https://orcid.org/0000-0002-6507-9237
![Sample our Economics, Finance,Business & Industry journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5931/TOC-Subject-Sample-2021-Economics-Finance-Business-Industry.jpg"})
ABSTRACT
The conferment of tenure at a United States university provides substantial job security to its recipients. Tenure is designed to allow a professor the ability to explore new and risky research questions without fear of losing their position due to lack of publications. At the same time, this policy creates an incentive system with an ambiguous effect on how the professor performs in the classroom. Professors may no longer care about teaching evaluations since future evaluations are unlikely to affect their job security. Alternatively, tenured professors, no longer having strenuous research priorities, may devote more resources to the teaching component of their job. This paper investigates the impact of the conferment of tenure on student evaluations of teaching. Data comes from a large, flagship state university and spans 22 semesters (2006–2017). We use an instructor-level fixed effects structure to compare end-of-semester course survey scores before and after an instructor receives tenure. We find that conditional on being granted tenure, professors experience a small, but persistent, decrease in student course evaluations. This effect is driven by professors in the top half of the course evaluation distribution.
Acknowledgments
The authors would like to thank Angela Dills, Fulya Ersoy, Gregory Gilpin, Cassandra Hart, Mikhail Kouliavtsev, Daniela Morar, Ben Ost, Lauren Russel, Katerina Sergi, Austin Smith and seminar participants at the University of New Haven, California State Chico, 2018 WEAI International Conference, 2019 AEFP Conference, and 2019 WEAI Conference. All remaining errors are our own.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 It was estimated in 1994 that around 50 tenured professors a year lose their jobs (Mooney Citation1994).
4 The state of Wisconsin, for example, effectively abolished the job protections associated with tenure in 2016.
5 Service expectations almost always increase after a professor is granted tenure, often because some committees require tenure for eligibility.
6 The University of Colorado system is made up of several universities, with the University of Colorado Boulder as the flagship campus. Although officially denoted UCB, it is often referred to as simply ‘Colorado’.
7 See Table 6 of Cheng (Citation2015) for the most comparable results.
8 Some universities have begun to use peer evaluation as well.
9 For more information, see https://www.colorado.edu/fcq/
11 See https://www.aaup.org/issues/tenure for more information. This is an admittedly biased source.
12 For complete information see https://www.colorado.edu/facultyaffairs/career-milestones/reappointment-promotion-and-tenure/reappointment-tenure-and-promotion
13 We are ignoring the service component to tenure, which only impacts a tenure decision in the rarest of circumstances.
14 According to the University of Colorado Boulder, in recent years 44 percent of all successful promotion and tenure candidates are given excellence in research and meritorious in teaching, 25 percent are given meritorious in research and excellence in teaching, and 31 percent achieve excellence in research and teaching (https://www.colorado.edu/facultyaffairs/career-milestones/reappointment-promotion-and-tenure/reappointment-tenure-and-promotion).
15 These data come in two files. The university provides a master data set, as well as data that can be manually downloaded from the university's FCQ website. When merging these two data sets, we gave priority to the manual data when any inconsistencies arose. We gave the manual data priority as it was the broader and more complete of the two data sets.
16 Data was most recently accessed on 9 September 2017.
17 These were most likely taught by non-tenure track employees and comprise of less than 0.1 percent of all observations.
18 Instructors who were tenured, either as associate professors or full professors, at the beginning of the study period are not included.
19 All medical school classes are taught at the University of Colorado Anschutz Medical Campus and are not part of the sample.
20 Beginning in Fall 2017, which is after the study period, an online-only evaluation policy began.
21 The intuition is students will feel more free to give a bad score when taken online versus in the classroom, or that students who do not like the class are more likely to not attend and thus not be present during in-class evaluations.
22 While it is impossible to verify this, anecdotal evidence suggests that instructors do follow this procedure. Students are familiar with the process and would know that an instructor staying in the room while the forms are being filled out is breaking the rules.
23 Graduate student instructors can be immediately banned from teaching if they receive less than a 4.0.
24 This provides evidence that students consider each question independently and are able to differentiate between subject material and instructor quality.
26 These classes also provide graduate students as teaching assistants, who receive their own FCQ evaluations.
27 This set of students is commonly referred to as the post-census roster or grade roster.
28 We treat a course as a new preparation as the first time an instructor has taught a class during the time frame of our sample.
29 We lose four instructors from our main sample due to lack of information regarding when their highest degree was obtained.
30 Both these results stay consistent in subsequent specifications.
31 Students who withdraw from the university during the middle of the academic year are more likely to be poor students, and worse students are thought to give lower evaluations.
32 The largest class at the University of Colorado Boulder, Principles of Microeconomics, has up to 490 students.
33 Two percent of our sample, for example, earned their terminal degree in the 1980s.
34 The results of this specification and its similarity to Column 3 are likely due to the inclusion of subject-level fixed effects. Non-traditional academic career paths are likely a function of field-of-study. For example, at a research university it is very rare for someone with an Economics PhD to have started in a non-academic position while this is relatively more common for someone with a natural sciences PhD.
35 Full results are available upon request.
36 Recitations are usually taught by graduate students but are occasionally taught by professors.
37 Prior interest in a course and the instructor overall score have a correlation of .
38 Because only 1 year is used, the indicator fall variable, professor fixed effects, and the linear years since highest degree variable are collinear. Instead of the linear and quadratic experience control variables, indicator variables for each year of experience are used.
39 Either of these options is possible, for the same reasons discussed above in the Tenure and Incentives Discussion.
40 Results available upon request.
41 For more information, see Langbein (Citation2008).
42 It is reassuring to see that students do not perceive professors to be less respectful after they earn tenure, despite the additional job security.
43 These measures are also useful in that they are not self-reported, but objective data.
44 This is technically an event study, as all professors who are denied tenure will not be part of the sample.
45 Results are available upon request.
46 See Online Appendix C for the tables that correspond to these figures.
47 A segment of the faculty may also have increasing evaluations, meaning another segment's evaluations decline by larger amounts than what was found in the baseline.
48 This idea is explored in Figlio, Schapiro, and Soter (Citation2015).
49 The average GPA across all courses in the sample is 3.29.
50 Even the latter can be forced out by aggressive boards.
51 These demands are also heterogeneous across departments.
52 The issue with the latter, however, is the small number of tenure applications.