No 'real responses' to changes in higher tax rates

Oisín Gilmore: Last week on Morning Ireland, IBEC’s chief economist Fergal O’Brien issued some warnings about the impact of taxes in Ireland on high income earners. He stipulated that Ireland’s taxes were so high that they were leading to shortages in the supply of high-skilled labour and in particular he implied that Ireland’s supposedly high tax rates would lead to high-skilled mobile labour leaving the country. And he argued that the focus on workers earning below €70,000 was misplaced.

In their Quarterly Economic Outlook IBEC state rather definitively that: “High marginal rates of tax disincentivise people from taking on extra work, from increasing their skills and from working in Ireland at all.” (p.9)

This would appear to be a case of Econ101. It appears intuitive that an increase in the tax rate means lower net pay, which should lead to workers working less. However, the empirical evidence on this is much less clear.

As with most issues in economic policy, the question of the responsiveness of labour supply to taxation has been more thoroughly studied in the US. Therefore, I am going to focus on US evidence.

Since the 1970s, the potentially negative economic impact of high tax rates has been a growing field of research. Famously in 1974 over in a meal in Washington Hotel with a Donald Rumsfeld and Dick Cheney, Arthur Laffer proposed his now famous Laffer Curve by drawing it on a napkin (see image above).

Laffer argued taxes were so high that the reduction in economic activity caused by them outweighed the tax increase. Therefore tax increases would lead to lower rather than higher government revenue. This proved to be quite a popular belief amongst right wing economists and has been used to justify a series of tax cuts, in particular on the very wealthy.

One major study by Martin Feldstein “The Effect of Marginal Tax Rates on Taxable Income” (1995) found that the cuts in marginal taxes in 1986 lead to a “dramatic increase in taxable income” in particular amongst high-income earners. However, this study and the many preceding studies that had similar findings have been heavily criticised for their methodology, with many arguing that they failed to distinguish between the increasing income experienced by high income earners and the impacts of the tax cuts on those incomes.

In the intervening 20 years there has been a substantial research programme that has fundamentally over turned the earlier consensus. Three leading economists of taxation (Emmanuel Saez, Joel Slemrod, and Seth H. Giertz) noted in a recent literature review “There is no compelling evidence to date of real responses of upper income taxpayers to changes in tax rates.” (p.35)

It is perhaps worth explaining what is meant by “real responses” here. This paper draws on the earlier work of Joel Slemrod (1990, 1995) where he draws a distinction between three levels in a hierarchy of responses to changes in taxation.

The first is a change in ‘timing’. Slemrod and other authors have found that if it is possible to change the timing of declaring income in order to avoid taxation, then incomes earners will do this. The second level in the hierarchy is ‘avoidance’. If income earners are able to alter their behaviour in order to avoid tax, they will do this. For example, if a CEO’s earnings are taxed at a different rate if they are bonuses to if they are an annual wage, then the CEO will structure their income in such a manner to pay the minimum amount of tax. However, Slemrod and other writers found that the evidence for this ‘avoidance’ response is weaker than for the ‘timing’ response. The third level in the hierarchy is the ‘real’ response. This response is basically what changes income earners make to their real economic activity in response to tax changes, after the timing and avoidance responses are removed from consideration. In other words, the real response address if high taxes “disincentivise people from taking on extra work, from increasing their skills and from working in a [given location] at all”, as IBEC claims they do.

There are a number of papers reviewing this literature. In addition to the Saez, Slemrod, and Giertz paper, there is a less technical and highly readable paper by Chye-Ching Huang at the Center on Budget and Policy Priorities in Washington DC. Interested readers can review these at their own leisure so, I will not go into too much depth here but merely restate the testimony of Leonard Burman, former head of the Urban-Brookings Tax Policy Center to the US Senate Committee on Finance: “Overall, evidence suggest [high-income American’s] labour supply is insensitive tax rates.”

On the specific issue of migration, the research is somewhat sparser. Bakija and Slemrod (2004) find that taxes do cause migration between states, but the number that move is relatively small. Young and Varner (2011) look at the “natural experiment” of the impact of the so-called “millionaire tax” in New Jersey, introduced in 2004. They find that this tax hike lead to a very small amount of migration. (The elasticity of responsiveness of migration to the state tax rate was found to be very less than 0.1 and not statistically different from zero.) But even this very small amount of migration was concentrated amongst the people living on retirement income, not wage earners.

Despite this substantial amount of evidence that modest changes in income taxation does not have a substantial impact on the economic behaviours of high income earners, some countervailing evidence should be noted. Firstly, Nickell (2004) has found that taxes have a higher effect on labour supply in Europe than in the US, with Ireland having one of the highest response. However, this paper looks at aggregate labour supply and doesn’t focus on high-income earners.

It has been found elsewhere that low and middle income earners are more responsive to tax changes than high income earners. The second piece of countervailing evidence comes from research conducted by Emmanuel Saez (mentioned above), Henrik Kleven and Camille Landais. They examine the effect of taxation on high income migrations. In their first paper on this topic, they look at the European market for superstar footballers and find that superstar footballers do move to countries with lower tax rates. In their second paper, along with Esben Schultz, they look at the impact of Denmark’s tax scheme to attract high income earners. They find strong evidence that it successfully attracted foreign high income earners. They use these findings to argue against tax competition and a ‘race to the bottom’ and highlight the need for tax coordination. It should be noted here that it is likely that pull effect of tax incentives for high income earners to immigrate might have a much larger effect than push effect of high taxes on domestic residents. (Intuitively, if you are a high income earner looking to migrate, the appeal of a country with a tax incentive aimed at you is likely to be substantial. Alternatively, if you are simply a high earner in a given country, you are likely to have friends, family, and numerous other ties to the place where you are living and moderate tax increases are unlikely to cause you to emigrate.)

To conclude, the overall evidence strongly suggests that, contrary to the claims of IBEC, moderate changes to the rate of tax on high earners do not “disincentivise people from taking on extra work, from increasing their skills and from working in Ireland at all”. However, the overall effect of taxation on labour supply is more ambiguous and the evidence on migration is contradictory.

For more on the need to preserve the higher rate of income tax, see this TASC policy brief.

Oisín Gilmore is a PhD candidate in economics at the University of Groningen.