Youngclaus, James A. MS; Koehler, Paul A. PhD; Kotlikoff, Laurence J. PhD; Wiecha, John M. MD, MPH
Most medical school graduates have education debt, or debt incurred both during medical school as well as any premedical/undergraduate debt. In addition, the average amount of education debt per medical school graduate is increasing. Among medical school graduates in 2011, 86% had education debt at graduation, averaging $161,290—the highest total to date. Among these indebted graduates, 23% of those at private medical schools graduated with loans of $250,000 or more.1 Controlling for inflation, the average education debt for an indebted medical school graduate in 2011 was nearly 3.5 times greater than it was in 1978. In the mid-1990s, very few indebted medical school graduates had education debt even approaching $162,000; for the class of 2011, this amount was the median.2
A common assumption about specialty choice is that education debt plays a key role—specifically, pulling physicians away from the relatively lower-paying primary care fields and toward the more lucrative specialties. However, the scientific literature includes no definitive proof that education debt significantly influences the career choices of most medical students and residents. Some research suggests that debt may have a marginal influence but that it is typically overwhelmed by other factors, including demographics, personal interest in a specialty’s content and/or level of patient care, desire for the “controllable lifestyle” offered by some specialties, and exposure to a strong role model in a specialty.3–12 Other research reports that education debt has no discernible influence on specialty choice.13–16 The evidence is stronger that the income potential of various specialties influences specialty choice, but, again, other factors carry more weight.17–22 The few articles that consider the actual economics of education debt repayment for a physician are either outdated23,24 or based on a household with a single wage earner following only a standard 10-year repayment plan.25
In this study, we examined the actual economics of loan repayment in the context of a physician’s household income and expenses to determine whether education debt level should affect specialty choice. We hope that our study will offer new, and perhaps surprising, insights into the economic viability of a primary care career for a physician with today’s standard level of education debt. Our analysis should inform policy makers and medical students alike, as misperceptions about the feasibility of education debt repayment can lead to faulty assumptions about specialty choice.
In 2010–2011, to gauge the ability of recent medical school graduates to repay their current levels of education debt without incurring additional debt, we analyzed the household finances of a fictional couple over their lifetime, using sophisticated financial planning software.
For our fictional indebted physician, “Dr. Median,” we created three career tracks: one as a primary care physician and two as higher-paying specialists (comparable to physicians in psychiatry or obstetrics–gynecology and in general surgery), each with a comparable starting salary (see Appendix 1). Dr. Median was married to an employed, college-graduate spouse of the same age. They had two children and remained married until their deaths at 85 years old. We compared scenarios with various combinations of education debt and participation in federal loan forgiveness/repayment programs, such as the National Health Service Corps (NHSC), while accounting for the impact of a high-cost residential area such as Boston, Massachusetts, versus a more moderate-cost area such as Denver, Colorado.
To clearly demonstrate the impact of medical school debt only, Dr. Median did not have undergraduate/premedical debt or consumer debt, such as credit card debt. National data support this scenario: Only 35% of medical school graduates in the class of 2011 reported premedical debt, and those who did reported a median amount of $18,000.1 Additionally, only 26% of medical school graduates reported having noneducation debt (including car loans, residency search loans, or other consumer debt that was not a home mortgage), reporting a median amount of $9,150.1
Dr. Median entered medical school in the fall of 2006 and graduated in the class of 2010 at age 27. In the primary care career track, she then entered a three-year residency, earning the median resident stipend (see Appendix 1). In the specialist career tracks, Dr. Median spent the appropriate length of time in residency then earned the median starting salary for that specialty. Her spouse graduated from a four-year college in 2006 with education debt and a starting salary typical for a college graduate in that year. Their first child was born during Dr. Median’s residency when she was 28 years old, and their second was born three years later after she entered private practice. Both children entered four-year colleges at 18 years old.
