The Rise of Healthcare Jobs

This is my summary of Professor Neale Mahoney’s (Stanford University) keynote at the 4th Health Economics Conference at TSE.

In his lecture, he presented joint work with Joshua Gottlieb, Kevin Rinz, and Victoria Udalova titled “The Rise of Healthcare Jobs”. Using restricted-access Decennial Census and American Community Survey microdata spanning 1980 to 2022 alongside NPPES practitioner registries and MarketScan claims, the study analyzes the structural growth, wage distributions, demographic composition, and geographic reallocation of the United States healthcare workforce.

1. Macro trends and counter-polarization in the labor market

Healthcare has steadily expanded into the single largest employment sector in the United States, surpassing 19 million workers and exhibiting recession-proof, linear growth over four decades. Across the broader macroeconomy, labor markets have experienced hollowed-out middle-skill employment driven by automation and deindustrialization. Healthcare diverges sharply from this pattern: real wage growth has been concentrated between the 50th and 95th percentiles of the national income distribution, driven primarily by registered nurses. Semi-parametric Dinardo-Fortin-Lemieux reweighting demonstrates that this real wage growth is not an artifact of compositional changes in education, but represents genuine real wage gains within clinical middle-class occupations.

2. Workforce demographics and gender convergence

The healthcare sector has remained 80 percent female in aggregate since 1980, but within-sector occupational sorting has changed substantially. Female representation among physicians has risen rapidly from single digits to over 40 percent of active doctors and more than half of current medical school graduates, flattening the historic gender gradient across clinical roles. Simultaneously, the foreign-born worker share exhibits a U-shaped distribution across skill levels: foreign-born labor is heavily represented among entry-level aides and licensed physicians, while middle-tier clinical positions remain predominantly domestic.

3. Supply constraints and the rise of midlevel practitioners

Severe statutory and institutional bottlenecks in physician training—specifically binding caps on medical school admissions and residency funding—have restricted physician growth to two percent annually, far below the healthcare demand generated by the doubling of the elderly population. In response, healthcare delivery has substituted toward midlevel clinicians, including Nurse Practitioners (NPs), Physician Assistants (PAs), and Certified Registered Nurse Anesthetists (CRNAs), whose employment grew at 6.7 percent annually from 2010 to 2022. Midlevels require 6 to 8 years of training compared to 11 to 15 years for physicians, allowing lower-cost and faster workforce deployment. MarketScan claims reveal that midlevels deliver an increasing share of care, including high-severity anesthesia and primary care visits, particularly across underserved rural areas and Appalachia where physicians are scarce.

To evaluate whether state-level policy drove this supply expansion, the authors estimate an event-study difference-in-differences specification evaluating state scope-of-practice (SOP) deregulation:

\[NP_{ist} = \sum_{\tau=-6}^{6} \beta_{\tau} (SOP_{s} \times Year_{t-\tau}) + \gamma_{s} + \lambda_{t} + \Theta X_{ist} + \epsilon_{ist}\]

The econometric estimates show that SOP reforms explain only 4.83 percent of the increase in NP claims within treated states and a negligible 0.41 percent of the national rise in NP utilization. The midlevel boom is a national structural response to physician shortages rather than a localized regulatory artifact.

4. Empirical test of the “manufacturing to meds” reallocation

The lecture directly evaluates the common economic narrative that industrial regions successfully replaced lost manufacturing employment with healthcare jobs. Regressing regional changes in healthcare employment against historical manufacturing shares across commuting zones yields an offset rate of approximately eleven percent. Because healthcare represents roughly 10.8 percent of baseline employment, healthcare grew at its ordinary baseline share without disproportionately clustering in deindustrialized markets. While the transition to healthcare employment provided stronger offsets for Black and female workers (30 to 35 percent), male manufacturing workers did not retrain into nursing or clinical care, demonstrating that healthcare institutions do not serve as localized employment engines capable of reversing Rust Belt economic decline.