Please note that the supporting documentation for this report is contained in linked tables that will appear in resizable pop-up windows so that they may be consulted alongside this report; these pop-up windows should be closed as you finish consulting each table. The full report is available as an Adobe Acrobat file that may be downloaded and printed (printable version of report); all the placement tables and figures for 2001-2002 are available in separate Acrobat files (printable version of tables and printable version of figures). It is best to use the “Print as Image” command when printing Acrobat files.
The CSWMG is mandated by the APA to monitor actively both the fairness of the employment process and hiring institutions’ compliance with professional ethical standards. In this capacity, the CSWMG produces an annual report on the previous year’s job placements. Because of the difficulty in obtaining and correlating data, these reports now lag nearly a full year behind (that is, the report on the 2002-03 hiring year is produced in the spring of 2004). The CSWMG would like to thank Barbara McManus and the APA office for all their work in collecting, collating, and producing this data. We would also like to thank all the candidates who took time to complete the fall survey. To put this discussion in broader context, readers are invited to consult previous CSWMG placement reports and the concurrent report of the APA-AIA Joint Committee on Placement on this web site.
I. Data sources
Three types of information have been collected and collated in completing the following analysis: an initial candidate form filled out when candidates first enroll in the APA Placement Service, a “Professional Matters Division Survey” sent out to candidates the following fall, and statistical placement information provided by the APA office. The database produced from all these types of information is correlated by use of an ID number rather than names so that candidate anonymity is preserved.
The data from the Placement Service provides the most reliable and complete information on the demographics of job seekers, since it has information on 100% of the candidates and institutions who used the Service. It provides us, therefore, with a solid statistical sample, although we have far less information about the 49 individuals who obtained 2003-2004 positions without registering with the service. The Survey includes some types of data not available through the Placement Service, especially data about the post-interview process (namely, the number of campus interviews and job offers per candidate, and the salary range of jobs received). The Survey also provides an opportunity for candidates to comment on hiring practices and violations of APA guidelines during interviews.
The Placement Service data draws on information from 359 subscribers, of whom 291 attended the annual meeting. The Surveys were sent to all 359 subscribers, and were returned by 127 (35%). This return rate is down a bit from previous years and is troubling. It is significant, however, that in terms of gender and age, the data from the Placement Service and from the Survey break into demographic groups that are statistically very close. (In terms of race/ethnicity the numbers are so small that variations exceed statistical significance.) In this regard, the numbers from the Placement Service suggest that the Survey data is highly reliable and not skewed by arbitrary return patterns.
II. Demographics
Placement Service
Excluding those who signed up as “Subscribers Only,” 359 people registered with the Placement service in 2002-03. Of these, 291 attended the annual meeting. Demographic breakdowns are consistent across these two groups (Table 2).
The gender ratio of candidates and candidates attending the meeting remains essentially unchanged since 1986, at roughly 60% males to 40% females (Figure 1). For 2002-3, 40% of all candidates were female (142 of 359), and 37% of candidates attending the meeting were female (107 of 291; see Table 2). While this statistic has been remarkably consistent, it is also disheartening, as it means that the field is moving no closer to gender parity.
This is consistent with the figures from the National Center for Education Statistics. In their table 255, showing degrees granted in Classics and related programs for the 2000-01 academic year, women and men are at parity for Bachelor’s and Master’s degrees. When we reach the Ph.D. level, however, women make up almost exactly 40% of the pool (20 out of 51). It is not clear whether this drop takes place because women are not going into Ph.D. programs or because they tend not to complete them at the same rate as men, but this appears to be the critical point for the existing gender imparity in the field.
The domestic status of applicants in general and at the meeting is also an area of concern for the CSWMG. Here there is some gender-based variation. Of all candidates, 41% of women are married, 46% are single, and 9% are listed as other long-term relationship (LTR). For men, 50% are married, 38% single, and 8% in other LTR (Table 11A). This suggests that it is somewhat easier to be in the academic job market – from the start – as a married man than as a married woman. With some minor variations, these same proportions hold for the domestic status of candidates at the meeting (Table 11B).
