North American universities have always competed for academic status. The selection of tenure-stream faculty is a major component in that competition, especially in Canadian universities where most tenure decisions are positive. In the early nineteen seventies the introduction of public advertisement of all positions brought about a significant change in the hiring process. This requirement at least eliminated the cruder versions of the "old boys' network" in determining appointments. In principle, this reform increased the possibility that academic merit (i.e., what a candidate knew) rather than favoritism (i.e., whom a candidate knew) would be the basis of selection for academic positions.
By the early nineteen eighties, in addition to the goal of merit in appointments, the goal of "Aaffirmative action" (in the USA) or Aemployment equity" (in Canada) was adopted to a greater or lesser extent by North American universities. Since then, many university faculty and almost all administrators have argued that there is no conflict between merit and equity. (Recently, for instance, the University of Toronto has adopted a slogan of "Excellence through Diversity".
Logically, however, the claim that there is no conflict between merit and equity is undermined by those many high-level university administrators who claim that they seek a proper "balance" between merit and equity: balance is required only between countervailing units. The claim is also undermined empirically when, in studies like that described here, the patterns of results obtained by emphasizing merit are different from those obtained by emphasizing equity.
In the ways that they deal with such problems as finding the proper "balance" between merit and equity, universities can be viewed as organisms that, guided by their academic administrators, differ both with respect to internal motivational and external environmental conditions.
For instance, with regard to motivation, York University has a stronger commitment to equity than the University of Toronto (witness: it is the only Canadian university to have won a federal Equity Award). It is reasonable to expect, therefore, that York's tenure-stream hiring policies would place more emphasis on equity, and its tenure-stream advertisements might contain phrases like Aespecially welcome applications from women "rather than Awelcome applications from both women and men."
An example of different adaptations to differing environmental conditions (strength of the economy) is the contrast between tenure-stream advertisements in Western and Eastern Canada. A phrase like "outstanding record of research publications" is a sign of a stronger merit requirement than Ainterest in developing a research program." The former sort of phrase is more evident in advertisements from universities in the West than those in the less affluent East of Canada. This West/East difference in merit emphases appears to reflect reasonable adaptations to differing environmental conditions.
Studying Institutional Adaptations Through Job Advertisements
It is possible to systematically study the differential adaptations of universities (and even of such sectors of universities as the hard sciences versus the huma-nities). The method is to examine the phraseology of tenure-stream advertisements in order to indirectly assess differences in degrees of emphasis on merit and equity as factors in hiring.
To measure these emphases, the study (conducted by the first author with assistants) employed the method of Ajudgmental content analysis. 1 Raters use 7-point scales to assess the degree of merit and of equity in tenure-stream advertisements for arts and science positions (taken from University Affairs) from which institutional and departmental identification has been removed. Previous SAFS Newsletter articles have reported on some aspects of this Canada-wide study [SAFS Newsletter, 27 (January 2001, 6-7); SAFS Newsletter, 30 (January 2002, 1-4)].
The present study is a closer analysis of the situation in Ontario, primarily contrasting York University and the University of Toronto. As in the Canada wide-study, the period 1992-1994 was compared with 1996-1998 because, in 1995, a "political earthquake" occurred in Ontario, with the left-wing New Democratic Party (NDP) government of Bob Rae being replaced by the right-wing conservative government of Mike Harris (whose "common sense revolution" included abolition of employment equity laws, at least in the private sector, laws which the NDP had strengthened during its 1990-1995 tenure in government).
On the face of it, University of Toronto and York appeared to react differently to the 1995 left-to-right shift in the Ontario government. For example, whereas, at least on the administrative level, University of Toronto maintained cordial relations with the Harris government, Harris's first minister of education was pelted with tomatoes in a rowdy protest when he visited York, York=s administration did not publicly apologize. Also, these two large Ontario universities differ in the strength of their commitment to employment equity, and the position of their average faculty on the left-right political continuum. So an expected result for this study was that there would be a differential effect of the Rae-to-Harris political change on the two universities' emphasis on equity in their tenure-stream advertisements.
