Ronald G. Ehrenberg and Joyce B. Main*
*Ehrenberg is the Irving. M. Ives Professor of Industrial and Labor Relations and Economics and a Stephen H. Weiss Presidential Fellow at Cornell University, and Director of the Cornell Higher Education Research Institute (CHERI). He also currently serves as an elected faculty trustee of Cornell University. Main is a PhD student in education at Cornell and a graduate research assistant at CHERI.
Under the auspices of the Association of Governing Boards of Universities and Colleges (AGB), the Cornell Survey Research Institute (SRI) conducted a survey for the Cornell Higher Education Research Institute (CHERI) of 4-year colleges and universities to obtain annual information for the 1981 to 2007 period on the gender of voting members and of chairs of boards of trustees. The goals of the survey were to document trends in the gender of board members and leaders, to learn whether the gender composition of board members and leaders influences the gender composition of chancellors/presidents and chief academic officers, and whether the gender composition of board leaders and members, and key academic administrators influence the rate at which academic institutions are diversifying their faculty along gender lines. This brief paper reports the results of the survey; later papers will describe our other findings.
Under strict confidentiality conditions, the AGB provided CHERI with contact information for a sample of 745 4-year colleges and universities. Survey questionnaires were mailed on March 21, 2008 to the institutions, with a cover letter explaining the project and the institutions were invited to respond by either fax or mail. An initial round of follow up phone calls took place in the second half of April, a second round took place in the first half of June, and a final round took place in the second half of July 2008. When data collection was completed on August 29, 2008, 509 completed responses had been received, 221 institutions either refused to participate or did not respond to the survey, and 15 institutions proved to be either duplicates (more than one address for the same institution) or branch campuses of an institutions whose board was the same as the board of the main campus. Hence, the response rate to the survey was 69.7% (509/730).
To say that the response rate was 69.7% is not to say that all institutions reported data for all the years between 1981 and 2007. Institutions were more likely to report data on the gender composition of their boards and the gender of the chair of their boards for more recent years. The dashed line in Figure_1.pdf (PDF, 9 KB) shows the number of institutions that reported data on the gender of their board chair in each year; the number was 502 in 2002 and it declined to 376 for 1981.The solid line presents data for the number of institutions that reported data on the share of their trustees that were female in each year; this number declined from 503 in 2007 to 341 in 2007.
This decline in the response rate to the survey as one goes further back in time may lead to be concerns that changes over time in the share of boards with female chairs and in the average share of board members who are female may be distorted by the changing composition of the institutions in the sample who are reporting data. However, it turns out that this is not the case. The trend we find using data for all of the institutions that report data in a given year are very similar to the trends we find when we restrict the analyses to the subsample of institutions that report data in every year.
Figure_2.pdf (PDF, 20 KB) illustrates this for the average shares of members of boards of trustees that are female. The solid line now traces the share in each year for all institutions that reported the share in that year, while the dotted line traces the share in each year for the subsample of institutions that reported data to us for every year between 1981 and 2007. Both tell the same story: The average share of board members that is female rose steadily from about 0.20 in 1981 to about 0.31 in 2007; steady but slow progress. Figure_3.pdf (PDF, 20 KB) similarly illustrates this for the shares of board chairs that are female. This share rose from slightly under 0.1 in 1981 to about 0.18 in 2007; again, steady but slow progress.
Research on corporate boards has found that while having one or two females on a board can make a difference in board operations, it usually takes three or more women on a board to achieve the “critical mass” needed to cause fundamental change on the board and to enhance the contributions that female board members make (see Alison M. Konrad, Vicki Kramer, and Sumru Ekrut, “Critical Mass: The Impact of Three or More Women on Corporate Boards”, Organizational Dynamics 37 (April 2008): 145-164). So in Figure_4.pdf (PDF, 10 KB) we plot each year for our entire sample of institutions the shares of institutions with at least three, four and five female board members. The share of boards in our sample with at least three female members rose from about 0.6 in 1981 to about 0.9 in 2007. Similarly, the share with at least five female members rose from about 0.4 in 1981 to about 0.6 in 2007. Our subsequent research will test whether these shares of female trustees prove to be important in explaining the rate at which academia diversifies its faculty across gender lines.
Once we disaggregate our entire sample of institutions by form of control (private or public) and institution type (baccalaureate, masters, doctoral), the sample sizes become smaller and figures similar to those presented above are much less “smooth”. Nonetheless, the following additional conclusions emerged from our survey:
- The average shares of board members who are female does not vary much between public and private institutions and has increased over time in both sectors (Figure_5.pdf (PDF, 18 KB)).
- The proportion of boards with female chairs has increased over time in both sectors and in most years was higher in public than it was in private institutions (Figure_6.pdf (PDF, 18 KB)).
- The average shares of board members who are female has increased over time for baccalaureate, masters, and doctoral institutions, but is lower in each year at the doctoral institutions than it is at the other two categories of institutions (Figure_7.pdf (PDF, 9 KB)).
- The proportion of boards with female chairs has trended upwards over time for all three institutional types, but it is lower at the doctoral institutions than it is at the baccalaureate institutions Figure_8.pdf (PDF, 19 KB).