Periodic Reporting for period 2 - Evidence-VAW (Generating New Evidence to Address Violence Against Women: Realizing Women’s Rights)
Reporting period: 2023-04-01 to 2024-09-30
Evidence-VAW is a fundamental cause of gender inequality and a substantial constraint on women’s lives, but it has drawn significantly less intellectual curiosity from economists than aspects of gender inequality such as the gender pay gap. I propose to redress this by acquiring, generating and analysing unavailable or under-used data using techniques for causal identification with a view to producing a rich tapestry of scientific evidence to guide policy and further scholarship. This is an important time to do this as there is a growing consensus around prioritizing VAW in international policymaking but the evidence base is thin. There are no systematic data on VAW, partly because women are inhibited in reporting. Creative new strategies are being used in many countries to reduce VAW, but without scientific evaluation. The main objective of the project is to improve our understanding of the causes of VAW by generating rigorous evidence and, based on this, to inform the design of policy. We will investigate the role of information, training, firm policy, state policy and institution design. We will implement experiments, gather new survey data, and leverage natural experimental variation from policy reform or institutional features. Where possible we will use longitudinal administrative data providing unparalleled opportunities for research. Although focused on women’s rights, this project addresses issues on the frontier of economics research, including inequality, productivity, institutional design, political economy, and legislative change.
Domestic violence (DV) - The literature on DV has focused upon interventions that increase the relative economic power of women, assuming that the problem reflects power issues. I argue that a first-order issue that dominates the power balance is unemployment. I've published innovative research identifying how job loss of men affects the chances that they perpetrate DV, how job loss of women affects the chances that they are victimized, and whether unemployment benefits can be re-imagined as a policy tool for DV. I used longitudinal Brazilian and Chilean administrative data sets and state of the art econometric techniques to identify causal effects. My work provides the first causal estimates of these parameters, a critical input to policy design. It proposes a policy that has not been discussed in the context of DV. I have written a paper on the evolution over time of the gender pay gap across the income distribution which is relevant as the first economics paper identifies the gender pay gap as the root cause of DV.
Continuing on the theme of employment and DV, I have spent a lot of time investigating the new Italian policy of offering a payroll tax cut to employers who hire women who have been abused. I'm currently structuring a pilot study to meet the challenges of accessing women in public shelters and employers. This work is in its early stages. The final aim of this work is to provide evidence-based advice on the structuring of subsidies like this, their likely take up and the chances that they provide sustainable employment for victims.
A problem with addressing violence against women is that women do not report because they typically suffer retaliation for reporting, whether from employers or husbands, and not much is done to fix the underlying issues. It has been argued that reporting of domestic violence can be improved by creating police stations that have only female police officers. We evaluated the roll out of these in India and we find this is the case. We provide evidence to show that the increased reports reflect changes in reporting rather than changes in incidence. In a different paper still in progress I am looking at how trust in police and courts influences women’s beliefs and partnership decisions in Chile.
I have drafted a paper on male fertility preferences as a driver of sexual violence in India, and another on alcohol bans, political mobilization and domestic violence. I am currently studying unsafe abortion (a form of VAW) in response to US foreign aid policy.
Workplace sexual harassment - In most countries, workplace sexual harassment (SH) is the responsibility of the employer. While HR teams exist to address harassment, they typically collude with the employer to protect the reputation of the institution. I have drafted a paper investigating how reporting can compromise the careers of women. I set this paper in the context of an Indian law that mandated that firms take responsibility for encouraging reporting and redressal in SH cases. I leverage the timing and the scope of the law to identify its impacts. I find that, after the law, firms are less likely to hire women (as compared with men). In this sense, the well-intended law misfired. The new UK Worker Protection Act is similar and my findings call for more careful calibration of policies like this. To pursue this, I'm now modelling firm behaviour using structural techniques. This work will deliver the first estimates of the extent to which SH in the workplace hinders productivity.
I have completed the intellectual conception of a project that will analyse whether gender-specific management training can improve worker wellbeing, including sexual harassment risk, and productivity. There is some evidence that female managers are more likely to follow through on SH complaints but there is no evidence of whether this comes at the cost of productivity; or whether men can be trained in the methods that successful women use, or indeed, whether women can be trained in the methods whether men use. This is the gap we propose to fill. This project is held up for the funds transfer to Colombia.
