The Taliban have devastated the rights of women in Afghanistan by limiting their ability to participate in society, potentially causing permanent harm to women’s autonomy within their own marriages. We posit that women married during the Taliban rule may see reduced autonomy based on changes in gender norms from Taliban policy. We estimate the effect of being married during various levels of Taliban rule in 2002, 2009, and 2012 on outcomes in 2015. We use variation in the Taliban’s control in a province at the time of marriage to estimate the effects of norms at the time of a marriage on outcomes measured in 2015. We find that women’s decision-making power was higher in 2015 for those who married when their province was not under Taliban control, but surprisingly, their experience of intimate partner violence (IPV) also increased, suggesting that intrahousehold conflicts potentially increased from regime changes. We link these results to a previous literature that suggests that retaliatory IPV may occur as women gain more power and lower their fertility, with the finding that when the Taliban lost control, fertility fell substantially. These results suggest potential policies that address the long-term negative impacts of the Taliban rule on women’s autonomy should the regime lose power again, with special attention given to potential increases in IPV.
Research Article
Marriage, Taliban Rule, Norms and Women’s Empowerment in Afghanistan
https://doi.org/10.21203/rs.3.rs-3239895/v1
This work is licensed under a CC BY 4.0 License
Version 1
posted
You are reading this latest preprint version
The recent return of Taliban rule has devastated the rights of women in Afghanistan. Taliban leaders are presently implementing sets of discriminatory measures to effectively erase the participation of women in civil life in Afghanistan (McAuliffe 2022). Afghan women are currently facing long-term sociocultural factors that restrict their basic rights and access to education, work, and social services (Human Rights Watch 2022). In addition, most women have lost their employment due to the Taliban restricting women’s access (Human Rights Watch 2022). The Taliban’s takeover of Afghanistan has undoubtedly jeopardized women’s rights and increasingly threatens their rights. A survey conducted in 2015 by the Demographic and Health program of USAID showed that 90% of women in the country have experienced violence from their husband. The threat of violence continues to be a harsh reality for women under Taliban rule, as their voices continue to be silenced (Mannell, 2021). The Taliban had nearly total control of Afghanistan from the late 1990s until the U.S. invasion in 2002. Over the last twenty years, Taliban control varied in part due to U.S. military intervention before the Taliban took back control in August 2021.
What is less known is how the Taliban’s discrimination against women may have longer-term impacts on women’s empowerment even if the Taliban loses power again. This paper examines whether the Taliban’s influence on social norms has long-term impacts on women’s decision-making power, likelihood of experiencing and views on intimate partner violence (IPV), fertility, and contraception use. Specifically, we look at how variance in Taliban control from 2002 to 2012 at the time of a woman’s marriage influences her empowerment at a later date, 2015. We hypothesize that women married during times of stronger Taliban rule will have less decision-making power and will have more children.
Previous research suggests that community norms influence women’s empowerment (Jejeebhoy & Sathar, 2001; Koenig et al., 2006; Yount et al., 2016) and that Taliban rule as described above clearly shifted these norms. Previous work on the influence of the Taliban’s first regime (1996 to 2001) suggests that women living in Taliban-controlled provinces were less educated and less likely to work outside the home (Noury & Speciale, 2016). The impacts on marriage are mixed, as Noury & Speciale (2016) show, women who grew up under Taliban rule married younger, while Maity & Shukla (2021) suggest that women in the ethnic groups opposed to the Taliban (Tajik/Uzbek) married slightly later. The Taliban had various levels of control during the prolonged conflict with US military forces from 2002 to 2021, which is the focus of this paper.
When governments become more favorable to women, such as when more female leaders are elected, one potential outcome is decreasing acceptance of IPV, as seen in Rwanda (Burnet, 2011) and Indonesia (Kuipers, 2020). However, these changes may create friction and increase IPV, as suggested by (Burnet, 2011) and in India (Anukriti et al. 2022). Anukriti et al. (2022) suggest a potential mechanism is friction created by women’s increasing autonomy lowering their fertility when their husbands still prefer to have more children. Anukriti et al. (2022) find that more female leaders are associated with a higher probability of experiencing IPV, increased contraception use, longer birth spacing and an increased probability of men wanting more sons than they have.
Conflict has been shown to have long-term impacts on women’s empowerment (Mattina, 2017; Shemyakina, 2009). Additionally, many of the circumstances in Afghanistan, such as low education and child marriage, are responsible for one of the highest rates of IPV in the world and lowest women’s empowerment (Akbary 2020; Blum et al., 2019.; Alexander &Thelwell 2020).
This study uses variation in Taliban control at the provincial level at the time of a woman’s marriage to investigate the influence of the Taliban on women’s empowerment, specifically regarding women’s household decision-making power at the time of marriage and experience of intimate partner violence and fertility. Previous papers have looked at the first Taliban regime from 1996 to 2001 (Noury & Speciale 2016; Maity & Shukla 2021), while we examine variation in Taliban control during the US conflict. We use data from the 2015 Afghanistan Demographic Health Survey (DHS) taken at a time of minimal Taliban rule.
We find that women married in provinces that had higher Taliban control at the time of their marriage had reduced decision-making power compared to women married when the US had more control in their province when measured in 2015 during a time of higher US control. Surprisingly, women married under Taliban rule reported less IPV than those married when the US had more control. We suggest that this could be due to retaliation, as, similar to Anukriti et al. in India when women gain power (2022), we find that when the Taliban lost control, women’s fertility declined, and her number of births relative to her husband’s desire also decreased. We do not find any changes in acceptance of IPV, suggesting that measurement error based on likelihood of reporting is not an issue. Unlike Anukriti et al. (2022), we do not see changes in contraceptive use. Finally, we do not observe a relationship between marriage age and Taliban control, which contrasts Noury & Speciale’s (2016) findings during the first Taliban regime.
