Project Four: Gender, Time Allocation and the "Wage Gap"
Principal Investigator: Jonathan Gershuny (University
Researcher: Man-Yee Kan (University of Oxford)
This project will consist of a series of investigations organised around a key hypothesis: A substantial part of the gender gap in wages that persists, beyond the successful operation of workplace-based equal opportunities policy, is to be explained in terms of day-to-day practices of unequal division of responsibility for production and caring services within private households.
The hypothesis that a gendered division of domestic labour leads to a gendered wage gap relies on the assumption that women who specialise in non-waged work, as a result reduce their paid work hours and participation rates, leading to a reduction in their rate of human capital formation. This is a recursive process: initial differentials in human capital, and attitudes to gender-roles, may be the starting basis for bargaining over the distribution of paid and unpaid work roles within a heterosexual partnership. If one partner differentially specialises in unpaid work, her (rarely his) human capital declines relative to the more paid-work-specialised partner, leading to an intensification of work-role differentiation over time.
The degree and continuity of commitment to the labour
market is dependent also on public regulation. A given division of domestic
labour has varying potential effects on partners' paid work participation,
depending on various regime attributes. Normal weekly hours of paid work,
levels and costs of childcare provisions, temporal service accessibility,
parental leave rights, parental-leave-related employment protection, and
so on, all have major consequences for participation in paid work. Regime
provisions interact with the private household norms and circumstances
to determine the outcome of negotiations over work-roles.
The proposed analyses have three key elements:
1. Establishing associations between gender (im) balance in unpaid work-time and wage differentials. This is essentially cross sectional: considering couple households, controlling for educational attainment, current employment mix etc, to establish the extent of the negative relation between women's share of domestic work and shadow wage differential. The cross-sectional approach has various evident problems: processes of selection into couples, and within couples into joint employment statuses - disentangling requires longitudinal data.
2. Disentangling causal processes using longitudinal data. Two longitudinal data sources (BHPS and HoL) will be fused to produce well-grounded panel estimates of gender time use balances throughout the 1990s. The objective is to construct models of (i) selection into particular household and employment statuses, (ii) relationships between inter-temporal changes in (or constancy of) attitudes to parental responsibilities and change in these statuses, and (3) changes in (relative, shadow) wages of husbands and wives associated with employment and family status changes, controlling for (changes in) parenting attitudes.
3. Estimating regime effects on association between gender balances and wage differentials. The obvious question is: do national patterns of cross-sectional domestic divisions of labour show variation associated with regime provisions? Less obvious, and potentially more revealing, however, we shall look at "clock-time" evidence from the diaries (and the whole-week work schedules in the HETUS studies). How do couples' schedules through the day match with each-other and with the times of service provisions? Can the aggregate time use patterns and associations with wage differentials be traced to regime-related differences in patterns of task-sharing? So, for example, do countries with short work-hours and high levels of childcare provision have more two-job couples staggering their work times so that one partner delivers child to childcare and the other collects, while countries with longer work hours and scarce childcare, have more conventionally gendered arrangements?
The project will analyse a) the BHPS which has extensive
work history, wage and other materials, repeated measures of attitudes
to parental responsibilities, and stylised measures of domestic labour
distributions (task responsibilities and time use); and b) the Home-on-Line
(HoL) study (1999/2001) consists of 1000 households, three waves, whole
household 7-day ("light") diary collection, with some identical
stylised estimators and income measures as in the BHPS. To investigate
regime differences requires cross-national comparisons the project will
also use: The Harmonised European Time Use Study (HETUS), including more
than 20 large national random-sampled whole-household time-diary collections.
Professor Jonathan Gershuny
Department of Sociology
University of Oxford
+44 (0)1865 286175