We analyzed four education debt levels: $150,000, $200,000, $250,000, and $300,000. Among indebted medical school graduates in the class of 2011, in overlapping categories, 59% had education debt of $150,000 or more, 33% had $200,000 or more, 15% had $250,000 or more, and 5% had $300,000 or more.1,2
In addition, because of capitalizing interest, the loan repayment balance would be much greater than the original amount borrowed. For example, with forbearance (i.e., no payment, but interest accrues) during a three-year residency, $150,000 borrowed increases to over $200,000 to repay after residency, whereas $300,000 becomes nearly $425,000. With a longer residency, these repayment totals would be even larger.
The loan repayment scenarios that we modeled included the standard 10-year plan and the extended 25-year plan. An extended plan would result in a much lower monthly loan payment than a standard plan but much more total interest paid over the longer repayment period. We assumed forbearance during residency with these plans because, currently, indebted medical residents have two options for addressing their education debt during residency: forbearance or payment, most likely via income-based repayment (IBR). If an indebted medical school graduate selects forbearance, then a standard or extended repayment plan would be most likely after residency. If he or she selects IBR, then continuing in that program would be most likely after residency. As a side note, selecting forbearance or an extended 25-year repayment plan would have no major impact on the borrower’s credit score.
We also modeled a number of scenarios in which Dr. Median participated in the NHSC loan repayment program, a federal program in which primary care and other health providers receive funds to repay their education loans in exchange for service of at least two years full-time (see Table 1). For both the standard and extended repayment plans, we modeled two to seven years’ participation in the program (a dozen separate scenarios).
The final repayment scenarios that we modeled included two new programs: IBR and public service loan forgiveness (PSLF). Introduced in 2009, these federal programs are available to all borrowers in the Direct Loan program and are particularly appealing for physicians with high education debt levels. IBR is a repayment plan that links a borrower’s repayment to his or her income, meaning that a physician could begin repayment during residency with a feasible monthly payment. PSLF cancels a borrower’s remaining loan balance after he or she has made 10 years’ worth of monthly payments while employed at a nonprofit. A physician could start in this program during residency and continue afterward by working at a nonprofit. While participating in PSLF, a borrower’s monthly payment can follow IBR guidelines, ensuring a manageable monthly payment regardless of debt or income levels.
In total, we analyzed 384 different scenarios: 16 repayment plans at 4 different debt levels for 3 different career tracks in 2 different locations (Boston, Massachusetts; and Denver, Colorado).
We modeled all our scenarios using ESPlanner (Economic Security Planner Inc., Boston, Massachusetts), an economics-based personal financial planning software program developed by one of us (L.J.K.) and other economists.26 The program has over 10,000 household users, is used by dozens of financial planners nationwide, and has been highly rated in numerous articles over the last 13 years, including by Money Magazine and the Washington Post.27,28 ESPlanner calculates a household’s sustainable living standard by considering how much the household will earn over its lifetime and appropriating spending and saving accordingly. For example, the software recommends saving when the household has relatively more disposable income, so it does not suffer a living standard decline even with the onset of increased financial obligations.
Household spending model
We began our analyses with the fixed annual, after-tax sums that, based on ESPlanner’s rigorous economic calculations, would be available to Dr. Median’s household after accounting for three major expenditures: education debt repayment, financing a house, and saving for college for their children. Next, we constructed a realistic, comprehensive household spending model based on median data in federal statistics for households at this income level and encompassing dozens of typical expenditures (see Appendix 1). We applied this household spending model to the fixed annual sums available to determine how much discretionary income would remain for Dr. Median’s household after accounting for the cost of food, clothing, health care, entertainment, etc.
For example, in Scenario 1 ($150,000 borrowed, forbearance during residency, standard repayment afterward) for the primary care career track, when the couple are 32 years old, their gross income is roughly $194,000 ($150,000 for Dr. Median, $44,000 for her spouse, in 2010 dollars), with a net of $81,500 after taxes and the three major expenses (education debt repayment, financing a house, and saving for college for the children). From that total we subtracted the household spending model sum of $68,100 for that year ($5,675 monthly) and were left with $13,400 ($1,100 monthly). We defined discretionary income as the amount that the household had remaining after accounting for the cost of both the three major expenses and our household spending model. In the above-mentioned scenario, at age 32, Dr. Median’s discretionary income would be $13,400 ($1,100 monthly) and, at age 40, $42,000 ($3,500 monthly).
