Report | Jobs and Unemployment

Paying for the Plan | American Jobs Plan

Download PDF

Paying for the American Jobs Plan with a financial transactions tax

Enactment of the American Jobs Plan would have immediate benefits for employment, but as long as the impact of the recession continues to weigh on the job market next year and through 2011, government economic policy must remain expansionary. When unemployment moderates and the economy has rebounded, it will be appropriate to address the federal budget deficit. The spending required by the American Jobs Plan would occur within the first two years after its enactment; in years three through 10, all of this spending could be recouped through a financial transactions tax.

Proposals for a financial transactions tax (FTT), such as one by Congressman Peter DeFazio (D-Ore.), have been met with strong opposition, especially from Wall Street. This opposition is misplaced. An intelligently designed financial transactions tax should be a key item on the policy menu. Those concerned about the state of the job market today and the state of the deficit tomorrow should embrace a proposal that calls for increased action to boost employment in the next two years that is paid for with the implementation of an FTT. The economic bottom line is that a financial transactions tax is a progressive revenue-raiser that is likely to be either efficiency-neutral or even efficiency-enhancing. Few other revenue-raisers can make this claim.

What would a well-designed financial transactions tax look like?

A financial transactions tax would levy a small “sales tax” on the transfer of ownership of a financial asset. For stock transfers, a rate of 0.5% (possibly split between buyer and seller) would raise significant revenue while not causing harmful economic effects. The rate would apply to and vary for other asset classes to provide disincentives for investors to arbitrage away the tax by moving out of stocks and into other instruments.

To put a 0.5% transaction tax in perspective, a typical state sales tax on goods and services is about 5.0%, or 10 times our proposed tax.

Taxing the simple transaction, rather than calculated proceeds, reduces the scope for disguising proceeds to avoid the tax. Also, taxing each transaction at a fixed rate implicitly imposes a higher tax on short “roundtrips,” that is, rapid (hourly or daily) turnovers of asset ownership. These short roundtrips generally do not provide large profits for investors, so the fixed transactions costs will impose a comparatively heavy fee. Conversely, if an asset is bought and held for an extended period, the transactions fee will likely constitute a small percentage of the total proceeds. This structure usefully biases the incidence of the tax toward speculators and away from long-term investors.

Financial transactions taxes exist today

Several countries have implemented examples of well-functioning financial transactions taxes. In the United States, the Securities and Exchange Commission currently levies a 0.00257% fee on stock transactions to fund its operating expenses. Japan had a 0.3% sales tax on stock transactions that raised substantial revenues (up to 4% of GDP during the stock market bubble of the late 1980s) before it was repealed in 1999. The United Kingdom currently imposes a 0.5% stock “stamp tax” on each trade on the London stock exchange.

The existence of the U.K. stamp tax belies an important argument often made against the tax in the United States—that it will harm the international competitiveness of U.S. stock exchanges and send companies fleeing to foreign ones. The fact that the London stock exchange is one of the largest in the world (in the top five in value of domestic market capitalization and total value of share trading, according to the World Federation of Exchanges Database) suggests that it has not been hurt unduly by the stamp tax. Furthermore, the U.K. tax means that the United States has more room to impose its own tax without losing competitive ground.

Lessons on designing a tax

The United States can learn from the rest of the world’s experience with financial transaction taxes. In the U.K., payment of the stamp tax is necessary to prove that ownership of the asset has legitimately changed hands. This feature gives it a built-in enforcement device: those who try to evade the tax will have trouble enforcing their claim to asset ownership.

On the other hand, the U.K. stamp tax is often criticized for taxing only stock transfers and thereby leaving a wide range of other asset markets unaffected. Those looking to evade the tax can engineer financial substitutes for stock transactions in these other asset markets. A well-designed tax would tax other asset classes as well as emerging asset classes at rates in line with the stock rate.

How much money could a financial transactions tax raise?

The U.K. stamp tax raised an amount equal to 0.2% of U.K. gross domestic product in the latest year analyzed; a similar share in the United States would equate to $30 billion. A 2004 Congressional Research Service study found that a 0.5% tax based on stock transactions could raise $79 billion annually. A review for the Canadian government in 1996 found that six of the nine countries surveyed collected revenue greater than 0.6% of their total GDP, equivalent to $85 billion in the United States in 2009.

A financial transactions tax could raise considerably more than these estimates—0.8% to 1.6% of GDP according to a 2002 study—by taxing a wider range of assets than stocks. In 2009 that range would amount to $113-226 billion. In short, the tax can be a significant revenue-raiser.

Who would pay the financial transactions taxes?

Probably the best guess as to who will be subject to the tax is to look at who owns financial assets now. A 2009 analysis of data from the Survey of Consumer Finance shows that the mean holdings of financial assets by the wealthiest 10% of households is 45 times greater than the mean holdings of the bottom 75% of households. In terms of wealth classes, the effect of the tax would therefore be extremely progressive and would help to reverse somewhat the decline in the progressivity of the federal income tax code that occurred between 2000 and 2006.

To make the incidence of an FTT more comparable to other taxes, one can look at the distribution of financial assets across income, not wealth, classes. By this measure, an FTT scores very well on progressivity even when stacked against overall federal taxes and even what is commonly regarded as the most progressive element of the current tax code – the federal estate and gift tax (see Figure A).

[Figure A]

What would be the economic impact of an FTT?

