Looking beyond GDP when measuring welfare
Gross domestic product often gets confused for a measure of economic well-being. That’s understandable given how often GDP or GDP growth rates gets held up as the key measure of economic health. But should policymakers really care about how much an economy produces every year in and of itself? Economists (and many other people) would agree that increasing economic output is good in so much that it allows people to consume things, live longer, and have some leisure time. GDP, which is simply the value of all final goods and services produced in a country in one year, doesn’t fully capture those conditions. A new paper takes a stab at creating a measure of economic welfare.
The paper, by economists Charles Jones and Peter Klenow of Stanford University, was published in the latest issue of American Economic Review. The two economists try to move beyond GDP and build a measure of economic welfare that uses data on consumption, leisure, any inequality evident in those two variables, and life expectancy in a country. Why those data? In Jones and Klenow’s model, a person’s welfare is based on how much they can actually consume and how much time they can actually take off. If overall consumption is a small percent of income or worker hours are long, then economic welfare will be lower for a given level of economic output. The same goes for consumption or leisure inequality—if they are high (consumption and leisure are concentrated among a few people) or if life expectancy is low (a person might not be around long enough to enjoy consumption or leisure) then the average economic welfare will be lower than income alone would suggest.
Before digging into the specific results from the paper, let’s be clear that the measure of well-being developed by Jones and Klenow is dependent upon the assumptions of their model. The economists run the model with a number of other assumptions as a check on their results and they hold up. But the model-based nature of the results should be noted.
The results end up being quite interesting. While there is a very strong correlation between the two authors’ measure of welfare and GDP per person, comparisons between countries are different than if we used GDP per person. If policymakers were to compare the United States to France by average consumption, for instance, then France’s living standard is only 60 percent of the U.S. level. But using a measure that includes France’s lower inequality, lower mortality rate, and more leisure, France’s welfare-based living standard is about 92 percent of the U.S. level, with inequality, mortality, and leisure all boosting the relative standard equally.
All Western European countries in the data set developed by Jones and Klenow end up looking closer to the United States based on this measures. Lower-income countries, however, appear further away from the high-income countries due to much lower life expectancy and high rates of inequality.
Jones and Klenow’s measure also can be used to make comparisons over time. Their results show that standards of living measured by welfare have grown much quicker than standards based on income growth. The difference—3 percent per year versus 2 percent per year from the 1980s to the mid-2000s—might not seem large, but it’s the difference between standards of living doubling every 24 years instead of in 36 years.
The new paper by Jones and Klenow makes for very interesting reading, but it’s only the beginning of a new effort. The two economists point out that with more and better data they could loosen some of their assumptions in the model and that there are other applications of their model for building measure of welfare. Hopefully, they and others looking at how to better capture trends in living standards will continue to work hard on these questions. But maybe, for their welfare’s sake, not too hard.