Our results also suggest that Taliban control could have long-term impacts through norms and education. We test for the transmission of norms for decision making, IPV, and marriage age across generations by estimating the relationship between outcome variables for women married from 2002 to 2012 and those women married in her province before the first Islamic government took over in 1992. These elderly women were married under the norms established within the province pre-Taliban. We see a strong relationship across generations for the outcome variables, suggesting that the Taliban’s level of influence from 2002 to 2012 may have longer-term impacts. Finally, the Taliban, as mentioned above, have reduced women’s education. We find, as expected, that women’s education is a key factor in increasing decision-making power, reducing IPV, and increasing marriage age.
The Taliban in their 2021 takeover have strongly discouraged women’s autonomy, limiting their mobility, decision-making power, and ability to leave domestic violence situations (Monir 2022). Taliban rule has led to numerous negative outcomes for women, including lowering marriage age, decreasing women’s ability to travel, and reducing support for survivors of IPV, which can potentially change community norms around gender equality within Afghanistan (Noury & Speciale 2016). Increased Taliban control has been shown to increase the number of child marriages in Afghanistan, as many families force their daughters into marriages as a way of protecting them from violence (de Silva de Alwis, 2021; Noury & Speciale 2016). Additionally, since the Taliban took over, women in Afghanistan have been banned from traveling more than 72 kilometers without a male relative (Monir 2022). Noury & Speciale (2016) suggest that this may have long-lasting impacts, as women exposed to more years of Taliban control during their youth were less likely to work outside of the household in 2008.
Community norms have a substantial influence on how women are treated in terms of their autonomy within marriage, their likelihood of experiencing IPV, and the choice of who she marries. These community norms may vary by ethnic group, and as Maity & Shukla (2021) suggest, ethnic groups’ relationship to the Taliban may affect how these norms change. Specifically, Maity & Shukla (2021) find that ethnic groups opposed to the Taliban, Tajik/Uzbek, saw an increase in marriage age. Outside of Afghanistan, one study of Pakistan, India, Malaysia, Thailand, and the Philippines concluded that women’s autonomy (e.g., mobility, decision making, and experience of IPV) is highly influenced by community norms and values (Jejeebhoy & Sathar, 2001). Second, a study of Egypt concluded that women’s agency depends on attitudes about gender, specifically regarding violence against wives (Yount et al., 2016). Similarly, a study of North India concluded that intimate partner violence was significantly related to community-level norms concerning wife-beating (Koenig et al., 2006). These studies suggest that community norms can limit women’s control over resources and freedom of movement, which reduces their ability to leave the household when they experience violence.
In India, Anukriti et al. (2022) find that women who live in districts with more female elected officials have an increased likelihood of experiencing IPV. They suggest the mechanism of misaligned fertility preferences driving conflict, as mentioned above. Similarly, Eswaran & Malhotra (2011) suggest that increasing women’s autonomy is also associated with increased IPV in India. In Zambia, providing women with contraceptives reduced happiness due to conflict with their husbands over fertility (Ashraf et al. 2014). This work in sum suggests that the downfall of the Taliban may increase autonomy, leading women to have fewer children but also to experience more IPV.
Conversely, two studies suggest that increasing women’s decision-making power tends to decrease IPV. Pakistan Mavisakalyan & Rammohan (2021) find that as women’s autonomy increases, IPV tends to decrease. In a study of 12 African countries using a similar survey instrument, Donald et al. (2023) find that women with more decision-making power are typically less likely to experience IPV.
War has also been shown to affect marriage choices and potentially women’s autonomy in other countries. Given the decades of conflict in Afghanistan, security and conflict also likely affect women’s marriages in Afghanistan. Mattina (2017) found that the 1994 Rwandan in areas that had more genocide caused an increase in domestic violence and decreased women's decision-making power for women who married after the genocide relative to women who married before. Mattina’s findings indicate that the “detrimental effects of genocide on women’s well-being were still sizeable as far as 16 years after the end of the mass killings.” Shemyakina, (2009) shows conflict that when a woman in Tajikistan is exposed to regional conflict during her most marriageable years, her probability of entering her first marriage is reduced by fourteen percentage points, although this delay may be a result of fewer men and it’s unclear what happened to IPV in this case.
Education tends to increase women’s autonomy and reduce the chance of IPV. Consistent with this result, Afghan women have one of the lowest rates of education in the world for women and the highest levels of IPV (Rammohan & Johar, 2009; Rizkianti et al., 2020; Sougou et al., 2020). The literacy rates in Afghanistan due to the spread of the Taliban regime and devastating wars are among the lowest in the world at approximately 45 percent for men and 17 percent for women (Alexander &Thelwell 2020). Noury & Speciale (2016) show that women exposed to more years of Taliban rule from 1996 to 2001 had reduced education. Afghanistan has one of the highest prevalence of IPV in the world (Coll et al. 2020). Previous work shows the expected negative relationship between education and experience of IPV in Afghanistan with the same survey data used in this paper’s analysis (Akbary 2020).
Child marriage and education are strongly linked, as women who stay in school longer typically get married later (Nguyen and Wodon, 2014). Child marriage tends to decrease women’s decision-making power and increase IPV, potentially in part due to their increased education, as discussed above (Kidman, 2017; Ahinkorah, 2021). Child marriage was relatively common during the first Taliban rule pre-2001; however, the attitudes of parents in Afghanistan when the Taliban lost control tended toward women staying in school and marrying later (Blum et al., 2019). Noury & Speciale (2016) find that marriage age decreased for women exposed more to Taliban rule.
Data and Descriptive Statistics
This paper links data from three sources: the 2015 Afghanistan Demographic and Health Survey (DHS), the Foundation for Defense of Democracies (FDD) Long War Journal, and the Global Terrorism Database to construct a data set of over 2,000 women’s responses questions on decision making and IPV. The 2015 Afghanistan DHS is a large-scale household survey that asks women a myriad of questions, including their level of decision-making power in their household and their experience with IPV. Included in the DHS are 21,324 women aged 15–49 spanning all 34 provinces in Afghanistan who were married between 1975 and 2015. Important to the use of provincial-level variation, the survey is representative at the provincial level. We use approximately 10% of the sample for women married in the early 2000s that we can pair with data on Taliban control after the first US invasion in 2002. We use this data set to measure outcomes related to women’s decision-making power, experience of IPV, and marriage age. Summary statistics are presented in Table 1 below.
