THE RELOCATION DECISIONS OF ACCOUNTING FACULTY

THE RELOCATION DECISIONS OF ACCOUNTING FACULTY

Michael E. Bitter, Ph.D., CPA

Associate Professor of Accounting

M.E. Rinker, Sr. Institute of Tax and Accountancy

School of Business Administration

Stetson University

Note: The author wishes to thank the Stetson University Summer Research Grant program for its financial support of this project and Jenny VanGemert for her assistance in data collection and entry. Requests for data should be addressed to the author.

THE RELOCATION DECISIONS OF ACCOUNTING FACULTY

Abstract

Several factors could potentially contribute to an accounting professor's decision to relocate--an imbalance in the supply of and demand for accounting academics (Schultz, 1989), salary compression (Jacobs & Herring, 1987; Schultz, 1989), the failure to receive tenure (Schultz, 1989; Schultz, et al., 1989), and nonwork (personal) issues (Holland & Arrington, 1987). Given conditions that encourage relocation and because faculty recruiting is an expensive and time-consuming endeavor, employer institutions should understand the factors that influence an accounting faculty's decision to relocate.

This study explores the relocation decisions of experienced accounting faculty--the variables considered and characteristics of the current and previous positions. A survey was mailed to 238 accounting professors who had relocated to a U.S. institution during 1996 or 1997. Ninety-eight usable responses were received, yielding a response rate of 41 percent.

A number of variables were of some importance to the relocation decision. Personal and faculty/administrative-related variables were most prominent among the 26 variables positively impacting the relocation decision. Surprisingly, all but one of the compensation variables were absent from this group. The ratings of faculty at doctoral-granting schools (DOC) for compensation, research resources, research support, and doctoral education variables were higher than those of faculty at schools not offering doctoral degrees (NONDOC).

THE RELOCATION DECISIONS OF ACCOUNTING FACULTY

Several factors could potentially contribute to an accounting professor's decision to relocate to a new academic institution. These factors include an imbalance in the supply of and demand for accounting academics (Schultz, 1989), salary compression (Jacobs & Herring, 1987; Schultz, 1989), the failure to receive tenure (Schultz, 1989; Schultz, et al., 1989), and nonwork (personal) issues (Holland & Arrington, 1987).

Given conditions that encourage relocation and because faculty recruiting is an expensive and time-consuming endeavor, it is important for employer institutions to understand the factors that influence an accounting faculty's decision to relocate. This study examines the relocation decisions of experienced accounting faculty. Information was obtained from 98 experienced faculty members about the variables that influenced their decision to relocate as well as about their current and previous positions.

Prior research and the motivation for the study are outlined in the next section. The sections following describe the research methodology, discuss the results, and summarize the study.

MOTIVATION FOR STUDY AND PRIOR RESEARCH

Over the past several years, the number of new accounting doctorates has decreased dramatically (Bitter, 2000). This decline, coupled with increasing market demand due to position backlog, the creation of new programs, turnover, and retirements, has resulted in a shortage of qualified accounting professors (AACSB, 1998). When accounting programs cannot meet their needs by hiring new graduates, they have little choice but to lure away experienced professors from other institutions with better offers (AACSB, 1998). In many instances, an experienced faculty member (with teaching experience, a proven research record, and/or a strong service record) is preferable to an inexperienced one.

The heated market for new doctorates has pushed starting salaries ever higher and has made the reality of salary compression even harsher for experienced faculty. On average, those with new doctorates in accounting were paid $88,700 in 2000 (AACSB, 2001). As a reference point, the average assistant, associate, and full professor earned $76,800, $80,600 and $98,900, respectively. Certainly, accounting faculty who have fallen victim to significant salary compression and have maintained their skills have incentive to change institutions.

Some accounting faculty have little choice but to change institutions when they are denied tenure or when they perceive their chances for tenure as unlikely. Schultz, et al. (1989) noted that success rates for those applying for tenure significantly declined between the early 1960s and the mid-1980s; this downward trend was expected to continue into the 1990s.

Salary and tenure issues are not the only reasons many faculty have chosen to move. In a study of experienced accounting faculty that had relocated between 1980 and 1985, Holland & Arrington (1987) found that personal factors were most important to relocation decisions. These included the spouse's and family's happiness, the overall nonprofessional quality of life, and geography. When Holland and Arrington partitioned the sample, they found that faculty relocating to schools with a doctoral program valued professional variables over personal variables. The responses of faculty relocating to schools without a doctoral program mirrored those of the aggregate sample.