As a committee, CSWMG discussed the difficulties facing women candidates (married or not, with or without children) at the January 2004 meeting in San Francisco. Classics would seem to be neither better nor worse than other academic disciplines in achieving greater gender-equity in hiring. The CSWMG feels, however, that as a profession we should strive to lead other disciplines in this regard.
In terms of age, the breakdown of all candidates is again consistent with that of candidates at the meeting. For all candidates, the percentages are as follows: aged under 30, 18%; aged 30-39, 61%; aged 40-49, 14%; aged 50-59, 4%; aged 60 and older, less than 1% (Table 7A). For those attending the meeting, the percentages are nearly the same: aged under 30, 20%; aged 30-39, 62%; aged 40-49, 12%; aged 50-59, 2%; no candidates over 60 attended the meeting (Table 7B).
For citizenship/residency status, we have presented the data for the 291 candidates who attended the meeting. By far the largest group are US citizens (74%), followed by Canadian citizens (7%) and US residents (5%). Non-North-American nationals holding degrees from foreign institutions made up 7% of the field, while non-North-American nationals with degrees from US institutions made up 4% (Table 10A).
For race and ethnicity, we continue to have so little data that it is difficult to draw any conclusions. Of all 359 candidates registered, 15 self-identified as Asian/Pacific Islander (4%), 4 as Hispanic/Latino (1%), and 326 as White (91%) (Table 12A; see also Figure 2). These percentages remain virtually the same for minorities attending the meeting. Our profession continues to lag behind other Humanities in attracting minority candidates.
Professional Matters Division Survey
The Survey was sent to all candidates but “Subscribers Only” (359); 127 surveys were returned, for a return rate of 35%. It is significant, therefore, that the demographic breakdowns of the surveys are in every category quite close to the data from the Placement Service.
The gender breakdown maintains the percentages that we have seen before: 54 of the 127 respondents were women, or 42.5% (Table 16A) In terms of domestic status, we see some slightly different proportions than in the other set of data. For women, 46% were single, 48% married, and 6% in other LTR. For men, 48% were single, 38% married, and 14% in other LTR (Table 16A). Briefly put, married women were slightly more likely to return the survey than single women, and married men had a surprisingly low return rate.
The age of survey respondents breaks down in proportions quite close to the placement service data. Of those returning the surveys, 19% were aged under 30, 62% aged 30-39, 11% aged 40-49, 2% aged 50-59, and none aged over 60 (Table 16A). The Survey does not ask for data on citizenship/residency status.
With questions of race/ethnicity, our statistical base becomes even smaller and less representative. Eight respondents identified themselves as Asian/Pacific Islander, or 6% of respondents. This means that Asian/Pacific Islanders have a higher return rate than other ethnic groups (over 50%), but with a sample this small, individual responses can have a large effect on the sample. The vast majority of candidates (93%) identified themselves on the Survey as White; no other race/ethnic group was represented with a response (Table 16B).
III. Interviews at the Annual Meeting
Placement Service Data
When analyzing the data concerning interviews and hires, it is important to keep in mind the broad outlines of demographic breakdowns above. Since, for example, women made up roughly 40% of the candidates, as well as 37% of the candidates attending the meetings, we would expect them to obtain roughly 40% of the available interviews and jobs. Other factors may also affect the number of interviews per candidate in certain demographic groups; for example, as the age of candidates increases, the number of jobs that they apply for tends to decrease (Table 7B). We expect, therefore, that older candidates will obtain somewhat fewer interviews, and in this case we need to consider the interview rate (that is, interviews per application.)
We also need to consider the fact that the data on interviews only considers the 291 applicants who attended the annual meeting. As we have suggested above, however, this group is consistent with the fuller group of all candidates. It should be noted that 26 candidates who returned the survey reported more interviews on the survey than were recorded on their official records with the Placement Service. This may have been due to some institutions that had not advertised with the Service conducting interviews on their own at the convention. Only the official Placement Service records have been used for all the statistics in this report that include interviews.