The DHS survey has information on women’s participation in household decisions regarding health care, spending, and travel, as well as their experience of IPV, all measured at the time of the survey in 2015. Approximately 60% of the 2,280 women married between 2002 and 2012 included in our sample have decision-making power for one of the four decisions, and almost 50% have experienced some type of violence from their husband/partner. More specifically, slightly more than half of the women have some level of decision-making power in visits to family or relatives (51%), and less than half have some level of decision-making power in their health care (48%) or large household purchases (42%). Interestingly, however, more than half of these women have some level of decision-making power in what is done with their husband’s income (56%).
Regarding IPV, almost half of the women in our sample experienced less severe violence in the last year (47%). In addition, approximately 16% of women have experienced severe violence, and approximately 9% have experienced sexual violence in the last year within their household. Last, the average age of cohabitation for the women in our analysis is 18 years old. We assume that women’s age of cohabitation and marriage age are synonymous, as it is customary for women to move in with their partner at marriage (Afghan Culture, n.d.).
To mirror the work of Anukriti et al. (2022), we also test for changes in the number of children a woman has ever given birth, with an average in the sample of approximately 4.5. It is worth noting that these women are of child-bearing age (15–49) and may have more children in the future. Anukriti et al. (2022) suggest that conflict arises as women increase their use of contraceptives and lower their fertility when governments are more favorable to women. Contraception use is low in Afghanistan, with only 21% of women reporting using any form of contraceptive, including traditional, nonmodern methods such as withdrawal. In the sample, 30% of the men wanted more children than the family had. Finally, Anukriti et al. (2022) rule out changing attitudes about IPV by examining questions if a husband beating his wife is ever justified. The survey asks if beatings are justified if a wife does one of five actions: goes out without telling her husband, neglects children, argues with her husband, or burns food. Most women (85%) believe that at least one of these actions justifies beating, and women believe that on average, 2.5 of these five actions justify beating. Going out without telling her husband is the most commonly justified reason at 70%, and burning food is the least common at 20%.
Additionally, we use the provincial average of the outcome variable for elder women married before the Islamic rule of Afghanistan as a proxy for norms within a province related to women’s decisions, IPV experience, and marriage age before Taliban rule. We use 1992 as the cutoff point, as the Democratic Republic of Afghanistan (1978–1992) was a period of more equal rights for Afghan women but was overthrown in 1992 and transitioned to the Islamic State of Afghanistan (Nunan, 2016). The goal of the leaders of the Islamic State was to Islamize society, which included abolishing women’s freedoms.
We see in the summary statistics (Table 1) that younger women have less decision-making power, with approximately 10% fewer women from the newer generation having some decision-making power in their own health care, large household purchases, and travel. The greatest change is in travel, with only 51% of women from the newer generation having some decision-making power compared to 64% of the elder generation. The result is consistent with the Taliban’s focus on limiting women’s travel during their rule.
In terms of IPV, the difference between generations is slightly smaller, and in the opposite direction, with 5% fewer women from the newer generation reporting less severe or severe IPV. Sexual violence stayed relatively constant, from 8% of the older generation experiencing sexual violence to 9% of the newer generation experiencing such violence.
To measure our key variable of Taliban control at women’s time of marriage, we use data from the Foundation for Defense of Democracies (FDD). These visualizations identify five levels of Taliban control (insignificant, minimal, moderate, high, and full) for each Afghan province in the years 2000, 2002, 2009 & 2012. The data for these visualizations were collected by an FDD senior fellow, Bill Roggio, based on his discussions with leaders in the field and his independent research (B. Rodrigo, personal communication, February 9, 2022). We are unable to include the year 2000 in our analysis, as it too closely mirrors the effect variable “high Taliban control” (as 32 out of 34 of the provinces were fully controlled by the Taliban at that time). Thus, the variable “Taliban Control” in our analysis indicates the varying levels of Taliban control across Afghan provinces in 2002, 2009 and 2012. It is important to note that the lack of data for Taliban control outside of these three years significantly limits our sample, as we are required to restrict our analysis only to women married in 2002, 2009, and 2012. This decreases our sample size from 21,324 women to 2,280.
The Long War Journal visualizations show that Taliban control was greatest in early 2000, with the Taliban controlling 32 out of 34 provinces. Following 9/11, in early 2002, the U.S. had the largest influence in the country, and the Taliban lost control of all provincial capitals. Later, in early 2009, the Taliban expanded into the south, east, and north, contesting seven provincial capitals. In response, by 2012, the U.S. doubled its military presence in the country and reclaimed all provincial capitals. In the descriptive statistics, we can see that approximately 14% of women in our sample were married under high Taliban control, 28% under moderate control, 43% under minimal control and 14% under insignificant control.
A graphical descriptive look at the relationship between Taliban power and outcomes shows that where Taliban had insignificant control, women had more decision making (see Fig. 1A) and were less likely to experience IPV (Fig. 1B), although there were negligible differences in age of cohabitation. For example, in Fig. 1A, 63% of women who married in 2002 in a province with insignificant Taliban control had some level of decision-making power, compared to the mere 31% of women who married in 2002 in a region with full Taliban control. This trend holds true for the years 2009 & 2012 as well for decision-making power. We do not show such a clear relationship between Taliban control at women’s time of marriage and women’s age of cohabitation, as seen in Fig. 1C.
Variation in Taliban control could potentially occur at times of varying degrees of violence and conflict. Next, we measure conflict in Afghan regions with the Global Terrorism Database, which is an open-source database with information on terrorist attacks around the world. The database includes terrorist attacks in Afghanistan and lists their date and the province in which they took place. We use this database to measure the number of terrorist attacks in a woman’s province in the year she was married to proxy conflict at the time of her marriage. The mean number of terrorist attacks in a province in a year was 20, but the median was only 8, signifying that there may be some outliers in the data.
