From Basic Research to Policy Recommendations

We live in unusual times. Who would have thought that some of the super-wealthy would be clamouring for higher tax rates for themselves? The trend was started by Warren Buffett, believed to be the world’s third-richest man, who asserted that "a billionaire-friendly Congress" had "coddled" the wealthy for too long and should raise taxes on the rich. His income-tax rate, he said, amounted to 17.4% last year, less than his secretary's. And that despite the fact that in the US the top 1% raked in fully 23.5% of total income in 2007, compared to 9% in 1970, the highest level on record since 1928.

Mr Buffett has now been joined in his call by scores of other well-heeled people in several European countries. True, they want their tax rates raised as a way to help their countries cope with debt. But taxes on the rich help to preserve programmes while holding down the taxes of the rest of the population. And taxes affect the evolution of the distribution of wealth.

There is a controversial political debate in the United States on whether taxes on the rich should increase. But does the Republican aversion to raising taxes make economic sense? For the answer, turn to a timely CESifo Working Paper that sheds much light on this.

Written by two CESifo Research Network members, Nobel laureate (and Distinguished CES Fellow) Peter Diamond and CESifo Distinguished Affiliate Emmanuel Saez, the paper casts a critical glance at the literature on optimal taxation to explore the path from basic research results to formulating policy recommendations, and what criteria should a proposal fulfil in order to have a chance of success.

As they rightly point out, a tax system should maximise a social welfare function within its budget constraints, taking into account that individuals respond to taxes and transfers: both can negatively affect incentives to work, save, and earn income in the first place. This creates the classical trade-off between equity and efficiency that lies at the core of the optimal income tax problem.

The authors consider multiple models, since each highlights a different aspect of reality, and try to draw inferences simultaneously from them. In their view, a theoretical result can be fruitfully used as part of forming a policy recommendation only if three conditions are met. First, the result should be based on an economic mechanism that is empirically relevant. Second, the result should be reasonably robust to changes in the modelling assumptions. Third, the tax policy prescription needs to be socially acceptable and not too complex.

Messrs Diamond and Saez come up with three policy recommendations from basic research that they believe can satisfy these three criteria reasonably well.

First, very high earners should be subject to high and rising marginal tax rates on earnings. The share of total income going to the top 1 percent (those earning more than US$400,000) has increased from 9 percent in 1970 to 23.5 percent in 2007, much higher than in European countries. Despite an enviable (for Europeans) low income tax rate of 22.4 percent, they accounted for over 40 percent of total federal individual income taxes in 2007. This points to their importance for raising revenue: without behavioural responses, a 7% hike in the tax rate would raise revenue by 1% of GDP. Even a 21% hike would leave top-earners pocketing an after-tax share of total income more than twice as high as in 1970. (Of course, higher tax rates would trigger some behavioural responses that could drive down tax intake, responses that are analysed in the paper.)

The second recommendation that Messrs Diamond and Saez make is that the tax and transfer policy toward low earners should include subsidisation of earnings, and that this subsidisation should be phased out at a relatively high rate. In this income group, as they point out, behavioural responses on labour force participation play a crucial role. The US’s Earned Income Tax Credit, for instance, which concentrates benefits on low-income working families instead of those with no earnings, has encouraged labour participation. This is a response along the extensive margin, i.e., the decision of whether or not to work.

However, the behavioural response along the intensive margin, i.e. hours of work on the job, is also important. If, as empirical studies show, the extensive elasticity is large for those with low incomes relative to the intensive elasticity, initial subsidisation combined with high phase-out rates further up the distribution would be the optimal profile.

Some high-income countries concerned that traditional welfare programmes overly discouraged work have followed this approach in recent decades, exhibiting a marked shift toward lowering the average tax rate at the bottom through a combination of a) introduction and then expansion of in-work benefits (such as the Earned Income Tax Credit in the US); b) reduction of the statutory phase-out rates in transfer programmes for earned income; and c) reduction of payroll taxes for low-income earners.

The third recommendation the authors make is that capital income should be taxed. And significantly. This flies in the face of the repeated calls for not taxing capital income, as well as of the two usual theoretical analyses upon which critics of this type of taxation base their arguments: those of Chamley and Judd (the optimum has no asymptotic long-run taxation of capital income), and Atkinson and Stiglitz (the optimum has no taxation of capital income).

Messrs Diamond and Saez point out that these theoretical approaches, however, rely critically on assumptions that are not empirically supported. The zero-tax result of Chamley and Judd, which the authors analyse, requires that rational intertemporal decision-making hold not only for entire lifetimes, but extend across dynasties. One may be forgiven for suspecting that the recent financial crisis may have put a bit of a dent on that notion. When the infinite-horizon decision-maker is replaced with a sequence of finite-horizon decision-makers, the picture regarding the tax question changes dramatically. And the Atkinson-Stiglitz result depends on uniformity in saving behaviour, which is not supported empirically.

The authors then posit four arguments for positive taxation of capital income: the difficulty of distinguishing between capital and labour incomes; the positive correlation between earnings opportunities and savings propensities; the role of capital income taxes in easing the tax burden on those who are borrowing-constrained; and the role of discouraging savings in encouraging later labour supply in the presence of uncertain future wage rates and higher savings for earlier retirement.

Their analysis of these four arguments in detail leads them to conclude that capital income should indeed be taxed.

US lawmakers could do worse than read the arguments of Peter Diamond and Emmanuel Saez before they sit down to devise suitable ways to reduce the deficit. And, as an added bonus, maybe reduce wealth inequality.

Some European countries could do with the same advice as well.

Pater Diamond, Emmanuel Saez: The Case for a Progressive Tax: From Basic Research to Policy, CESifo Working Paper No. 3560

• Other CESifo Working Papers by Peter Diamond
• Other CESifo Working Papers dealing with tax policy
• CESifo Working Papers published in the past three months