Chapter

Appendix. Measuring Capital Account Restrictiveness: A Survey of the Literature

Author(s):
International Monetary Fund. Monetary and Capital Markets Department
Published Date:
October 2010
Share
  • ShareShare
Show Summary Details

How can the most benefit be gained from an open capital account, despite the risks that come with unhindered international capital flows? This is the question prompted by a series of financial crises. The jury is still out among policymakers and scholars on questions such as the growth-enhancing nature of financial globalization, the efficiency of capital controls, and the relationship between resilience to a crisis and capital account liberalization. In part, the questions are hard to answer because a satisfactory way to measure the intensity of capital account restrictions for a sufficiently large number of countries and long time series has not been established yet. This appendix takes a look at researchers’ attempts to measure the extent of capital account restrictiveness.

The IMF’s Annual Report on Exchange Arrangements and Exchange Restrictions (AREAER) is a primary source of information regarding de jure controls on the capital account. Since 1950, the AREAER has included a description of the foreign exchange arrangements, exchange and trade systems, and capital controls of all IMF member countries. Researchers rely on this database because the data cover a long time period and a large number of countries, including developing and emerging economies.

An additional but less frequently used source of information is the OECD Codes of Liberalisation of Capital Movements and Current Invisible Operations (Codes). These legal instruments represent the liberalization commitments of Organization for Economic Cooperation and Development (OECD) member countries, which currently number 32. The Codes include a detailed description of capital transactions and the respective controls imposed by OECD member countries. The categorization of capital transactions is similar to that in the AREAER, but the Codes are limited in terms of the period of time and the number of countries they contain. The Codes by definition cover only OECD member countries and only for the duration of their membership. Most of the member countries are advanced economies that eliminated their capital controls long ago, and so the Codes’ information is useful only for specifically targeted research.

Beginning with the 1967 issue, the AREAER has provided a binary variable indicating whether a country imposes controls on payments for capital transactions by residents.22 Since 1996, to overcome the limitations of the binary aggregation, it has extended the coverage in a comprehensive manner and provides binary indicators across subcategories of capital transactions with differentiating controls according to the direction of flows and residency.

In the AREAER’s post-1996 reporting system, the capital account transactions category is broken down into 11 categories of capital transactions and 2 categories of provisions specific to the financial sector.23 Some categories are further divided into subcategories. Furthermore, for each transaction subcategory, a distinction is made between restrictions on inflows and outflows and between transactions by residents and nonresidents. In total, more than 40 types of transactions related to the capital account are coded on a binary basis in the current tabular system.

Table 10.Summary of AREAER–Based Indices
IndexPeriodCountryDisaggregation

by

direction

of flows
Disaggregation

by

type

of flows2
Description
Aggregated

indices
Advantages: Extensive coverage

Disadvantages: Highly aggregated. A potential effect from the methodological break in underlying indicators at the 1996 revision.
IMF dummy1966–All1NoNoBinary indicator on the existence of restrictions on capital transactions
IMF share1970–AllNoNoProportion of years in an examined window for which countries had liberalized capital accounts
Chinn and Ito (2008)1970–2007184NoNoThe first standardized principal component of four IMF binary indicators (on the existence of multiple exchange rates, controls on current accounts, controls on capital accounts, and repatriation requirements). The index is updated annually by the authors.
Glick and Hutchison (2005)1975–9769NoNoIMF dummy for the pre-1996 period, and a dichotomous classification based on the coverage of controls for the post-1996 period
Disaggregated indicesAdvantages: Finer breakdowns by direction of flows, or by type of flows.

Disadvantages: Indicates only the existence.

Difficult to extend to pre-1996 period, because text information in the pre-1996 period is not always detailed enough to distinguish controls on inflows and outflows.
Johnston and Tamirisa (1998)1995345YesYes (13)Average of binary indicators of controls on sub-categories of transactions
Miniane (2004)1983–200023NoYes (13)Average of binary indicators of controls on subcategories of transactions. Underlying indicators for the pre-1996 period are coded based on the text information.
Brune (2006)1970–2007184YesYes (8)Average of binary indicators of controls on subcategories of transactions (five types of capital transactions, repatriation requirements, and multiple exchange rates). Underlying indicators for the pre-1996 period are coded based on the text information.
Schindler (2009)1995–200591YesYes (6)Average of binary indicators of controls on sub-categories of transactions. Disaggregation corresponds to major components in the Balance of Payment statistics. The index does not cover some components of capital flows.
Intensity indicesAdvantage: Measures the intensity of capital controls

