Potential earnings and labor supply of disability insurance recipients
Andreas Steinsvall Myhre
Disability Insurance (DI) programs have long been criticized for their apparent work disincentives. As such, there has been a large debate of whether individuals on DI receipt are capable of work. In this thesis, I provide new estimates of the additional earnings and labor supply potentials of Norwegian DI recipients. By using a propensity score matching method, I reweight the sample of rejected DI applicants in order to estimate the counterfactual outcomes of DI beneficiaries if they had not been awarded DI benefits. I find that Norwegian DI recipients exhibit limited earnings and labor supply potentials. I also find evidence of rejected applicants not necessarily serving as a valid control group for allowed applicants. In the years after the final allowance decision, I find that the fraction of recipients participating in the labor force would at most have been 20-25 percentage points higher if the recipients were not awarded DI benefits. I also find greater responsiveness to the allowance decision of younger applicants, suggesting greater responses to economic incentives of younger applicants.The liberalizations of the disability screening processes in most countries have led to a debate of whether the DI programs do not only serve as insurance against health shocks, but also have become a gateway for voluntary early retirement. Therefore, I investigate the behavior of Norwegian DI recipients around the transition to the retirement program. Around this transition, the individuals are eligible for the same amount of benefits, but work incentives change significantly. I find that approximately 2% of the recipients who were 62-65 years of age when awarded DI benefits increase earnings as they are transferred to the retirement program, and are no longer bound by the earnings restrictions implied by the DI program. I also find strong evidence of bunching behavior of DI recipients below the maximum allowed level of earnings for DI beneficiaries.