Before 1994, postsecondary institutions could require retirement at age 70; however, after the exemption from the Federal Age Discrimination in Employment Act lapsed, retirement was no longer mandatory.1 Consequently, the mean age of full-time faculty members in U.S. medical schools accredited by the Liaison Committee on Medical Education (LCME) has increased from 44.7 in 19872 to 49.5 in 2017.3 Also, the percentage of full-time medical school faculty members who are 55 or older increased from 19% in 19872 to 32% in 2017.3 Notably, adults 55 or older constituted, by comparison, only 28.5% of the entire U.S. population in 2017.4
Although professional medical societies have published studies on this population,5,6 and another recent publication has summarized the findings of a survey of faculty 50 or older at 1 medical school,7 the academic medicine community knows little about the characteristics and work–life expectations of faculty members 55 or older working at U.S. medical schools. Additionally, little is known about how to support late-career and retiring faculty. Cain and colleagues7 examined the cultural barriers to successful retirement at the University of Massachusetts and explored a framework for aligning institutional and individual needs relative to retirement policies, programs, and resources. Their process resulted in innovative solutions that reduced cultural barriers and led to a new institutional policy to meet the needs of both the institution and its late-career faculty, but their research is from a single institution and constitutes just a first look at late-career faculty members’ experiences.
We conducted a survey of faculty members 55 or older and of faculty affairs and faculty development officers to better understand not only the characteristics and expectations of late-career faculty but also the resources and policies in place at medical schools to support this population. In an accompanying article in this same issue of Academic Medicine,8 we have presented the results of our latter survey, providing an inventory of institutional resources for late-career-stage retirement planning and development available at U.S. medical schools.
Here, we present a first attempt to characterize and assess the status of late-career faculty members who work full-time in LCME-accredited medical schools across the United States. The objectives of our study were twofold: (1) to characterize late-career, full-time faculty members; and (2) to assess late-career-stage faculty members’ work–life expectations.
We surveyed full-time, late-career faculty members at 14 U.S. LCME-accredited medical schools from May through September 2017. We defined “late career” as faculty members 55 or older. After careful consideration, we chose a convenience sample strategy because a population-based survey of all U.S. medical schools quickly became quite complex, in part because of individual institutions’ research ethics review processes. We identified a convenience sample of 14 schools from the steering committee, subcommittees, and membership of the Group on Faculty Affairs (a member group of the Association of American Medical Colleges [AAMC]). We chose the medical schools in our sample to maximize representativeness of the U.S. medical school population based on region, public/private status, community-based status, financial relationship with parent university, and full-time faculty size (Table 1). Coinciding with our project start date of December 2016, we used data from the 2016 AAMC Organizational Characteristics Database and from a December 31, 2016 snapshot of the AAMC Faculty Roster to define our study sample.9,10
We developed a questionnaire that included standard sociodemographic questions on, for example, age, sex, race/ethnicity, academic appointment (or primary role), length of appointment, and years in academic medicine. The questionnaire also included items assessing late-career work–life expectations in terms of time to expected retirement, personal and professional factors associated with retirement decisions, interest in part-time retirement, mission areas to continuing working in upon retirement, nonacademic activities to pursue upon retirement, and items or services desired from the institution upon retirement (see Supplemental Digital Appendix 1, available at http://links.lww.com/ACADMED/A704).
The study was approved by the American Institutes for Research’s institutional review board. Survey respondents were informed—both in the introductory email from their institution’s faculty affairs and development office leaders and in the survey instructions—that the survey was anonymous, that participation was voluntary, and that they could skip any question.
We conducted, first, univariate analyses to evaluate the data for the aggregated total and, then, bivariate analyses to examine within-group differences by age and academic role. For the bivariate analyses, we used independent-samples t tests to compare mean differences, and chi-square and McNemar tests to compare proportionate differences. For age, we compared faculty members who were 55 to 64 years old with faculty members 65 or older. For academic role, we compared basic scientists, clinicians, clinician–scientists, and “others” who, for example, identified as administrators or full-time educators. We analyzed data using SPSS for Windows (version 24, Chicago, Illinois).
