American Economic Journal: Applied Economics 2022, 14(2): 1–22
https://doi.org/10.1257/app.20200447
1
Will Studying Economics Make You Rich? A Regression
Discontinuity Analysis of the Returns to College Major
†
By Z B A M*
We investigate the wage return to studying economics by leverag-
ing a policy that prevented students with low introductory grades
from declaring a major. Students who barely met the grade point
average threshold to major in economics earned $22,000 (46per-
cent) higher annual early-career wages than they would have with
their second-choice majors. Access to the economics major shifts
students’ preferences toward business/nance careers, and about
half of the wage return is explained by economics majors working in
higher-paying industries. The causal return to majoring in econom-
ics is very similar to observational earnings differences in nationally
representative data. (JELA22, I26, J24, J31)
F
orty-year-old US workers with undergraduate degrees in economics earned
median wages of $90,000 in 2018. By comparison, those who had majored
in other social sciences earned median wages of $65,000, and college graduates
with any major other than economics earned $66,000. Relative to workers with
lower-wage majors, the observational premiums earned by workers with high-wage
majors like engineering, nursing, and economics are similar in size to the wage gap
between college graduates and nongraduates (Altonji, Blom, and Meghir 2012).
These gaps have motivated a large literature examining the determinants of stu-
dents’ major choices (Zafar 2013; Stange 2015; Arcidiacono, Aucejo, and Hotz
2016; Wiswall andZafar 2018; Patnaik etal. 2020). However, average wage differ-
ences between majors do not necessarily reect the causal effect of choosing one
* Bleemer: University of California, Berkeley (email: [email protected]); Mehta: University of California,
Santa Barbara (email: [email protected]). David Deming was coeditor for this article. Thanks to Joseph Altonji,
David Card, Carlos Dobkin, Laura Giuliano, Hilary Hoynes, Peter Kuhn, Enrico Moretti, Jesse Rothstein,
Christopher Walters, Matt Wiswall, Basit Zafar, and seminar participants at UC Berkeley for helpful comments; to
the UC Santa Cruz Ofce of the Registrar and the UC Berkeley Center for Studies in Higher Education for help in
obtaining the data used in this study; and to Alia Roca-Lezra and Dan Ma for excellent research assistance. This
study was granted exemption by UC Berkeley’s Ofce for Protection of Human Subjects. Bleemer was employed
by the University of California in a research capacity while conducting this study and acknowledges nancial
support from the National Academy of Education/Spencer Dissertation Fellowship and UC Berkeley’s Center
for Studies in Higher Education. Both authors hold undergraduate degrees in economics. See Bleemer andMehta
(2022) for the code and public data used in this study. Any errors that remain are our own.
†
Go to https://doi.org/10.1257/app.20200447 to visit the article page for additional materials and author
disclosure statement(s) or to comment in the online discussion forum.
2 AMERICAN ECONOMIC JOURNAL: APPLIED ECONOMICS APRIL 2022
major over another. This study directly analyzes the treatment effects of earning an
undergraduate degree in the popular high-earning eld of economics.
1
Estimating the causal effects of earning specic college majors is challenged
by students’ nonrandom assortment across majors: most students self-select their
college major, and many universities and departments use admissions and grade
requirements to restrict entry into certain majors. As a result, observational wage
differences across majors may reect selection bias. We overcome this challenge by
using a regression discontinuity (RD) design that exploits a fuzzy discontinuity in
economics major access at a large, moderately selective public university (Angrist
andLavy 1999).
2
We implement this design to estimate the effect of studying eco-
nomics on students’ early-career earnings and industries as well as how the major’s
effect on earnings is mediated by changes in students’ other educational outcomes,
career preferences, and early-career industries. We then characterize and estimate
the biases that arise when using observational average wage difference between
economics and other majors as a proxy for the treatment effect of majoring in
economics.
The specic case we analyze is the economics department at the University of
California, Santa Cruz (UCSC). UCSC Economics imposed a grade point average
(GPA) restriction policy in 2008: students with a GPA below 2.8 in Economics1
and 2 were generally prevented from declaring an economics major.
3
Students
who just met the GPA threshold were 36percentage points more likely to declare
the economics major than those who just failed to meet it. Most of these students
would have otherwise earned degrees in other social sciences. Students just above
the threshold who majored in economics were surprisingly representative of UCSC
economics majors on observables; for example, their average SAT score was at the
forty-rstpercentile of economics majors.
Comparing the major choices and average wages of above- and below-threshold
students shows that majoring in economics caused a $22,000 (46percent) increase
in the annual early-career wages of barely above-threshold students. It did so without
otherwise impacting their educational investment—as measured by course-adjusted
average grades and weekly hours spent studying—or outcomes like degree attain-
ment and graduate school enrollment. The effect is nearly identical for male and
female students, may be larger for underrepresented minority students, and appears
to grow as workers age (between ages 23 and 28). About half of the wage effect
can be explained by the effect of majoring in economics on students’ industry of
employment: relative to students who did not qualify for the major, economics
majors became more interested in business and nance careers and were more likely
to nd employment in higher-wage economics-related industries like nance, insur-
ance, and real estate (FIRE) and accounting. Most of the barely above-threshold
1
Economics is a particularly popular major at highly selective universities. The 2020 federal College Scorecard
shows that economics was the most-earned major at 11 of the top 20 highest-ranked American universities (as
ranked by US News and World Report) and was among the top 5 majors at 34 of the 50 highest-ranked universities.
