WORKING PAPERS

Personnel Management and School Productivity: evidence from India (R&R at the Economic Journal)
with R. Lemos and K. Muralidharan

[Online appendix] [NBER WP 28336]

This paper uses new data to study school management and productivity in India. We report four main results. First, management quality in public schools is low, and ~2σ below high-income countries with comparable data. Second, private schools have higher management quality, driven by much stronger people management. Third, people management quality is correlated with both independent measures of teaching practice, as well as school productivity measured by student value added. Fourth, private school teacher pay is positively correlated with teacher effectiveness, and better-managed private schools are more likely to retain more effective teachers. Neither pattern is seen in public schools.

Organizational capacity and profit shifting (under review)
with Kat Bilicka

[Appendix] [NBER WP 29225]

Good organizational capacity drives productivity and potential taxable profits, but may also enable multinationals (MNEs) to more efficiently re-allocate profits across tax jurisdictions, lowering actual taxable profits. We show that MNE subsidiaries with better organizational capacity report significantly lower profits and have a higher incidence of bunching around zero reported profitability in high-tax countries. This pattern is not present in low-tax countries. Further, responsiveness to corporate tax rate changes in terms of profit reporting is driven by firms with good organizational capacity. We show our results are consistent with profit-shifting behavior and rule out key alternative channels.

Strategy and management practices: revisiting the World Management Survey (R&R at Strategy Science)
with S. Wolfolds

The academic field of Strategy has a strong history of theoretical literature, but a relatively younger empirical literature testing these theories. Partly due to the nature of questions in Strategic Management, scholars have often relied on collecting their own data or using specialized, and often expensive, proprietary data. This limits the possibility of replication exercises, which are a key step to refining and reinforcing the theories that are most supported in practice. To support this effort, we revisit the World Management Survey (WMS): a cross-country, cross-industry survey dataset with over 20,000 observations at the establishment-level that is collected through a rigorous and well-documented process and made free and accessible to researchers. While it is well-cited in the Strategy literature, we propose it is underused and better exposure to this data’s offerings has the potential to add significant value to the field.


The Value of Preserving Job Matches During a Crisis
(under review)
with M. Bennedsen, B. Larsen and I. Schmutte
[online appendix]

Covered on: The Economist, VoxEU
First draft: COVID Economics, Issue 27, 9 June 2020

It is generally difficult to measure the importance of preserving worker-firm relationships, particularly for low-wage jobs that involve general skills. The COVID-19 pandemic led to the sudden and seemingly temporary disruption of millions of otherwise productive employment relationships around the world. Using novel administrative and survey data from Denmark, we study a policy where firms paid up to 25\% of wages to furlough instead of firing workers. We find firms derive value from maintaining ties to low-wage and blue collar workers and that preserving those matches is beneficial to firms, suggesting policies that preserve job matches may help speed-up recovery.

Understanding school management with public data: A new measurement approach and applications (under review)
with C. Leaver and R. Lemos

[CEPR Discussion Paper] [Online Appendix]

Why do students learn more in some schools than others? One consideration receiving growing attention is school management. To study this, researchers need to be able to measure school management accurately and cheaply at scale. We introduce a new approach to measuring management practices using existing public data and exemplify the methodology with OECD’s PISA and Brazil’s Prova Brasil. Both indices show a strong, positive relationship between management and learning. We highlight two example applications: an extension of the Ahktari et al (2022) analysis of political turnover and student learning, and an exploration of the mechanisms behind the key performance relationship.

The ties that bind: implicit contracts and the adoption of management technology in the firm
with R. Lemos

Covered on: World Bank’s Development Impact blog, The Economist, LSE Business Review, VoxDev

Awarded Accessit Best Paper at IOEA 2018.
[project website] [CEPR Discussion Paper]

We investigate how implicit contracts between firm managers and employees are linked to the adoption of productivity-enhancing organizational practices. We collect new data on ownership successions and show the first causal evidence that maintaining family control leads to lower adoption of managerial best practices. We use gender composition of the outgoing CEO’s children as identifying variation at the succession point. We explore firm “reputation costs” as a novel mechanism constraining investment in management, and build a new proxy using data on eponymy — firms named after the family name. We find suggestive evidence that implicit contracts matter for management adoption.

Blame the manager? Displacement Events and Job-Driven Scarring [draft available upon request]
with I. Schmutte

We study how job loss affects the earnings, employment, and wage of managers. Using event studies around mass layoff events for Brazilian firms we find that, on average, displaced workers lose around 60 percent of their earnings in the year after job loss, and after three years have earnings 25 log points below comparable non-displaced workers. The earnings effect comes from a sharp, but temporary drop in the number of months worked per year, combined with a sharp but persistent drop in wages of around 20 log points. For managers, the wage penalty is over twice as large relative to non-managers. Less than 15 percent of the wage losses for managers can be explained by movements to employers that pay all workers less. Taken together, these results are consistent with a model in which managers either accumulate significant job-specific rents that are destroyed on separation, or they face worse market conditions and scarring after displacement, perhaps because they held responsible for the mass displacement event that precipitated their termination.

Pay Transparency and Mental Health [draft available upon request]
with M. Bennedsen, E. Simintzi, M. Tsoutsoura and D. Wolfenzon

This paper explores the impact of pay transparency on worker well-being. We exploit a pay transparency legislation in Denmark that requires firms with more than 35 employees to disclose pay information by gender. Using detailed employee and mental health prescription administrative data, we use two empirical strategies, Difference-in-Differences and Regression Discontinuity Design. Our results suggest that the legislation led to a short-run decline in the relative growth rate of anti-depressant use for women in affected firms. We consider two competing mechanisms behind this result: the potentially negative impact of horizontal (peer) comparisons and the potentially positive impact of organizational changes leading to the reduction of pay disparity. Taken together, our results are not consistent with a horizontal wage comparison mechanism and the evidence points to the second mechanism being the dominant force.

Firms and Mental Health [slides available upon request]
with M. Bennedsen, M. Tsoutsoura and D. Wolfenzon

This paper explores the extent to which firm-based factors affect workers’ mental health. We use a novel dataset on prescription anti-depressant and anxiety drugs for the population of workers in Denmark over 13 years, and link it to the employer-employee dataset tracking 2.6 million workers across over 35,000 firms over time. We use an event study design to estimate the degree of convergence in drug use after a move into a new firm. After three years, the usage of movers converges to the usage in the destination firm by as much as 20%.

SELECTED WORK IN PROGRESS

Golden skirts or dark horses: female CEOs in India
with Namrata Kala

School management and principal training: evidence from Puerto Rico
with Gustavo Bobonis and Marco Gonzales-Navarro

Entrepreneurial Manufacturers in Emerging Economies
with Angela Aristidou and Sarah Wolfolds

Natural Laws of Management
with Scott Ohlmacher and “world MOPS” leaders

Financial Management and School Performance: evidence from Mozambique
with Sandra Sequeira and Guo Xu

I have a research program with Ian Schmutte on exploring management and personnel questions with the Brazilian WMS and RAIS data. Reach out if you are interested in these topics.