May 01, 2017 | By Liz Lierman, Eleanor Townsley, Jenny Watermill, and Dyanne Rousseau
TAGS: graduate outcomes, internships, surveys, journal
NACE Journal, May 2017
The number of internships completed by U.S. college students has increased over the last 20 years.1 The uncertain and competitive job market that emerged in the wake of the 2008 world financial crisis refocused public attention on the role of internships in undergraduate education as a way to prepare for career success. High-quality internship experiences are increasingly understood to be an integral part of an excellent undergraduate education, with prominent educational organizations advocating internships as a pathway to career success for college students.2,3,4 Research about the impact of internship experiences on educational and career outcomes within the first six months of graduation, however, is equivocal. Both the Council for the Advancement for Standards in Higher Education5 and the National Association of Colleges and Employers (NACE) have focused on the conditions of internships for undergraduate students, raising questions about the educational value of internships—especially that of unpaid internships—in the context of labor law.6,7,8,9,10
While completing a paid internship is positively correlated with receiving a full-time job offer prior to college graduation,11 evidence for the impact of unpaid internships on postgraduate outcomes is less clear cut. Phil Gardner’s research12 for Intern Bridge, for example, indicated that unpaid internships carried educational benefits, but the immediate career dividend was less apparent. Gardner’s work also showed that women were significantly more likely than men to be engaged in unpaid internships, and liberal arts majors in the social science and humanities fields were more likely to engage in unpaid internships than those in engineering, computer science, business, and communications. More recent research by Andrew Crain13 confirms these general findings. Crain analyzes student survey data from a large, public southeastern university to show that, while unpaid internships have a positive impact on students’ academic experiences and career-seeking behaviors in college, there is no evidence of a direct impact of unpaid internships on postgraduate career outcomes. Since unpaid internships are more likely to be found among female, liberal arts, education, and health students,14,15 this difference in internship experience interacts with gender and field of academic major to produce inequality in postgraduate outcomes by field of study.
In this context, our response to the request for research proposals16 is an effort to assess the impact of experiential learning on career outcomes. We explore the relationship between internship participation and educational and career outcomes using the case study of a small New England liberal arts college with a large-scale internship program.
As a liberal arts college for women, Mount Holyoke College has watched the national conversation about internships closely. Starting in 2012, we began a series of conversations about what kind of internships matter for career success among liberal arts graduates. As we assessed our own student data, we were struck both by the economic obstacles and the lack of information that prevented many students from pursuing internships. Together, as a community of faculty, students, staff, and trustees, we sought an approach to internships that would embed high-quality, educationally substantive internships within the deep intellectual experience provided by our traditional liberal arts curriculum.
The result of these conversations was the introduction of a centralized, clearly communicated, universal internship funding program, which we named the Lynk.17 Since 2014, the Lynk internship program has guaranteed a flat rate stipend of $3,000 for domestic internships and $3,600 for international internships to every student at least once during her undergraduate career. The program was designed intentionally to ensure access to high-quality, substantial summer internships for all students, regardless of financial background. In addition to flat stipends, Lynk provides additional grant support to offset the summer earnings burden for students with very high financial need, and it includes additional stipends for internships in expensive world cities or internships that involve international travel costs.
The Lynk program also aims to deliberately connect internships to liberal arts courses, and our application and evaluation system for Lynk internships mandates that students complete a reflective curriculum as part of their internship experience. This includes one-on-one internship advising; resume reviews; workplace preparation; and online modules on goal setting, financial management, and human subjects protections, before the internship, as well as credit-bearing reflection and public speaking requirements upon return. The preparation and reflection elements of the internship program draw from the literature on reflection and capstone courses that suggests the value of learning in an internship can be better leveraged if students are well prepared for internships and if they reflect on their learning when they return to campus following an internship.18,19 These elements are also in line with the concern expressed both by the Department of Labor20 and NACE21 that internships should provide an educational benefit to students.
Our research model assumes that postgraduate outcomes are produced at the complex social intersection between students’ individual social and academic characteristics, students’ career-seeking and internship behaviors, the opportunities students are able to access for internships during college, and, following graduation, the opportunities they are able to access in employment and higher education. (See Figure 1.)
