The Prevelance of Induced Abortion in Turkey and the Factors Affecting Its Decision
Tuğba Adalı, Alanur ÇavlinPregnancy outcomes other than live births not only attract the attention of health workers due to potential health risks but also of social scientists as these outcomes can provide clues regarding social injustice, gender inequality, and women’s fundamental human rights. Compared to previous surveys, the results from the 2013 Turkey Demographic and Health Survey indicate a change in abortion levels. Accordingly, the proportion of pregnancies ending in abortion had dropped from 10% in 2008 to 5% in 2013. This study examines this change and the factors affecting abortion. The data mainly come from the 2013 Turkey Demographic and Health Survey. The changes in the prevalence of abortion between 2008 and 2013 have been investigated using selected characteristics of women. The factors affecting abortion between 2008 and 2013 have been modeled using logistic regression analyses and have been addressed in two groups, namely demographic and social-economic/cultural determinants. The decline in abortions from 2008 to 2013 is statistically significant. Considering the changes in abortions services provided in public hospitals with the impact of recent political discourses, this finding is expected. Analysis shows that the odds of having an abortion are affected by demographic factors more than socio-economic ones. Pregnancies to women who have ever used a modern contraceptive, which were due to contraceptive failure, and which took place after a son was already born are more likely to end in abortion. Moreover, religiosity is a significant determinant of abortion.
Türkiye’de Kürtajın Yaygınlığı ve Kürtaj Kararını Etkileyen Faktörler
Tuğba Adalı, Alanur ÇavlinDoğumla sonuçlanmayan gebelikler, sadece sebebiyet verebileceği sağlık riskleri açısından sağlık çalışanlarının değil, aynı zamanda sosyal adaletsizlik, toplumsal cinsiyet eşitsizliği ve kadınların temel insan hakları açısından önemli veriler sunabileceğinden sosyal bilimler alanındaki araştırmacıların da ilgisini çekmiştir. 2013 Türkiye Nüfus ve Sağlık Araştırması’nın sonuçları, kürtaj düzeyinde önceki araştırma sonuçlarına göre önemli bir farklılığa işaret etmektedir. Araştırmaya göre son 2008-2013 yılları arasında kürtaj oranı ise %10’dan %5’e düşmüştür. Bu çalışmada, kürtaj düzeylerindeki değişim ve kürtaj risklerini etkileyen faktörler incelenmiştir. Çalışmada temel olarak 2013 Türkiye Nüfus ve Sağlık Araştırması verileri kullanılmıştır. Analizlerde 2008-2013 yılları arasındaki kürtaj yaygınlığındaki değişim kadınların seçilmiş özellikleri bazında incelenmiştir. Ardından 2008-2013 döneminde gerçekleşen kürtajları etkileyen faktörler lojistik regresyon analizi ile modellenmiştir. Bu modellerde riski etkileyen faktörler demografik ile sosyo-ekonomik ve kültürel belirleyiciler olarak iki grupta incelenmektedir. Kürtaj yaygınlığında 2008-2013 dönemi arasında görülen düşüş istatistiksel olarak anlamlıdır. Son yıllarda siyasi söylemin etkisiyle kamu hastanelerindeki kürtaj hizmetinin kısıtlanmış olması dikkate alındığında, bu beklenen bir bulgudur. Çalışmadaki lojistik regresyon analizlerine göre, gebeliklerin kürtaj ile sonuçlanması sosyo-ekonomik özelliklerden çok demografik özellikler ile belirlenmektedir. Modern yöntem kullanmış; gebeliği yöntem başarısızlığıyla oluşmuş; veya halihazırda erkek çocuğu olan kadınların gebeliklerinin kürtajla sonlanma riski daha yüksek bulunmuştur. Ayrıca dindarlığın da kürtajın önemli bir belirleyicisi olduğu görülmüştür.
Apart from live births, induced abortion (hereinafter referred to as abortion) is the most widely studied topic among pregnancy terminations from a multidisciplinary perspective. It is a main topic in demography, is affected deeply by micro-level fertility decisions, and is a topic of population policies that affect this demographic event at a macro level.
The 1994 ICPD Conference in Cairo recognized that women should have access to reproductive health services that would also allow them to control their fertility. The provision of safe abortion services during the first 10-12 weeks of a pregnancy is recognized as part of this access. Eliminating unwanted pregnancies and maternal deaths due to unsafe abortion were parts of the Millennium Development Goals and currently part of Sustainable Development Goals (Sedgh, Singh, & Hussain, 2014; World Health Organization [WHO], 2015).