The specific amounts for various categories in the household spending model were less important than having a framework to evaluate Dr. Median’s ability to repay her education debt while meeting normal expenses and saving for major future expenses. We kept the household spending model constant across the three career tracks, despite the income differences, to be able to accurately compare the impact of each track.
Because ESPlanner specifies a fixed annual amount available to the household, its output can be compared with the results from a variety of other household spending models. Indeed, we conducted several sensitivity analyses adjusting spending in various categories, and the results validated the robustness of our household spending model. We also constructed a separate spreadsheet with independent calculations for taxes and the three major expenses, and the resulting yearly household net income was in alignment with the ESPlanner output.
Tables 2 and 3 summarize the repayment scenarios across the three career tracks that we modeled for Dr. Median’s household.
In Scenario 1 ($150,000 borrowed), Dr. Median chose forbearance during residency, followed by a standard 10-year repayment plan, requiring her household to make the maximum monthly payment after she completed her residency. In this scenario, while living in Boston with a $145,000 primary care starting salary, her household would have $1,100 per month in discretionary income two years into practice at age 32. In Denver, her household would have $1,500 per month (see Table 2, Scenario 1). Because this scenario requires the largest monthly loan repayment and our model suggests that her household could manage it, we did not include the results of the other repayment plans at this borrowing level in this report as these would allow even more monthly discretionary income.
Scenario 1 would result in an even higher level of monthly discretionary income if Dr. Median pursued a non-primary-care specialty. For example, psychiatry or obstetrics–gynecology would afford her $3,200 per month with a $195,000 starting salary, and general surgery would afford her $5,100 per month with a $245,000 starting salary (see Tables 3 and 4, Scenario 1).
In Scenario 2 ($200,000 borrowed, forbearance during residency, standard 10-year repayment plan), her household would have much less discretionary income if Dr. Median chose primary care: just $200 per month at age 32 living in Boston. To repay this level of borrowing without falling into debt or reducing spending in other areas, Dr. Median’s household might need to consider other repayment options, such as living in a more moderate-cost area like Denver, where the same scenario would result in $600 per month in household discretionary income (see Table 2, Scenario 2). Alternatively, if Dr. Median were to make a two-year commitment to the NHSC loan repayment program while living in Boston, her household would also have $600 per month (see Table 2, Scenario 3). If she chose an extended 25-year repayment plan living in Boston, her household would have $1,500 per month (Table 2, Scenario 5). They would have roughly $1,000 per month if she chose IBR in that scenario (Table 2, Scenario 8), although IBR would require them to make an $800 monthly loan repayment during her residency. At this level of borrowing, even if Dr. Median pursued primary care while living in Boston—the specialty choice and living combination that generates the most economic challenges among all the models—many repayment plans would be economically feasible. However, extended repayment would result in a hefty interest payment over its 25 years.
Examining the scenarios at the level of $250,000 borrowed showed a different story. The only scenarios where a standard 10-year repayment plan seems economically realistic involve at least a four-year commitment to the NHSC loan repayment program, whether living in Boston or Denver (Table 2, Scenario 12). Using the extended 25-year repayment plan living in Boston at this level of borrowing does result in $900 in monthly discretionary income at age 32, although full repayment for this scenario totals over three-quarters of a million dollars (Table 2, Scenario 13). If Dr. Median chose IBR at this borrowing level, which requires repayment during residency, her household would also have $900 in monthly discretionary income at age 32 (Table 2, Scenario 16).
Fewer than 1 in 20 indebted medical school graduates face the final borrowing level that we modeled—$300,000. This borrowing level would be difficult for Dr. Median to comfortably repay if she were a Boston-area primary care physician. The few repayment scenarios in which her household discretionary income would be above $300 a month involved either IBR (Table 2, Scenario 24), PSLF (Table 2, Scenario 25), or a four-year NHSC commitment while on an extended 25-year repayment plan (Table 2, Scenario 23). If Dr. Median pursued primary care in Denver at this borrowing level with an extended 25-year repayment plan, her household would have $700 in monthly discretionary income and over $920,000 in total repayment (Table 2, Scenario 21).