2
This design was recommended (but not implemented) by both Altonji, Blom, andMeghir (2012) and Altonji,
Arcidiacono, andMaurel (2016).
3
Like many universities, UCSC has multiple “tracks” for its economics major. Students just above the GPA
threshold mostly chose its “business management economics” (BME) track, in which about one-third of required
courses are taken in business- and nance-related subdisciplines.
VOL. 14 NO. 2 3
BLEEMER AND MEHTA: WILL STUDYING ECONOMICS MAKE YOU RICH?
economics majors would have otherwise earned degrees in lower-earning elds like
psychology and sociology, and differences in either OLS-estimated average wages
by major (with or without controls) or median wages by major (estimated at the
university, state, or national level) slightly underestimate the estimated local average
treatment effect. This suggests that the net magnitude of selection bias and treatment
effect heterogeneity is small in this context.
4
Our data include comprehensive 2000–2014 UCSC student and course records
linked to biannual administrative student surveys, National Student Clearinghouse
(NSC) educational outcomes, and annual California unemployment insurance (UI)
employment records. These highly detailed records allow us to test several alterna-
tive explanations for above-threshold students’ higher postgraduate earnings. We
show that detailed student characteristics are smooth across the GPA threshold and
that grade distributions in economics courses remained unchanged in the period.
There is no evidence of students bunching above the threshold, as might be expected
if threshold-crossing was somehow manipulated. We also show that wages were
smooth across the grade threshold prior to the policy’s implementation but slightly
discontinuous during an interstitial period with a less binding major restriction
policy, generating similar (but noisier) instrumental variable estimates to the main
specication. While our main empirical strategy estimates linear RD models with
standard errors clustered by GPA (Lee andCard 2008), we conrm the estimates
using a number of other specications, including “honest RD” estimates following
Kolesár andRothe (2018).
5
A small number of previous studies have analyzed major-specic returns in
other countries by exploiting centralized eld-specic enrollment assignment rules
(Kirkeboen, Leuven, andMogstad 2016; Hastings, Neilson, andZimmerman 2014;
Daly andLeMaire 2021). However, the external validity of those estimates in the
United States may be limited: American universities offer a broader core liberal arts
curriculum, permit students to choose their majors years after their initial enrollment,
and provide students with more discretion over their courses, all of which could nar-
row eld-specic returns.
6
A large literature has employed selection-on-observables
methods and structural estimation to identify major-specic returns (James etal.
1989; Rumberger andThomas 1993; Black, Sanders, andTaylor 2003; Arcidiacono
2004; Hamermesh andDonald 2008), generally arguing that selection bias explains
a substantial portion of US wage variation across majors.
This study’s reduced-form RD design provides unusually transparent evidence
of postsecondary education’s heterogeneous and persistent role in shaping stu-
dents’ labor market outcomes. Our estimated early-career wage return to econom-
ics rivals the baseline return to a college degree, implying that major choice is a
4
Our results mirror the well-known nding that causal estimates of the return to schooling slightly exceed the
mean differences recovered from OLS (Angrist andKeueger 1991; Card 1999), with our study focusing on hetero-
geneity in the return to schooling.
5
Because of the small number (20) of discrete GPAs available to students, these latter estimates are likely
conservative.
6
The only known quasi-experimental study to previously identify heterogeneous returns by college major in the
United States is by Andrews, Imberman, andLovenheim (2017), who analyze the return to majoring in business by
exploiting a GPA threshold policy at several University of Texas campuses. Their suggestive nding of a large wage
return to business majors closely parallels our own estimates with regard to economics.
4 AMERICAN ECONOMIC JOURNAL: APPLIED ECONOMICS APRIL 2022
rst-order heterogeneity component in the return to higher education.
7
A related
literature has used quasi-experimental research designs to highlight university selec-
tivity as another important dimension of heterogeneous university treatment effects
(Hoekstra 2009; Zimmerman 2014; Cohodes andGoodman 2014; Bleemer 2021,
2022). However, even students who are quasi-randomly switched to enrolling at
universities with 25percentage points higher graduation rates—a large increase in
selectivity—receive an early-career wage return 30percent smaller than the return
to majoring in economics at UC Santa Cruz (Bleemer 2021).
8
These ndings imply
that widespread but understudied university policies that shape student major
choice—like GPA restrictions, variable tuition, and grade ination—have important
long-run efciency and social mobility ramications.
9
While prior studies have documented that students select majors partly based on
career preferences (Wiswall andZafar 2018), we present quasi-experimental evi-
dence that major choice causally affects students’ career preferences and industry of
employment. The correlation between college graduates’ majors and their occupa-
tions and industries of employment is notably weak: fewer than 60percent of most
majors’ students work in the top 10 highest-employment (5-digit) occupations for
that major (Altonji, Blom, andMeghir 2012).
