Henry Aaron, of Brookings, has a column today in which he purports to demonstrate that federal spending isn’t a problem.  His method is to pose four true or false questions, assert that most people would answer “true” to one or more of them, and then explain why all are false, thereby showing how ill-informed people are.  If Aaron is going to write a column where he proclaims other peoples’ ignorance, then he ought to make sure that his own analysis isn’t flawed.  The first question he poses is: “[t]he federal government is spending a larger share of national income than at any time since World War II – true or false?”  His answer is as follows:

Let’s start with government spending. According to the Congressional Budget Office, the Federal Government will spend 21.7 percent of GDP next year under current policy. Were the U.S. economy operating at capacity, that share would be less than 20.6 percent, because output would be higher and spending for such items as unemployment insurance would be lower. For the preceding three decades government spending averaged 21.1 percent of national output. In brief, the numbers flatly contradict the assertion that spending is “out of control.”

Taking his numbers at face value, Aaron puts forth two facts in his response:

1. Over the last three decades, average spending was 21.1% of GDP; and

2. Next year, under current policies, the government will spend 21.7% of GDP.

The only logical conclusion, based on these numbers, is that the response to his first question is “true.”  To arrive at his “false” conclusion, Aaron has to insert a counterfactual – that spending would be 20.6% of GDP if the U.S. economy was operating at full capacity.  This is a ridiculous argument.  One could just as easily say that spending would only be 10% of GDP if the economy grew by a factor of two.  That may be a true statement, but it is certainly irrelevant to answering the question of whether actual spending next year will be higher or lower than the thirty year average.  The answer to that question, as Aaron so succinctly demonstrates above, is “true.”