The considerable economic benefits to pursuing the higher-paying specialties are apparent when comparing scenarios across Tables 2, 3, and 4. For example, Scenario 18 ($300,000 borrowed) seems economically infeasible if Dr. Median were a primary care physician living in Boston, as it results in −$1,300 per month in discretionary income (Table 2, Scenario 18). However, if Dr. Median had a starting salary of $195,000, this same scenario would result in $400 per month in discretionary income, and, with a starting salary of $245,000, her household would have $2,200 per month(Tables 3 and 4, Scenario 18). In Tables 3 and 4, we show only selected Boston scenarios for comparison purposes because these higher-income households seem economically capable of living in an expensive city and managing the most financially challenging repayment option that we modeled—$300,000 borrowed and repaid via a standard 10-year repayment plan.
The scenarios that we modeled indicate that recent medical school graduates with the median amount of education debt ($160,000) can enter primary care, raise a family, live in an expensive urban area, and repay their debt in 10 years without incurring additional debt, provided their household income is comparable to the spending model that we used, which was based on median statistics. To achieve all of these goals at the $200,000 debt level, however, a primary care physician would need to consider an extended repayment plan and/or a federal loan repayment program, such as IBR or the NHSC. To achieve these goals at debt levels of $250,000 or more, a primary care physician would need to consider living in a more moderately priced area and/or choosing an extended 25-year repayment plan or a multiyear commitment to a federal loan forgiveness/repayment program.
In general, the more a medical student borrows, the more likely it is that he or she will need to consider repayment options other than a standard 10-year repayment plan if he or she lives in an expensive urban area. Not surprisingly, each such choice involves trade-offs. For example, an extended repayment plan results in a lower monthly payment but much more interest paid over the repayment period. IBR bases the monthly payment on the household’s income but, for maximum benefit, requires payment during residency when the household’s income is at its lowest. PSLF offers the possibility of loan forgiveness but only after 10 years of working in a government or nonprofit position. The NHSC offers generous loan repayment sums but requires at least two years of service in an NHSC-approved position in a health professional shortage area, which can be urban or rural.
If Dr. Median pursued a higher-paying specialty, we found that her household could manage her education debt repayment without incurring further debt while living in any city and using any type of repayment plan for any level of education debt included in our models.
For all levels of borrowing across the three career tracks, regardless of the repayment plan or the discretionary income available early in Dr. Median’s career, our models show her household with ample finances for retirement.
Our results demonstrate that a primary care career is economically viable for today’s medical school graduates, with the caveats that we noted above (a nonstandard repayment plan or living in a less expensive area) for the higher debt levels. At the highest levels of borrowing, if Dr. Median pursued a primary care career, she faced a challenge in repaying her education debt. Dr. Median could meet the challenge of repaying that extreme level of education debt while pursuing primary care, but she would likely need to consider options such as living in a lower-cost area and choosing a 25-year extended repayment plan with its larger interest amounts, or participating in a federal loan forgiveness/repayment program. Although physicians have historically shown little interest in loan repayment programs, such as the NHSC, recent federal budgets have included considerably more overall NHSC funding, which may trigger increased interest among new medical school graduates.
We recommend that future research efforts focus more directly on the influence of education debt on specialty choice by exposing undecided medical students to the economic details of the education debt repayment options to determine whether they report that this information influences their decision making.
Potential repayment challenges
Although education debt levels for indebted medical school graduates in the classes of 2009–2011 remained stable, if they were to increase in the coming years at a rate comparable to those that we have seen historically, an increased number of indebted graduates, particularly those pursuing primary care, might need to consider repayment options beyond the standard 10-year repayment plan, which could make primary care less attractive to some graduates.
If a household’s income were less than that in our model, which is based on median statistics, because of a lower income level for either spouse, the household would face challenges managing their education debt repayment, particularly if they chose a standard 10-year repayment plan while living in an expensive city. In addition, if a household were to maintain a living standard above that of our model, which for some items is based on federal statistics for households in the “$70,000 and above” category, it would be faced with challenges to balance loan repayment with living expenses.
