Like my parents at home? Gender differences in children’s housework in Germany and Spain

Prior research has documented the adverse impact of delayed cohabitation on a number of children’s outcomes, including individual motivations and ambitions, reservation wages, labor market entry and geographical mobility (Billari and Tabellini 2011).

Boeri and Van Ours (2008) show that, at the time of the surveys, the percentage of children under age 3 using formal childcare facilities is 34 % in the Germany and 5 % in Spain, indicating a significant difference in the availability of childcare services between the two countries.

Information on the variables, and on how to access the data, is available on the MTUS website: http://​www.​timeuse.​org/​mtus. See Fisher et al. (2011) for a full description of the MTUS documentation.

Separation has increased in Europe in recent decades, as shown by the increase in crude divorce rates in almost all countries (EUROSTAT 2015). The fact that we do not choose monoparental families eliminates around 25 % of the children. In the case of Germany, from 3272 children, we choose 2518 of these children, which represents a decrease of 23 % of the sample. In the case of Spain, from 11,330 children, we choose 8155 of these children, which represents a decrease of 28 % in the sample. Focusing on households, for Germany and Spain the samples are reduced by 25 % (from 2115 households to 1575 household) and 30 % (from 7500 households to 5244 households). However, we do not include monoparental households in our analysis, given that the share and the amount of housework done by each member can be different in these households.

The characteristics or preferences of adult children still living with their parents may differ from the characteristics of those individuals not living with their parents. We have compared individuals in the same age range living (i.e., being a child of the reference person of the household) and not living with their parents (i.e., not being a child of the reference person of the household) in the two countries. We observe that individuals not living with their parents participate in and devote comparatively more time to housework activities, in some cases with a threefold difference (e.g., women in Spain). Furthermore, and in comparison with individuals living with their parents, individuals not living with their parents are older, have higher labor force participation and live in larger households. These differences may indicate that those living with their parents are able to enjoy a “better living.” This evidence indicates that those children staying at home may want to benefit from their parents’ input, having different preferences than those who decide to leave home. Thus, sample selection may be biasing our results, and the reported correlations can be seen as a lower bound, as we could expect a higher relationship once children leave home.

The analysis of childcare time may be interesting, since it is one of the larger times that teenagers spend when they have younger siblings. Looking at the time devoted to childcare in the current data, 96.83 % of the sample report no time in childcare activities, and given this low variation in childcare time we cannot analyze this activity. To the extent that children may not be doing childcare on a daily basis, it may be that the data at hand is not the best dataset to analyze this issue. We leave this analysis for future research.

The MTUS has been widely used across the social sciences (Gershuny 2000, 2009; Gershuny and Sullivan 2003; Gauthier et al. 2004; Guryan et al. 2008; Gimenez-Nadal and Sevilla 2011, 2012; Giménez-Nadal and Molina 2013).

Given the significant rate of non-participation in housework time of children, if the number of non-participants differs greatly across countries or gender, tobit models are helpful as they take into account the distribution of non-participants. Thus, we have alternatively estimated tobit models on the (log of) time devoted to housework by children. Results obtained from tobit models are the same as the results obtained from OLS models, and given that OLS coefficients can be directly interpreted we follow OLS results. Results for tobit models can be found in Table 6 of online “Appendix 1.”

See online “Appendix 2” for a description of the distribution of housework time of parents and their children.

To the extent that there is positive assortative matching by education (Oppenheimer 1988; Mare 1991; Pencavel 1998; Lewis and Oppenheimer 2000; Blossfeld and Timm 2003), we consider that parents have similar levels of education, and we thus exclude from the analysis those couples where the members of the couple have different levels of education.