The Effect of Tuition and Opportunity Cost on the Pursuit and Completion
of a Graduate Management Degree
Mark Montgomery and Irene Powell
Dept of Economics
Grinnell IA, 50112
Fax: (641) 269-4985
This study uses multivariate statistical analysis to examine the impact cost on the
likelihood a person will enroll in and complete, respectively, an MBA program. It considers both
the explicit cost of paying tuition and the implicit cost, or opportunity cost, of earnings foregone
while in school. Results suggest that tuition is not a major impediment either to pursuing or to
completing an MBA. High opportunity cost, however, tends to reduce enrollment and degree
Acknowledgments: The authors are grateful to Jason Bent for valuable research assistance and
to Mary Kay Dugan and Mark Gritz for helpful suggestions. This study was supported by the
Battelle Memorial Institute in Seattle, Washington, and by a grant from the Grinnell College
Faculty Grant Board.
The few studies that exist on the demand for graduate business education (Jantzen ,
Montgomery ) have focused on the impact of tuition cost on the demand for a particular
institution. This research examines the impact of the cost of obtaining a degree on the decision
to enroll in a graduate management program, and on the likelihood of completing once enrolled.
The analysis considers more than just tuition cost. Tuition is only the explicit cost of pursuing a
Completion of a Management Degree, page 2
graduate management degree. There is also a substantial implicit cost of lost earnings while
attending school. The latter is usually called the “opportunity cost” of obtaining the degree.
Two questions are considered in this research: 1) how do the explicit and implicit costs influence
the decision to pursue a management degree, and 2) how do these costs affect the likelihood of
completing the degree? The cost effects are compared across demographic groups and between
more-competitive and less-competitive business schools. The data for this study came from a
unique longitudinal survey of registrants for the Graduate Management Admission Test [GMAT]
REVIEW OF THE LITERATURE
The research that is most relevant to the present study is that done on graduate school
attendance or completion decisions. Of these, almost all examine determinants of matriculation
only (Weiler ; Eide, Brewer, and Ehrenberg ; Fox ).
Two studies look at determinants of graduate school completion for samples of students
who have already made the decision to attend. For example, Baker  used a set of
applicants for fellowships for science and engineering graduate school who all were going to
attend their respective programs. Ehrenberg and Mavros  used data on students already
enrolled in graduate programs in four disciplines. By using data on students who have already
made the decision to attend, sample selection might result in biased estimates of the effect of
factors such as ability, gender, or race, on the probability of completion. For example, if ability
is correlated with the probability of attendance – which it surely is – then the effect of ability on
Completion of a Management Degree, page 3
completion for the group of attendees will be biased. Additionally, by using data only on
students enrolled, earlier studies could not take account of cost on the behaviors of people who
had abandoned plans to attend precisely because the costs were prohibitive. The current study
looks at such people.
The research just discussed analyzes attendance or completion either in Ph.D. programs,
or in both Ph.D. and professional schools combined, none looked at business schools. They also
use data on individual students. Jantzen  attempts to “identify the factors that influence
the demand for U.S. graduate business programs,” using data on business schools and their
enrollments. He found that enrollment at individual business schools was negatively affected by
tuition costs and positively affected by quality and being in a generally high demand area. But
because Jantzen used data on schools, not individuals, his research could say nothing about the
effect of demographic characteristics on demand for business education, especially how they
interact with costs. Individual data, on the other hand, can tell us whether, for example, minority
students are more sensitive to costs than white students.
The research reported in this article takes advantage of a unique longitudinal survey of
GMAT registrants to develop a multivariate statistical analysis of a sequential set of decisions
made by individuals who contemplated attending business school. The first decision is a choice
among three enrollment alternatives: to pursue an MBA full-time, to pursue it part-time, or not to
pursue one at all. The second decision, conditional upon having enrolled, is whether to complete
the degree, to drop out, or to still be pursuing it at the time of the last interview. The
methodology used here, multivariate analysis of sequential decisions, follows that of a small
number of previous studies that examine college-level attendance and completion (e.g., Light and
Completion of a Management Degree, page 4
Strayer ; Light and Strayer ; Ganderton and Santos ; and Venti and Wise
). In particular, Light and Strayer  use data on individual students and their schools
to focus on the effect of race on college attendance and completion in a sequential choice model.