When considering the number of interviews per candidate, it is possible to make several broad divisions. A full 45% of candidates received no interviews or only one. The next 35% received 2 (14%), 3 (11%), or 4 (10%) interviews. The following 9% received 5 (4%) or 6 (5%) interviews. And the remaining 10.6% of candidates received 7 interviews or more (Table 4A). Not surprisingly, this presents a picture of interview feast and famine: the top 31 candidates obtained 298 interviews, the next 26 received 145, the next 101 received 291, and the bottom 133 only obtained 65 interviews total. Even though the job market was better this year than several years ago, therefore, the relatively kind ratio of 2.08 candidates per position did not result in many interviews for candidates in the lower 80%; and the lower 45% face grim prospects.
Women obtained interviews at a somewhat greater rate than did men. Women make up 32% of those who received 0-1 interviews, 37% of those with 2-4 interviews, 46% of those with 5-6 interviews, and 45% of those with 7 or more interviews. Since women made up only 37% of the candidates at the meeting, they appear to have a slight edge at this stage of the hiring process. In this regard it is worth noting that only 17 of 68 (25%) candidates with no interviews were women.
This advantage is confirmed by the data in Table 11B; women averaged 3.2 interviews at the convention, where men averaged only 2.5. Women thus obtained 42.5% of all interviews, roughly 5 percentage points higher than their representation at the meeting (37%). This perceived advantage does not appear to be the result of higher application rates. Broken down by gender, 23% of women applied for 20 or more jobs, the same as men; 46% of women applied for 10-20 positions, compared to 53% of men; and 54% of women applied to 9 or fewer positions, as compared to 47% of men (Table 11B).
The advantage for women appears to be even stronger when we look at interview rates per application. As Table 11B shows, women who applied for 20 or more positions averaged 6.3 interviews, compared to 4.6 for men from the same group. Women who applied for 10 or more averaged 4.9, as compared to 3.5 for men. Within these groupings, domestic status seems to have a greater effect on women’s chances than on men’s. While single, married, and other LTR men who applied for more than 20 jobs obtained interviews at nearly the same rate, women in this same category had a distinct advantage if they were married. Married women obtained 7.3 interviews, as compared to 6.6 for other LTR and 5.3 for single women. These disparities may be greater among women because we have a smaller statistical sample than for men (22 women total, 43 men). The discrepancy is considerably less marked when we look at all candidates, rather than the small percentage who applied for 20 or more jobs.
Interview rates at the Annual Meeting are influenced by the field of specialization. The areas with the highest number of interviews in 2003 were Latin Literature (270 interviews, an average of 3.8 per candidate); Greek Literature (170 interviews, average 2.3 per candidate), Art/Archaeology (149 interviews, average 3.0 per candidate). We cannot, however, trace the apparent advantage for women to this breakdown. Although there is a higher percentage of women in Art/Archaeology (51%) than their representation in the field as a whole, their proportion in Latin Literature (42%) is only slightly higher, and their proportion in Greek Literature (25%) is significantly lower. If we take these three fields together, women make up 38% of the candidates, virtually the same as their ratio in the field as a whole (Table 6A).
Publication records also affect interview rates, although in somewhat counterintuitive ways. From Table 13C, it appears that publishing a book reduces one’s chance of obtaining interviews. The 59 candidates who had published a book averaged 1.6 interviews; those who had not averaged 3.0. Obviously some other factors come into play here; candidates with a book were applying for fewer jobs (Table 13A) and may also have been applying for more senior jobs, in which fewer candidates were interviewed. Interestingly, however, the gender advantage is also evident when we break candidates down by publication rate; women obtained more interviews than men when looking at candidates with a book or without, and the same is true for the categories of 0 articles, 1-2 articles, 3-4 articles. Only when we reach the categories of 5-9 articles and 10+ articles are male candidates interviewed at a higher rate; here again, our statistical sample is fairly small (50 candidates total).
