The Benefits of Financial Transactions Taxes
PDF | View on original website
I appreciate the opportunity to express my views to the members of the Bundestag on financial transaction taxes (FTT). I apologize for not being able to attend in person. A long-planned vacation made it impossible to make this meeting date, so I am especially thankful for your willingness to accept a written statement from me instead.
I will make three main points in my statement.
I will address these points in more detail below.
The Economic Argument for FTTs
There are several reasons why an FTT is desirable from an economic standpoint. First and foremost it is an effective way to raise a large amount of revenue. Since the tax is being imposed on a relatively small number of actors, it is relatively easy to administer. This best shown by the example of the United Kingdom, which has imposed a tax on stock trades for decades. It requires that the broker who conducts the trade collect the tax. The UK’s Inland Revenue Service reports that the tax on stock trades is administratively the most efficient tax to collect. Its costs are less than 0.05 percent of the revenue collected. By comparison, the administrative costs of collecting the income tax are more than ten times as high, being equal to 0.7 percent of the revenue collected.
The experience of the UK with its stock transfer tax belies the often-repeated claims that financial taxes will be difficult or impossible to collect. Prior to the downturn, the UK collected 4 billion pounds a year from this tax, an amount that is just under 0.3 percent of GDP. Japan was also able to raise a substantial amount of revenue through a set of financial transactions taxes that it had in place in the 80s, at the peak of its stock bubble drawing 4.0 percent of federal tax revenue through this source. Other countries have had comparable success in raising revenue through financial transactions taxes. In short, there really is no basis for the debate about the administrative feasibility of FTTs since many countries have raised large amounts of revenue over long periods through such a tax.
In addition to presenting the possibility of raising large amounts of revenue, FTTs may also improve the operation of financial markets. The most obvious way in which they can have this effect is by reducing the volume of trading. A large and growing share of the economy’s resources are being used in the creation and trading of financial assets. In the United States, the share of private sector output that goes to the narrowly defined financial sector (security and commodity trading and investment banking) quintupled between the mid-seventies and the last decade, going from 0.5 percent of private sector output to more than 2.5 percent. There have been comparable increases in the finance share in most other countries, even if not quite as large as in the United States.
Finance is an intermediate good, like trucking. Unless this enormous expansion of the financial sector was associated with a better end result (e.g. more effective allocation of capital or more secure savings) then it is difficult to see any economic benefit from the increased resources being consumed by this sector. It is as though we had five times as many people and trucks used in the trucking sector, but no improvement in delivery times or other identifiable service benefit.
If a financial transactions tax can reduce the volume of trading without interfering with the sector’s ability to allocate capital, then it will have made the economy more efficient. The resources that would have otherwise been committed to the financial sector can instead be employed in a sector where they can have measurable economic benefit. This is potentially a very important if overlooked benefit. Taking the example of the United States, if a financial transactions tax could reduce the size of the financial sector by a third, then it would effectively increase economy-wide productivity by 0.8 percent. This is an impressive accomplishment. Policymakers often labor at great length over policies that are not expected to have an effect even one-tenth as large.
The other way in which a financial transactions tax can improve the operation of financial markets is by reducing the potential rents in short-term trading that serve no productive purpose. By its nature, FTTs will have far more impact on short-term trading than long-term investing. The logic is straightforward: the expected profits on a short-term transaction (hours, days, or weeks) are much smaller than on a longer-term investment that is expected to be held for a period of years. This means that an increase in transactions costs, as a result of a financial transaction tax, will represent a much larger share of the expected profits from short-term trades than longer-term investments.
The consequence of the change in incentive structure should be to get actors in financial markets to focus more on longer-term investment opportunities rather than opportunities for short-term gains. In principle, financial markets will more effectively allocate capital in ways that support growth if major actors focus on seeking long-term investment opportunities. For this reason also, FTTs can be expected to improve the operation of financial markets.