They find that, controlling for other variables, “minorities are more likely than whites to attend
college” and “more likely than observationally equivalent whites to graduate from college.”
However, they found that the unconditional likelihood of earning a degree is lower for blacks
than for whites. This finding emphasizes the importance of using a nested choice model to
examine attendance and completion of graduate school programs.
The GMAT registrant survey used in this study makes it possible to examine more
carefully the effect of both the explicit and implicit costs of obtaining an MBA degree on the
probability of attendance and completion than could any previous studies.
This study uses the GMAT Registrant Survey, sponsored by the Graduate Management
Admission Council (the organization that administers the GMAT). The Battelle Memorial
Institute in Seattle, Washington, designed the survey instruments and collected the data. The
survey contacted individuals who registered to take the GMAT on test dates between June 1990
and March 1991. The registrants were surveyed in three waves between 1991 and 1994. There
were 4333 registrants who completed all three waves of the survey. A large fraction of those
interviewed never pursued graduate business education; many never even bothered to take the
GMAT. This study relies on a subsample of 1988 GMAT registrants who applied to, and were
accepted by, at least one graduate management program. Table 1 shows the proportion of
sample members who applied to, attended, and completed business school respectively.
Data from the survey questionnaire were supplemented with information from the
Graduate Management Admissions Council=s own registration and test records. Information
Completion of a Management Degree, page 5
about sample members= undergraduate schools was obtained from Barron=s Profiles of American
Colleges (1992). Finally, data on the characteristics of graduate business programs were taken
from Barron=s Guide to Graduate Business Schools (Miller, 1994).
Outcomes of Business School Application & Attendance
4333 Wave 3 Respondents
Application Status Attendance Status Completion Status
Went Full-Time Still Enrolled
681 (34%) 144 (21%)
Applied & Accepted 215 (27%)
1988 (46%) Went Part-time Still Enrolled
806 (40%) 404 (50%)
Didn’t Apply/Not Accepted
ESTIMATI G THE COST OF ATTE DI G GRADUATE MA AGEME T SCHOOL
Opportunity Cost Estimates
The explicit cost of going to business school, the tuition, clearly might have a direct effect
on the decision to attend school; however, the theoretical effect of opportunity cost is slightly
more subtle. The key element in opportunity cost is the income foregone through lost earnings.
There are several ways in which going to graduate school reduces earnings. Most obviously, of
course, it is quite difficult to be a full-time student while trying to hold down a full-time job. But
Completion of a Management Degree, page 6
the majority of graduate management students do not attend full-time; many pursue the degree
while working a regular job. There are still opportunity costs, however. To the extent that more
time devoted to the job translates into promotion and pay raises, even part-time students may lag
behind coworkers who are unencumbered by classes, exams, and homework. Moreover, the
decision to take night classes can constrain the kind of day job available to the student.
Opportunity cost is difficult to measure. It cannot be observed directly, like salary or
income, because it involves an alternative that was not chosen, that is, a situation that the survey
respondent is not in. Therefore it is necessary to predict opportunity cost. This is done using the
standard econometric technique of estimating a multivariate regression equation that relates
earnings to individual characteristics such as education, work experience, race, gender, etc. This
study uses regression analysis to relate earnings to personal characteristics under each of the three
regimes: 1) a full-time student, 2) a part-time student, or 3) not in school. By estimating that
relationship, it is possible to predict what a given person would earn in the job market from
observing what similar people tend to earn.
Most studies that use predicted earnings have to estimate the wage equations with the
same sample to which the coefficients are applied. The current research is able to avoid this by
taking advantage of the 1986 MBA New Matriculants Survey conducted by the National Opinion
Research Center of the University of Chicago and (also) sponsored by the Graduate Management
Admission Council. Regressions from the New Matriculants Survey provide earnings
coefficients that are then applied to the individual characteristics of respondents to the GMAT
survey to predict earnings for GMAT registrants. The predicted earnings created by applying the
regression coefficients to the GMAT survey data represent our measures of opportunity cost.