Understanding the factors that influence the relocation decisions of accounting faculty is important not only for institutions seeking to hire experienced accounting faculty, but also for those that desire to retain their faculty. Recruiting institutions could use information about faculty relocation decisions to evaluate their competitive position and potentially improve the overall effectiveness of their recruiting efforts. Institutions concerned with retention could use this information to determine what steps, if any, they can take to improve retention by providing valued faculty with incentives to stay.

Accounting faculty contemplating a re-entry into the market in the near future (especially those who have been "out of the market" for several years) could benefit from the experiences of relocating faculty. Certainly, there can be significant trade-offs associated with any job relocation.

The purpose of this study is to explore the relocation decisions of experienced accounting faculty, thus updating the work of Holland & Arrington (1987). Faculty relocations are examined in two ways. First, variables that were most important to the decision process will be determined. Second, information about both the current and the previous position will be obtained, including data on teaching loads, work load mix, research mix, and tenure/promotion criteria. This information can potentially provide additional insight into the reasons experienced faculty relocated.

RESEARCH METHODOLOGY

Subjects

A two-page survey was mailed to experienced accounting professors who relocated to a U.S. institution during 1996 or 1997. The mailing list was compiled from Hasselback's 1998-1999 Accounting Faculty Directory. To encourage response, subjects were provided with a postage-paid return envelope, guaranteed anonymity, and offered an executive summary of the results.

Research Instrument

The research instrument contained three sections. The first listed 54 variables that may be important to the decision to change jobs. [1] For presentation purposes, the 54 variables were judgmentally categorized into one of two groups--institutional variables and personal variables. The institutional variables group was further subdivided into eight categories: general, compensation, teaching, research support, service, faculty/administration, facilities, and resources.

Subjects were asked to rate the relative importance of each variable using the five point scale developed by Holland & Arrington (1987). The scale is as follows:

5. Primary positive factor. The decision to relocate was

made in part because you perceived the level of job

satisfaction associated with this factor to be much

greater at your current school.

4. Important positive factor. The decision to move was

made in part because you perceived the level of job

satisfaction associated with this factor to be

somewhat greater at your current school.

3. Not an important factor. You perceived no difference

between your current school and your former school

OR any difference was not important to your decision

to relocate.

2. Important negative factor. The decision to move was

made in spite of the fact that you perceived the

level of job satisfaction associated with this factor

to be somewhat greater at your former school.

1. Primary negative factor. The decision to move was

made in spite of the fact that you perceived the

level of job satisfaction associated with this factor

to be much greater at your former school.

0. No opinion or not applicable.

At the end of the section, subjects were asked to identify, in order of importance, the five

variables most important to their relocation decision.

Section two included questions about both the respondent's current and previous position and institution. [2] Section three requested demographic information.

Analysis

Mean importance ratings for each variable were computed for the overall sample as well as for two partitions of the sample (hereafter called subsamples): respondents relocating to schools with an accounting-related doctoral program (DOC) and those relocating to schools without a doctoral program (NONDOC). Responses from those at DOC and NONDOC schools were expected to differ in many cases due to varying missions and faculty expectations. The Mann-Whitney U (MW) test was used to examine group differences. [3]

RESULTS

Characteristics of Respondents

Of the 238 surveys mailed to relocated accounting professors, 98 usable responses were received, yielding a response rate of 41 percent. The majority of the respondents were white (90 percent) and married (77 percent); the average age of the respondents was 44. Most held an appointment at the assistant (47 percent) or associate (30 percent) professor rank at their current institution. See Table 1.

Seventy of the respondents (71 percent) relocated to a NONDOC institution, with the majority (50 of the 70) moving from another NONDOC school. Eighteen respondents (18 percent) moved from one DOC school to another, while only eight (8 percent) moved from a NONDOC school to a DOC school. Interestingly, only 23 respondents (24 percent) received a promotion upon relocation to their new school.

Respondents were asked to rate their level of satisfaction with their previous position at the time they left on a scale from one (very dissatisfied) to five (very satisfied). The mean satisfaction rating was 2.57, indicating that the typical respondent was mildly dissatisfied with their previous position. There were no significant differences between the mean satisfaction ratings for respondents moving from DOC (2.63) and NONDOC schools (2.53). See Table 2.

The profile of the typical respondent's current employer was as follows: a public (65 percent), regionally-focused (67 percent), teaching-oriented (86 percent) institution located in the southeast (22 percent) or midwest (27 percent). Generally, there tended to be a migration away from research-oriented schools toward teaching-oriented schools.

