Women’s bargaining power and contraception use in post-Soviet Tajikistan

ABSTRACT

This article aims to examine the relationship between women’s household bargaining power and their adoption of modern contraception in post-Soviet Tajikistan using the 2012 Demographic and Health Survey. The study uses direct measures of bargaining weights: a woman’s ability to make decisions about her own health care; visits to her family or relatives; and contraceptive use. An additional measure defining a woman’s financial capability to receive medical treatment for herself is added in the analysis to understand its correlation to women’s contraceptive-use behaviour. The probability of using contraception is 187 percentage points higher for a woman who has both control over her own health care and financial means to get medical help than a woman who does not have these choices. Having a say in the decision to control births increases the probability of using contraception by 98 percentage points. Our findings reveal that certain aspects of a woman’s household decision-making and financial freedom are relevant to explain her contraceptive-use behaviour.

KEYWORDS:

• Tajikistan
• decision-making power
• contraception
• family planning
• fertility

Introduction

It is estimated that meeting the unmet need for family planning of women in developing countries could prevent 54 million unintended pregnancies, 79,000 maternal deaths and 1.1 million infant deaths (Singh and Darroch 2012). Access to family planning services is essential to securing the well-being and autonomy of women. A study from long-term Matlab experiments in Bangladesh and Ghana showed that women’s health, earnings and assets and children’s schooling and well-being substantially improved in areas with improved access to family planning services (Canning and Schultz 2012).

Despite global efforts to increase awareness and improve access to family planning services, unplanned pregnancies remain high in low- and middle- income countries. High fertility rates raise the physical costs of childbearing and childrearing for women. A woman exposed to four or more pregnancies faces greater risks of morbidity and mortality than a woman experiencing only two or three pregnancies (Mason and Taj 1987). The unmet need for contraceptives and excess fertility are generally explained by the costs of methods, distance to healthcare providers, low autonomy, and a lack of schooling and contraceptive knowledge (Bulatao 1998; Skirbekk 2008).

A number of factors have been associated with contraceptive-related decisions. One of these factors is a woman’s decision-making ability within the household.¹ Studies suggest that women who actively participate in domestic decision-making are more able to control their fertility through the adoption of modern contraception than women who do not participate in domestic decisions (Hameed et al. 2014; Hogan, Berhanu, and Hailemariam 1999; OlaOlorun and Hindin 2014; Woldemicael 2009). In this article, we examine which determinants of decision-making power are relevant to explain women’s use of modern contraception in post-Soviet Tajikistan.

We use the 2012 Tajikistan Demographic and Health Survey (TjDHS [Republic of Tajikistan 2013]) to understand the relationship between women’s power position and their contraceptive behaviour. A woman’s decision about her health care, visits to her family or relatives, and contraception use, and her financial capability to receive medical treatment for herself, are used as proxy measures for bargaining power. We also examine if financial means to receive medical treatment is associated with different contraceptive-use behaviour for women who have autonomy over their health care. Variables such as financial capability and healthcare decisions are used as interaction terms to define the relationship with women's contraceptive choice. Further, we control for other indicators of power, such as education, employment, household structure and land ownership. We hypothesize that these factors will be significant in understanding women’s ability to access and use family planning in Tajikistan.

This study adds to the literature on factors determining decision-making authority and its relationship with a woman’s ability to control fertility through access to modern contraception in post-Soviet Tajikistan. It also provides insight into the Tajik government’s policy intervention programmes on family planning and suggests possible ways to improve reproductive health care in similar countries.

Empirical evidence on women’s decision-making power and family planning services in Tajikistan is quite limited. Tajikistan provides a unique case to assess women’s decision-making power and the government’s role in promoting reproductive health-care services. The prevailing traditional cultural restrictions on women often position males as the household decision makers. The government has taken several legal measures to inform the population of ways to control fertility and increase their access to and use of modern methods of contraception. The conceptual framework that informs the analysis of decision-making power on women's contraception behaviour is influenced by the personal experience of one of the authors, who was brought up in Tajikistan, as well as by her commitment to promoting gender equality and female empowerment in the country.

We find that contraceptive use is lower among women who have autonomy in the decision-making process regarding their own health care and visits to family or relatives. However, interacting a woman’s healthcare decisions with her financial ability to access medical treatment for herself provides an interesting result, showing that the probability of using modern contraception is 187 percentage points higher for a woman who has both the ability to make health-care decisions and the financial means to receive medical services compared to a woman who does not have both or either of these choices. This suggests that financial independence enhances a woman’s healthcare decision-making power and assists her in accessing family planning services. Further, the likelihood of using modern contraceptives is more than 90 percentage points higher for a woman who makes contraceptive decisions alone or in communication with her husband, compared to a woman who does not participate in family planning decisions. Overall, findings of this study suggest that a woman can be a decision-maker on household issues but having financial control and spousal support improves her capability to access and use birth control. These findings are important to consider when designing policies on effective implementation of reproductive health-care strategy in a society where lack of economic opportunities and patriarchal norms limit women’s autonomy to exercise modern methods of family planning.

The next section of the article outlines a bargaining framework, including measures of power. Section 3 briefly discusses recent scholarship on the relationship between decision-making power and the use of modern contraception. Section 4 provides background on Tajikistan, including information on marriage, fertility and family planning. Sections 5 and 6 elaborate our methodology, outcome and predictor variables, and estimation strategies. In Section 7 we present the results, and the conclusion in Section 8.

