Essays on estimating effects of star arrival in small open economies: An event-study analysis
Yadav, Anil
Yadav, Anil
Loading...
Repository DOI
Publication Date
2024-03-14
Type
Thesis
Downloads
Citation
Abstract
As the economy’s growth becomes more dependent on innovation, science policy is increasing in importance as an instrument for shaping the direction and impact of science on economic development, especially in small open economies. Policymakers in small open economies have shown increased interest in various science policies to stimulate the development of national innovation systems. A potentially attractive strategy in the science-policy mix is the targeted recruitment of star scientists designed to catalyse the development of research clusters in targeted research areas. However, there is little evidence on how stars impact the productivity of their peers in small open economies. In this context, the objective of the thesis is to provide an evidence-based evaluation of the impact of star scientists on their peers. It addresses five specific goals aligned with the objective in five separate empirical and methodological papers. A better understanding of the knowledge spillovers from the star can support the development of star recruitment policies intended to recruit stars and integrate them into local networks to catalyse productivity. Thus, the first paper of this thesis focuses on the co-authorship channel, examining how forming a co-authorship relationship with a co-located star affects the productivity of the co-author, both including and excluding star co-authored output. The latter effect provides a measure of the extent to which star collaborations crowd out/in other output. The analysis is conducted on a matched sample for treated and control authors obtained using coarsened exact matching. An event-study model is employed on the matched sample to estimate the dynamic effect of the star co-authorship on quality-adjusted productivity. In addition, I explore the heterogeneity by period, field, and whether the authors have multiple star co-authorships. While the first study offers valuable insights into the co-authorship relationship with the star, the second paper investigates the dynamic effects of star arrival on the arriving department. It is hoped that the arriving star can positively impact department productivity through channels such as access to better training and knowledge, changing norms relating to research, and even creating collaboration opportunities. An event-study model estimates the dynamic effects of a star arrival on quality-adjusted research output at both the department and matched individual incumbent levels. The matched individual-level sample allows us to make more credible causal inferences about the star arrival effect. To help further understand star impacts on peers, the third paper examines how relatedness to the arriving star modifies the size of the productivity effect on co-located peers. The paper assumes a non-linear relationship between the level of relatedness and the productivity effect, given that higher relatedness, on the one hand, can increase absorptive capacity and, on the other hand, aggravate potential knowledge redundancy. A difference-in-difference model is utilised to estimate the treatment effects for different intensities of treatment, where the intensity is proxied by the relatedness measure. In addition, an event-study model is employed to provide evidence of a causal effect on incumbent productivity from star arrivals, where I use coarsened exact matching to identify a control for each treated incumbent. The previous three papers provide an evidence-based evaluation of star impact using difference-in-difference and event-study research designs. In addition, difference-in differences and event-study research designs are one of the most common methods to estimate treatment effects in the economics of science literature. However, recent developments in the literature on difference-in-differences and event-study models have raised important concerns that when effects are heterogeneous and treatment is staggered, the two-way fixed effects (TWFE) difference-in-differences/event-studies regression can provide biased estimates. Thus, policy formed based on studies using these biased approaches may not be effective. To this end, the fourth paper (a methodological paper) addresses this concern by exploring the relative performance of five alternative estimators with the TWFE estimator for estimating treatment effects on binary and count outcomes within a simulation exercise. Following this, the fifth paper (a replication paper) implements the alternative estimators on a prior published study and compares the results with the original study, highlighting the need for policymakers to pay particular attention to research design before drawing strong policy-related conclusions. The fourth and fifth paper indirectly affects evidence-based policymaking and addresses the broad objective of the thesis. The thesis is a novel contribution to the Irish and international economics of science empirical and policy literature, given the paucity of information currently available on the impact of star scientists in small open economies. This thesis has significant implications for policymakers and institutions designing recruitment strategies and post-arrival support. I hope the results of the five studies undertaken for the thesis will be particularly useful for policymakers and funding agencies in Ireland and other small open economies to form evidence-based star recruitment policies.
Funder
Publisher
NUI Galway