Volatility of Asset Prices
It is also possible that FTTs will reduce the volatility of financial markets, although the evidence on the relationship between volatility and transactions taxes is quite mixed. As noted above, an FTT will raise transactions costs and therefore reduce the volume of trading.
Some of the transactions that will be discouraged by an FTT will be stabilizing. For example, arbitragers trade to take advantage of small differences in prices between different markets. Insofar as arbitragers reduce their trading as a result of the tax, there can be somewhat larger gaps in prices between markets that go uncorrected. This can lead to more volatility in prices and further divergences between market prices at a point in time and fundamental values.
However, higher volume trading can also be destabilizing. If traders act, not on their own assessment of fundamentals, but rather in response to the trades of others, then more trading can amplify fluctuations away from fundamental values. In this case, an FTT can be expected to reduce volatility. There is an extensive literature on this theory of “noise trading” which dates back to Keynes. More recently it has been developed in a series of papers that included Lawrence Summers, the former U.S. Treasury Secretary and current head of the National Economic Council, as a co-author. Other scholars have continued to follow in this framework.
Efforts to measure the impact of FTTs, transactions costs, and trading volume on volatility have produced mixed results. Several studies have found that higher transactions costs are associated with higher volatility. However, other studies have found no significant relationship between volatility and transactions taxes. In an interesting study, Roll (1989) found that markets with transactions taxes had no greater volatility in the period around 1987 stock market crash than markets without a tax. Another study found that high transactions taxes increased volatility, however smaller transactions taxes (less than 0.5 percent) had no significant effect on volatility.
One notable study found evidence that more trading does increase volatility by comparing the volatility of the New York Stock Exchange over three-day periods where the market was open every day with three-day periods in 1968 that included a Wednesday where the NYSE was shut to deal with paperwork. Volatility was considerably higher in the former set of days than the latter, suggesting that trading, rather than events in the world, was leading to market fluctuations. This study concluded that at least a portion of this volatility was due to noise trading, where traders responded to market momentum or other noise rather than making an independent assessment of market fundamentals. If this noise trading is discouraged by higher transactions costs, then an FTT would reduce volatility.
It is possible that an FTT can both increase some types of volatility and reduce others. Specifically, it may lead to a situation in which somewhat larger divergences from fundamentals may go uncorrected. However, it may also lead to situations that are less prone to the sort of speculative run-ups that we have seen in markets in recent years. The latter would be difficult to test, since unambiguous examples of speculative price movements are relatively rare.
Finally, it is important to remember that the tax rates currently being considered are relatively modest. Trading costs have fallen sharply over the last three decades due to improvements in computer technology. Therefore, even with FTTs in place, transaction costs would only be rising to their level of two decades ago in most markets. Insofar as high transactions costs actually do increase volatility, the set of FTTs being considered would have the same impact on volatility as did the 80s level transactions costs.
Critics of FTTs have argued that they will reduce the levels of liquidity in financial markets, which could make it more difficult for investors to sell assets. There is some truth in this claim, but it is likely to be of very little consequence in almost all cases.
An FTT will raise transactions costs and therefore reduce trading volumes. This means that markets will be somewhat less liquid (i.e. people may have to wait longer before they can sell shares of stock, bonds, or other assets). However, it is important to recognize that since transactions costs will just be driven up to their 80s level, there is no reason to believe that markets will be any less liquid than they were in the 80s. Of course, the United States and other wealthy countries already had vibrant capital markets in the 80s, so it would be difficult to argue that the lower levels of liquidity would impose any major hardship on ordinary investors.
It is also worth noting that the tax will have a smaller impact on less liquid assets than on highly liquid assets for the simple reason that less liquid assets already have higher transactions costs. For example, the transactions costs for trading the stock of small companies listed on the NASDAQ can be several times larger than the cost of trading the stock of large companies. This means that adding a fixed amount to the transaction cost through a tax (e.g. 0.25 percent) will impose a much smaller percentage increase in the transaction costs for a less often traded stock, therefore it should have less impact on its liquidity.