Disadvantage: Subjective judgment, limited country coverage
Quinn (1997), Quinn and Toyoda (2008)1950–9994YesNoScores (0, 0.5, 1, 1.5, or 2) based on the severity of restrictions (existence of approval requirements and frequency of approval)
Potchamanawong and others (2008)1995–200426YesYes (13)Scores (0, 0.25, 0.5, or 1) based on the severity of restrictions (requirements for reporting, registration, or approval; existence of quantitative restrictions)
Montiel and Reinhart (1999)1990–9615NoNoScores (0, 1, or 2) based on the existence of prudential regulations and explicit capital control measures.
Source: Published IMF staff reports.

Because of limited availability before 1996, restrictiveness indices covering long time series are highly aggregated. Some researchers use the IMF binary variable on capital controls directly. An early example is found in Alesina, Grilli, and Milesi-Ferretti (1994). Klein and Olivei (2008) convert the binary variable to the proportion of years during which countries had open capital accounts. Grilli and Milesi-Ferretti (1995) suggest including other measures, such as multiple exchange rates and current account restrictions, that might affect investors’ ability to circumvent capital controls. Along these lines, Mody and Murshid (2005) construct an index for 60 countries during 1977–99 by taking the sum of four dummy variables on the presence of multiple exchange rates for capital account transactions, controls on current account transactions, controls on capital account transactions, and the stringency of requirements for the repatriation and/or surrender of export proceeds. With four similar dummy variables, Chinn and Ito (2008) come up with a financial openness indicator by taking the first principal component. Their index has been updated to cover the period 1970–2007 for 182 countries.

More disaggregated indices have become available since the pioneering work of Johnston and Tamirisa (1998). They built an index for 45 countries that treated subcategories of transactions separately, depending on the direction of capital flows. Their index was constructed only for 1995, but later was developed further by Miniane (2004). Schindler (2009) takes a similar approach and constructs an index that differentiates between various transactions according to the direction of the flows. His data set covers 91 countries during 1995–2005. An important difference between Johnston and Tamirisa (1998) and Schindler (2009) is in the types of transactions the indices cover. The latter focuses on restrictions on shares, bonds, money market instruments, collective investments, financial credits, and direct investment, whereas the former is more comprehensive, covering other capital transactions (derivatives, real estate transactions, and personal capital transactions) as well as restrictions on accounts, repatriation requirements, and some provisions specific to the financial sector.

Some researchers try to connect pre- and post-1996 AREAER data to produce longer time series.

  • (1) Miniane (2004) constructs an index similar to that of Johnston and Tamirisa (1998) but extends it to cover 1983–2000 for 23 countries and 13 major categories. For the period 1996–2000, his index follows the method described in Johnston and Tamirisa (1998). For the pre-1996 period, he uses the text information in the AREAER to induce backward. His index does not differentiate between controls on inflows and outflows, because the pre-1996 AREAER does not always contain enough information to make the distinction.

  • (2) Glick and Hutchison (2005) develop a dichotomous classification based on the coverage of controls for the 1996–97 period and connect the index to the pre-1996 AREAER binary indicator. Their index includes 69 countries for the period 1975–97. A country is classified as restricted if more than half the subcategories are subject to controls in the 1996–97 observations; otherwise it is considered unrestricted.

  • (3) Brune (2006) constructs a disaggregated index with long time series for an extensive list of countries. According to the description in Brune and Guisinger (2007), the index covers 184 countries during 1970–2007 for 12 categories of current and capital account transactions. The categorization is more aggregated than the one in Johnston and Tamirisa (1998), but some of the categories differentiate between controls on inflows and outflows. The index assigns a binary value to each category based on the descriptive information in the pre-1996 period.

