Does annual real gross domestic product per capita overstate or understate the growth of individual welfare over the past two centuries?

• Author: Van Den Berg, Hendrik

The decrease in the death rate, and the attendant increase in life expectancy--more than doubling--during the last two centuries in the richer countries, and in the twentieth century in the poorer countries, is [sic] the most stupendous feat in human history.

--Julian Simon, introduction to The State of Humanity (26)

Angus Maddison's (2001) historical estimates of annual real gross domestic product (GDP) per capita show that nearly all the economic growth in annual real GDP per capita that has ever occurred in the world occurred after 1820. Maddison's estimate of 1820 real GDP per capita of $667 at 1990 prices and others' estimates that the real output per capita at which humanity can survive is approximately $250 per year at 1990 prices imply that average world real GDP per capita little more than doubled during the first several hundred thousand years of human existence. (1) During the past two hundred years, the world's average real GDP per capita has increased nearly ninefold. In the more developed countries, the growth of annual GDP per capita has been much greater. Figure 1 illustrates the recent explosion of the world's average annual real GDP per capita.

The accuracy of the stunning picture of recent economic growth shown in figure 1 depends on whether annual real GDP per capita correctly reflects human welfare and economic growth. Many economists have accepted real GDP per capita as a reasonable proxy for human welfare. For example, Simon Kuznets explicitly defined economic growth as "a sustained increase in per capita or per worker product" (1966, 1), and Douglass C. North and Robert Thomas wrote that economic growth "implies that the total income of society must increase more rapidly than population" (1973, 1). However, although few people deny that Maddison's carefully researched historical data indicate what happened to annual real GDP per capita over the past two hundred years, some authors have objected to using real GDP per capita to measure human welfare.

A common theme in the economic development literature has been that the growth in average annual real GDP per capita overstates the improvement in human welfare. (2) Some economists have suggested alternative measures of economic growth that focus on health, education, political rights, and many other conditions that are equated with the quality of life. Julian Simon (1995a) and others have pointed out that virtually all measures of human welfare have improved greatly over the past two hundred years, in which case the general impression given by figure 1 is not misleading. (3) However, because most of the other measures have not improved by as great a multiple as real GDP per capita, the perception remains that the growth in annual GDP per capita overstates the improvements in human welfare during the past two centuries.

In this article, I argue that the various alternative measures of human welfare do not remedy a serious fundamental shortcoming of annual real GDP per capita as an indicator of human welfare, which is that it fails to measure the lifetime welfare of individuals. The many alternative measures of human welfare that have been proposed are, like GDP, also annual flows or point-in-time stocks, none of which tells us anything about individuals' lifetime welfare. A simple measure of average individual lifetime welfare fortunately can be constructed easily from readily available data. This measure indicates that the recent growth of individual welfare actually has exceeded the growth indicated by annual real GDP per capita.

Alternative Measures of Human Welfare

Economists have suggested many alternative measures of human welfare, including life expectancy, infant mortality rate, caloric intake, access to safe water, adult literacy rate, school enrollment, the distribution of income, hospital beds per capita, and a long list of other similar measures of the "quality of life." One especially popular approach to improving on GDP measures of growth has been to use a weighted average of several alternative measures of human welfare. Typical of this approach is the well-known Human Development Index (HDI) published annually by the United Nations Development Program (UNDP). The HDI is a weighted average of real GDP per capita, a measure of health, and a measure of education.

Representing many development economists' belief that the welfare effect of marginal income declines as income per capita rises, the HDI decreases the weight of real GDP per capita relative to the other measures as real output per capita rises. In fact, until recently, increases in real output per capita carried weight in the HDI only for low levels of annual output per capita. (4) Specifically, before 1999, the relationship between annual real GDP per capita and human welfare that the UNDP assumed was the one depicted by the solid "kinked" line in figure 2. Thus, increases in GDP per capita up to the world average of about $5,000 had a strong effect on the value of the overall HDI index because, as the solid line in figure 2 shows, in that range small increases in GDP per capita were assumed to bring large increases in human welfare. After the world's average GDP per capita of $5,000 was reached, however, further increases in output were assumed to matter only slightly for human welfare, and hence further growth of output per capita was deemed to be of little importance. This approach meant that the great differences in real output per capita between Japan and a middle-income country such as Mexico made almost no difference for the two countries' relative HDI indices. The idea behind this approach was that people have certain "basic needs," and once these needs are satisfied, further increases in consumption will be largely superfluous. (5)

Beginning with the 1999 HDI, a more "curved" relationship between income and welfare was assumed, at the urging of Sudhir Anand and Amartya Sen (1999). Sen writes: "we have excellent reasons for wanting more income or wealth. This is not because income and wealth are desirable for their own sake, but because, typically, they are admirable general-purpose means for having more freedom to lead the kind of lives we have reason to value.... The usefulness of wealth lies in the things that it allows us to do--the substantive freedoms it helps us to achieve" (1999, 14). In other words, economic growth increases positive economic freedom, and that increase has real welfare implications. As a result of such reasoning, the 1999 HDI was modified so that the relationship between real GDP per capita and individual welfare was assumed to be the one shown by the dashed curve in figure 2 (UNDP 1999). Note, however, that although the revision of the HDI accepts that welfare always is related positively to real GDP per capita, the HDI still does not assume a constant one-to-one relationship.

Apart from the possibility that welfare gains from GDP growth might be nonlinear, the growth of annual real GDP per capita may exaggerate improvements in human welfare for other reasons. A well-known shortcoming of GDP is that it counts only legal market transactions, omitting household production, illegal or "underground" production, and leisure despite their effects on human welfare. The omission of household and underground activity, for example, causes the growth in official GDP to overstate true growth in production if the unrecorded economic activity shrinks as a percentage of total economic activity. As Scott Fuess and I (Fuess and Van den Berg 1996, 1998) found for the United States and Mexico, this development probably has occurred in many economies as women increasingly have entered the formal labor market and have substituted paid work for previously unvalued but valuable housework. GDP may also overstate real growth in output because it does not account for the depletion of nonrenewable resources. Martin Weitzman (1999) has estimated the cost of resource depletion in order to derive net domestic product (NDP), and he has found that the annual depletion of the world's major minerals is equivalent to approximately 1 percent of real GDP per year.

Growth of Real GDP per Capita May Understate...