Skip to main content


Log in

Medical education reforms and the origins of the rural physician shortage

Cliometrica Aims and scope Submit manuscript

Despite an increased production of the total number of physicians, a persistent geographic maldistribution of physicians has characterized the past 70 to 80 years...The opportunity for medical education in this century is to recapture the diversity and relevance of distributed training, even as patient care, education, and research are further improved. Distributed medical education that is uniquely adapted and responsive to the needs of rural underserved communities has the potential to reclaim medicine’s social contract with the public. (American Academy of Family Physicians 2013).


In the first two decades of the twentieth century, medical schools increased standards for admission and added basic science to their curricula. During this time period, the probability a new medical school graduate located in a rural area declined by 40%. Using novel data from the American Medical Directories, we find that physicians trained in more rigorous programs with higher admission standards were less likely to set up practice in rural areas. While all physicians were being drawn to metropolitan areas during this period, the pull was stronger for graduates of the higher quality schools. We also find some evidence that physicians trained in the more scientifically and clinically based programs were more strongly attracted to places with more hospitals. These findings suggest that the medical education reforms of the early twentieth century contributed to the urban–rural disparity in access to physician care.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1

Source: “Medical Education in the United States,” Journal of American Medical Association, Aug. 27, 1932: p. 746

Fig. 2

Source: Pre-medical college requirements are from American Medical Association Council on Medical Education (1919, 1923). Medical schools and graduates from American Medical Association Council on Medical Education (1905–1910; 1911–1914; 1915–1920)

Fig. 3

Source: Pre-medical college requirements are from American Medical Association Council on Medical Education (1919, 1923). Fees are from American Medical Association Council on Medical Education (1905–1910; 1911–1914; 1915–1920)


  1. Future research should examine how the rural–urban gap in physician quality affected health outcomes. The difficulty in obtaining adequate data at the county-level for the period before 1915, however, makes this a challenging research agenda. Studies conducted using state-level data do demonstrate a relationship between better-trained physicians and improved health outcomes. Law and Kim (2005) examine how changes in state-level physician licensure affect mortality using data from 1900, 1910, 1920, 1930 and 1940. They find an association between licensure and mortality for a small number of measured health outcomes, suggesting that physician quality may have improved health outcomes over time, even though existing physicians were grandfathered as licensure standards became more rigorous. Lichtenberg (2011) finds that in the 1990s, life expectancy increased more rapidly in states with higher fractions of physicians trained in top-ranked medical schools.

  2. In the early 1800s, local medical societies lobbied state legislatures to enact licensing laws and give local medical societies the power to license physicians. While many states enacted licensing laws, they were “...unwilling to enact laws that would have seriously deterred unlicensed practitioners” Rothstein (1972, p. 76). Moreover, these laws were generally unenforceable. As a result, physicians could easily found medical schools and augment their income with student fees.

  3. Professional standardization occurred earlier in Europe than in the USA. In France, medical education was centralized under the national government. Medical degrees were only awarded by a Faculté de Médecine, and students were required to complete a clinically based medical curriculum, pass a series of examinations, and write a thesis. Would-be physicians in Germany first had to pass a rigorous exam to get into medical school, follow a prescribed plan of study that was uniform across the country, pass several exams and complete a year-long hospital internship before being permitted to practice. In England, the Medical Act of 1858 created the General Medical Council, a central body dealing with medical education and licensure. The law and its later amendments sought to ensure uniform training, curriculum, and examinatios for students enrolled in licensing corporations (which focused on training by apprenticeship) and universities. A second act that further strengthened the first was enacted in 1886. For greater discussion, see Commission on medical education (1932).

  4. As described in the 1905 report from the AMA Council on Medical Education, an applicant to Johns Hopkins was required to have either “(a) completed the chemical-biologic course which leads to the A.B degree in the university” or “(b) graduated at an approved college or scientific school and can furnish evidence of an acquaintance with Latin and a fair reading of French and German, and a knowledge of physics, chemistry and biology, such as may be obtained from a year’s course, including laboratory instruction.” (p. 556).