Any newly introduced tax will change behavior and affect economic efficiency. The impact of an FTT depends greatly on how one views the current structure of U.S. financial markets. If one assumes that financial markets are efficient and allocate capital optimally, then the tax will reduce this efficiency by curbing financial activity at the margin. But the current upheaval in global financial markets has undermined this “efficient markets” view of finance and given credence to the “noise trader” approach, which argues that financial markets are prone to speculation, herd behavior, and excess volatility. If this is true, then marginal transactions in financial markets may have zero or even negative social utility.

This noise trader view is bolstered by examining two trends of the past three decades (see Figure B). While the financial se
ctor takes up a larger and larger share of the U.S. economy over that time, increasing by almost 3 full percentage points of GDP (its share was 75% larger in 2009 than in the late 1970s), real investment in plants and equipment (plus financial services exports) is actually less. Given that one of the core functions of the financial sector is allocating capital efficiently, it seems odd that more and more resources are demanded by this sector to do the same amount of capital allocation.

[Figure B]

The academic research to-date shows mixed results of raising transactions costs (including FTTs) on asset trading and overall volatility. While there is little in this evidence to argue strongly that a first-order benefit of an FTT would be to rationalize financial markets dominated by noise traders, there is also nothing in this evidence to suggest that an FTT would cause any economic harm.

Policy recommendation

We recommend a modest 0.5% tax on financial transactions to pay for the American Jobs Plan. The tax would take effect three years after the implementation of the plan, and at year 10 it would cover the plan’s entire cost, making it deficit-neutral.

Jobs created by the American Jobs Plan

Each of the five points of the American Jobs Plan would have a significant impact on job creation. Table 1 shows the estimated first-year impact of each element of the plan.

1. Safety net spending. The investments in safety net programs (unemployment insurance and COBRA) would create jobs indirectly because those receiving assistance will have greater disposable income to spend on goods and services. An additional $110 billion in these programs would increase nationwide employment by 931,000 jobs.

2. Relief to state and local governments. Fiscal relief to state and local governments would prevent many public sector layoffs and create jobs in the private sector. States are currently facing several budget crises. For fiscal year 2011 (which starts in July 2010), state and local governments are facing a $182 billion shortfall, even after including $38 billion in Recovery Act assistance for that year. Without assistance, they will be forced to fire public sector employees (such as teachers and first responders), cut programs, raise taxes, or some combination of the three. All of these adjustments would also harm private sector employment, since the reduction in disposable income for those laid off would lead to less consumption of goods and services. Private sector employment would be further harmed as private firms that directly deliver services based on state funding—organizations such as hospitals, nursing homes, and construction contractors—are forced to cut back as well. Additional assistance of $150 billion to state and local governments would increase employment by about one million jobs. About half of this employment impact would be in the private sector.

3. Investments in schools and transportation. The Recovery Act contained additional funding for the nation’s infrastructure, including roads, bridges, and waterways. However, one element that was initially included but then stripped out of the final version was money for improvements to school buildings. Adding $30 billion to restore school facility funding and adding to other infrastructure investments would boost employment by 239,000 jobs.

These three components of the American Jobs Plan—each of which is an extension or augmentation of the Recovery Act—would in total boost employment by 2.2 million jobs. However, given the severity of the downturn, we must also look to new policy to spur employment in the private sector and to employ people directly to provide public services.

4. Public service jobs. The United States has a long history of direct public service employment in times of economic crisis, and a new program today could be targeted at distressed, high-unemployment communities and provide services such as environmental clean up, community policing, before- and after-school care of children, demolition or boarding up of abandoned houses and buildings, and parks improvements. A $40 billion investment in public service employment would employ an estimated 1 million people.

5. Job creation tax credit. The job creation tax credit would provide a temporary credit to firms that expand employment. The approach would give businesses a credit of 15% of expanded payroll costs in 2010 and 10% in 2011. The credit could spur an estimated 1.4 to 2.8 million new jobs in 2010 and 1.1 to 2.3 million in 2011.

Taken together, the gross cost of the five components of the American Jobs Plan is roughly $400 billion in the first year. However, the net cost of the plan would be much lower; we estimate that roughly 40 cents of each dollar spent on the plan would be recouped through higher revenue induced by increased economic activity, as well as reduced spending on existing safety net programs.

In total, the extension and augmentation of the Recovery Act together with the public service employment proposal would add about 3.2 million jobs to the economy that would otherwise not exist. A final estimate of the American Jobs Plan is more difficult because of how the tax credit might interact with other aspects of the jobs proposals; however, a conservative estimate would suggest that the total package, using the lower bound estimate of the job creation tax credit, would lead to at least an additional 4.6 million jobs over the first year.

[Table 1]


To help the tens of millions of Americans who are unemployed or underemployed, Congress and the Obama administration should take bold, decisive action to create jobs. The Recovery Act passed earlier this year helped pull the economy out of its sharp descent. But the American people, who see unemployment as one of the most important economic problems facing the country, overwhelmingly favor additional action to create jobs. The American Jobs Plan would create at least 4.6 million jobs in the first year alone. Over a 10-year period, the entire cost of the plan would be paid for with a financial transactions tax.

Further Reading: Paying for the Plan

Tracking the recovery: Voters’ views on the recession, jobs, and the deficit

Generating Jobs for a Robust Recovery by Lawrence Mishel (Oct. 20, 2009 / Policy Memo #151)

Download entire plan (PDF)

American Jobs Plan main page

See related work on Job creation | Jobs and Unemployment