Over the past twenty-five years, a number of countries, mostly in Latin America, have imposed taxes on banking transactions. Empirically, these taxes have been shown to be effective in generating revenue in the short term. Recent studies, however, question their ability to be a reliable source of revenue in the medium term and show that they result in significant financial disintermediation.
When, in 1898, the U.S. government introduced a two-cent tax on bank checks to finance the Spanish-American War, little did it know that a century later many Latin American countries would use similar taxes to fight their own fiscal battles. Since 1976, taxes on bank transactions have been introduced repeatedly in Argentina, Brazil, Colombia, Ecuador, Peru, and Venezuela. As of the end of 2006, such taxes were in effect in seven Latin American and Caribbean countries: Argentina, Brazil, Bolivia, Colombia, the Dominican Republic, Peru, and Venezuela.
There is considerable variation across countries in the design of these taxes. Usually, taxes on bank transactions are levied on withdrawals from bank accounts, including the clearance of checks and the use of automated teller machines (ATMs), as well as on payments of loan installments. In addition, in Argentina, Colombia, and Ecuador, these taxes have been levied on all, or some, bank credit transactions. There is also considerable diversity in what these taxes are called, with names including bank debit taxes, bank account debit taxes, and financial transaction taxes.
Bank transaction taxes (BTTs) were typically introduced at a time of crisis to quickly generate a burst of revenue. There are several reasons why these taxes are quite appealing as fiscal instruments, especially at times of fiscal distress. First, collection and administration costs of bank transaction taxes are minimal, since the banks simply transfer a portion of each qualified transaction to a government account. Second, the government enjoys an immediate and continuous revenue stream, since these taxes are collected from transactions in real time. Third, there is not much organized popular opposition to these taxes, since they are typically introduced as a “tax on the rich.” Finally, and most importantly, the short-term revenue performance of bank transaction taxes is usually quite strong. In their descriptive paper, Coelho, Ebrill, and Summers (2001) confirm that BTTs have been successful in raising revenue in the short term. They also document abundant anecdotal evidence, however, that these taxes have resulted in financial disintermediation. The authors argue that BTTs should be avoided unless the country has significant fiscal needs that cannot be met by more appropriate tax instruments.
Kirilenko and Perry (2004) estimate the degree of disintermediation (a permanent erosion of the tax base) resulting from the introduction of a BTT. They construct monthly series for real BTT revenue (nominal revenue adjusted for inflation and changes in the tax rate) for six countries (Argentina, Brazil, Colombia, Ecuador, Peru, and Venezuela) assuming that during the first month after the introduction of the tax, there is no change in the behavior of providers and users of banking services. This assumption allows them to use the decline in revenue from its starting value (in the first month the tax is in effect) as an estimate of financial disintermediation. The authors show that, on average, the introduction of a BTT results in disintermediation of between 4 and 44 cents for every dollar in revenue. According to their estimation, for every dollar raised through these taxes, financial disintermediation has reached maximum values of 46 cents in Argentina, 58 cents in Brazil, 64 cents in Colombia, 48 cents in Ecuador, 66 cents in Peru, and 49 cents in Venezuela. The authors also find that disintermediation effects tend to cumulate as the taxes remain in place, providing additional evidence for the hypothesis that these taxes quickly erode their own tax bases.
Baca-Campodónico, de Mello, and Kirilenko (2004) employ the persuasiveness of a full-fledged panel study to analyze the revenue-raising ability of bank transaction taxes. The authors use a panel of quarterly data from six Latin American countries that have had BTTs during the last nine years—Argentina, Brazil, Colombia, Ecuador, Peru, and Venezuela. They find that, for a given tax rate, BTT revenue declines in real terms over time (owing to erosion of the tax base). Therefore, in order to meet a given revenue target, the tax rate needs to be raised repeatedly. They also find, however, that a 0.1 percentage point increase in the statutory tax rate reduces the revenue base (or productivity) by 0.18–0.30 percent. Thus, increasing the tax rate erodes the tax base by more than it raises revenue and accelerates the speed at which the tax base is being eroded. They conclude that these taxes should be used only as a temporary means to mobilize revenue in situations of fiscal duress.
While the studies already discussed look at a cross section of countries, a complementary strand of literature focuses on experiences of individual countries with bank transaction taxes. For example, Albuquerque (2003) argues that a BTT increases the cost of government borrowing in Brazil. He calculates the inflection point of a Laffer-type curve adjusted for the higher cost of government borrowing and argues that an 80 percent increase in the actual average tax rate during the period (0.34 percent) to the estimated maximum rate (0.62 percent) would have yielded only about 36 percent more in revenue in 2000. The author also estimates that the losses for the actual and the calculated maximum tax rate are 21.5 percent and 57.8 percent of revenue (net of estimated higher government borrowing costs), respectively. In light of this evidence, he seeks to discourage the use of this tax.
Suescun (2004) presents a multisector, dynamic-general-equilibrium model that is used to evaluate the distortions owing to a transaction tax. Using calibrated data from Brazil, he argues that “a transaction tax imposes an intermediate level of distortions. In some cases, it is the best option after a consumption tax … in all model specifications, the growth toll of a transaction tax is small, even comparable to that of a consumption tax.”
