One Metric of the Size of Government

Looking at civilian employment in the executive branch since 1940.

Ok, so yesterday I ended up engaged in an interchange on Twitter* over the size of the federal government as based on the number of civilian employees in the executive branch.  This was in the context of how big the government is and it growth over time.  The interchange started, actually, with a rather gross misreading of the chart in a Retweet from one of my Tweeps, about which I commented and corrected.  Ultimately, what struck me about the table was that while a comparison of the first data point in 1940 to the last in 2011 shows a substantial increase (699,000 civilian employees in the executive branch in 1940 to 2.15 million in 2011) the bottom line is that the outlier over the 71 period is 1940, not 2011.  The table can be found here, and here’s a basic line graph of the data:



Of course, the initial fluctuation is clearly linked to WII, but after that event the overall level of civilians employed by the executive branch (the largest part of the federal government by far) has stayed within a range hovering around 2 million for many, many decades.

There are two points of consequence that I think are significant.

1.  Contrary to a prevailing narrative, government (at least by this metric) has not radically grown of late.  Indeed, while there has been a slight uptick in the last several years, the size of the government in this area is only slightly above both the seven decade average (2.02 million) and the median (2.06 million).

2.  As a function of population, the relative size of government is actually smaller now than it was after the WWII era explosion took place (and receded).  Consider:  in 1950 there were 1.98 million civilian employees in the executive branch and the national population was 161 million.  In 2011 the employee count was 2.15 million for a population of over 300 million.  Therefore, relative to the population, the government is actually far smaller now than it was in 1950.

Now, this is just one metric of the size of government, but it is interesting to note that despite the narrative of uncontrolled growth, on this measure the government can be seen to be either relatively steady (in absolute terms) or actually smaller (in terms of employees as a function of the US population).

Now, certainly, this tells us nothing about whether this is a good (or bad) level of employment.

For kicks, here’s total federal employment, including the military and the legislative and judicial branches from 1962-2011 (source):



The overall trend is downward, especially from its Viet Nam era peak.  Interestingly, there is also a general downward trend from the late 1980s until the mid-2000s, with an uptick in recent years.  The average for the range is question is 4.94 million (and the median is 5.00 million).  The 2011 figure is 4.40 million, substantially below the average and the median.

*One thing this interchange reminded me: it is impossible to actually discuss anything over Twitter.

FILED UNDER: US Politics, ,
Steven L. Taylor
About Steven L. Taylor
Steven L. Taylor is a Professor of Political Science and a College of Arts and Sciences Dean. His main areas of expertise include parties, elections, and the institutional design of democracies. His most recent book is the co-authored A Different Democracy: American Government in a 31-Country Perspective. He earned his Ph.D. from the University of Texas and his BA from the University of California, Irvine. He has been blogging since 2003 (originally at the now defunct Poliblog). Follow Steven on Twitter


  1. Franklin says:

    I imagine the trend of privatization will get brought up.

    EDIT: Holy crap you guys got Edit/Delete to work in Chrome. Thumbs up, many less posts with dumb typos from me from now on.

  2. Ernieyeball says:

    Note at the bottom of the table titled Executive Branch Civilian Employment Since 1940

    Data comes from agency 113-A monthy submissions and covers total end-of-year civilian employment of full-time permanent, temporary, part-time, and intermittent employees as of September 30th.

    Civilian employment. Does this include employees of private contractors a government agency (NASA comes to mind) hires to do work the US Govt. pays for?
    Or does the count only include Civil Servants that draw a paycheck directly from the US Govt?

  3. mantis says:


    EDIT: Holy crap you guys got Edit/Delete to work in Chrome. Thumbs up, many less posts with dumb typos from me from now on.


  4. JKB says:


    I doubt it. Nor does it deal with the transformation of the federal workforce, especially from 1962. The elimination of file clerks, most blue collar workers and even secretaries replacing them with more meddling regulators.

  5. Tsar Nicholas says:

    – In 1940 the federal government spent 118 billion (inflation-adjusted) dollars.

    – In 2012 the federal government spent 3.1 trillion dollars.

    – In 1969, the apex of the Vietnam War, federal government spending was about 19% of GDP.

    – In 2004, the height of the 2nd Iraq War, federal government spending was under 20% of GDP.

    – In 2012, the fourth consecutive year of “recovery,” according the media-academe complex, federal government spending was 24% of GDP.

  6. bill says:

    um, lessee- computers good! they do seem to be rising as of late, what would prompt that?!