10
Nevertheless, majoring in econom-
ics causes students to report a stronger preference for business and nance careers
prior to labor market entry—likely in part as a result of perceived job availability—
and to be more likely to ultimately work in related industries like FIRE and account-
ing. These changed industry preferences could reect the fact that knowledge and
skills acquired in the economics major may be particularly useful in these industries,
providing students with industry-specic human capital (Altonji, Kahn, andSpeer
2014; Kinsler andPavan 2015).
I. Background
The University of California, Santa Cruz is a moderately selective public research
university in northern California. In 2010, UCSC admitted 64percent of freshman
applicants, resulting in a 3, 290-student class largely split between White (38percent),
Asian (27percent), and Hispanic (24percent) students. Nearly all (98percent) of its
7
One reason for the economics major’s large return is the relatively low return to economics majors’
second-choice social science elds, highlighting the importance of counterfactual student choices in measuring
educational returns (Kirkeboen, Leuven, andMogstad 2016).
8
As in nearly all previous studies on the return to education and university selectivity, we are unable to distin-
guish whether the observed returns result from changes in human capital or signaling. We discuss this further in
SectionV. Other recent papers on heterogeneous university returns by university quality include Sekhri (2020) and
Canaan andMouganie (2018).
9
The close correspondence between observational and causal estimates of major-specic returns also suggests
the potential for private pecuniary gains resulting from providing students with locally relevant information about
average wages by majors, which has been shown to increase students’ enrollment in high-wage majors (Berger
1988; Beffy, Fougre, andMaurel 2012; Hastings, Neilson, andZimmerman 2015; Wiswall andZafar 2015). See
Bleemer andMehta (2021) on GPA restrictions, Andrews andStange (2019) on variable tuition, and Ahn etal.
(2019) on grade ination. Policies encouraging economics major choice (e.g., Porter andSerra 2020) are particu-
larly likely to provide students with substantial pecuniary returns.
10
A substantial academic literature studies how university policies shift students toward science and engineer-
ing majors (Sjoquist andWinters 2015; Denning andTurley 2017; Castleman, Long, andMabel 2018), though none
directly investigate whether this actually bolsters the STEM labor force.
VOL. 14 NO. 2 5
BLEEMER AND MEHTA: WILL STUDYING ECONOMICS MAKE YOU RICH?
students were California residents. In many ways, UCSC is relatively representative
of the average US university; among four-year US universities in the 2010 Integrated
Postsecondary Education Data System database (weighted by enrollment), UCSC is
at the forty-second percentile in admissions rate, the fty-ninth percentile in average
student SAT scores, the forty-second percentile in middle-income students’ average
net price of attendance, and the fty-third percentile in student-to-faculty ratio.
11
The
UCSC Department of Economics had 25 ladder-rank faculty and 7 lecturers in 2010
and taught 8,800 student enrollments that academic year, implying that each faculty
member taught an average of 91 students per quarter, among the highest loads at the
university.
12
The UCSC Economics Department’s 2003 GPA restriction was the university’s
rst policy limiting enrolled students’ access to a particular college major (Bleemer
andMehta 2021). The restriction was rst recorded in UCSC’s 2003 course catalog,
which stated that students with a GPA in Economics1 and 2 (EGPA) below 2.8
would only be allowed to declare the major “at the discretion of the department.”
If students retook one of the courses, only the initial grade was used to calculate
EGPA. This policy hardly changed de jure over the following ten years, though
the 2012 course catalog is the rst to note that for students with below-2.8 EGPAs,
“appeals are rarely granted.” Starting in 2013, calculus grades were added to the
EGPA calculation.
However, the department’s “discretion” left substantial room for year-over-year
de facto differences in below-2.8 students’ access to the major.
13
The difference
in the probability of majoring in economics above and below the EGPA threshold
remained small (below 15percentage points) until the 2008 entering cohort and then
ranged from 25 to 60percentage points until 2012.
14
As a result, this study focuses
on these latter ve cohorts of freshman UCSC students.
II. Data
The student database analyzed in this study ( University of California ClioMetric
History Project (UC-CHP) 2020) was collected from the UCSC Ofce of the
Registrar as part of the UC ClioMetric History Project (Bleemer 2018). The sample
covers all freshman-admit students who rst enrolled at UCSC between 1999 and
2014.
15
For each student, we observe gender, ethnicity, cohort year, ( pre-enrollment)
home address, California residency status, high school, and SAT score as well as
11
Calculations from the Integrated Postsecondary Education Data System. Average SAT calculated as the
summed averages of the twenty-fth and seventy-fth percentiles of each SAT test component. Average net price
dened over federal nancial aid recipients with family incomes between $48,000 and $75,000.
12
Altonji andZimmerman (2019) show that economics and business degrees have below-average educational
costs.
13
Online Appendix Figure A-1 shows 2000–2014 UCSC students’ likelihood of majoring in economics
by EGPA for each cohort.
14
This change was likely driven by increased demand after the 2007–2008 nancial crisis; see online Appendix
Figure A-2.
15
Community college transfer students are omitted from our analysis because they followed a different admis-
sion rule into the economics major.
