Lies, damn lies, and conventional measures of the growth of government.

• Author: Higgs, Robert
• Position: Etceteras ...

Many of us who believe that governments continue to grow relentlessly, at least in the economically advanced countries, have been criticized by analysts who claim that in fact the growth of government has petered out or slowed substantially. The latter group perceives us to be needlessly alarmed and faults us for a failure to acknowledge the decisive turn of events associated with the so-called Reagan and Thatcher revolutions of the 1980s. Not to worry, they exhort us; the statists are on the run, and a brave new world of market-oriented liberalism shimmers on the horizon.

My thesis here is that these seemingly level-headed realists are the ones who have failed to perceive correctly the ongoing growth of government. (1) A major reason for their failure is their reliance on certain conventional measures of the size and growth of government. Some of these measures have a built-in tendency to exhibit deceleration even when a more compelling representation displays continuing steady growth. Often the conventional measures miss the growth of government because it has been diverted into channels beyond the scope of their measurement. To some extent, governments have been growing in important but unmeasured or poorly measured ways all along, and they continue to grow in these ways, perhaps more menacingly than ever before. Off-budget spending, for example, is a well-known resort of political scoundrels, but it is only one example among many of how governments employ hard-to-measure means to achieve their usual ends, especially when tax revolts, formal spending limits, or borrowing limitations frustrate their chronic desire to spend at a greater rate.

Government's Share of Gross Domestic Product

The most common measure of the size of government is the amount of government spending relative to gross domestic product (GDP). In a recent monograph on the growth of government, for example, Vito Tanzi and Ludger Schuknecht present much of their data in the form of government-spending variables relative to GDP. A major theme of the book is: "Government spending [measured in this way] increased most rapidly until about 1980. Since the early 1980s, it has been growing more slowly and in some instances has even declined" (2000, 3).

Now, the first thing to notice is that a sure-fire way to make nearly any economic magnitude appear small is to divide it by GDP, because the latter, which purports to be the total value at market prices of all final goods and services produced within a country in a year, is always an enormous dollar (or euro or peso or other currency unit) amount. Government spending of $2,855,200,000,000, as in the United States in fiscal year 2001, seems to be an astronomical amount, but just divide it by the value of GDP and, voila, it is a mere 28 percent--surely nothing to be alarmed about, especially in comparison with corresponding figures for many European countries that exceed 50 percent. (2)

The next thing to notice is that because government spending for currently produced final goods and services is itself a component of GDP, the ratio of the former to the latter is immediately compromised. Any addition to such government spending increases the denominator as well as the numerator of the ratio. Suppose that in year one the government spends $100 dollars for currently produced final goods and services, and the GDP in that year is $500. Now suppose that in year two the government spends twice as much--that is, it increases its purchase amount by 100 percent--but nothing else changes. In year two, the government's share of GDP will be 33.33 percent (or $200/$600), as compared to 20 percent in year one. An analyst focusing on the government's spending share concludes, then, that government has grown not by 100 percent, as it plainly...