Appendix I. Market Access Developing Countries in Sample
Azerbaijan, Rep. of
China, P. R.
Iran, I. R. of
Trinidad and Tobago
Venezuela, Rep. Bol.
Cardenas, M. and F. Barrera, 1997, “On the Effectiveness of Capital Controls: The Experience of Colombia during the 1990s,” Journal of Development Economics, Vol. 54, Issue 1, pp 27–58.
Cardoso, Eliana and Ilan Goldfajn, 1998, “Capital Flows to Brazil: the Endogeneity of Capital Controls,” IMF Staff Papers, Vol. 45, No. 1, pp 161–202.
Cordella, Tito, 2003, “Can Short-term Capital Controls Promote Capital Inflows?” Journal of International Money and Finance, Vol. 22, pp. 737–745.
Dooley, Michael, 1995, “A Survey of Academic Literature on Controls over International Capital Transactions,” NBER Working Paper No. 5352.
Ghosh, Atish, 1995, “Capital Mobility Among the Major Industrialized Countries: Too Little or Too Much?” Economic Journal, 105, pp. 107–28.
International Monetary Fund, 2000, “Capital Controls: Country Experiences with Their Use and Liberalization,” International Monetary Fund, Occasional Paper No. 190.
Schadler, Susan, Maria Cakovic, Adam Bennett, and Robert Kahn, 1993, Recent Experiences with Surges in Capital Inflows, International Monetary Fund, Occasional Paper No. 108.
Ul Haque, Nadeem, Donald Mathieson, and Sunil Sharma, 1997, Causes of Capital Inflows and Policy Responses to Them, Finance and Development (March), pp. 3–6.
Periods of high global liquidity are identified based on the indicators in World Economic Outlook (April 2007).
The Chair’s Concluding Remarks on “Globalization, Financial Markets, and Fiscal Policy.”
The discussion is based on simple averages for 50 “market access” EMEs; see Appendix I for the list of countries.
See World Economic Outlook (October 2007) for more details.
The exposition does not discuss unorthodox policy responses; e.g., the possible role of tax policies not only in tightening fiscal policy, but also in terms of their effect on resource allocation. Some of these measures could target certain types of capital flows—as opposed to targeting all capital inflows (for example, real estate investment). The merits of such specific tax measures are two-fold: targeting overheating in a specific economic activity and reducing broader speculative activity and volatility. The drawbacks however, are also two—fold: they might lead to distortions in resource allocation, and they act as the equivalent of capital controls, with the usual attendant costs.
To the extent that the capital inflow is fueling the current account deficit (making the country a “blend” between Cases 1 and Case 2 above), reducing the availability of foreign savings (by imposing controls) or tightening prudential regulations on bank lending (if this is the source of external finance) could play a useful role.
Schadler et al. (1993) argue that fiscal consolidation was an important factor attracting capital flows to developing and emerging market countries in the late 1980s and early 1990s.
A recent World Economic Outlook chapter on capital inflows (October 2007) shows that fiscal restraint works well as a policy response to large inflows. Schadler et al. (1993) show that while fiscal restraint was advisable to prevent overheating and appreciation, but only one country (Thailand) out of the six in their study did so.
In Case 3 countries, if exports are mainly of nonrenewable resources and there are hysteresis effects of an appreciation on nontraditional exports, then it is likely that the authorities might seek to limit appreciation.
The activity index is constructed by combining inflation rates and output gaps into one index, where each component enters the constructed index with mean 0 and variance 1.
The results are based on an unbalanced panel dataset of 50 EMEs using annual data over the period 1999–2007.
Given the small sample size of Case 4 and Case 5 (see Table 1), these are excluded from the econometric estimations. Moreover, as previously noted, Case 1 through Case 3 are the most interesting for the purposes of this paper as they represent either cases with positive BOP pressures and/or cases with positive capital flows. In effect we are carrying out econometric estimations where the lagged values of other policy variables are added together with controls for the country’s cyclical position—debt-to-GDP ratio, short—term debt-to-reserves ratio, inflation, and exchange rate regime. The estimations include both country and time dummies.
Periods of high global liquidity are identified as described in the footnotes to Figure 4 and WEO (April 2007).
The exchange rate volatility indicator is defined as the absolute value of the monthly percentage change of the nominal exchange rate over the previous 12 months and averaged over a 12-month horizon (Ghosh and others, 2002). The flexibility indicator reflects the change in the exchange rate but without the absolute value.
The effect of monetary tightening on the current account balance is ambiguous as it would appreciate the exchange rate—leading to lower exports—but also dampen economic activity, lowering imports.
Monetary policy is measured by the change in economic growth and inflation relative to the change in broad money—an increase represent a tightening of monetary conditions.
Sterilization is represented by the effect of the change in net foreign assets (scaled by the level of reserve money) on the change in net domestic assets (also scaled by the level of reserve money).
Fiscal policy is represented by the fiscal impulse; that is, the change in the difference between the regular and the cyclically-adjusted general government balance—an increase represents a tightening of fiscal conditions.
The exchange rate volatility indicator is defined as the absolute value of the monthly percentage change of the real exchange rate over the previous 12 months and averaged over a 12-month horizon. The flexibility indicator also reflects the change in the real exchange rate but without the absolute value.
The results described in this section hold even more strongly when a reduced sample is used. Such sample is intended to focus the empirical work on the “most extreme” observations for each case. The sample includes the one-third of all the observations in each case that are farthest from the boundaries of each case. In effect, as noted in Figure 3, the reduced sample drops observations that are on either side of the diagonals or close to the origin.