We have developed a discrete choice experiment on women’s perceptions of safety in the workplace and whether this limits their work participation. We have done a pilot and will do a full experiment in the Spring exploring social mobility and ability gradients. We have drafted a pre-registered report for survey experiments that investigate citizen, manager and policymaker beliefs over prevalence, harms and policy effectiveness. We will evaluate the scope for information in these domains to improve policy in the sexual harassment space. We have drafted a pilot study designed to understand how individuals perceive sexual harassment, what traits of the incident would lead them to report and whether this varies with identity concordance. Possibly the most ambitious of these projects is a discrete choice experiment designed to understand the phenomenon of women getting into and persisting in abusive partnerships.
Continuing on the theme of employment and DV, I have spent a lot of time investigating the new Italian policy of offering a payroll tax cut to employers who hire women who have been abused. I'm currently structuring a pilot study to meet the challenges of accessing women in public shelters and employers. This work is in its early stages. The final aim of this work is to provide evidence-based advice on the structuring of subsidies like this, their likely take up and the chances that they provide sustainable employment for victims.
A problem with addressing violence against women is that women do not report because they typically suffer retaliation for reporting, whether from employers or husbands, and not much is done to fix the underlying issues. It has been argued that reporting of domestic violence can be improved by creating police stations that have only female police officers. We evaluated the roll out of these in India and we find this is the case. We provide evidence to show that the increased reports reflect changes in reporting rather than changes in incidence. In a different paper still in progress I am looking at how trust in police and courts influences women’s beliefs and partnership decisions in Chile.
I have drafted a paper on male fertility preferences as a driver of sexual violence in India, and another on alcohol bans, political mobilization and domestic violence. I am currently studying unsafe abortion (a form of VAW) in response to US foreign aid policy.
Workplace sexual harassment - In most countries, workplace sexual harassment (SH) is the responsibility of the employer. While HR teams exist to address harassment, they typically collude with the employer to protect the reputation of the institution. I have drafted a paper investigating how reporting can compromise the careers of women. I set this paper in the context of an Indian law that mandated that firms take responsibility for encouraging reporting and redressal in SH cases. I leverage the timing and the scope of the law to identify its impacts. I find that, after the law, firms are less likely to hire women (as compared with men). In this sense, the well-intended law misfired. The new UK Worker Protection Act is similar and my findings call for more careful calibration of policies like this. To pursue this, I'm now modelling firm behaviour using structural techniques. This work will deliver the first estimates of the extent to which SH in the workplace hinders productivity.
I have completed the intellectual conception of a project that will analyse whether gender-specific management training can improve worker wellbeing, including sexual harassment risk, and productivity. There is some evidence that female managers are more likely to follow through on SH complaints but there is no evidence of whether this comes at the cost of productivity; or whether men can be trained in the methods that successful women use, or indeed, whether women can be trained in the methods whether men use. This is the gap we propose to fill. This project is held up for the funds transfer to Colombia.
We have developed a discrete choice experiment on women’s perceptions of safety in the workplace and whether this limits their work participation. We have done a pilot and will do a full experiment in the Spring exploring social mobility and ability gradients. We have drafted a pre-registered report for survey experiments that investigate citizen, manager and policymaker beliefs over prevalence, harms and policy effectiveness. We will evaluate the scope for information in these domains to improve policy in the sexual harassment space. We have drafted a pilot study designed to understand how individuals perceive sexual harassment, what traits of the incident would lead them to report and whether this varies with identity concordance. Possibly the most ambitious of these projects is a discrete choice experiment designed to understand the phenomenon of women getting into and persisting in abusive partnerships.
I am making progress beyond the state of the art in three dimensions. 1. I am asking questions that have not been asked – to take two examples: Would extension of unemployment benefits reduce domestic violence?; Would mandating firms to prevent sexual harassment improve working conditions for women? Does subsidising employers to hire female victims of abuse put them on a sustainable new economic trajectory? 2. I am collecting data on questions for which no data exist. For example, to what extent do women trust in the police? Do individuals over or under-estimate the prevalence and harms of workplace sexual harassment? Do these perceptions matter for how effective policy is? 3. I am using new or state of the art methods including RCTs, discrete choice and survey experiments, and analysis of administrative data.