The high costs of volatility have prompted a search for ways to discourage short-term speculative capital movements without harming longer-term investments more in tune with good fundamentals. Taxes offer one possible tool. In the 1930s, John Maynard Keynes, disturbed by the casino-like behavior of the U.S. stock market, proposed a financial transactions tax to increase the cost of speculative activities and redirect energies toward productive objectives. In response to considerable currency and capital flow turbulence in the late 1970s, James Tobin suggested imposing a tax (the Tobin tax) on all transactions involving currency conversion.
In an IMF Working Paper titled Retarding Short-Term Capital Inflows Through Withholding Tax, Howell H. Zee of the IMF’s Fiscal Affairs Department adapts the narrower goals of the Keynes proposal and suggests that countries pursuing sound economic policies and seeking to cope with large and volatile capital flow movements impose a withholding tax on all private capital inflows. Such a tax, Zee argues, would be easy to administer and difficult to evade, and would thus be more effective than the reserve requirements employed by a number of countries to impede short-term capital flows.
Zee explains that the Tobin proposal, expressly designed to slow hyperactive international capital movements, taxes the amount of currency converted rather than the investment’s rate of return. The burden of such a tax varies inversely with the time period of the investment. Even a low nominal rate (1 percent or less) provides a substantial deterrent to short-term transactions but is of little or no consequence for longer-term investments.
Concerns about the Tobin tax arise, however, in terms of the practicalities of its application. The tax is designed to be administered on a universal and uniform basis on all currency conversions. Its critics contend the tax would generate large economic distortions on transactions unrelated to capital flows, require substantial international coordination to put an enforcement mechanism in place, and raise troublesome issues about the distribution of potentially large revenues among countries. Zee suggests these may be valid criticisms of a financial transactions tax that has a global goal. The criticisms are less apt if the transactions tax has the narrower national objective (and thus is more akin to the nature of Keynes’s proposal) of moderating the degree of volatility of capital flows into or out of a country.
The principal risk a country faces from liberalizing its capital account is a sudden and significant reversal in capital inflows that has little to do with the country’s own policies. Private speculators do not pick up the bill for the destabilizing impact of their decisions—the country does. In economic terms, this constitutes a negative externality for the country and, according to Zee, calls for a classic economic remedy—a tax on the activity that is generating the externality. He also noted that policymakers have increasingly acknowledged the potentially useful role that price-based measures (of which well-designed taxes are a prime example)—rather than quantitative controls on capital movements—can play in addressing volatile capital movements unrelated to economic fundamentals.
A financial transactions tax that is national in nature—what Zee terms a cross-border capital tax—would avoid the concerns about international coordination, enforcement, and revenue sharing that are associated with the Tobin proposal. A cross-border capital tax would not differentiate between capital flows that generate externalities and those that do not, but the tax burden on flows that do not generate externalities would be insignificant, as they typically have a much longer time horizon.
Cross-border capital tax
Zee proposes a cross-border capital tax that would subject all private financial inflows to a withholding tax at the point, and time, of their entry into the country. The amount withheld on such inflows that is unrelated to capital movements (primarily trade and income flows) would be credited against domestic tax liabilities, and excess credits would in principle be refundable. Most of the burden of the tax would fall on the short-term foreign borrowings of residents; export receipts and income from foreign sources would largely escape the tax.
A cross-border capital tax, Zee suggests, could be readily designed to require all financial institutions that handle funds transferred from abroad to serve as withholding agents; all taxes withheld on export receipts are refunded to exporters under the refund mechanism of a value-added tax; all taxes withheld on interest, dividends, royalties, repatriated profits, and other income flows are credited against domestic income tax liabilities; and the rate of the cross-border capital tax could be adjusted to the degree of disincentive needed.
The proposed tax could be implemented easily. It relies on a withholding mechanism that can be found in one form or another in almost all tax systems. By piggybacking on effective existing infrastructures, the new tax avoids the administrative headaches associated with designing and operating new systems and keeps start-up and administrative costs to a minimum.
Clearly, Zee admits, there will be minor administrative complications, notably in determining export credits and refunds in countries without value-added taxes and in handling paperwork for those who receive funds from abroad but are not otherwise required to file income tax returns. But this added burden would likely be counterbalanced by benefits beyond the obvious one of discouraging volatile short-term capital flows. A cross-border capital tax could at least partially address the increasingly vexing question of how to tax foreign-source capital income (for example, interest and dividends). With globally integrated financial markets, capital income from abroad is easy to earn but difficult to detect unless taxpayers voluntarily report it. A cross-border capital tax would serve, Zee says, as a final withholding tax, or a minimum income tax, on unreported income.
There is no hard-and-fast rule for determining the optimal rate of a cross-border capital tax, Zee adds. A rate as low as 1 percent imposes a heavy penalty on short-term investments and, if capital flows are high, yields significant revenue. Authorities might be tempted to view this tax as a revenue source, but he strongly cautions against it. Indeed, Zee suggests that some or all of the revenue should be shared among the withholding institutions. This would reimburse them for administrative costs and remove the temptation to use the tax for other objectives.
The withholding proposal offers an interesting contrast to the use of reserve requirements on short-term capital flows (such as Chile employed), Zee says. The two systems take fundamentally different approaches. Chile’s reserve requirements, which implicitly serve as a tax (in terms of the interest lost on reserves), necessitate the identification of types of capital flows liable for the reserve requirement. In practical terms, this targeted coverage has been fraught with problems, as borrowers and officials looked to find, or close, loopholes in the system.
A withholding tax largely sidesteps enforcement issues, because it covers all financial inflows and places the burden of proof on those who file for credits and refunds. It is inherently harder to evade (the fungibility of capital flows is immaterial when the tax is applied to all inflows), and easier and less costly to administer (since it relies on existing tax infrastructures).
Ultimately, Zee concludes, a cross-border capital tax offers an effective means of increasing the transaction costs of short-term capital movements without imperiling the attractiveness of longer-term investments. It will not shield a country from the consequences of unsustainable policies. A cross-border capital tax, he insists, would correct for market failures, not propagate policy failures.
Copies of IMF Working Paper No. 00/40, Retarding Short-Term Capital Flows Through Withholding Tax, by Howell H. Zee, are available for $7.00 each from IMF Publication Services. See page 162 for ordering information.