The financial costs of medical education in the United States are high.1 From 1990 to 2003, medical school tuition and fees grew 167% for residents at public schools to US $15,865, and 83% for nonresidents at private schools to US $34,968.2,3 Due to the high and rising fees, 81% of graduating medical students in 2003 had outstanding loans for their medical school education, with average medical school indebtedness exceeding US $100,000. The average medical school debt for all students in 2003, which included students without debt, was US $83,361, an increase of 154% since 1990.4
While tuition and fees have grown between 83% and 167% since 1990, the mean stipend for first-year housestaff has increased only 54%,5 and residency and fellowship duration in many medical disciplines has increased.6,7 By comparison, median net income for physicians from 1990 to 2000 increased only 35%,8 or close to the rate of inflation.9 Figure 1 shows the recent changes in tuition and fees, medical school debt, housestaff stipends, physician income, and inflation.
The high and growing financial costs of medical education and the increasing duration of residency and fellowship training significantly affect medicine. Although the number of applicants to medical school increased from 2002–03 to 2003–04, overall interest in medicine, as measured by the number of applicants to U.S. medical schools, dropped 26% from its peak in 1996–97 to 2003–04.10,11 Economic factors, such as the expectation of incurring a large debt obligation and prolonged deferral of full entrance into the profession, deter some individuals from pursuing a particular career.12,13 A recent survey conducted for the Association of American Medical Colleges (AAMC) of college students qualified for medical school found that the cost of attending medical school and the time it takes to become a doctor were the top two reasons for not applying to medical school.14
The large upfront costs of medical education15 are especially challenging to applicants of limited economic means.14,16,17 In the United States., the number of Black, Mexican American, Native American, and mainland Puerto Rican applicants declined by 17% from 1996 to 2002,18,19 despite efforts by the AAMC to increase the number of students from groups currently underrepresented in the medical profession.20 Rising tuition prices and student debt levels have provoked concern about medical school class composition in Canada21 and cries that “only the rich can become doctors” in the United Kingdom,22 two countries that have significantly lower average medical student debt levels than the United States.
Medical school debt burden also influences career decisions in medicine. While the literature on the effect of debt burden on specialty choice is mixed,23 some economic research does indicate that medical school indebtedness plays an important role in specialty choice,24 and debt may negatively affect primary care specialties, such as family practice.25 In addition, debt may also have an important impact on long-term career plans, such as the likelihood of physicians entering academic medicine and pursuing research.26
Finally, the large debt burden also weighs on young physicians psychologically. A recent study found that increasing educational debt was associated with depressive symptoms, increased cynicism, and decreased humanism.27
Recent proposals to decrease the financial burden of medical education and training have included reducing the cost of medical school,28,29 decreasing the duration of medical education,6,30,31 and increasing residents’ compensation,29 but to date, a systematic financial comparison of these alternatives has not been performed. In this study, we evaluated the relative financial impact of these potential solutions to decreasing the financial burden of medical education and training.
To determine the financial impact of changes in medical education, we constructed a financial model for calculating the net present value of the following four proposed solutions:
* ▪decrease medical school tuition and fees by 25%;
* ▪decrease the duration of medical school from four years to three years;
* ▪increase housestaff stipends to match those of first-year physician assistants; and
* ▪decrease the duration of graduate medical education for medicine subspecialists by one year.
Table 1 summarizes the data inputs and assumptions used in the model.
Net present value is a financial measure that calculates the current value of money that is either received (e.g., physician income) or paid out (e.g., medical school tuition) in the future. Because income or payments that occur at different points in the future are not directly comparable, a present value for future income or payment must be calculated.32
Present values are calculated by adjusting downward the value of future earnings or payments by a “discount rate,” which reflects the fact that a future dollar is worth less than a dollar today. A dollar in the future is worth less primarily because inflation erodes the buying power of a dollar over time. By using net present value calculations, all future payments or earnings are converted to values that are comparable in today’s dollars.
