Tax policy in Switzerland: a race to the bottom?

Tax policy is notoriously a controversial issue in Switzerland. Cantonal autonomy in this area is extensive, deeply rooted, and widely believed to lead to competition among cantons to attract wealthy taxpayers. Some think this is good: citizens can choose among different combinations of taxes and expenditures, and there is a restraining effect on public deficits. But others think that competition is bad, because it leads to a race to the bottom that undermines the capacity of cantons to deliver important public services. One of the latest incarnations of this longstanding debate is the popular initiative “Für faire Steuern: Stopp dem Missbrauch beim Steuerwettbewerb”, on which the Parliament should take position soon.

In a recent paper, Fabio Wasserfallen and I compared cantonal personal income tax rates and found an interesting pattern. To the extent that cantonal tax rates converged towards lower levels, the trend was stronger for lower incomes. We can see it in this figure, which shows the evolution of tax rates for four income categories (CHF 25,000, 50,000, 100,000 and 200,000 per year) (click on the image for a larger version):

The horizontal axis shows the tax rate for an unmarried person without children in 1990, while the vertical axis shows the difference in rates between 2007 and 1990. Negative values mean that tax rates decreased, positive values that they increased, and a negative relationship between the starting point (rates in 1990) and change (2007 minus 1990) is indicative of convergence (what is sometimes called “beta” convergence). Unfortunately the data are not corrected for inflation (it is not straightforward for technical reasons), which implies that changes within income levels were actually less dramatic than it seems. However, we can still safely compare trends across income levels, and we then see that convergence was stronger for lower income levels, and decidedly more limited for higher income groups. This conclusion rests on the fact that the negative relationship between tax rates in 1990 and change between 2007 and 1990 becomes weaker as income increases.

This is not really what one would expect on the basis of tax competition arguments. Part of the story is probably that competition is limited to a specific region (Zurich and some of its neighbors) and does not have a national dimension. But still, the pattern across income levels is quite striking. Any ideas what might be going on here?

4 thoughts on “Tax policy in Switzerland: a race to the bottom?

  1. That’s a very interesting graph. But isn’t this finding simply due to the progressivity of taxation? There might be greater convergence, but there is also much less variance in taxation levels for lower incomes.

    The margin of maneuver (and competition across cantons) in taxation policy is essentially located at the upper end of income levels and not at the lower end. As a state seeking revenues, you don’t have much interest in changing taxation for lower incomes because their taxation base is small anyway and it doesn’t make much of a change for state revenues. Also, if you lower taxes for low incomes they won’t come to your canton because of this, and if they do, they may cost more in social benefits than what they pay in taxes.

  2. Good point. But that’s exactly what’s puzzling. If cantonal governments are interested in wealthy taxpayers, why is convergence stronger for 100k than for 200k, despite the greater room for maneuver and incentives to cut in the latter?

  3. I guess the greater margin of maneuver for high incomes means a greater degree of possible variation in tax rates (due to the nature of this policy), but does not mean that all cantons can use it (because of different incentives/preferences/constraints).

    Greater potential of variation for high incomes: Say one needs 18k a year to live. If somebody earns 20k a year, you can tax him only up to 10% of his income, because if you tax him more he won’t have enough to live. So the potential variation is between 0 and 10%. If somebody earns 200k a year, you can tax him up to 90% until he reaches the survival threshold. The potential variation is between 0 and 90%. So it’s somehow natural that there’s greater convergence for low incomes because the possible variation is much smaller.

    Differentiated ability to make use of this potential variation: In Schwyz or Uri you can lower taxes for high incomes, get lower tax revenues at the outset and bet on the fact that you’ll attract wealthy taxpayers. In Zurich you have a lot more public services to finance with tax money, and probably cannot make those kinds of risky policies.

    Maybe it’s this combination of greater potential variation and differential incentives/preferences across cantons that explains lower convergence (and therefore greater diversity) as to the taxation of high incomes?

  4. Makes sense. We certainly need to do more work to explain the trends. But regardless of the reasons, the descriptive point remains: less downwards convergence for high incomes than one would expect.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s