6 AMERICAN ECONOMIC JOURNAL: APPLIED ECONOMICS APRIL 2022
UCSC course enrollments and grades.
16
The EGPA running variable is calculated
by averaging students’ GPAs in Economics1 and2, using their earliest letter grades
if they retook either course.
These student records are linked by name and birth date to the NSC StudentTracker
database (NSC 2019), which contains undergraduate and graduate enrollment and
degree attainment records for nearly all American colleges and universities, and
by social security number to employment records from the California Employment
Development Department (EDD 2019), which include annual wages and six-digit
North American Industry Classication System (NAICS) industry code.
17
We proxy
family income by the mean adjusted gross income in the student’s home zip code in
their rst year of enrollment (IRS 2018).
18
UCSC students are also linked to survey responses from the biannual UC
Undergraduate Experience Survey (UCUES), conducted online in the spring of
even-numbered years (Student Experience in the Research University (SERU)
2019). The second/third and third/fourth year response rates among the 2008–2012
students in the main sample were 29 and 28percent, with the response rates and
respondent characteristics smooth across the GPA threshold.
19
Among the survey’s
many questions are responses about number of hours per week spent studying and
students’ intended careers.
20
Non-economics majors are categorized into four disciplines: humanities,
social sciences, natural sciences, and engineering. Combining the three tracks
of the economics major—economics, BME, and global economics—it was the
second-most-popular major at UCSC for the 2008–2012 cohorts (11.7percent of
students), below psychology (12.9 percent) but ahead of environmental studies
(6.1percent) and sociology (6.0percent).
Table1 presents descriptive statistics for 2008–2012 UCSC freshman-admit stu-
dents. Relative to the full sample of 15,400 UCSC students, the 3,053 students who
complete Economics1 and2 are more likely to be male and Asian and come from
slightly higher-income neighborhoods. Of those students, the 55percent who actu-
ally declare the economics major are 41percent female (compared to 56percent
across UCSC) and 44 percent Asian (compared to 27 percent) and have similar
average SAT scores to the average UCSC student (1716 out of 2400).
16
ACT test scores (submitted by 4percent of applicants instead of SAT scores) and SAT scores on a 1600 point
basis are converted to 2400-point SAT scores using standard concordance tables.
17
NSC match quality is near complete but missing for some students who opt out of coverage. For example,
97 percent of UCSC undergraduate degrees awarded to the 2008–2012 cohorts appear in NSC (see AppendixC
of Bleemer 2021). EDD NAICS code reects the industry of employment from the year’s latest nonmissing quar-
ter (US Census Bureau 2017). EDD employment records exclude out-of-state, federal, and self-employment. All
EDD-related analysis was originally conducted for the purpose of institutional research (see Bleemer andMehta
2020).
18
Income statistics are from the IRS Statistics of Income (SOI). Wage and income statistics are winsorized at
the top and bottom 2percent and CPI ination adjusted to 2019 (BLS 2019).
19
See online Appendix Figure A-3. UCUES data were provided by the SERU Consortium at UC Berkeley’s
Center for Studies in Higher Education and linked by student ID.
20
Full questions and responses are provided in the survey Appendix.
VOL. 14 NO. 2 7
BLEEMER AND MEHTA: WILL STUDYING ECONOMICS MAKE YOU RICH?
III. Empirical Design
We identify the relationship between economics major choice (the treatment) and
resulting outcomes ( Y ) by exploiting a discrete fuzzy grade discontinuity in economics
major access (Hahn, Todd, andvander Klaauw 2001). Figure1 shows the rst-stage
estimate of the impact of meeting the 2.8 GPA threshold on economics major choice
for the 2008–2012 cohorts. Above-threshold students were about 36 percentage
points more likely to declare the economics major. Some below-threshold students
were nevertheless able to declare the major—“at the discretion of the department”—
and about 20percent of above-threshold students chose not to declare the major.
Each bubble is scaled by the proportion of students who earned that EGPA ; because
the EGPA is calculated over only 2 letter grades, students could earn only 14 com-
mon or 6 uncommon EGPA s.
Let Y
i
(1) denote the outcome that UCSC student i would experience if they majored
in economics, and let Y
i
(0) denote the outcome they would experience if they did
not. Outcomes of interest include (for example) postgraduation earnings, industry of
employment, study time, and graduate school attendance. Let C be the group of policy
compliers: the subset of students who major in economics if they are above the GPA
threshold but do not if they are below it. The effect of the major on policy compliers
whose EGPA was near the threshold (the local average treatment effect) is given as
(1) LAT E
RD
(
Y
)
≡ lim
EGPA
↓
2.8
E
[
Y
i
(
1
)
| EGPA, i ∈ C
]
− lim
EGPA
↑
2.8
E
[
Y
i
(
0
)
| EGPA, i ∈ C
]
so long as E
[
Y
i
(
1
)
| EGPA, i ∈ C
]
and E
[
Y
i
(
0
)
| EGPA, i ∈ C
]
are smooth at
EGPA = 2.8 .
