Examining the Labor Market Effects of Public Pension Design
Publication Date
1-1-2016
Grant Type
Early Career Research Award
Description
Teacher retirement funds fall short of their liabilities by an estimated $1 trillion, causing many states to consider cost-saving changes including restructuring their public pension plans. Most public school teachers are eligible for a defined-benefit pension plan that creates abrupt changes in retirement incentives as teachers age. Restructuring pensions could have unintended consequences on student learning outcomes and further complicate staffing in hard-to-staff schools. Using a Cox proportional hazard model, I leverage the plausibly exogenous variation in retirement incentives driven by pension rules to identify the causal effect of retirement incentives on exit behavior. I use a detailed, 17-year panel of data on all North Carolina public school teachers to study the retirement behavior teachers with different characteristics as they approach and pass full retirement eligibility. Of particular importance are the effects on retention decisions of teachers with relatively large or small impacts on student achievement, as indicated by ‘value-added’ measures. I also explore whether teachers at hard-to-staff schools are more sensitive to pension incentives, potentially making these schools even more difficult to administer. To my knowledge, this is the first study of its kind to (i) use a hazard model to study teacher retirement, (ii) measure retirement incentives by the proximity to pension eligibility thresholds, and (iii) to differentiate the retirement behavior of teachers by teacher-level quality. My results concerning the heterogeneous retirement behavior of teachers applies directly to the pension policy debate currently ongoing in many states, and may apply more broadly to the retirement decisions of other types of workers, where worker quality and working conditions are difficult to measure.
Grant Product
Are Teacher Pensions "Hazardous" for Schools?
Upjohn Institute Working Paper No. 18-281, 2018