Is Economic Empowerment a Protective Factor Against Intimate Partner Violence? Evidence from Turkey

Abstract

This paper analyzes the relationship between women’s economic empowerment and the incidence of intimate partner violence (IPV) using data from the National Survey on Domestic Violence against Women in Turkey (2008, 2014). Two hypotheses are tested: (i) women’s economic independence reduces the risk of partner violence as suggested by household bargaining models; (ii) women’s vulnerability to IPV increases when their economic situation improves relative to their partner’s as suggested by a male backlash model. Women’s employment has a positive effect on the exposure to IPV but it is not statistically significant after controlling for the simultaneous causality between employment status and IPV. Earning more income than their partners, on the other hand, lowers the risk of IPV by 9.3%, providing evidence for the household bargaining model. The protective effect of income differs according to class positions of women. Earning more income than their partners lowers the risk of physical (7.5%) and sexual violence (6.4%) for women in poor households while it lowers the risk of psychological violence (11.5%) for women in medium-wealth and rich households.

Résumé

Cet article analyse le lien entre l'autonomisation économique des femmes et l'incidence de la violence entre partenaires intimes (VPI) à l'aide des données de l'Enquête nationale sur la violence domestique à l'égard des femmes en Turquie (2008, 2014). L’on teste deux hypothèses: (i) l'indépendance économique des femmes réduit le risque de violence conjugale comme le suggèrent les modèles de négociation des ménages; (ii) la vulnérabilité des femmes aux violences exercées par un partenaire intime augmente lorsque leur situation économique s'améliore par rapport à celle de leur partenaire, comme le suggère un modèle de réaction masculine. Pour les femmes, le fait d’avoir un emploi a un effet positif sur leur exposition aux violences exercées par un partenaire intime, mais cet effet n'est pas statistiquement significatif lorsque l’on prend en compte la causalité simultanée entre le statut de l'emploi et la VPI. En revanche, le fait de gagner plus que leur partenaire réduit le risque de VPI de 9,3% pour les femmes, ce qui appuie le modèle de négociation des ménages. L'effet protecteur du revenu diffère selon la classe sociale des femmes. Pour les femmes issues des ménages pauvres, le fait de gagner plus que leurs partenaires réduit le risque de violence physique (7,5%) et sexuelle (6,4%), tandis que cela diminue le risque de violence psychologique (11,5%) pour les femmes issues des ménages moyens et riches.

Appendix

List of Variables

Outcome Variables

See Tables 7, 8, 9, 10, and 11.

Table 7 Diagnostics for weak instruments: first stage results

Table 8 Instrumental variable estimations, women in poor households

Table 9 Instrumental variable estimations, women in medium-wealth and rich households

Table 10 Instrumental variable estimations with the birth cohort instrument, 2008

Table 11 Job status and financial control behavior among working women with rural backgrounds

Intimate Partner Violence (IPV): A dummy variable equal to one if the respondent experienced at least one type of violence, physical, sexual or psychological.

Physical violence: A dummy variable equal to one if the respondent experienced at least one of the 6 physical violence acts from her spouse:

• Slap or throw an object that would hurt

• Push, shove, or pull hair

• Hit with his fist or in a way that hurts

• Kick, pull on the ground, or beat

• Choke or burn

• Physical violence during pregnancy

Psychological violence: A dummy variable equal to one if the respondent experienced at least one of the 3 psychological violence acts from her spouse:

• Insult

• Humiliate

• Threaten or scare

Sexual violence: A dummy variable equal to one if the respondent experienced at least one of the 3 sexual violence acts from her spouse:

• Forced sexual act

• Forced sex due to fear

• Humiliated sexual act

Financial control behavior: A dummy variable equal to one if respondent experienced financial control by her partner through (i) taking income from her despite her disapproval or (ii) refusing to give her money for household spending.

Independent Variables

Individual Characteristics

• Age: The age of the respondent.

• Education: Respondent’s level of education in four categories, primary school, secondary school, high school, and university and higher education.

• Mother tongue (Non-Turkish speaker): A dummy variable equal to one if the respondent’s mother tongue is not Turkish.

Partner Characteristics

• Employment status: A dummy variable equal to one if the respondent’s partner is currently employed

• Bad habits: Four dummy variables each equal to one if the respondent’s partner

  • a. Use alcohol

  • b. Use drugs

  • c. Gambles

  • d. Has affairs or another partner (through informal arrangements or religious marriage)

Household Characteristics

• Number of children: Number of living children in the household

• Wealth index: an index constructed by taking the average of 15 dummy variables, each of which equals to one if the respondent owns the asset. The assets included are: refrigerator, gas/electric oven, microwave oven, dishwasher, washing machine, plasma-TV (LCD), cable-TV(Digiturk etc.), cellphone, computer, internet, air-conditioner, car, taxi/mini-bus, tractor, and motorcycle.

• Household wealth status: Household asset ownership in three categories, poor, medium and rich:

  • o Poor if wealth index < 6

  • o Medium if wealth index > 5 and < 11

  • o Rich if wealth index > 10

Marriage Characteristics

• Women’s age at marriage: The age of the respondent at the time of her first marriage.

• Marriage decision: A dummy variable equal to one if the respondent decided on marriage together with her husband, instead of being decided by her or his family.

• Age gap between spouses: The difference between ages of husband and wife (husband’s age minus wife’s age).

HBM Variables-Women’s Employment

• Employment status: A dummy variable equal to one if the respondent is currently employed

• Social security: A dummy variable equal to one if the respondent has social security.

• Personal income index: An index constructed by taking the average of six dummy variables, each of which equals to one if the respondent has income from the ownership of following assets: land, house, company, vehicle, bank account, other income.

MBM Variables-Relative Resources

• Woman earns more money: a dummy variable equal to one if the respondent earns more money than her spouse.

• Schooling gap between spouses: The difference between husband’s and wife’s years of schooling (husband’s years of schooling minus wife’s years of schooling).

Abuse History

• Mother experienced IPV: a dummy variable equal to one if the respondent’s mother experienced IPV

• Mother-in-law experienced IPV: a dummy variable equal to one if the respondent’s mother-in-law experienced IPV

• Partner faced domestic violence: a dummy variable equal to one if the respondent’s partner experienced domestic violence.

Gender Role Attitudes

• Gender role attitudes index: an index constructed by taking the arithmetic average of seven dummy variables each of which equals to one if the respondent gives a progressive response to the following statements. The higher index numbers are associated with pro-gender equality attitudes.

  • a. A woman should not argue with partner if she disagrees with him (= 1 if no)

  • b. A woman should be able to spend her money as she wishes (= 1 if yes)

  • c. Men can beat their partners in certain situations (= 1 if no)

  • d. It may be necessary to beat children for discipline (= 1 if no)

  • e. Men should also do housework such as cooking and cleaning (= 1 if yes)

  • f. Men in the family are responsible for a woman’s behavior (= 1 if no)

  • g. It is a woman’s duty to have sexual intercourse with her husband (= 1 if no).