From 1996 to 2002, eligible physicians voted approximately 9 percentage points less than the general population.1 Since then, physician voter engagement has not been reported. We investigated physician voter participation, voter registration, and voter turnout from 2006 through 2018 in California, New York, and Texas, which are states with the largest number of physicians.
We merged the National Provider Identifier (NPI) registry with state voter files from L2, a nonpartisan political data corporation,2 and identified physicians registered to vote in general elections by using a matching process based on names, NPI enumeration dates, dates of birth, and occupational data from commercial reports. We identified the number of active physicians using state workforce profiles from the Association of American Medical Colleges. We obtained general population voting data and determined eligibility using the Voting and Registration Supplement to the Current Population Survey3 by the US Census Bureau and Bureau of Justice Statistics.4
The primary outcome was voter participation (ie, the proportion of physicians who voted among those eligible to register) compared with the general population using χ2 test analysis. The secondary outcomes were voter registration and voter turnout (ie, the proportion of registered physicians who voted). We modeled voter turnout using logistic regression, adjusting for age, sex, race/ethnicity, education, income, household size, and conducted sensitivity analyses. We used marginal effects to convert odds ratios into adjusted percentages. We estimated pooled and election year results.
This study was considered exempt by The University of Texas Southwestern Medical Center Institutional Review Board, and informed consent was not required owing to use of publicly available information. Voter file matching and analysis details are available in the Supplement.
We identified 112 032 physicians registered to vote in 2018: 50 854 in California, 39 046 in New York, and 27 578 in Texas. A total of 73 893 physicians (66%) were male; 51 570 (46%) were White, 9837 (9%) were Hispanic, and 3001 (3%) were African American. A total of 45 133 (40%) were baby boomers (individuals born between 1946 and 1964), and 39 668 (36%) were generation X (individuals born between 1965 and 1979). Primary care was the largest specialty with 49 550 (44%) physicians.
Physician voter participation and registration were lower than the general population for all elections from 2006 through 2018 (P < .001; see Figure 1 for supporting data). Pooled physician voter participation was 14 percentage points lower than the general population (486 671 [37%] vs 182 982 000 [51%]; P < .001), driven primarily by differences in pooled voter registration (670 489 [50%] vs 236 244 000 [66%]; P < .001).
Unadjusted physician voter turnout was higher than the general population for all election years (2018: 83 890 [75%] vs 22 168 398 [62%]; P < .001; and 2006: 40 995 [54%] vs 14 799 137 [48%]; P < .001), and adjusting narrowed the gap to 1 to 7 percentage points higher than the general population (2018: 76 243 [69%] vs 22 175 064 [62%]; P < .001; and 2006: 37 416 [49%] vs 14 804 302 [48%]; P < .001; Figure 2). Findings remained consistent after adjusting for political party.
From 2006 through 2018, voter participation among eligible physicians in California, New York, and Texas was 14 percentage points lower than the general population. This is similar to research from 1996 through 2002.1 Half of eligible physicians were not registered to vote, even though wealthier, more educated voters generally have at least 50% higher registration rates.5 However, after adjusting for characteristics of physicians that are associated with turnout, registered physicians had narrower but still significantly higher turnout than the general population. The reason for this pattern of physician voter engagement is unclear, but low participation may be because of the fear of seeming political while practicing medicine, in addition to other administrative and psychological barriers.6 It is unclear if unregistered physicians would have comparable voter turnout if they were to become registered.
Limitations of this study include potential mismatch of physicians from the NPI registry to state voter files, residual confounding owing to unmeasured factors, and uncertain generalizability to other states. Future efforts to improve physician voter participation should explore the influence of both increasing voter registration and election turnout.
Accepted for Publication: October 3, 2020.
Published Online: October 22, 2020. doi:10.1001/jamainternmed.2020.6887
Corresponding Author: Arthur S. Hong, MD, MPH, The University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas TX, 75390-9169 (arthur.hong@UTSouthwestern.edu).
Author Contributions: Dr Hong had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Lalani, Johnson, Hong.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Lalani, Johnson, Maddineni.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Maddineni, Hong.
Obtained funding: Lalani, Johnson.
Administrative, technical, or material support: Lalani, Johnson, Halm, Hong.
Supervision: Johnson, Halm, Hong.
Conflict of Interest Disclosures: None reported.
Funding/Support: This research was supported by the Seldin Scholar fund from the Department of Internal Medicine at The University of Texas Southwestern Medical Center.
Role of the Funder/Sponsor: The funding organization had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approvals of the manuscript; and decision to submit the manuscript for publication.