PSID-SHELF accentuates the PSID’s strengths through (1) its household panel structure that follows the same families over multiple decades; and (2) its multigenerational genealogical design that follows the descendants of panel families that were originally sampled in 1968, with immigrant sample refreshers in 1997–1999 and 2017. Every individual who has ever been included in the PSID’s main study is included in the PSID-SHELF data, with over 80,000 people observed, some of them across more than 40 survey waves (1968–present). The current version of PSID-SHELF includes 41 waves of survey data, ranging from 1968 to 2019.

 

The file contains measures on a wide range of substantive topics from the PSID’s individual and family files, including variables on demographics, family structure, educational attainment, family income, individual earnings, employment status, occupation, housing, and wealth—as well as the essential administrative variables pertaining to key survey identifiers, panel status, sample weights, and household relationship identifiers. PSID-SHELF thus covers some of the most central variables in PSID that have been collected for many years. PSID-SHELF can easily be merged with other PSID data products to add other public-use variables by linking variables based on a survey participant’s individual and family IDs.

 

Despite a focus on longitudinally consistent measurement, many PSID variables change over waves, e.g., thanks to new code frames, topcodes, question splitting, or similar. PSID-SHELF provides harmonized measures to increase the ease of using PSID data, but by necessity this harmonization involves analytic decisions that users may or may not agree with. These decisions are described at a high level in the PSID-SHELF User Guide and Codebook, but only a close review of the Stata code used to construct variables in the data will fully reveal each analytic decision. The Stata code underlying PSID-SHELF is openly accessible not only to allow for such review but also to encourage users, as they become more comfortable with PSID, to use and alter the full code or selected code snippets for their own analytic purposes. PSID-SHELF is entirely based on publicly released data and therefore can be recreated by anyone who has registered for PSID data use.

 

Despite careful and multiple code reviews, it is possible that the code used to produce PSID-SHELF contains errors. The authors therefore encourage users to review the codes carefully, to report any mistakes and errors to us ([email protected]), and take no responsibility for any errors arising from the provided codes and files.