Net present value has everyday applications. For example, a physician practice may want to purchase an ultrasound device. The practice will want to know if the future income generated by the ultrasound will offset the initial cost of the device. Net present value calculations are helpful to make such financial decisions.
For this study, we calculated net present values from the standpoint of a student beginning medical school in 2004–05. In the base case, the net present value of a career in medicine was calculated under the current medical education system with cash flow projections carried 40 years into the future. We then evaluated the four proposed solutions and compared them with the base case and with each other. We determined the financial impact of each of the proposed solutions by calculating the difference in net present value between the current system and the proposed one.
We modeled several factors including undergraduate medical education costs, housestaff compensation, the duration of graduate medical education, and physician compensation. The weighted average tuition and fees for matriculating medical students in 2003–04 was calculated from the actual mix of medical students in public and private schools paying resident and nonresident tuition and fees.2 Housestaff compensation was determined from the AAMC Survey of Housestaff Stipends, Benefits & Funding.5 The duration of residency for general internal medicine and the internal medicine subspecialties was modeled as three and six years, respectively.33 We chose the latter because six years is the minimum accredited length for the four largest medicine subspecialties (as measured by number of residents in 2002): cardiovascular disease, gastroenterology, pulmonary disease and critical care medicine, and hematology and oncology.34 The median net income (including deferred compensation) after expenses but before taxes for general internists and internal medicine subspecialists was determined from the 2003 edition of Physician Socioeconomic Statistics published by the American Medical Association (AMA).8 The mean annual income for full-time physician assistants for 2003, who graduated in 2002, was obtained from the 2003 American Academy of Physician Assistant Census Report.35
To project values into the future, we had to make a few assumptions. The future annual growth rate in tuition and fees for medical schools was estimated to be 6.3%, which is the calculated average compound annual growth rate for tuition and fees for nonresidents at private schools and residents at public schools from 1990 to 2003.3 The growth rate for stipends for housestaff and income for physicians and physician assistants was estimated to be 3% per year, which closely approximates inflation as measured by the Consumer Price Index from 1913 to 2003.9 The incremental growth in housestaff stipend reflected both inflation (3%) and a step increase in payment (4%) for each successive year of training, which is approximately the incremental difference between each postgraduate year of training in 2003–04.5 To reflect the lower net income earned by younger physicians, physicians were assumed to earn 80% of the median compensation for their selected specialty immediately upon completion of their residency.8 This proportion increased to 100% of the median over five years.
In the base model, cash flows (payments and earnings) were discounted at 6% per year and were assumed to occur at the end of the year. Other researchers examining the educational costs and incomes of physicians and other professionals have used discount rates ranging from 5% to 10%.15 Economists suggest a real (excluding inflation) interest rate of around 3%.32,36 Including the effects of inflation leads to a (nominal) discount rate of approximately 6%. Higher discount rates (e.g., 10%) give less weight to cash flows in later years while lower discount rates (e.g., 5%) give relatively more weight to cash flows in later years. We rounded off results to the nearest US $10,000.
To examine the effect of using a range of reasonable values for model inputs and to perform additional analyses, we adjusted assumptions and variables in the model. The discount rate was varied from 6% to 10%, and the reduction in the price of medical tuition was increased from 25% to 100%. The increase in residents’ compensation was raised to the mean income of all physician assistants in full-time clinical practice ($76,039).35
Reduce medical school tuition
Reducing average medical school tuition and fees by 25% resulted in a savings of US $30,000 in today’s dollars to future physicians.
Decrease medical school duration
Reducing the duration of medical school from four years to three years resulted in a net present value of US $160,000 for general internists and of US $230,000 for internal medicine subspecialists. The one-year saving in tuition and fees contributed US $30,000 in today’s dollars to the total while the majority of the benefit was due to the earlier realization of income as a practicing physician.