Conceptual contexts of bargaining power

The construct ‘women’s bargaining power’ has been conceptualized and measured by researchers in numerous ways and remains ambiguous. Studies use different terms, including ‘autonomy’, ‘power’, ‘empowerment’, ‘agency’, and ‘status’ (see Malhotra 2002 for a detailed discussion). Thus, in this review, we do not limit ourselves to studies using only ‘bargaining power’ or ‘decision-making power’.

Agarwal (1997) provides a well-founded explanation for considering of the links between extra-household and intra-household factors in the study of bargaining process and gender relations. Intra-household interactions include elements of cooperation and conflict, which produce different outcomes for household members. These outcomes depend on the relative bargaining power of household members, which is affected by the strength of members’ fallback or exit position (a threat that each member or partner has). The position is influenced by a wide range of factors: individual economic assets, communal support systems, social norms and institutions, and perceptions of deservedness and needs (Agarwal 1997). Furthermore, Kabeer (1999) suggests viewing an individual’s ability to exercise power from three interrelated dimensions: resources, agency and achievements. She argues that these dimensions are important in defining indicators for measures of bargaining power as they explain the source, process and outcomes of power in connection to social norms and institutions.

Scholars seeking to understand intra-household dynamics and their dependency on resources, risks and environments have produced a rich empirical literature on measures of women’s bargaining power. Researchers distinguish direct and indirect measures of power in the study of intra-household dynamics. Direct measures refer to the relevant decision-making abilities that women have with respect to their partners. Variables include the extent of deciding alone, together or not at all on purchases, seeking health care or getting permission to visit relatives. Women are most empowered when they make these decisions alone, and least when these decisions are made by their husband alone or someone else (Allendorf 2007; Fielding and Lepine 2017). Indirect measures include women’s education, employment and marital status, and well-being outcomes such as life expectancy, contraceptive use, self-esteem and domestic violence (Laszlo et al. 2020; Mabsout and van Staveren 2010).

Several studies have used household surveys to examine correlation between women’s decision-making abilities and on their own and their family’s well-being. Women’s higher decision-making power is positively associated with rural girls’ school enrolment and utilization of maternal health-care services in Pakistan (Hou 2016; Hou and Ma 2013). Fielding and Lepine (2017) use Senegalese household survey data to model the effects of empowerment within the home on married women’s well-being. They find that a wife who cannot leave the house without her husband’s permission is more likely to report poor mental and physical health than women who can. This shows that women’s mobility and autonomy influence their health-care decision-making.

Working outside the home and earning income are other indicators of bargaining power. Income from work improves a woman’s exit position by exposing her to new situations and new information, which may further improve her bargaining power (see Doss 2013 for an excellent review). Income earned by women results in a greater investment in human capital such as education, health, and nutrition (Qian 2008). Women engaging in the labour market are more likely to experience freedom of mobility and respect in their household and community, to be aware of their rights, and to participate in community life (Mahmud and Tasneem 2014; Laszlo et al. 2020).

Women’s education is often included in empirical analyses of intra-household bargaining power. An educated woman is expected to be more informed and to make better decisions for herself as well as for the household. Furthermore, education appears to enhance women’s employment opportunities and their capacity to deal with the outside world, including government officials and service providers (Doss 2013; Duflo 2012; Kabeer 2005). Higher education improves women’s participation in all decisions (Meurs and Ismaylova 2019), the odds of infant survival (Hossain 2015), children’s nutritional status (Imai et al. 2014), and women’s final say in childrearing decisions (Yount 2005).

Researchers have claimed that household structure may influence women’s decision-making roles and lives. Allendorf (2012, 189) writes,

Within the joint family system married women gain power as they move from being daughters-in law to mothers-in-law. As mothers-in-law they have established their value and security by continuing the family line through having sons and have their own daughter-in-law to command.

Using data from the Women’s Reproductive Histories Survey in India, Allendorf (2013) finds that women who have better relationships with their husbands and mothers-in-law have more agency. Women also gain agency when they live in family structures that are supportive of female power. Analyzing the impact of kinship systems on Indonesian women’s decisions, Rammohan and Johar (2009) find that women in matrilineal kinship systems are likely to exercise greater control over personal and child-related issues than women in patrilocal communities. Physical autonomy is lower for married women living in communities practicing patrilocality.

Studies have demonstrated the importance of women’s land rights or asset ownership in strengthening women’s bargaining power and ability to make choices about their lives and environment (Mishra and Sam 2016; Panda and Agarwal 2005) and their children’s well-being (Allendorf 2007; Quisumbing and de la Briere 2000).

Reviewing previous studies, we find no evidence of examining the relationship between power related factors in interaction with other variables and outcomes which allow more hypotheses to be tested. Household decisions can be made by women but achieving desired outcomes including health depends on resources they have. Thus, we included an additional variable in our analysis which captures a woman’s financial autonomy to receive health-care services for herself and its interaction with her decision on health care. In doing so, this expands the understanding of the relationship between the variables of interest in the model, which we discuss in section 4.

Literature review: women’s decision-making power and contraceptive-use behaviour in post-Soviet Tajikistan

To our knowledge, there is no published research assessing the relationship between Tajikistani women’s decision-making and their use of modern contraception. In this section, we discuss studies that have examined the relationship between women’s decision-making power and their contraceptive choice. This review helps us develop a conceptual framework for our empirical analysis. Next, we focus on studies that analyse women’s decision-making power and their maternal and reproductive health-care services in Tajikistan.

A body of literature suggests that women who are actively involved in decision-making processes are able to control their fertility through the adoption of modern contraception (Hameed et al. 2014; Hogan, Berhanu, and Hailemariam 1999; Woldemicael 2009). Using the 2008 Nigerian DHS, OlaOlorun and Hindin (2014) find that older Nigerian women who are actively involved in making household decisions are able to make fertility-related choices.