Globally speaking, abortion levels have been going down in developed countries since the 1990s (down from 46 abortions per 1,000 women to 27 per 1,000); however, this does not necessarily hold for developing countries (fluctuating between 39 to 36 per 1,000 women). In Turkey, the trend has been a downward one since the mid1980s, an era marked by the enactment of a population law that legalized abortion on demand up to the tenth week of pregnancy. An increase in the percentage of pregnancies ending in abortion was observed in demographic surveys following this law, but a consistent trend has been observed since (Table 1). Since the beginning of the 1980s, the use of modern contraceptives has been increasing, and total fertility rate has been decreasing; both of which are in line with the abortion trend.
On the other hand, the drop in the level of abortions from 2008 to 2013 was somehow sharper than expected, where the number of induced abortions per 100 pregnancies dropped by half from 10 to 5. The confidence intervals of these estimates are non-overlapping (Adalı, Çavlin, & Berktaş, 2015). A closer look into these two years reveals that contraception use and total fertility rate remained at similar levels. Thus, this study aims to investigate whether the decline can be attributed to a specific group of women and to find out which pregnancies end in abortion based on data from 2013.
Data and Method
The study mainly uses data from the 2013 Turkey Demographic and Health Survey, a large-scale household survey that interviewed 11,794 households and 9,746 women. The survey is based on a multistage, stratified cluster sample. Analysis has been done using the Complex Samples module of SPSS 23, which allows for these design variables and sample weights to be taken into account. All the interviewed women were asked detailed information about their reproductive history over the five years preceding the survey date using a calendar module. Monthly information on the use of contraceptives, discontinuation of contraception and the reasons for it, pregnancies, and pregnancy outcomes were recorded in this module. The data used in this study is based on this calendar data, where each case is an event defined by a terminated pregnancy (through a live birth, an abortion, a miscarriage, or a stillbirth). To study the determinants of abortion, an indicator variable that takes the value of 1 if pregnancy ended in an abortion has been defined as the dependent variable.
Theoretical Framework
The determinants of abortion have been categorized under two main headings in this study: demographic determinants and socio-economic/cultural determinants. The variables under these two headings have been selected according to the literature. Among the demographic determinants are: the woman’s age at time of abortion, whether the current pregnancy will cause her to exceed her ideal number of children, whether she has had a son, whether she has previously had an abortion, whether she has ever used modern contraceptives, and whether the current pregnancy was a result of contraceptive failure. The selected socio-economic and cultural determinants are: type of place of residence as (i.e., urban or rural), region, education level, household wealth index, woman’s employment status, mother tongue, and religiosity (as defined by certain practices).
Results
A comparison of two surveys from 2008 and 2013 reveals that abortion declined in all subgroups of women, regardless of age, socio-economic status, or parity. Looking at the magnitude of the decline, the decline appears more rapid for more educated women and women in wealthier households. The logistic regression analysis has been carried out in two steps, with only demographic determinants as the first step and with all variables combined in the second step. The significant determinants in the first model are age (higher odds of abortion for higher ages), previous abortion experience (higher odds if yes), modern contraceptive use (higher odds if yes), and contraceptive failure as cause of pregnancy (higher odds when pregnancy unplanned). When demographic variables were controlled for, no socio-economic variables proved significant. Only religiosity is signficant among this block of variables, with the pregnancies to women who are less religious having higher odds of being terminated through an abortion. The addition of this block of variables also changed the significance of one demographic variable that compares the odds of having an abortion for women with at least one son, and women with daughters only against having no children; and the odds of abortion significantly increase for the foremost category.
Conclusion and Discussion
The findings imply that the rapid decline in recent years could not be attributed to a specific socio-economic group. The regression analysis shows that such determinants are not significant when controlling for demographic variables. The fact that contraception use has not changed remarkably between 2008 and 2013 leads to questioning the supply side of abortion services. In the last decade, Turkey has witnessed a change in population policies where government officials have made public anti-abortion statements. Despite no changes to the 1983 law, multiple studies by NGOs and universities have shown signs of abortion services vanishing from public hospitals (Bianet, 2015; Mor Çatı Kadın Sığınağı Vakfı, 2015; O’Neil, 2017; Topgül, Adalı, Çavlin, & Dayan, 2017). The decline in abortions in Turkey might be attributed to this situation, which would call for measures to ensure women regain their right to reproductive health services.
