Adoption in Two Southern-East States: Exploring Why More Kids Get Adopted and Sooner in Arkansas?

In 2015, Mississippi (MS) had 1,437 children waiting to be adopted, and of those were 312 adopted (U.S. DHHS, 2016b).  If the percentage of children of adopted were calculated from the previous two figures, the resulting percentage for Mississippi would have been 21.7%.  Using the same formula the percentage for the Unite State would have been 48%, and Mississippi would have placed number 50 of 52 compared to all 50 states, including the District of Columbia and Puerto Rico.  This paper explores the characteristics that influence adoptions from the foster care system in Mississippi and Arkansas (AR).  To examine the differences between both state the Adoption and Foster Care Analysis and Reporting System (AFCARS), Foster Care File data for these two states was used.  Propensity Score Matching (PSM) was conducted, however the difference between the mean propensity scores from the treatment (i.e., MS) and the matched (i.e., AR) were not significantly different pre- and postPSM, therefore the original state samples were used.  Survival analysis (i.e., Cox Regression) was used to examine the differences in adoption between the two states.  As the first step in the survival analysis, Life Tables analysis was used, which is a descriptive technique for observing the distribution of time-to-event variables. Life Tables indicated that the survival curves were different across both states.  Following, Life Tables, a Kaplan-Meier survival analysis was conducted.  The Kaplan-Meier analysis is a nonparametric technique, where no assumption for the shape of the hazard function is assumed. The results suggested that children in AR may have faster adoption periods than the children in MS.  The final step was a Cox regression, this regression was used to test the hypothesis regarding the effect of two different states “scenarios” on the occurrence of adoption.  There were several state variable interactions that ended up in the final model, suggesting that black children in MS children can expect an adoption hazard 1.6 times greater than their AR counterparts.  Both urban and rural children in MS would experience higher adoption hazard that their counterparts in AR, 1.69 and 2.69 times, respectively.  Factors that contributed to significantly reduce the adoption hazard in MS were stable placements (i.e. 40.3% reduction), and children place in either relative care or in pre-adoption homes (i.e. 75.8 and 92.9% reduction, respectively).  Social Work and Child Welfare implications of these findings and other are discussed.