Completion of a Management Degree, page 7
Table 2a reports the predicted opportunity cost for full-time and part-time study,
respectively. To attend graduate management school, a full-time student sacrifices more than
$18,000 per year (in 1994 dollars). This represents about 56% loss of income when compared to
what she could expect were she not in business school. Part-timers sacrifice much less, about
$3000 per year, or roughly 9% of what they would be earning if not a business school student.
Average opportunity cost varies considerably by sex. Full-time management school costs men
$21,239 in earnings versus $14,215 for women. Because women earn less in the job market,
however, their proportionate sacrifice is much closer to men’s: 51% of non-student earnings
compared with 56% for men. Comparing by race/ethnicity, whites sacrifice more, in absolute
dollars, than blacks, Hispanics or Asians, to attend full-time. In proportionate terms, however,
blacks have the highest opportunity cost of full-time attendance, but the lowest opportunity cost
of part-time attendance.
Tuition Cost Estimates
To estimate tuition and financial aid, we found tuition cost for each individual averaged
over the schools for which each was accepted for admission. The cost per credit hour was
estimated from figures reported in Barron’s Guide to Graduate Business Schools (Miller ).
We then multiplied the cost per credit hour times the minimum number of credits required for a
degree. For schools that reported only annual tuition, we multiplied annual tuition times the
number of years required to obtain a degree. Similarly, we used the average financial aid award
for each school to measure the effect of financial aid.
Table 2b shows the minimum tuition cost of obtaining a graduate management degree,
Completion of a Management Degree, page 8
and the average financial award. The
table shows that full-time costs are Variables Included in the Statistical Models
slightly lower on average than part- Variable Sample Mean
time costs: $10,838 versus $12,195.
This may seem odd, why are part- Black 16%
time costs higher on average than Hispanic 17%
full-time costs? One possible Number of Kids at Home 0.35
1-3 Years Work Experience 31%
explanation is that demand for full- 1
3-7 Years Work Experience 27%
> 7 Years Work Experience 1 23%
time slots is more elastic than
demand for part-time slots and BA in Humanities 8%
BA in Science 26%
therefore have a lower equilibrium BA in Social Science 15%
GMAT Score 508
price. (This is consistent with the
% of Sample from a Less 37%
Selective BA School
observation that part-time attendance Undergraduate GPA (4 point scale) 3.06
is by far the most popular choice.)
GMAT Cutoff > 650 2 11%
The table suggests that average GMAT Cutoff 600-650 2 25%
GMAT Cutoff 500-600 2 45%
tuition costs vary little by sex or Cost Measures
race/ethnicity. On the other hand, Tuition to Obtain Degree Full-Time 3 $10471
Tuition to Obtain Degree Full-Time 3 $11352
blacks appear to be accepted at Predicted Earnings If Not in School $34895
Before taking GMAT.
schools offering, on average, lower 2
These are for the management school attended, there
weren’t included in the attendance models.
financial aid awards. For those not attending this was measured for their first
Completion of a Management Degree, page 9
MODELS ESTIMATI G PREDICTED EFFECTS OF COST O ATTE DA CE A D
To obtain estimates of the effect of cost on the probability of attending graduate
management school and the probability of completing school, we estimated multinomial logit
models by maximum likelihood. The variables included in the models are presented in Table 3
below, along with the mean values for the sample.
Using these variables, we first estimated the effect of cost on the decision to attend
graduate management school and the effect on the probability of completing a graduate
management degree conditional upon having pursued one. We next combined those estimates to
predict the effect of cost on the unconditional probability of completing a management degree,
that is, on the probability of both entering a degree program and completing it. To do this we
estimated models of completion that were conditional upon attending school part-time or full-
time. The probabilities from these models were then multiplied by the probability of attending in
the relevant way: part-time or full-time. From the products of attendance and the conditional
completion probabilities, we could estimate the unconditional likelihood of a particular outcome
like completing or dropping out.
