This project addresses one of the current data needs of retirement income research by providing a clean, user friendly panel database of detailed employer-specific pension data from the Forms 5500 for the period 1984-2005. This project has two fundamental tasks. First, it will transform the raw data from the Department of Labor (DOL) into usable panel database for the research community. To do this, the raw data will be cleaned; dictionary files and codebooks will be created; the variable names will be harmonized to maintain consistency across all available years; and each year of data will be merged to create a longitudinal dataset. The resulting data will contain adjusted measures of pension participation following traditional DOL assumptions. The second task of this project is to make the resulting data available to researchers via the World Wide Web, and provide aggregate pension information on a timely basis. To do this, this project will use the current internet infrastructure of the Center for Retirement Research at Boston College. The data and documentation will be freely available to researchers. This project will also produce aggregate tabulations of the data that will improve the timeliness of the official DOL releases by at least three years. The resulting dataset will give researchers immediate access to firm-level data that will help them address previously unanswered questions in the retirement research field, such as understanding the potential demand for older workers; the reasons why firms offer pensions; the measurement of pension coverage and participation; how pensions affect the firm's valuation; the most cost-effective manner of providing coverage; why employers use their own stock to compensate their employees; and why different industries offer different types of pension plans. ? ? ? ?
