haiti	Ghana	Gabon	Jordan
Iraq	Guyana	Guatemala	Kuwait
Lebanon	Honduras	India	Latvia
Liberia	Indonesia	Iran	Libya
Malawi	Kenya	Jamaica	Lithuania
Mozambique	Madagascar	Kazakhstan	Malaysia
Myanmar	Mali	Papua New Guinea	Mexico
Nicaragua	Moldova	Paraguay	Morocco
Niger	Mongolia	Peru	Namibia
Nigeria	Romania	Philippines	Oman
Pakistan	Senegal	Russia	Panama
Sierra Leone	Sri Lanka	South Africa	Poland
Sudan	Suriname	Syria	Qatar
Togo	Tanzania	Thailand	Saudi Arabia
Turkey	Uganda	Ukraine	Trinidad and Tobago
Zambia	Venezuela	Uruguay	Tunisia
Zimbabwe	Yemen	Vietnam	United Arab Emirates

Note: Time-invariant groups by composite risk. The countries in bold are used in the regression analysis. All the countries listed are used in computing τ^S−i.

View Table

Table 1.

Country List by Risk

Hish Risk	Hish-Moderate Risk	Low-Moderate Risk	Low Risk
Angola	Albania	Argentina	Bahamas, The
Burkina Faso	Algeria	Azerbaijan	Bahrain
Congo, Republic of	Armenia	Bolivia	Botswana
Côte d’Ivoire	Bangladesh	Brazil	Chile
Dem. Rep. of the Congo	Belarus	Bulgaria	China
Ethiopia	Cameroon	Dominican Republic	Costa Rica
Guinea	Colombia	Egypt	Croatia
Guinea-Bissau	Ecuador	El Salvador	Hungary
Haiti	Ghana	Gabon	Jordan
Iraq	Guyana	Guatemala	Kuwait
Lebanon	Honduras	India	Latvia
Liberia	Indonesia	Iran	Libya
Malawi	Kenya	Jamaica	Lithuania
Mozambique	Madagascar	Kazakhstan	Malaysia
Myanmar	Mali	Papua New Guinea	Mexico
Nicaragua	Moldova	Paraguay	Morocco
Niger	Mongolia	Peru	Namibia
Nigeria	Romania	Philippines	Oman
Pakistan	Senegal	Russia	Panama
Sierra Leone	Sri Lanka	South Africa	Poland
Sudan	Suriname	Syria	Qatar
Togo	Tanzania	Thailand	Saudi Arabia
Turkey	Uganda	Ukraine	Trinidad and Tobago
Zambia	Venezuela	Uruguay	Tunisia
Zimbabwe	Yemen	Vietnam	United Arab Emirates

Note: Time-invariant groups by composite risk. The countries in bold are used in the regression analysis. All the countries listed are used in computing τ^S−i.

We use Schindler’s (2009) index of capital controls, which is a de jure measure constructed from information contained in the Annual Report on Exchange Arrangements and Exchange Restrictions (AREAER) provided by the IMF. The main advantage of this index relative to others commonly used in the literature such as Chinn and Ito (2006) is that the Schindler index distinguishes between controls on capital inflows and outflows. In particular, Schindler’s index calculates inflow controls as the average of the restriction dummies on the following series of international transactions: purchase of financial assets locally by nonresidents, sale or issue of financial assets abroad by residents, collective investment by nonresidents to residents, financial credits by nonresidents to residents, inward direct investment, and liquidation of direction investment.¹⁴ The index varies between 0 (i.e. no restriction) and 1 (i.e. restrictions on all international transactions). Schindler also computes outflow controls in a similar way.¹⁵ We will use this second set of restrictions in our robustness analysis.

The data on capital controls are depicted in Figure 5. Panel 5.A plots the inflow controls by regions, with higher numbers indicating more stringent restrictions.¹⁶ There are noticeable differences as well as some common trends across regions. Asian and former Soviet bloc countries are more likely to have higher inflow restrictions than Latin American and central and east European countries. Countries across all regions tightened inflow controls during the Asian financial crisis, and the high level persisted in most regions except for central and east European countries that were going through capital account liberations in preparation for EU membership. During the 2007 financial crisis, most regions tightened inflow controls except for countries of the former Soviet bloc whose restrictions were nevertheless very stringent compared to other regions except for Asia.

Figure 5 also depicts inflow restrictions when we group countries by economic characteristics rather than geographic location. Specifically, we focus on export specialization (Panel 5.B), growth rate (Panel 5.C) and composite risk (Panel 5.D). By dividing countries into groups, we aim at capturing characteristics that render them close substitutes from the perspective of international investors. In this regard, the natural way to group countries together is based on the similarity of risk and return, as they are the two main determinants of investment decisions. Forbes et al. (2013) find that countries exporting to China are affected by Brazil’s capital inflow restrictions. Since these countries are mostly commodity exporters, we also consider export specialization as a relevant country grouping.¹⁷ Just like geographic location, groups by economic characteristics also display some interesting common patterns. In particular, countries with lower composite risk tend to display more stable and lower capital controls relative to riskier countries. Primary goods and service exporters tend to have lower inflow restrictions than fuel, manufacturing and diversified exporters. Finally, no clear pattern emerges when we group countries by growth rate.

