What Data Can – And Can’t - Do
Columnist David Brooks recently wrote a provocative piece, “What Data Can’t Do,” in the New York Times. As the leader of a data-rich health policy research institute, I paid close attention. Brooks lists several issues that explain why data are constrained and limited in how they can help solve today’s policy and social problems.
In the light of Colorado’s Medicaid Expansion debate and policy decision, Brooks’ column resonated with me. The Colorado Health Institute has been asked by dozens of individuals and organizations to “tell us what the data say we should do.” If only this were an easy task.
Data – with its numeric and quantifiable characteristics – suggests that one answer can be retrieved. One cost estimate, one savings estimate, one precise number of how many lives can be improved or saved. To those not buried deep in analytics, data suggests this precision. To those of us working day in and day out with data, we know answers are elusive and complex.
What resonates most with me is Brooks’ contention that data obscures values. This is certainly true in the Medicaid Expansion debate. At the heart of the matter are several questions that are not based on the data:
- What’s the role of government in providing insurance to the working poor?
- Is public insurance a better vehicle than private insurance to provide this coverage?
- What’s the role of personal responsibility?
- What’s the state’s role in creating healthy communities?
These questions reflect our values – and data, no matter how big or how precise, will never adequately address these issues.
Policy decisions are informed by our values - and our data. They are different animals and we need to consider both.