Presentation is loading. Please wait.

Presentation is loading. Please wait.

Can Mentoring Help Female Assistant Professors

Similar presentations


Presentation on theme: "Can Mentoring Help Female Assistant Professors"— Presentation transcript:

1 Can Mentoring Help Female Assistant Professors
Can Mentoring Help Female Assistant Professors? An Evaluation by Randomized Trial Janet Currie, Princeton University Donna K. Ginther, University of Kansas Francine Blau, Cornell University Rachel Croson, Michigan State University This research was funded by the NSF (SBE ) and (SES ) and the American Economic Association. We are grateful to the many women who volunteered to serve as mentors, workshop participants, and organizers.

2 Overview Problem: Dearth of women at the highest ranks in academic economics Potential solution: Does the CEMENT mentoring workshop enhance untenured academic woman economists’ career outcomes? Our results indicate: Mentor can work, especially for women with elite training.

3 Committee on the Status of Women in the Economics Profession
CSWEP is a standing committee of the American Economic Association charged with promoting women economists in academia, government agencies and elsewhere. CSWEP conducts an annual survey of PhD granting and liberal arts institutions Organizes sessions at AEA & regional meetings Runs the CEMENT mentoring program

4 CSWEP data demonstrate the leaky pipeline:
32.9 percent of new PhDs in Economics were female, consistent with numbers since the 1990s. In PhD granting departments, 28.8 percent of assistant professors were female. 23 percent of tenured associate professors were female 13.9 percent of tenured full professors were female.

5 Previous literature suggests women are less likely to get tenure than men (Ginther and Kahn 2004,2009): Women may lack networks (e.g. McDowell, Singell and Slater (2006) find they are less likely to coauthor) Women may lack mentors (Blau, Ferber and Winkler (2006)) Women are disadvantaged in achieving tenure when there are gender-neutral stop clock policies (Antecol, Bedard & Stearns 2016) Women economists’ papers take about six months longer in peer review (Hengel 2017) Women are not given equal credit for coauthored work (Sarsons 2017) Wu (2017) documented hostility towards women on the Economics Job Market Rumors website.

6 History of CeMENT 1998--trial economics mentoring workshop funded by the NSF: Creating Career Opportunities for Female Economists. 2004—American Economic Association applied for a grant to the National Science Foundation to fund bi-annual mentoring workshops and proposed to evaluate them using a randomized controlled trial. 2007—AEA committed to funding additional workshops above & beyond NSF funding. 2010—”Can Mentoring Help Female Assistant Professors? Interim Results from a Randomized Trial” showed that the treatment group had more publications, more publications in top journals & more federal research grants. 2014—American Economic Association committed to funding workshops on an annual basis.

7 The CeMENT Intervention
Each two day workshop was held after the AEA meetings (in 2004, 2006, 2008, 2010, 2012, 2014, 2015, 2016, 2017, 2018). They were aimed at research faculty. We will mainly focus on the first five cohorts ( ) >80 applications for each wave. Applicants sorted into groups by field, and then randomized into treatment. There were 4-5 participants with 1-2 mentors in each group.  Fields included labor, health, macro, experimental, development/international, theory, econometrics.

8 The CeMENT Intervention
Each participant circulated a research paper or related work (e.g. grant proposal) before the workshop.  These were discussed by the small groups (~ 1 hr. for each participant). Each person was required to read all of the submitted papers and provide comments. Plenary sessions were also held with the senior mentors covering topics such as research and publishing, getting grants, networking, teaching, the tenure process and work-life balance.  

9 Participant Reaction was positive:
On a Seven Point Scale (where 7= extremely helpful; 1 = not helpful at all)

10 Participant Reaction was positive:
“It was an incredible experience and I found it extremely helpful.” “I learned a lot from the workshop and I wish I would have attended 2 years ago.” “I had a really fantastic experience at the CeMENT workshop. So much information and networking packed into the 2 days!” “Although I have been teaching…for more than five years, I still found many of the discussions and much of the advice extremely helpful.”

11 Some comments suggested that effects lasted past the workshop:
“My experience was very positive. Our group kept in touch regularly after the initial meeting. This led to organizing and appearing in conference sessions together, reading each other's work, sharing helpful things like grant applications, and inviting each other to speak at our respective institutions. When I went up for tenure, it was very helpful to know two successful senior people, whom I could suggest to my department as letter writers.” “The workshop has had a huge impact on my career. I was at an Economics Department with very few women, and I was the only experimental economist At the ASSA meetings two years later, the whole team sat down for two hours to go through my NSF proposal paragraph by paragraph. It greatly improved the quality of the proposal. . .This workshop was the best thing that happened to me since my marriage.”

12 Follow Up Data We collected CVs from treatments and controls using web searches and s. We coded the current position, publications, and NSF and NIH grants. If current vitas were not available, we updated info from publicly available sources. Applicants were also surveyed in earlier cohorts, but attrition among the controls was a problem.

