What We Can’t Say about Contingent Faculty

You can’t draw any conclusions about the conditions of contingent labor until you have the data on wages.
By Teresa Tam and Daniel Jacoby

The effects of reliance on part-time faculty in higher education have been much discussed of late. Most observers now agree that the increasing reliance on contingent academic labor has worrisome consequences for both students and faculty. We recently attempted to provide needed analysis of what drives the current reliance on part-time faculty and how policy makers can best respond.

Our research illustrated the inadequacy of the existing data on part-time faculty wages. We were able to find most of the data we believed were essential to an evaluation of the market for parttime labor, including rates of faculty unionization, an indicator of the potential supply of graduate students, and the percentages of students who are part time at the institutions we examined. Most of the data we used were obtained from the Integrated Postsecondary Education Data System (IPEDS), the core higher education data collection program for the Department of Education’s National Center for Education Statistics (NCES). But a reliable measure of part-time faculty wages was missing. Such data are not currently collected nationwide at the institutional level. Without reliable data on wages, the validity of key arguments in the debate about part-time faculty cannot be assessed.

The stakes of the debate are rising as numerous states consider legislation prompted by the Faculty and College Excellence campaign of the American Federation of Teachers (AFT). The changes proposed by this legislation, which are designed to achieve pay equity for contingent faculty and increase the number of undergraduate classes taught by full-time tenure-track faculty, hinge upon understandings of the supply and demand for parttime faculty that cannot be verified without better wage data. We cannot know, for example, whether raising the wages of parttimers leads to a reduction of part-time faculty as we would expect; some observers argue that this reduction does not occur. Competing claims exist side by side. Some argue that unionization improves wages for part-time faculty, while others assert that it protects full-time faculty wages, creating fewer tenure-track positions and more contingent academic labor. Without a trustworthy measure of part-time wages, we cannot determine whether college administrators’ desire for flexibility is so important that they will continue to employ part-time faculty even when part-timers’ wages approach those of their full-time colleagues. According to Craig Smith, deputy director for higher education at the AFT, “Politically, you have to have data to go with your story. Having the data may not move your agenda, but not having [them] will definitely hinder it.”

AAUP and IPEDS Data

The AAUP’s annual survey of faculty compensation, the Annual Report on the Economic Status of the Profession, is the most relied-upon source of faculty earnings data. Though the AAUP provides data at the individual institution level, it reports only on full-time faculty. Given the increasing reliance on contingent academic labor—contingent faculty are estimated to comprise two-thirds of community college faculty and half of appointments in higher education overall—the omission of data on part-time wages constitutes an oversight so large that it threatens the usefulness of this essential wage collection effort.

According to John Curtis, the AAUP’s director of research and public policy, a number of obstacles have prevented the AAUP from collecting such data. First is the likelihood of inconsistent data. Institutions do not keep information on part-time faculty in a centralized location as they do data on full-time faculty. Curtis notes that the lack of a federal mandate on part-time wage reporting is a “big factor” in inconsistent record keeping by institutions. Second, the AAUP’s data collection survey could not be easily adapted to include part-time faculty wages. Currently, AAUP staff members calculate average salaries for the survey based on aggregate measures of full-time faculty salaries. Collecting part-time faculty wages would be even more complicated because the number of part-time faculty, the number of credit hours, and rates of pay by  hour or course would be needed to complete the calculations. Third, the AAUP’s research operation is not sufficiently staffed to enable it to gather this information. Curtis says that both members and leaders of the AAUP would like to have data on part-time faculty wages. He has proposed to the AAUP Executive Committee that this be given a high priority among new research projects.

IPEDS, the preeminent source of national higher education data, was critical to our study. Because most hiring is conducted at the level of the individual institution, we used the institution as the unit of analysis. IPEDS includes data from all primary providers of postsecondary education in the country in areas such as enrollments, program completions, graduation rates, faculty, staff, finances, institutional prices, and student financial aid. It also collects information on the number of part-time faculty by race or ethnicity and gender. The NCES previously collected data on parttime earnings through its National Study of Postsecondary Faculty, but neither that study nor IPEDS provides reliable earnings data by institution. Additionally, the NCES has stopped conducting the faculty study. A technical review panel created to analyze the IPEDS human resources survey—which contains information on faculty salaries— was scheduled instead, but this panel has been postponed as a result of changes brought about by the recent Higher Education Opportunity Act.

