‘Missing girls’ in interwar Poland: child sex ratios and their correlates across multiple borderlands

ABSTRACT

Our testing of the relationship between child sex ratios (CSRs) and demographic, socioeconomic, and cultural diversity across nearly 300 districts of interwar Poland around 1931 yields a picture more complicated than common explanations of high masculinization of offspring. In line with existing literature, we found district-level CSRs to be positively associated with the extent of agriculture but negatively related to the relative spread of female employment outside farming and less hierarchical gender and generational household arrangements. At the same time, several of the classical modernization variables (e.g. industrialization, urbanization, female literacy or fertility) either did not result in lower sex ratios or turned out irrelevant. In this article, we attempt to reconcile these diverging results by putting them in the context of the country’s relative backwardness, the specific labor demands created by modernization, and the structure of the agricultural labor market. Altogether, our results add a new stimulus to study gender discrimination in infancy and childhood in East-Central European context and to contemplate universal explanations thereof.

KEYWORDS:

• Child sex ratios
• missing girls
• gender mortality differentials
• interwar Poland
• modernization

Acknowledgments

We would like to thank Francisco J. Beltrán Tapia and Francisco J. Marco-Gracia for the comments on the early version of the paper as well as the two anonymous Reviewers and the Editors for their suggestions, which were very helpful for refining the original manuscript. Michael Timberlake and Piotr Puchalski are acknowledged for language editing. Financial support from the National Science Centre, Poland is also gratefully acknowledged.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Notes

1. Importantly, a level of 102 boys per 100 girls below 5 years of age is to be expected in the context of non-discrimination towards girls and high infant mortality.

2. Polish studies of historical CSRs have so far been carried out without explicit reference to major debates surrounding the ‘missing girls’ topic, and mainly as tests for data quality (e.g., Holzer, 1960; Kuklo, 2009; Szołtysek, 2015, v. 2). Recently, Liczbińska and Sobkowiak (2020) analyzed sex ratios at birth in the Poznań province in the 19th and early 20th centuries from the point of view of social and economic inequalities, but they did not pay attention to the issue of potential female infanticide or mortal neglect.

3. This notion pertained to some extent also to families of nobility (Kostrzewska, 2021).

4. E.g. IGS (1935, pp. 36–38), Chałasiński (1969, p. 386). These examples show the anger of a husband towards his wife for the fact that she had given birth to a daughter instead of a son, as shown by this notable excerpt from a peasant memoir: ‘The daughter evokes subconscious aversion from the day she makes first sound’. Other instances of cultural misogyny could be the Polessyan proverb ‘kozy nie tawar, a baby nie ludzie’ (goats are not cattle, women are not human); the Polish label ‘dziurawe wojsko’ (leaky army), which was used to describe women; or the Jewish morning prayer of men: ‘Blessed are you, Lord, our God, ruler of the universe, who has not created me a woman’ (Engelking, 2007). Similar, prejudicial meaning could be attributed to the custom practiced in rural areas of Małopolska, where during the first bath of a newborn, the father tossed some coins into the water. The sum thrown into boy’s bath was twice as large as in the case of a girl (Trawińska-Kwaśniewska 1957).

5. We assume that the relation between ASDR 1–4 and IMR in the districts of the eastern province is equal to that observed in former Galicia, a region of similar economic development but with much better registration (Fogelson, 1936).

6. According to Szulc (1936), Jewish population tended to delay the formal registration of newborn children (and particularly girls) until ‘administrative, educational, familial or economic’ necessity arose, even up to seven years after birth.

7. This variable could be sensitive to the fact that the census of 1931 used the inherited legal status as the indicator of being urban. In this regard, the difference between the former Russian partition, on the one hand, and the German and Austrian partitions, on the other, was reflected in the census instructions (GUS, 1932) and regulations of self-government (Kotulski, 2019). To account for this problem, we use the partition dummy variables (see below).

8. An annuitant (dożywotnik) is understood here (and in the census of 1931) as a person who obtains annuity or maintenance in some other form from the coresident kin owner of the farm (and presumably the household head) in exchange for relinquishing proprietorship of the farm (GUS, 1937–38). This then reflects Ausgedinge, a form of generational succession or inter-generational intervivos transfer that was once widely practiced in preindustrial Europe (see, Held, 1982). According to Gruber and Szołtysek (2016, p. 140), in intensely patriarchal areas, no young man was permitted to become a household head as long as an older male household member was alive due to the strict seniority principle that regulated power relations within domestic groups.

9. Note that the diffusion of innovations and work opportunities that this variable pertains to are also accounted for more directly by other measures considered below.

10. These parts are Galicia (former Austrian partition), Kingdom of Poland (partially autonomous state within the Russian Empire), ‘taken lands’ (territories incorporated directly into the Russian Empire), former German partition (Upper Silesia, Greater Poland, Pomerania).

11. For an absolute majority of the fourteen late 18^th and early nineteenth-century regions of Poland-Lithuania analysed by Szoltysek, it has been found that the observed variation in CSRs can be attributed primarily to random noise associated with a small sample size (i.e., the population of children under age five from which the child sex ratios are derived) as well as structural explanations related to infant mortality differentials and census quality.

12. There is positive and significant correlation between the share of croplands and the percentage of farms smaller than 2 hectares in the region (R = 0.25***).

13. Mędrzecki quotes a peasant woman author who states that at the beginning of the 20 c. ‘[…] girl from the youngest age is useful at home: she plucks feathers, peels potatoes, spins, and babysits the siblings. By sending her to school, her mother loses a helper, and for this reason – even if she decides to send the daughter to school – her education does not last long’ (Various authors, 1904).

Additional information

Funding

This work was supported by the National Science Centre, Poland, under Grant The Fertility Decline in the Interwar Poland, No. 2019/35/D/HS3/00721.