INTEGRATING APPROACHES AND IDENTIFYING GAPS

Rand Corporation, Santa Monica CA

Investigators

Linked publications & trials

Abstract

The primary aim of this project is to exploit the synergies between the other component projects and to identify in which directions new research is likely to be most fruitful. Thus the project has a strong comparative and methodological aspect in that it attempts to evaluate and integrate the contribution of different approaches and different types of information. As part of that approach it seeks to make contributions by experimenting with new types of data and modeling approaches that will remedy weaknesses in existing models. To achieve its integrative and evaluative aim, the project seeks to guarantee that definitions, sample selections and specification choices in the other projects are made in a coherent way. Furthermore the component projects will all be asked to estimate or simulate a number of key policy parameters of interest. Thus we will obtain these parameters for different datasets from countries with varying institutional backgrounds and for different models. A framework will be developed for comparing the contributions of different variables (e.g. health, economic incentives, and workplace conditions) to the understanding of retirement decisions. This will involve quantitative measures of goodness of fit and predictive performance in and out of sample, as well as available datasets, in particular subjective data on expectations, choice opportunities and constraints, stated preferences, etc., to validate or extend the models used. Gaps in data and knowledge that remain will be identified and new (experimental) data will be collected to address as yet unanswered questions. We will propose the addition of new types of data to some of the most important surveys in the United States and Europe.