Green Eric P, Blattman Christopher, Jamison Julian, Annan Jeannie
Duke Global Health Institute, Box 90519, Durham, NC 27708, USA.
Columbia University SIPA, 420 W 118th St, New York, NY 10027, USA.
Soc Sci Med. 2015 May;133:177-88. doi: 10.1016/j.socscimed.2015.03.042. Epub 2015 Mar 28.
Intimate partner violence is widespread and represents an obstacle to human freedom and a significant public health concern. Poverty alleviation programs and efforts to economically "empower" women have become popular policy options, but theory and empirical evidence are mixed on the relationship between women's empowerment and the experience of violence. We study the effects of a successful poverty alleviation program on women's empowerment and intimate partner relations and violence from 2009 to 2011. In the first experiment, a cluster-randomized superiority trial, 15 marginalized people (86% women) were identified in each of 120 villages (n = 1800) in Gulu and Kitgum districts in Uganda. Half of villages were randomly assigned via public lottery to immediate treatment: five days of business training, $150, and supervision and advising. We examine intent-to-treat estimates of program impact and heterogeneity in treatment effects by initial quality of partner relations. 16 months after the initial grants, the program doubled business ownership and incomes (p < 0.01); we show that the effect on monthly income, however, is moderated by initial quality of intimate partner relations. We also find small increases in marital control (p < 0.05), self-reported autonomy (p < 0.10), and quality of partner relations (p < 0.01), but essentially no change in intimate partner violence. In a second experiment, we study the impact of a low-cost attempt to include household partners (often husbands) in the process. Participants from the 60 waitlist villages (n = 904) were randomly assigned to participate in the program as individuals or with a household partner. We observe small, non-significant decreases in abuse and marital control and large increases in the quality of relationships (p < 0.05), but no effects on women's attitudes toward gender norms and a non-significant reduction in autonomy. Involving men and changing framing to promote more inclusive programming can improve relationships, but may not change gender attitudes or increase business success. Increasing women's earnings has no effect on intimate partner violence.
亲密伴侣暴力行为普遍存在，是对人类自由的一种阻碍，也是一个重大的公共卫生问题。扶贫项目以及在经济上“赋予”女性权力的努力已成为流行的政策选择，但关于女性赋权与暴力经历之间的关系，理论和实证证据并不一致。我们研究了2009年至2011年一个成功的扶贫项目对女性赋权、亲密伴侣关系及暴力行为的影响。在第一个实验中，这是一项整群随机优势试验，在乌干达古卢和基特古姆地区的120个村庄（n = 1800）中，每个村庄确定了15名边缘化人群（86%为女性）。通过公开抽签将一半的村庄随机分配到立即接受治疗组：为期五天的商业培训、150美元，以及监督和咨询。我们通过伴侣关系的初始质量来检验项目影响的意向性治疗估计值和治疗效果的异质性。在首次拨款16个月后，该项目使企业所有权和收入翻了一番（p < 0.01）；然而，我们表明，对月收入的影响受到亲密伴侣关系初始质量的调节。我们还发现婚姻控制权（p < 0.05）、自我报告的自主权（p < 0.10）和伴侣关系质量（p < 0.01）略有增加，但亲密伴侣暴力行为基本没有变化。在第二个实验中，我们研究了一项低成本尝试的影响，即让家庭伴侣（通常是丈夫）参与到这个过程中。来自60个候补村庄（n = 904）的参与者被随机分配为个人或与家庭伴侣一起参与该项目。我们观察到虐待和婚姻控制权有小幅但不显著的下降，关系质量有大幅提高（p < 0.05），但对女性对性别规范的态度没有影响，自主权有不显著的下降。让男性参与并改变框架以促进更具包容性的项目可以改善关系，但可能不会改变性别态度或增加商业成功。提高女性收入对亲密伴侣暴力行为没有影响。