In its last major assessment of part-time earnings in 2004, the NCES relied upon the National Study of Postsecondary Faculty. The NCES report concluded that part-time faculty overall averaged just over $11,000 in annual earnings from basic salary, with faculty at public community colleges earning $9,200. Including additional income from institutions increases these earnings to $11,900 and $10,000, respectively, at all institutions and at two-year colleges. Even if accurate, these data mask considerable variation among individuals and across colleges. In its 2005–06 Annual Report on the Economic Status of the Profession, the AAUP indicated that the majority of part-time community college faculty received between $1,400 and $2,250 a course during the 2003–04 academic year.

Other Data Sources

The lack of adequate data on part-time faculty wages from IPEDS prompted us to search for other sources of usable data for our research on two-year public institutions. Table 1 summarizes what we found.

These scattered efforts at data collection in various states are insufficient for the needed analysis of supply and demand conditions for part-time faculty. Some studies base their results on assumptions about what part-time faculty would earn if they worked full time. Other studies indicate what part-time faculty earn when they are paid by the credit or course. Actual earnings depend on the number of courses taught as well as any additional payments that may occur. Such variations are indicative of the challenges facing data collectors. Further challenges include accounting for differences between institutions on the semester system and those on the quarter system as well as for differences in how basic pay is provided (by hour of class contact, credit hour, or a percentage of full-time work).

Wage Estimation

Given the inadequacy of existing data, we were forced to produce our own estimates of part-time faculty earnings to examine supply and demand conditions. IPEDS provides data on full-time faculty salaries and total instructional salaries, and we concluded that the difference between the two amounts might approximate the total amount paid to part-time faculty. Once we determined the difference between these amounts, we divided it by the total number of part-time faculty to estimate the average annual pay for part-time faculty members at each institution. Though a large percentage of our institution-wide estimates seemed credible, too many were not. We estimated part-time earnings at a sizable number of institutions to be well over $30,000 a year, a figure that might reflect the earnings of few individuals but is highly unlikely to reflect typical earnings. Earnings estimates were so high as to exceed $250,000 for a few institutions, while other estimates were negative.

The problems we encountered reinforce the argument for improving data collection on a number of fronts. First, we were surprised that many institutions did not report data in one or more of the categories needed for a part-time wage calculation; roughly 35 percent of the two-year public colleges in the total sample were eliminated from our study as a result. Second, without a standardized definition of “part-time faculty,” we could not be sure that all institutions were reporting numbers of part-time faculty in a consistent manner. Third, discussions with select institutions led us to believe that “total instructional salaries” may have contained data other than full-time and parttime faculty wages, such as full-time overloads. Perhaps most worrisome is the fact that the data do not permit separation of summer wages for fulltime faculty from those of part-time faculty. These problems are indicative of the fact that the new majority faculty—those employed on contingent appointments—remain an afterthought in data collection.

Perverse results in our study led us to believe that even the NCES estimates (cited earlier) of the percentage of faculty working part time may be understated. We initially tried to study demand by limiting ourselves to those institutions for which our estimates of individual earnings were less than or equal to $25,000. Because few studies report part-time faculty earning more than 60 percent of what full-time faculty earn when earnings are measured by the number of courses taught, and because salaries of full-time faculty at community colleges are roughly $50,000 a year, it is unlikely that part-time faculty, who typically work less than half time, would earn annual wages in excess of $25,000. Still, the possibility of summer pay and wrongly included benefits made an average annual earnings figure of $25,000 plausible.

We conducted telephone interviews with select institutions in our sample to investigate possible sources of error. The interviews revealed that some institutions did not fully account for all of the part-time faculty members they employed. For example, some included only part-time faculty who were considered part of the tenure-track or permanent faculty.