We use four additional control variables from the DHS that potentially influence women’s decision-making power and experience of IPV: wealth, age, education, and rurality. The wealth quintile ranges from the lowest household wealth at 1 to the highest at 5, and the average is approximately 3. The effect of the respondent’s current age was examined, with the average being 26 years old. Education was also a factor, with the mean education being 1.58 years, the minimum being 0 years and the max being 14 years. Last, each household is defined as a rural or urban residence. Approximately 77% of households were classified as rural.
Table 1: Summary Statistics
Econometric Model
The econometric model is written out in equation #1 below. The unit of observation is a woman, i, living in a household, h, in a province, j, and married in year t. Recall we have women only married in 2002, 2009 or 2012. We have multiple outcomes corresponding to decisions of health care, purchases, travel, and spending of the husband’s income (decideHlthCare, decidePurchases, decideTravel, & decideHusbIncome). These outcomes are all measured in 2015, the time of the DHS survey. We also measure if she has a role in any four of those decisions (anyDecision). We have three measures of IPV (severeViolence, lessViolence, & sexualViolence) and another if she experiences any form of IPV (anyIPV). We analyze her age at first cohabitation (ageCohab), a proxy for marriage. We use a linear probability model with 1 indicating that she has some input in the decision (or experiences some IPV) and with 0 indicating she does not, while age at first cohabitation.
We also test for changes in fertility, contraception, fertility preferences and views on IPV, which are potential mechanisms for the increase in IPV. We measure fertility with (births) the total number of births the respondent has had, contraception (Contraception) a binary variable if the respondent uses contraception, and a binary indicator if the husband wants more children, then the family has (HusbandMoreKids). In terms of justification of IPV, respondents are asked for five scenarios if a beating is justified: if she goes out without telling her husband (Goesout), neglects children (NeglectsKids), argues with her husband (Argues), refuses to have sex (RefusesSex), or burns food (BurnsFood). We estimate regressions for each individually, the total number of the five she says are justified, and if any of the five are justified. All estimations for mechanisms are also linear probability models, except in the case of births, which is continuous.
The first predictor variable (ElderY) measures norms using the average of the outcome variable of interest in 2015 for all elder women married before 1992. As noted, these women were married before the Islamic takeover of Afghanistan. The Elderly variable varies only at the province level, denoted by the subscript j.
Taliban control is measured at the time (t) of a woman’s marriage (either 2002, 2009 or 2012), which is defined as either Insignificant, Minimal, Moderate, or High. This variable represents the level of Taliban control in a woman’s region when she was married and is thus measured at the province level, denoted by the subscript j and time t. The level of High control is the omitted comparison group. We also control for conflict at the time of a woman’s marriage using attacks, which indicates the number of terrorist attacks in province (j) in the year of her marriage (t = 2002, 2009 or 2012).
There are six more control variables included in the regressions: marriage year, education, age, wealth, rurality, and province. We control for national time variant trends with indicators for women married in 2009 and 2012 (2009 & 2012), with the year 2002 omitted for comparison in the econometric model, as the model is comparing results from the other two years to 2002. The variable yrsEducation represents the years of education attained by a woman and is measured at the individual level. The woman’s age (age) at the time of survey in 2015 controls for trends in the population. The variable wealth represents the wealth quintile of a woman’s household, which ranges from the lowest household wealth at a coefficient of 1 to the highest at a coefficient of 5. This variable is measured at the individual level, denoted by the subscript h. The variable rural represents whether a household is in a rural or urban setting, with a coefficient of 1 indicating rural and a coefficient of 0 indicating urban.
The results suggest two main findings for women married from 2002 to 2012. Women married in provinces during times of higher Taliban control reported having less decision-making power and being less likely to experience IPV when surveyed in 2015. As we will discuss below, the results on IPV run counter to theory that being married during the Taliban should reduce women’s empowerment and increase IPV. We present two hypotheses for this result that it could be retaliation or measurement error and find that retaliation for misaligned fertility preferences is more likely. Second, there is a strong positive relationship between decision-making power and experience of IPV for the elder generation within a province (those married before 1992) and the newer generation (those married during the early 2000s).
We find weak statistical effects of the Taliban control at the time of marriage on the age of first cohabitation (column 6 in Table 2). Those women married with minimal Taliban control were roughly half a year older, but the coefficient is statistically significant at the 10% level. Previous works discussed above find positive and negative impacts on marriage age, which is another potential channel for the Taliban to influence decision making and IPV. The lack of strong effects on marriage age helps reduce the worries around the endogeneity of the types of people who get married under the Taliban rule being a potential omitted variable.
We find that women married under lower levels of Taliban control (i.e., insignificant or minimal) had more decision-making power (see Table 2), particularly for deciding on their own health care, travel, and large household purchases. Women married in moderately, minimally, and insignificantly Taliban-controlled provinces at the time of marriage are 6.8, 9.9, and 18.8 percentage points more likely to have decision-making power on travel compared to those married under high Taliban control (see column 3 of Table 2). These results are statistically significant at the 10%, 5% and 1% levels, respectively. Additionally, the percentage of women with decision-making power in their own health care and large household purchases are 8.7 and 7.7 percentage points higher for those married under insignificant Taliban rule compared to those married under high Taliban rule (see columns 2 and 4 of Table 2). Both results are statistically significant at the 10% level.
