Abstract
This study examines the number of unique unions and characteristics of unionization elections among physicians in training in the US.
Physicians in training (interns, residents, and fellows) with government employers have always had the right to unionize, whereas those with private employers regained it in 1999.1 Given recent concerns about work conditions, burnout, and patient safety, physicians in training appear increasingly interested in collective activism,2 but little is known about their unionization efforts.3 We assessed the number of unique unions and characteristics of unionization elections among physicians in training.
Methods
Labor unions represent a bargaining unit—employees with common interests and work responsibilities—against a specific employer, which may include more than 1 hospital or health care system and physicians in training from multiple programs. Physicians in training can unionize through state employment relations boards (ERBs; if a government employer) or the National Labor Relations Board (NLRB; if a private employer). To determine the number of existing physicians-in-training unions, on July 14, 2023, we web scraped all internal medicine, family medicine, pediatrics, and psychiatry programs accredited by the Accreditation Council for Graduate Medical Education (ACGME) and identified unique unionized bargaining units at public and private institutions through review of institutional websites, press releases, news reports, and union contracts (eAppendix in Supplement 1). We chose these specialties because they contain the most physicians in training, cover all patient age ranges, and capture stand-alone institutions. We characterized location, employer ownership, presence of right-to-work laws (prohibiting requirements for all individuals of the bargaining unit to pay union dues),4 and representing labor organizations. To evaluate completeness of capture, we compared identified bargaining units represented by the Committee of Interns and Residents, Service Employees International Union (CIR/SEIU)—the largest physicians-in-training labor organization—with CIR/SEIU’s published list; 47 of 48 (97.9%) unions were captured.
State ERBs do not publicly publish their union election results. Therefore, to examine unionization activities, we used NLRB’s election reports, limited to private institutions, from January 2011 through June 2023; reports before 2011 did not allow identification of physicians in training. Using keyword search and manual evaluation, we selected union elections pertaining to physicians in training, nurses, and other health care workers (Supplement 1). We calculated the proportion of eligible individuals who voted (voter turnout), proportion of voters who voted to unionize, outcome, margin of victory or defeat in each election, and overall union win percentage. Characteristics examined included representing labor organization, location, bargaining unit size, and presence of right-to-work laws. Analyses were conducted using RStudio version 2023.03.0. This study was approved by the Harvard Pilgrim Health Care institutional review board.
Results
We evaluated 1909 programs from 1188 public and private institutions, identifying 67 unique physicians-in-training unions against their employer. Most unions were in non–right-to-work states (n = 65; 97.0%), were against private employers (n = 40; 59.7%), and were represented by CIR/SEIU (n = 61; 91.0%). California (n = 22) and New York (n = 19) had the most unions.
Of 15 366 union elections through the NLRB since 2011, 18 were among physicians in training (Table), occurring in 6 of 13 years, with most in 2023 (n = 7). A mean (SD) of 17 (6) elections occurred yearly among nurses, and 64 (27) among other health care workers; 2017 had the most elections for nurses (n = 26) and other health care workers (n = 119). CIR/SEIU represented physicians in training in 16 elections (88.9%). All physicians-in-training elections occurred in non–right-to-work states.
Table. Union Elections Among Physicians in Training With Private Employers, 2011-2023a.
| Yearb | Employerc | Eligible voters | Voter turnout, %d | No. (%) | Election result | Labor organization | |
|---|---|---|---|---|---|---|---|
| Votes for union | Votes against union | ||||||
| 2013 | Dignity Health–California Hospital Medical Center | 25 | 88.0 | 22 (100) | 0 | Win | CIR/SEIU |
| 2014 | Beth Israel Medical Center | 417 | 75.3 | 110 (35.0) | 204 (65.0) | Loss | CIR/SEIU |
| Icahn School of Medicine at Mount Sinai at Elmhurst Hospital Center | 142 | 82.4 | 65 (55.6) | 52 (44.4) | Win | CIR/SEIU | |
| 2015 | Howard University Hospital | 263 | 82.1 | 110 (50.9) | 106 (49.1) | Win | CIR/SEIU |
| St Mary Medical Center (Long Beach) | 37 | 43.2 | 9 (56.2) | 7 (43.8) | Win | CIR/SEIU | |
| Palisades Medical Center | 79 | 79.7 | 47 (74.6) | 16 (25.4) | Win | CIR/SEIU | |
| 2018 | Bayonne Medical Center | 55 | 78.2 | 14 (32.6) | 29 (67.4) | Loss | CIR/SEIU |
| 2022 | Greater Lawrence Family Health Center | 40 | 97.5 | 29 (74.4) | 10 (25.6) | Win | CIR/SEIU |
| Keck School of Medicine of USC | 56 | 94.6 | 53 (100) | 0 | Win | CIR/SEIU | |
| Stanford Health Care | 1450 | 72.3 | 835 (79.6) | 214 (20.4) | Win | CIR/SEIU | |
| University of Vermont Medical Center | 347 | 77.2 | 209 (78.0) | 59 (22.0) | Win | CIR/SEIU | |
| 2023 | California Pacific Medical Center | 109 | 77.1 | 71 (84.5) | 13 (15.5) | Win | CIR/SEIU |
| George Washington University | 420 | 64.0 | 253 (94.1) | 16 (5.9) | Win | CIR/SEIU | |
| Loma Linda University Health Education Consortium | 805 | 62.7 | 361 (71.5) | 144 (28.5) | Win | UAPD | |
| Mass General Brigham | 2304 | 70.6 | 1215 (74.7) | 412 (25.3) | Win | CIR/SEIU | |
| Montefiore Medical Center (Moses Campus) | 1188 | 63.7 | 620 (81.9) | 137 (18.1) | Win | CIR/SEIU | |
| University Medical Resident Services PC (University at Buffalo) | 799 | 47.6 | 270 (71.1) | 110 (28.9) | Win | UAPD | |
| University of Pennsylvania Health System | 1394 | 71.9 | 892 (89.0) | 110 (11.0) | Win | CIR/SEIU | |
Abbreviations: CIR/SEIU, Committee of Interns and Residents, Service Employees International Union; UAPD, Union of American Physicians and Dentists.
