Shorter Hours Raise Living Standard of the Poor

09/01/1998
Summary of working paper 6419
Featured in print Digest

In the past, the lowest paid workers toiled away the longest while now it's the high wage workers that put in the most hours.

Many Americans feel they are working way too hard and for far too many hours. The stress is real enough. But the length of the workday has dropped considerably over the past century. The typical worker in the 1880s toiled away for ten hours a day. His 1940s peer labored for eight hours. Time diary studies suggest that today's average employee logs in less time at work than a strict 9-to-5 schedule. The steep decline in hours worked is an under-appreciated improvement in American living standards.

Yet the gains in hours worked haven't been distributed equally. In the past, the lowest paid workers toiled away the longest while now it's the high wage workers that put in the most hours, according to NBER researcher Dora Costa. In The Unequal Work Day: A Long Term View (NBER Working Paper No. 6419), Costa finds that workers in the 1890s earning less than 90 percent of all workers labored nearly 11 hours while those making more than 90 percent of all workers labored for almost 9 hours. In 1991, the comparable figures were roughly 7.5 hours and 8.5 hours, respectively. "I find that although hours of work have fallen for all workers, the decline was disproportionately large among the lowest paid workers," Costa writes.

Like the distribution of income, the distribution of hours worked has implications for trends in income inequality. In essence, the common wage and wealth data for measuring inequality may underestimate long-run improvements in living standards among low paid workers. For instance, Costa calculates that between 1973 and 1991, 26 percent of earnings inequality among male workers between the top and bottom groups described earlier could be attributed to differences in hours worked. "In the past an inegalitarian distribution of work equalized income whereas today it magnifies earnings disparities," says Costa.

A related paper, The Wage and the Length of the Work Day: From the 1890s to 1991 (NBER Working Paper No. 6504), delves deeper into time at work and inequality. For example, in addition to the role of hours worked in earnings inequality among men from 1973 to 1991, Costa finds that the same phenomenon accounts for more than half of the earnings inequality among women and 17 percent of the increase in total household earnings inequality among husband and wife households.

What is behind changes in hours worked? In both papers, Costa casts doubt on the idea that state legislative actions limiting the workday had much impact. The big shifts in hours worked was in effect by the 1920s, yet laws limiting the workday applied mostly to women and to a few men in dangerous industries until the 1930s. Instead, changes in the willingness of workers to supply their labor dominated. In the past, worker hours were very sensitive to changes in pay. From 1890 to 1919, real wages increased by 43 percent and the work day fell from 10 hours to 8 hours.

In contrast to a century ago, increases in the hourly wage no longer encourage people to take more time away from the office or factory. For one thing, workers are not as time poor as they once were. For another, with incomes higher than before, the impact of a wage hike is smaller. "Regardless of the reason for the change in the distribution of work hours the results of this paper imply that although the rich and the poor will always differ in terms of income, income differences no longer mean that the poor have less time for fun," says Costa. Now that's progress!

-- Chris Farrell