  7. Dave says:

    @Tsar Nicholas: Are you meaning to tell us that spending as a percentage of GDP was higher in the midst of a double dip recession? Holy shit say it isn’t so. Dishonest way to look at spending which previously was reported in dollars and then as percentage of GDP. Afghanistan and Iraq both of which were totally unnecessary wars are what bankrupted this country, not deficit spending in the midst of a recession. I recall the republicans arguing that deficit spending was great for the economy and pointing to Reagan and it failed to grow the economy, but did grow the debt. So yes spending will be higher as a percentage of GDP when the country takes in less money due to a bum economy. There are lies, damn lies and statistics and your intellectual dishonesty is proof of that.

  8. Jc says:

    Tsar, remember when you bow to your bronze St Ronnie statue. The national debt tripled under his administration, deficits at 5% of GDP were the norm and if you look at the highest fed employees total in the past 30 years, well Ronnie has got that one too! For a guy who rails against government, you sure like guys who run big government, borrow a ton (at high rates too) and employ a lot of gubmint inefficient do nothings.

    Should also look at total gov including state and local to see which gubmint has grown the most.

  9. Unsympathetic says:

    1) Be sure to include all .gov — not just federal. If a task is shifted from federal to state, that action doesn’t change the size of government, it just changes the name on the paycheck of the same person doing the same task.
    It is impossible to go back in time to estimate the total number of state and local employees who have worked under federal mandates. It seems reasonable to conclude, however, that the number has increased dramatically as Congress and the President figured out how to encumber state and local governments at no cost to the federal Treasury.
    2) Be sure to include the SHADOW employment — the contract and grant workforce. If government “outsources” a task that once was performed by the government, that contract must employ people to accomplish the task.. so include those FTE’s. As above, the outsourcing by itself didn’t cut jobs.

    Instead of engaging in an endless effort to keep the civil service looking small, counting federal FTE’s in the ways I have described would force people to consider just how many of the 16.9 million federal “producers” should be kept in-house and at what cost. One can easily argue that the answers would lead to a larger, not smaller, civil service, or at least a civil service very differently configured. There are many good reasons to hire a shadow employee, whether to improve government performance, protect taxpayer interests, or protect small business. But maintaining an illusion of smallness is not one of them.

    Congress and the President would also have to ask tougher questions about who works in the shadow and under what conditions. It makes no sense, for example, to enforce a rule-bound merit system for one set of federal producers, while allowing an entirely different system for contractors, grantees, and state and local surrogates.

    Counting all the jobs might also lead to a long-overdue discussion of public service in non-government settings. More than a third (4.7 million) of the federal government’s 1996 shadow worked for state and local government, which makes them public employees, not private.

    Even if they don’t share a common purpose, shadow workers still have public obligations. Just because they receive their paychecks and identification cards from private firms or nonprofit organizations does not mean they can be ignored as distant pieces of the federal public service. They may be motivated by profit, and may serve multiple customers, but they accept a public service obligation whenever they act on behalf of the federal government.

  10. OzarkHillbilly says:


    Be sure to include the SHADOW employment — the contract and grant workforce. If government “outsources” a task that once was performed by the government, that contract must employ people to accomplish the task..

    True. It just introduced all kinds of “For Profit” inefficiencies that had the lone positive effect of lining the pockets of GOP donors (Halliburton any one?).

  11. OzarkHillbilly says:

    and yes, I am being a completely partisan hack here, but the best example I can think of, is all of the outsourcing of military jobs done during the Bush II years, tho it is possible that is only the most recent example.

  12. superdestroyer says:


    I believe that the total number of civil servants is set by Congress. In the 1980’s, there was a push to reduce the number of activity duty military. It just results in more civil servants working for the DoD. IN the 1990’s the push in DoD was to reduce the number of civil servants, so more contractors were hired. Of course, all the contractors did was hire ex-military and retired civil servants who had been trained directly on the government dollar.

    What is amazing is if one looks at A place like Aberdeen Proving Grounds, the test engineers are GS-13 and above civil servants who started out as activie duty military Lt’s there in the 1980’s. Then the engineer positions were civilianized and then the technicians and subordinates were contracted out. Now the managers have had a career path that cannot be reproduced in the future since the subordniates are outside the civil service and there to military with the needed skills.

  13. superdestroyer says:


    The contractors lined the pockets of many of the core Democratic groups. Do you really think that the contractors really play partisan politics?

