The U.S. government spends large sums on biomedical research: NIH, the single largest funder of biomedical research in the United States, has a budget of over $40 billion in 2021, far larger than any other federal research funder. The core purpose of this funding is to generate knowledge that leads to inventive new treatments and breakthroughs across the medical spectrum. But just how much of a health benefit does this budget generate for local populations? A new study attempts to find out.
What We Know On Research Impacts So Far
Most of research on funding impacts has focused on “easily” quantifiable metrics like papers and patents, or on the pharmaceutical products approved. The results of these studies are generally favorable – for instance, one study found that a funding boost of $10 million at NIH generates an additional 2.3 biomedical patents, while another found that government funding has played an indirect role in nearly two thirds of priority review drugs. While this work has merit, it doesn’t identify the actual impact on disease treatment and quality of life improvements. Other attempts have been made to quantify the connection between research, medical progress, and health on a large scale. For example, it has been estimated that the decrease in mortality made possible by medical progress in the 20th century saves us $3.2 trillion a year. While this number is impressive, it is based on all medical progress – which is not terribly useful when trying to make granular decisions on research policy.
On the other hand, breaking medical progress into finer detail, such as narrowing the inquiry in terms of grants and particular disease categories, can allow for greater insight into the correlation of biomedical research spending and actual human health impacts. A recent working paper published by the National Bureau of Economic Research, “Does Research Save Lives? The Local Spillovers of Biomedical Research on Mortality,” attempted just such an analysis, looking at how NIH grants impact the local community. The authors, Rebecca McKibbin of the University of Sydney and Bruce A. Weinberg of Ohio State University, sought a different approach from prior work by looking not at patents or pharmaceuticals, but instead at the morbidity of patients in the same geographic area as the grant. This enables analysis of the connection between NIH grant funding and the research’s life-saving impact.
A core insight underlying the study is that grants are allocated to a specific researcher who has location, patient and resource constraints. That means that innovations are first implemented locally, and then spreads outward as researchers share their knowledge with their networks, both across their field and in their local area. This idea is grounded in previous literature, for example when Agha and Molitor noted patients physically located in the region of a new cancer drug discovery were much more likely to receive it in its first two years following FDA approval.
The Impact of Grants on Morbidity
The authors’ analysis involves a few steps. First, the authors calculated life years lost to specific diseases: they examined the deaths recorded by hospitals per cause over the years, and subtracted the age of death from average life expectancy overall. Then they matched those causes of death to general disease research categories, and created a link between the topic of each NIH grant and biomedical publication to the disease category. With these data sets in hand, they were able to compare the life years lost to a disease against NIH grants received or articles published in the geographic region surrounding a hospital system.
To analyze the data, the authors started with the geographic distribution of grant funding. They found that in the event of a funding windfall – either a ‘small’ one like increased funding for a particular disease or a blanket increase in NIH funding like ARRA – some institutions consistently received increased funds as compared to others. They attribute this in part to the Matthew Effect – the concept that the rich get richer through accumulated advantage. However, they note that over time the health impact of that funding does spread geographically, as the local advantage to health outcomes dissipates and the innovation is adopted by doctors across the country. Patients near the hospital where the discovery was made no longer have an increased survival advantage when other hospitals start implementing that discovery too – now everyone gets to live a little longer.
After exploring funding and morbidity, they looked at the correlation of deaths and scientific publications, and found that deaths were also more delayed in the geographical location of published findings on the disease, the same finding as funding. This implies that publications, which are funded by NIH grants, have a perceptible relationship with the lives of the people who live near the researchers and the clinicians they share their findings with. As one can imagine, the spread of these innovations is not immediate. First, the researcher’s close network of clinicians learn of the new discovery, and then the clinician’s broader networks, and so on – the authors calculate it takes about 10 years for a new innovation to be pervasive across the US, though it can vary by disease area. This suggests that a more even geographic distribution of funding may have a direct impact on the lifespans of Americans.
Lastly, the authors attempted to quantify the impact of funding on mortality by analyzing the amount of funding per disease as well as the impact of publications on the disease of study. They estimate that each 1% increase in local funding decreases mortality from the disease in the local region by 0.22%. Based on this, they translate this to suggest that every $100,000 increase in funding generates an estimated 401 life years saved. When it comes to publications, the average peer-reviewed article generates about 294 life years saved.
To be sure, there may be other factors at play in these findings. These may include local resource density as well as socioeconomic, ethnic, and gender factors that the authors have not yet touched upon, but could conflate some of the impacts. While these variables should be investigated in the future, the current findings nevertheless have implications for future NIH and other biomedical funding. In sum, the knowledge spillovers from biomedical funding appear to contribute to a decrease in local mortality, which eventually spreads to encompass all areas as the new discovery is implemented. This adds an additional consideration when allocating funding across disease field and geographic lines, as well as when setting the NIH budget overall.
 Calculated in terms of gains in life expectancy by medical advances - both decreasing mortality and increasing quality of life – and applies the economic value of a life-year to the total.
 Cancer appears to be the slowest to adopt new experimental techniques, perhaps because it has the more incremental improvements to morbidity. It’s time to uptake is longer than 10 years.