We test several implications of this smoothness assumption. First, we nd that the
empirical grade distribution does not spike at or near the 2.8 EGPA threshold and
T1—D S 2008–2012 UCSC E C
Freshman Econ 1 and 2 Economics Near-threshold
students enrollees majors economics majors
(SE)
Female (percent)
55.7 41.3 40.9 35.6
(7.3)
White (percent)
40.8 32.4 32.8 27.9
(6.5)
Asian (percent)
26.5 41.4 43.7 41.1
(8.1)
Hispanic (percent)
24.3 19.2 16.7 18.3
(7.1)
Black (percent)
2.9 1.9 1.7 6.2
(1.8)
CA resident (percent)
97.1 97.4 97.2 99.7
(2.5)
SAT score (2400 scale)
1720 1697 1716 1667
(14)
Mean zip code inc. ($)
92,060 95,819 99,477 86,770
(7,309)
Number of students 15,423 3,053 1,689
Notes: This table presents mean demographic and socioeconomic statistics for 2008–2012 UCSC freshman-admit
students, those who take Economics1 and Economics2, and those who then declare the economics major. The
nal columns present the average characteristics of the students who majored in economics because of their barely
above-threshold EGPA s, estimated following equation(1) by treating the interaction between each characteristic
and economics major indicator as the outcome (Abadie 2002). Mean zip code income measures the mean adjusted
gross income of tax lers in the student’s home zip code in the year they graduated high school.
Source: UC-CHP student database and IRS SOI
8 AMERICAN ECONOMIC JOURNAL: APPLIED ECONOMICS APRIL 2022
that the 2008–2012 distribution is highly similar to the 2003–2007 grade distribu-
tion, years when the EGPA threshold was loosely enforced.
21
This pattern implies
that students did not manipulate their course grades to meet the GPA threshold.
Second, we nd that detailed student socioeconomic characteristics are smooth
across the GPA threshold, as is a one-dimensional summary of student character-
istics generated by exibly predicting each student’s 2017–2018 average wages by
socioeconomic observables. This indicates that effects estimated across the thresh-
old are unlikely to be driven by anything other than qualication for the major.
22
Finally, as a placebo test, we nd that economics major selection and early-career
wages are smooth across the 2.8 EGPA threshold in 2000-2002, before the GPA
restriction was introduced.
23
21
See online Appendix Figure A-4. Both distributions share the same shape as the 2000–2002 grade distribution
(prior to the EGPA restriction’s implementation), though average EGPA s trended downward over time. Students’
Economics2 grades are smooth across the threshold.
22
See online Appendix Figure A-5. Predicted wages are estimated by OLS on the 2017–2018 wages of
2008–2012 UCSC students who did not complete Economics1 and2. Predicted wages are imputed only for stu-
dents with observed 2017–2018 wages to match our main labor market estimation sample.
23
See online Appendix Figure A-6. We also exploit the small increase in economics major choice across the
less binding 2003–2007 GPA threshold to noisily replicate the instrumental variable wage results in the main
F1. T E UCSC E GPA T M E
Notes: Each circle represents the percent of economics majors (y-axis) among 2008–2012 UCSC students who
earned a given EGPA in Economics1 and2 (x-axis). The size of each circle corresponds to the proportion of stu-
dents who earned that EGPA . EGPA s below 1.8 are omitted, leaving 2,839 students in the sample. Fit lines and
beta estimate (at the 2.8 GPA threshold) from linear RD specication; standard error (clustered by EGPA ) in
parentheses.
Source: UC-CHP student database
β = 36.1 (2.7)
2.0 2.5 3.0 3.5 4.0
Average GPA in Economics 1 and 2
100
0
20
40
60
80
Percent in major
VOL. 14 NO. 2 9
BLEEMER AND MEHTA: WILL STUDYING ECONOMICS MAKE YOU RICH?
Our baseline specication for estimating equation (1) is linear in the running
variable ( EGPA ) on either side of the threshold and clusters standard errors by the
20 observed EGPA s above 1.8 (Lee andCard 2008). We also check that our results
are robust to using a number of alternative specications. These include (i)allowing
quadratic running variable terms, (ii)adding demographic controls and high school
xed effects, (iii)narrowing the bandwidth to 0.5 EGPA points on either side of the
threshold, and (iv)estimating “honest” local linear RD coefcients with optimal
bandwidth and triangular kernel following Kolesár andRothe (2018).
24
We note
below the rare occasions in which any of the alternative specications result in coef-
cients that differ substantially or statistically from those presented in the gures.
25
The last columns of Table1 present estimated characteristics of the students who
majored in economics as a result of their barely above-threshold EGPA s (estimated
following Abadie 2002). These students’ observable characteristics are surprisingly
similar to those of the average UCSC economics student: 36percent are female,
41percent are Asian, and essentially all of them are California residents. Despite
their low introductory course grades, there is no indication that they were much less
prepared for success than other economics majors: their mean SAT score is at the
forty-rstpercentile of all economics majors, while the mean income of their zip
codes of residence is at the forty-eighth percentile of their economics peers.
26
The
representativeness on observables of our above-threshold policy compliers suggests
that our estimated local average treatment effects may be similar to the average
treatment effect of majoring in economics at UCSC.
