Q: What’s your background?
A: I am a social scientist (PhD in Social Psychology) with primary interest in homelessness. I have studied homelessness for almost 40 years. My principal approach has been through the analysis of administrative data, and the linkage of homeless shelter data with a wide range of datasets across health, education, employment, housing, criminal justice, and social services.
Q: How did you get into the linkage space?
A: I discovered in 1992 that Philadelphia and New York City were the only cities in the US with centralized shelter registration and placement systems. These data enabled the first longitudinal studies on the prevalence and patterns of homelessness. By 1995, I was linking the shelter records with data from local jails, state psychiatric hospitals and Medicaid, examining both how these systems impacted homelessness, and how homelessness impacted these systems. A major benefit was the ability to “monetize” these services to estimate the costs of homelessness. By 2000, my colleagues and I were using these data to study the impact of permanent housing interventions on homelessness trajectories, usage of other acute care systems, and costs (and cost offsets). This work led to a major federal program begun in 2004 to reduce “chronic homelessness” by the Bush administration. In 2000, the Congress also mandated that all cities in the US were to adopt administrative data systems to track homeless services and outcomes as a condition of federal funding.
Recognizing the time and resource intensive work required to access these various data systems (some taking up to four years to negotiate), I worked with John Fantuzzo at Penn to develop a systematized approach to getting more ready access to these data through the creation of the Kids Integrated Data System (KIDS) in Philadelphia. Originally intended to study service needs and use in early childhood, the system was quickly expanded to include all children and youth serving agencies, as well as adult service systems. We began to network with other communities doing this work, including the states of Washington and South Carolina, Allegheny (Pittsburgh) and Cuyahoga (Cleveland) counties, and the pioneering work of colleagues at Chapin Hall at the University of Chicago.
In 2008, the MacArthur Foundation approached us to formalize a network of communities with integrated data systems, both to learn and establish best practices, and to promote the growth and use of such systems across the country. The Actionable Intelligence for Social Policy initiative at Penn became the means by which we established and grew this network. Today the network includes 43 states and local governments who we convene regularly with core support from the Annie E. Casey Foundation.
Q: How did you get involved with IPDLN?
A: I learned about IPDLN through research into the nature and scope of IDS work in Europe, in particular through the work of David Ford at Swansea University. Upon learning of IPDLN, I started to attend their biennial conferences. Michael Schull of the University of Toronto invited me to join the board of IPDLN.
Q: How do you see the future population data linkage?
A: Population data linkage is going to be at the forefront of social policy analysis, planning and evaluation in the very near future. As the systems become more embedded in government and at universities, their use will grow and become more firmly established. I expect that the academic and policy literatures will increasingly feature administrative data, which is already rapidly growing in health services, homelessness, and education research.