The implementation of our new internship program has created a natural laboratory to study the relative impact of these different factors on postgraduate outcomes. We compare paid and unpaid internships across industries and consider the effects of guaranteeing an internship to students versus asking students to compete for funding. We also compare different kinds of paid internships on student outcomes. The following questions focus our research on the impact of internships, both in sharpening educational and career choices, and supporting postgraduate career success:
We analyze the impact of internships on student educational and career outcomes in a series of linked regression analyses using data on the Mount Holyoke graduating classes of 2013, 2014, and 2015 (n=1,818). Data were combined from multiple sources, including anonymized student records, internship reports, utilization data from the career development center, and surveys combined with a “knowledge rate” strategy to collect postgraduate outcomes information. Regression is used to isolate the independent effects of student characteristics on the probability of completing different kinds of internships. This enables us to see the net effect of internships on postgraduate first destinations, and to assess the impact of several different kinds of internships, including employer-paid, unpaid, and college-funded internships. We can also distinguish the industry of internship, and we can compare internships before and after the implementation of the new Lynk internship funding program in 2014.
Mount Holyoke’s student population is distinctively diverse, which facilitates comparisons between student sub-groups that are sometimes difficult to study because of their smaller size in other samples of undergraduate students. International students, in particular, constitute a larger percentage of Mount Holyoke students relative to the national undergraduate student population. In addition, nearly 100 percent of Mount Holyoke students are women, creating a “natural control” for gender in this study.
For internships themselves, we count the total number of summer internships, distinguishing between employer-paid, college-paid, and unpaid internships. We also distinguish between internships in the for-profit, government, and nonprofit sectors. We observe that about one-quarter of all students do reflective curriculum associated with their internships. Tracking previous research, we find that internships taken up by students at our small private college are more likely to be in the nonprofit sector (53.9 percent) than in the for-profit (21.4 percent) or government (15 percent) sectors. While more than three-quarters of all students participated in at least one internship (78.3 percent) and nearly half participated in two or more internships (46.1 percent), only 24.8 percent of all internships were employer-paid. This compares with 41.3 percent of internships that were unpaid. An additional 29.9 percent of internships were supported by a college-funded Lynk internship stipend. The total number of internships during an undergraduate career and the payment status of internships are dependent variables in our analyses.
The first destination of students six months after college graduation is the third dependent variable in this analysis. We followed NACE coding guidelines22 to construct the variable, which shows that 96.7 percent of all graduates were either employed (75.3 percent) or in graduate education programs (21.5 percent) six months after graduation, while 3.3 percent of graduates were seeking employment. Our sense is this may overestimate employment and graduate school, since response bias on surveys of initial career outcomes six months after graduation is positively skewed.
To explore our research questions, we use three sets of regression analyses to model initial career outcomes, the total number of internships, and the source of funding of internships. We conclude with an investigation of whether or not the introduction of a universal internship funding program has improved student access to internships.
Our principal finding is that GPA and the total number of internships a student participated in during her undergraduate career are the major predictors of initial career outcomes. Graduates with more internships in college had much higher odds of being employed at six months relative to those who never participated in an internship. Similarly, graduates with higher GPAs had much higher odds of being employed compared to those with lower GPAs.
An exploratory analysis of competing alternative models examined many other possible predictors of employment, including industrial sector and source of funding of internships, race, citizenship, and financial need. None of these factors significantly affected the odds of being employed versus the odds of seeking employment six months after graduation. That said, an expanded model that investigated all possible interaction effects did marginally improve the model’s explanatory power. It showed that the effect of the total number of internships on first-destination outcomes six months after graduation depends on frequency of visits to the career center, although the main effect of career center use is not significant. This means that the impact of career center visits on initial career outcomes is fully expressed through its impact on total number of internships. In short, the relative odds of being employed versus seeking employment is jointly produced by GPA on the one hand, and the interaction of career center use and number of internships on the other. GPA remains, by far, the strongest predictor in the model.
Measures of academic strength are most strongly associated with the odds of being in graduate school at six months relative to the odds of seeking employment. Students in graduate school had higher GPAs, and were more likely to have declared a double major or a second major in a different academic division rather than two humanities majors or two science majors. A higher number of internships during an undergraduate career was also associated with higher relative odds of being in graduate school compared to job seeking at six months.