These indices have some imperfections in common. First, the AREAER’s binary coding does not allow for measurement of capital controls’ intensity or coverage or of their gradual liberalization or intensification. This is because all controls have the same value, whether they prohibit, impose a tax on, set a ceiling on, or subject a transaction to the authorities’ approval. Similarly, the same value is attributed to controls that apply to all transactions included in the category and to those that affect only certain types of transactions in a category. Hence, the binary indicators reflect only a distinction between the presence or absence of controls. Countries typically liberalize controls gradually, differentiating between inflows, outflows, and types of transactions within one group of transactions or gradually increasing the ceiling for specific transactions. Countries also often tighten controls by imposing additional requirements or setting lower limits for specific transactions. These adjustments, which do not change the binary value of the transactions but increase or decrease their intensity, are not captured by any of the indices above.

Researchers have proposed alternative measures to address these limitations. Quinn (1997) and Quinn and Toyoda (2008) build indices that reflect the intensity of controls based on the text information of the AREAER. Each country receives a value between 0 and 2 in increments of 0.5, depending on the extent of the restrictions. Although they do not differentiate among the types of transactions, restrictions on inward flows and outward flows are coded separately. Their data set covers 1950–99 for 94 countries. Potchamanawong (2007) and Potchamanawong and others (2008) similarly evaluate the intensity of controls. This index, which takes a value between zero and 1 in increments of 0.25, covers each of the 13 major categories of the post-1996 AREAER, with a distinction between inflows and outflows. The data set covers 1995–2004 for 26 emerging market economies. Another example is found in Montiel and Reinhart (1999), who measure the intensity of controls by assigning a value of zero, 1, or 2, based on whether 15 emerging economies had prudential regulations and explicit capital control measures during 1990–96. These intensity indices do a better job of capturing gradual changes in capital account restrictions, but are considerably limited because their construction involves researchers’ subjective judgment of regulations.

Furthermore, none of these indices can measure de facto capital controls—only the presence or absence of de jure controls. In reality, enforcement of the controls has considerable influence on the “controlled” nature of the transaction. Often, actual implementation of the controls differs from the measures as described in the relevant laws and regulations, either setting more stringent conditions or not enforcing the controls at all, rendering the de jure controls ineffective. As a result, some researchers use de facto measures associated with outcomes of actual capital mobility. Among others, the most widely used measure is the sum of external assets and liabilities expressed as a percentage of GDP (Lane and Milesi-Ferretti, 2007). An example of measures that are both de jure and de facto in nature is the share of equities available to foreign investors (Edison and Warnock, 2003).

Although there has been little effort to construct cross-country comparable indices based on the “changes” section of the AREAER, the information it contains can help overcome some of the weaknesses described above. It allows chronological tracking of changes in capital controls in a given year, regardless of their significance. The information in the AREAER also allows more precise calibration of controls, since it includes the precise date of the change (month/day/year). Because the significance of changes is not easily comparable across countries, the use of the information contained in this section is more suitable to country-specific studies. Examples of capital control indices based on the changes section are found in Cardoso and Goldfajn (1998) for Brazil covering the period 1983–95; in IMF (2010) for Brazil, Colombia, Croatia, Korea, and Thailand for 2000–08; and for a larger set of liquidity-receiving economies for 2003–mid-2009.

Despite its limitations, the AREAER is the only database that can be used to measure capital account restrictiveness across a large number of countries. Various indices have been proposed, but the actual choice of index depends on the specific purpose of the application, because each index is different in its coverage and strength. When the existing indices fail to meet a specific research need, a look at the information in the text can help fill the gap.

References

    Alesina, Alberto, VittorioGrilli, and Gian MariaMilesi-Ferretti, 1994, “The Political Economy of Capital Controls,” in Capital Mobility: The Impact on Consumption, Investment and Growth, ed. by LeonardoLeiderman and AssafRazin (Cambridge, United Kingdom: Cambridge University Press), pp. 289313.

    Brune, Nancy E., 2006, “Financial Liberalization and Governance in the Developing World,” PhD dissertation, Yale University.

    Brune, Nancy E., and AlexandraGuisinger, 2007, “Myth or Reality? The Diffusion of Financial Liberalization in Developing Countries” (unpublished).

    Cardoso, Eliana, and IlanGoldfajn, 1998, “Capital Flows to Brazil: The Endogeneity of Capital Controls,”IMF Staff Papers, Vol. 45, No. 1, pp. 161202.