  5. Medical educators recognized that raising admission standards may hurt them financially since their applicant pools would be smaller, but competition between them for prestige led them to do so anyway. In describing the University of Pennsylvania’s decision to raise entrance requirements, for example, Ludmerer (1985) states that the university recognized it would face a decrease in revenue from student fees, but that it was the right decision, and the school “...could no longer consider ‘the financial or commercial side of any question in reference to medical education.” Ludmerer (1985) notes similar decisions at other schools (p. 85).

  6. It is important to note, however, that this relationship may not be due to the constraints of the legislative restrictions but rather to the changes in the admissions standards of medical schools that predated, and perhaps, led to the new laws. As the narrative above describes, many medical schools adopted more rigorous curricula and standards before state laws required them to do so. Law and Kim found that two types of state laws had negative effects on the numbers of physicians, pre-medical school educational requirements and 4-year medical school course requirements. Baker (1984) notes that most medical schools adopted 4-year programs before state laws were enacted that required physicians to attend such programs. He goes on to argue that, “Four-year legislation enacted after 1900 may have been aimed at the stragglers, or may simply have been redundant confirmation of existing practice” (p. 191).

  7. The same observer suggested that the reliance of medical education on hospitals and new laboratory techniques overemphasized rare and unusual diseases, with the result that students may have been ill-prepared to “ with the more usual types of illness which are so prominent in everyday practice” (Commission on Medical Education 1932, p. 114).

  8. Our discussion closely follows the simple model presented for migration choices in Wozniak (2010), but is adjusted to capture the initial location choices of newly minted graduates of medical school.

  9. See Gaynor (1994) for survey of the literature on the competitive nature of physician services.

  10. The model can easily be extended to include future migration by newly minted doctors, or to include the migration decisions of established physicians. For simplicity, we do not directly investigate migration decisions to focus on the first-order issues. However, the results from the analysis of equilibrium outcomes include any migratory behavior undertaken or planned for by physicians.

  11. Other potential measures of medical school quality suffer from a number of limitations, and are all highly correlated with the adoption of pre-med college requirements. The AMA introduced its rating system for medical schools in 1907, and thus this system is not applicable to physicians in our sample who graduated prior to that date. Additionally, while there is significant variation in ratings in the early years, by 1918 an A-rating had become synonymous with requiring pre-medical college education. In fact, at no time in our dataset (1905–1920) did a B or C rated school ever require college education for admission. In any event, the inclusion of AMA ratings in regressions does not substantively change the coefficient on pre-medical education, nor provide much additional explanatory power. The indicator variable for attending a medical school that closed by 1923 appears to be supplying the same information contained in the AMA ratings; the two variables are highly correlated. We report results from these regressions in Appendix Table 12. We choose not to use the Flexner report descriptions as a measure of quality because the report only provides his general impression of the worthiness of each medical school’s program, not a set of systematic ratings. Moreover, the Flexner report consists of a snapshot taken in 1909, which does not capture important changes in quality over time, nor does it provide information for graduates prior to 1909.

  12. The AMA published the first AMD in 1906 with the subtitle, “A Register of the Legally Qualified Physicians of the United States and Canada.” The AMA compiled this first listing by consulting the records of state licensing boards, medical colleges, and medical societies. The foreword notes that the goal was to include only “legally qualified” physicians.

  13. Results are robust to including term length in our model and the model fit does not appreciably change. We choose not to include a measure of fees because it appears that medical school fees mainly capture cost-of-living differences across the states. Moreover, we believe that the fees variable mechanically picks up the secular increase in both the probability of facing pre-med college requirements and the probability of urban practice. As evidence of this, the implied effect of nominal fees on physicians’ location decisions is too large in economic terms. For example, average nominal fees increased by 30% from 1905 to 1920, but nominal GDP per capita increased by 2.5 times and real GDP per capita increased by 13% (Sutch 2006). The real value of fees declined over much of the period because of war-related inflation.

