Wealthy Are Dumping Assets Fast Before Year's End... Including Their Business Or Maybe Your Job, page 2

This explains things well:

Is Raising Marginal Tax Rates on Higher Income Individuals a Good Idea? Becker

Newly elected French President Francois Hollande campaigned on a promise to raise the marginal tax rate to 75% on individuals making more than 1 million euros. President Obama wants to increase marginal tax rates on high-income individuals to their level before the Bush tax cuts. Obama defines “high” as anyone making over $250,000 per year. Neither the modern history of high tax rates, economic analysis, or their consequences for the budget deficit and income redistribution indicates that raising taxes on higher income individuals are a good idea.

A few decades ago, tax rates on higher incomes were at least 70% not only in high taxing Scandinavian countries, but also in the United States (where the top rate was 91% in 1960) and many other countries where government spending took a relatively small share of GDP. The widespread avoidance and evasion of such high taxes through sophisticated accounting methods, reduced work effort, and even in countries like Sweden through outright barter for services and goods, led to a world wide revolution toward flatter and much lower tax rates. The bi-partisan tax accord in the US in the 1980’s reduced the top income tax rate to 28% (the top rate was 33% for a fraction of high income persons), and even Sweden lowered its top income tax rate to about 55%. Since aside from the Great Recession, GDP has grown quite rapidly during the decades subsequent to the tax reduction movement, why is there growing pressure to raise tax rates again on the so-called “rich”?

From the academic side, support for raising taxes on higher income individuals mainly comes from evidence on individual behavior over the lifecycle. This evidence suggests that labor supply does not respond very much to changes in after-tax earnings (see a good survey of this issue in Meghir and Phillips, “Labor Supply and Taxes”, 2010). Yet evidence on aggregate labor supply, such as the differences in hours worked among countries with different levels of taxes, suggests that workers spend considerably more hours working when marginal tax rates on their incomes are lower.

A recent article in the Journal of Economic Literature tries to reconcile the micro and macro based conclusions by arguing that the micro evidence gives a biased picture of aggregate labor supply responses (see Keane and Rogerson, “Micro and Macro Labor Supply Elasticities: A Reassessment of Conventional Wisdom”, June 2012). One of the factors in their reconciliation comes from the importance of on the job investments in human capital. These investments bias downwards the micro estimates of labor supply elasticities that are based on variations in hours worked over the lifecycle. This bias is especially large at younger ages where most of these investments occur since the true earnings at these ages is much larger than the observed wages at these ages because younger workers are raising their future wages through investments in their human capital. This bias helps explain why computed elasticities of labor supply are usually greater for older workers.

A second important bias shows up especially in the difference between the labor responses of men and women. It has long been known that women respond more to higher tax rates on their (family) earnings because many women then leave the labor force entirely rather than simply adjusting their hours worked. As Keane and Rogerson show, such decisions to enter or leave the labor force can greatly increase the aggregate labor response to changes in tax rates.

These and other corrections to simple interpretations of the micro evidence on labor responses to changes in tax rates clearly suggest that aggregate labor responses to tax rates may be quite large. Moreover, as welfare economics shows, raising tax rates by only a few percentage points on a sizable tax rate base- as in President Obama’s proposal- will tend to have large costs in efficiency even when the elasticity of response to the tax increase is relatively small.

I would not argue, however, that the evidence conclusively proves that the higher taxes proposed would do significant damage. Suppose then to be conservative that there is only a 50-50 chance (I believe the true probability is much more than 50-50) that the tax increases proposed by Hollande and Obama would sizably reduce hours worked and the effort put into work relative to the magnitude of the tax increases, and would cause sizable loses in efficiency relative to the additional revenue raised. Would such probabilities justify much higher tax rates?

To answer this question, one has to consider the potential benefits and costs of raising taxes on higher income individuals, and determine whether expected benefits exceed expected costs. If higher taxes on the rich only slightly affected their work effort (the usual assumption in revenue calculations), tax revenue would rise, but not by a lot since the great majority of revenue comes from taxes on the other 98% of taxpayers. With only a little increase in revenue under the most favorable conditions about labor supply responses, such tax increases would do little to close the budget deficit, and not much additional revenue would be available to redistribute to lower income families.

Higher taxes on the so-called “rich” would likely reduce after tax income inequality. The sizable growth in earnings and income inequality since 1980, especially at the high-income end, is one of the forces behind the movement toward higher taxes on the “rich”. In particular, many people are upset about CEOs getting high bonuses and stock options even when there companies are doing badly, or bankers getting their very high options and bonuses even after they made such bad decisions that contributed to the financial crisis. I have sympathy with these concerns, but the way to attack these problems is not to raise the taxes on everyone earning more than $250,000. For professionals and small business owners, not workers in the financial and banking sectors, constitute the great majority of persons in that higher income bracket. The best way to meet these concerns is by further improving corporate governance, and by formulating effective rules-based banking regulations that discourage the too big to fail banks from taking on excessive risks, and hence unjustified compensation.

So the gain in tax revenue from higher taxes on richer individuals would not be great even in the chance that these taxes only slightly discourage their hours worked and effort at work. But the cost to the economy in the chance that higher taxes greatly discourage their effort (relative to the magnitude of the tax increase) is likely to be substantial in terms of fewer hours worked and less work effort by high income individuals, reduced incentives to start businesses, less investments in their human capital, investing abroad rather than in the US or other countries that raised these taxes, and even migration abroad, especially in countries like France where many talented Frenchmen are already working in Britain, the US, and other countries.

So I conclude that even with considerable uncertainty about how much higher taxes on higher-income individuals would reduce their work effort and their investments, the expected gain from raising these taxes is likely to be negative. The trend toward lower marginal tax rates during the past 50 years was perhaps mainly the result of interest group pressure from higher income individuals, but it also receives support from a benefit-cost analysis of the expected effects of tax increases on behavior.