Blackstone (2017) uses specific measures of women’s autonomy: the sole decision-making power in large and daily household purchases, the degree to which women can visit relatives or friends without their husband’s permission, women’s communication with their partner regarding family planning, and women’s attitude to wife beating. She finds that decision-making is positively associated with contraceptive use and met need for contraception in Ghana, and women who justify wife beating in one or more instances are less likely to use contraception. Adding to the literature, Hameed et al. (2014) analysed data on 2,133 women in three districts of Punjab, India, who were either using any form of contraceptive or living with unmet need for contraception. They found that couples’ joint decision-making is a stronger determinant of contraceptive use than women-only decision-making.

Spousal communication is also a significant explanatory variable in fertility-related preferences and decisions. Women’s ability to communicate with their partners about the number and timing of their children is associated with the greater use of contraceptives (Meekers and Oladosu 1996). Woldemicael (2009) finds that Eritrean women with higher education and financial autonomy were more able to discuss family planning with their partner than women who did not have these choices.

Education improves women’s knowledge and use of modern contraception. Studies from Vietnam and Sierra Leone show that education is highly correlated with women’s contraceptive-use behaviour (Dincer, Neeraj, and Michael 2013; Mocan and Cannonier 2012; Thang and Anh 2002).

Paid employment brings women economic independence, enhances their decision-making abilities, and improves their access to health care and contraception (Phan 2013). The evidence from Turkey, Bangladesh and Ghana illustrates a positive relationship between women’s employment and their use of contraception (Blackstone 2017; Laskar et al. 2006; Pekkurnaz 2020). However, some authors argue that education and employment are not key factors in declining fertility or using contraception, especially in societies with strong son preferences. These scholars suggest that government policies are the most influential factors in the fertility decline. For example, family programmes adopted by the Chinese government in 1970 and Vietnam from the late 1980s to the early 2000s contributed significantly to the fertility decline in those nations (see Phan 2013 for excellent discussions).

A few studies have examined the relationship between women’s decision-making abilities and utilization of reproductive or maternal health care using the 1999, 2003 and 2007 Tajikistan Living Standard Surveys (Falkingham 2003; Fan and Habibov 2009; Kamiya 2011; Meurs and Giddings 2012). Research has also been undertaken on family planning practices in the region, such as the effects of family planning messages on modern contraceptive-use behaviour (Habibov and Zainiddinov 2017) and reviews of Soviet and post-Soviet family planning (Henry and Juraqulova 2020). Based on these studies, we postulate that if women are able to access and utilize reproductive health or maternal health-care services, they are more likely to be exposed to family planning messages and services provided at medical facilities and to use modern methods of contraception.

Falkingham (2003) uses the 1999 Tajikistan Living Standard Survey to examine the links between poverty and women’s educational status and the use of maternal health-care services. Her study shows that women with less educational are less likely to consult a doctor during pregnancy and to give birth at health-care facilities. Kamiya (2011) finds that a woman’s influence on decision-making regarding household finances increases her likelihood of receiving maternal health services, but reduces her chance of receiving adequate antenatal care. This might be due to the higher opportunity costs of a woman’s work within the household, which we think may also affect her access to family planning services. Meurs and Giddings (2012) measure the impact of a household member’s decision and other factors on women’s use of maternal health-care services. They find that the decisions of both the husband and the oldest woman in the household are associated with a woman’s ability to access health-care services. Their findings also document that a woman’s education is positively related to her use of maternal health services, suggesting that her role in the decision is larger than that of her husband.

Background: marriage, fertility and family planning in Tajikistan

Marriage and fertility in Tajik culture

Tajiks live in a conservative society grounded in a culture with strict customs and values about gender identities and roles, and a deep-seated tradition of intergenerational family control. Men are raised to demonstrate their power and authority in the family by controlling women and younger family members. Men are meant to focus on a professional life in business or politics, outside of the home; women are to be dependents and caretakers, remaining in the private sphere (Haarr 2007; Harris 2004, 2006).

Civil affairs related to marriage are guided by the Family Code. According to the Code, marriage age is 18 years, though in exceptional cases it can be reduced by the court’s decision. Tajik society still practises arranged marriages and a bride choice is usually made by a groom’s mother. ‘In the marital relation the power is on the side of husbands and their families, who can thus bend the situation to their advantage’ (Harris 2004, 101). A daughter-in-law living in husband’s familial house is expected to be a hard worker and obey her mother-in-law to avoid power struggles within the household. Further, not so many nuclear families can be observed or found in the country, and major household decisions are usually made in consensus with husbands’ parental family, not by individual nuclear units (Harris 2004). In multigenerational homes the young wife is expected to assent to financial and household decisions made by her husband and mother-in-law.² This disempowerment extends to decisions on the young wife’s health care, including her use and choice of family planning methods and her fertility. The 2012 TjDHS reports that a staggering 36.7% of married women’s health decisions are made by their husband or someone else, and 46% of married women make decisions about their own health jointly with their husbands.

In Tajikistan, tradition expects couples to have a child within their first or second year of marriage. As Harris (2004) notes, ‘Although the desire for children is very strong and a daughter-in-law must prove her fertility as soon as possible, parents-in-law and even husbands often do not seem very interested in the outcome of pregnancy’ (109). On average, Tajik couples have three or four children (2012 TjDHS), though couples with 6–12 children can be found in rural areas of Tajikistan.³ Many prefer to have a son as the first child or to have at least one son among daughters. A son is expected to bear the responsibility for his parents’ well-being in old age, even if he does not live with them (Grogan 2007).