Schindler’s Inflow Control Index by Groups

Citation: IMF Working Papers 2014, 145; 10.5089/9781498383202.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Schindler’s Inflow Control Index by Groups

Citation: IMF Working Papers 2014, 145; 10.5089/9781498383202.001.A001

• Download Figure
• Download figure as PowerPoint slide

Schindler’s Inflow Control Index by Groups

Citation: IMF Working Papers 2014, 145; 10.5089/9781498383202.001.A001

• Download Figure
• Download figure as PowerPoint slide

Following Forbes and Warnock (2012) we compute gross capital inflows as the sum of net portfolio investment liabilities, other liabilities, and net foreign direct investment. All components of gross capital inflows are available from the IMF’s International Financial Statistics (IFS) database as reported in BPM5 format.¹⁸ Table 2 shows summary statistics of inflow controls, gross inflows, and other variables used in the regression analysis. Table 3 reports the data sources. Note that we have two measures that capture the risk of investing in a country: composite risk and law and order. Higher values indicate lower risk. The composite risk index is a measure of a country’s combined economic, financial, institutional, and political risks, while law and order index captures mainly the institutional risk of a country. For this reason, we focus on the composite risk index as the proxy for country risk in our main regressions. The other measure of country risk is used in the robustness analysis. The other variables in the spillover analysis are standard taken from the push-pull factor literature. We use real US interest rate as a proxy for world real interest. This is computed from annualized US three month treasury rate adjusted by US inflation. The VIX index is used as a proxy for global volatility. Inflation and other country specific variables come from the WEO, while we use Ilzetzki et al. (2010) for exchange rate regime classifications.

Table 2.

Summary Statistics by Country Groups

Table 2.

Summary Statistics by Country Groups

Variable	Obs	Mean	S.D.	Min	Max
Schindler index (inflows restrictions)	1170	0.49	0.34	0.00	1.00
Gross inflows (% GDP)	1108	4.84	8.08	−38.99	63.85
VIX	1170	21.54	6.45	12.58	34.04
Real US interest rate	1170	0.94	1.63	−2.36	3.18
Inflation	1135	15.01	54.01	−9.86	1061.21
Real GDP growth rate	1169	4.12	4.72	−24.79	62.19
Nominal GDP per capita in USD (logged)	1159	7.42	1.17	4.38	10.66
Nominal GDP in USD (logged)	1170	10.10	1.70	6.47	15.42
Real effective exchange rate	1170	106.78	33.93	56.09	597.37
Composite risk	1138	66.24	8.72	28.00	86.70
Law and order	1138	3.38	1.14	1.00	6.00
de facto exchange rate regime	1134	2.17	1.15	1.00	6.00

View Table

Table 2.

Summary Statistics by Country Groups

Variable	Obs	Mean	S.D.	Min	Max
Schindler index (inflows restrictions)	1170	0.49	0.34	0.00	1.00
Gross inflows (% GDP)	1108	4.84	8.08	−38.99	63.85
VIX	1170	21.54	6.45	12.58	34.04
Real US interest rate	1170	0.94	1.63	−2.36	3.18
Inflation	1135	15.01	54.01	−9.86	1061.21
Real GDP growth rate	1169	4.12	4.72	−24.79	62.19
Nominal GDP per capita in USD (logged)	1159	7.42	1.17	4.38	10.66
Nominal GDP in USD (logged)	1170	10.10	1.70	6.47	15.42
Real effective exchange rate	1170	106.78	33.93	56.09	597.37
Composite risk	1138	66.24	8.72	28.00	86.70
Law and order	1138	3.38	1.14	1.00	6.00
de facto exchange rate regime	1134	2.17	1.15	1.00	6.00

Table 3.

Data Sources

Table 3.

Data Sources

Data	Source
Capital flows	International Financial Statistics (IFS)
Capital controls	Schindler (2009)
VIX	Yahoo Finance
US three month treasury rate	Federal Reserve Economic Data (FRED)
Inflation rate	World Economic Outlook (WEO)
Real GDP growth rate	WEO
Nominal GDP per capita in US$	WEO
Nominal GDP in US$	WEO
Real Effective Exchange Rate	Information Notice System (INS)
Composite risk index	Political Risk Service (PRS)
Law and Order	PRS
de facto Exchange rate regime classification	Reinhart et al. (2008)
Election Year	Database of Political Institutions (DPI)
Political Orientation of the Ruling Party	DPI

View Table

Table 3.