A simple schematic for this calculation is presented in Diagram 1. The boxes on the left-
hand side represent probabilities of attending part-time, full-time, or not at all, respectively.
Completion of a Management Degree, page 10
Probability of Conditional Probability Unconditional Probability
Attending of Completing of Completing
(Boxes of same shade are added together)
Didn’t Go Completed Full-Time Completed
Still in Full-Time Degree
Went Quit Full-Time Still Working
Full-Time Towards Degree
Not a Degree
Went Still in Part-Time
The boxes in the middle represent probabilities of completing conditional on the part-time or
full-time attendance. If we multiply the left and center probabilities, we can sum over outcomes
(boxes of the same shade) to get unconditional probabilities: the boxes on the right.
Effect of Cost on Probability of Attendance and Conditional Completion
Figures 1a and 1b show the predicted effect of changes in opportunity cost and tuition
cost on the probability of attending graduate management school, and of completing a degree
once enrolled.1 We consider two scenarios. In Scenario 1, we assume that non-student earnings,
Completion of a Management Degree, page 11
our measure of opportunity cost in the statistical model, drops by $20,000 per annum. In
Scenario 2, the tuition cost of obtaining a degree drops by $5000. Each of these values
represents roughly half of the average level of earnings or tuition, respectively, for the average
The top figure, (1a), shows the impact of these changes on attendance. The effect of the
change in opportunity cost is in the expected direction: lowering the implicit cost of attendance,
by lowering opportunity cost, tends to encourage a shift toward full-time attendance. The
disincentive to stay out of full-time school seems to be slightly reduced. The probability of not
going to school drops from about .09 to .07. There is a somewhat larger migration from part-
time to full-time attendance: the probability of going part-time drops from .55 to .48. The
combination of these influences is that reducing non-student earnings by $20,000 would increase
the proportion of full-time students from .37 to .45, a small but nontrivial effect.
The third set of columns in the top figure represents the effect of a reduction in explicit
cost, that is, a reduction in tuition cost. In contrast to other studies (e.g., Jantzen  on
business schools, and Ganderton and Santos  on college attendance decisions), these
results suggest a small impact. A $5000 dollar tuition reduction (about half the average cost)
appears to have minimal influence on the attendance decision.
The bottom figure, 1b, examines the effect of a reduction in cost on the probability of
completing a degree for those students who matriculated in a graduate management program
sometime during the survey period. Here the predicted effects of opportunity cost are somewhat
surprising. Note from the right hand side of the figure that lower opportunity cost tends to
increase the probability of dropping out. The probability of quitting the program rose from about
Completion of a Management Degree, page 12
.16 to about .23. This increase in dropouts appears to come at the expense of those who stay in,
but haven’t finished, rather than those who finish. In other words, lower job earnings tend to
convert some of those who take a relatively longer time to earn their degrees into dropouts. It
does not seem to affect those who get their degrees relatively more expeditiously.
This result may seem counterintuitive. One would expect those with relatively low
forgone earnings to be less likely to drop out, to be more willing to keep plugging along in
school, because they are sacrificing relatively less by being there. There are two potential
explanations. The first is that this result – that low earners are more likely to drop out – is an
artifact of the way we have measured completion. For our study, “completing a degree” means
completing it within the four years of our survey period. It is possible that our higher earners are
less likely to have dropped out yet, that is, by the time of the second follow up interview. If
higher earnings cause them to postpone entering management school, it could appear that those
with high opportunity cost drop out less frequently. In this case, among the lower earners we
would observe more students who have been in school long enough to know they do not want to
A second potential explanation for why low non-student earnings increase the drop-out
rate is that opportunity cost is likely to act as a screen. Some students are more motivated than
others to work hard and to complete management school. Presumably, those with lower
motivation are more likely to quit. But they are also less likely to enroll in the first place. It is
plausible that high opportunity cost keeps less motivated students out of business school and
therefore, indirectly, reduces the quit rate. For example, one can imagine some low earners,
especially those who are not employed, attending school because they have little else to do. Such
Completion of a Management Degree, page 13
low incentive may not sustain them through a rigorous graduate program, so they would more
likely drop out. This view is consistent with our finding that lower earnings do not increase the
proportion that completes the degree. Those who finish in a timely fashion are presumably more
highly motivated (especially those who attend full-time) and higher opportunity costs do not
tempt the highly motivated students to drop out of school.