Political support of family planning programmes

In Tajikistan, abortion is considered a predominant method of family planning, reflecting the history and traditions that remain in place from Soviet times. According to the 2012 TjDHS, induced abortion is common among Tajik women with many children, older women, educated women, and those from wealthier households. About 83% of abortions occurred after no contraceptive method was used. Some women may have unsafe abortions, resulting in complications or disabilities, though this has not been systematically documented or studied in independent Tajikistan.

Over the past 20 years, the country’s fertility rate has dropped from about seven children per woman to an average of four.⁴ Rates of contraceptive use are low across Tajikistan, in spite of 83% of Tajik women reporting knowledge of at least one modern method. Just one in four Tajik married women ages 15–49 uses modern birth control, and one in five married women experiences an unmet need for family planning (2012 TjDHS).

The government has taken several legal measures to increase access to and use of family planning methods. The Law on Reproductive Health and Rights (2002) and a ten-year Strategic Plan for Reproductive Health (2005–2014) are two examples of legislative efforts to improve reproductive health services and outcomes for Tajiks. The government of Tajikistan, jointly with local and international organizations, has conducted national family planning outreach efforts to encourage safe motherhood and uptake of modern contraceptives.

The government formed the Republican Centre for Reproductive Health within the Ministry of Health and Social Protection of Population to administer and manage such projects, including family planning programmes, at the national level. Each oblast, or region of the country, has a reproductive health-care centre, established by the Republican Centre for Reproductive Health, to implement and monitor projects at the local level. With the financial support of development partners, centres of reproductive health and NGOs conduct meetings with beneficiaries in rural areas. Including men and women in discussions of family planning at the community meetings is a priority. Women’s mothers and mothers-in-law are also invited to these meetings, as they influence their daughter’s or daughter-in-law’s health decisions.⁵ Harris (2004, 71) uses interviews and observations of Tajik women living with extended family to document the behaviour of daughters-in-law obeying their mothers-in-law, who in turn influence the daughters-in-law’s mobility and decision-making.

To support the work of reproductive health-care centres on family planning services, UNFPA Tajikistan founded the Association of Family Planning. The association conducts meetings and trainings for representatives of centres to discuss ways to develop and deliver resources for awareness-raising campaigns on family planning. Family planning methods have been provided free of charge to women in both rural and urban areas during the awareness campaigns.⁶

In 2005 a reproductive health logistics management and information system was introduced with the support of development partners. The system was piloted in northern Tajikistan (Sughd Region), where UNFPA provided a cascade approach to training family practitioners in providing modern methods of contraception and effective family planning counselling. In 2014, a result, 65% of primary care facilities provided at least three types of contraceptives (up from 47% in 2010).⁷

To continue supporting the population with affordable family planning services, the Ministry of Health and Social Protection of Population collaborated with national and development partners to review and finalize the Family Planning Action Plan (2017–2020). According to the plan, over 700,000 people will get access to modern methods of contraception. But while increasing access to health services lowers an important and potentially overwhelming barrier to care for many women, access to modern methods may not directly influence the use of methods. The distinction is important.

Understanding the relationship between of women’s household decision-making power and their choice of contraceptive method may provide an additional area to consider in designing family planning programmes.

Methodology

Data and descriptive statistics

To study how factors determining women’s decision-making ability affect their family planning practice, we use the 2012 Tajikistan Demographic and Health Survey. This first DHS of Tajikistan and was conducted by the Statistical Agency and the Ministry of Health and Social Protection of the Population. Technical assistance for the survey was administered and funded by the USAID-funded MEASURE DHS programme. The data set is a nationally representative survey containing information on population and health issues for 9656 women between the ages of 15 and 49 from 6674 households. The survey took place between June and September 2012. The 2012 TjDHS survey is ideal for this study because the questionnaire is designed to provide detailed information on fertility preferences and outcomes, family planning practices, and individual and household characteristics for a nationally representative number of women.

Outcome variable

The outcome variable is whether a woman is currently using any type of modern contraception. We restrict the sample to women of reproductive age, between 15 and 49, and currently married or living in a union (6381 women). Pregnant and declared infecund and sterilized women are excluded since they are not in the position of making contraceptive choices. The final sample includes 5091 women. About 26% of these currently use a modern method of contraception. Table 1 presents descriptive statistics for all variables.

Table 1. Summary statistics for variables used in the analysis (n = 5,091).

Variable	Mean	Std. error	Type of variable
Decision-making variables
Woman solely makes decision about her own health care	0.151	0.008	Binary
Woman jointly with husband or someone else decides on her own health care	0.463	0.015	Binary
Woman solely makes decision on visits to her family or relatives	0.095	0.007	Binary
Woman jointly with husband or someone else decides on visits to her family or relatives	0.543	0.014	Binary
Woman has say on contraception decision	0.115	0.006	Binary
Woman jointly with husband make contraception decision	0.165	0.007	Binary
Woman has no financial constraints to get medical treatment for herself	0.563	0.015	Binary
Woman solely makes decision about her own health care AND has no financial constraints to get medical treatment for herself	0.084	0.006	Binary
Woman together with husband or someone else decides about health care AND has no financial constraints to get medical treatment for herself	0.275	0.012	Binary
Individual characteristics
Head of household	0.040	0.0035	Binary
Wife	0.457	0.012	Binary
Daughter-in-law	0.447	0.012	Binary
Woman’s years of schooling	9.864	0.091	Binary
Woman works outside	0.208	0.011	Binary
Woman owns land	0.390	0.018	Binary
Woman’s age (years)	32.621	0.141	Continuous
Woman's age squared	1137.736	9.442	Continuous
Marriage duration	12.359	0.148	Continuous
Last sexual intercourse was two months ago	0.216	0.009	Binary
Age difference in couple (years)	3.918	0.076	Continuous
First child is son	0.497	0.007	Binary
Children at home	2.723	0.036	Continuous
Husband works	0.974	0.003	Binary
Husband years of schooling (years)	11.442	0.083	Continuous
Family planning initiatives
Family planning message received from a visit by family planning worker	0.258	0.011	Binary
Family planning message received at a health-care facility	0.267	0.010	Binary
Other control variables
Dushanbe (Ref: RRS region)	0.085	0.005	Binary
GBAO	0.020	0.001	Binary
Sughd	0.319	0.013	Binary
Khatlon	0.338	0.016	Binary
Rural (Ref: urban)	0.760	0.021	Binary
Wealth: poorer (Ref: poorest)	0.193	0.011	Binary
Wealth: middle	0.200	0.010	Binary
Wealth: richer	0.212	0.012	Binary
Wealth: richest	0.206	0.014	Binary

