Enter into any political debate and it is likely that it will devolve into a debate about the size of government. It is not enough to say things like “well the leftists want more government,” or vice versa, because both sides want either option. The number of useful bills that went unsupported because they “increased the size of government” is likely very large.
Whether a particular bill increases – or decreases – the size of government is the wrong question to ask. If a particular idea solves a perpetual issue, why should the finality of that idea depend on whether it increases or decreases the size of government? Relying on such an indicator adds inefficiency to an already burdened system, but serves only as an appeal to a rigid philosophy.
Measuring The Size of Government
How one measures the size of government matters. There is – and must – be a distinction between the philosophy of things, and the reality of things. Often, these two conceptualizations of the same issue conflict. Additionally, what ought to be may not be what is, and vice versa. What ought to be is also subject to debate.
For example, the “design” of the federal government to be “as small as possible” matters little if necessity begs to differ. Again, “small” depends on how you measure your system. Similarly, big taxes require context – like most other issues – in order to objectively determine if this is a valid concern.
All too often debates get so entrenched in the philosophy of things that reality gets pushed aside. This is evident even when we look at seemingly objective analyses.
For example, the Brookings Institute measures the size of “big government” by how much it spends, how much it does, and how many people it employs.
If we measure the size of government by the dollar amount spent – as is often done – then yes, the size of the federal government is astronomical.
The amount of money being spent is often what is cited to reflect the preposterous size of the federal government. Indeed, federal outlays in 1950 were a little over $42-billion dollars. By 1986, federal spending passed $1-trillion, and by 2018 that number has quadrupled to more than $4-trillion dollars. The difference in spending between 1950 and 2018 is nearly a 10,000% increase (this isn’t an exaggeration).
Dollar amount isn’t a complete picture, however. When we look at federal spending as a proportion of GDP, we see far less growth. Put another way, the size of government seems vastly deflated when we look at spending proportionally.
In 1950, federal spending was about 14.2% of GDP. In terms of 2018, net spending was 20%. The 68-year average is 19.14%. You can see a few trends in this graph.
While Eisenhower was president, federal spending as a percentage of GDP fell by 13% from his first day in office to his last. Ronald Reagan saw his term with a net 4% decrease, but from peak spending it was a 9% cut. Gerald Ford increased spending as a percentage of GDP by 14%. Bill Clinton saw spending decrease by more than 15% from 20.5% to 17.5%. George W. Bush immediately reversed Clinton’s trend by increasing spending by 15%. Interestingly, Obama’s first year in office saw the highest percentage of spending benched to the GDP, at 24.4%. However, by the time Obama left office, spending had decreased by 15% again, closing at about 20.6%, which is about where it was at when Clinton entered office. Although members of the political may feel a strong need to criticize this explosion in spending, the net impact on the economy was positive.
All of the above is to point out that the political leanings of the president in power seemingly have little to do with whether federal spending as a percent of GDP goes up or down. Additionally, under every president, the amount of money spent has increased.
Returning however to whether or not the government is too big or too small depends on how one looks at the data. Dollar value suggests that the government is huge, whereas the percentage of spending per GDP suggests we’re about where we should be.
Turning to the issue of federal employees, the numbers also show a sense of stagnation.
This graph does not support the idea that the government has grown substantially, or even continues to do so.
The graph above puts the previous one in context of population. It goes to show that as the population has increased, the number of federal employees per 1,000 civilians has decreased. In effect, the number of federal employees has gone from 10 per 1,000 in the 1970’s to about 6.5.
To this end, while members of the Brookings Institute lament that the federal government has created seven new executive departments since 1953, along with “dozens of sub cabinets” – this seems over hyped. With such an ‘explosive’ number of cabinets, we ought to have expected significantly more employees. We have not seen such a result.
A caveat, however, is that the government has shifted to utilizing non-profit organizations and contractors. It is disingenuous to classify these workers as federal employees. It is a valid argument that the number of contractors can lead to a ‘crowding out’ effect, though.
The United States Code
Another measure often employed is by looking at the size of the United States Code. Arguably, measuring the size of government on the size of the United States Code is also weak. While it’s true that the code is large, a better solution to sizing up government is making it more efficient. Regulation is an important element to a thriving economy.
One possible solution to moderating the growth of the United States Code would be to implement sunset clauses. This would be a proactive measure to ensure that laws don’t remain on the books that are unnecessary. It would free up resources when creating new laws. These also add efficiency by ensuring that laws are passed with the intent to resolve short term problems.
