To measure a society's living standard, economists usually consult income data. But economist Dora Costa of the Massachusetts Institute of Technology thinks a better measure would be what people are spending on recreation.

Costa does that in a new study which looks at surveys of consumer spending from the late 1800s to the present "to determine whether trends in real income per capita are consistent with trends in recreational budget shares" -- by which she means the percent of income that people spend on leisure: reading, going to the movies or playing sports, for example.

• She found that in the late 1880s, less than 2 percent of household expenditures were devoted to recreation

• By the mid-1930s, the recreational budget share had risen to 4 percent -- reaching 6 percent by 1991.

If reported income falls while the recreational budget rises, that's a sign reported income is understating living standards. Income changes likely underestimate the increase in living standards -- particularly during times of innovation in consumer goods and reduction in work hours such as in the 1920s and the 1970s and 1980s, she found.

• Although reported per-capita income fell 1.2 percent per year from 1919 to 1935, by looking at changes in recreation spending during that period, Costa concludes that per-capita income actually grew 1.2 percent annually.

• For the period 1972 to 1991, reported income rose 1.8 percent per year, but consulting recreational-spending figures convinces Costa that the true figure is more like 3.6 percent increase per year.

Moreover, lower-income households experienced a larger increase in living standards than higher-income ones. One reason is that in the 1890s, workers in the lowest decile of wage distribution labored nearly 11 hours each day, while those in the highest decile worked nine-hour days. Now, it's the highest-paid workers who put in the longest days.

Source: Macroscope, "How Much Better?" Investor's Business Daily, June 28, 1999.