A similar relationship applies in currency markets. The transactions costs on trades between euros and dollars may be less than 1 basis point for very large trades. By contrast, transactions costs on less widely traded currencies can easily be twenty of thirty times larger. In this context, the imposition of a modest transaction tax (e.g. 1-5 basis points) would lead to a substantial percentage increase in costs in the euro-dollar trade, but still likely leave it a highly liquid market. It will lead to a much more modest percentage increase in transactions costs or less frequently traded currencies. Therefore its impact on liquidity in these markets is likely to be limited.
As a basic proposition the greatest reduction in trading volume will occur in assets that have the most liquid markets, leaving them still quite liquid. The impact of the tax on trading volumes in markets that are already not very liquid will be considerably smaller.
Critics of FTTs have also argued that they will reduce the efficiency of financial markets. This means that they may be slower in responding to changes in market fundamentals. Insofar as markets are less efficient, this could mean, for example, that the price of oil stocks will rise less rapidly in response to news of a disruption in supplies from a major producer. (Baltagi, et al., 2006) have found some evidence for this sort of reduction in efficiency.
While lower trading volume could in fact lead to less efficiency (it can also lead to more efficient markets if it reduces noise trading), it is important to realize the limited impact of this effect. Again, since trading costs would just be pushed up to their 1980s level, markets will be no less efficient than they were in the 80s. As a practical matter, the issue is likely to be a question (using the oil stock example) of whether prices adjust in a single day or over a couple of days. It is difficult to imagine that this sort of delay in price adjustment would have any major repercussions for investment and real economic activity.
Support for Financial Transactions in the Economics Profession
As noted earlier, there have long been prominent proponents of financial transactions taxes in the economics profession. Keynes argued the case for FTTs in The General Theory, believing that markets worked best when investors sought long-term investment opportunities rather the pursuit of short-term profit. He famously argued that when capital markets begin to resemble casinos, they are unlikely to do serve their economic function well.
More recently economists have picked up and advanced this argument. Nobel Laureate James Tobin argued for what came to be known as a “Tobin Tax” in the context of taxing foreign exchange transactions. Tobin wanted to “put sand in the wheels,” slow down the volume of trading in these markets. He was concerned that movements away from equilibrium could become self-reinforcing, leading to unnecessary instability in financial markets. In 1989, Nobel Laureate Joseph Stiglitz wrote an article outlining the case for a financial transactions tax. His main concern was to reduce the volume of short-term trading. Lawrence Summers, the future Treasury Secretary, wrote an article in the same journal issue also arguing the case for a FTT.
There is a substantial body of ongoing research outlining the case for financial transactions taxes. It continues to be the case that a substantial number of economists support FTT, including many of the most prominent members of the profession. In addition to Stiglitz and Tobin, up until his death, Nobel Laureates Paul Krugman and Dennis McFadden also have publicly endorsed the tax. Jeffrey Sachs, the director of the Earth Institute at Columbia University and one of the world’s leading development economists, is another prominent supporter of financial transactions taxes. A recent economists sign-on letter, circulated in the United States, gathered more than 200 signatures from economists at colleges, universities and research centers across the country.
Support for Financial Transactions Taxes from Civil Society
In the wake of the financial crisis there has been a huge surge in support for financial transactions taxes from civil society groups around the world. Large grassroots organizations in support of financial transactions taxes have arisen in most major wealthy countries. These organizations see these taxes as a way to raise revenue to meet aid commitments to developing countries and also to cope with the cost of reducing greenhouse gas emissions. In addition, a portion of the tax can also be used to offset the costs of financial bailouts resulting from the crisis as well the increased public indebtedness resulting a sustained period of high unemployment.
In the United States, FTTs have received a substantial amount of renewed interest as a result of the crisis. Both of the countries major labor federations, the American Federation of Labor-Congress of Industrial Organizations and Change to Win, have endorsed FTTs. Many of the member unions have also made FTTs a top agenda item. Dozens of smaller organizations across the country have also endorsed FTTs, hoping to have the revenue both as a source of financing for general government and also to meet important development goals. There also have been bills introduced in both houses of Congress calling for FTTs.