  14. Appendix Figure 4 provides an example of one page of the AMD.

  15. Estimates are calculated using the 1% IPUMS samples (Ruggles et al. 2015). Incorporated cities and towns with below 10,000 population are grouped with “unincorporated areas”, which are places that were outside municipal jurisdictions. “Unincorporated” areas tended to be sparsely populated, and the Census Bureau classified them as rural in their population counts.

  16. Results reported throughout the paper are robust to using a population measure of 2500. Results are also robust when running our models using a county-based classification scheme based on the concept of metropolitan areas (a county with a large city as its economic center). We use the IPUMS definition of metropolitan counties, and divide rural (non-metropolitan counties) into those adjacent to metropolitan areas or not adjacent to metropolitan areas).

  17. Modern research suggests that physicians trained in rural areas may be more likely to return to rural areas (Rosenblatt and Hart 2000). During the early twentieth century, medical education was not urbanizing as fast as physicians. Many of the large state-funded medical schools were located outside of major metropolitan areas, and most of the closures between 1900 and 1920 were of schools in large cities (Mayers and Harrison 1924, 143). In our four-state sample, the fraction of new physicians who attended rural medical schools increases slightly between 1909 and 1921.

  18. The results are robust when using a linear probability model, Appendix Table 10, and using a 2500 population cutoff as a definition of rural, rather than 10,000 in population, Appendix Table 11.

  19. The IPUMS defines a metropolitan area to be a county or group of contiguous counties that contained at least one city of 50,000 or more residents. For a county to be included in a metropolitan area, it had either to contain the central city or be metropolitan in character, meaning that it contained a large non-agricultural workforce. This classification scheme is similar to that used in modern studies of the geographic distribution of medical professionals.

  20. In 1909, the estimated odds ratio on the interaction between the number of doctors and medical school requirements is not statistically significant in CA, likely because very few schools had established those requirements in CA by 1909. A handful of students who graduated from Johns Hopkins and chose to settle in California dominate the 1909 results—as witnessed by the implausibly large estimated odds ratio on the interaction between ln(miles to medical school) and pre-medical college requirement in CA in 1909.

  21. We include the linear Oaxaca–Blinder decomposition results in Appendix Table 13; the results show that all the explanatory power is loaded onto the pre-med requirements, which explains 100% of the change over time in the likelihood of rural practice location.

  22. Pilot searches of physicians practicing in Los Angeles had very low match rates.

  23. For 27 physicians, we were unable to find any records in For 29, we were able to find the draft card but the information on town of birth was missing or illegible.

  24. To be consistent with the earlier models, we also estimate the North Carolina model using a definition of rural as less than 10,000 inhabitants. As shown in Appendix Table 16, this only strengthens our findings. Additionally, the results are quantitatively and qualitatively similar when using a linear probability model instead of a probit. Results are reported in Appendix Table 15.


  • American Academy of Family Physicians (2013) Graduate medical education for rural practice

  • American Medical Association (1906) Standards of medical education adopted by the American Medical Association, July 1905. J Am Med Assoc 47(8):627

    Google Scholar 

  • American Medical Association Council on Medical Education (1905–1910) Medical Schools of the United States. J Am Med Assoc

  • American Medical Association Council on Medical Education (1911–1914) Medical Colleges of the United States. J Am Med Assoc

  • American Medical Association Council on Medical Education (1915–1920) Medical education in the United States. J Am Med Assoc

  • American Medical Association Council on Medical Education (1919) Medical education in the United States. J Am Med Assoc 73(7):507–512 (Table 4 Existing and Extinct Medical Colleges)

    Google Scholar 

  • American Medical Association Council on Medical Education (1923) Medical education in the United States. J Am Med Assoc 81(7):571–576 (Existing and Extinct Medical Colleges (un-numbered table))

    Google Scholar 

  • (2005) U.S., World War I Draft Registration Cards, 1917–1918. Database on-line

  • Andersen R, Anderson O (1967) A decade of health services. University of Chicago Press, Chicago

