Can Training Improve Organizational Culture? Experimental Evidence from Ghana's Civil Service

joint with Michel Azulai (IFS), Daniel Rogger (World Bank) and Martin.J.Williams (Oxford).

October 2020

Abstract Organizational culture is an important driver of organizational performance, but evidence on how to improve performance-oriented organizational cultures is scarce - especially in the public sector. We partnered with Ghana's Civil Service to design a new innovation training module geared towards such culture change and deliver it on a randomized basis to mid-level bureaucrats in central government. The intervention was delivered at full scale by integrating it into the Civil Service's standard training routine for one year. We find that the training improved organizational culture and performance 6-18 months post-training. Our design was split between an individual-focused training arm and one in which officials from an entire organizational unit were trained together. Our results are completely driven by the individual-focused arm, with the team-based treatment arm having no impact on culture or performance. We discuss potential explanations for this difference in effectiveness. Simple and scalable training interventions can thus have significant impacts on culture and performance, but their design matters.

Do School Closures During an Epidemic have Persistent Effects? Evidence from Sierra Leone in the Time of Ebola

joint with Oriana Bandiera (LSE), Niklas Buehren (World Bank), Markus Goldstein (World Bank) and Andrea Smurra (UCL)

August 2020

Abstract School closures are a common short run policy response to viral epidemics. We study the persistent post-epidemic impacts of this on the economic lives of young women in Sierra Leone, a context where women frequently experience sexual violence and face multiple economic disadvantages. We do so by evaluating an intervention targeting young women that was implemented during the 2014/15 Ebola epidemic in Sierra Leone. This provided them a protective space where they can find support, and receive information on health/reproductive issues. We document the impacts of the intervention on 4,700 young girls and women tracked from May 2014 on the eve of the Ebola crisis, to the post-epidemic period in 2016. In control villages, school closures led young girls to spend significantly more time with men, teen pregnancies rose sharply, and school enrolment among young girls dropped by 17pp post-epidemic, long after schools had re-opened. These adverse effects on enrolment are halved in treated villages because the intervention breaks this causal chain: it enables girls to allocate time away from men, reduces out-of-wedlock pregnancies by 7pp, and so increases re-enrolment rates post-epidemic. A long term follow up in 2019/20 shows persistent impacts of the intervention on the human capital accumulation of young girls, time they spend with men, and quality of partners matched with. Our analysis has important implications for school closures in response to the current COVID-19 pandemic in contexts where young women face sexual violence, highlighting the protective and lasting role safe spaces can provide in such times.

The Anatomy of a Public Health Crisis: Household Responses Over the Course of the Zika Epidemic in Brazil

joint with Ildo Lautharte Junior (World Bank)

June 2020

Abstract The global frequency and complexity of viral outbreaks is increasing. In 2015, Brazil experienced an epidemic caused by the Zika virus. This represents the first known association between a flavivirus and congenital disease, marking a `new chapter in the history of medicine' [Brito 2015]. We use tens of millions of administrative records to document household responses to a public health alert linking the emerging Zika virus and congenital disease. We find a 7% reduction in pregnancies post-alert, a response triggered immediately after the alert, and driven by higher SES women. On responses during pregnancy, we find an increased use of ultrasounds (9%) and abortions (5%), especially late term abortions. However, these impacts are driven by mothers that conceived post-alert -- there is no response to the public health alert during pregnancy among mothers that conceived just pre-alert, despite their unborn children also being at risk. The primary welfare cost of the epidemic is disease avoidance, as households move away from planned fertility paths. We conclude by discussing the extent to which our findings extend to household responses to public health alerts on other emerging viral threats.

Development Policy through the Lens of Social Structure

joint with Oriana Bandiera (LSE), Robin Burgess (LSE), Erika Deserranno (Kellogg), Ricardo Morel (IPA) and Munshi Sulaiman (BRAC).

May 2020

Abstract This paper studies how the social structure of village economies affects policy implementation by local agents. We randomly select one of two viable candidates to deliver an agricultural extension program in rural Ugandan villages. We show that delivery agents favor their own social ties over ex-ante identical farmers connected to the other (non-selected) candidate and that this is inconsistent with output maximization or targeting the poorest. Favoritism disappears when the potential delivery agents belong to the same social group. Using the randomized allocation of the program across villages, we show how unobserved social structures explain the variation in delivery rates and program effectiveness that we often observe in the data.

Identifying Network Ties from Panel Data: Theory and an Application to Tax Competition

joint with Aureo de Paula (UCL) and Pedro CL Souza (PUC-Rio)

April 2020

Abstract Social interactions determine many economic behaviors, but information on social ties does not exist in most publicly available and widely used datasets. We present results on the identification of social networks from observational panel data that contains no information on social ties between agents. In the context of a canonical social interactions model, we provide sufficient conditions under which the social interactions matrix, endogenous and exogenous social effect parameters are all globally identified. While this result is relevant across different estimation strategies, we then describe how high-dimensional estimation techniques can be used to estimate the interactions model based on the Adaptive Elastic Net GMM method. We employ the method to study tax competition across US states. We find the identified social interactions matrix implies tax competition differs markedly from the common assumption of competition between geographically neighboring states, providing further insights for the long-standing debate on the relative roles of factor mobility and yardstick competition in driving tax setting behavior across states. Most broadly, our identification and application show the analysis of social interactions can be extended to economic realms where no network data exists.

Revisions requested, Review of Economic Studies.

Parental Responses to Information About School Quality: Evidence from Linked Survey and Administrative Data

joint with Ellen Greaves (Bristol), Iftikhar Hussain (Sussex) and Birgitta Rabe (Essex).

November 2019

Abstract Multiple inputs determine children's cognitive achievement. We study the interaction between family and school inputs by identifying the causal impact of information about school quality on parental time investment into children. Our setting is England, where credible information on school quality is provided by a nationwide school inspection regime. Schools are inspected at short notice, with school ratings using hard and soft information. As such soft information is not necessarily known to parents ex ante, inspection ratings provide news to parents that plausibly shifts inputs into their children. We study this using household panel data linked to administrative records on school performance and inspection ratings. We observe some households being interviewed pre school inspection (control), and others being interviewed post inspection (treatment). Treatment assignment is determined by a household's survey date relative to the school inspection date, and shown to be as good as random. We find that parents receiving good news over school quality significantly decrease time investment into their children (relative to parents that will later receive such good news). Hence on average, beliefs over school quality and parental inputs are substitutes. Our data and design allow us to provide insights on the distributional and test score impacts of the nationwide inspections regime, through multiple margins of endogenous response of parents and children. Our findings highlight the importance of accounting for interlinked private responses by families to policy inputs into education.

Revisions requested, Economic Journal.

Please reload