The majority of the respondents' institutions offered an accounting-related bachelor degree (93 percent), a MBA degree (86 percent), and an accounting master's degree (50 percent) and were AACSB accredited (58 percent). Twenty-seven percent offered an accounting-related doctoral degree. Respondents' current schools were less likely to offer an accounting-related doctoral program or hold AACSB accreditation than their previous schools.

The enrollments for current employer institutions ranged from 1,400 to 60,000 students, with an average of 12,093 students. The average accounting faculty numbered ten full-time, tenured/tenure-track members. See Table 3.

Descriptive Results

Overall Sample. Twenty-six of the 54 variables had a mean importance rating above three ("not an important factor"), but none received an average rating above four ("important positive factor"). Given this, it appears the average respondent did not perceive the new position as a significant improvement over their previous position.

Of the 26 variables positively impacting the relocation decision, particular importance was placed on personal (nonwork) and faculty/administrative-related variables. Personal variables such as happiness of the spouse and family, the geographic location of the school, overall nonprofessional quality of life, and location: urban-suburban-rural ranked first, fourth, sixth, and fifteenth, respectively, in importance. Faculty/administrative-related variables such as the collegiality of and compatibility with the faculty, ranked second and third, while quality of the faculty, compatibility with the dean and the department chair ranked seventh, eighth, and thirteenth, respectively. Notably absent among the 26 variables were all but one of the compensation-related variables; base salary tied for eighth. See Table 4. Twenty-two respondents chose to write-in an additional variable that had been important to their relocation decision. [4]

The high relative ratings of the personal variables is consistent with Holland & Arrington (1987), where four of the top five variables were personal. Faculty-related variables, however, were rated much higher in this study than in the Holland and Arrington (H & A) study.

Subsamples. The ratings of faculty moving to a DOC school differed in many respects from those of the overall sample and those of the DOC group in the H & A study. In comparison to the overall sample, happiness of the spouse and family remained the highest rated variable, but the base salary variable climbed to second. Following closely were collegiality of the faculty, overall nonprofessional quality of life, quality of the faculty, and reputation of the department. Four of the top five variables for the DOC group in H & A's study were professional (salary, chairperson, perceived changes in market value, and research opportunities and support).

The two doctoral education variables, ranked at the bottom in the overall sample, climbed significantly, ranking thirteenth and seventeenth. Research-related variables such as the research interests of the faculty, research support, and library resources were also ranked much higher by faculty moving to DOC schools than by the overall sample.

The ratings of faculty moving to NONDOC schools were very similar to those of the overall sample, which is not overly surprising given that 71 percent of the respondents moved to a NONDOC school. The ratings were also consistent with H & A, where faculty at NONDOC schools rated four personal variables among the top five.

Subsample Differences. The mean importance ratings of respondents moving to DOC schools differed from those of respondents moving to NONDOC schools for 13 of the 54 variables. For 11 of the 13 variables, mean ratings were higher for DOC school respondents. Generally, DOC school faculty considered compensation, research resources, research support, and doctoral education variables as more important than those at NONDOC schools. See Table 4.

Most Important Hiring Variables

It was expected that many variables would be rated similarly, in aggregate, by respondents. As such, in order to determine which variables were the most important to the relocation decision, respondents were asked to list, in order of importance, the five variables most important. Frequencies were tallied and a rank score was calculated for each variable by awarding five points each time the variable was selected as most important, four points each time the variable was selected next most important, etc.

Rank scores indicate that the happiness of the respondent's spouse and/or family was the most important variable, scoring 131 points. This is especially impressive since only 75 of the 98 respondents (77 percent) are married or have family residing with them; presumably, the other 23 respondents would not have ranked this variable in their top five.

The second most important variable was base salary (99 points). Interestingly, this variable tied for 8th based upon its mean importance rating. This may indicate that base salary is very important to many respondents (those ranking this variable in their top five), but not to all of them. The same can be said for the variable ranked third, the likelihood of obtaining tenure (89 points). Despite its high ranking, the variable's mean importance rating placed it seventeenth among all variables. Since only 13 of the 98 (13 percent) respondents were tenured at their new school, one might expect the likelihood of obtaining tenure to be important to most respondents. The conflicting evidence suggests that the likelihood of obtaining tenure is very important to many, but not all respondents. Perhaps some respondents intend to remain mobile and have little interest in tenure.

