Trends on the Marriage Market

Figure 4.2 plots the estimated marriage surplus for different nationality combinations of spouses over time. We focus on marriages where at least one partner is German. To increase visibility, we separately show the surplus for marriages in which both partners are German (black line) one spouse is an EU15 citizen (blue line), one spouse is an EU10 citizen (orange line). We pool all other nationalities in the “Rest” (of the world) category (gray line).

From 1998 until the announcement date of the labor market reform (black dashed line), we observe a slow but steady decline in marriage surplus for all combinations of spouses. Afterwards, trends diverge. We observe a steep decline in the estimated surplus for marriages in which one partner is of EU10 or “Rest of World” type.¹³⁰ According to our hypothesis, these marriages were disproportionately negatively affected by the reform due to higher unemployment risk. Conversely, the marriage surplus for German couples and between Germans and EU15 citizens, arguably the groups with lower unemployment risk, increases slightly.

It is challenging to isolate the marriage market impact of the labor market reform in January 2003 due to the Eastern enlargement of the EU. Ten new member states joined in May 2004 (red dashed line in Figure 4.2) and this was anticipated. For citizens of these countries, marrying a German was no longer necessary to obtain access to the German labor market following the enlargement.¹³¹ Thus, the EU expansion could have had an impact on the marriage market in itself, as argued by Adda et al. (2019) in the Italian context. To take this confounding factor into account, we control for the EU expansion

129This is only an approximation since the MC data is survey-based and individuals could potentially get married in the same year after reporting in the survey that they are singles. The numbers of individuals might thus be somewhat upward biased due to double counting.

130By far the largest group of EU10 immigrants are Poles. Figure C.1 in the Appendix separates the surplus of German-Polish marriages and German-Other-EU10 marriages, The trend is broadly the same, the decrease somewhat less steep for Poles.

131Although access to the German labor market for migrants from the new EU member states was initially restricted.

Figure 4.2: Development of Marital Surplus over Time

2000 2005 2010

−8−7−6−5−4−3−2

Time Trend: Marriage Surplus

Year

Marriage Surplus

Hartz I Reform EU Expansion

DE−DE DE−EU15 DE−EU10 DE−Rest

Notes: Marriage surplus for marriages where at least one spouse is German by nationality of the non-German spouse.

The black dashed vertical line indicated the year in which the labor market reform became effective, the red dashed vertical line marks the year 2004 in which the EU expansion took place.Data: RDC of the Federal Statistical Office and Statistical Offices of the Federal States, Marriage Registry, survey year(s) 1997-2013, own calculations.

in our main analysis by using separate treatment and control groups to isolate its effect in our differences-in-differences setup.

In Figure 4.2, the trends described above did not change around the date of the EU expansion. On the contrary, the divergence between German/German-EU15 and German-EU10/German-Rest marriages became more pronounced. The surplus of mar-riages between Germans and between Germans and EU15 citizens continued to increase before flattening out in 2005. EU10 citizens were directly affected by the EU expansion and, accordingly, their marriage surplus with Germans decreased at a steeper rate as compared to German-Rest marriages, especially after 2007. This makes intuitive sense:

arguably, the surplus from marrying a German for EU10 citizens reflected to some extent

Figure 4.3: Expected Gains to Entering the Marriage Market

2000 2005 2010

0.000.050.100.150.200.250.30

(I) Men

Year

Expected Gain to Entering the Marriage Market

Hartz I Reform EU Expansion

DE EU15EU10 Rest

2000 2005 2010

0.000.050.100.150.200.250.30

(II) Women

Year

Expected Gain to Entering the Marriage Market

Hartz I Reform EU Expansion

DE EU15EU10 Rest

Notes: Expected gains to entering the marriage market for men (left sub-figure) and women (right sub-figure).

the obtainment of a residence permit and labor market access before 2004. Due the EU expansion, this part of the marital surplus was erased.

The declining marriage surplus for interethnic marriages with non-EU15 spouses is also reflected in the estimated expected gains to entering the marriage market, see Fig-ure 4.3, which plots these gains for men (left sub-figFig-ure) and women (right sub-figFig-ure).

Recall that, according to Equation 4.11, the estimated gains of marriage only depend on the total number of individuals relative to singles of any particular type. As before, they are shown for German men/women in black, for EU15 men/women in blue, for EU10 men/women in orange and gray for all remaining nationalities.

Notably, the gains to marriage decrease more substantially for foreign women than for foreign men and this is mainly driven by EU10 women, who had very high gains in the beginning of our sample that rapidly decreased thereafter. This is in line with the fact that we observe relatively many women from Poland and other EU10 countries who get married in Germany as compared to men from these countries, recall Table 4.1. For EU10 men the gains are essentially flat before the labor market reform, decrease in 2003/2004, and reach another relatively stable level. Rest-of-the-world men’s and women’s gains are very similar. They start decreasing before the reform, but the negative trend accelerates in 2003–2005.

Table 4.4: Treatment and Control Groups

Type of “Hartz” Treatment EU Treatment “Hartz”-Effect Controls

Marriage (Jan 01, 2003) (May 01, 2004) (EU-Effect)

German-German No No Control —

German-EU15 No No Control —

German-EU10 Yes Yes Treatment Treatment

German-nonEU Yes No Treatment Control