We also reach a peculiar finding that women married in regions with insignificant Taliban control experienced higher levels of IPV in 2015 than those married in regions with high Taliban control (see Table 3). These results could be due to retaliation for changing norms or if Taliban control influences reporting could be driven by measurement error. The higher IPV results for insignificant Taliban control are for severe and sexual violence with no difference in less severe types of IPV. For severe violence regression (column 2 in Table 3), women married in regions with minimal or insignificant control were 7.6 and 7.9 percentage points more likely to experience severe IPV compared to those married during high Taliban control. These results are statistically significant at the 1% and 5% levels, respectively. Additionally, women married in regions with insignificant Taliban control were 4.4 percentage points more likely to experience sexual violence than those married in regions with high control (column 4 in Table 3). This result is statistically significant at the 10% level.
| Dependent variable: | ||||||
|---|---|---|---|---|---|---|
| Any Health Care | Travel | Purchases | HusbIncome | Age Cohab | ||
| (1) | (2) | (3) | (4) | (5) | (6) | |
| ElderDecision | 0.760*** | 0.891*** | 0.768*** | 0.848*** | 0.853*** | 0.719*** |
| (0.047) | (0.048) | (0.049) | (0.042) | (0.048) | (0.074) | |
| Insignificant Taliban Control | 0.077 | 0.087* | 0.188*** | 0.077* | 0.069 | 0.447 |
| - | (0.050) | (0.048) | (0.050) | (0.046) | (0.045) | (0.359) |
| Minimal Taliban Control | -0.051 | 0.010 | 0.099** | 0.036 | -0.046 | 0.524* |
| (0.041) | (0.039) | (0.040) | (0.037) | (0.037) | (0.300) | |
| Moderate Taliban Control | 0.017 | 0.051 | 0.068* | 0.051 | -0.009 | 0.101 |
| (0.036) | (0.035) | (0.036) | (0.035) | (0.034) | (0.274) | |
| attacks | -0.00002 | 0.0003 | 0.0003 | 0.0003 | -0.001* | -0.005* |
| (0.0004) | (0.0004) | (0.0004) | (0.0004) | (0.0003) | (0.003) | |
| age | 0.001 | 0.001 | 0.002 | 0.002 | -0.004 | |
| (0.003) | (0.003) | (0.003) | (0.003) | (0.003) | ||
| Married 2009 | -0.014 | 0.003 | 0.001 | -0.035 | -0.056* | 0.966*** |
| (0.035) | (0.034) | (0.035) | (0.034) | (0.032) | (0.225) | |
| Married 2012 | -0.020 | -0.044 | -0.088** | -0.083** | -0.050 | 1.459*** |
| (0.038) | (0.038) | (0.038) | (0.037) | (0.035) | (0.224) | |
| wealth | 0.009 | 0.009 | 0.002 | 0.011 | 0.023*** | 0.073 |
| (0.009) | (0.009) | (0.009) | (0.009) | (0.009) | (0.072) | |
| yrsEducation | 0.017*** | 0.016*** | 0.013*** | 0.013*** | 0.019*** | -0.027 |
| (0.003) | (0.003) | (0.003) | (0.003) | (0.003) | (0.023) | |
| Rural | 0.006 | 0.013 | -0.046* | -0.037 | 0.008 | -0.136 |
| (0.029) | (0.028) | (0.028) | (0.028) | (0.026) | (0.221) | |
| Constant | -0.167 | -0.152 | -0.095 | -0.090 | 0.051 | 5.042*** |
| (0.107) | (0.105) | (0.105) | (0.103) | (0.098) | (1.346) | |
| Observations | 2,200 | 2,200 | 2,200 | 2,199 | 2,187 | 2,228 |
| R2 | 0.182 | 0.209 | 0.216 | 0.226 | 0.216 | 0.065 |
| Adjusted R2 | 0.178 | 0.205 | 0.212 | 0.222 | 0.212 | 0.061 |
| Note: | *p**p***p < 0.01 | |||||
| Dependent variable: | ||||
|---|---|---|---|---|
| Any IPV | Severe Violence | Less Severe Violence | Sexual Violence | |
| (1) | (2) | (3) | (4) | |
| ElderIPV | 0.848*** | 0.734*** | 0.844*** | 0.964*** |
| (0.039) | (0.046) | (0.039) | (0.044) | |
| Insignificant Taliban Control | 0.010 | 0.079** | 0.017 | 0.044* |
| (0.044) | (0.034) | (0.044) | (0.026) | |
| Minimal Taliban Control | 0.059 | 0.076*** | 0.061 | 0.029 |
| (0.037) | (0.029) | (0.037) | (0.022) | |
| Moderate Taliban Control | 0.013 | 0.026 | 0.012 | 0.027 |
| (0.035) | (0.027) | (0.035) | (0.021) | |
| attacks | 0.001** | 0.0003 | 0.001** | 0.001** |
| (0.0004) | (0.0003) | (0.0004) | (0.0002) | |
| age | 0.004 | 0.004* | 0.004 | 0.001 |
| (0.003) | (0.002) | (0.003) | (0.002) | |
| Married 2009 | -0.030 | -0.009 | -0.026 | 0.002 |
| (0.033) | (0.026) | (0.033) | (0.020) | |
| Married 2012 | -0.067* | -0.030 | -0.067* | -0.008 |
| (0.037) | (0.029) | (0.037) | (0.022) | |
| wealth | -0.004 | -0.002 | -0.003 | -0.002 |
| (0.009) | (0.007) | (0.009) | (0.005) | |
| yrsEducation | -0.013*** | -0.002 | -0.013*** | -0.0004 |
| (0.003) | (0.002) | (0.003) | (0.002) | |
| Rural | 0.085*** | 0.021 | 0.090*** | 0.004 |
| (0.028) | (0.022) | (0.028) | (0.016) | |
| Constant | -0.113 | -0.144* | -0.133 | -0.037 |
| (0.104) | (0.081) | (0.105) | (0.061) | |
| Observations | 2,227 | 2,226 | 2,227 | 2,225 |
| R2 | 0.229 | 0.119 | 0.224 | 0.189 |
| Adjusted R2 | 0.225 | 0.115 | 0.220 | 0.185 |
| Note: | *p**p***p < 0.01 | |||
There is a strong relationship between the provincial average for elder women married before 1992 and all outcomes of interest (decision making, IPV, and marriage age). This suggests relatively consistent norms across generations. We find that a 1 percentage point increase in decision making of the elder generation is associated with between .72 and .89 higher degree of the generation married from 2002 to 2012 probability of having a role in that decision with all results statistically significant at the 1% level. The results are stronger for IPV, with an additional one-point increase in the elder likelihood of experiencing IPV being associated with increases of 0.73 to 0.96 likelihood of the same IPV measure in the 2002 to 2012 married cohort.