Completing the top five were the geographic location of the school and compatibility with the faculty. The remainder of the top ten included two personal factors, two faculty/administration-related variables, and promotion/tenure criteria. Noticeably absent from the top ten were any research-related variables. See Table 5.

Additional analyses were performed to determine the importance rankings for respondents at DOC schools and NONDOC schools. The top three variables for respondents at DOC institutions were the same as those for the overall sample, except that the base salary variable was ranked as the most important (45 points). The importance placed on base salary may indicate that some respondents relocated to DOC schools, in part, to overcome salary compression at a previous institution. Also included among the top five were the research interests of the faculty (tied for third with 26 points) and the reputation of the department and compatibility with the faculty (both tied for fifth with 24 points).

Similar to the overall sample, respondents from NONDOC schools ranked happiness of the spouse/family as, by far, the most important (104 points). Two additional personal factors, the geographic location of the school and ties to the region, were ranked second and third (75 and 65 points), respectively. The likelihood of obtaining tenure and promotion/tenure criteria were ranked fourth and fifth, respectively. Note that three of the top five variables were personal and the remaining two related to promotion and tenure.

In summary, respondents from DOC schools were more interested in base salary and research-related variables, while those from NONDOC schools had special concern for the happiness of their families and the location of the school. Both groups placed a high level of importance on the likelihood of obtaining tenure.

Analysis of Faculty Movement

For respondents moving from a NONDOC school to another NONDOC school, the five variables with the highest mean importance ratings were compatibility with the faculty, the geographic location of the school, collegiality of the faculty, quality of the faculty, and the opportunity to teach desired courses. For those moving from DOC to NONDOC schools the five highest rated variables were compatibility with the dean, collegiality of the faculty, happiness of the spouse and family, university/department/school mission, and the opportunity to teach desired courses.

There were several significant differences between the ratings of those moving from a DOC to a NONDOC school and those moving from a NONDOC to another NONDOC school. These variables include quality of the faculty, reputation of the department, ties to the region, student placement, research interests of the faculty, faculty governance, research support, library resources, and availability of summer teaching. In each case, "NONDOC to NONDOC" respondents rated their current position more favorably relative to their previous position than did "DOC to NONDOC" respondents. In fact, the mean ratings for these variables by "DOC to NONDOC" respondents were all below three, indicating that the job satisfaction associated with this variable was somewhat better at the previous DOC institution. This finding also supports that possibility that the relocations from DOC to NONDOC schools were not voluntary.

For respondents moving from a DOC to another DOC school, the five highest rated variables were the overall nonprofessional quality of life, collegiality of the faculty, base salary, research interests of the faculty, and the geographic location of the school. It is interesting to note the importance placed on geographic location by these respondents. In contrast, the mean importance rating for this variable by respondents now at DOC schools (coming from both DOC and NONDOC schools) ranked this variable twelfth. The ratings of respondents moving from NONDOC to DOC schools are not discussed because of the low number of respondents (eight) making such a move.

Profile of Teaching Assignments

Respondents at schools on a semester (quarter) system averaged five (five) classes and three (three) preparations per academic year (excluding summer) and taught three (three) days per week. Teaching loads at DOC schools on a semester (quarter) system averaged four (five) classes and two (three) preps per academic year versus six (six) classes and three (three) preps for those at NONDOC schools. On average, teaching loads at current and previous schools were virtually identical, indicating that respondents apparently did not relocate simply to reduce their teaching load. See Panel A of Table 6.

Overall, respondents taught primarily upper division courses (48 percent), splitting the remainder of their teaching load between introductory (23 percent) and graduate courses (29 percent). Respondents at DOC schools spent more time teaching upper division and graduate courses and less time teaching introductory courses than respondents at NONDOC schools. Given the emphasis of DOC institutions on graduate education and the likely availability of graduate assistants to teach introductory courses, this finding is not surprising. Compared with teaching assignments at their previous institutions, respondents were now teaching less at the upper division level and more at the graduate level. See Panel B of Table 6.

Forty-nine percent of the respondents teach, on average, two courses in summer school. Respondents now at NONDOC schools are much more likely (59 percent) to teach summer school than those at DOC schools (27 percent). See Panel C of Table 6.

Profile of Work Load and Tenure/Promotion Criteria

Respondents were asked to estimate their current and previous work load mixes. Overall, respondents spent about half (51 percent) of their time on teaching activities and 35 percent of their time on research activities. Compared to their previous position, respondents now spend slightly more time on teaching and less time on research.