In contrast to previous studies (e.g., Ehrenberg and Mavros  on graduate school
completion, and Light and Strayer  on college completion), Figure 1b suggests that tuition
appears to have little impact on the completion rate. This is consistent with the attendance
Effect of Cost on Unconditional Completion
Figure 1a presents the estimated effect of opportunity cost on the decision to attend
graduate management school; Figure 1b considers the effect of opportunity cost on the
probability of completing a graduate management degree conditional upon having pursued one.
In Figure 2, we combined the estimates in Figures 1a and 1b to predict the effect of opportunity
cost on the unconditional probability of completing a management degree, that is, on the
probability of both entering a degree program and completing it.
Lowering opportunity cost appears to substantially improve the likelihood of obtaining a
degree (in a given time period): from about .24 to .40. It also reduces the chances of still being
enrolled and increases the chances of dropping out.2 Given the results in Figure 1 we expect that
opportunity cost reduces the expected completion rate by shifting some potential full-timers into
part-time programs, and/or by causing some students to delay entry into school. Note also that
Completion of a Management Degree, page 14
with high opportunity cost, more people choose not to be candidates for the degree. This finding
is consistent with the expectation that opportunity cost discourages attendance, particularly
among less motivated students.
The Effect of Opportunity Cost by Sex, Race and Part-time/Full-time Status
Next, we examine whether and how the effect of opportunity cost differs among students
of different race, gender, or full-time/part-time status. Figure 3 gives the results for attendance
and Figure 4 for completion. Figure 3a compares the effect of opportunity cost for women and
men respectively. The differences between men and women appear to be very small: for both
sexes, lowering opportunity cost increases the proportion of full-timers, mainly at the expense of
part-timers. The effect for men is slightly larger. It is interesting to note that when we segregate
the sexes both models predict a larger effect of opportunity cost than did the joint model for the
two sexes. This is because the joint model imposes the restriction that all variables have the
same effect for men and women, an assumption that may be too restrictive.
Figure 3b looks at differences between two aggregate racial groups: 1) African Americans
and Latino/Latinas, and 2) Whites and Asian Americans.3 Between races, the effect of
opportunity cost has a similar pattern but the magnitude varies somewhat. For whites lower
opportunity cost shifts a larger fraction of students into full-time attendance: a proportion of
about .2 as opposed to .12 for blacks and Hispanics. Moreover, for blacks and Hispanics, lower
opportunity cost increases full-time attendance by pulling more people from the Not-Go category
and less from the Part-Time category than for whites. Apparently, higher opportunity cost tends
to discourage proportionately more minority candidates from pursuing graduate business
Completion of a Management Degree, page 15
education than white students.
Figure 4 presents the predictions for completion, differentiated by part-time/full-time
status (Figure 4a), by sex (Figure 4b) and by race (Figure 4c). Note that, as in Figure 1, these
completion rates are conditional upon having enrolled in a program. From Figure 4a, we see that
opportunity cost affects full and part-time students in somewhat different ways. In both cases,
opportunity cost moves a relatively small proportion of people out of the category Still-in-
School. Among full-timers these people tend to finish, among part-timers they tend to drop out.
This finding reinforces our earlier interpretation of the impact of opportunity cost on completion:
it acts partly as a screen. Surely less motivated students will disproportionately attend part-time,
so a screening effect should be more noticeable among part-timers. Indeed, among the full-
timers, higher opportunity cost has virtually no effect on the dropout rate.
As with the attendance models, the impact of opportunity cost on completion seems to be
noticeably different. For both racial groups, lowering opportunity cost increases drop out rates,
which differ very little between the sexes (Figure 4b). With race, however (Figure 4c), the
effects are proportionately greater for minorities. But for whites and Asians, opportunity cost
also tends to increase the probability of completing the degree versus still being in school. For
minorities the effect is reversed. Thus, for blacks and Hispanics, higher earnings tend to
discourage completion, partly by extending the length of study, but more significantly by
increasing the quit rate.
