Another “How Big?” Question: U.S. Federal Spending

Just about everyone and their dog has seen those scary graphs of government spending shooting skyward along an exponential curve.  They look like this:

This runs from 1913 until 2011 and was retrieved from  In 1913 Federal spending was $715 Million.  In 2011 it was $3,603,061 Million ($3.6 TRILLION).  However, if you’ve seen my last few posts you know how important it is to adjust for inflation and population before assigning any meaning to these kinds of statistics.  Well, that’s just what I did.  I grabbed my population numbers from a previous post and used an online inflation calculator to convert everything to 2012 dollars.  We end up with real (inflation-adjusted) Federal spending per person from 1913 to 2011.  How’s that one pan out?  Again, I was stunned by the result (which means my guess and reality were *again* very different).

Wow!  I kinda thought the folks screaming about how much government spending has grown were exaggerating in a major way.  It looks like I was wrong.  In 1913 the federal government spent $172 per person per year (in 2012 dollars!).  In 2011, the federal government spent $11,799 person person per year (in 2012 dollars).  I’m blown away!  At the height of WWII we spent only $8515 per person in 2012 dollars.

Here are my thoughts:

  • Since around 1930 we’ve seen an almost linear growth in federal spending per person.  The nineties represent a departure from that trend, but 00’s soar to bring us right back to the trend line.
  • We get a LOT more from our government today in terms of social safety nets, research, public health, education and other programs than we did in the early 1900’s.  Is that good or bad?  That’s a matter of whether or not that is what we want in terms of what we give the government and what we expect to get in return. There isn’t a clear right or wrong — but we need to recognize that fact as a society and make an enlightened choice.
  • Federal spending really is increasing dramatically over the long-term trend and seems to do so in a fairly predictable way.
  • I don’t think that trend line is sustainable.  How far can it go?  How will it end?  I have no idea.

I’d really like to look at federal spending vs. median household income, but that is a topic for another day.

Hat tip to Kenny Felder for getting me to pursue this set of calculations!

In Perspective, How Big is the U.S. Government?

This is another post in which I’m trying to separate fact from rhetoric for my own sake, and I thought I’d share my results. There are a lot of politicians making promises to shrink “Big Government” in the U.S. What I wanted to know was simply how big has that government actually become.

Let me say right off that there are LOTS of ways to consider whether the federal government is “BIG” or not — budget-wise, intrusiveness into local/regional issues, etc. As a start I’m just looking at how many employees the federal government actually has. There are real problems with this — for example, contractors are not, I believe, counted as employees and I can’t say how the number of contractors has changed over the years. Likewise, this gives no glimpse into private-sector workers whose livelihoods are completely dependent on federal work (defense industry, for example).  This is just another glimpse that might give me a bit of insight.

As in my previous posts, I’ve analyzed the data by adjusting for population (i.e.: per capita results). Why? If the population has doubled, but the government is the same size, then are government has HALF as many employees per citizen as it did previously. That indicates a much leaner government. On the other hand, if population doubles but government employment quadruples, then we have TWICE AS MANY employees per citizen as we did previously.  Population adjustments are absolutely necessary if you want to make sense of historical trends.

A second caveat:  this is another “first order” approximation.  I’m not looking to publish this for economic modelling purposes, so I’m not worried about being off a few percent here or there due to interpolations I made on census data, for example. I just wanted to make a quick-and-dirty reality check to see how big a problem government growth actually is. Employment data were drawn from this table:

Once again, the data and my expectations were pretty darn different!  Here’s total U.S. Federal Employment (Civilian + Military) per capita from 1962 through 2010:

Much to my surprise, the size of the federal government in terms of employees per capita has been on a general downtrend since the end of the 60’s. In fact, it’s now under half the it’s peak size for this time period. I absolutely did not expect that.  Digging through the data reveals that we’ve seen big drops in civilian employees per capita and still bigger drops in uniformed military per capita. The drops in military employment as the cold war wound down and we refocused on stand-off engagements (cruise missiles, smart bombs, drones) make sense, but I didn’t think that through beforehand. What really surprised me, though, is that the number of civilian employees (what we can consider to be the bureaucracy) has actually shrunk per capita.  I really did think government was much bigger in terms of employees now than it was in the 60’s or 70’s.

Well… now I know better!  Just for kicks here is one more graph stacking up all federal employees (executive branch employees, legislative/judicial employees and military):

So what does all of this say about whether or not our federal government is too big?  Not a thing.  It does not tell us what the best size would be for the most successful government for our society.  It only says that our federal government has been shrinking since the 60’s in both the civilian and military side of things.  We may or may not have a “BIG” government, but it *is* a much smaller government than it used to be.  The only other thing that I notice is the lack of any convincing claim that one party or the other has a significantly different impact on size of government from their rival.  The downward trend is fairly uniform.