Privatization of the Foster Care System through the Lens of Efficiency and Effectiveness

Background and Purpose

Privatization of the public child welfare (PCW) system has become increasingly popular since its introduction in the early 1990s.  State governments that initiate the privatization of foster care services rationalize the changes with claims of effectiveness and/or increased efficiency of the system.  There has been no real focus on identifying what efficiency of the system means for children and their families, nor what aspects of effectiveness focuses on children in foster care.  As a result, the unintended consequences of this total restructuring of foster care bureaucracy, through the privatization of the state foster care system are as yet unknown. The primary aim of this study is to investigate whether or not there are differences between state foster care systems and their levels of privatization, as well as the differences in states’ rates of efficiency and effectiveness with regard to a child’s trajectory of experience within the foster care system.

Methods

A secondary data analysis was conducted using the Adoption and Foster Care Analysis and Reporting System (AFCARS) data for this study (n=683, 031). Using a purposive sampling method, states (n=51) were grouped into three groups: fully privatized, partially privatized, and non-privatized. Bivariate and multivariate comparisons were conducted by level of privatization to determine differences between privatized, partially privatized, and non-privatized foster care systems and their rates of efficiency and effectiveness in categories related to removal manner, removal reason, and children exiting to emancipation.

Results

A discriminant function analysis (DFA) was conducted to determine any differences between the three levels of privatization of the foster care system as well as each model’s rate of efficiency and effectiveness. Box’s M test was utilized to test the assumption that groups have equal variance-covariance matrices. Box’s M equaled 61, 015.50, F (110, 4978768827) =, p< .001, which means that equality of variance-covariance matrices cannot be assumed. The three-group discriminant analysis produced two discriminant functions.  The first function had a canonical correlation of .314, whereas the second function had a canonical correlation of .150.  For functions one through two, Wilk’s Lambda equaled .881, Chi-square (20, N= 73, 193) 9,258, p< .001.

These results suggest that the models slightly to moderately discriminate among the three groups of states based on levels of privatization and their rates of efficiency and effectiveness. Specifically, non-privatized foster care systems differentiate the most from fully privatized and partially privatized foster care systems and are both more efficient and more effective, with standardized coefficients for function 1 of 0.015 and 0.712 respectively.

Conclusions/Implications

The finding of differences between the levels of privatization of the foster care system is one of the first in research that is supported by empirical evidence from a national database inclusive of public child welfare information. Ultimately, the results of the study demonstrated that non-privatized foster care systems demonstrate higher rates of efficiency and effectiveness. While further research is warranted to extend these findings, the results from this study demonstrate the necessity of stronger data regarding privatization of public child welfare.