Increase residency compensation
Matching housestaff stipends with the average salary of a first-year physician assistant increased the net present value of future earnings by US $60,000 for general internists and US $100,000 for internal medicine subspecialists.
Decrease residency duration
Reducing the duration of graduate medical education by one year for internal medicine subspecialties returned a net present value of US $170,000. Figure 2 summarizes the financial impact of the four potential remedies evaluated.
A higher discount rate, which decreases the weight given to income earned in later years, did not alter the relative financial attractiveness of the four proposed solutions but did decrease the magnitude of the net present values of the proposed remedies, especially for proposals that reduce the duration of medical education. For example, using a 10% rather than 6% discount rate decreased the net present value of reducing medical school by one year for subspecialists from US $230,000 to US $170,000.
Increasing the reduction in medical school tuition and fees to 50%, 75%, and 100% (full subsidy) resulted in a net present value of US $50,000, US $80,000, and US $100,000, respectively. Increasing residency compensation to the mean income for all physician assistants increased the net present value to US $90,000 for general internists and US $160,000 for medicine subspecialists. Table 2 summarizes the results of the sensitivity analyses.
Recent efforts by the AMA and the AAMC to reduce medical student debt have focused on adjusting student loans, capping medical school tuition, and expanding loan repayment programs.14,29 By expanding the analysis beyond medical school tuition, we have demonstrated that the most effective financial means for decreasing the large financial burden of medical education is not to reduce the price of medical education but rather to reduce the duration of training.
Reducing the duration of medical school by one year had the largest financial benefit, which ranged from US $160,000 to US $230,000 in today’s dollars, depending on the specialty chosen. Similar results would be expected for any one-year decrease in the current eight-year duration of premedical and undergraduate medical education.
The primary reason for the large benefit from eliminating a year of medical school is not the cash savings from tuition but rather the additional year of income gained as a physician. In economic terms, this benefit is due to the high opportunity cost each year of training holds. The opportunity cost is the highest value an alternative course of action would provide.37 In the case of medical training, the opportunity cost of each year of premedical and undergraduate medical training is the alternative of delaying by one year and foregoing one year of earning relatively high income as a practicing physician.
Shorter medical school duration has precedence in the United States. In the early 1970s, a three-year medical school curriculum flourished. While medical students and faculty gave mixed reviews,38,39 at least one study found no difference in objective measures of student performance between students from a three-year and four-year curriculum.38 In Western Europe, undergraduate medical education is often combined with the U.S. equivalent of college, and the result is a shorter duration of training. For example, in the United Kingdom, the equivalent of undergraduate medical education is completed six to seven years after high school.40 Few such combined programs are available in the United States. For 2004–05, only 18 of the more than 120 U.S. medical schools offered combined college and medical school programs that could be completed in fewer than eight years.41
More recently, the American Board of Internal Medicine and the American Board of Family Practice sponsored pilot programs with accelerated residencies that allowed select fourth-year medical students to incorporate the fourth year of medical school into an internal medicine or family practice residency.42 An evaluation of internal medicine trainees in the accelerated and traditional residency programs found “no statistically significant differences between the two groups in mean scores on the Internal Clinical Evaluation Exercise, standardized in-service exam, monthly attending evaluations, or in Board pass rates.”43, p. 361 An assessment of the University of Tennessee’s accelerated family residency program from 1992–2002 concluded, “Accelerated residents have performed as well as or better than nonaccelerated residents on standardized testing,”44, p. 178 and that the program has been an effective means for allowing medical students to complete their family medicine training in six (rather than seven) years.44 Other researchers have reached similar conclusions in evaluating other programs.45,46 While these pilot programs have demonstrated good results in select trainees, the pilots were discontinued in part due to Accreditation Council for Graduate Medical Education (ACGME) requirements that trainees receive their medical degree prior to beginning residency.42