The total number of full-time faculty 55 or older in the 14 participating institutions was 5,204. Of these, 2,126 faculty members (40.8%) responded. Below we have presented demographic information and the work–life expectations of respondents. Because faculty members could skip any question, the number of respondents for each question varies. We have presented only the raw number and percentage of faculty members providing specific personal information or endorsing a response (see also Tables 2–4).
Demographics of faculty who are 55 or older
The average age of the respondents was 62.3 years (standard deviation [SD] = 5.4; range = 55–88). A little more than one-third of the respondents were 55 to 59 (753; 36.0%), about a third were 60 to 64 (738; 35.3%), and the remainder were 65 or older (598; 28.6%). We detected no significant differences in age by academic role (Table 2). The majority of the faculty respondents were male (1,425; 67.2%). Male faculty members responding to our survey constituted a statistically significant larger proportion of respondents in the 65 or older age category (489; 81.8%) compared with the 55 to 64 category (915; 61.5%). Additionally, male respondents constituted larger proportions among the clinicians (593; 70.6%) and clinician–scientists (395; 70.9%) compared with the basic scientists (278; 61.1%) and those in other roles (122; 59.5%). The majority of respondents self-identified as white (1,841; 88.3%). A much smaller percentage self-identified as Asian (131; 6.3%); as Hispanic, Latino, or Spanish (55; 2.6%); as black or African (28; 1.3%); or as other mixed race/ethnicity (30; 1.4%). We detected no race/ethnicity differences by age; however, by primary role, those self-identifying as Asian were disproportionately represented among basic scientists (51; 11%). Most of the respondents (1,803; 85.5%) were married or partnered, and we detected no statistical differences in relationship status by age or primary role.
About two-thirds of our respondents (1,376; 65.6%) reported that they were professors, and about a quarter (455; 21.7%) reported that they were associate professors. Respondents who were 65 or older had a higher proportion of professors (471; 80.0%) compared with their relatively younger (55–64) peers (882; 59.9%). Clinician–scientists were also disproportionately professors (441; 79.6%) compared with clinicians (455; 54.2%). Slightly more than half (1,088; 51.9%) of the respondents reported that they were either tenured or on a tenure track. Respondents 65 or older were more likely to report tenure (364; 61.8%) compared with those between the ages of 55 and 64 (702; 47.7%). Likewise, basic scientists (352; 77.4%) and clinician–scientists (369; 66.6%) were more likely to report tenure compared with their clinician (236; 28.2%) and other (110; 53.7%) peers.
Work–life expectations of faculty who are 55 or older
A little more than a third of respondents (760; 37.2%) reported having worked in academia between 30 and 39 years, a similar number (720; 35.3%) reported between 20 and 29 years, and 10% each reported 10 to 19 years (212; 10.4%) and 40 to 49 years (206; 10.1%). In general, clinicians reported fewer years working in academia compared with basic scientists, clinician–scientists, and those in other roles. One-third (664; 32.6%) of all respondents indicated that they had worked at their current institution for 20 to 29 years, and a third of respondents 65 or older (193; 33.6%) indicated working at their current institution for 30 to 39 years. Clinicians and other respondents reported similar lengths of employment, generally fewer years than basic scientists and clinician–scientists (data not shown).
When asked if they had thought about or had begun to think about retiring from full-time employment in academic medicine, 915 respondents (45.2%) answered yes, and collectively they indicated that they might retire at an average age of 67.8 (SD = 4.3; see Table 3). One-fifth (397; 19.6%) responded that they were not sure when they would retire but would likely do so in the next 5 years, and one-quarter (500; 24.7%) similarly reported that they were not sure when they would retire but would likely do so in the next 10 years. A little over a tenth (213; 10.5%) indicated that they had not yet begun to think about retiring. The respondents over 65 were slightly more likely to respond that they were planning to retire or thinking about retiring (278; 49.0%)—likely at age 70.8—than those between 55 and 64 (626; 44.0%) who thought they would retire at age 66.5. Clinicians and those in other roles were more likely to report thinking about retirement (406 [49.3%] and 101 [49.5%], respectively) compared with basic scientists (172; 38.9%) and clinician–scientists (230; 42.1%). Correspondingly, clinicians and others also reported a slightly younger average planned retirement age (67.3) compared with basic scientists and clinician–scientists (68.1 and 68.7, respectively). Respondents were asked whether, upon considering retirement, they would accept an option to move to a part-time appointment, and 50.0% (1,004) agreed (data not shown).