For all levels of borrowing, the PSLF repayment scenarios (Scenarios 9, 17, and 25) seem to be among the most financially advantageous, because the household’s monthly IBR loan repayment would be tied to its income, not its debt, and, per the program’s guidelines, a sizeable portion of the loan balance would be forgiven after 10 years of repayment. However, there is an opportunity cost to this program in that it requires employment at a nonprofit where the income potential for a physician would likely be lower than if the physician were in private practice. Thus, over 10 to 20 years, the short-term financial gain afforded by the program’s loan forgiveness potential could be outstripped by the long-term financial advantages of having a higher annual salary each year. Although a full accounting of this opportunity cost is beyond the scope of our study, the scenarios that we modeled mask this trade-off because the starting salaries that we used were based on data from physicians in private practice, not those working for a nonprofit. Physicians considering PSLF would likely benefit from consultation with certified financial planners to evaluate the program in the context of their individual loan portfolios and potential career paths.
First, our findings are based on a financially savvy household that follows the ESPlanner guidelines on optimal spending and saving levels to minimize the financial shock of major future expenses, such as purchasing a home or having a child. In scenarios with forbearance during residency, our models had the household saving during residency to prepare for loan repayment and other expenses after residency. For residents and early-career physicians who are not as financially savvy or disciplined as our fictional household, our findings may represent an optimistic assessment of a primary care physician’s ability to repay such education debt levels without increasing her household’s debt. However, if the household were to maintain a standard of living below the assumptions of our household spending model, augment the household income, or choose a location with a lower cost of living, managing debt repayment would be easier than we predicted.
Second, in our sensitivity analyses, family planning decisions had the greatest impact across scenarios. For example, we found the decision to start a family later or space children farther apart to be financially advantageous. Alternatively, we found that having more than two children strained some scenarios where high levels of borrowing were repaid via a standard 10-year repayment plan in an expensive city. Thus, a household that makes different family planning decisions than Dr. Median’s might not follow the scenarios that we modeled.
Third, we assumed continuous NHSC support, if any NHSC support, in our scenarios. The loan repayment program, though, offers no guarantee beyond the initial two-year commitment, as participants must amend their contract each additional year, which is subject to the availability of appropriated funding for the program. Thus, our scenarios that modeled NHSC support beyond two years would not hold true for physicians who receive only the initial two-year commitment.
Fourth, all our scenarios included a married working couple with children, and our conclusions followed from that baseline assumption. The following scenarios, which are beyond the scope of our study, would generate different outcomes and conclusions: an indebted MD who remains single; a married MD whose spouse does not have an income, works part-time, or has significantly more debt than the amount we modeled; a married MD who experiences a divorce; an MD married to another MD with medical school debt of his or her own; or an MD who is much older at the start of medical school. These scenarios provide interesting opportunities for future research.
Our economic modeling of a physician’s household income and expenses across a range of medical school borrowing levels in high- and moderate-cost living areas shows that physicians in all specialties, including primary care, can repay the current median level of education debt. At the most extreme borrowing levels, even for physicians in comparatively lower-income primary care specialties, options exist to mitigate the economic impact of education debt repayment. These options include an extended repayment term or federal loan forgiveness/repayment program, such as IBR, PSLF, and the NHSC. Each option, however, has its own trade-offs. Our findings are important because, regardless of the degree to which education debt and/or income expectations influence specialty choice, medical students and new physicians should understand the long-term financial implications of their career choices.
Acknowledgments: The authors thank Julie Fresne, Henry Sondheimer, and Paul Jolly, all of the Association of American Medical Colleges, for their careful review of earlier drafts and helpful suggestions.
Funding/Support: The work of Dr. Wiecha and Dr. Koehler was supported, in part, by a Primary Care Training grant (#D56HP10305-03-00) to Dr. Wiecha from the Health Resources and Services Administration, U.S. Department of Health and Human Services.
Other disclosures: None.
Ethical approval: Not applicable.