Reducing the duration of graduate medical education for internal medicine subspecialties also resulted in a large financial impact, US $170,000 in today’s dollars. In addition to its financial attractiveness, others have called for its consideration on educational grounds in both medicine and surgery.6,7,30 In medicine, Dr. Lee Goldman, chair of the Department of Medicine at the University of California, San Francisco, and others47 have argued that “training simply takes too long” and have proposed shorter and more tailored medicine training programs that could “better prepare (trainees) for what they actually want to do.”7, p.134 Goldman has suggested “modernizing” internal medicine training by reducing the length of core internal medicine training to two years. After the “core,” medicine trainees could either pursue one year of primary care training, one year of inpatient training, or enter subspecialty training directly.7
While medicine considers such options, some surgical subspecialties have acted. In plastic surgery, where a combined general surgery and plastic surgery residency reduced total duration by two years, a study concluded that a combined program is “effective, appropriate, resource-efficient” and produced excellent outcomes in trainees.48 In October 2001, the American Board of Thoracic Surgery changed the training requirements for its certification that was in part motivated “by a total duration of medical school and of (graduate medical education) that obligates the thoracic surgeon entering practice to be older, as compared to peers in other professions, and to be constrained by greater indebtedness.”49, p.1433 Additional pathways to certification created by the Board include options that decrease the total duration of training by one or two years.50 Recently, similar changes have been called for in vascular surgery.51
Reducing the duration of graduate medical education in face of the recent ACGME duty-hour restrictions (to 80 hours per week)52 may be challenging; however, it, too, has precedence. In the United Kingdom, as a result implementing the Calman reforms, the duration of training for many specialties decreased.53,54 Shortening the duration of training also took place in an environment where the work week for trainees had been shortened with plans to shorten the actual hours of duty even further to 48 hours per week.54 While an evaluation of such changes is beyond the scope of this paper, these reforms at least attest to the feasibility of decreases in the duration of training in the setting of (modest) limits on the work week and provide excellent case studies for future investigation.
Increasing housestaff compensation levels to those of a physician assistant also generated a significant financial impact that was a function of residency duration and sensitive to the level of physician assistant income used. We chose physician assistants’ compensation as a benchmark because physician assistants (and nurse practitioners) increasingly are working with, and in place of, residents as a result of the ACGME duty-hour restrictions.52,55
Of the four proposed solutions evaluated from a financial perspective, reducing medical school tuition and fees had the smallest financial impact. Only a full subsidization of all medical school tuition and fees begins to approach the magnitude of reducing medical school or subspecialty training duration by one year.
The proposed solutions that we evaluated are not mutually exclusive and can be combined. For example, decreasing medical school and subspecialty training each by one year generates a financial benefit of US $400,000 for future internal medicine subspecialists.
While the focus of our study was on internal medicine and its subspecialties, the analysis and implications can be extended to other specialties. In general, the higher a year of foregone specialty income, the greater the financial impact of reducing pre-, undergraduate, or graduate medical education duration. For example, the impact of shortening education and training duration for general pediatrics will be less than for general internal medicine, to the degree that income for pediatricians is lower than that for general internists. The results for pediatric subspecialties will be similar to those for the medical subspecialties that are not procedure-oriented as income and training duration are fairly comparable. Relatively high incomes for surgical subspecialists, such as vascular surgeons, will lead to substantial financial benefits from a one-year reduction in training duration.
Our study had several limitations. We did not consider all variables possible for reducing the financial burden of medical education. Reducing the duration of premedical education was not evaluated but could be modeled in a similar manner and would have a comparable effect as reducing the duration of medical school. In addition, we did not examine changes to the complex structure of student loans; however, even a full (100%) subsidization of medical school tuition had a smaller financial impact than decreasing the duration of medical school by one year. We did not model scholarships, which often target individuals or groups with the greatest need; the cost of living; benefits; income taxes; the tax treatment of interest on debt; and loan forgiveness programs.
