Essays in unemployment insurance

Desbuquois, Alexandre (2022) Essays in unemployment insurance. PhD thesis, London School of Economics and Political Science.

Text - Submitted Version Download (3MB)

Abstract

This thesis consists of three chapters. In the first chapter (Optimal Unemployment Insurance with Liquidity Provision), I exploit the joint provision, in many countries, of Unemployment Benefits (UB) and a Severance Package (SP) post redundancy to suggest a new and welfare enhancing organisation of Unemployment Insurance (UI). To do so, I use French administrative data and a combination of regression kink and discontinuity designs. While the SP and UB paid early in the unemployment spell should have roughly the same consumption smoothing value, I show that UB are a lot more distortive. All of my results point in the same direction and indicate that the SP would be a good substitute for UB paid early in the unemployment spell. As a consequence, the optimal UI organisation should be made of a SP at the onset of unemployment followed by a waiting period with no UB, and by the payment of UB in case of long-term unemployment. In the second chapter (Can Unemployment Insurance Reduce Job Stability?) I analyse the impact of a specific eligibility condition on unemployed individuals’ incentives to accept short-time contracts. The analysis relies on a reform in France that reduced the necessary number of days worked to open eligibility to UI from 122 to 30 days. By exploiting administrative data and a regression kink design strategy, I show that this reform made unemployed more likely to accept short-time contracts, and to repeat the unemployment experience after their end date. I estimate this increased likelihood to become unemployed again to generate an additional cost per unemployed of 500 to 700e. In the third chapter (There is Only One First Time: Behavioural Responses and Unemployment Experience) I explore the impact of one dimension of heterogeneity, namely unemployment experience, on unemployed individuals’ behavioural 4 response to UB. This paper leverages a key advantage of the MCVL, a Spanish administrative dataset: it tracks the full employment history of a 4% randomly selected sample, i.e. that it provides full details about every job held by these individuals since their very first employment contract. I then analyse, through a regression kink design strategy, a discontinuity in the UB schedule, and differentiate this approach as a function of individuals’ unemployment experience. Only individuals in their very first unemployment spell significantly respond to the exogenous change in the UB level. This finding is robust to a large set of validity checks and to multiple specifications.

Item Type:	Thesis (PhD)
Additional Information:	© 2022 Alexandre Desbuquois
Library of Congress subject classification:	H Social Sciences > HC Economic History and Conditions H Social Sciences > HD Industries. Land use. Labor
Sets:	Departments > Economics
Supervisor:	Landais, Camille
URI:	http://etheses.lse.ac.uk/id/eprint/4419

Actions (login required)

Record administration - authorised staff only