For better or for worse: the long-term effects of postwar mobilization on family formation

ABSTRACT

This article estimates the long-term legacies of female labour force mobilization on women’s family formation outcomes such as marriage, age at first marriage and divorce. We identify the long-term marriage effects of female labour force mobilization by exploring postwar mandatory employment in Germany. Using difference-in-differences analysis, we find that participation in postwar reconstruction efforts increased women’s probability of being currently married, ever married and marrying at younger ages. We also find that postwar employment had no differential effect on divorce rates of the affected cohorts of women. These results persist after accounting for the potential changes in the composition of the population, demand for female labour, war relief payments and postwar state-specific policies.

KEYWORDS:

• Labour market mobilization
• marriage
• divorce

JEL CLASSIFICATION:

• J12
• J16
• J21
• J48

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ See Blattman and Miguel (2010) for a review of the literature.

² This is applied to both men and women. However, the application of this law to women is the novelty here, which mandated women to participate in the labour force for the first time. Individuals who failed to register would lose the right to receive food ration cards, and employers who did not comply risked imprisonment, fines and criminal prosecution (Meiners 2011). This law was repealed in West Germany in 1955.

³ Related to this, in their detailed study of employee records from the Ford Motor Company, Kossoudj Sherrie and Dresser (1992) analyse the firm-wide occupation distribution to understand the details of the disappearance of women from the US industrial workforce after WWII.

⁴ Allied Forces released prisoners of war gradually between 1946 and the early 1950s (Meiners 2011).

⁵ See Akbulut-Yuksel (2014) for detailed information on WWII destruction.

⁶ An individual not assigned to rubble removal and reconstruction received a level five food ration card, which was known as the ‘hunger card’ or ‘ascension pass’ because it provided minimal food (Heineman 1996).

⁷ This article provides evidence on the impact of postwar mandatory employment using city-by-cohort variation in female labour force mobilization rates within Germany; therefore this approach may yield lower bound estimates for the aggregate nation-wide effects.

⁸ In 1946, German marriage law permitted females to marry if they were at least 16 years old.

⁹ Results remain statistically similar when the rubble women dummy is replaced with a continuous measure of length of exposure. Table 7 provides analyses with the length of exposure variable.

¹⁰ Cameron and Trivedi (2005) suggest that in difference-in-differences analysis if there remains observable differences in the distribution of the characteristics between the treatment and control groups, the standard solution is to include such control variables in the regression. Thus, following Cameron and Trivedi, we account for other potential confounding factors and variables in our article that could have impacted the affected and the control cohorts differentially.

¹¹ We also account for the possibility of selective mortality in a robustness check, results are reported in appendix Table A1.

¹² However, including women born between 1935 and 1939 to the affected cohorts yields quantitatively and statistically similar results, as reported in appendix Table A2.

¹³ The wartime destruction measure varies at the ‘Raumordnungsregionen’ level (ROR or city). RORs are ‘spatial districts’ determined by the Federal Office for Building and Regional Planning based on economic interlinkages and commuting flows of areas. The area of former West Germany had 38 different RORs in 1978.

¹⁴ Research Data Center (RDC) (1978), authors’ own calculations.

¹⁵ When we include women who were students in our sample, results remain very similar, since the number of women who were still students then is very low.

¹⁶ Hochstadt (1999) also states that shorter distance moves and migration within provinces followed exactly the same pattern, and that this period was perhaps the longest span of stable migration rates in the past 200 years of German history. Until 1980, net population changes due to migration for all cities over 20 000 approaches 0.

¹⁷ The SOEP provides information on the respondents’ RORs starting from 1985; therefore we use the 1985 wave for the internal migration analysis.

¹⁸ The German displaced persons law passed in 1953 defines refugees as individuals who migrated from the Soviet Zone/GDR (Fluechtlingsausweis C, refugee card C). This law also applied to the individuals born in refugee and displaced households after the displacement (Luettinger 1986, 21–22).

¹⁹ We also estimate the effects of the mandatory employment law on rubble women’s fertility rates. However, these analyses warrant caution since the Microcensus only provides information on the number of children still residing within the same household as their mothers. Most children of the rubble women may have already moved out by this time as these women were married earlier and probably had children at younger ages; therefore it is likely that we focus on a selective group in our fertility analysis, which renders statistical inference difficult.

²⁰ This variable takes a value of 9 for women born between 1920 and 1931, since the mandatory employment law affected these women during the entire time it was implemented. It takes a value of 8 for women born in 1932, a value of 7 for women born in 1933, a value of 6 for women born in 1934 and finally a value of 0 for women born between 1940 and 1954.