The Effect of Cost at the Top Tier Schools
Our final set of figures, Figure 5, looks at the influence of opportunity cost and tuition at
Completion of a Management Degree, page 16
top tier and second tier business programs. Schools are ranked on the basis of competitiveness of
admissions. We use a variable4 that reflects the minimum acceptable GMAT score for a person
to be admitted to the program in question. Using this GMAT-cutoff criterion, schools were
divided into four categories as indicated in Table 3. It is clear from the tables that schools that
are more selective tend to charge higher tuition indeed, substantially higher tuition. For the
average registrant, tuition at a top tier school (cutoff > 650) was nearly three times that at a
lowest tier school (cutoff < 500). Students at the top tier schools also faced higher opportunity
cost, presumably because the superior credentials and/or abilities required for acceptance by the
top business school are also rewarded in the job market.
Figure 5 shows predicted effects of tuition and opportunity cost on completion rates at the
top tier and second tier schools, respectively. Because elite schools are more expensive, in
Figure 5 we use a much larger decrement to tuition than was used in Figure 1: $20,000 compared
with the $5000 in the earlier figure. The figure shows that even this very large change in tuition
cost has a fairly small impact on the likelihood of finishing a degree at an elite school. Lower
tuition raises the completion rate at the top schools by only about .06. At second tier institutions,
the predicted completion rate actually falls slightly with lower tuition. The most noticeable
effect in Figure 5 is that of opportunity cost on completion at the highest ranked schools.
Decreasing opportunity cost raises the drop out rate very sharply, from .05 to .25. Most of the
extra dropouts are shifted over from completers (rather than those still enrolled.) This finding is
consistent with that reported in Figure 2. As above, we attribute this finding to the tendency of
earnings to screen out degree aspirants who are not highly motivated. The effect appears to be
even stronger at elite institutions where the rigor of the program might be particularly
Completion of a Management Degree, page 17
discouraging to less-motivated students.
CO CLUSIO S
This paper looked at the effect of tuition and opportunity cost, respectively, on the
decision to attend graduate management school and, for those who attended, on the probability of
completing a degree within a given time period. One important caveat should be stated regarding
these results. Our survey was directed only at people who registered to take the Graduate
Management Admission Test (GMAT). This limits our ability to observe the extent to which
tuition and/or opportunity cost discourage attendance in a broader population. Because they
registered for the GMAT, we can presume that our sample members all seriously considered
graduate management school. It is quite conceivable that other people were sufficiently
discouraged by opportunity cost or tuition that they never seriously contemplated business
school. Such people would never show up in our sample. Our results, therefore, could
understate the influence of cost on the decision to pursue a business degree.
The statistical models yielded several interesting results. First, tuition is not a major
impediment either to attending management school or to completing a degree. This is true even
at elite institutions where tuition costs are much higher than average.
A more substantial impediment to getting an MBA (within a fixed period of time) appears
to be the implicit cost of forgone earnings while in school. These opportunity costs seem to
operate in three ways to delay or forestall completion of a degree. First, opportunity cost seems
to have a small tendency to discourage attendance. Second, high opportunity cost shifts a modest
Completion of a Management Degree, page 18
proportion of students from full-time to part-time attendance, thus slowing down completion of
the degree within any given time frame. Third, high opportunity cost lowers rather than raises
the drop out rate among those who enroll in a degree program. It seems to do this by acting as a
screen: discouraging students with low motivation from entering programs. This screening effect
appears most pronounced at the elite institutions.
The impacts of tuition and opportunity cost do not vary greatly by gender. However,
higher opportunity cost tends to discourage proportionately more minority candidates from
pursuing graduate business education than white students. Also, the screening effect of
opportunity cost appears more pronounced among blacks and Hispanics. For these minority
students, higher earnings seem to discourage completion, partly by extending the length of study,
but more so by increasing the quit rate.
