https://orcid.org/0000-0002-7098-686X
![Sample our Global Development journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5929/TOC-Subject-Sample-2021-Global-Development.jpg"})
ABSTRACT
Son preference is known to influence fertility decisions, but very little is known about the prevalence of son preference in Turkey and its consequences for fertility behaviours. We use data from five waves of the Demographic and Health Survey and the Survey on Income and Living Conditions to show that son preference results in differential stopping behaviours and shorter birth spacing. Despite many public reforms promoting gender equality in Turkey, women and children are put at risk in the quest for a son. This result holds in all regions, in both urban and rural areas, and across the socio-economic spectrum.
RÉSUMÉ
L’influence que joue la préférence marquée pour les fils sur les décisions liées à la fertilité est connue, mais l’on en sait encore peu sur la prévalence de cette préférence en Turquie et les comportements de fécondité dans ce pays. Nous utilisons des données tirées de cinq vagues de l’Enquête démographique et de santé et de l’Enquête sur les revenus et les conditions de vie afin de montrer que la préférence pour les fils mène à des comportements de prévention de grossesse différentiels et à des grossesses plus rapprochées. En dépit de nombreuses réformes publiques encourageant l’égalité des genres en Turquie, les femmes et les enfants sont mis en danger afin d’assurer la naissance de fils. Ce résultat s’observe dans toutes les régions, en ville comme à la campagne, et à tous les paliers de l’échelle socio-économique.
Notes
1 Examples are the law supporting women’s entry into the public sphere, the restructuration of the legal system, and the implementation of policies to transform social traditions and values, which were inherently discriminatory before the 1923 constitution. Although Turkey has become one of the very few countries in the region with such provisions in its legal system, gender equality is not always reflected in practice due to a strong patriarchal culture and gender discriminatory traditions and values (Glpker-Kesebir Citation2016).
2 For example, while India is infamous for the high prevalence of sex-selective abortions, its large Muslim community has normal sex-ratio at birth (Bhat and Francis Citation2007).
3 Except in the case of life-threatening dangers in which case an abortion can take place further along in the pregnancy.
4 To give insight into the importance of breastfeeding, Victora et al. (Citation2016) estimate that near-universal breastfeeding would prevent 823,000 annual deaths among children less than five years old.
5 Thus, having a son is essential for the survival of one’s surname.
6 Still nowadays, only 34% of women aged 15 years or more participate in the labour force in Turkey compared to 52% among the OECD countries (World Bank Citation2019).
7 For a discussion of the recent turnaround of the government see Zillman (Citation2017).
8 For the long-lasting consequences of gender bias see, for example, Alesina, Giulianoa, and Nunn (Citation2013).
9 Specifically, the relevant questions in TDHS are: “If you could go back to the time you did not have any children and could choose exactly the number of children to have in your whole life, how many would that be?” and “How many of these children would you like to be boys, how many would you like to be girls and for how many would it not matter?”
10 Consequently, even women desiring an odd number of children can state a gender-equal preference.
11 Or that they, or their partner, are sterilized or infecund.
12 To ascertain the results are not driven by a specific time period, the model was re-estimated by decade at first birth (1970s, 1980s, 1990s and 2000s). We reach similar conclusions for each decade at first birth than for the pooled sample. Results available on request.
13 The only exception is for the first wave in which women aged 12–49 years old were interviewed.
14 It was not possible to match all households with the geographic location. Households for which no geographical location was available are clubbed together in a missing geographical data dummy. As a robustness check, these observations are excluded in an alternative specification. The main conclusions remain ( and , Column 3).
15 Only civil marriages are recognized in Turkey. Thus, someone who marries only religiously is likely to have a more traditional set of mind.
16 This is, however, not possible with SILC data as we only know the current gender composition of older siblings.
17 As SILC data is restricted to more recent years, and given the sharp fertility drop experienced by Turkey in the past decades, the SILC sample is restricted to children born at parity 5 or less.
18 When a household has only one previous child, the sex of the previous child and the sex composition of previous children are equivalent. Thus 2B (specification 1) and 2S (specification 2) are equivalent and so are 2G (specification 1) and 2D (specification 2). This is however not the case for more numerous families.
19 Remember, the hazard rate is the number of times a person will fail during one-time unit – here one month.
20 As we have no information on how the decision is made within the household, we simply look at the final decision. An interesting avenue for further research would be to shed light on the bargaining process between husband and wife, and potentially also their parents, when it comes to fertility decisions.
21 More details about the survey can be found at https://catalog.ihsn.org/index.php/catalog/4610.
22 Such a high prevalence of short birth spacing may be surprising given the ubiquity of breastfeeding in Turkey. Indeed, in TDHS 2013, only 2.5% of women claim not to have breastfed their most recent birth. However, to be effective as a contraceptive method, breastfeeding should be exclusive. The data available does not point towards exclusive breastfeeding as being the norm in Turkey. While we do not know for how long a child was exclusively breastfed, we know when the respondent’s menses returned after delivery. Limiting the sample to the last birth, to minimize recall bias, 50% of women claim to have had their menses return within 2 months, while nearly 75% claim to have had their menses return within 6 months.
23 The results for specification 2 are available on request.
Additional information
Notes on contributors
Marie-Claire Robitaille
Marie-Claire Robitaillee is an Associate Professor in Economics at Saint Mary’s University. Her research mostly focuses on son preference, children and education issues in developing countries.
Yigit Aydede
Yigit Aydede is a professor of economics at Saint Mary’s University. His main research interest is economics of population. He teaches machine learning, econometrics, and data analytics courses both at the undergraduate and graduate levels. His current research projects are related to machine learning applications on health care and chronic disease surveillance systems.