Over half of the respondents identified the following 4 personal factors (listed in descending order) that might affect their retirement decisions: their personal health (endorsed by 1,275 respondents [62.5%]), plans for postretirement (1,187 respondents [58.2%]), spouse’s/partner’s plans and opinion (1,154 respondents [56.6%]), and personal finances (1,104 respondents [54.1%]).
Respondents also identified professional factors that might affect their retirement. The top 8—all endorsed by at least one-fifth of the faculty members—were as follows:
- institutional/departmental offerings for phased retirement or part-time appointment (endorsed by 1,139 respondents [55.8%]);
- changes in institutional leadership (844 respondents [41.4%]);
- presence of a successor (593 respondents [29.1%]);
- challenges with grant funding (566 respondents [27.7%]);
- change in allocation of effort (559 respondents [27.4%]);
- challenging relationship with supervisor (513 respondents [25.1%]);
- lack of academic role clarity or certainty (456 respondents [22.4%]); and
- willingness of department to continue to support those who depended on them (421 respondents [20.6%]).
See Table 3 for the full list of factors that might affect respondents’ retirement plans.
Late-career faculty members responding to our survey reported that if they could choose 1 academic mission area in which to continue to spend their time after retiring from full-time employment, they would choose teaching/education (775; 39.3%), research/scholarship (586; 29.7%), patient care/clinical services (370; 18.8%), and administration/institutional service (242; 12.3%). See also Figure 1. Respondents 65 or older were more likely than their younger counterparts to indicate research/scholarship (191 [34.0%] vs 380 [27.6%]). Basic scientists were more likely to indicate research (231; 53.2%) and teaching (146; 33.6%), clinicians more likely to choose teaching (389; 48.6%) and patient care (252; 31.5%), clinician–scientists more likely to endorse research (260; 49.2%) and teaching (139; 26.3%), and those in other roles more likely to select teaching (99; 49.5%) and administration (48; 24.0%) (data not shown).
When asked to indicate 3 of the most important items they would like from their institution upon full-time retirement (Table 4), the most common responses, all endorsed by at least a third of respondents, were (in descending order) as follows: health insurance (endorsed by 1,369 respondents [68.2%]), email (949 [47.3%]), part-time teaching roles (839 [41.8%]), and library services (683 [34.0%]).
Through our research, we characterized late-career, full-time faculty members 55 or older across 14 U.S. LCME-accredited medical schools and assessed their work–life expectations. The results from our multi-institution sample survey of 2,126 faculty members provide important information that will allow faculty affairs and development leaders to create policies, programs, and resources to support late-career transitions.
On average, late-career faculty members in U.S. medical schools are 62 years old, male, white, and married or partnered. About 40% are clinicians, and approximately one-quarter are clinician–scientists or basic scientists (the remaining 10% or so work in other roles). While 11% reported that they had not even begun thinking about retirement, 44% were not sure when they might retire, and only 45% had begun thinking about retirement. This high level of uncertainty and low level of retirement preparedness indicate ample opportunity for institutions to act. For example, the findings that half of our respondents would consider moving to part-time status and that more than a third wanted to continue working in education/teaching or research/scholarship indicate a need for part-time policies and benefits for late-career and retired faculty. Faculty between 55 and 64 who are thinking about retirement anticipate an average retirement age of 67, whereas faculty members 65 or older anticipate an average retirement age of 71. Interestingly, however, only 29% of those 65 or older plan to retire in the next 5 years, and only 16% plan to retire in the next 10 years. If these faculty members’ plans become reality, then U.S. medical schools will employ a large number of full-time faculty who are 80 or older—a situation that will require an entirely new set of faculty development tools. Indeed, our findings support an earlier demographic analysis showing that the current North American workforce includes a growing number of individuals older than 65.11 Given age-related physical and cognitive decline11 and the absence of a national mandatory retirement age,1 the academic medicine community must begin to consider and develop policies and guidelines around medical license renewals,11 workplace accommodations, academic roles and responsibilities, and enhanced institutional support.12
A majority of respondents identified 4 personal factors that might affect retirement decisions: their health, their plans post retirement, their spouse’s or partner’s plans and opinions, and their personal finances. Ironically, even though health was the most-endorsed personal factor, we believe that insufficient time may be faculty members’ biggest barrier to participating in health programming, in addition to the possible stigma of perceived frailty. Thus, medical schools need to embrace new technologies that support cultures of health and wellness, and they need to normalize the aging process. Myriad apps for mobile devices, as well as podcasts and other fitness technologies in eHealth,13 are designed to be seamlessly embedded into busy lives. These innovations present an opportunity for faculty development leaders to actively engage with faculty, as well as with institutional wellness and work–life offices, to identify useful tools and best practices. A central repository for such tools and best practices, and for policies and resources—particularly those developed by medical organizations and specialties—would be a valuable asset to the faculty development community.