Women married more recently have less decision-making power, but women’s age at first cohabitation has increased. Women married in 2012 had less decision-making power, were less likely to experience IPV and were married later than those married in 2002. Specifically, those married in 2002 had 8.8 and 8.3 percentage points less decision-making power in travel and large purchases than those married in 2012 (statistically significant at the 5% level) and were approximately 6.7 percentage points less likely to experience IPV (statistically significant at the 10% level). Additionally, those married in 2002 had 5.6 percentage points less decision-making power in how to spend their husband’s income than those married in 2012. Last, compared to women married in 2002, women married approximately 1 and 1.5 years later in 2009 and 2012, respectively.
Terrorist conflict in the year of marriage seems to reduce women’s empowerment. There appears to be some impact of the number of terrorist attacks in the year of marriage on IPV and reducing marriage age but not on decision making. The standard deviation in attacks in a province in the year a woman was married was 33. For less severe and sexual violence, a coefficient of 0.001 suggests that one standard deviation in attacks would increase IPV by approximately 3.3 percentage points, similar for less severe and sexual violence. To put this coefficient in context, approximately 50% of women experience any violence and 9% experienced sexual violence, so the magnitude is relatively small for less severe violence and higher compared to sexual violence. The coefficient on marriage age is -0.005, suggesting that a one standard deviation increase in attacks is associated with marriage approximately 0.17 years earlier.
Our findings also support the existing literature on education increasing women’s decision-making power and lowering IPV. In our models, an additional year of education leads to a 1.3 to 1.9 percentage point increase in the probability a woman can make any one of the four major decisions, can decide on travel or large purchases, can decide on her own health care, and can decide on how to spend her husband’s income, respectively. Similarly, an additional year of education leads to a 1.3-percentage point decrease in experiencing any IPV or less severe violence. Interestingly, there does not appear to be a statistically significant relationship between years of education and age of cohabitation.
We find that women married when the Taliban have insignificant control have substantially fewer children and are more likely to have husbands who want more children than they have. This result is consistent with the mechanism of falling fertility as the Taliban lose power, leading to higher incidences of IPV. Specifically, husbands of women married under insignificant or minimal Taliban control are 15 and 11 percentage points more likely to want more children than the family has than those in the omitted group, married under full Taliban control. The women married in insignificant and minimal Taliban also have 0.66 and 0.35 fewer children than the omitted group. We do not find effects on contraceptives, although use is very low already in Afghanistan.
| Dependent variable: | ||||
|---|---|---|---|---|
| Husband Wants More Kids | # of Kids | Currently Using Contraception | ||
| (1) | (2) | (3) | ||
| Elderhusmore | 0.604*** | 0.152*** | 0.788*** | |
| (0.079) | (0.027) | (0.068) | ||
| Insignificant Taliban Control | 0.150*** | -0.659*** | -0.019 | |
| (0.048) | (0.108) | (0.036) | ||
| Minimal Taliban Control | 0.114*** | -0.345*** | -0.018 | |
| (0.040) | (0.092) | (0.031) | ||
| Moderate Taliban Control | 0.020 | -0.141 | -0.039 | |
| (0.036) | (0.086) | (0.028) | ||
| attacks | -0.0004 | -0.0003 | -0.001** | |
| (0.0004) | (0.001) | (0.0003) | ||
| age | 0.002 | -0.006 | -0.001 | |
| (0.003) | (0.007) | (0.002) | ||
| Married 2009 | 0.044 | -2.636*** | -0.076*** | |
| (0.035) | (0.083) | (0.027) | ||
| Married 2012 | 0.030 | -3.725*** | -0.080*** | |
| (0.038) | (0.091) | (0.031) | ||
| wealth | 0.023** | 0.017 | 0.038*** | |
| (0.009) | (0.023) | (0.007) | ||
| yrsEducation | 0.007** | -0.015** | 0.007*** | |
| (0.003) | (0.007) | (0.002) | ||
| Rural | -0.036 | -0.005 | -0.014 | |
| (0.028) | (0.069) | (0.023) | ||
| Constant | -0.059 | 4.308*** | -0.013 | |
| (0.105) | (0.335) | (0.084) | ||
| Observations | 2,182 | 2,228 | 2,228 | |
| R2 | 0.080 | 0.659 | 0.107 | |
| Adjusted R2 | 0.075 | 0.657 | 0.103 | |
| Note: | *p**p***p < 0.01 | |||
We find little evidence of Taliban control at the time of marriage influencing beliefs on when IPV is justified. If anything, women married when the Taliban lacked control are more likely to believe that IPV is justified. This runs counter to the prediction that the Taliban would decrease the probability that IPV was justified due to changing norms. It is worth noting that we find null effects in three of the five scenarios. Women married when the Taliban had insignificant control were nine and thirteen percentage points more likely to believe a husband is justified in beating his wife if she refuses sex or burns food.
