Much of health services research seeks to inform particular policy choices and is best characterized as policy-driven research. The reverse, research-driven policy, occurs when studies alter how people perceive reality, which eventually leads to new policy. An example of the latter is nearly four decades of work by John Wennberg and colleagues. Observing variations in practice across small geographic areas led to the notion that some care is preference sensitive, whereas other care is supply constrained. For the former, patient, rather than physician, preferences should be honored, after acquiring and effectively communicating the best available information on the benefits and risks of treatment options. Finding that areas with high use of services have no better quality or outcomes than do areas with lower use led to the notion of accountable care organizations (ACOs). Eventually, both patient engagement and ACOs were written into the Affordable Care Act of 2010.