Differences between those in employment and those in graduate school six months after graduation are more subtle, and GPA again has the strongest influence. This is followed by citizenship, with international students more likely to be in graduate school than domestic students. The difference between those in graduate school and those in employment is also associated with a range of other characteristics, including academic major (natural science majors had higher odds of being in graduate school) and financial need (those with higher need were more likely to be employed than in graduate school). Being a frequent user of the career center was associated with being in employment rather than graduate school. There is no difference in internship participation between graduates who entered employment and those who entered graduate school.
In summary: the major predictors of being employed or in graduate school on the one hand, compared to seeking employment on the other, are GPA and total number of internships in a college career. The differences between those in employment and those in graduate school are more fine-grained. Citizenship status, academic major, financial need, and using the career center distinguish between those who are employed and those in graduate school six months after graduation.
Since the total number of internships a student completed as an undergraduate is a significant predictor of first destinations, our second analysis examines the factors associated with participation in at least one internship compared with no internships, and with participation in multiple internships compared with one internship.
International student status is associated with higher odds of internship participation, a finding that does not appear in the existing research literature, likely because international students often form too small a percentage of an overall student population for clear comparisons. International students who are also frequent users of the career center have even higher odds of participating in one or more internships.
Declaring a natural science major, having a high GPA, participating in reflection curriculum, and using the career center frequently are also associated with higher odds of internship participation. Traditional-aged, four-year students are more likely than transfer students and much more likely than nontraditional-aged students to participate in one or more internships.
In addition, being a frequent user of the career center not only improves the odds of a second internship compared to one internship, but also this effect becomes more pronounced for a second internship than the first internship.
Being an international student or declaring a natural science major both increase the odds of participating in two or more internships compared with one internship. Having a higher GPA and doing reflection curriculum also improve the odds of participating in multiple internships compared to one. Although there were no significant racial differences between those who participated in a first internship rather than no internships, students who identify as Asian-American or Latina had higher relative odds of participating in multiple internships versus one internship.
Being a transfer student or nontraditional-aged student decreased the odds of participating in two or more internships compared with one internship. However, in further analysis, student type is found to interact with using the career center. That is, if transfer students and nontraditional-aged students (Frances Perkins Scholars) use the career center, their odds of a second internship are greatly increased.
For students with first internships that were employer-paid or college-funded, the odds of participating in a second internship were lower than if the first internship was unpaid. There was no significant effect of the industry of the first internship on the likelihood of a second internship.
Although the research literature suggests a positive impact of employer-paid internships on postgraduate career outcomes, we find no independent effect of internship payment source on career destinations at six months. Rather, the effect is of total number of internships completed as an undergraduate student. We therefore investigate how different types of internships contribute to internship participation and total number of internships.
Students with at least one employer-paid internship were twice as likely as those with a college-funded internship to have two or more internships. Students with at least one unpaid internship were similar to those with employer-paid internships, in that they were also twice as likely as those with college-funded internships to have two or more internships.
Academic major and industry of internship are the best predictors of whether or not a student participated in employer-paid internships. Participation in unpaid internships is correlated with academic major, and those less likely to participate in unpaid internships were students with higher GPAs, nontraditional-aged students, and frequent users of the career center.
Finally, we found that as students graduated in classes that had access to a universal internship funding program from the college, there was a decline in the number of students who never completed an internship, an overall decline in the number of students participating in unpaid internships, and a decrease in the proportion of students completing two or more internships. Significant differences between the GPAs of students participating in internships disappeared, and there is now less of a difference between students of different financial standing in their likelihood of receiving internship funding.
Our study underscores the importance of increasing student access to internship experiences and improving student access to higher total numbers of internships during a college career, as well as the central importance of academic strength—both in its impact on internships and independently—on postgraduate outcomes at six months. Universal internship funding at Mount Holyoke has increased access to first internships for students, which raises the question of whether this might be an effective approach at other institutions as well. For institutions that have already improved access to first internships for their students, our findings raise further questions about which strategies may be most effective in building on this access to increase the total number of high-quality internships per student. Additionally, our findings make clear the need for outreach and programming specifically aimed at increasing use of the career center and supporting internship participation among transfer and nontraditional-aged students.