Data Sources

Data	Source
Capital flows	International Financial Statistics (IFS)
Capital controls	Schindler (2009)
VIX	Yahoo Finance
US three month treasury rate	Federal Reserve Economic Data (FRED)
Inflation rate	World Economic Outlook (WEO)
Real GDP growth rate	WEO
Nominal GDP per capita in US$	WEO
Nominal GDP in US$	WEO
Real Effective Exchange Rate	Information Notice System (INS)
Composite risk index	Political Risk Service (PRS)
Law and Order	PRS
de facto Exchange rate regime classification	Reinhart et al. (2008)
Election Year	Database of Political Institutions (DPI)
Political Orientation of the Ruling Party	DPI

3.2 Evidence of Capital Flow Deflection

In this subsection, we present evidence that capital controls deflect capital flows to other countries. We first present our empirical strategy, which extends the standard push-pull factor model to account for spillover effects. Then we present the regression results for the entire sample of developing economies and for groups of countries that are likely substitutes from the perspective of international investors. Finally, we undertake a series of robustness tests.

3.2.2 Benchmark Regression

Table 4 shows the OLS estimation results of equation (11) for the entire sample (i.e. when S = Ω). For expositional clarity, variables are divided into three blocks: global push variables that only vary across time and are invariant across countries; domestic pull factors that vary across both time and countries; and the spillover variable -the rest of world’s capital inflow control- which also varies both across time and countries. Column (1) reports the OLS coefficients for equation (11). To reduce possible omitted variable bias, column 2 reports the OLS coefficients when year fixed effects are included. Since global push factors only vary over time but do not vary across countries, they drop out in the specification with year fixed effects.

Table 4.

Capital Flow Deflection (Entire Sample)

Note: Robust standard errors clustered at country level in parentheses. *** p< 0.01, ** p < 0.05, * p < 0.1

Table 4.

Capital Flow Deflection (Entire Sample)

	(1)	(2)	Expected Signs
Global Push Factors
Real US interest rate	−0.337**		−
	(0.152)
VIX	−0.132***		−
	(0.0490)
Domestic Pull Factors (all lagged)
Real GDP growth rate	0.595***	0.504**	+
	(0.188)	(0.194)
Real GDP growth rate shock	−0.434**	−0.406*	−
	(0.209)	(0.207)
Real GDP per capita (logged)	0.652	0.481	+
	(0.571)	(0.565)
De jure Capital inflow control	−2.468*	−2.455*	−
	(1.341)	(1.440)
Composite risk index	0.150**	0.143**	+
	(0.0647)	(0.0646)
Spillovers
ROG’s inflow control	23.00***	8.013	+
	(5.807)	(27.50)
Year fixed effects	No	Yes
Observations	1,007	1,007
R-squared	0.135	0.167

Note: Robust standard errors clustered at country level in parentheses. *** p< 0.01, ** p < 0.05, * p < 0.1

View Table

Table 4.

Capital Flow Deflection (Entire Sample)

	(1)	(2)	Expected Signs
Global Push Factors
Real US interest rate	−0.337**		−
	(0.152)
VIX	−0.132***		−
	(0.0490)
Domestic Pull Factors (all lagged)
Real GDP growth rate	0.595***	0.504**	+
	(0.188)	(0.194)
Real GDP growth rate shock	−0.434**	−0.406*	−
	(0.209)	(0.207)
Real GDP per capita (logged)	0.652	0.481	+
	(0.571)	(0.565)
De jure Capital inflow control	−2.468*	−2.455*	−
	(1.341)	(1.440)
Composite risk index	0.150**	0.143**	+
	(0.0647)	(0.0646)
Spillovers
ROG’s inflow control	23.00***	8.013	+
	(5.807)	(27.50)
Year fixed effects	No	Yes
Observations	1,007	1,007
R-squared	0.135	0.167

Note: Robust standard errors clustered at country level in parentheses. *** p< 0.01, ** p < 0.05, * p < 0.1

Both specifications are able to explain more than 10% of the variation in gross capital inflows. All variables except for real GDP per capita are significant determinants of gross capital inflows and their coefficients have the expected signs.²² Our key variable of interest, the rest of group’s capital inflow controls $\left({\tau }_t^{S-i}\right)$, is significant in the standard push-pull factor model. But the variable becomes insignificant once we control for the year fixed effects, implying that the spillover effect found in column (1) is likely to be due to omitted variable bias. In particular, global push factors other than the ones included in regression (11) may cause gross capital inflows in a developing country and capital controls in the rest of the world to move in the same direction. The lesson from Table 2 is, therefore, that the pull-push model (11) explains gross capital inflows to developing countries reasonably well, but the evidence of spillover effects from the rest of the world is not robust.

The lack of evidence of capital flow deflection for the entire sample is hardly surprising and is in line with other studies (Forbes et al., 2013). In particular, note that this exercise is not a good test of the model in Section 2. An underlying assumption of the model is that countries are identical, if not for the capital controls they implement. In other words, there is a single perfectly integrated world capital market, where all these countries are perfect substitutes from the perspective of investors. This assumption clearly does not hold when we consider the entire sample, where countries are highly heterogeneous, but it could be more reasonable when we look at a smaller group of countries with similar characteristics. Specifically, capital flow deflection may be significant within well defined groups of developing economies, while still be on average irrelevant for the broad (and highly heterogeneous) sample. This is precisely what we investigate next.