Do Family Support Centers Reduce Maltreatment Investigations? Evidence from Allegheny County

Background: The Family First Prevention Services Act encourages states to develop their child abuse prevention efforts by implementing services that are grounded in evidence. For that evidence-base, FFPSA and other policies usually look to individual or family-level interventions. Public policy and intervention research is less often focused on community-level interventions, despite decades of research on neighborhood effects and a growing body of evidence that suggests communities are an important element of effective interventions (e.g., Communities that Care). Family Support Centers (FSCs) are another example of a community-level intervention. Among other aims, FSCs strengthen communities and thereby mitigate the risks that bring families to the attention of the child welfare system.  However, FSCs are still early in the evidence-building cycle.  To advance that evidence-building process, this study applies an ecological, social determinants framework to examine whether there is a plausible link between the presence of neighborhood-based FSCs and lower maltreatment investigation rates.

Method: We used three sources of data from one urban U.S. county: 1) child maltreatment investigation records from 2009 to 2013, 2) child population counts and eight measures of social disadvantage from the 2010 Census, and 3) geo-locations of the FSCs. We aggregated individual investigations to census tracts constituting 192 neighborhoods, 55 of which were served by an FSC. We then created an index of social disadvantage from the census measures in order to statistically compare, using a multilevel fixed effect Poisson count model with variable exposure, ecologically similar neighborhoods, the location of FSCs and investigation rates.

Results: There are three main findings: 1) FSCs were more likely to be located in areas with higher investigation rates (event rate ratio=2.15, p=0.001), which matches the strategic intent of locating FSCs in high-need neighborhoods; 2) although FSCs are located in socially-disadvantaged areas, the neighborhoods where FSCs are located are ecologically diverse; and 3) when comparing ecologically similar areas served by an FSC with those areas that are not served by an FSC, areas with an FSC had lower investigation rates (event rate ratio=.79, p=0.005).

Conclusions: Located in areas where natural support systems need strengthening (judging from the level of social disadvantage), FSCs were associated with lower investigation rates.  From an evidence-building perspective, given the emphasis on well-supported interventions, these findings are suggestive but not dispositive.  Nevertheless, the findings support the idea that protective mechanisms operate at the community-level.  Theory suggests that FSCs, by their presence, alter the risk and protective character of communities in ways that affect individual-level risk.  In our study, we found evidence that supports this underlying thesis.  Further research could, using an experimental design, explicitly test this theory.  Moreover, well-designed studies might also test the extent to which more traditional interventions, i.e., interventions that target individuals within families (e.g., parenting programs), depend on ecological effects to achieve the potency found in well-controlled intervention research.  The authors will discuss opportunities for bringing ecological research into the intervention research arena, which could broaden the range of evidence-based prevention strategies.