Last year’s report raised the point that age discrimination appears to be a problem at the interview stage, and this year’s figures support such a conclusion. Candidates under the age of thirty averaged 3.3 interviews, and those aged 30-39 averaged 3.0. By contrast, candidates aged 40-49 averaged 1.1, and those aged 50-59 averaged only 0.1. While some of this sharp drop can be explained by the fact that candidates tend to apply for fewer jobs as they advance in age (Table 7A), that is not the sole cause. Those candidates over the age of 40 who applied for more than 20 positions averaged far fewer interviews than their younger counterparts, as did those who applied for more than 10 positions (Table 7B). It appears that institutions are hesitant to interview candidates over the age of 40.
The figures for race and ethnicity continue to be so low as to be within a statistical margin of error. Only 19 candidates self-identified as any type of minority, 15 of them as Asian/Pacific Islander. Of these, 16 attended the meeting (13 Asian/Pacific Islander). They averaged between 2.6 and 3.0 interviews apiece(Table 12A), a figure virtually identical with the average for white candidates of 2.8. From these figures, all we can say with certainty is that minorities do not hold any clear advantage in obtaining interviews and continue to be badly underrepresented in the field.
Professional Matters Division Survey
The Survey asked respondents to identify inappropriate interview topics, and to comment on any violations of APA guidelines during their interviews. As in previous years, several candidates did feel that some inappropriate questions were asked. We do not see any clear patterns in these reports with regard to gender. The most common of such questions appear to be in the areas of race, religion, marital status, and partner’s willingness to relocate (Table 20A).
Twenty-one candidates indicated that they witnessed, but did not report, violations of APA guidelines (Table 20B). The specific violations included inappropriate remarks or questions, inappropriate behavior (including use of alcohol), and the perennial favorite, inappropriate place. The tight scheduling of interview rooms continues to be a problem: candidates end up standing in the hallway while the previous interview concludes or are kicked out of the room and have to conclude the interview in the same hallway (Table 20C). The CSWMG strongly urges that this difficulty be addressed.
The most common reason given for not reporting such violations was “personal indifference.” Under the category of “other,” candidates tended to indicate that they felt that the violation was reasonable under the circumstances, e.g., “Questions seemed reasonable given nature of institution.” Two comments by males indicated a lack of faith in the integrity of the hiring process as a whole.
IV. Hiring
Our most detailed data comes from the 126 jobs that were filled by members registered with the Placement Service. An additional 49 jobs (38% of total) were filled by candidates not registered with the service, so that we have very limited data about those positions (Table 15A and Table 15B). Of the 175 positions total, 92 (52%) were temporary (adjunct or non-tenure track), while the other 81 (48%) were tenure-track or tenured. When we look at the full group of jobs, we see somewhat less of an advantage on gender lines than in the interview process. Women obtained 37% of temporary positions, 41% of permanent positions, and 39% of all positions(Figure 3). These numbers are very close to the percentage of female candidates (37%) and also correspond closely to the percentages for the 126 jobs filled by candidates registered with the Placement Service, for which we have fuller data. Readers should bear in mind that the discussion that follows is based on the 126 jobs obtained by Placement Service candidates, not the full 175 jobs.
Of the 126 jobs filled by candidates registered with the Placement Service, 32 were not announced to the APA. Because of the correlation of Survey respondents with Placement Service data, however, we are able to include information about these 32 jobs in all the tables that pertain to hiring. These 126 jobs, then, were of the following types: 73 (58%) were temporary, and 53 (42%) were tenure-track or tenured.
The ratios of hires are consistent, in terms of gender, with the ratios for interviews (see Table 4B and Table 4C for information about gender, number of interviews, and new academic jobs). Female candidates obtained 41% of all positions and 41.5% of all tenure-track or tenured positions (Table 11C). Put another way, 37% of women obtained a position (as compared to 34% of men), and 15% of women obtained a tenure-track or tenured position (as opposed to 14% of men). While these numbers show a very slight advantage for women on the job market (compared to their 40% representation in the candidate pool), it is important to note that these numbers show a marked drop in the success rate from the previous year for women obtaining tenure-track jobs. In the previous year, women received 49% of all permanent jobs obtained by candidates (see Figure 5 for a comparative gender breakdown of various aspects of the placement process for the years 2002-2003 and 2001-2002). It now appears that the 2001-02 data represented a blip rather than a trend.