Predictor variables

Variables defining women’s decision-making choices

To examine the relationship between a woman’s decision-making power and her ability to control birth, we consider these variables: control over her own health care, decision to visit her family and relatives, access to money needed for treatment, and contraceptive decision. The selection of these direct indicators is based on the structure of 2012 TjDHS. The literature suggests that these indicators are likely to be important predictors of women’s reproductive behaviours (Blackstone 2017; Woldemicael 2009)

The survey asked women whether they were involved in decision-making in four different areas: their own health care; large household purchases; visits to her family or relatives; and what to do with the money their husband earns. We chose women’s participation in health-care decisions as it is the more likely indicator of women’s health-care decision-making. Another variable of interest is visits to relatives or friends, which suggests a woman’s influence over her social life. Decision-making on large household purchases is not considered in the analysis as it involves inputs of both husband and wife in the decision, according to norms and gender role expectations in Tajik society. Within marriage, a man is expected to provide for the family economically and a woman is expected to maintain the household and ensure that her husband is aware of what necessities need to be purchased for the family. Decisions about how to handle money that the husband earns is not included in the model as we believe it is an earner's or couple's decision how to spend. However, our data shows that it is more likely an indicator of joint decisions (47%) compared to husband's decision alone (29%) and wife's solely decision (7%).

We created variables for a woman’s control over her own health care and decision-making on visits to relatives and families for each response of ‘alone’ and ‘jointly with husband or someone else’ to minimize variations in the relationship between responses and the outcome variable. The reference group for each variable includes decisions made only by the husband or others. Household decision-making binary variables are:

• Woman solely makes decisions about her own health care = 1, 0 otherwise;

• Woman decides jointly with husband or someone else = 1, 0 otherwise;

• Woman solely makes decisions about visits to family or relatives = 1, 0 otherwise.

• Woman decides jointly with husband or someone else of her visits to family or relatives = 1, 0 otherwise.

Table 1 shows that household decisions are more often made in negotiation with the husband or someone else rather than solely by women. Over 45% of respondents reported that they make health-care decisions or decisions to visit family or relatives in communication with their husband or someone else. Few women (10%) are able to make decisions on visits to their family or relatives without the husband’s input in final say.

A woman can have access to family planning services if she has no financial barriers to care. The TjDHS asks women a set of questions on ‘major problems preventing them from getting medical advice or treatment’. One of these questions is ‘getting medical help for her self: getting money needed for treatment’, with two responses: ‘a big problem’ and ‘not a big problem’. We created a binary variable, coded as 1 if she responded that getting money for treatment was not a big problem (about 56% of responses). A woman may be able to make health-care decisions, but a lack of financial independence may prevent her from receiving reproductive health-care services, including family planning. To explore this we interact a woman’s health-care decision with the financial-constraint variable. Eight percent of women make health-care decisions alone and have no financial barriers to care, while 28% make decisions with their husbands and do not experience financial barriers to care.

Another variable of interest is contraception decision. Two binary variables were created to investigate this decision-making: wife makes decisions alone; and wife makes decisions jointly with husband. About 12% of women report making contraception decisions alone, whereas 17% make decisions with the husband.

Individual characteristics

We include additional indicators that influence women’s decision-making power, such as years of schooling, employment outside of household (with work for family members as a reference group), household structure (defined in terms of a woman’s relationship to the head of the household), and land ownership. On average, women in the sample have 10 years of schooling, indicating that their education level is generally secondary, or completion of a high school diploma. In the education system of Tajikistan, general secondary education is between 10 and 11 years of study, and it includes primary (grades 1–4), general basic (grades 5–9), and secondary (grades 10–11). In our sample, 21% of women maintain work outside of the household, which provides an opportunity for them to earn and freedom to move.

A woman’s position in a family or a household may influence her decision-making power. Our binary variable wife is 1 if a woman claimed to be a wife of the household head (46% so identified), indicating that she does not live in an extended family. Similarly, 45% identified as daughter-in-law. Next, a variable for a woman who identified herself as a head of household (4 percent) is added to observe her contraceptive behaviour. The reference group includes women themselves as the daughters, mothers, sisters, granddaughters and other relationships to the head of the household. In Tajikistan, young brides usually move to the house of their husband’s family after marriage. This arrangement affects women’s opportunity cost of time for child and elderly caring, and limits her time to receive adequate health-care services, including family planning.

Another binary variable in the model refers to female land ownership. Previous studies use control over economic resources, measured by studying a woman’s asset ownership; this shows its positive effect on various aspects of women’s reproductive health (Beegle, Frankenberg, and Thomas 2001). In our sample, about 40% of women own land, alone or with their husband.