One token of comfort to offer up is the fact that the number of bills being passed by Congress has been declining since the 100th Congress. Polarization will at the very least be a stop gap in terms of the growth of the United States Code. The expense, of course, will be that nothing will get resolved, which inflates other issues. To put this in perspective, the rate at which laws were passed in the 100th Congress was a little over 8%. As of the 114th it is less than 2%.
Center for Freedom and Prosperity
Reasons offered as to why “big government” is bad are myriad, but the Center for Freedom and Prosperity offers four reasons: 1) crowding of the private sector, 2) higher taxes, 3) higher debt, and 4) government was designed to be small.
Taxes, Debt, and the “Small Design”
First, taxes. The same trend appears for taxes as it did for federal spending. In other words, when looking at the total dollar amount, it appears that the United States is spending a lot in taxes, and it is. From 1965 to 2017, the amount of actual dollars that the federal government accrued increased by over 3,000 percent.
But again, the number of dollars is not the whole picture as this reflects more of a raw outcome in terms of data.
Tax revenue, as a proportion of GDP, is again more stable. The minimum percent of GDP was 22.5% in 1965 with an average of 20% over the whole time period. Put another way, from 1965 to 2017 the tax burden has been 20% of GDP plus or minus 3%. Whether or not 20% is too high or too low depends on the situation.
Second, whether the government was designed to be small or not is up for debate. A better question asks how flexible we ought to allow the federal government in adjusting its size according to necessity. Additionally, the size of the government in present time may still be proportionally small, yet bigger than was ‘designed’ 200 years ago. Appeals to bygone eras are a poor way of addressing government inefficiencies.
Third, higher debt. This is a valid criticism against large government as debt is certainly a symptom of government spending. Government spending is a symptom of larger governments. However, fiscal responsibility and better management of programs may be a better solution than simply “spending less.”
For example, Americans spend about $2-trillion dollars in medical care. What’s more staggering than that price tag is the estimated 18 to 37% that is figured to be waste. In other words, because of redundancies and inefficiencies, Americans waste a significant amount of money on medical care. This isn’t just a problem for the medical field, though. Simply cutting funding would thus not fix the solution. Cutting funding, for the sake of limiting government, may well exacerbate problems. Those problems would then feed into economic turmoil, and stagnation.
Crowding of The Private Sector
An often made argument against big government is that it overcrowds the private sector. This is called the ‘crowding out effect‘. Put another way, when the government floods the market with dollars, private investors lose out on opportunities. Causes for this are tied to basic economic principles, social welfare spending, and infrastructure. This is the short version, of course, and it’s a well known economic phenomenon.
Advocates for smaller government have a firm hold when arguing this case. Bolstered by well known economic principles is the fact that the federal government is contracting out non-profits much more than before.
On the one hand this means that the government is “smaller,” as measured by the number of federal employees. On the other hand, by doing this the government accrues more debt. The more debt that the government accrues the more that it must rely on selling government bonds. The more government bonds there are in the market, the fewer corporate bonds will then be invested in. This has the effect of ‘pushing out’ investment options in favor of government bonds. It’s important to note, however, that this is not a dollar for dollar match, and interest rates matter here.
However, as is the theme, this is not the whole picture. The reverse of this effect is the ‘crowding in effect‘. This phenomenon is most pronounced in times of economic decline. The less complicated description of this effect is that the government can help spur on economic growth by said saturation. It’s the same principle that leads to the crowding-out, with the functional impact of healing the economy instead of harming it.
All of the above is to say this: rigidity in the model of government can have detrimental impacts on both efficiency, function, and growth. If you believe that government should always be small, even when flexibility is warranted, and even necessary, then what good is such a model?
The Armey-Rahn Curve is a statistical model that attempts to provide an objective measure of an optimum size of government. The model does three things. First, it suggests that for new and developing countries, federal spending has the effect of speeding up economic growth. Second, there is a peak performance, which is represented as the peak point of an inverse-U shape, and depends on a variety of factors. Third, too much federal spending (or taxation) has the effect of slowing down economic growth.
The Armey Curve can be calculated in a variety of different ways, and has been applied to over one hundred countries in twice as many studies. The size of government is usually determined by the proportion of expenditures to GDP. Economic growth is typically measured as the growth of rate of GDP over a given time.
