This thesis studies the 1998-2003 doubling of the National Institutes of Health budget to evaluate how assertions about the impact of research investments compare with actual health and research outcomes. Stakeholders in the doubling noted a variety of outcomes intended to result from the effort. Using public value mapping (Bozeman and Sarewitz, 2005), I have compared stakeholders' stated intentions of what the doubling ought to achieve with the health and research outcomes actually produced. In applying public value mapping, I first conducted interviews and reviewed press releases, Congressional record, news, and other data from the doubling period. Six public values were commonly represented in this data: (1) improving health outcomes (2) reducing the cost of healthcare (3) producing application-relevant knowledge (4) building biosecurity and biodefense capabilities (5) developing the research enterprise (6) economic growth I then inferred causal logic chains by which increasing funding could lead to achievement of the public values and identified four investment intermediaries through which funding would pass in advancing public values. Finally, using proxies, I evaluated if the public values had advanced in a way directly attributable to funding increases. This analysis identified (5) as achieved. (1), (3), (4), and (6) were indeterminate in one of the two components necessary for evaluating public value achievement: either no clear advancement or no direct link between outcomes and doubling investments. (2) was a failure due to the increase in healthcare costs throughout and following the doubling period. These results indicate that complex societal outcomes used to justify incremental research investments are challenging to causally attribute to those same investments, and thus uncertain premises on which to base policy.