With the United States government facing substantial budgetary pressures in the years ahead, it is very likely that FTTs will stay in the public debate. Most of the proposals for dealing with the deficit involve substantial cuts in Social Security and other popular programs as well as increases in regressive taxes, such as a value-added tax. In this context, a FTT is likely to have considerable appeal and remain on the national agenda in the years ahead.
There is a long history of financial transactions taxes around the world. They have often raised substantial amounts of revenue. There are no major administrative obstacles to implementing a FTT.
The effect of FTTs on financial markets is likely to be limited in terms of affecting their liquidity, volatility, and efficiency. The size of the taxes under consideration will only raise transactions costs back to the levels that they were at 20 to 30 years ago. For this reason, the increase in transactions costs will not make the markets operate better nor worse than they did in these decades. The tax will undoubtedly lead to a substantial decline in trading. As a result, much or all of the cost of the tax will be offset by a reduction in the volume of transactions, leaving the total cost of trading to non-financial actors little affected.
Finally, if Germany were to implement FTTs through the cautious path currently being considered in the Bundestag, it is likely to find itself part of a worldwide movement in this direction. There are substantial forces pushing for FTTs in most major wealthy countries including the United States. The actions of the German government can add momentum to this movement.
 See Bond, S., M. Hawkins, and A. Klemm. 2004. “Stamp Duty on Shares and Its Effect on Share Prices,” London: The Institute for Fiscal Studies, WP04/11 (p 4).
 Japan Securities Research Institute, 1992. Securities Markets in Japan, 1992. Tokyo: Japan Securities Research Institute.
 A fuller discussion of the research on this issue can be found in Pollin, R., D. Baker, and M. Schaberg, 2002. “Financial Transactions Taxes for the U.S. Economy,” Amherst: MA: Political Economy Research Institute, [available at http://www.peri.umass.edu/236/hash/aef97d8d65/publication/172/].
 See for example DeLong, J.B., A. Shleifer, L. Summers, and R. Waldmann, 1987. “The Economic Consequence of Noise Trading,” National Bureau of Economic Research, WP # W2395.
 See Schulmeister, Stephan, 2010. “Boom-Bust Cycles and Trading Practices in Asset Markets, the Real Economy and the Effects of a Financial Transactions Tax,” WIFO Working Papers 364/2010, available at http://www.wifo.ac.at/wwa/jsp/index.jsp?fid=23923&id=38641&typeid=8&display_mode=2.
 For example, see Jones, C. and P. Seguin, 1997. “Transactions Costs and Price Volatility: Evidence from Commission Deregulation, American Economic Review, 728-37.
Hu, S. 1998. “The Effects of the Stock Transaction Tax on the Stock Market: Experiences from Asian Markets,” Pacific Basin Financial Journal, 347-364.
 Roll, K. 1989. “Price Volatility, International Market Links, and Their Implications for Regulatory Policy,” Journal of Financial Services Research, 211-246.
 Lindgren, R. 1994. “Transactions Taxes and Stock Market Volatility,” Stockholm School of Economics, Working Paper Series in Economics and Finance, # 25.
 French, K. and R. Roll, 1986. “Stock Return Variances: The Arrival of Information and the Reaction of Traders,” Journal of Financial Economics, 5-26.
 Tobin, J., 1978. “A Proposal for International Monetary Reform,” Eastern Economic Review, pp 153-159.
 Stiglitz, J., 1989. “Using Tax Policy to Curb Speculative Short-Term Trading,” Journal of Financial Services, Research, pp 101-115.
 Summers, L. and V. Summers, 1989. “When Financial Markets Work Too Well: A Cautious Case for a Securities Transactions Tax,” Journal of Financial Services, Research, pp 261-286.
 Economists Sign-on Letter, 2009. “An Open Letter in Support of Financial Transactions Taxes.” Available at http://www.cepr.net/documents/FTT-support.pdf.