How important is adjustment cost for individuals who face a change in work incentives induced by a policy change? I provide the first estimate of adjustment cost in the context of a Disability Insurance (DI) program by exploiting a policy change that increased work incentives by increasing the exemption threshold. I use bunching at the thresholds to estimate adjustment cost and earnings elasticity. The estimates of adjustment cost are quite sizeable and heterogeneous. The estimates of earnings elasticity, from both static and dynamic models, are larger than the elasticity estimated with no adjustment cost, suggesting that adjustment cost attenuates the responses to work incentives. Policies designed to increase labor supply will work if the work incentives are large enough to offset the adjustment cost. Accounting for adjustment cost then might explain the disparate findings on the effects of an increase in work incentives on the labor supply of beneficiaries of DI programs. (C) 2020 Elsevier B.V. All rights reserved.