The “Ever Expanding” Government Illustrated, Part III

Following up on these two previous posts [1] [2], I decided to confront some views commonly held in certain circles with some actual data. To summarize:

• Real government nondefense spending on goods and services is declining, and is declining relative to real GDP.
  
• The ratio of government outlays to nominal potential GDP is declining.
  
• Total civilian government employment is declining, and is declining as a share of total nonfarm employment.
  

Real government nondefense spending on goods and services is declining

Figure 1: Real spending on goods and services at state and local levels (blue area), and at Federal (nondefense) level (red area), in billions of Ch.2005$, SAAR. NBER defined recession dates shaded gray. Source: BEA, 2011Q2 second release.

Clearly, after rising throughout the 2000’s, total government nondefense spending picked up a bit in 2009, after which point is has declined. It’s currently at about 2006 levels.

Real nondefense spending on goods and services is declining relative to GDP

Since these series are in Ch.2005$, one cannot literally add the series up. However, as an approximation which should be pretty good over the last decade or so, I’ll add up government spending at state and local levels to nondefense spending at the Federal level. Then, since I cannot divide that series by GDP in chained dollars, I’ll log them both and rescale them to 2000Q1=0. Those series can then be compared.

Figure 2: Real spending on nondefense goods and services at government levels (teal), and real GDP, in billions of Ch.2005$, SAAR, logged and rescaled to 2000Q1=0. NBER defined recession dates shaded gray. Dashed line at 2009Q1. Source: BEA, 2011Q2 second release, NBER, and author’s calculations.

As is clear from the graph, real nondefense spending is falling, driven mainly by the decline in spending at the state and local levels. Example: See Wisconsin [3] [4].

The ratio of government outlays to nominal potential GDP is declining

Government outlays include transfers (unemployment insurance, SNAP, social security payments, etc.), as well as spending on goods and services. Once one takes these expenditures into account, what does the picture look like? I have plotted this series, normalized by potential GDP, in the figure below.

Figure 3: Nominal outlays at the state and local levels (blue area), and at Federal (nondefense) level (red area), SAAR, divided by potential GDP. NBER defined recession dates shaded gray. Note that Federal nondefense proxied by taking current expenditures and subtracting defense spending on goods and services. Source: BEA, 2011Q2 second release, NBER, and author’s calculations.

Since I didn’t have ready access to Federal nondefense outlays, I proxied nondefense outlays by subtracting nominal defense spending on goods and services from total Federal current expenditures. To the extent that there are defense transfers, I am overstating Federal nondefense outlays. The ratio in 2011Q4 is just about the same level as in 2008Q2.

Total civilian government employment is declining, and is declining as a share of total nonfarm employment

In the figure below, I plot the absolute level of government employment at all levels (blue line) and the ratio of government employment to total nonfarm payroll employment (red line).

Figure 4: Total government employment, in 000’s (blue line, left scale), and ratio of government employment to nonfarm payroll employment (red line, right scale), both seasonally adjusted. NBER defined recession dates shaded gray. Source: BLS, January employment release, NBER, and author’s calculations.

These are all graphs based on data readily available at BEA and BLS.