The latest in a series examining persistent macroeconomic myths 
Reader Bruce Carman writes:
total gov’t spending in the US, including personal transfers, is equivalent to 54% of private GDP
This is merely the latest installment in demonstrations of innumeracy in defense of reducing the size of government. Time to look at some data. First, even if we believed outlays as a share of non-government GDP was a reasonable measure, could we replicate the 54% number cited? No.
Figure 1: Ratio of government current expenditures (all levels) divided by non-government GDP (GDP minus government consumption spending) (blue) and divided by GDP (red). NBER defined recession dates shaded gray. Source: BEA, 2012Q3 third release, Table 3.1 for expenditures, NBER and author’s calculations.
So, total current government expenditures (on goods, transfers) is less than 44% of “private GDP”; whether that is a denominator that makes sense is a relevant question. More importantly, I just cannot understand why people assert things as facts without checking them (or even providing a link to a source; didn’t they teach this sort of thing in high school?).
Notice that 2012Q3 levels of outlays to non-government GDP are less than in 1982Q4.
Second, note that government spending in real terms has been increasing at a slower pace than real GDP.
Figure 2: Log real government consumption and investment to real GDP ratio (blue), and log nondefense government consumption and investment to GDP ratio (red), all in Ch.2005$. Real nondefense government calculated using Tornqvist approximation. NBER defined recession dates shaded gray. Source: BEA, 2012Q3 third release, NBER and author’s calculations.
Note that because the real variables are measured in Chain weighted dollars, one cannot simply take a ratio and compare levels. Rather one can look at how fast real aggregates have grown relative to other real aggregates by examining the slope of the curves.
Total government spending on goods and services has trended more slowly than GDP since 1987, and has been growing more slowly since 2009Q3. Nondefense government spending has exhibited similar behavior. One might suspect it’s employment, not output, that is most visible, thereby driving the “expanding government meme”. Below is the ratio of government employment to total nonfarm employment.
Figure 3: Ratio of government employment to total nonfarm payroll employment, s.a. NBER defined recession dates shaded gray. Source: BLS, November 2012 release, NBER, and author’s calculations.
While employment peaked at a ratio comparable to that recorded in 1992M08 (after taking into account temporary Census workers), it is now even below the previous peak recorded in 2003M07.
Finally, while current expenditures as a ratio to potential GDP did rise to high levels during the recession and immediate aftermath (as one would expect from a counter-cyclical stabilization policy), they are now below the ratio recorded in 2002Q3. What is remarkable is how much below average total current revenues are.
Figure 4: Ratio of government current expenditures (all levels) divided by potential GDP (blue), government current expenditures minus Federal defense consumption and investment (light blue), and total current receipts (red). NBER defined recession dates shaded gray. Source: BEA, 2012Q3 third release, Table 3.1 for expenditures and receipts, CBO Budget and Economics Outlook (August 2012) for potential GDP, NBER and author’s calculations.
Note that defense spending accounts for 5 percentage points of potential GDP worth of total expenditures (spending on goods and services, transfers, at all levels of government), as of 2012Q3.