VoxEU Column Labour Markets

How prices, incomes, and discrimination affect the ways we use time

People combine goods and time in household production, and theory suggests that as their wage rates rise, they will substitute goods-intensive for time-intensive activities. However, it is not clear how activities that take essentially no, or minimal, amounts of spending, such as sleeping or watching TV, fit into the theory. This column uses data from time diaries for the US, France, and Germany to demonstrate that not all non-work time is the same, and different components of non-work time respond differently to changing incentives.

The ways that people use their time have changed over the last century in wealthy countries – less paid work, less time spent in household chores, more time in leisure. The formal economic theory of time use for analysing these and other changes goes back to an article by Gary Becker published in 1965. The crucial insight was that people don’t simply consume goods (spend their incomes) and use their non-work time – they combine goods and time in household production. This insight leads to the conclusion that activities that take little time and lots of money will be disproportionately undertaken by high-wage people, while those activities that take a lot of time, such as sleep (Biddle and Hamermesh 1990), will be neglected by such people. The assumption underlying the theory is that people will be induced to substitute goods- for time-intensive activities as their earnings ability – i.e. their wage rates – rises. But what about activities that take essentially no, or minimal, amounts of spending?

To examine this and other questions about time use, we use data from time diaries for the US from 2003–2015, for France from 2009–2010, and for Germany 2012–2013 (Hamermesh and Biddle 2018). In a time diary, the respondent is asked to complete on the next morning what he or she was doing at each moment of the previous day. Americans complete one day’s diary, the French two days’, and the Germans three days’ diaries. Most of these diaries provide information on detailed categories of time use. These can be aggregated into six useful categories: paid work, home production (cooking, cleaning, etc.), sleeping, personal activities (sex, washing up, etc.), watching television, and other leisure. With the large samples from each country, along with detailed demographic and economic information on each diarist, we can get quite precise estimates of how differences in incomes and wages alter the ways that people spend time.

For most adults, decisions about how much to work are affected by the returns to work – their wage rates – and their household’s other income; they must choose among the six types of activity. Some adults, however, choose not to work. They allocate their time among the other five activities, with their decisions about spending time determined only by their preferences, the nature of the goods they might buy, and their household’s other income (their partner’s earnings, any inheritance, pension/old-age support, etc.). The theory predicts that every activity that takes the same amount of money per minute will respond the same way to changes in income. Sleep, which costs almost no dollars per hour spent, and TV-watching, also extremely low-cost, would respond the same way. 

Figure 1 Income elasticity of demand for different uses of time: Non-workers in the US, France, and Germany

The figure shows the income elasticities of all five activity groups among those who do no paid work. Clearly, people’s sleep and television-watching to not respond to the same extent when their incomes rise. They do sleep less and watch less TV when income rises, but the declines in time spent watch TV are much larger per dollar, or euro, increase in income than are their cutbacks in sleep time, with income elasticities over five times as large. Not all time-intensive activities elicit the same response to changing economic incentives.

How can this be, since the theory suggests that the effect of differences in incomes on people’s choices of consumption between these two activities should be the same because they both take essentially no money? The answer is that, in addition to how people choose how much of different activities to consume, they must also choose how to combine the time and goods that generate the activities – they must choose a method of home production. The ease with which goods and time can be substituted in home production isn’t the same for all activities – and not for sleep and watching TV. Sleep is more biologically determined than TV-watching so that changes in incentives don’t alter it very much. With rising incomes, it is easier to switch to spending less time watching TV but to more expensive television (streaming series and watching them on a 140cm screen rather than watching network television on a small screen).

With 24 hours in a day, if sleep and TV-watching decline as incomes rise, some other activities must see increases in the time devoted to them. The most consistent increases are seen in time spent in other leisure activities (cinema-going, concerts, exercising, and others). In all three countries, time spent in other leisure activities increases as people’s incomes rise. The income elasticities are not large, but in all cases are positive. The income elasticities of home production and personal activities are mixed – positive in at least one country, negative in at least one.

The theory underlying these results assumes that people pay identical prices for each good that they purchase to combine with their time. We know, though, that racial/ethnic discrimination in pricing of some products means that minority groups often pay more for the same product than do members of the majority. This has been demonstrated numerous times in the US for auto purchases, home loans, and home rentals (e.g. Ayres and Siegelman 1995). Evidence for France and Greece suggests that immigrants, the minorities of those countries, also face price discrimination in their purchases (Acolin et al. 2016). With minorities of the same total income as whites facing higher product prices, the amount of goods that they can purchase is smaller. When combining goods and time, they will choose to spend more of their time doing things that require fewer goods per hour of time – those activities that are more time-intensive.

The data for sleeping and TV-watching for African Americans and for French immigrants bear this prediction out: for both activities, and in both countries, if a minority has the same income and demographic characteristics as a majority-group member, he or she sleeps more and watches more TV. The effects are not tiny: an African-America watches about 15% more television than an observably identical majority-group member. The difficulty is that this difference is much bigger than differences in prices, and thus in effective real incomes, that are produced by racial/ethnic discrimination would imply. At least half of the extra sleep and TV-watching of minorities must arise from something other than discrimination in product pricing. Perhaps fear of being hassled leads minorities to engage in such activities as sleeping or watching TV that is done at home, or perhaps there are simply racial/ethnic differences in preferences for engaging in these activities.

Economists have long distinguished between work and leisure; that’s a fundamental dichotomy in both macroeconomics and labour economics. As this discussion shows, and as discussed at length in Hamermesh (2019), that distinction is far too broad. Not all non-work time is the same, and different components of non-work time (lumped together and called leisure in common economic parlance) respond differently to changing incentives – to changing incomes and changes in wage rates. This conclusion shouldn’t be surprising. Ask yourself, if your income rises, would you spend the same extra time going to movies as you would changing diapers?


Acolin, A, R Bostic and G Painter (2016), “A field study of rental market discrimination across origins in France”, Journal of Urban Economics 95: 49-63.

Becker, G (1965), “A theory of the allocation of time”, Economic Journal 75: 493-517.

Biddle, J, and D Hamermesh (1990), “Sleep and the allocation of time”, Journal of Political Economy 98: 922-43.

Hamermesh, D, and J Biddle (2018), “Taking time use seriously: Income, wages and price discrimination”, NBER Working Paper 25308.

Ayres, I, and P Siegelman (1995), “Race and gender discrimination in bargaining for a new car”, American Economic Review 85: 304-21.

Hamermesh, D (2019), Spending time: The most valuable resource, New York: Oxford University Press.

1,050 Reads