Other control variables

We include control variables for women’s age and marital duration. On average, respondents were 33 years old and marital duration was 13 years, as of the survey date. The average gap between a husband and wife’s ages is four years.

Overall, in 2012 TjDHS, about 67% of women report that they currently live with a husband or a partner. We expected this number to be lower, given that the country has a relatively high male labour migration rate. Tajikistan is an exporter of seasonal low-skilled migration, with 96% of migrants heading to Russia. About 52% of households in Tajikistan have at least one family member who is or has been in labour migration (Danzer and Ivashchenko 2010; Danzer, Dietz, and Gatskova 2013). The long absence of a husband might reduce a wife’s need for family planning services; on the other hand, his return home may increase her chance of getting pregnant if she is not using any contraception. The DHS documents women’s reports of the time elapsed since their last sexual encounter (in days, weeks, months and years). We created a binary variable equal to 1 if a woman reported that their last sexual intercourse was 2 months ago. This variable can serve as a proxy for the husband’s possible absence due to labour migration and his wife’s need for contraception.

Fertility characteristics are represented by a continuous variable for the number of children at home. Most women in our sample have three children. Another fertility feature that affects the use of contraception is having a son as a first child. In a patriarchal society, sons are preferable to daughters. If a woman has a son with her first birth, she may space or limit births afterwards. Kugler and Kumar (2017) showed that couples in India tend to have more children if the firstborn is girl. We created a binary variable, first child is son, to explore the effect of this variable on Tajik women’s further decision to use contraception. About 50% of women report that their first child was male.

To account for husband characteristics, we create variables for the husband’s years of schooling and employment. The average schooling is about 12 years, which is longer than the average for women. Accordingly, he might have at least a college degree. Furthermore, about 97% of men participate in the labour force, which is notably high. The Tajik government has not published official (un)employment statistics. A 2017 World Bank jobs diagnostic report for Tajikistan states that only 43% of eligible working-age adults are employed or currently seeking employment (Strokova and Ihsan 2017). The number from our sample may include men who are in labour migration as well.

A set of binary variables was created regarding how family planning information is delivered to women in Tajikistan: family planning message received in health facility and family planning message received from the visit of family planning doctors. These variables also provide information on the government’s intervention in the promotion of reproductive rights in the context of its family planning strategy. About 26% of women reported receiving family planning messages from doctors’ visits, and 27% were told about the method of birth control.

Women’s contraception use varies across households and income levels due to residence. About 76% of respondents live in rural areas. The DHS working team has constructed a wealth index based on information they obtained on dwelling and household characteristics and assets (for details, see the 2012 report). The wealth index has five equal categories: poorest, poorer, middle, richer and richest. We created binary variables for these five categories to control for variation in household wealth.

Empirical approach

In this section, we turn to the estimation of the effects of women’s decision-making power on the use of contraception. The outcome variable is whether a woman is currently using a modern method of contraception. Since this variable is binary, the probit model is applicable for the analysis. The model is(1) $C_i=B_i{a}_1+F_i{a}_2+X_i{a}_3+Z_i{a}_4+M_i{a}_5+F{M}_i{a}_6+{ϵ}_i$(1) where $C_i$ is the self-reported use of modern contraception by woman i (i = 1,2, … , N), $B_i$ is a vector of variables measuring her decision-making power; F_i is the financial capability to get medical treatment for herself; X_i and $Z_i$ are sets of control variables for the woman, her husband and household characteristics, $F{M}_i$ is a set of variables capturing where family planning messages are received; and the a’s are coefficients that measure the effects of the explanatory variables on the outcome variable.

Results and discussion

Relationship between women's decision-making power and their use of contraception

We perform two probit estimations to examine the relationship between determinants of women’s power positions and the use of modern contraception. In the second estimation, we interacted two explanatory variables, such as woman’s say on health care, with her financial capability to get medical treatment. The results of both estimations are reported in Table 2. Model 1 has the results of probit estimation for Equation (1), and Model 2 reports results for the estimation which includes an interaction term. Both estimations and models are accompanied by columns containing marginal effects.

Table 2. Probit estimates of relationship between women's decision making abilities and contraceptive use.