IV. Baseline Return to the Economics Major
Figure 2 shows that 2008–2012 UCSC students with above-threshold EGPA s
had far higher early-career wages than their below-threshold peers.
27
Measuring
average California wages in 2017 and 2018—when students in the sample were
23 to 28 years old— above-threshold students earned about $8,000 higher wages
than below-threshold students, with a standard error of $1,900.
28
Given that they
were also 36percentage points more likely to major in economics, the IV estimator
suggests that students who just met the GPA threshold earned higher early-career
wages by about $22,000 if they declared the economics major, rising from $37,000
to over $59,000. Measuring wages in log dollars provides a similar 0.58 log dollar
specication below ( rst-stage 6.2percentage points (2.9SE), IV $32,500 ($19,600)).
24
The small number of running variable values suggests that these last estimates will be conservative. Online
Appendix Tables A-1 to A-4 present regression coefcients from these alternative specications for all main results.
25
All OLS and IV regressions are estimated using the felm function in the lfe R package, version 2. 8-5. Honest
local linear regressions are estimated by the RDHonest R package, version 0.3.2.
26
This absence of signicant positive selection may result from the substantial noise in introductory course
grades, which reect a host of professor, teaching assistant, and extracurricular determinants (e.g., Sacerdote
2001; Fairlie, Hoffmann, andOreopoulos 2014). A linear regression of EGPA on high school xed effects and
gender-ethnicity indicators interacted with SAT score, mean zip code GPA, and cohort provides an adjusted R
2
of
only 0.15.
27
Impacted students mostly graduated between 2012 and 2016, so their early-career earnings and industries
were not shaped by a postgraduate recession (Altonji, Kahn, andSpeer 2016).
28
Students with earnings in only one of the two averaged years are assigned their observed year’s wages; stu-
dents with no observed wages in either year are dropped. Some RD specications provide somewhat larger wage
return estimates.
10 AMERICAN ECONOMIC JOURNAL: APPLIED ECONOMICS APRIL 2022
estimated treatment effect, though that estimate is statistically noisy in the Kolesár
andRothe (2018) specication.
The estimated returns to majoring in economics are nearly identical when
estimated separately by student gender: $21,700 (SE $8,800) for men, $22,600
($5,700) for women. The unexpectedly high observed earnings of students with
EGPA = 2.35 visible in Figure2 obtains only for male students, driving those esti-
mates’ higher standard errors. The return is also similar in magnitude among under-
represented minority (Black, Hispanic, and Native American) students: $27,600
($13,500).
29
29
See online Appendix Figure A-7. California wages are observed for 80–90percent of the sample, likely the
result of nearly all UCSC freshman students being California residents. There is some evidence that students’ like-
lihood of 2017–2018 California employment rises at the GPA threshold, though the estimates are not robust across
different specications; see online Appendix Figure A-9.
F2. T E UCSC E GPA T A W
Notes: Each circle represents the mean 2017–2018 wages (y-axis) among 2008–2012 UCSC students who earned
a given EGPA in Economics1 and2 (x-axis). The size of each circle corresponds to the proportion of students who
earned that EGPA . 2017–2018 wages are the mean EDD-covered California wages in those years, omitting zeroes.
Wages are CPI adjusted to 2018 and winsorized at 2percent above and below. EGPA s below 1.8 are omitted, leav-
ing 2,446 students with observed wages. Fit lines and beta estimate (at the 2.8 GPA threshold) from linear RD spec-
ication and instrumental variable specication (with majoring in economics as the endogenous variable); standard
errors (clustered by EGPA ) in parentheses.
Sources: UC-CHP student database and CA Employment Development Department
2.0 2.5 3.0 3.5 4.0
Average GPA in Economics 1 and 2
70,000
40,000
50,000
60,000
Dollars
β = 7,989 (1,885)
IV = 22,123 (5,847)
VOL. 14 NO. 2 11
BLEEMER AND MEHTA: WILL STUDYING ECONOMICS MAKE YOU RICH?
These estimates do not appear to be solely driven by college graduates’ rst
employment after graduation. Figure3 presents estimates of the annual wage return
to majoring in economics four to nine years after graduating high school for three par-
titions of our baseline sample: the 2008–2009 cohorts, 2010 cohort, and 2011–2012
cohorts. It shows suggestive evidence that the wage return grows larger as workers
age from 23 to 28, though the small number of cohorts challenges separate identi-
cation of age and cohort effects. Online Appendix Figure A-8 contextualizes this
nding by using American Community Survey (ACS) wage data (Ruggles etal.
2020) to visualize the median wages of US economics majors annually from ages
22 to 62 along with the weighted median wages of US college graduates who earned
the second-choice majors that UCSC’s policy-complying economics majors would
have earned if economics had been unavailable (discussed further below). The rel-
ative observational return to economics increases with age in workers’ twenties
and thirties and remains large throughout workers’ careers, resulting in a $536,000
observational net present value of majoring in economics.
30
30
The observational wage return to economics shrinks (though remains large) after age 50, possibly reecting
informational obsolescence (Deming andNoray 2020).
