Measuring the Inflation of Government

Those who follow the political discourse in Rhode Island with some regularity will have a sense of Tom Sgouros’s conclusion with his “Budgeting for Disaster” series before he begins: Under all the complexities of numbers, the principle is very simple, and the solution to Rhode Islanders’ pain is more tax revenue and bigger government.

A key component of his argument comes in “Part II“: “no matter how you slice it, state and local taxes are taking up a smaller share of the state’s economy today than 20 years ago.”  Using data from the federal Bureau of economic analysis, one can dispute Sgouros’s specific calculations.  For example, he measures total government revenue against total personal income, which may be a reasonable way to compare one state to another, depending what one is intending to discover, but it’s less reasonable when dealing with a single state over time.  The BEA categorizes taxes separately as “personal income” and “taxes on production and imports” (including sales and property taxes) for a reason.  Different arguments apply to the growth of government compared with how much its people possess and what its people are doing.

It also isn’t clear why Sgouros mixes BEA income projections with the governor’s tax revenue projections.  He finds that applying the governor’s growth estimates to “the 2011 BEA income number” (which doesn’t actually appear to have been released, yet) projects $49.7 billion in personal income in RI for 2013, but the governor’s budget projections predict $48.5 billion.  He also appears to have mistyped the governor’s projected total revenues, which Sgouros cites as $3.27 billion, but the budget calculates at $3.37 billion.  If we apply Sgouros’s higher income projection to the percentage of income that the governor plans to claim in taxes and add back in the missing $0.1 billion, we find total general revenue of $3.39 billion, which amounts to 6.9% taken out of the economy, not 6.6%, which Sgouros compares with 8.25% in 1992.

That’s not a huge change, to be sure, but it’s indicative of the variations that are available in this sort of analysis when one jumps from one data set to another.  As it happens, the BEA data also offers figures for the amount of taxes collected in Rhode Island, in all its forms, each year, and using its data exclusively, it’s possible to argue a variety of contrary percentages contrasting with the 20% reduction that Sgouros claims for the government’s percentage of the economy.

The BEA doesn’t project future taxes, and not all 2010 tax numbers are available, but if we estimate 2010 revenue (using the the data that is provided) and compare the same number of years, from 1989 to 2010 we could state the following:

  • If we only consider the taxes that the BEA itself counts against personal income, as opposed to GDP, such as income, capital gains, and personal property taxes, state government’s proportion of the local economy only decreased by 2.6%.
  • If we consider total taxes taken out of total GDP (adjusting for a change of accounting methodology in the late ’90s), we find that the government increased by 0.9%.
  • Or, if we do the closest thing to what Sgouros has done and count total taxes against personal income, the government’s take has increased by 3.5%.

But let’s stipulate that there are some ways to cut the numbers that allow us to say that the government’s share of the economy did go down; that finally brings us to the most important question in this entire analysis, which Sgouros breezes through in a single paragraph:

The inflation rate is calculated based on a household’s expenses, and it looks at the rise in prices of the things that people buy: groceries, gasoline, auto insurance, medical services, TVs, washing machines, and so on. Your state government does not actually buy a lot of groceries, and I’m betting that you do not buy a lot of services from corrections officers or park rangers. That is, there’s no reason to think that the inflation rate tells you anything useful about how much government spending should grow.

Using Sgouros’s numbers, from July 1992 to June 2013, the annual tax collections of the state of Rhode Island increased by 88%.  According to’s calculator, price inflation was 73% from July 1990 to June 2011 (the closest fiscal year match possible).  As he says, that looks like “government spending is out of control.”  What he wants to do, instead, is to use personal income as the gauge of government inflation, in order to argue that the state should actually be taking almost another billion dollars from the local economy.

Even using his own examples, though, that argument is deeply flawed.  An individual household may not hire many “corrections officers or park rangers,” but it does spend money on various forms of security and landscaping.  The government may not “buy a lot of groceries,” but it does spend a whole lot on “medical services.”  The two entities are more parallel than he insists.

More importantly, Sgouros departs from the question that he’s ostensibly answering:  As a portion of personal income, should government services increase more like other goods and services that people purchase or more like their income?  Government may have decreased as a percentage of total wealth, but all of the other items that Sgouros lists as components of inflation have decreased more.

That’s precisely what one would expect.  A key way in which the economy grows is by expanding the set of things for which people are willing to work, while making that work more productive.  We don’t expect every expense in our budgets to grow at the same rate as our income as we find ways to earn more; some go down or disappear, while others go up or materialize.  Your cell phone bill has likely skyrocketed with the development of smartphones and data services, but you’re getting more from the device, and being made more productive.

The question, therefore, is whether government is more like consistent expenses, like home maintenance, that keep up quality of life, or more like dynamic expenses, like technology, that contribute to continued advancement.  That’s a matter of philosophy, and there are sincere disagreements about it.  The budgetary pain about which Sgouros complains has resulted because the expansion of government activities that people were willing to permit when times were flush requires a reduction when times are not.  Some would argue that a wealthier society needs fewer government services even in absolute terms, as living conditions improve and private fortunes create more resources for private investment and charity.

Whichever way one leans, Sgouros begins with a point that must be argued, not asserted.