    Chinn, Menzie, and HiroIto, 2008, “A New Measure of Financial Openness,”Journal of Comparative Policy Analysis, Vol. 10, No. 3, pp. 30922.

    Edison, Hali J., and FrancisE. Warnock, 2003, “A Simple Measure of the Intensity of Capital Controls,”Journal of Empirical Finance, Vol. 10, No. 1–2, pp. 81103.

    Glick, Reuven, and MichaelHutchison, 2005, “Capital Controls and Exchange Rate Instability in Developing Economies,”Journal of International Money and Finance, Vol. 24,No. 3, pp. 387412.

    Grilli, Vittorio, and Gian MariaMilesi-Ferretti, 1995, “Economic Effects and Structural Determinants of Capital Controls,”IMF Staff Papers, Vol. 42, No. 3, pp. 5488.

    International Monetary Fund (IMF), 2010, “Global Liquidity Expansion: Effects on “Receiving” Economies and Policy Response Options,”in Global Financial Stability Report, World Economic and Financial Surveys (Washington, April).

    Johnston, R. Barry, and NataliaT. Tamirisa, 1998, “Why Do Countries Use Capital Controls?”IMF Working Paper 98/181 (Washington: International Monetary Fund).

    Klein, Michael W., and GiovanniP. Olivei, 2008, “Capital Account Liberalization, Financial Depth, and Economic Growth,”Journal of International Money and Finance, Vol. 27, No. 6, pp. 86175.

    Lane, Philip R., and Gian MariaMilesi-Ferretti, 2007, “The External Wealth of Nations Mark II: Revised and Extended Estimates of Foreign Assets and Liabilities, 1970–2004,”Journal of International Economics, Vol. 73, No. 2, pp. 22350.

    Miniane, Jacques, 2004, “A New Set of Measures on Capital Account Restrictions,”IMF Staff Papers, Vol. 51, No. 2, pp. 276308.

    Mody, Ashoka, and AntuPanini Murshid, 2005, “Growing Up with Capital Flows,”Journal of International Economics, Vol. 65, No. 1, pp. 24966.

    Montiel, Peter, and CarmenM. Reinhart, 1999, “Do Capital Controls and Macroeconomic Policies Influence the Volume and Composition of Capital Flows? Evidence from the 1990s,”Journal of International Money and Finance, Vol. 18, No. 4, pp. 61935.

    Potchamanawong, Pariyate, 2007, “A New Measure of Capital Controls and Its Relation to Currency Crises,” PhD dissertation, Claremont Graduate University.

    Potchamanawong, Pariyate, ArthurT. Denzau, SunilRongala, JoshuaC. Walton, and ThomasD. Willett, 2008, “A New and Better Measure of Capital Controls,” in The Design and Use of Political Economy Indicators: Challenges of Definition, Aggregation, and Application, ed. by KingBanaian and BryanRoberts (New York: Palgrave MacMillan), pp. 81102.

    Quinn, Dennis P., 1997, “The Correlates of Change in International Financial Regulation,”American Political Science Review, Vol. 91, No. 3, pp. 53151.

    Quinn, Dennis P., and A. MariaToyoda, 2008, “Does Capital Account Liberalization Lead to Growth?”The Review of Financial Studies, Vol. 21, No. 3, pp. 140349.

    Schindler, Martin, 2009, “Measuring Financial Integration: A New Data Set,”IMF Staff Papers, Vol. 56, No. 1, pp. 22238.

The summary features table of pre-1996 AREAER issues provides binary indicators for the following eight categories: bilateral payments arrangements with members and with nonmembers, restrictions on payments for current transactions, restrictions on payments for capital transactions, import surcharges, advance import deposits, repatriation requirement of export proceeds, and surrender requirement of export proceeds. In addition, there is an indicator for the existence of separate exchange rates for some or all capital transactions.

These are the categories of capital transactions: capital market securities, money market instruments, collective investment securities, derivatives and other instruments, commercial credits, financial credits, guarantees, sureties and financial backup facilities, direct investment, liquidation of direct investment, real estate transactions, and personal capital transactions. Provisions specific to the financial sector include measures related to commercial banks and other credit institutions and those related to institutional investors. Beginning in the 2006 issue, institutional investors are further broken down into insurance companies, pension funds, and investment firms and collective investment funds.

    Other Resources Citing This Publication