For comparative purposes, the study looked at a third group of university advertisements for the "rest" of Ontario2 so the design of our factorial study had Location as a three-level factor (York, Toronto, and Rest). The second factor was Time (1992-1994 vs. 1996-1998), while the third factor was Discipline Hardness (hard sciences, social sciences, and humanities). Rated Merit and Rated Equity were the dependent variables. Finally, there was a Rater factor, the four raters (Yaniv Morgenstern, Joanna Mostowka, Carol Okamoto, and Joanna Renata) providing the four levels of this factor.
Reliability And Absence Of Bias From Rater Differences
The inter-rater correlations for merit ranged from 0.55 to 0.76; for equity, the correlations were higher, ranging from 0.83 to 0.93. There were statistically significant3 differences among raters in their mean merit (ranging from 4.30 to 5.02) but not equity ratings. However, most importantly, no interactions emerged between the four-level Rater factor and the other three factors of Location, Discipline Hardness, and Time. That is to say, we can be confident that differences among raters did not confound the interpretation of what the raters were rating with respect to the three factors of Location, Time, and Discipline Hardness.
A potential source of bias was that, especially in the case of the most experienced rater who had worked on prior studies, raters may have been able to identify York advertisements even though, as in previous studies, the university and departmental identification had been removed. To the extent that this identification was possible, raters may have given higher equity ratings to York advertisements not on the grounds of actual content, but on the grounds of expecting York advertisements to be more equity-oriented.
To investigate this source of potential bias, raters were asked to guess the origin of each advertisement (York, Toronto, or Rest), and to express the degree of confidence in their identification on a 5-point scale. Mean identification accuracy was not above chance level, and the two raters whose confidence in their ability to identify advertisements was highest were also the two raters who performed at significantly below chance level in correctly identifying York advertisements. Accordingly, it is unlikely that this source of bias confounded the results.
Results: Expected And Unexpected Differential Adaptations
Another distinction between adaptations is to classify them into those readily explicable or expected in terms of institutional policies and/or the environment, and those that are not so readily understandable in those terms.
An example of an expected difference that has already been anecdotally mentioned is the York/Toronto difference in their emphasis on equity. The results of the present study's judgmental content analysis were in line with this expectation. York's mean equity rating (3.75) was significantly higher than Toronto's (1.80).
Taken together, the remaining Ontario Universities (Rest) had an average equity score that equaled that of York (i.e., 3.75), suggesting that, on the average, those other universities followed York's rather than Toronto's degree of emphasis on equity. It bears emphasis that the Rest category contained universities that probably vary considerably in terms of their equity emphasis. A more recent example of an extreme equity emphasis was Wilfrid Laurier's psychology department which, in 1999, advertised a developmental psychology position that was restricted to women only (see SAFS Newsletter, 23, November 1999, 1-3).
However, contrary to what was expected in terms of differences in policy regarding equity, the Rae-to-Harris shift had no effect on mean equity ratings, either in terms of the main effect of Time, or the specifically predicted Location x Time interaction. So, despite the marked York/Toronto differences in their commitment to equity position on the left-right political continuum, and reactions to the Harris government, the relative equity emphasis of the tenure-stream ads of these institutions remained unaffected by the provincial left-to-right political shift.
Still, when the merit results were analyzed, a significant Location x Timeinteraction did emerge as depicted in Figure 1.
As inspection of Figure 1 suggests, the source of this interaction was that, whereas over time (from the Rae to the Harris period) York and the Rest increased their merit emphasis, Toronto uniquely decreased its merit emphasis. So, contrary to expectations, the Rae-to-Harris shift differentially affected Toronto and York (together with the Rest) not in their equity but in their merit emphases.