  14. Andy says:

    It’s been a while since I’ve looked at this, but as I recall, most of the “growth in government” as measured by # of workers or even spending occurred at the local and state level. To be sure, there has been a lot of privatization in the federal sector, especially since the mid-1990’s, but I have not found any solid numbers to quantify that. As usual, analysis based on single metrics should be taken with a substantial grain of salt….

  15. @Andy:

    As usual, analysis based on single metrics should be taken with a substantial grain of salt….

    Indeed. I am pretty sure that that is why the author of the piece noted:

    Now, this is just one metric of the size of government

    I am fairly certain nothing was claimed in terms of this being a comprehensive metric.

  16. (And really, it isn’t so much that a grain of salt is needed, as much as recognition that there are more metrics needed for a complete picture).

  17. @Tsar Nicholas: Ok, so based on cherry-picking you still underscore that over a long period of time (the later 60s to the mid 90s) that the numbers have not changed much, save in the context of the largest economic downturn since the Great Depression.

    You are ultimately making the opposite of the case you think you are making.

  18. I would agree that things like government contractors over time would add a lot to this picture.

    However, I must say that what of folks are doing in this thread are rationalizing away the numbers presented. Interestingly, they are doing so by wishing for numbers that they don’t have, but just know confirm their beliefs.

  19. Rodney Dill says:

    I wonder how the automation due to computers during the last few decades actually affect the need for government employee’s (mostly non-military). There should be far less paper pushers, clerks, etc.. needed. The same number of gov’t employees should actually be able to get more value added work done now with automation than in the past. Presumably there would be more higher paid positions capable of doing more. Regardless of whether you value the “more” or not.

    (Yes I know I’m trying to talk about value added work, efficiency, automation, and the government all at the same time, but the more or less flat trend got me thinking)

  20. Rodney Dill says:

    @Steven L. Taylor:

    I would agree that things like government contractors over time would add a lot to this picture.

    On the surface I would’ve assumed that this chart meant the total government funded workforce, including contractors. If it doesn’t that could change things, though I couldn’t tell you either whether this is a good or bad level of gov’t workforce.

  21. Mikey says:

    Interesting, and relevant, and it even includes my Congressman.

    Rep. Gerry Connolly says federal workforce hasn’t grown since 1990

    This is consistent with Steven’s graph. However, there’s a catch…

    While official workforce figures have gone down in recent decades, use of government contractors has exploded, especially in the defense sector following the 9/11 terrorist attacks.

    In fiscal 2000, the government spent $208 billion on contractors. In fiscal 2009, that figure was up more than 150 percent, to $540 billion.

  22. @Rodney Dill: @Mikey: I have done some poking about for numbers on contractors, and I haven’t been able, yet, to find a an adequate time series. Indeed, I have not yet found a current number.

    There are a lot of them, however.

  23. Tsar Nicholas says:

    Doh, major typo alert. Yikes. Should read:

    In 1940 the federal government spent 118 billion (inflation-adjusted) dollars.

    In 2012 the federal government spent 3.6 trillion dollars.

    Oh, well.

    Speaking of banana republics, below are public debt-to-GDP ratios for various third-world countries:

    24% – Indonesia.
    65% – Brazil.
    70% – India.

    And here’s another noteworthy debt-to-GDP ratio:

    103% – USA.

  24. Rob in CT says:

    The contractors thing strikes me as both a key factor and something that’s hard to nail down. Also, how much state & local government employment has changed due to things done at the federal level. Last time I looked, a lot of governmental growth (basically all of it during my lifetime) was at the state/local level. Theoretically one could tease out how much of that growth was due to federal mandates, and assign it to the feds.

  25. @Tsar Nicholas: Comparative analysis doesn’t work this way. You can’t just grab three or four numbers and proclaim QED.

    Using 1940 as a point of comparison is especially problematic given that soon there after quite a lot changed for the federal government.

  26. Mikey says:

    @Steven L. Taylor: Apparently it’s well-nigh impossible to get a total, because nobody in the Fed Gov actually adds up all the contractors.

    And that might be difficult, anyway, because some contractors are on-site people who bill 100% of their time to a single contract and report to work at a Fed Gov facility, and some are people who report to work at a company location and bill portions of their time to several different contracts. How do we count the latter? If I bill 10 hours/week each to four different projects, I might get counted as four people.

