tax plan is an across-the-board 20 percent cut in marginal tax rates. This cut, along with a few other tax changes Mr. Romney has endorsed – such as repeal of the estate tax and the alternative minimum tax – would reduce federal tax revenue from personal income and payroll taxes by an estimated $3.6 trillion to $3.8 trillion over 10 years.
The total is closer to $5 trillion when Mr. Romney’s proposed cut in the corporate income tax rate to 25 percent is included. About two-thirds of this amount would go to taxpayers making $200,000 a year or more – about 5 percent of all taxpayers.
Extending the Bush tax cuts for high-income earners, as Mr. Romney proposes, adds another trillion in lost revenue and increases the share of the benefits going to the top 5 percent. Even if the cost of the Romney tax cuts for the top 5 percent is covered by base-broadening measures, as Mr. Romney promises – but as President Obama and many others assert is mathematically impossible – does it make sense to devote trillions of dollars to lowering income taxes for the top 5 percent? Is this an effective way to create jobs?
Mr. Romney appears to think so. His plan rests on the assertion that lower taxes for high-income taxpayers will increase economic activity and employment – that lower taxes for job creators create jobs and will do so quickly. This assertion, while superficially convincing and ideologically compelling, is not supported by the evidence.
If tax cuts for high-income earners generate substantial real economic activity and job creation, then we should expect to see two things in the data. First, employment growth should be stronger in the years after tax cuts for these earners. Second, parts of the country with a larger share of high-income earners should experience stronger employment growth after national tax cuts for these taxpayers, because the places where they live receive a larger share of the national tax cuts.
What do we actually see after combing through a half-century of economic data? Neither of these predictions is borne out.
The graph below, based on our research, shows the relationship between the cumulative change in income and payroll tax liabilities for the top 5 percent over a two-year period as a share of gross domestic product and employment growth in the two years after the change.
The graph and the regression analysis on which it is based reveal that there is no link between income tax cuts for the top 5 percent and subsequent job creation. (We also examined the relationship between tax cuts for the top 10 percent and subsequent job creation and found the same result.)
The table below highlights three of these tax changes — the Reagan tax cut of 1982, the Clinton tax increase of 1993 and the Bush tax cut of 2003 – and subsequent employment growth. Strong employment growth followed the Reagan cut, but the employment growth following the Clinton tax increase exceeded the employment growth following the Bush tax cut, which was comparable in size to the Reagan cut.
Job growth at the state level after national tax cuts for high-income earners confirms the absence of a strong link between such cuts and the pace of job creation in the next two years.
The next graph shows no substantial link between tax cuts for the top 10 percent and the pace of job creation at the state level. Employment growth in states with a large share of rich people, such as Connecticut or New Jersey, was not much faster, on average, than it would have been otherwise after the Reagan and Bush tax cuts for the top 10 percent and wasn’t much slower, on average, after the Clinton tax increase on this group.
If there really were a strong link between job creation and tax cuts for high-income “job creators,” we should be able to see the effects somewhere. But we have found no evidence that such cuts lead to substantially faster employment growth at the national, state or even ZIP-code level.
Tax cuts for everyone else are a much more effective path to job creation. Our research found a statistically significant and positive relationship between tax cuts for the bottom 95 percent and job growth at both the national and state levels. The graph below shows the relationship for the national data. Our results indicate that almost all of the stimulative effect of income and payroll tax cuts on job creation in the short to medium run result from such cuts for the bottom 95 percent.
Lower-income taxpayers spend a higher share of their tax cuts. Many of these taxpayers often have more difficulty borrowing money and tapping into their housing wealth than higher-income individuals. These demand-side forces explain why consumption goes up much more after tax cuts for the bottom 95 percent than after equivalently sized cuts for the top 5 percent. An increase in consumption, which still accounts for about 70 percent of G.D.P., fuels increases in demand, and that leads companies to create more jobs. In survey after survey, businesses confirm that changes in demand are the primary determinant of their employment decisions.
Investment also increases after tax cuts for the bottom 95 percent, suggesting that shifting moderately size tax cuts to the bottom 95 percent from the top 5 percent isn’t a zero-sum trade-off between consumption and investment.
Instead, an increase in demand and economic activity because of an increase in consumption also makes investment more attractive, especially in difficult economic conditions.
Over all, our research shows that tax cuts for the bottom 95 percent are much more effective than tax cuts for the top 5 percent at increasing job creation in the subsequent two years.
Other analysts reach similar conclusions. For example, the Congressional Budget Office and Mark Zandi, Moody’s chief economist, find that tax cuts for lower-income recipients generate larger increases in employment per dollar cost to the federal budget than comparable tax cuts for high-income taxpayers in the short run.
What about the long run? A recent report by the Congressional Research Service found no clear relationship between cuts in marginal tax rates that primarily benefit high-income taxpayers and economic growth and job creation. A recent review by three distinguished academic economists also found no convincing evidence that real economic activity responds materially to tax-rate changes on top income earners, although such rate changes do affect their tax-avoidance behavior.
Cross-country comparisons also do not show a close link between top marginal rates and growth. While these studies don’t find large effects in the long run, we note that these long-run effects are harder to measure and are thus more uncertain.
Nevertheless, if the priority is to create a substantial number of jobs over the next presidential term, evidence from the last half-century strongly suggests that tax cuts for the top 5 percent won’t work. Tax cuts for working families, tax cuts directly aimed at expanded hiring or increases in infrastructure investment would have much more bang for the buck and would cost much less in terms of forgone revenue and deficit reduction in the future.
With elevated unemployment, weakness in Europe and slowing growth in emerging economies, fiscal measures that actually increase economic activity and employment in the near term are required. Our research shows that tax cuts for the rich do not meet this standard.