A year is an arbitrary unit of time – corporation tax edition
So, it turns out that corporation tax revenues from big companies in the US are lower than ever, despite resurgant corporate profits. I want to zero in on just one aspect to why we should expect corporate tax receipts to be low after a recession – and point out that it is a feature, not a bug. I am referring to companies being allowed to use losses in previous years to offset current taxes.
A simple example shows you why this makes sense: imagine a small business, say one that sells ice-cream from a truck. In the months of April-September the company makes a profit of $100 a month, but the other months it loses $100 as the truck sits idle (let’s say the company owns the truck, and uses the straight-line method for depreciation). Over the course of the year, there is no profit. But if corporation tax was assessed on a monthly basis, and the tax rate was 10%, the company would pay $60 ($100 * 6 months * 10%) a year. But if it was assessed on an annual basis, it would pay none.
Suppose instead the profit from April-September was $200 a month, therefore accumulating $600 in profit over the year. But there is a recession in the next year and instead of netting $200 in the summer months, it’s now only $50. Therefore the company loses $300 in that year. If you assessed corporate tax in individual years, $60 would be paid over the two years. If the tax was assessed over the two year period, it would be $30. From the standpoint of measuring profit, a year is an arbitrary unit of time.
So, whilst the implied corporate tax rate falls when measuring the tax receipts against profits in any given year after a recession, the relevant comparison is total tax receipts against total profits over time. That’s not to say that there are other less salubrious dynamics driving corporate tax receipts down, which there most certainly are. But it’s a curious feature of annual accounting that we should expect effective corporate tax rates to fall after a recession, and there is nothing wrong with that at all. Indeed, otherwise the tax would be prejudiced against businesses who tend to suffer most during recessions (which doesn’t seem like a very good idea to me), just as assessing the tax monthly would be prejudiced against seasonal businesses.
I think this reinforces a theme I alluded to last month, that it is much harder than you think to map political preferences onto data (e.g. higher levels of measured wealth inequality in Sweden than in the US). Corporation tax policy isn’t working, but even if it was the effective tax rate on profits as measured in a given year ought to decline after a recession.
I’ve talked before about the fallacy of analytical nationalism; that making nation-states the unit of analysis can lead to misleading comparisons. I haven’t yet thought of a catchy name for this one (the fallacy of analytical Gregorianism, anyone?), but I’m open to suggestions.