Friday, February 03, 2006
Labor, Leisure, and Inequality
by Tom Bozzo
A curiosity, in the minds of the authors, is that measured leisure increased most for the least-educated adults:
Meanwhile, phenomena such as automation and self-service methods in a variety of industries have made it increasingly suck to be less educated and/or lower-skilled over the time period in question. What seems to be happening, by and large, is that market labor supply among the less-educated is significantly lagging that of the better-educated, and the residual of market and non-market labor is leisure. I wonder how much of the observed labor/leisure phenomenon represents a smooth labor/leisure tradeoff, versus a more complex balancing of costs and benefits pertaining to trying to string together a workweek from multiple McJobs.
As for the well-to-do, the presence of social consumption externalities such as Robert Frank describes — à la, it's not good enough to spend a week at a lake for vacation, it has to be a week at a lake in Patagonia to generate an impressive-enough story for the next neighborhood potluck — may well make the labor/leisure picture a lot more complicated than the standard model suggests.
One of the more controversial claims in the paper is the suggestion that the faster growth of leisure among the less-educated may represent an offset to traditional measures of income and wealth inequality. Part of that would hinge on things like utility comparisons of sitting around watching TV versus upper-middle to upper-class extreme consumerist leisure, which are all but a lost cause empirically. It's also less than entirely clear what the decline in household labor means for the total economic consumption (i.e., of market and non-market goods) of households. Or, who's preparing the same meals more efficently within the household, versus substituting household production for measured market services.
Last, I could pick a nit with the authors' definitions of "leisure" — note that I said "measured leisure time" back in the lede. One of the data problems is consistency of measurement of child care; varying amounts of child care are shifted between household labor and leisure in the measures in part to address this. However, "recreational child care" (i.e., playing with one's children) is included even in the narrowest leisure definition. Now, I'll admit that given a choice between changing a diaper and building LEGO spaceships with John, I'll take the LEGO. But I don't call him Ming the Merciless for nothing. Kids' play and adults' leisure choices are not well aligned, and thus playing with one's children has a substantial element of work to it.
Cold Spring Shops' Stephen Karlson points to an interesting Boston Fed working paper by Mark Aguiar and Erik Hurst, "Measuring Trends in Leisure: The Allocation of Time Over Five Decades" [pdf]. The authors string together trend data on labor and leisure time allocations between 1965 and 2003. The not entirely shocking findings are that net measured leisure time has increased markedly, mainly reflecting decreased market labor supply by men and decreased time spent in non-market household work by women (in the latter case, offset in large part by increases in paid work time).
A curiosity, in the minds of the authors, is that measured leisure increased most for the least-educated adults:
Given that the least‐educated households experienced the largest gains in leisure, this growing “inequality” in leisure is the mirror image of the well‐documented trends in income and expenditure inequality. The fact that the least‐educated experience the most leisure poses an empirical puzzle for the standard model that relies solely on income and substitution effects: The time‐series evidence suggests that rising incomes induce greater leisure, while the recent cross‐sections suggest that higher incomes are associated with lower levels of leisure.I'm not sure how much of a puzzle this really should be to a careful analysis. A commonplace of the conservatarian argument that the working classes never had it so good is that the market penetration of various household electrics and electronics has soared, implying that only the truly impoverished lack them, mirroring sharply falling relative prices of those goods (not to mention the technology of easy credit). Though data presented in the paper shows that declines in household labor have been faster for the better-educated — defined in the paper as holders of high school diplomas or better — then for the lesser-educated, so it's not suggestive that there's actually any effect coming out of the poor catching up with labor-saving conveniences over the period in question.
Meanwhile, phenomena such as automation and self-service methods in a variety of industries have made it increasingly suck to be less educated and/or lower-skilled over the time period in question. What seems to be happening, by and large, is that market labor supply among the less-educated is significantly lagging that of the better-educated, and the residual of market and non-market labor is leisure. I wonder how much of the observed labor/leisure phenomenon represents a smooth labor/leisure tradeoff, versus a more complex balancing of costs and benefits pertaining to trying to string together a workweek from multiple McJobs.
As for the well-to-do, the presence of social consumption externalities such as Robert Frank describes — à la, it's not good enough to spend a week at a lake for vacation, it has to be a week at a lake in Patagonia to generate an impressive-enough story for the next neighborhood potluck — may well make the labor/leisure picture a lot more complicated than the standard model suggests.
One of the more controversial claims in the paper is the suggestion that the faster growth of leisure among the less-educated may represent an offset to traditional measures of income and wealth inequality. Part of that would hinge on things like utility comparisons of sitting around watching TV versus upper-middle to upper-class extreme consumerist leisure, which are all but a lost cause empirically. It's also less than entirely clear what the decline in household labor means for the total economic consumption (i.e., of market and non-market goods) of households. Or, who's preparing the same meals more efficently within the household, versus substituting household production for measured market services.
Last, I could pick a nit with the authors' definitions of "leisure" — note that I said "measured leisure time" back in the lede. One of the data problems is consistency of measurement of child care; varying amounts of child care are shifted between household labor and leisure in the measures in part to address this. However, "recreational child care" (i.e., playing with one's children) is included even in the narrowest leisure definition. Now, I'll admit that given a choice between changing a diaper and building LEGO spaceships with John, I'll take the LEGO. But I don't call him Ming the Merciless for nothing. Kids' play and adults' leisure choices are not well aligned, and thus playing with one's children has a substantial element of work to it.