| Dependent variable: | ||||||
|---|---|---|---|---|---|---|
| Any of 5 | Goes Out | Neglects Children | Argues | Refuse Sex | Burns Food | |
| (1) | (2) | (3) | (4) | (5) | (6) | |
| Elderjust | 0.715*** | 0.735*** | 1.050*** | 0.731*** | 0.820*** | 0.325*** |
| (0.056) | (0.055) | (0.072) | (0.071) | (0.058) | (0.048) | |
| Insignificant Taliban Control | 0.023 | -0.051 | 0.069 | 0.035 | 0.089* | 0.130*** |
| (0.033) | (0.044) | (0.051) | (0.049) | (0.050) | (0.043) | |
| Minimal Taliban Control | 0.021 | -0.022 | 0.001 | 0.036 | 0.076* | 0.006 |
| (0.028) | (0.038) | (0.043) | (0.043) | (0.042) | (0.036) | |
| Moderate Taliban Control | 0.028 | 0.008 | 0.006 | 0.023 | 0.039 | 0.011 |
| (0.026) | (0.035) | (0.039) | (0.039) | (0.039) | (0.034) | |
| attacks | -0.0001 | 0.00001 | -0.0002 | 0.00002 | 0.0003 | 0.0004 |
| (0.0003) | (0.0004) | (0.0004) | (0.0004) | (0.0004) | (0.0004) | |
| age | 0.002 | 0.002 | -0.002 | -0.001 | 0.004 | 0.002 |
| (0.002) | (0.003) | (0.003) | (0.003) | (0.003) | (0.002) | |
| Married 2009 | 0.048* | 0.009 | -0.012 | 0.015 | 0.058 | 0.066** |
| (0.025) | (0.033) | (0.037) | (0.037) | (0.037) | (0.031) | |
| Married 2012 | 0.037 | 0.048 | 0.030 | 0.0001 | 0.048 | 0.058* |
| (0.028) | (0.037) | (0.041) | (0.040) | (0.041) | (0.034) | |
| wealth | -0.011 | -0.0004 | 0.004 | -0.014 | -0.027*** | -0.026*** |
| (0.007) | (0.009) | (0.010) | (0.010) | (0.010) | (0.009) | |
| yrsEducation | -0.009*** | -0.014*** | -0.010*** | -0.015*** | -0.011*** | -0.009*** |
| (0.002) | (0.003) | (0.003) | (0.003) | (0.003) | (0.003) | |
| Rural | -0.0002 | 0.018 | -0.003 | -0.004 | -0.038 | -0.029 |
| (0.021) | (0.028) | (0.031) | (0.030) | (0.031) | (0.026) | |
| Constant | 0.207** | 0.170 | 0.022 | 0.223* | 0.011 | 0.057 |
| (0.088) | (0.109) | (0.117) | (0.119) | (0.114) | (0.097) | |
| Observations | 2,228 | 2,109 | 2,080 | 2,101 | 1,976 | 2,019 |
| R2 | 0.084 | 0.097 | 0.111 | 0.065 | 0.125 | 0.057 |
| Adjusted R2 | 0.079 | 0.093 | 0.106 | 0.060 | 0.120 | 0.052 |
| Note: | *p**p***p < 0.01 | |||||
The Taliban rule in Afghanistan has significantly contributed to IPV and critically reduced women’s decision-making power in the household. This research sheds light on the plight of women married during different levels of Taliban rule between 2002 and 2012. Specifically, we look at the impact on decision making, IPV and marriage age. For women married from 2002 to 2012, in regions with insignificant Taliban control, they experienced substantially more decision-making power and more IPV, as portrayed in a 2015 nationally representative survey. An additional analysis shows that reductions in fertility may be causing strife in the households, as regions with less Taliban control also have fewer children and are more likely to have husbands who want more children than they have. We also provide evidence counter to potential changes in reporting IPV, as women married when the Taliban controlled their province are either as or less likely to believe beatings are justified in five scenarios.
Our second core result is finding a very strong relationship between decision-making power and experience of IPV for the elder generation within a province (those married before 1992) and the newer generation (those married 2002–2012). This result suggests that norms transmit overtime and is suggestive that Taliban’s influence on norms at the time of marriage may have long-lasting effects.
One limitation is that there was only one survey year (2015) and only a few years of variation in Taliban control. A further concern could potentially be omitted variable bias related to who gets married when the Taliban takes control. We provide evidence to counter this concern specifically, that marriage age does not seem to be changing much based control during this time. We also find a counterintuitive result that women married when the Taliban had insignificant control are more likely to experience severe or sexual IPV. We suggest that this could be retaliation for growing women’s empowerment. On the other hand, there could be measurement error where women are less likely to report IPV where Taliban norms linger. Finally, it is also possible that the variation in the Taliban’s power between 2002 and 2012 may have different effects on the more recent Taliban takeover. This research will assist in providing targeted help for women who are married during the worse parts of Taliban control in terms of empowerment. It adds to the growing literature showing the long-term negative impacts of oppressive regimes on women. It also shows potential long-term negative effects of governments that are not supportive of women’s rights through the strong connections between cohorts on empowerment measures. Considering this, it is important that we understand how to improve women’s empowerment in the next post-Taliban regime or similar situations around the globe.
Conflict of Interest Statement
We have no conflicts of interest to disclose.
- Akbary MF, Ariyo T, Jiang Q. Sociocultural Determinants of Attitudes Toward Domestic Violence Among Women and Men in Afghanistan: Evidence from Afghanistan Demographic and Health Survey 2015. Journal of Interpersonal Violence. December 2020. doi:10.1177/0886260520980400
- Ahinkorah, B. O., Onayemi, O. M., Seidu, A. A., Awopegba, O. E., & Ajayi, A. I. (2021). Association between girl-child marriage and intimate partner violence in Sub-Saharan Africa: insights from a multicountry analysis of demographic and health surveys. Journal of interpersonal violence, 08862605211005139.
- Alexander, L., & Thelwell, K. (2020). Literacy rates in Afghanistan. The Borgen Project. Retrieved April 4, 2022, from https://borgenproject.org/tag/literacy-rates-in-afghanistan/#:~:text=Unfortunately%2C%20due%20to%20the%20spread,and%2017%20percent%20for%20women.
- Anukriti, L., Erten, Bilge & Mukherjee, Priya, 2022. "Women’s Political Representation and Intimate Partner Violence," Policy Research Working Paper Series 10113, The World Bank.
