Federal investments in research strongly impact the development of the country’s scientific workforce. A significant percentage of federal research spending takes place at universities, who recruit researchers at different levels such as graduate students, postdoctorates, and faculty. The National Institutes of Health (NIH) is the largest funder of non-classified research in the U.S. federal government. Between 1998 and 2003, Congress doubled NIH’s budget to over 30 billion dollars in order to boost discoveries in biomedicine and foster the development of a strong research workforce.

While in 1997 the increase in NIH budget was unanimously supported by Congress, there is current evidence that the rise in funding has had unintended effects on the research community. Studies demonstrate that the funding increase resulted in a much steeper rise in the number of grant applications, causing success rates to drop from 32% in 1998 to 21% in 2007. In turn, the drop in success rates has fueled competition and resulted in researchers spending more time writing grant proposals, reducing the time spent in actual research. Increased competition has also lowered the percent of early-career NIH grant awardees from 23% in 1998 to 15% in 2005, negatively affecting the development of the workforce. Looking back at the political process that led to NIH’s budget doubling suggests that the funding policy lacked support of a formal analysis outlining the most advantageous resource allocation and evaluating potential long-term effects in the system. The current study addresses this issue.

We build a differential equation-based dynamic model of the research workforce. The model includes variables for different stages in the academic research career of researchers. It provides a framework for how research funds are spent and how they affect the workforce ecosystem. We use the doubling of NIH’s budget from 1998 to 2003 as a case study to calibrate our model. We then take a normative approach, and examine how the amount of research spending and speed of funding increases affect long-term research production.

Our calibrated model replicates the historical trend of NIH funding, research workforce dynamics, and NIH grant success rates. Our simulation suggests: 1. A temporary and steep rise in research funding, in the long term, can result in reduced grant success rates and increased dropout rates from the workforce pipeline. 2. A smoother funding expansion is critical in improving research workforce development and results in a higher level of research production. 3. Funding expansion, and the subsequent growth in the pool of qualified professionals, should be met with a planned growth in the future downstream capacity, i.e., universities’ ability to grow their number of tenure-track positions. The study brings insights into designing an improved decision-making system for research spending in biomedicine as well as other research domains.