What are the implications of these findings for policy toward graduate management
education? One is that dollars invested in financial aid may less impact on student enrollment
and completion than was previously believed. A second is that schools need to recognize the
power of foregone earnings as an inhibitor of enrollment and completion. Schools would be
wise to find ways to mitigate the earnings loss for part-time students. More flexible class
schedule may be example of ways to accommodate the part-time student. Also, for an elite
school, it may be worth ensuring that the night (part-time) program is not a poor cousin of the
day (full-time program). Note also, that understanding the effect of opportunity cost is
especially important because it is likely the most talented students who have the highest potential
Completion of a Management Degree, page 19
Baker, Joe G. (1998). Gender, race, and Ph.D. completion in natural science and engineering.
Economics of Education Review,17, 179-188.
Ehrenberg, R. & Mavros, P. (1995). Do doctoral students financial support patterns affect their
times-to-degree and completion probability? Journal of Human Resources, 30.
Eide, E., Brewer, D.J., & Ehrenberg, R. (1998). Does it pay to attend an elite private college?
Evidence on the effects of undergraduate college quality on graduate school
attendance. Economics of Education Review, 17. 371-376.
Fox, M. (1992). Student debt and enrollment in graduate and professional school. Applied
Economics, 24. 669-677.
Ganderton, P. & Santos, R. (1995). Hispanic college attendance and completion : evidence from
the high school and beyond surveys. Economics of Education Review, 14. 35-46.
Jantzen, R.H. (2000). Price and quality effects on the demand for U.S. business programs.
International Advances in Economic Research, 6. 730-740.
Light, A. & Strayer, W. (2000). Determinants of college completion: school quality or student
ability? Journal of Human Resources, 35. 299-332.
Light, A. & Strayer, W. (2002). From Bakke to Hopwood: does race affect college attendance
and completion? The Review of Economics and Statistics, 84. 34-44.
Miller, Eugene, Barron=s Guide to Graduate Business Schools, (New York: Barron=s Educational
Series Inc.) 1994.
Montgomery, Mark (2002). A nested-logit model of choice of a graduate business school. The
Economics of Education Review, 21. 401-521.
Venti, S.F. & Wise, D.A. (1983). Individual attributes and self-selection of higher education :
college attendance versus college completion. Journal of Public Economics, 21.
Weiler, W.C. (1994). Expectations, undergraduate debt and the decision to attend graduate
school: a simultaneous model of student choice. Economics of Education Review,
Completion of a Management Degree, page 20
The Statistical Models of the Likelihood of Attendance and the Likelihood of Completion
The table below shows the mean values for variables included in the multinomial logit
analysis used to estimate the probabilities in Figures 1 through 5.
Variables Included in the Statistical Models
Variable Sample Mean
Number of Kids at Home 0.35
1-3 Years Work Experience 31%
3-7 Years Work Experience 27%
> 7 Years Work Experience 1 23%
BA in Humanities 8%
BA in Science 26%
BA in Social Science 15%
GMAT Score 508
% of Sample from a Less
Selective BA School
Undergraduate GPA (4 point scale) 3.06
GMAT Cutoff > 650 2 11%
GMAT Cutoff 600-650 2 25%
GMAT Cutoff 500-600 2 45%
Tuition to Obtain Degree Full-Time 3 $10471
Tuition to Obtain Degree Full-Time $11352
Predicted Earnings If Not in School $34895
Before taking GMAT.
These are for the management school attended, not included in the attendance models.
For those not attending this was measured for their first choice school.
Completion of a Management Degree, page 21
Opportunity cost is measured differently for these statistical models than in Table 2a. In the former, we
include only non-student earnings, because predicted part-time, full-time and non-student earnings are highly
correlated. However, experiments showed that different definitions of opportunity cost yielded similar results.
Note, however, that given how the unconditional probabilities are calculated, they may not sum to 1.
Many empirical studies suggest similar labor market behavior and results for African Americans and
Latino/as and for Anglo Whites and Asian Americans.
Batelle Memorial Institute developed this variable.