  27. Tsar Nicholas says:

    And here are rough deficit-to-GDP ratios are various noteworthy times in recent U.S. history:

    5% – 1983 (“Reaganomics”)
    4% – 2003 (“the Bush recession”)
    7% – 2012 (“the Obama recovery”)

    In any event, obviously measuring the size of government over the decades by comparing numbers of federal employees utterly is preposterous. So much of what the Feds do these days is contracted out, rather than done by the Feds’ own payroll employees. But public money spent is public money spent; makes no difference whether the recipients receive Forms 1099 as opposed to Forms W-2. And of course technology is a colossal factor. Not every G-Man has a lipstick and stockings-clad secretary. Now we have such things as computers and such.

    Ultimately this debate largely is academic, pun both intended and unintended. Profligate federal spending along with taxes and regulations already has wrecked the economy and combined with the entitlement time bombs ultimately will destroy the economy. Think PIIGS but with higher crime rates. The liberal media-academe cabal, however, will not at all be concerned. They’ll simply ignore it. Look, over there, it’s a poll showing the GOP losing further ground with racial minorities. Yay!

    The future looks bleak. Spending money you don’t have never works out over the long haul.

  28. rudderpedals says:

    Plotting population or GDP on the same axis as fed employment would really highlight the gap.

  29. george says:

    Why not include the military in the size of government? They’re still supported by tax dollars, and unless they’ve been privatized while I wasn’t looking, they are in fact part of the government.

  30. bk says:

    @Tsar Nicholas:

    Not every G-Man has a lipstick and stockings-clad secretary.

    This pretty much confirms my long-standing belief that, contrary to your repeated assertions that you have a “job” with a lot of “clients” who pay you a “lot of money”, you basically are posting from your mother’s basement in your underwear eating Cheetos.

  31. @george: The military is included in the second graph.

  32. PD Shaw says:

    @Rodney Dill: There are comparisons of the composition of the public sector across OECD countries that show that over the past twenty years all OECD countries have shifted significant portions of their directly employed workforce to private contractors or (in non-US countries) quasi- public corporations. Regardless of political and economic variations, all of the advanced countries are having similar problems coping with employee total compensation that grows at rates in excess of GDP.

    In addition, many federal government services are performed by the states under federal guidelines and federal funding. I know state employees who essentially work for the federal government (particularly when their manager is furloughed). If the State of Illinois goes on a lengthy strike, we will see the federal government take back many jobs.

  33. An Interested Party says:

    Profligate federal spending along with taxes and regulations already has wrecked the economy and combined with the entitlement time bombs ultimately will destroy the economy.

    Does anyone actually believe this horse$hit?

  34. Andy says:

    @Steven L. Taylor:

    Excuse me, but if this was aimed at me then that is a wrong assumption on your part.

  35. JKB says:

    Good luck finding a reasonable estimate of the number of contractors. I’d expect you’d have to go through every contract as the information doesn’t have a central clearinghouse. Then, you’d have to investigate to see how many were legitimate outsourcing, such as landscaping and janitorial services, as opposed to de facto “personal services” contracts which are technically highly limited. But I saw plenty of instances where the contract was set up as outsourcing with the contractors providing services on-site but the contractor were really assigned work and essentially managed by the government supervisor, i.e., personal services. BTW, the military has also now outsourced a lot of support services formerly performed by military members.

    I once had the odd situation of writing a feasibility study on conducting an A-76 competition to contract out a group of jobs. I learned to my surprise that my contact at headquarters was herself a contractor and the person who was overseeing all the studies for the office was also a contractor (retiree). Different jobs and functions so not really a problem but a bit surreal to be submitting my study on the feasibility of contracting out jobs to a contractor who would submit it to another contractor who would sign off on it before submitting it to the boss for his “official” sign off.

    So really, the number of government employees tells you little about the size and growth of government. Is that by happenstance or design? If you’ve worked in government you know it is the former which resulted from millions of tiny designs over the decades.

  36. @Andy: It was not aimed at you, in fact.

  37. Pharoah Narim says:

    @Tsar Nicholas:Sooooo…are you saying you’re buying a one-way ticket to Indonesia? Or that our Kenya-born Indonesian-schooled President is the right person to pull us back from the Abyss?

  38. mantis says:

    @Tsar Nicholas:

    Spending money you don’t have never works out over the long haul.

    And Tsar comes out against the very concept of credit. What would we do without this genius to school us every day?

  39. Dave Schuler says:

    @Rodney Dill:

    The chart is derived from agency 113-A monthly reports which expressly do not include private sector contractors.