- Ashraf, N., E. Field, and J. Lee (2014): “Household bargaining and excess fertility: an experimental study in Zambia,” American Economic Review, 104, 2210–37
- Blum, R. W., Li, M., Pasha, O., Rao, C., & Natiq, K. (2019). Coming of Age in the Shadow of the Taliban: Education, Child Marriage, and the Future of Afghanistan from the Perspectives of Adolescents and Their Parents. Journal of Adolescent Health, 64(3), 370–375. https://doi.org/10.1016/j.jadohealth.2018.09.014
- Burnet, J. E. (2011): “Women have found respect: Gender quotas, symbolic representation, and female empowerment in Rwanda,” Politics & Gender, 7, 303–334.
- Coll, C. V. N., Ewerling, F., García-Moreno, C., Hellwig, F., & Barros, A. J. D. (2020). Intimate partner violence in 46 low-income and middle-income countries: An appraisal of the most vulnerable groups of women using national health surveys. BMJ Global Health, 5(1), e002208. https://doi.org/10.1136/bmjgh-2019-002208
- De Silva de Alwis, R. (2021). Spike in child marriages may be ‘the most disturbing fallout of the Taliban takeover.’ University of Pennsylvania Carey Law School. https://www.law.upenn.edu/live/news/14262-spike-in-child-marriages-may-be-the-most
- Donald, A., Doss, C., Goldstein, M., & Gupta, S. (2023). Sharing responsibility through joint decision-making and implications for intimate-partner violence: evidence from 12 Sub-Saharan African Countries. Review of Economics of the Household, 1–32.
- Eswaran, M. and N. Malhotra (2011): “Domestic violence and women’s autonomy in developing countries: theory and evidence,” Canadian Journal of Economics/Revue canadienne d’´economique, 44, 1222–1263.
- Girl’s Not brides (n.d) https://www.girlsnotbrides.org/learning-resources/child-marriage-and-education/
- Human Rights Watch (2022). “Afghanistan: Taliban deprive women of livelihoods, identity.” Retrieved from https://www.hrw.org/news/2022/01/18/afghanistan-taliban-deprive-women-livelihoods-identity#
- Jain, S., & Kurz, K. (2007). New Insights on Preventing Child Marriage. 60.
- Jejeebhoy, S. J., & Sathar, Z. A. (2001). Women’s autonomy in India and Pakistan: The Influence of Religion and Region. Population & Development Review, 27(4), 687–712. https://doi.org/10.1111/j.1728-4457.2001.00687.x
- Kidman, R. (2017). Child marriage and intimate partner violence: a comparative study of 34 countries. International journal of epidemiology, 46(2), 662–675.
- Kohno, A., Techasrivichien, T., Suguimoto, S. P., Dahlui, M., Nik Farid, N. D., & Nakayama, T. (2020). Investigation of the key factors that influence the girls to enter child marriage: A meta-synthesis of qualitative evidence. PloS one, 15(7), e0235959. https://doi.org/10.1371/journal.pone.0235959
- Koenig, et al. (2006). Individual and contextual determinants of domestic violence in North India. American Journal of Public Health 96(1) https://mail.google.com/mail/u/0/#inbox/FMfcgzGmvTrlLWCwqrqHdTmKZdBLCmQV?compose=GTvVlcRwPkPJWnZfDMmlwMFXHMbDpkzptXSVjxvdqFgTsMTWHdKzkTNhPslmfbjPhTSZNTPNLLvKq&projector=1&messagePartId=0.1
- Kuipers, N. (2020): “The E↵ect of Electing Female Candidates on Attitudes toward Intimate Partner Violence,” The Journal of Politics, 82, 1590–1595.
- Maity, B. and S. Shukla “The Long-Term Effect of Exposure to the Taliban Regime on Women’s Age at Marriage in Afghanistan” unpublished mimeo.
- Mannell, J., Grewal, G., Ahmad, L., & Ahmad, A. (2021). A Qualitative Study of Women's Lived Experiences of Conflict and Domestic Violence in Afghanistan. Violence against women, 27(11), 1862–1878. https://doi.org/10.1177/1077801220935191
- Mannell, J. (2021). Afghan women's lives are now in danger from the Taliban – but they have always faced male violence. The Conversation. Retrieved from https://theconversation.com/afghan-womens-lives-are-now-in-danger-from-the-taliban-but-they-have-always-faced-male-violence-166768
- Mavisakalyan, A., & Rammohan, A. (2021). Female autonomy in household decision-making and intimate partner violence: Evidence from Pakistan. Review of Economics of the Household, 19, 255–280.
- Monir G. (2022). Taliban are revoking Afghan women's hard-won rights: DW: 02.01.2022. DW.COM. Retrieved April 4, 2022, from https://www.dw.com/en/taliban-are-revoking-afghan-womens-hard-won-rights/a-60283590
- McAuliffe, M. (2022). Struggling to survive: Gender, displacement, and migration in Taliban-controlled Afghanistan. Center for Strategic and International Studies. Retrieved from https://www.csis.org/analysis/struggling-survive-gender-displacement-and-migration-taliban-controlled-afghanistan
- Nguyen, M. C., & Wodon, Q. (2014). Impact of child marriage on literacy and education attainment in Africa. Washington, DC: UNICEF and UNESCO Statistics.
- Noury, A. G., & Speciale, B. (2016). Social constraints and women's education: Evidence from Afghanistan under radical religious rule. Journal of comparative economics, 44(4), 821–841.
- Raj, A., Salazar, M., Jackson, E.C. et al. Students and brides: a qualitative analysis of the relationship between girls’ education and early marriage in Ethiopia and India. BMC Public Health 19, 19 (2019). https://doi.org/10.1186/s12889-018-6340-6
- Rammohan, A., & Johar, M. (2009). The Determinants of Married Women’s Autonomy in Indonesia. Feminist Economics, 15(4), 31–55. https://doi.org/10.1080/13545700903153989
- UNFPA Afghanistan. Child marriage. (2016). Retrieved March 14, 2022, from https://afghanistan.unfpa.org/en/node/15233
- The authors would like to thank the Towson University Committee on Undergraduate Research and Creative Inquiry for financial support.
- The DHS asks who makes each decision, we code women as having decision making power if the respondent reports either she decides alone or with her partner the decision of interest.
No competing interests reported.