13 Data and Control Variables
Coded the rank of the PhD Program and first job using program rankings from Kalaitzidakis, Mamuneas & Stegnos (JEEA 2003) Publications ranked by journal quality. top Ranked Journals: American Economic Review, Journal of Political Economy, Quarterly Journal of Economics, Econometrica Control for cohort (2004, 2006, 2008, 2010, 2012) Years since PhD > 6 Sample Size = 303 observations 189 Treated 114 Controls

14 Outcome Data Outcomes for cohorts from 2004 – 2012:
Probability of achieving tenured rank (associate or full professor) Last observed working in academia Outcomes for cohorts from : Number of publications by years since treatment Probability of top publication Probability of federal grants (NSF or NIH)

15 Balancing Tests

16 Probability of Tenured Rank & Academic Job

17 Probability of Tenured Rank by Institution Rank

18 Estimation Approach Model 1: Treatment Dummy
Model 2: Treatment & Cohort Dummies, Years Since PhD Model 3: Treatment & Cohort Dummies, Years Since PhD, Rank of PhD Institution, Rank of First Job Outcomes: Tenured Rank Academic Job Publications Grants

19 Outcome=Probability of Tenured Rank
Model 1 Model 2 Model 3 Effect of Treatment 0.036 -0.023 -0.017 [0.059] [0.054] Cohort&Years Since PhD X PhD Rank First Job Rank # Observations 303 R-squared 0.001 0.232 0.241

20 Outcome=Probability of Tenure at Top 40 School
Model 1 Model 2 Model 3 Effect of Treatment 0.131*** 0.118*** 0.107*** [0.039] [0.035] Cohort&Years Since PhD X PhD Rank First Job Rank # Observations 303 R-squared 0.037 0.125 0.283

21 Outcome=Probability of Tenure at Rank 41+ School
Model 1 Model 2 Model 3 Effect of Treatment -0.100* -0.145** -0.124** [0.059] [0.058] [0.055] Cohort&Years Since PhD X PhD Rank&1st Job Rank # Observations 303 R-squared 0.009 0.120 0.245

22 Effect of Treatment on Probability of Tenured, Robustness to Different Rank Cutoffs
Outcome Model 1 Model 2 Model 3 Rank<=30 0.102*** 0.089** 0.076** [0.035] [0.032] Rank<=50 0.134*** 0.115*** 0.103*** [0.041] [0.038] Rank<=75 0.122*** 0.095** 0.084** [0.045] [0.044] [0.043] Rank 101+ -0.084 -0.112* -0.092* [0.057] [0.058] [0.055]

23 Treatment effect on remaining in academic job
Outcome Model 1 Model 2 Model 3 1. Any Academic Job 0.076 0.061 0.067 [0.051] [0.050] 2. Academic Job Rank<=40 0.113 0.144 0.119 [0.102] [0.113] [0.111] 3. Academic Job Rank 41+ 0.040 0.003 -0.003 [0.062] [0.061] Cohort & Years Since PhD X PhD Rank&1st Job Rank

24 Treatment effect on remaining in tenured or tenure-track job
Outcome Model 1 Model 2 Model 3 1. Any tenure/tenure track 0.095* 0.072 0.078 [0.053] [0.052] 2. Tenure/tenure track rank<=40 0.129*** 0.127*** 0.109*** [0.041] [0.037] 3. Tenure/tenure track rank>40 -0.034 -0.055 -0.031 [0.059] [0.055] Cohort & Years Since PhD X PhD Rank&1st Job Rank

25 Probability of Top Publications and Federal Grants
Outcome Publications FederalGrants Any Top Publications 0.101** 0.062* 0.051 [0.047] [0.033] Cohort & Years Since PhD X PhD Rank & 1st Job Rank # Observations 303 R-squared 0.040 0.246 0.145 0.221

26 Probability of a Top Publication, by year, all cohorts
Effect of Treatment 0.060*** 0.079** 0.086* 0.091* 0.124* 0.126 0.181 [0.022] [0.036] [0.045] [0.055] [0.064] [0.083] [0.113] Constant 0.045 0.141*** 0.218*** 0.269*** 0.262*** 0.275*** 0.241*** [0.030] [0.053] [0.058] [0.063] [0.073] [0.088] Cohort FE X Observations 506 389 329 245 195 130 74 R-squared 0.022 0.033 0.052 0.034 Cohorts 1-8,

27 Summary Women are more likely to be tenured in top 40 ranked departments and less likely to be tenured in departments ranked 101+. But there is no effect of treatment on academic employment. Publication results show evidence of cumulative advantage. Questions: Why is there a negative effect on women from lower ranked institutions? Does timing matter? Were there differential effects by field?

28 Questions?


Download ppt "Can Mentoring Help Female Assistant Professors"

Similar presentations


Ads by Google