Figure 1 also shows that the merit requirements for York and Toronto were approximately equal, and markedly higher than those for the Rest. This seems to be an expected result that stems from the greater funding available to both York and Toronto compared to the remaining Ontario universities (which include a number of quite small institutions where externally funded research is not undertaken on the scale of either York or Toronto).
The merit results also yielded a significant interaction between Discipline Hardness and Time, and hence another differential Rae-to-Harris effect not on equity but on merit. This interaction is depicted in Figure 2. It is clear that there is a marked and unique increase of the merit emphasis in the hard sciences from the Rae period (when there was essentially no difference among the three discipline-hardness categories) to the Harris period (where the marked difference is between the hard sciences and the other two categories). Moreover, as there was no three-way interaction between Location, Discipline Hardness, and Time, this unique increase in the emphasis on merit in the hard sciences appears to have occurred across all Ontario Universities.
The increase in merit hardness from the Rae period in the hard sciences alone is not readily explicable in terms of explicitly formulated policies. It may, however, reflect an "adaptation" to equity pressures by departments in the hard sciences. Perhaps one result of the Rae-to-Harris shift was to produce more explicit attention to the equity vs. merit issue in tenure-stream hiring, and faculty in the hard sciences, who share more agreement about criteria for merit, adjusted to the pressure from administrations and 'equity officers' for 'inclusive' language in advertisements by increasing the merit requirements for positions. An indirect way of testing this interpretation would be to survey faculty who were members of hiring committees in the 1992-1994 and 1996-1998 periods concerning their attitudes towards the merit/equity "balance" and departmental policies on tenure-stream appointments.
The third and final significant interaction that emerged from the merit results was that between Location and Discipline Hardness, an interaction that is depicted in Figure 3. This interaction arises from a contrast between York on the one hand, and Toronto and the Rest on the other hand. Whereas at York, social science had the highest merit requirements, at Toronto and in the Rest, social science had the lowest merit emphasis.
Some "local", admittedly anecdotal, knowledge may account for this apparently unique social-science emphasis on merit at York during the nineties. During the seventies and early eighties, a period of considerable expansion in the social sciences, there was a perception of 'softness' in social science appointments. In the mid- to late-1980s the Faculty of Arts administration revised and standardized procedures for tenure and promotion committees. These factors may have subsequently strengthened the emphasis on merit in advertisements for positions in the social sciences.
To more systematically test this account, it may be possible to conduct interviews with those in positions of administrative authority at the time. Another approach would be to apply the present judgmental content analysis to Ontario university advertisements during two earlier sets of three-year periods (e.g., 1978-1980 and 1982-1984) to see whether the sort of interaction depicted in the bottom panel of the figure emerges.
The method of judgmental content analysis of tenure-stream advertisements provides a systematic, empirical way of examining differences in the way in which merit and equity in tenure-stream hiring are emphasized. Some of these differences are either due to obvious differences in institutional policies, or to marked differences in the environment in which universities operate.
However, there are other differences that came to light that are not predictable on the basis of 'armchair' study. It is interesting that in this research these less expected differences emerged not in the equity but in the merit emphases, and there were unique aspects shown by the University of Toronto in the later, 1996-1998 period (Figure 1), by the hard sciences in the later, 1996-1998 period (Figure 2), and by York University in the social sciences (Figure 3). Finally, it is possible to generate hypotheses that may account for these unexpected effects, hypotheses that are themselves open to further empirical testing. These investigations, then, appear to be relevant to those Canadian academic administrators who are interested in finding the right "balance" between merit and equity, no matter where they are positioned on the merit-equity continuum.
- For details, see Technical Report #1 to Donner Canadian Foundation on JJF’s web site: http://www.psych.utoronto.ca/~furedy/judgm6.htm.
- The universities in this group included Algoma, Brock, Guelph, Lakehead, McMaster, Ottawa, Queen's, Redeemer College, Trent, Waterloo, Western Ontario, Windsor, and Wilfrid Laurier.
- Defined at a level of less than 0.05, i.e., a less than 5% chance that the observed sample difference is drawn from a population with no difference.