Colorado Department of Human Services

This report summarizes statistical models The Lewin Group developed to explain the determinants of caseload trends in the Colorado Works program from July 1998 through December 2007. Using regression analysis, equations were developed to estimate the size of the Colorado Works monthly caseload based on the Colorado economy, Colorado Works program characteristics, and other factors. Separate models of caseload size were developed for one-parent families, two-parent families, and child-only cases. In addition, models were also developed of monthly entries and exits for one-parent cases, and county-level models of the one-parent caseload for the five largest counties. Models were estimated using data from July 1999 through December 2007. Among the findings: Unemployment rate is the key factor affecting the size of the one-parent and two-parent caseloads. The caseload in a month is affected by the unemployment rate up to 24 months earlier. Several Colorado Works and federal policy variables also had a statistically significant effect on the size of the caseload.

"Throughout the engagement, Lewin consultants listened to us and delivered superb analyses and thoughtful guidance–not canned solutions."  NONPROFIT HEALTH SYSTEM