Although fewer and farther between than other types of internships, employer-paid internships have important value, as students with employer-paid internships were more likely to complete two or more internships in total. Further research on employer-paid internships is warranted; issues include gender differences in paid internship participation, the interaction between the availability and impact of paid internships in particular industries, and employers’ use of paid internships as a recruiting strategy. Knowing more about these issues would increase our understanding of who can access employer-paid internships, and to what extent the importance of internship payment source to postgraduate outcomes may vary according to the industry.
As we strive to maximize career success and postgraduate outcomes for all our students, we know that we need to support students to succeed academically. There is also evidence that internship participation (especially participation in multiple internships over the course of a college career), engagement in career development activities, and an accurate understanding of labor market realities are central to a student’s ability to respond to postgraduate opportunities and challenges.
1 Gardner, P. (2010). The Debate Over Unpaid College Internships. Intern Bridge, Inc. Retrieved from http://www.ceri.msu.edu/wp-content/uploads/2010/01/Intern-Bridge-Unpaid-College-Internship-Report-FINAL.pdf
2 Association of American Colleges and Universities & Peter D. Hart Research Associates (2007). How Should Colleges Prepare Students to Succeed in Today’s Global Economy? LEAP National Report, College Learning for the New Global Century. Retrieved from https://www.aacu.org/sites/default/files/files/LEAP/2007_full_report_leap.pdf
3 Association of American Colleges and Universities & Peter D. Hart Research Associates (2008). How Should Colleges Assess and Improve Student Learning? Employers’ Views on the Accountability Challenge. Retrieved from https://www.aacu.org/sites/default/files/files/LEAP/2008_Business_Leader_Poll.pdf
4 Association of American Colleges and Universities & Peter D. Hart Research Associates (2010). Raising the Bar: Employers’ Views on College Learning in the Wake of the Economic Downturn. Retrieved from https://www.aacu.org/sites/default/files/files/LEAP/2009_EmployerSurvey.pdf
5 O’Neill, N. (2010). Internships as a high-impact practice: Some reflections on quality. Peer Review, 12(4), 4-8.
6 Koc, E. (2014a). Factors Associated With Job-Search Success, NACE Journal, Vol. 75, No. 2, 14.Retrieved from http://www.naceweb.org/job-market/trends-and-predictions/factors-associated-with-job-search-success/
7 Gardner, P. (2010). The Debate Over Unpaid College Internships. Intern Bridge, Inc. Retrieved from http://www.ceri.msu.edu/wp-content/uploads/2010/01/Intern-Bridge-Unpaid-College-Internship-Report-FINAL.pdf
8 U.S. Department of Labor, Wage and Hour Division (2010). Fact Sheet #71: Internship Programs Under the Fair Labor Standards Act. Fact Sheets, April, 2010. Retrieved from http://www.dol.gov/whd/regs/compliance/whdfs71.htm
9 Basow, R. R., & Byrne, M. V. (1993). Internship Expectations and Learning Goals. Journalism Educator, 47(4), 48-54.
10 National Association of Colleges and Employers (2015a). Executive summary: 2015 internship & co-op survey. Retrieved from http://www.naceweb.org/store/2016/internship-co-op-survey-2016/
11 Koc, E. (2014a). Factors Associated with Job Search Success, NACE Journal, Vol. 75, No. 2, 14.Retrieved from http://www.naceweb.org/job-market/trends-and-predictions/factors-associated-with-job-search-success/
12 Gardner, P. (2010). The Debate Over Unpaid College Internships. Intern Bridge, Inc. Retrieved from http://www.ceri.msu.edu/wp-content/uploads/2010/01/Intern-Bridge-Unpaid-College-Internship-Report-FINAL.pdf
13, 14 Crain, Andrew. 2016. “Understanding the Impact of Unpaid Internships on College Student Career Development and Employment Outcomes.” NACE Foundation Report.