	Model 1		Model 2
	b (se)	Marginal effects	b (se)	Marginal effects
Woman solely makes decision about her own health care	−0.110	−0.033	−0.291**	−0.086
	(0.097)		(0.130)
Woman jointly with husband or someone else decides on her own health care	−0.189**	−0.056	−0.298*	−0.088
	(0.087)		(0.119)
Woman solely makes decision on visits to her family or relatives	−0.160	−0.047	−0.159	−0.047
	(0.112)		(0.112)
Woman jointly with husband or someone else decides on visits to her family or relatives	−0.062	−0.018	−0.059	−0.017
	(0.087)		(0.087)
Woman has no financial constraints to get medical treatment for herself	3.297***	0.064	0.091	0.890
	(0.097)		(0.103)
Woman solely makes decision on her own health care AND has no financial constraints to get medical treatment for herself			0.304*	1.870
			(0.163)
Woman jointly with husband or someone else decide on health care AND has no financial constraints to get medical treatment for herself			0.186	1.360
			(0.137)
Woman has say on contraception decision	3.135***	0.978	3.303***	0.979
	(0.084)		(0.098)
Woman jointly with husband make contraception decision	0.217***	0.930	3.139***	0.930
	(0.064)		(0.084)
Woman's years of schooling	0.005	0.002	0.004	0.330
	(0.012)		(0.012)
Woman works outside	−0.075	−0.022	−0.073	−0.950
	(0.077)		(0.077)
Head of household	0.118	0.035	0.108	0.560
	(0.192)		(0.192)
Wife	0.178	0.053	0.170	1.380
	(0.123)		(0.124)
Daughter-in-law	0.230	0.068	0.225**	1.920
	(0.116)		(0.117)
Woman owns land	0.122*	0.036	0.127*	1.830
	(0.069)		(0.069)
Woman’s age (years)	0.191***	0.057	0.190***	5.360
	(0.036)		(0.036)
Woman’s age squared	−0.003***	−0.001	−0.003***	−5.980
	(0.001)		(0.001)
Marriage duration	0.020*	0.006	0.020*	1.860
	(0.011)		(0.011)
Last sexual intercourse was two months ago	−0.049	−0.014	−0.050	−0.650
	(0.077)		(0.077)
Age difference in couple (years)	−0.001	0.000	−0.001	−0.310
	(0.005)		(0.005)
First child is son	0.081	0.024	0.084	1.390
	(0.060)		(0.060)
Children at home	0.051**	0.015	0.052**	2.080
	(0.025)		(0.025)
Husband works	0.056	0.017	0.039	0.200
	(0.196)		(0.195)
Husband years of schooling (years)	−0.010	−0.003	−0.011	−1.020
	(0.011)		(0.011)
Family planning message received from a visit by family planning worker	0.287***	0.085	0.292***	3.630
	(0.080)		(0.080)
Family planning message received at a health-care facility	0.118	0.035	0.115	1.420
	(0.081)		(0.081)
Dushanbe (Ref: RRS region)	−0.370***	−0.110	−0.379***	−3.120
	(0.122)		(0.122)
GBAO	0.208*	0.062	0.201*	1.830
	(0.111)		(0.110)
Sughd	−0.193**	−0.057	−0.200**	−2.130
	(0.095)		(0.095)
Khatlon	−0.114	−0.034	−0.114	−1.130
	(0.085)		(0.085)
Rural (Ref: urban)	−0.007	−0.002	−0.007	−0.080
	(0.093)		(0.093)
Wealth: poorer (Ref: poorest)	0.097	0.950	0.099	0.950
	(0.105)		(0.104)
Wealth: middle	0.015	0.004	0.022	0.210
	(0.106)		(0.105)
Wealth: richer	0.054	0.016	0.066	0.580
	(0.114)		(0.113)
Wealth: richest	0.099	0.029	0.109	0.870
	(0.125)		(0.125)
Constant	−5.145***		−5.035***
	(0.669)		(0.670)
Observations	5091		5091
Chi-squared	2335.224		2337.337
Pseudo-R^2	0.681		0.681
Log-likelihood	−1007.004		−1005.366

***1%, **5%, *10% significance. Robust standard errors in parentheses.

Comparing the results of these two estimations we observe that the coefficients for the main variables of interest change slightly, but signs or effects are consistent. Therefore, we focus on interpretation and discussion of the results of Model 2.

Variables of interest

Most of the variables of interest are statistically significant in explaining women’s use of modern contraceptives. Surprisingly, the probability of using modern contraception is lower for women who make decisions about health care or visits to family alone, compared to those who have no voice at all. Similar results were obtained for women who make decisions jointly with the husband or someone else, compared to those who do not actively participate in decisions on health care or visits to family or relatives.

Even if women have a voice in decisions, lack of financial autonomy may prevent them from receiving adequate health-care services, including family planning. Therefore, we interacted the variable defining a woman’s financial capability to get health care for herself with her choice regarding health care and visits to family. The probability of using contraception is 9 percentage points higher for women who do not experience financial barriers to medical treatment, but the effect is statistically insignificant. However, the interaction term is positively and significantly correlated with contraceptive use. The probability of using birth control increases by 187 percentage points for women having both a say in their own health care and the financial means to get medical treatment. The same relationship is observed for joint health-care decisions, although it is statistically insignificant.

These results suggest that decision-making abilities, along with financial resources, predict women’s use of modern family planning methods. Having a voice in the household but not having insufficient financial resources may limit a woman’s choice to receive reproductive health services, including birth control. Therefore, the presence of both factors is important for women to meet their contraception needs.

Contraceptive use is 98 percentage points higher among women making contraception decisions alone than among women who have no voice in this choice. Joint husband–wife decision making on modern contraception is also positively associated with contraceptive use. Consistent with previous studies, a desire to limit or space births and the intention to use contraception are strongly predicted by women’s ability to discuss family size and family planning with their husband. Husbands’ involvement in the discussion of family planning services appears to be an important indicator of support for women’s decision to adopt modern methods of contraception (Hogan, Berhanu, and Hailemariam 1999; Terefe and Larson 1993; Woldemicael 2009).

Other determinants

Other determinants of decision-making power included in the model are education and employment. A woman’s years of schooling appears to be positively associated with the use of contraception, whereas her employment outside household is negatively correlated with the contraceptive use, although both effects are statistically insignificant. In Tajik society, household structure may affect women’s decision-making ability, contributing to a woman’s choice to control births. Women who reported being head of household, wife or daughter-in-law were more likely to use contraception than women who are the daughter or mother or granddaughter, but both results appear to be statistically insignificant. On the other hand, we find that women’s land ownership is statistically significant and positively correlated with women’s contraceptive behaviour. The probability of contraceptive use is 183 percentage points higher for female landowners compared to those who do not own land.

Control variables

Contraceptive use increases with age. The older Tajik women get, the more likely they are to use contraceptives, but at a decreasing rate. The likelihood also increases with the length of a marriage. Having children and having a son as a first child are also associated with an increase in contraceptive use. Presence of children at home increases the likelihood of using contraception by 200 percentage points. But there is a negative coefficient associated with the women’s last sexual intercourse variable. That is, the use of contraception is lower among women who last had intercourse at least two months ago. We used this variable as a proxy to define whether women are using any modern methods of birth control while their husbands are currently away or in a labour migration. As for the husbands’ characteristics, we observe that male employment is positively associated with women’s use of contraception, but male years of schooling is not. These mixed results need further study and discussion.

