Are Public Child Welfare Worker Wages Associated with Caseload and Overtime Pay Ratios?

• In Event: Upstream Policies, Downstream Harm: Systemic Influences on Child Welfare

Abstract

Motivation: The public child protective services (CPS) workforce provides legally mandated support and casework services to vulnerable children and families. However, CPS agencies face chronically high worker turnover rates which incur significant financial costs to the agency and are linked to detrimental impacts on youth. The literature is sparse regarding relationships between CPS wage levels and turnover-related factors, despite qualitative evidence that financial well-being affects workers’ ability to care for clients on the job and their intent to stay. This study utilizes derived longitudinal wage and administrative data to test whether there is an empirical relationship between worker wages and turnover factors to establish baseline observations and provide direction for future work. This study highlights the need to empirically investigate under-studied workforce interventions which impact CPS agencies’ ability to provide equitable and quality services.

Methods: County worker wage data were downloaded from the California State Controller’s Office website spanning 2011-2023 and subsetted to 42 unique counties with more than ten workers reported per year (n=489). CPS workers were identified using names of county agencies then searching for non-managerial master’s level worker titles. Annual staff sizes were calculated from the number of rows per county per year, and caseloads were derived by dividing the number of children with investigations per county by total staff. Overtime pay ratios, or annual overtime earnings divided by total earnings, were calculated for each worker and averaged by county to measure the share of total income due to overtime wages. The wage data were converted to nominal hourly mean wages by dividing workers’ annual total wage by 2,080 hours. Finally, fixed effects and random effects linear regressions were used to estimate models where caseload was the dependent variable, and wages and overtime earnings ratios were the independent variables (transformed using the natural log function to aid estimation and interpretation).

Results: Both random intercept and fixed effects regressions yield estimates of an approximate 0.78 (SE = 0.12, p<0.01) to 0.99 (SE=0.13, p<0.01) percent decrease in average caseload for a one percent increase in the average hourly nominal wage, depending on whether overtime pay is accounted for. Both models estimate an approximate 0.11 (SE=0.03, p<0.01) percent increase in the average caseload for a one percent increase in the overtime earnings ratio. Random intercept ICC estimates indicate slightly higher between-county variance than within (ICC~0.6). These results demonstrate that decreases in caseload levels over time correlate strongly with nearly proportional increases in wages, while increases in overtime earnings ratios correlate with small increases in average caseload.

Implications: Preliminary analysis suggests a possible strong negative relationship between changes in wages and changes in average caseload for California CPS workers over time, while changes in caseloads are moderately positively correlated with changes in overtime earnings ratios. Future research could include investigating the causal relationship between these variables and whether these relationships stem from better retention and recruitment outcomes or increased resources in the community which correlate with higher pay, better case outcomes, higher job satisfaction, and subsequently lower worker turnover.

Author

• Sofia Guo, University of California, Berkeley
  Presenting Author