    Google Scholar 

  • Arrow K (1963) Uncertainty and welfare economics of medical care. Am Econ Rev dec:941–973

  • Baker SL (1984) Physician licensure laws in the United States, 1865–1915. J Hist Med Allied Sci 39:173–197

    Article  Google Scholar 

  • Bonner TN (1963) American doctors and German universities. University of Nebraska Press, Lincoln

    Google Scholar 

  • Burrows E, Suh R, Hamann D (2012) Health care workforce distribution and shortage issues in rural America. National Rural Health Association Policy Brief

  • Chen C, Petterson S, Phillips RL, Mullan F, Bazemore A, O’Donnell SD (2013) Toward graduate medical education (GME) accountability: measuring the outcomes of GME institutions. Acad Med 88(9):1267–1280

    Article  Google Scholar 

  • Commission on Medical Education (1932) Final report of the commission on medical education. Office of the Director of the Study, New York

  • Commission on medical education (1932) Final report of the commission on medical education. Office of the Director of Study, New York

  • Dhalla IA, Kwong JC, streiner DL, Baddour RE, Waddell AE, Johnson IL (2002) Characteristics of first year students in Canadian medical schools. Can Med Assoc J 166:1029–35

    Google Scholar 

  • Fairlie RW (2005) An extension of the Blinder–Oaxaca decomposition technique to logit and probit models. J Econ Soc Meas 30(4):305–316

    Article  Google Scholar 

  • Fairlie RW (2017, April) Addressing path dependence and incorporating sample weights in the nonlinear Blinder–Oaxaca decomposition technique for logit, probit, and other nonlinear models. Working paper 17-013. Stanford Institute for Economic Policy Research

  • Field J (1970) Medical education in the United States: late nineteenth and twentieth centuries. In: O’Malley CD (ed) The history of medical education. University of California Press, Berkeley

    Google Scholar 

  • Flexner A (1910) Medical education in the United States and Canada. Carnegie Foundation for the Advancement of Teaching, Princeton

    Google Scholar 

  • Gaynor M (1994) Issues in the industrial organization of the market for physician services. J Econ Manag Strategy 3(1):211–255

    Article  Google Scholar 

  • Gottlieb DS, Joseph G (2006) College-to-work migration of technology graduates and holders of doctorates within the United States. J Reg Sci 46(4):627–59

    Article  Google Scholar 

  • Haines MR (2001) The urban mortality transition in the United States, 1800–1940. Annales de demographie historique 101:33–64

    Article  Google Scholar 

  • Kessel R (1958) Price discrimination in medicine. J Law Econ 1:20–53

    Article  Google Scholar 

  • Kwong JC, Dhalla IA, Streiner DL, Baddour RE, Waddell AE, Johnson IL (2002) Effects of rising tuition fees on medical school class composition and financial outlook. Can Med Assoc J 166:1023–8

    Google Scholar 

  • Kwong JC, Dhalla IA, Streiner DL, Baddour RE, Waddell AE, Johnson IL (2005) A comparison of Canadian medical students from rural and non rural backgrounds. Can J Rural Med 10:36–42

    Google Scholar 

  • Law MT, Kim S (2005) Specialization and regulation: the rise of professionals and the emergence of occupational licensing regulation. J Econ Hist 65(3):723–756

    Article  Google Scholar 

  • Lichtenberg F (2011) The quality of medical care, behavioral risk factors, and longevity growth. Int J Health Econ Manag 11(1):1–34

    Google Scholar 

  • Ludmerer KM (1985) Learning to heal: the development of American Medical Education. Basic Books Inc, New York

    Google Scholar 

  • Mayers L, Harrison LV (1924) The distribution of physicians in the United States. General Education Board

  • McCarthy TR (1985) The competitive nature of the primary-care physician services market. J Health Econ 4(2):93–117

    Article  Google Scholar 

  • McFadden D (1974) Analysis of qualitative choice behavior. In: Zarembka P (ed) Frontiers in econometrics. Academic Press, New York