Domestic Status appears not to have had a significant effect on men’s or women’s chances of getting a job (Table 11C). The success rate for single women was 38%, for married women 39%. Both numbers are quite close to women’s overall 37% success rate. Although other LTR women were hired at a rate of only 23%, the statistical sample is very small (only 3 candidates.) For tenure-track positions we see more variation, but again, the samples are smaller. Only 9% of single women candidates (6 total) obtained permanent positions. Married women obtained permanent posts at a rate of 22%; other LTR at a rate of 15%. For men, the numbers in the groups with largest representation show more variation: 38% of single men obtained jobs, as compared to 30% of married men. A full 44% of other LTR men obtained posts (8 jobs total). The rates for tenure-track jobs show less variation: 15% of single men obtained permanent jobs, 12% of married men, and 22% of other LTR men. From these numbers it is difficult to pinpoint any specific trends, which perhaps suggests that domestic status is not a significant factor in hiring considerations.
When we look at the three areas of specialization that had the most positions, we do not see any particular advantage for women in obtaining positions (Table 6B). Women candidates made up 25% of the Greek Literature candidate pool, and obtained 31% of the jobs. This imbalance is mirrored, however, by Art/Archaeology, where the candidate pool is 51% female, but women obtained only 47% of the jobs. In Latin literature, where women make up 42% of the pool, they obtained 42% of the jobs. Again, the breakdown by area of specialization does not result in statistically significant trends.
The slight advantage for women that we saw for interview rates when broken down by amount of publication is less clear when it comes to hiring (Table 13E and Table 13F; see Table 13D for information on all hires by amount of publication). Women had higher success rates than men whether the candidates had published a book (30% of women obtained jobs, 22% of men) or not (39% of women, 37% of men). But while women with no articles did better than men (46% compared to 36% success), men with 1-2 articles did better than women (33% for men, 27% for women). At 3-4 articles, women had a 35% success rate compared to men’s 31% rate. In the categories of 5-9 articles and 10+, the numbers are again very small. Perhaps the most curious fact is that publishing a book seems not to be an indicator of success for men or women in obtaining tenure-track jobs; for both genders, the success rate is higher for candidates without books than for those with. One number, in addition, sticks out as anomalous: women candidates with no articles had a surprising 21% success rate at obtaining tenure-track jobs. The only groups to surpass this rate were men with 5-9 articles (38%) and women with 5-9 articles (31%). We will want to watch this category closely in future years to see if this is a trend or just an anomaly.
The discrimination based on age that appears at the interview stage also seems to be in effect at the hiring stage (Table 7C). Candidates under the age of 30 had the best chance of obtaining a job, with a success rate of 44%, followed closely by those aged 30-40 (37%). Candidates aged 40-49 had only a 29% success rate, and only 13% of candidates aged 50-59 obtained a job. These trends are less marked, but present, when we look at tenure-track jobs. The most successful age group at obtaining permanent positions was those aged 30-39, at 18%, followed by those under 30 (13%). Those aged 40-49 obtained permanent positions at the rate of 10%, and no candidates aged 50-59 obtained a tenure-track job. While it is true that application rates tend to go down as age goes up (Table 7B), this fact alone does not explain the difficulty that older candidates have in winning jobs, whether temporary or tenure-track.
In terms of race/ethnicity, our numbers of minority candidates in 2002-03 are so low that any anomalies could be a matter of statistical variance. We can say, however, that minority candidates did not hold any advantage in the job market (Table 12B). Five of the 126 placement service positions went to ethnic minorities (4%). Minorities made up 5% of the total candidates; in other words, they had a lower success rate than whites in obtaining jobs (Asian/Pacific Islander 27%; Hispanic/Latino, 25%; White 37%). This was simply an abysmal year with respect to the goal of ethnic/racial diversity of our field.
