The director of the Urban-Brookings Tax Policy Center, Donald Marron, had an interesting point in testimony yesterday before the Senate budget committee about how tax expenditures obscure the size of the federal government:
Analysts usually invoke official budget measures—revenues and outlays—when trying to measure the federal government. For example, we often hear that federal revenues have averaged about 18.1 percent of gross domestic product (GDP) over the past four decades, while outlays have averaged about 20.7 percent. But those measures are incomplete—and potentially misleading—if some tax breaks are effectively spending programs.
In some preliminary research, my Tax Policy Center colleague Eric Toder and I (2011) have tried to estimate how large the government is when we recognize that many (but not all) tax preferences are effectively spending programs. For fiscal 2007, we estimate that spending-like tax preferences amounted to 4.1 percent of GDP. Adding that to official outlays yields a broader definition of spending, 23.7 percent of GDP in 2007, about a fifth larger than the official 19.6 percent. Similarly, our broader definition of revenues—official revenues plus revenues foregone through spending-like tax preferences—is 22.6 percent of GDP rather than the official 18.5 percent.
These figures illustrate that conventional budget measures understate the extent to which federal fiscal policy affects economic activity.
That "preliminary research" to which Mr. Marron refers is a paper called "Measuring Leviathan: How Big Is the Federal Government?"