Tino Sanandaji on the P90 Threshold

by Reihan Salam

An interesting factoid from Tino:

An exciting new database by economists Facundo Alvaredo, Tony Atkinson, Thomas Piketty and Emmanuel Saez amongst others has data on the income level of the tenth percentile for several countries. This measure, P90, is the level of income that would make someone richer than exactly 90% of the population. 

I have adjusted the figures for inflation and PPP based on OECD data. For a couple of countries the data was a few years prior to 2005, so I assumed their income share remained constant and their income increased with GDP. These are not perfect estimates, because the data comes from different sources, have different units and different standards, you should just view them as illustrative.

These are pre-tax income. Payroll taxes are not included, nor is employer funded benefits such as health insurance and pensions. These are mainly based on the statistics gathered by the tax-agency, so unreported income will not be included. 

Recall that including health and pension benefits tends to markedly increase income for Americans, particularly those at the top end of the distribution. 

So what does Tino find? 

In 2005 the P90 threshold in the United States was $105.000, which means that ten percent of Americans had income at or above that level. 

The same year the P90 threshold in Sweden was $43.000 (37700 kr), so if you earned this paltry sum you were among the top ten percent highest earners. In fact, less than 1 percent of Swedes earn more than $100.000 per year, about 880.000 kronor per year purchasing power adjusted. 

This is confirmed by the Swedish Statistical Agency, which reports that in 2005 there were 61.000 Swedes who earned more than 800.000 kronor per year, or merely 0.8% of the adult population. 10.2% earned more than 360.000 kr, which again indicates that the figures are almost identical to the Top Income Database. 

No doubt, those at the bottom are better off in Sweden than in the United States. However the middle class in Sweden is still somewhat poorer than the middle class in the United States, while high-skilled Swedes are far poorer than high-skilled Americans.

Adding in the value of social insurance and high-quality public programs will no doubt increase the effective income of Swedes at P90, but of course these measurements exclude the value of privately-provided fringe benefits among Americans. 

It seems clear to me that American social insurance programs don’t work very well and that our public programs devoted to aiding the poor work even less well. That is a serious problem, and it is the central fixation of this blog. Among other things, we’ve advanced ideas like moving to a defined contribution model for social insurance programs like Medicare and, for a wide variety of public programs, making it easier for specialized providers to compete with incumbents, drawing lessons from Sweden and other market democracies. 

But would poor and middle class Americans be better off if our P90 threshold were half as high as it is now? This would mean that we’d have far less inequality. Yet it would have other consequences as well that would have a deleterious impact on the poor and the middle class, as Tino notes in his post.

The Agenda

NRO’s domestic-policy blog, by Reihan Salam.