F3. E W R E M A
Notes: This gure shows RD instrumental variable β estimates at the 2.8 GPA threshold of the effect of majoring in
economics on earnings in each of 4–9 years after high school graduation, splitting the sample into the 2008–2009,
2010, and 2011–2012 UCSC incoming-class cohorts. The bars show 95percent condence intervals from standard
errors clustered by EGPA . The black line shows the difference between the national median wages of economics
majors and those of college graduates with majors in barely above-threshold UCSC students’ second-choice majors,
as measured in the ACS; see online Appendix Figure A-8. Wages are CPI adjusted to 2018 and winsorized at 2per-
cent above and below.
Sources: UC-CHP student database, CA Employment Development Department, and ACS (Ruggles etal. 2020)
Years since high school graduation
Dollars
−20,000
20,000
60,000
RDIV est., 2008–2009 cohorts
RDIV est., 2010 cohort
RDIV est., 2011–2012 cohorts
Diff. in ACS medians
4 6 8 10 12 14
12 AMERICAN ECONOMIC JOURNAL: APPLIED ECONOMICS APRIL 2022
V. Why do Economics Majors Earn Higher Salaries?
A. Educational Performance, Resources, and Attainment
Figure4 shows how the characteristics of UCSC students’ postsecondary educa-
tions differed as a result of being provided access to the economics major. PanelsA
andB show that access to the economics major does not change students’ likelihood
of earning a college degree or enrolling in a graduate degree program (within seven
F4. T E E M A E A
Note: Each circle represents the mean educational outcome (y-axis) among 2008–2012 UCSC students who earned
a given EGPA in Economics1 and2 (x-axis). The size of each circle corresponds to the proportion of students who
earned that EGPA . Undergraduate degree attainment is measured in 2018. Graduate school enrollment indicates
enrollment at a four-year university after undergraduate degree attainment within seven years of UCSC matricula-
tion. Course-adjusted college GPA is calculated as the mean of the differences between students’ grades and each
course’s xed effect from a two-way student-course xed effect model (see online Appendix FigureA-10). EGPA s
below 1.8 are omitted, leaving 2,839 students in the sample. Fit lines and beta estimate (at the 2.8 GPA threshold)
from linear RD specication and instrumental variable specication (with majoring in economics as the endoge-
nous variable); standard error (clustered by EGPA ) in parentheses.
Sources: UC-CHP student database and NSC
75
80
85
90
95
100
Panel A. Degree attainment
Percent
Adjusted GPA
Percent
10
5
0
15
20
25
Panel B. Grad. school enrollment
2.8
3.0
3.2
3.4
3.6
3.8
4.0
Panel C. Course-adjusted GPA
Average GPA in Economics 1 and 2Average GPA in Economics 1 and 2
Average GPA in Economics 1 and 2
2.0 2.5 3.0 3.5 4.0
2.0 2.5 3.0 3.5 4.0
2.0 2.5 3.0 3.5 4.0
β = −0.4 (1. 5 )
IV = −1.0 (4.2)
β = −0.03 (0.02)
IV = −0.08 (0.06)
β = −2.3 (2.2)
IV = −5.9 (6.0)
VOL. 14 NO. 2 13
BLEEMER AND MEHTA: WILL STUDYING ECONOMICS MAKE YOU RICH?
years of matriculating).
31
Above-threshold students also have similar time to degree
as below-threshold students. Economics major access does not provide students
with smaller class sizes; if anything, average class sizes grow larger.
32
It does not
lead students to earn higher or lower grades when adjusted for course difculty
(panel C), nor does it change the weekly amount of time students report studying
outside of class.
33
Instead, the primary estimable difference in students’ postsecondary educations is
the content of that education. Barely above-threshold economics majors completed
13 more economics courses than nonmajors, for a total of 17 economics courses
on average. This caused the economics majors to take nine fewer courses in other
social sciences and about four fewer courses across other disciplines. About seven
of the additional economics courses were in traditional economics subdisciplines,
while almost six were in subdisciplines related to business, nance, and accounting
also offered by UCSC’s economics department. Access to the economics major did
not change the number of mathematics and statistics courses that students com-
pleted, but they did complete an average of two additional courses in quantitative
methodology.
34
If there was no signal value of economics degree attainment, then these estimates
would imply a wage elasticity of economics course taking of about 0.3.
35
However,
this estimate is likely upwardly biased by the potentially high signal value of eco-
nomics degrees relative to students’ second-choice majors. We are unable to directly
distinguish between the degree’s signal value and the value of additional human
capital accumulation in this setting.
36
B. Employment by Industry
Majoring in economics causally impacts the industries in which students are
employed in their early careers. This could reect either industry-specic human
capital formation or changes in students’ preferences across industries. PanelA
of Figure5 suggests that part of the effect arises from student preferences: survey
31
Near-threshold students had a 96percent bachelor’s attainment rate—including degrees earned at other insti-
tutions by 2018—compared to 94percent across the 2008–2012 UCSC freshman cohorts.
32
For plots showing estimates for additional educational outcomes like time to degree and class size, see online
Appendix Figure A-10.
