Do wage subsidies for disabled workers result in deadweight loss? – evidence from the Danish Flexjob scheme

We evaluate the effects of wage subsidy programs for the disabled, in particular, their potential for welfare-loss reduction vs. deadweight loss creation. We do this in the context of the Danish Flexjob scheme, a large, nation-wide scheme that was implemented in 1998 and targeted towards improving the employment prospects of the long-term disabled with partial working capacity. We analyse the hiring response to a shock in the wage reimbursement amount to certain firms using the program. Firms received a salary reimbursement for both current and new employees granted a Flexjob subsidy. In 2002, the reimbursement to government firms was lowered while the reimbursement to municipal and regional employers remained the same. We combine the reform with unique data on whether or not a new Flexjob hiree was previously employed in a regular (unsubsidized) job at the same firm. Thus, we can investigate whether the changes in the reimbursement amount to governmental units affected the share of Flexjobs within such firms that were allocated to retained employees versus to new hires. The findings show substantial substitution between “insiders” and “outsiders” after the reform. After the reform, governmental firms create fewer Flexjobs. At the same time, the composition of Flexjob hires within such firms changes substantially: the share of new Flexjobs allocated to retained employees is twice as large as it would have been in absence of the reform. The finding on deadweight loss seems to run counter to the theoretical prediction. A possible alternative mechanism for the finding could be that when subsidies are reduced and worker productivity is not known with certainty, employers have an economic incentive to fill Flexjob positions from inside the firm.