    Google Scholar 

  • Moretti E (2011) Local labor markets. Handbook of labor economics, vol 4. Elsevier, Amsterdam

    Google Scholar 

  • Newhouse JP (1970) A model of physician pricing. South Econ J 37(2):174–183

    Article  Google Scholar 

  • Page D, Baranchuk A (2010) The flexner report: 100 years later. Int J Med Educ 1:74–75

    Article  Google Scholar 

  • Polsky D, Kletke PR, Wozniak GD, Escarce JJ (2002) Initial practice locations of international medical graduates. Health Serv Res 37(4):907–928

    Article  Google Scholar 

  • Roback J (1982) Wages, rents, and the quality of life. J Political Econ 90(6):1257–1278

    Article  Google Scholar 

  • Rosen S (1979) Current issues in urban economics. Johns Hopkins University Press, Baltimore

    Google Scholar 

  • Rosenblatt RA (2010) Commentary: Do medical schools have a responsibility to train physicians to meet the needs of the public? The case of persistent rural physician shortages. Acad Med 85(4):572–74

    Article  Google Scholar 

  • Rosenblatt RA, Hart LG (2000) Physicians and rural America. West J Med 173(5):348–351

    Article  Google Scholar 

  • Rothstein WG (1972) American physicians in the nineteenth century. From sects to science. Johns Hopkins University Press, Baltimore

    Google Scholar 

  • Ruggles S, Genadek K, Goeken R, Grover J, Sobek M (2015) Integrated public use microdata series. Version 6.0 [Machine-readable database]

  • Sloan FA, Feldman R (1978) Competition in the health care sector: past, present, and future. Federal Trade Commission, Washington

    Google Scholar 

  • Sutch R (2006) Gross domestic product: 1790–2002 [continuous annual series]: Table Ca9-19. In: Historical statistics of the United States, earliest times to the present: millennial edition. Cambridge University Press, New York, NY

  • United States Department of the Interior, Bureau of Education (1921) Biennial survey of education, 1916–1918, Volume 1. Government Printing Office

  • U.S. Bureau of the Census. (1913) Thirteenth census of the United States taken in the year 1910, Volume III: Population 1910, Reports by states, with statistics for counties, cities, and other civil divisions, Nebraska-Wyoming, Alaska, Hawaii and Porto Rico. Government Printing Office

  • van Bibber AF (1929) The swan song of the country doctor. The North American Review

  • Wozniak A (2010) Are college graduates more responsive to distant labor market opportunities. J Hum Resour 45(4):944–70

    Google Scholar 

Download references


Thomasson acknowledges support from the Julian Lange Professorship. We thank seminar participants at Queens University, Oberlin College, and the University of Wisconsin-LaCrosse for helpful comments and suggestions.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Melissa A. Thomasson.

Ethics declarations

Conflict of Interest

The authors declare that they have no conflict of interest.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.



See Fig. 4 and Tables 10, 11, 12, 13, 14, 15, 16 and 17.

Fig. 4
figure 4

Example page from 1906 American Medical Directory

Table 10 Determinants of rural \((< 10,000)\) practice location choice: linear probability model.
Table 11 Determinants of rural \((< 2500)\) practice location choice: probit marginal effects.
Table 12 Rural practice location choice of new doctors: probit marginal effects (including AMA ratings as controls).
Table 13 Linear Oaxaca–Blinder decomposition of changes in probability of rural location of new doctors.
Table 14 Characteristics of recent medical school graduates by birthplace (rural defined as pop < 10,000), North Carolina, 1918.
Table 15 Results from linear probability model of practice in rural area, North Carolina 1918.
Table 16 Marginal effects from probit model of practice in rural area (defined as pop < 10,000), North Carolina 1918.
Table 17 Determinants of rural practice location choice for new doctors: probit marginal effects.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Moehling, C.M., Niemesh, G.T., Thomasson, M.A. et al. Medical education reforms and the origins of the rural physician shortage. Cliometrica 14, 181–225 (2020).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:


JEL Classification