Data and Computer Code for Results Reported in “Early Social Security Claiming and Slow Asset Decumulation: Experimental Evidence”

Description

Data and Computer Code for Results Reported in “Early Social Security Claiming and Slow Asset Decumulation: Experimental Evidence” Program required: Matlab (2024b was used) and Stata (version 16) The programs and data in each subdirectory are organized so that they can be run independently of one another. However, our analysis was completed in the following order: a. Solve the theoretical model using the code in the "Model/Rational choice model" subdirectory (e.g., ss_t0.m, ss_t50.m, ss_t100.m). These codes are written in Matlab. On a computer equipped with 24 cores, they typically run within 20 minutes. b. Analyze the experimental data using the code in the "Data" subdirectory (e.g., analysis_student.do, analysis_prolific.do). These codes are written in Stata and run in about 10 minutes. c. Solve the behavioral models using the code in the "Model/Wealth preservation model" and "Model/Misperception model" subdirectories (e.g., wealthpreservation.m). These codes are written in Matlab, were run on an 11-core cluster, and each completes within 5 days.

Institutions

• University at Albany, State University of New York

  New York, Albany

Categories

Licence