2. Youngclaus J. Author’s 2011 analysis of Association of American Medical Colleges Graduation Questionnaire and Consumer Price Index–All Urban Consumers (CPI-U) data from 1978 to 2011 [available on request]
3. Rosenblatt RA, Andrilla CH. The impact of U.S. medical students’ debt on their choice of primary care careers: An analysis of data from the 2002 medical school graduation questionnaire. Acad Med. 2005;80:815–819
4. Woodworth PA, Chang FC, Helmer SD. Debt and other influences on career choices among surgical and primary care residents in a community-based hospital system. Am J Surg. 2000;180:570–575
5. Marci CD, Roberts TG. The increasing debt of medical students: How much is too much? JAMA. 1998;280:1879–1880
6. Kassebaum DG, Szenas PL. Relationship between indebtedness and the specialty choices of graduating medical students: 1993 update. Acad Med. 1993;68:934–937
7. Rosenthal MP, Marquette PA, Diamond JJ. Trends along the debt–income axis: Implications for medical students’ selections of family practice careers. Acad Med. 1996;71:675–677
8. Colquitt WL, Zeh MC, Killian CD, Cultice JM. Effect of debt on U.S. medical school graduates’ preferences for family medicine, general internal medicine, and general pediatrics. Acad Med. 1996;71:399–411
9. Baker LC, Barker DC. Factors associated with the perception that debt influences physicians’ specialty choices. Acad Med. 1997;72:1088–1096
10. Berg D, Cerletty J, Byrd JC. The impact of educational loan burden on housestaff career decisions. J Gen Intern Med. 1993;8:143–145
11. Fox M. Medical student indebtedness and choice of specialization. Inquiry. 1993;30:84–94
13. Kahn MJ, Markert RJ, Lopez FA, Specter S, Randall H, Krane NK. Is medical student choice of a primary care residency influenced by debt? MedGenMed. 2006;8:18
14. McDonald FS, West CP, Popkave C, Kolars JC. Educational debt and reported career plans among internal medicine residents. Ann Intern Med. 2008;149:416–420
15. Spar IL, Pryor KC, Simon W. Effect of debt level on the residency preferences of graduating medical students. Acad Med. 1993;68:570–572
16. Frank E, Feinglass S. Student loan debt does not predict female physicians’ choice of primary care specialty. J Gen Intern Med. 1999;14:347–350
17. Thornton J, Esposto F. How important are economic factors in choice of medical specialty? Health Econ. 2003;12:67–73
18. Ebell MH. Future salary and US residency fill rate revisited. JAMA. 2008;300:1131–1132
19. Kassebaum DG, Szenas PL. Factors influencing the specialty choices of 1993 medical school graduates. Acad Med. 1994;69:163–170
20. Xu G, Rattner SL, Veloski JJ, Hojat M, Fields SK, Barzansky B. A national study of the factors influencing men and women physicians’ choices of primary care specialties. Acad Med. 1995;70:398–404
21. Newton DA, Grayson MS, Thompson LF. The variable influence of lifestyle and income on medical students’ career specialty choices: Data from two U.S. medical schools, 1998–2004. Acad Med. 2005;80:809–814
22. Rosenthal MP, Diamond JJ, Rabinowitz HK, et al. Influence of income, hours worked, and loan repayment on medical students’ decision to pursue a primary care career. JAMA. 1994;271:914–917
23. Petersdorf RG. Financing medical education. Acad Med. 1991;66:61–65
24. Hernried J, Binder L, Hernried P. Effect of student loan indebtedness and repayment on resident physicians’ cash flow. An analytic model. JAMA. 1990;263:1102–1105
25. Palmeri M, Pipas C, Wadsworth E, Zubkoff M. Economic impact of a primary care career: A harsh reality for medical students and the nation. Acad Med. 2010;85:1692–1697
26. . Economic Security Planner Inc. ESPlanner version 2.16.64 [software]. http://www.esplanner.com
. Accessed September 18, 2012
30. Physician Placement Starting Salary Survey 2009 Report Based on 2008 Data.. 2009 Englewood, Colo Medical Group Management Association
42. Sallie Mae.How America Pays for College: Sallie Mae’s National Study of College Students and Parents, Conducted by Gallup. http://www.salliemae.com/howamericapays
. Accessed August 21, 2008 [URL no longer available]