Arbeláez, Burman, and Zuluaga (2002) use panel-data analysis to estimate the effect of a BTT on interest margins and profitability of 43 financial institutions in Colombia during 1995–2001. They find that the BTT increased the cost of credit and led to significant disintermediation. As a result, profits of financial institutions declined in the short term by more than the amount of revenue raised by the government. The authors consequently recommend abolishing the tax.
Lastrapes and Selgin (1997) investigate the impact of a two-cent check tax on the U.S. economy during the 1930s. They estimate the impact of the tax on the currency-deposit ratio and the money stock using a vector autoregressive model and monthly data from August 1921 to December 1936. They show that the tax led to significant disintermediation. As a result, the monetary contraction in the United States during the 1930s is estimated to have been 15 percent higher than it would have been without the tax. The authors also argue that policymakers were aware of the likely adverse effect of the tax but deliberately chose to overlook it in order to raise revenue. They present the two-cent check tax as a typical example of the Depression-era policies that disregarded the impact of fiscal measures on monetary and financial outcomes.
There is also a broad body of literature on the taxation of financial intermediation. For example, a comprehensive volume edited by Honohan (2003) presents an analytical overview of the four main types of tax bases for the financial sector: income, expenditures, assets, and transactions. In the introduction to the volume, the editor points out many theoretical considerations and practical hurdles that need to be addressed in designing a tax system for the financial sector.
Finally, there is a large related body of literature on corrective financial transaction taxes. It argues that these taxes can correct distortions in financial markets to some extent. This argument runs counter to a view that that these taxes cause greater financial disintermediation and damage revenue-mobilization capacity more than they help ease the distortions that they are purported to rectify. Habermeier and Kirilenko (2003) examine research on market microstructure, asset pricing, rational expectations, and international finance with a view to assessing the impact of securities transaction taxes on financial markets. They argue that transaction taxes can obstruct price discovery and price stabilization, increase volatility, reduce market liquidity, and inhibit the informational efficiency of financial markets.
Overall, the literature on bank transaction taxes shows that these taxes have been quite effective in generating revenue in the short run. Empirical evidence suggests, however, that bank transaction taxes are distortionary, have contributed to significant financial disintermediation, and are not a reliable source of revenue in the medium run.
- Search Google Scholar
- Export Citation
AlbuquerquePedro H.2003“BAD Taxation” (unpublished; Laredo, Texas: Texas A&M International University).
ArbeláezMaria A.LeonardE. Burman and SandraZuluaga2002“The Bank Debit Tax in Colombia” (unpublished; Washington: Urban Institute).
Baca-CampodónicoJorgeLuiz deMello and AndreiKirilenko2004“The Rates and Revenue of Bank Transaction Taxes” (unpublished; Washington: International Monetary Fund)
CoelhoIsaiasLiamP. Ebrill and VictoriaSummers2001“Bank Debit Taxes in Latin America: An Analysis of Recent TrendsIMF Working Paper 01/67.
HabermeierKarl F. and AndreiA. Kirilenko2003“Securities Transaction Taxes and Financial Markets,”Staff PapersInternational Monetary FundVol. 50No. 2 pp. 165–80.
HonohanPatrick2003“Avoiding the Pitfalls in Taxing Financial Intermediation” in Taxation of Financial Intermediation: Theory and Practice for Emerging Economiesed. by PatrickHonohan (Washington and New York: Oxford University Press for the World Bank) pp. 1–28.
KirilenkoAndrei and VictoriaJ. Perry2004“On the Financial Disintermediation of Bank Transaction Taxes” (unpublished: Washington: International Monetary Fund).
LastrapesWilliam D. and GeorgeSelgin1997“The Check Tax: Fiscal Folly and the Great Monetary Contraction,”Journal of Economic HistoryVol. 57 pp. 859–78.
SuescunRodrigo2004“Raising Revenue with Transaction Taxes in Latin America—Or Is It Better to Tax the Devil You Know?” World Bank Policy Research Paper No. 3279 (Washington).
Jacques Polak Seventh Annual Research Conference
The International Monetary Fund will hold the Jacques Polak Seventh Annual Research Conference at its headquarters in Washington on November 9–10, 2006. The conference is intended to provide a forum for discussing innovative research in economics, undertaken by both IMF staff and outside economists, and to facilitate the exchange of views among researchers and policymakers.
The theme of this year’s conference is “Capital Flows.” Topics include (but are not restricted to) (1) the role of capital flows in economic development and growth; (2) booms and busts in capital flows to emerging market countries, and policies that can mitigate their effects; (3) global financial imbalances and their resolution; (4) the impact of capital mobility on domestic policies and institutions; and (5) the determinants of the direction, size, and structure of international capital flows.
This year, the Mundell-Fleming Lecture will be given by Olivier Blanchard (Massachusetts Institute of Technology). Further information on the conference program will be posted on the IMF website (www.imf.org). The Organizing Committee includes Gianni De Nicoló, André Faria, Olivier Jeanne (chair), Jaewoo Lee, Andrei Levchenko, Eduardo Ley, and Nikola Spatafora.