As might be expected, the research component of the work load was highest at DOC schools (48 percent) and was similar to the work load of respondents who previously held a position at a DOC school. Also as expected, faculty at NONDOC schools spent the majority of their time (55 percent) on teaching activities. See Panel A of Table 7

Of the research that was being conducted by the respondents, the majority (58 percent) was targeted toward academic journals. Academic journals were the focus of respondents at DOC schools (92 percent), while respondents at NONDOC schools targeted their research for both academic (46 percent) and practitioner (40 percent) journals. A comparison of current and previous positions indicated a shift away from research for academic journals toward research for practitioner journals and other types of research. See Panel B of Table 7.

Respondents were asked to indicate their perception of the importance of teaching, research, and service effectiveness in tenure and promotion decisions at their current and previous schools. Overall, and for respondents at NONDOC schools, teaching effectiveness was perceived as most important. This finding is consistent with H & A, but is inconsistent with prior research (e.g., Hermanson, et. al., 1996; Schultz, et. al., 1989; and Street, et. al., 1993) that found that research effectiveness was of primary importance to promotion and tenure decisions. Consistent with both H & A and other prior research, however, respondents at DOC schools perceived research effectiveness to be most important.

A comparison of current and previous positions shows a marked shift in the promotion/tenure importance weightings away from research effectiveness and toward teaching effectiveness. This was true for both DOC and NONDOC institutions. See Panel C of Table 7.

It was suspected that the type of research performed by respondents and promotion/tenure weightings might vary depending upon whether a respondent had been promoted by the new institution. Consistent with the findings of H & A, there were no significant differences in the research mixes reported by promoted respondents and those that had not been promoted.

Differences, however, were noted in the promotion/tenure importance weightings. Teaching effectiveness was weighted as more important by respondents that had been promoted, while research effectiveness was perceived as more important (although not quite as important as teaching effectiveness) by those that had not been promoted. This is contrary to H & A, where the importance weightings of promoted faculty were almost identical to those of faculty who had not been promoted. See Table 8.

SUMMARY

There were a number of variables important to the decision to relocate, although none received an average rating above four ("important positive factor") on the five point scale. Personal and faculty/administrative-related variables were most prominent among the 26 variables positively impacting the relocation decision. Alternatively, all but one of the compensation variables were absent from this group.

The variables rated highest by faculty from both DOC and NONDOC schools were similar to those rated highest by the overall sample. Doctoral education and research-related variables, however, were rated higher by DOC faculty than by the overall sample. In comparison, the ratings of DOC faculty for compensation, research resources, research support, and doctoral education variables were higher than those of NONDOC faculty.

When asked to identify the five variables most important to their relocation decision, respondents now at DOC schools were more concerned with base salary and research-related variables, while those now at NONDOC schools showed greater concern for the happiness of their families and the location of the school.

The majority of respondents relocated to a NONDOC school. NONDOC respondents relocating from another NONDOC school rated their new position much more favorably relative to their previous position than did those relocating from a DOC school. The typical respondent taught five classes per year with three preparations. The teaching load at NONDOC schools, on average, was greater than that at DOC schools. Respondents at DOC schools teach more at the upper division and graduate levels and less at the introductory level than those at NONDOC schools.

The work load of respondents at DOC schools was weighted toward research, while the respondents at NONDOC schools spent the majority of their time on teaching activities. Regardless of affiliation, respondents reported that their research was aimed more toward academic journals than practitioner journals. Teaching effectiveness was perceived as most important to tenure and promotion decisions by the overall sample and by faculty at NONDOC schools. Not surprisingly, research effectiveness was perceived as most important by faculty at DOC schools.

What can institutions do to retain their accounting faculty? Apparently, not much. Survey results point to personal and faculty/administrative variables as most influencing the decision to relocate. Most of these variables are not controllable, at least in the short-term, by the institution. Perhaps the best advice that can be given to institutions is to spend more time and expend greater effort in the recruitment process. Evaluate the match between job candidates and the faculty, the administration, the university's culture, and the local community. Poor matches mean a much higher likelihood of turnover.

The findings of this study are subject to certain limitations. First, information provided on the survey was self-reported and may contain errors or omissions. Second, as with most survey research, the results may not be representative of the population due to nonresponse bias. Finally, the survey was sent to faculty that had relocated during 1996 and 1997. Thus, it is impossible to know whether the findings will hold true in the future.