Second, to help faculty think about their lives after retirement, programming should start early in each faculty member’s career. Transition planning both takes time and evolves over time. Coaching and mentoring, annual reviews, and developing individual career plans are all means of engaging with faculty members around life transitions—and are vital to their future.
Third, pre- and postretirement programming should include faculty members’ partners and spouses. In our experience, faculty are often torn between wanting to engage in institutional events or offerings and not wanting to infringe on personal time; therefore, inviting spouses or partners and family members to participate in programming may result not only in higher attendance and participation but also in improved faculty morale. Additionally, preretirement programming should provide anticipatory socialization and conceptual mapping for life after full-time employment. The freedom that faculty members experience after fully, or even partially, retiring has both advantages and disadvantages. For some late-stage faculty and early retirees, we hypothesize that the perceived lack of structure may be unnerving, and some may be reluctant to retire fully for fear of “not knowing what to do with myself.” Faculty affairs and faculty development officers need to identify mental models to help faculty “adopt lifestyles that give coherence, structure, and meaning to their leisure.”13 Personnel in faculty affairs and development must also develop tools that specifically encourage faculty members to explore activities and roles that they may have neglected during their peak career years (e.g., extracurricular interests and hobbies, volunteer work, self-development, fitness, travel, engaging with family and friends). For that matter, helping all faculty members develop their whole selves, regardless of their rank, would surely buttress a vital academic community.
Fourth, junior and early-career faculty should learn about personal financial investing upon entering their academic medicine career, and they should receive regular financial health checkups. By midcareer, faculty members should have a heightened awareness about financial planning for retirement, taxes, health care, housing options, estate management, etc. Although not necessarily the purview of faculty affairs, a personal financial and wealth management program certainly calls for active partnership with human resources (HR) offices and other stakeholders, including hospital partners.
Professional factors likely to influence retirement decisions showed greater variability compared with personal factors. At least one-fifth of respondents identified 8 factors: institutional or departmental offerings for phased retirement, changes in institutional leadership, the presence of a successor, challenges with grant funding, changes in allocation effort, changing relationship with supervisor, lack of role clarity or certainty, and willingness of the department to continue to support those who depend on the faculty member. We identified 4 common themes among these factors: policy, flexibility and innovation, leadership and succession planning, and funding. Additionally, these 8 factors most endorsed by our faculty respondents highlight the need for partnerships between faculty affairs/faculty development offices and other institutional offices or departments, including finance. Faculty affairs and development officers should collaborate with finance experts to develop and disseminate retirement policies. They should also collaborate with department leaders to create succession planning models that include personnel and leadership strategies, as well as financial models, for supporting staff. Faculty development officers should also consider partnering with HR departments to develop and invest in deliberate, formal succession planning programs for department chairs and other leaders. If institutional leaders become more skilled at succession and transition planning, a culture of growth and renewal may encourage faculty members who are reluctant to engage in conversations about their own transition planning. Ironically, current workforce projections point to a substantial physician deficit in the next 10 years,14 so physicians nearing retirement might be asked or encouraged to keep working full- or part-time. These trends further justify the need for institutional and national attention focused on late-career faculty. One important caveat in promoting active partnerships with HR offices is to recognize the unique value that faculty affairs and development offices bring to faculty members. The medical and scientific training that many in faculty affairs and faculty development have may afford these professionals increased credibility among faculty.