We also did not consider the cost of the proposed remedies. Medical school revenues would be decreased by reducing tuition or the duration of medical school (unless offset by other sources). However, even if total medical school tuition remained constant, a one-year reduction in the duration of medical school still yields a financial benefit of US $100,000 or more to future physicians. Large increases (more than US $20,000 per year as modeled in our study) in residency compensation for either the nearly 30,000 residents in internal medicine and its subspecialties or the approximately 100,000 U.S. residents34 will be expensive for teaching hospitals or funders of graduate medical education. Conversely, reductions in the duration of residency will reduce federal spending on graduate medical education but leave teaching hospitals with a smaller supply of relatively inexpensive labor. Additional financial analyses are needed to evaluate these costs and trade-offs in more detail.
While our focus was on the financial impact of the proposed remedies, educational considerations are paramount. However, as detailed here, a shorter duration of undergraduate medical education has precedence in the United States and abroad, and, increasingly, leaders in medicine and surgery are advocating for and implementing reductions in the duration of U.S. graduate medical education.
In this study, we demonstrate that the primary driver for the high financial burden of medical education is not large cash outlays for tuition or even low resident compensation, but rather the long duration of training, which carries a high opportunity cost. Consequently, efforts aimed at reducing the duration of medical education are the most likely to generate substantial financial benefit for future physicians. The educational impact and implementation of such remedies, which are within the control of medical leaders, warrant further investigation.
1Morrison G. Mortgaging our future: the cost of medical education. N Engl J Med. 2005;352:117–19.
3AAMC Data Book: Statistical Information Related to Medical Schools And Teaching Hospitals: January 2003. Washington, DC: Association of American Medical Colleges Division of Medical School Services and Studies, 2003.
5AAMC Survey of Housestaff Stipends, Benefits & Funding: November 2003 Report. Washington, DC: Association of American Medical Colleges Division of Health Care Affairs, 2003.
6Johns MM. The time has come to reform graduate medical education. JAMA. 2001;286:1075–76.
7Goldman L. Modernizing the paths to certification in internal medicine and its subspecialties. Am J Med. 2004;117:133–36.
8Wassenaar J, Thran S (eds). Physician Socioeconomic Statistics: 2003 Edition. Chicago: American Medical Association, 2003.
10Barzansky B, Etzel SI. Educational programs in US medical schools, 2002-2003. JAMA. 2003;290:1190–96.
12Freeman RB. The Market for College-Trained Manpower: A Study in the Economics of Career Choice. Cambridge, MA: Harvard University Press, 1971.
13Bok DC. The Cost of Talent: How Executives and Professionals Are Paid and How it Affects America. New York: Free Press, 1993.
15Weeks WB, Wallace AE, Wallace MM, Welch HG. A comparison of the educational costs and incomes of physicians and other professionals. N Engl J Med. 1994;330:1280–86.
16Hughes RG, Barker DC, Reynolds RC. Are we mortgaging the medical profession? N Engl J Med. 1991;325:404–7.
17Jolly P. Medical school tuition and young physicians’ indebtedness. Health Aff (Millwood). 2005;24:527–35.
20Medical School Admission Requirements: United States and Canada, 2004–2005. Washington, DC: Association of American Medical Colleges, 2003.
21Kwong JC, Dhalla IA, Streiner DL, Baddour RE, Waddell AE, Johnson IL. Effects of rising tuition fees on medical school class composition and financial outlook. CMAJ. 2002;166:1023–28.
23Spar IL, Pryor KC, Simon W. Effect of debt level on the residency preferences of graduating medical students. Acad Med. 1993;68:570–72.
24Thorton J. Physician choice of medical specialty: do economic incentives matter?. Appl Econ. 2000;32:1419–28.
25Rosenthal MP, Marquette PA, Diamond JJ. Trends along the debt-income axis: implications for medical students’ selections of family practice careers. Acad Med. 1996;71:675–77.
