The rise of the US health-care sector over the past several decades has been remarkable. As Figure 1 [page 3, top] shows, in 1970, the country devoted slightly more than 6 percent of GDP to health care, about 1 percent more than other nations. Today, the nation devotes almost 18 percent of GDP to health care, which is larger than spending on cars, clothing, food, furniture, housing, fuel, and recreation combined--and is a full 8 percent above the average in comparable countries.
Health outcomes haven't kept up, as shown in Figure 2 [page 2, bottom left]. US life expectancy was slightly below the average of comparable countries in 1980. Today it has fallen far below that of these other countries, with life expectancy actually declining for the first time in decades.
These striking facts have motivated a sharp increase in the quality and quantity of work in the NBER Health Care Program. From a handful of working papers in 1992, this program has grown to produce an average of more than 100 working papers a year in the last three full years. These papers reflect the larger interest of the economics profession in health issues. In 1990, the American Economic Review published just two articles about health; now it publishes about five a year. In the American Economic Journals in Economic Policy and Applied Economics, major new general-interest journals that cover health topics, about one in eight articles published in 2017 focused on health. The Health Care Program has expanded and drawn in a new generation of health economists.
[Text incomplete in original source.]tor, with 674 working papers posted in the program since 2012. These studies have covered a broad array of topics, and it is impossible to do them justice in this short review. Instead, I will highlight a few key areas of study by NBER researchers, with apologies to the large number of authors of studies that I am excluding.
The Affordable Care Act
The ACA is the most significant government intervention in the US health-care system since the introduction of Medicare and Medicaid. Moreover, it was introduced both in a data-rich environment in which many datasets can be used to analyze its impacts, and in a manner that generated quasi-experimental variation that can be used to convincingly estimate those impacts. In particular, the enormous expansion of the Medicaid program to all those whose income is less than 133 percent of the poverty line, which occurred only in a subset of states and over time in those states, provides a natural case study for understanding the impact of expanded insurance coverage. This has provided a wonderful environment for economic research.
Health Care Program affiliates' research on the ACA has covered a wide variety of areas, but has focused primarily on the impacts of the ACA on insurance coverage, health-care utilization, and health, as reviewed by Benjamin Sommers and me. (1) Studies show that the ACA clearly has expanded coverage [Figure 3] through provisions such as extending coverage of dependents up to age 26, (2,3) expanding Medicaid, (4) and subsidizing premiums in the new exchange. (5) Notable is the finding of that last paper that much of the increase in Medicaid enrollment was not from those who were newly eligible, but from those previously eligible who had now enrolled in the program.
There has also been a clear increase in health-care utilization in response to broadened insurance coverage. (6) Early studies have generally found positive impacts of the ACA on population health, but more work is needed to assess the long-term impacts on health. (7)
A particularly notable area of research on the ACA has been focused on the impact of the law's provisions on labor market behavior, with mixed results. Research on a large restriction on health insurance coverage in Tennessee before the ACA showed an associated significant rise in labor force participation, suggesting that expansions under the ACA might reduce the supply of labor. (8) But studies of both the expansion of insurance to young adults (9) and the overall effects of the ACA exchanges and Medicaid exchanges (10) do not find significant impacts on labor supply.
A common refrain in health economics is that the most expensive piece of medical technology is the physician's pen, yet there is relatively little understanding of the physician behaviors that drive medical spending. A set of recent papers has made enormous progress in helping us understand physician decision-making and its implications for the health-care system.
One of the enduring mysteries in health care is the enormous variation among physicians in treatment styles. These differences emerge in physician training. (11) David Cutler, Jonathan Skinner, Ariel Dora Stern, and David Wennberg use surveys of physicians to show that much of the variation reflects physician beliefs unsupported by clinical evidence. (12) There is mixed evidence on the welfare implications of physician treatment variation. Gautam Gowrisankaran, Keith Joiner, and Pierre-Thomas Leger find that physicians randomly assigned to different emergency department doctors who are more skilled see higher resource use, but not necessarily better outcomes. (13) In contrast, Janet Currie, W. Bentley MacLeod, and Jessica Van Parys find that for heart attack patients, there is large variation in treatment intensity across providers, and those who treat more intensively deliver better outcomes. (14)
A related question is whether more information provided to patients can improve outcomes and performance. Jonathan Kolstad finds that when "report cards" were introduced on surgeon outcomes in Pennsylvania, surgeons responded strongly to poor performance relative to their peers, suggesting a strong role for "intrinsic motivation." (15) At the same time, Erin Johnson and M. Marit Rehavi, (16) and in another study, Michael Frakes, Anupam Jena, and I find that when physicians are themselves patients, they receive a quality of care similar to that of comparable non-physician patients. (17)
surgeon outcomes in Pennsylvania, surgeons responded strongly to poor performance relative to their peers, suggesting a strong role for "intrinsic motivation." (15) At the same time, Erin Johnson and M. Marit Rehavi, (16) and in another study, Michael Frakes, Anupam Jena, and I find that when physicians are themselves patients, they receive a quality of care similar to that of comparable...