The results of this survey point to exciting opportunities for faculty development leaders to engage with late-career faculty members. About half of the faculty members indicated that they would consider part-time employment. Further, more than one-third indicated that they would enjoy remaining in education or teaching, and another third indicated that they would enjoy continuing in research or scholarship. Engaging retired faculty with learners and younger faculty would require careful and thoughtful orchestration involving diverse stakeholders but would likely have tremendous payoff for students and other trainees, junior faculty, departments, and various program directors—as well as for the retirees. We know that late-career faculty identify very strongly with their academic life and crave ongoing intellectual stimulation. Faculty development leaders should work with undergraduate medical education and residency curricula leaders to design programs that would train, vet, and certify retirees for various roles throughout the institution. For example, retirees could learn about current best practices for teaching and mentoring or become certified in coaching.
Faculty development pre and post retirement may occur in conjunction with an emeritus college, retiree academy, or retiree center. Some institutions have created these and similar organizations for retirees. Not only do these efforts provide ongoing intellectual and social stimulation for retirees but they also offer the potential to become vibrant centers, vital to institutions for preserving the vast intellect and skills of faculty. Retiree academies provide the physical, mental, and emotional space for retiring and retired faculty members to assist one another with career transitions via mentoring, retooling, legacy building, and continued engagement in academia. Building and training a cadre of “retirementors”—retirees imparting their wisdom on how to live life with vitality and vigor15—would be one way to embed program sustainability and foster community. The return on institutional investment for an emeritus college or similar effort would likely be high.
In addition to teaching and mentoring roles and opportunities, faculty most commonly want, upon full-time retirement from the institution, health insurance, email, and library services. Health insurance is understandably a critical component to retirement because of its high cost, especially for faculty members who may be ineligible for Medicare. Depending on the future status of Medicare and Social Security, universities may need to reexamine their provision of health benefits in retirement. Finally, institutional email and library services are relatively easy, symbolic gestures that require nominal institutional investment.
To our knowledge, this is the largest multisite survey of full-time, late-career faculty members in U.S. LCME-accredited medical schools; however, we acknowledge some limitations. Our findings may not be generalizable to faculty at some individual medical schools because we surveyed a representative, convenience sample of faculty from only 14 of the 145 LCME-accredited U.S. medical schools. We acknowledge that the institution-specific policies of the 14 schools may have biased the responses of the faculty survey participants. Furthermore, the majority of our respondents were white, married/partnered men; thus, these results may not be generalizable to faculty from other demographic strata. Also, these data were collected during a snapshot in time in 2017, and our response rate was only 41%. Nevertheless, our findings highlight the urgent need for national-level attention to, and possibly monitoring processes for, late-career and retired U.S. medical school faculty.
This survey represents an important first step in learning about late-career faculty members and what they need and want. We believe that the traditional view of retirement as an abrupt exit from employment has lost its relevance; rather, anticipated transition from full-time employment for medical school faculty has become more gradual and complex. This cultural and retirement construct disruption demands reinvention and innovation.13,16 The academic medicine community needs to continue these conversations—to learn more from late-career faculty and retirees and to better understand how institutions can help them. Faculty affairs and development leaders must identify and share best practices; develop and evaluate policies, programming, and resources; and measure and report the outcomes of their initiatives for late-career and retired faculty members. Importantly, work in this area should not remain only in faculty affairs or faculty development offices. Instead, dean’s office leaders and department chairs should lead the effort to develop comprehensive workforce plans (i.e., recruitment, retention, and succession). Furthermore, any efforts emanating from faculty affairs and development offices should also involve leaders in the parent university to help them understand the impact of policies and practices that both support and impede a robust school of medicine faculty life cycle. Future work that engages late-career and retired faculty will benefit these valuable faculty members and their institutions alike.
The authors would like to thank Valarie Clark, MPA, former director of constituent engagement at the Association of American Medical Colleges, for her expertise in managing this project.