26Fox M. Medical student indebtedness and the propensity to enter academic medicine. Health Econ. 2003;12:101–12.
27Collier VU, McCue JD, Markus A, Smith L. Stress in medical residency: status quo after a decade of reform?. Ann Intern Med. 2002;136:384–90.
28Johns MM. Mandatory national health service: an idea whose time has come. JAMA. 1993;269:3156–57.
30Friedman SG. Anyone in the O.R.? New York Times. 2003 June 10: A29.
31Zweifler J. Why we should reduce family practice training to two years. Acad Med. 2003;78:885–87.
32Gramlich EM. A Guide to Benefit-Cost Analysis 2nd ed. Englewood Cliffs, NJ: Prentice Hall, 1990.
33Graduate Medical Education Directory: 2003–2004. Chicago: American Medical Association, 2003.
34Graduate medical education. JAMA. 2003;290:1234–38.
36Cutler DM, Richardson E. The value of health: 1970-1990. AEA Papers Proc. 1998;88:97–100.
37PH W, Freeman RE (eds). The Blackwell Encylcopedic Dictionary of Business Ethics. Cambridge, MA: Blackwell Publishers, 1997.
38Kettel LJ, Dinham SM, Drach GW, Barbee RA. Arizona’s three-year medical curriculum: a postmortem. J Med Educ. 1979;54:210–16.
39Woodward CA, Ferrier BM. Perspectives of graduates two or five years after graduation from a three-year medical school. J Med Educ. 1982;57:294–302.
40Dangerfield P. Medical student debt in the United Kingdom. Med Educ. 2001;35:619–21.
41Moller M (ed). Medical school admission requirements, 2004–2005, United States and Canada. Washington, DC: Association of American Medical Colleges, 2003.
42Puffer J, president and chief executive officer, American Board of Family Medicine, Lexington, Kentucky. Personal communication, July 22, 2005.
43Chang LL, Grayson MS, Patrick PA, Sivak SL. Incorporating the fourth year of medical school into an internal medicine residency: effect of an accelerated program on performance outcomes and career choice. Teach Learn Med. 2004;16:361–64.
44Delzell JE Jr, McCall J, Midtling JE, Rodney WM. The University of Tennessee’s accelerated family medicine residency program 1992–2002: an 11-year report. Fam Med. 2005;37:178–83.
45Petrany SM, Crespo R. The accelerated residency program: the Marshall University family practice 9-year experience. Fam Med. 2002;34:669–72.
46Galazka SS, Zweig S, Young P. A progress report on accelerated residency programs in family practice. Acad Med. 1996;71:1253–55.
47Blackwell TA, Powell DW. Internal medicine reformation. Am J Med. 2004;117:107–8.
48Scholten DJ, Anderson KD, Beernink J, Passinault WD. Experience with surgical resident training in a combined program in plastic surgery. Am Surg. 1995;61:594–601; discussion 601–2.
49Olinger GN. “Change in the wind”: report from the 2000 Thoracic Surgery Directors Association retreat on thoracic surgery graduate medical education. Ann Thorac Surg. 2001;72:1433–7.
50The American Board of Thoracic Surgery Certification 〈http://www.abts.org/doc/6678
〉. Accessed 28 November 2005. Evanston, IL, American Board of Thoracic Surgery.
51Berguer R. Problems facing vascular surgery in 2004. Vascular. 2004;12:39–41.
52Steinbrook R. The debate over residents’ work hours. N Engl J Med. 2002;347:1296–302.
53Wills AJ. Assessment of professional development of neurology trainees: a UK perspective. Lancet Neurol. 2005;4:254–57.
54Cassar K. Development of an instrument to measure the surgical operating theatre learning environment as perceived by basic surgical trainees. Med Teach. 2004;26:260–64.
55Kuttler H. At Yale, new rules create jobs for PAs. AAPA News. 2002 July 30;1:6.