Family planning messages received either at a health-care facility or on a visit by family planning workers increase Tajikistani women’s use of contraception. The probability of using contraception is 363 percentage points higher among women who were visited by family planning workers, and the coefficient is statistically significant. These factors explain the effects of family planning intervention programmes on women’s reproduction control in Tajikistan.

Finally, we observe regional differences in the use of contraception. Women living in the capital, Dushanbe, are less likely to use contraception than women living in Districts of Republican Subordination (DRS). A similar correlation is observed for women residing in Sughd and Khatlon. However, contraceptive use is higher for women in Gorno-Badakhshan Autonomous Oblast compared to DRS. We also learned from the 2012 TjDHS that women in GBAO have greater decision-making power than women in other regions. This might explain the greater contraceptive practice in the region. On the other hand, both qualitative and quantitative studies are needed to evaluate reproductive health-care programmes, including family planning services, and women’s beliefs and perceptions about modern contraceptive use in all regions of Tajikistan. This would help the government allocate and distribute family planning resources efficiently and effectively across regions, and increase awareness of the benefits of family planning, targeting areas with low contraceptive use.

Conclusion

In this article, we examined factors related to women’s decision-making power and their relationship with contraceptive-use behaviour using the 2012 TjDHS. Our model includes variables on a woman’s use of contraceptives, decisions on her own health care, visits to her family or relatives, and financial autonomy to utilize health services. We control for determinants such as a woman’s status in the household, years of schooling, outside employment, and land ownership. All these decision weights are found to have effects on women’s use of contraception. The probability of using contraception is lower among women who have a say on their health care and their visits to family or relatives. However, interacting a woman’s ability to make health-care decisions with her financial autonomy, we observe that the probability of using contraception increases by 187 percentage points, holding other factors constant. Contraceptive use is 98 percentage points higher among women who make contraceptive decisions alone, and about 93 percentage points higher for women who have inputs in the decisions jointly with the husband, compared to women who do not have a say in the decision. The husband’s involvement in family planning and the decision to control births is crucial in supporting the wife’s use of contraception. Therefore, initiatives focused on increasing awareness of family planning services and their benefits for family well-being should target both Tajik women and men in Tajikistan.

Furthermore, our findings show that contraceptive use is higher among women receiving family planning messages either at a health-care facility or from a visit by family planning workers. Exposure to family planning messages through health facilities or workers is likely to increase men and women’s knowledge and use of modern methods. Annually, up to 100,000 unwanted pregnancies have been prevented in Tajikistan as a result of actions taken by family planning programmes.⁸

Our findings suggest that prioritizing investment in improving women’s autonomy and decision-making ability empowers them to negotiate fewer children and acquire information on the benefits of family planning for their health, the health of their children, and the well-being of their family. Policy makers may consider supplementing family planning policies with the provision of financial support or subsidy to defray the cost of family planning services for women who cannot afford them. More research is needed to explore female perceptions about family planning methods in Tajikistan, how those vary geographically, and how they may impact a woman’s choice of family planning method.

Acknowledgements

We would like to thank the anonymous referees, whose comments and feedback substantially improved the quality of this contribution. We are grateful to Professor Charles M. Becker for his invaluable comments on previous versions of this article.

Disclosure statement

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

Notes

1 ‘Bargaining power’ and ‘decision-making power’ will be used interchangeably throughout the article. We also discuss other terms used in previous studies.

2 Zakluchitelnie zamechaniya Komiteta po likvidasii diskriminasii v otnoshenii jenshin: Tajikistan [Concluding comments of the Committee on the Elimination of Discrimination against Women: Tajikistan], (CEDAW/C/TJK/CO/3 February 2007), <http://www.unhchr.ch/tbs/doc.nsf/898586b1dc7b4043c1256a450044f331/16313018419e3b4ac12572a400337a3e/$FILE/N0724356.pdf>.

3 Kak v Khatlonskoi oblasti uchili planirovat sem'u [How family planning was taught in Khatlon region], <http://cso-central.asia/kak-v-xatlonskoj-oblasti-uchili-planirovat-semyu/>.

4 UNICEF’s The State of the World’s Children 2010; WHO World Health Statistics.

5 Kak v Khatlonskoi oblasti uchili planirovat sem'u [How family planning was taught in Khatlon region], <http://cso-central.asia/kak-v-xatlonskoj-oblasti-uchili-planirovat-semyu/>.

6 UNFPA Tajikistan Press Release, ‘Reproduktivnoe zdorov’e - osnova blagopoluchiya naseleniya' [Reproductive health is a foundation for populations' wellbeing]’, <https://tajikistan.unfpa.org/ru/news/репродуктивное-здоровье-–-основа-благополучия-населения>.

7 Review of the National Programme on the Development of Family Medicine 2011–2015 in Tajikistan, <http://www.euro.who.int/__data/assets/pdf_file/0018/314433/Review-of-the-National-Programmeme-on-the-Development-of-Family-Medicine-2011-2015-in-the-Republic-of-Tajikistan-rev1.pdf>.

8 Vlasti Tajikistana reshili umenshit rojdaemost v strane [The government of Tajikistan has decided to reduce births in the country], <https://rus.ozodi.org/a/30103238.html?fbclid=IwAR2XIPvod6CKk357PwU_Z7im1OeWvOaUx_nPqnu9gNkP3obH8GTBRdvJCTA>.