15 Gardner, P. (2010). The Debate Over Unpaid College Internships. Intern Bridge, Inc. Retrieved from http://www.ceri.msu.edu/wp-content/uploads/2010/01/Intern-Bridge-Unpaid-College-Internship-Report-FINAL.pdf
16 National Association of Colleges and Employers Foundation (2015b). Call for proposals: Unpaid internships and career outcomes. Retrieved from www.naceweb.org/uploadedFiles/Pages/about-us/nace-foundation-call-for-proposals.pdf. The NACE Foundation activity was transferred to the NACE Center for Career Development and Talent Acquisition.
17 Townsley, Eleanor, Becky Packard and Eva Paus (2015). “Making the Lynk at Mount Holyoke: Institutionalizing Integrative Learning”. Peer Review. American Association of Colleges and Universities. 17(1):24-29.
18 Henze, B.R. (2006). The Research-experiential Internship in Professional Communication. Technical Communication, Vol. 53, No. 3, 339-347.
19 Fernald, P.S., & Goldstein, G.S. (2013). Advanced Internship: A High-Impact, Low-Cost, Super-Capstone Course. College Teaching Vol. 61, No. 1, 3-10.
20 Mersol, Greg. (2016). “New York District Court Grants Summary Judgment for Employer in Gawker Intern Case. Litigation Over Interns Dries Up Internship Opportunities.” BakerHostetler Blog distributed by JDSupra Business Advisor. www.jdsupra.com/legalnews/new-york-district-court-grants-summary-21412/
21 National Association of Colleges and Employers (2011). Position statement: U.S. internships. Retrieved from http://www.naceweb.org/about-us/advocacy/position-statements/position-statement-us-internships/
22 National Association of Colleges and Employers. 2014. Standards and Protocols for the Collection and Dissemination of Graduating Student Initial Career Outcomes Information for Undergraduates. Retrieved from http://www.naceweb.org/uploadedfiles/pages/advocacy/first-destination-survey-standards-and-protocols.pdf
Note: Forthcoming Study on Outcomes for the Liberal Arts
This article provides highlights from a forthcoming study conducted by the authors through the NACE Center for Career Development and Talent Acquisition.
The study joins three other NACE Center research projects that are focused on internships: Understanding the Impact of Unpaid Internships on College Student Career Development and Employment Outcomes, published in December 2016; a forthcoming study from researchers at Endicott College—highlights of which appear in the May 2017 issue NACE Journal ; and a study exploring the outcomes associated with internships across multiple institutions (to be published by the end of 2017).
Liz Lierman is director of the Career Development Center (CDC) at Mount Holyoke College. A dedicated advocate for the liberal arts, Lierman previously founded the first career services office for Bard College at Simon's Rock, and led career development programs for Oberlin College. She currently serves on the Principles for Professional Practice committee for NACE, and is an alumna of the NACE Leadership Advancement Program.
Eleanor Townsley is professor of sociology and director of the Nexus Curriculum to Career Program at Mount Holyoke College. Her recent research includes “Making the Lynk at Mount Holyoke: Institutionalizing Integrative Learning” published in Peer Review (2015), and “Science, expertise and profession in the post-normal discipline” in American Sociologist (2016). Her new work focuses on the humanities foundations of experiential learning.
Jenny Watermill is the associate director for internships and student employment at Mount Holyoke’s CDC. She supervises experiential learning and has been a critical collaborator in the implementation of Mount Holyoke’s Lynk Initiative, which connects academic work with practical applications of the liberal arts and sciences. She currently serves as the chair of the Employer Development Committee for the Liberal Arts Career Network.
Dyanne Rousseau is the program manager at Mount Holyoke’s CDC. She is responsible for data systems and analytics related to student and alumnae advising and manages the CDC’s knowledge rate research project. She has a B.A. in politics from Mount Holyoke College.
Overall unemployment rate
Bureau of Labor Statistics
Unemployment rate, bachelor’s degree grads age 20 – 24
Bureau of Labor Statistics
Average projected starting salary, 2017 humanities major
Winter 2017 Salary Survey
Average starting salary, Class of 2016 bachelor’s degree graduate
Spring 2017 Salary Survey
Percent of Class of 2015 bachelor’s degree grads employed or pursuing more education
Class of 2015 First-Destination Survey report