It saddens me to inform our readers of the death of Steven Sandell (see page 13), Director of SSAs Division of Policy Evaluation. Steve made great strides in evaluating the effects of public policy on participants in America's social welfare programs, particularly older workers and retirees. On a personal note, several of us in ORES had grown personally and professionally close to Steve in recent years. We feel his loss very deeply.
This issue includes an article co-authored by Steve on the distributional effects of changing the averaging period used to calculate Social Security benefits and on establishing a minimum benefit provision. The authors assess the impact of increasing the required earnings for a benefit from 35 to 38 and to 40 years of earnings.
Our second article examines the changes in earnings distributions over the 1981-95 period. Because the article uses data from SSNs Continuous Work History Sample (that comes directly from W-2 forms filed by employers), it generally avoids the self-reporting errors and other problems often faced by researchers involved in this type off analysis.
Next is a description of the comprehensive administrative record database developed for the evaluation of the Project NetWork return-to-work experiment targeting people with disabilities. The results show that it is feasible to simulate complex eligibility rules using administrative records, and to create a rich data file for a comprehensive evaluation. Useful lessons can be learned from this pioneering experience for the design and implementation of future demonstration evaluations. The results of the Project NetWork experience show that it is feasible to use administrative records to create control or comparison groups in a versatile manner that would be either not feasible or prohibitively expensive using survey data alone.
A brief overview of the more important analyses of lifetime distribution under the Social Security program-analyses that are sometimes referred to as "money's worth" studies-is presented in our next article. Such studies are essential to understanding how …