The Misguided Campaign Against Payroll Taxes

Like many countries, the United States funds a significant share of its social insurance through so-called "payroll taxes." Sadly, the term "payroll taxes" is not used with much precision in our national debate, but generally it refers to the Social Security Tax (12.4% on first $118,500 of earned income, split evenly between employer and employee), the Medicare Tax (2.9% of earned income, split evenly between employer and employee), and (less commonly) the Unemployment Insurance Tax (6% of first $7,000 of earned income, paid entirely by employer).

Recently, Jeff Spross of The Week has decided that these payroll taxes are very bad, and further decided they should be replaced with a consumption tax:

It's basically the worst possible species of tax. The payroll tax is actually just another income tax, but applied at a flat rate without the progressive brackets, making it regressive in the same way a national sales tax would be. Moreover America's payroll tax is capped, so the income you earn above a certain level isn't taxed. With a national sales tax, by contrast, no matter how much money you make or how big your expenditures are, you'll still pay the tax on them.

Moreover, sales taxes hit consumption, but payroll taxes hit employment — they tax job creation. In a comparison of different Western countries' tax mixes, economist Lane Kenworthy found that different combinations of payroll taxes, consumption taxes, and income taxes had about the same effect on economic growth rates. But a heavier reliance on payroll taxes was associated with a pretty significant drop in employment rates.

I think Spross is very wrong here, especially when it comes to the cross-country bit of it. The US tax level is extremely low by international standards. And this is true of its "payroll tax" level as well. The US should be looking to expand its overall tax level significantly, perhaps through consumption taxes, not looking to dial any of its taxes back.

Here is the overall tax level for various OECD countries in 2012 (only included developed countries in Europe plus US, Australia, Canada, and New Zealand):

At 24.3%, the US has the lowest tax level of these countries. Remember that the next time people drone on about our supposed fiscal difficulties.

The argument here is about payroll taxes. So here is payroll tax levels for the same countries (derived by adding social security contributions to "other" payroll taxes):

When we look at payroll tax levels, we find the US has improved greatly from last to seventh from last. It nonetheless finds itself with a payroll tax level way below most of these countries, as you can see.

As for the countries with lower payroll taxes, it wouldn't be correct to say (with the narrow exception of Ireland) that they have simply traded those taxes off for consumption taxes. This is evident in the following graph where I have combined payroll tax levels with (non-payroll) income taxes to create an overall income tax level measure:

Countries with low payroll taxes make up for them with higher income taxes. Denmark is the most dramatic example, as it has almost no payroll taxes but the highest income tax level in the world. These countries often have higher consumption taxes as well, but they are almost always additive to higher payroll and income taxes, not replacements for them.

You'll notice that Spross' explicit argument cites not to payroll tax levels but to payroll tax shares, which refers to the percentage of the tax level that is made up of payroll taxes. Here is a graph of that:

Once again, the US shows up pretty low on this ranking. But, more importantly, this particular stat makes very little sense if it is being used to predict such things as employment levels. If the US kept the payroll taxes it has now, but significantly increased its income taxes, its payroll tax share would fall. But how exactly would jacking up income taxes while holding all else equal cause employment levels to increase?

On the flipside, if the US kept the payroll tax level it had now but eliminated income taxes and consumption taxes (so a radical tax level reduction), its payroll tax share could conceivably spike to the highest in the OECD. But how would that cause employment levels to fall (as Spross' analysis would predict)? By what mechanism? This whole line of argument makes no sense because you can't coherently compare payroll tax shares between countries with dramatically different tax levels.

The more serious argument for why payroll taxes supposedly reduce employment levels (which Kenworthy quotes another scholar making) has nothing to do with payroll tax shares, and everything to do with the usual wage floor disemployment debate. The first dollar paid by an employer to a new employee is subject to employer-side payroll taxes of at least 13.65% (6.2% for Social Security, 1.45% for Medicare, and 6% for Unemployment). Normally, it is assumed that this 13.65% is passed on to the workers through lower wages. But as you move down the wage ladder and approach the minimum wage line, this becomes impossible. The net effect is that the minimum wage (from the employer's perspective) is at least 13.65% higher than the actual minimum wage line.

So, this is not actually an argument against payroll taxes. It's an argument for taking those taxes into consideration when deciding where to set the wage floor. Even if it were an argument directly against employer-side payroll taxes (perhaps the obscuring of those costs leads, politically, to suboptimally high wage floors), the US employer-side payroll tax level is also quite low by international standards:

Given all of this, it's just extremely difficult to understand why on earth anyone would be talking about the US having some overriding need to cut its payroll taxes and replace them with consumption taxes. The US has one of the lowest payroll tax levels in the developed world as well as one of the lowest overall tax levels. There is no particular reason to worry about the payroll tax share in itself (as Spross does), but even if there were, you can easily bring that share down by holding payroll taxes exactly where they are and jacking up consumption and income taxes. Not sure what that accomplishes for such things as employment levels, but, hey, it brings the share down!

At this point in time, what the US needs to do is increase its tax level significantly (I'd say by at least 50%). If it currently had a tax level equal to 45% of GDP (like the Nordics), then you might start wondering about how you can hold that level constant while optimizing the tax mix. But it actually has a ridiculously low tax level and that's true even for the level of each particular type of tax it uses. There are some who quite like that and others who think the tax level should be lower still. But if you are not one of those people (say because you care about inequality and poverty and social security), talking about anything but expanding the tax level is a waste of time, and talking about swapping out the payroll tax for an even more regressive consumption tax is downright foolish.