Can buy me love: the effect of child welfare expenditures on maltreatment outcomes

While the association between adverse economic circumstances and child maltreatment outcomes is well documented in the literature, few have bothered to control this association for child welfare expenditures. Malcolm (2012) presents evidence that increases in state child welfare expenditures are strongly associated with reduced child maltreatment victimization and even fatality rates over time.

As child welfare expenditures are positively correlated with good economic circumstances, much of the measured effect of economic circumstances on child maltreatment appears to be a statistical artifact rather than a causal link. This assumption is tested by measuring the impact of the 2000 recession on the rate of maltreatment victimization in the United States while controlling for state child welfare expenditures. The multivariate regression analysis covers administrative data from the annually collected National Child Abuse & Neglect Data System (NCANDS) between 1996 and 2002. The total number of substantiated child maltreatment victims in a given state and the count of child deaths due to maltreatment were analyzed as dependent variables. Data on child welfare expenditures were available through the Urban Institute’s biennial analysis and cover the whole range of expenditures on child welfare systems (including services contracted out to private agencies, salaries and other overhead). To further control for the economic environment separately from child welfare spending at the state level, the author introduces excess revenues, education spending per child, and generosity of income transfer programs. Poverty and economic insecurity as potential contributors to increased child maltreatment were covered as covariates by including the states’ poverty and unemployment rate.

The study findings indicate that higher expenditures on child welfare were convincingly associated with less child abuse. A $100 increase in child welfare expenditures is associated with a decline in the victimization rate by 15.5% of its mean value. The effect on fatalities was less pronounced with a decline in the arrival rate of 10.1% of its mean value. Second, and more surprisingly, the well-documented effect of poverty and economic insecurity on child maltreatment incidence seems to be overstated if one fails to control for policy differences. Finally, the effect of adverse negative circumstances diminishes further upon controlling for social attitudes correlated with income, as measured by aggregates of the General Social Survey (included attitudes towards harshness of courts, religious attendance, spanking and importance of obedience as a parenting outcome).