Physicians-in-training union elections were identified through the National Labor Relations Board (NLRB) election reports, which provide a comprehensive list of union elections conducted by the NLRB each year. Physicians-in-training union elections were identified in 2 steps: (1) selecting election cases containing relevant keywords (interns, residents, resident physicians, fellows, or house staff); and (2) manual evaluation of all election cases provided by NLRB to identify any missed elections.
There were no NLRB union elections among physicians in training in 2011, 2012, 2016, 2017, 2019, 2020, and 2021; 2023 included only data through June.
The organization listed is the legal employer reported in the NLRB union election. However, the election referenced may involve multiple Accreditation Council for Graduate Medical Education–sponsoring institutions that are legally registered under the same employer.
Voter turnout is the proportion of individuals who voted divided by the number of eligible voters.
Mean voter turnout for physicians in training was 73.8% (range, 43.2%-97.5%) vs 84.7% (range, 49.1%-100%) for nurses and 78.8% (range, 12.5%-100%) for other health care workers. Physicians in training elected to unionize in 16 of 18 elections (88.9%) vs 83.6% among nurses and 77.0% among other health care workers. Margin of victory for physicians in training ranged from 0.9% to 50.0%; margin of defeat ranged from 15.0% to 17.4%.
Discussion
Most ACGME-sponsoring institutions in internal medicine, family medicine, pediatrics, and psychiatry did not have physicians-in-training unions as of July 2023. Among private union elections since 2011, participation was variable, but most voted to unionize. Relative to other groups in health care, physicians in training had a higher percentage of union election wins but lower voter turnout, notable because NLRB determines election results based on votes cast. Most unions were located in non–right-to-work states, consistent with literature documenting low union density in such states.5
Study limitations include potential underestimation of unionization efforts if unionized physicians in training were missed during review or employers voluntarily recognized a labor union, bypassing the need for elections. Unionization activities reflected only 2011 through mid-2023 and physicians in training with private employers.
Research is needed to understand why physicians in training decide to unionize and the effect of unionization on physicians in training, training programs, and patients.
Section Editors: Jody W. Zylke, MD, Deputy Editor; Karen Lasser, MD, and Kristin Walter, MD, Senior Editors.
eAppendix. Supplemental Methods
Data Sharing Statement
References
- 1.Kobdish RM. National Labor Relations Board rules that interns and residents may unionize. Health Law. 1999;12:18. [Google Scholar]
- 2.Lin GL, Ge TJ, Pal R. Resident and fellow unions: collective activism to promote well-being for physicians in training. JAMA. 2022;328(7):619-620. doi: 10.1001/jama.2022.12838 [DOI] [PubMed] [Google Scholar]
- 3.Ahmed AM, Kadakia K, Ahmed A, Shultz B, Li X. Trends in labor unionization among US health care workers, 2009-2021. JAMA. 2022;328(24):2404-2411. doi: 10.1001/jama.2022.22790 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Ichniowski C, Zax JS. Right-to-work laws, free riders, and unionization in the local public sector. J Labor Econ. 1991;9(3):255-275. doi: 10.1086/298268 [DOI] [Google Scholar]
- 5.Ozkan E, Ozbeklik S. What do right-to-work laws do? evidence from a synthetic control method analysis. J Policy Anal Manage. 2016;35(1):173-194. Accessed July 14, 2023. http://www.jstor.org/stable/43866565 [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eAppendix. Supplemental Methods
Data Sharing Statement