Professional Matters Division Survey
One hundred and twenty-seven candidates returned the Survey, of whom 79 obtained a position of some kind (63%). Obviously, then, successful candidates were more likely than unsuccessful ones to return to survey. Within this group, however, hires are consistent with the larger set of data from the Placement Service. Those candidates, for example, who received either one or two job offers were 40.5% women, a familiar ratio. Of those who did not receive a job offer, 44% are women (Table 19A). Forty-two percent of all these jobs went to women (Table 19D). This set of data gives us, moreover, some information about campus visits and salary range that is not available from the Placement Service Data. For example, 67 respondents (53%) reported campus interviews on the survey, but 14 of these 67 (21%) did not obtain new academic positions (Table 18)
Interestingly, most of the survey respondents who got jobs reported that they had not been interviewed for that specific position at the annual meeting (61% of women and 63% of men; see Table 19B).
Within this data set, some interesting trends appear. For the salary ranges where we have a relatively large statistical sample, women’s representation is proportional to their percentage of the larger pool. That is, women obtained 42% of the jobs paying $30K-$40K; 40% of those paying $40K-$50K; and 44% of those paying $50K-$60K (Table 19C; see also Figure 4). When we look at the type of position (Table 19D, however, men had a striking success rate in departments offering the Ph.D: 76% of those 29 jobs went to men. The lower the highest degree offered by the department, the more likely women were to obtain the job. Women obtained 47% of the jobs in departments offering the M.A., 54% of the jobs in those offering the B.A., and 67% of the jobs in departments without a classics major. (This last category consisted of only 3 jobs, however). This upward slope is the reverse of what appeared in the 2001-02 survey data. Again, this is a category we will want to watch closely. If it marks a consistent trend, then it appears that women are being pushed towards jobs in small, liberal arts colleges where they will not teach graduate students; this will not serve the profession well in our pursuit of gender equity.
V. Conclusions
In last year’s report, the authors wrote, “women had a banner year in 2001-2002.” In that year, women were hired at a greater rate than their representation in the full candidate pool and in the pool of those attending the meeting (see Figure 5). We cannot say the same for 2002-03. Although women did appear to have a slight advantage in obtaining interviews (42.5% of interviews, as compared to their 37% representation in the meeting pool), their hiring rate (41%) was only marginally higher than their representation in the entire candidate pool (40%). It was certainly not a bad year for women, but it must be admitted that we are making little progress, and that at a snail’s pace, towards the goal of gender equity. As noted earlier in this report, data from the National Center for Educational Statistics suggests that the crucial filter point is graduate school; it is in the granting of Ph.D.’s that women’s representation in Classics falls from near equity to roughly 40%. This, then, is the point at which we must address attention.
Particularly disturbing in this year’s figures are the statistics showing that men were disproportionately hired by institutions offering the Ph.D. This figure will need to be monitored closely over the next several years.
Ethnic and racial minorities continue to be embarrassingly under-represented in the discipline of Classics. We suspect that this trend begins well before graduate school, and will not be corrected without a wholesale effort at the undergraduate and high-school levels. This remains a key area of focus for the CSWMG.
Age discrimination, noted in last year’s report, also appears to be a significant problem for 2002-03. Candidates over the age of 40 suffered clear disadvantages at both the interview stage and the hiring stage. We must address this issue, as a committee and as a profession.
In sum, although great progress has been made in the area of combating gender imparity over the last thirty years, we now appear to be stuck at a 40% plateau. In the areas of age discrimination and racial/ethnic under-representation, the road ahead is both steep and rocky. These areas will not improve without a major initiative on the part of the profession.
We welcome comments and suggestions from APA members regarding any aspect of placement, or other practices of the APA. Such comments should be sent to the present Chair of CSWMG, Kirk Ormand, Dept. of Classics, Oberlin College, Oberlin, OH 44074, or kirk.ormand@oberlin.edu.
Report submitted by Kirk Ormand and Cashman Kerr Prince, CSWMG
Division of Professional Matters
American Philological Association
July 2004