Humorously enough, this curve is by some measure an objective strike against anarchists who would see to it that government is abolished. Under such a system, economic growth would cease to exist. The same would be true for anarcho-something-somethings that would like to retain capitalist free trade without a government in place. The curve also puts a cap on other systems that would over inflate federal spending as well, though, as this would have the same effect as having no government at all.
Having said that, the Armey Curve can give very different results based on the criteria that is plugged into it. For example, Robert Barro’s 1990 analysis in the Journal of Political Economy suggests that peak performance occurs when expenditures, as a ratio of GDP, are a little bit above 25%. Richard Vedder and Lowell Gallaway’s analysis, which was presented to the Senate’s Joint Economic Committee in 1998, determined that expenditures should be a little bit below 18%.
Another way to generate an Armey Curve is be replacing federal expenditures with average federal, state, and local taxes as a percentage of GDP. For example, Gerald Scully’s 1998 report from the National Center for Policy Analysis determined that economic growth would be nominal if average taxation was between 21.5 and 22.9%. As seen above, federal tax revenues in the United States have been 4 to 10% above that for quite some time.
According to Scully, if average taxes were within his calculated ranges, growth rate of GDP would have been approximately 5.56 percent per year. Strikingly, Scully does not address the issue of inflation, as high GDP growth rates often produce. Economists generally agree that long term sustained growth at that level leads to recessions, not just inflation. An annual growth of between 2 and 4 percent – but somewhere around 3 – is considered ideal annual growth.
Considerations On Measuring The Size of Government
There are two observations to consider when looking at models such as the Armey Curve. First, such models often misconstrue correlation with causation. While relationships may be linear, and statistically significant, they still may not accurately reflect causation.
A second consideration is that when we focus on a “peak optimal performance” of a government, or economy, small fluctuations can be seen as devastating. For example, the average expenditures as a ratio of GDP since 1950 has been a little over 19%. If peak performance according to Vedder and Gallaway is 18%, is a 1% different so dramatic that we must cut entitlement programs by 15% (or some other arbitrary number)?
The same argument suggests that if peak performance is 25%, then we’ve been lagging behind based on a 70-year average. Presently, this ratio is 20%, so in order to reach peak, the government should actually spend more money. But again, this is entirely based on the assumption that government expenditures cause the economy to go up or down.
It’s also important to point out that the Armey Curve is more a reflection on how to obtain peak economic growth. This means that a 1% difference in either direction simply means faster or slower economic growth. It does not mean that we ought to go into a panic.
Moreover, because the curve relies entirely on government spending and gross domestic product, the curve does not take into account how that money could be spent in a way that would maximize economic growth. Such minutiae of spending would not appear in this model, and as such efficient spending that maximizes growth may well still be calculated as having surpassed optimal size of government.
Weaknesses of this model in other ways. For example, De Witte and Moesen’s 2009 piece in Public Choice found similar results to that of Scully by using average taxation. However, upon further inspection, the authors found that more than 30% of the rate on taxes could be explained by the size of a family. Additionally, countries that have larger families also tend to spend more on social programs. This, in part, was able to explain why some OECD countries had far larger peak economies with taxes being somewhere between 30 and 40% in contrast to the United States’ suggested range in the 20’s – because American families are on average smaller.
One of the most significant observations with the Armey Curve, however, is that while the curve seems universally applicable, it is immensely subject to variation. For the United States alone, studies on optimal economic growth on the percent of expenditures varies from 15 to 30%. Measures replacing expenditures with taxes as a percent of GDP range from 19 to 32%. Facchini and Melki’s 2011 article goes in to great detail about the differences among countries. Openness to free trade, date ranges, and even social attitudes towards larger governments can significantly impact this model.
Verdict: Big Government or Small Government?
When we combine things such as wasteful spending in healthcare, and government spending, we can get different results. Consider that the government were to reform the system, and thus save upwards of $900-billion dollars, the result would reflect in the Armey Curve that we’re not spending enough money. In other words, the size of government matters very little if how it performs its functions is wasteful.
Additionally, countries such as the Netherlands, and Sweden, that have public expenditures in the 30 to 40% of GDP range are considered to be in optimal performance based on the Armey-Curve. Yet, such governments are measurably larger using the same standards used here to measure the size of governments.
Rather than focusing on the size of government, a focus on function is more relevant. The idea that government ought to be small, or nonexistent, is not supported by more objective measures.