When our calculations produced perverse results, we realized that if institutions had underestimated the numbers of part-time faculty teaching on their campuses, our wage estimates would be too high. We found that the more systematic the undercount, the more statistically significant the estimate of the relationship between high part-time wages and low part-time employment. In short, we were forced to admit not only that some of our wage estimates were likely to be unreliable, but also that the widely used count of part-time faculty based on NCES and IPEDS data was likely to be significantly understated.

Challenges

The lack of comprehensive part-time faculty wage information at the institutional level begs a question: what is preventing these data from being collected? No nationally recognized definition of “part-time faculty” exists, which makes nationwide collection of part-time wages difficult. New Mexico’s Report on Part-Time Faculty Compensation and Salary Survey states, “There [is] significant variance in the meanings and use of the terms ‘part-time’ and ‘full-time’ faculty between institutions.” Craig Smith of the AFT has said that “without a common set of definitions for contingent faculty it will be difficult to gain the level of transparency we should have on how contingent faculty are being employed and treated.”

Institutions will oppose any additional reporting burden. Parttime faculty are generally paid by the credit hour, which makes their earnings more difficult to document than those of their full-time colleagues. Therefore, requests that institutions add such data are met with resistance. Numerous groups have asked the Department of Education to require part-time salary reporting in the IPEDS survey. Smith says that the department will “not move on the issue, claiming the IPEDS survey is ‘too long already.’” Compounding this problem is the lack of political pressure. According to Elise Miller, the IPEDS program director, there “hasn’t been an interest by Congress in collecting part-time faculty wages. [Members] are usually focused on student-aid issues instead.” Miller says that the IPEDS human resources data collection will be examined within a year, at which point items may be added or eliminated based on panel review. Even if changes were made, however, it would take approximately three years to implement them.

The Way Forward

A standardized definition of part-time faculty must be agreed upon at the national level. Part-time wage data will not be useful unless institutions are consistent in how they report data. A federal mandate may be a way to create this standardization, though enforcing it to a point where the data obtained from institutions are usable will be challenging.

Government intervention may be a long time in coming. Still, part-time faculty pay represents a vital area for data collection. Omitting wage data on part-time faculty—no matter how many difficulties the collection of those data pose—is akin to eliminating food and oil prices from inflation statistics. We may want to treat them separately in particular analyses, but to continue to omit them altogether will leave us driving blind.

Teresa Tam is an analyst for the Government Accountability Office. Daniel Jacoby formerly held the Harry Bridges Endowed Chair in Labor Studies at the University of Washington and has been the acting assistant vice chancellor for institutional research at the University of Washington Bothell Campus.


Comments

As typically occurs with issues of contingent faculty, the topics raised in this article radiate contradiction.  I’m torn between gratitude that someone is pointing out these enormous gaps in recordkeeping and disgust that yet again the call for better data trumps the call to action.

At the outset, my suspicions were raised by the noted obstacles.  Aren’t “rates of faculty unionization” specifically gathered by the National Center for the Study of Collective Bargaining in Higher Education?  Aren’t the numbers of graduate students part-time students available within institutional data?  The labeling of “omission of data on part-time wages” as an “oversight” seemed unnecessarily generous.  How many excuses can we make for people who are supposedly expert?  Whether the “data mask considerable variation” seems besides the point already – the salaries are low, low, low and we’ve had “scattered efforts” at data collection for HALF A CENTURY.

When the authors point out that “Such variations are indicative of the challenges facing data collectors,” I want to protest that consistently under-paid part-time faculty as the rule is indicative of the challenges facing higher education.  Yes, it would be great to have a standardized definition of part-time faculty but do we have such a highly standardized of full-time faculty?  Do we need to worry about full-time faculty overloads fouling up the numbers as much as we need to worry about administrators teaching overloads that both foul up the numbers and undermine faculty work?  Never mind that “the new majority faculty” . . . “remain an afterthought in data collection” – even an afterthought is visible to some extent.  Legislators, the public, parents, students, are all seldom aware of this majority.

So, yes, institutions oppose comprehensive reporting, and yes, there is a lack of political pressure, and, yes, we need a federal mandate.  It’s very simple especially with all this talk of transparency and accountability:  all employees must be reported, no hidden employees, no employees left behind.

Karen T.