33
Above-threshold students earn slightly lower unadjusted GPAs than below-threshold students as a result of rela-
tively lower grading standards in UCSC’s economics department; see online Appendix FigureA-10.
34
Quantitative methodology courses include any course that mentions “statistics,” “econometrics,” “psycho-
metrics” or “quantitative/math/research/information methods” in its title. See online Appendix Figures A-11 and
A-12.
35
Arteaga (2018) nds that in the setting of a Colombian university, a policy change that resulted in a 15per-
cent reduction in course taking among economics majors caused a 16 percent decline in students’ early-career
wages, implying a unit wage elasticity of economics course taking. It is unsurprising that we estimate a lower elas-
ticity given that (i) below-threshold UCSC students excluded from the economics major took other courses instead
of economics courses, whereas the Colombian students graduated having completed fewer aggregate courses, and
(ii) below-threshold UCSC students earned a different college major instead, which could change the signal value
of their degree.
36
One potential strategy to directly estimate the signal value of UCSC’s economics degree would be to compare
the wages of economics majors and nonmajors who took comparable numbers of economics courses. Unfortunately,
as at many US public universities, many UCSC economics courses were formally or informally restricted to eco-
nomics majors. Online Appendix FigureA-13 shows that there is essentially no overlap between the distribution
of economics courses completed by 2008-2012 UCSC economics majors and nonmajors, thwarting that design.
14 AMERICAN ECONOMIC JOURNAL: APPLIED ECONOMICS APRIL 2022
F5. E E M A I P E
Notes: Each circle represents the mean outcome measure (y-axis) among 2008–2012 UCSC students who earned a
given EGPA in Economics1 and2 (x-axis). The size of each circle corresponds to the proportion of students who
earned that EGPA . Intended career in business/nance indicates selecting “Business, nance-related professions”
on a survey asking “Career hope to eventually have after education complete” (see the online survey Appendix)
among the 834 in-sample second- and third-year UCUES respondents. Employment in FIRE and accounting indi-
cates 2017 or 2018 employment in the FIRE (NAICS codes 52 and 531) or accounting (541211) industries; see
online Appendix FigureA-5. Imputed wages by industry ( six-digit NAICS) are calculated as the mean 2017–2018
wages of all 2008–2012 freshman-admit UCSC students. Imputed wages are CPI adjusted to 2018 and winsorized
at 2percent above and below. Fit lines and beta estimate (at the 2.8 GPA threshold) from linear RD specications
and instrumental variable specications (with majoring in economics as the endogenous variable); standard error
(clustered by EGPA ) in parentheses. Six 2012 sophomore respondents were omitted from estimation; see online
Appendix FigureA-14.
Sources: UC-CHP student database, SERU database, and CA Employment Development Department
2.0 2.5 3.0 3.5 4.0
2.0 2.5 3.0 3.5 4.0
Panel A. Intend career in bus.
/
n.
Panel C. Imputed wages by industry
Average GPA in Economics 1 and 2Average GPA in Economics 1 and 2
Average GPA in Economics 1 and 2
Panel B. Emp. in FIRE or accounting
Percent
Percent
30,000
40,000
50,000
60,000
100
80
60
40
20
0
30
25
20
15
5
10
0
Dollars
β = 3,642 (1,196)
IV = 10,220 (3,422)
β = 16.1 (6.9)
IV = 51.8 (23)
2.0 2.5 3.0 3.5 4.0
β = 9.1 (2.3)
IV = 25.1 (5.8)
VOL. 14 NO. 2 15
BLEEMER AND MEHTA: WILL STUDYING ECONOMICS MAKE YOU RICH?
responses from students’ sophomore and junior spring quarters (prior to labor
market entry) show that barely above-threshold economics majors became more
than 50percentage points more likely to report an interest in a business or nance
career than nonmajors, though this could in part reect increased employment
opportunity in those industries.
37
Panel B shows that economics major access
increases students’ early-career likelihood of working in the most impacted FIRE and
accounting industries by 25percentage points, split two-thirds/ one-third between
the two. Economics majors became 17percentage points less likely to work in the
education, health-care, and social assistance industries in 2017–2018.
38
PanelC of Figure5 shows the effect of majoring in economics on the average
wages earned in students’ industries of employment. Industries are dened by
six-digit NAICS codes, and industry mean wages are measured using the 2017–2018
wages of all 2008–2012 UCSC students. Barely above-threshold economics majors
work in industries with higher mean wages by about $10,000, implying that just
under half of the $22,000 wage return to majoring in economics can be explained by
economics majors working in higher-paying industries.
39
VI. Average Wage-by-Major Statistics
Differences in the average wages earned by college graduates with different
majors are often presented as useful for students’ major selection (Carnevale,
Cheah, and Hanson 2015; US Department of Education 2019), but they could
be misleading as a result of self-selection into majors. To examine this concern
empirically, this sectioncompares the causal return to majoring in economics at
UCSC to observational differences in wages by major estimated using data from
various reference populations (e.g., all UCSC graduates or college graduates in
California).
Denote the average wage of college graduates in reference population R who
completed major m by w
̃
m
R
. Among students at UCSC who have taken Econ1 and2,
let m
i
be student i ’s chosen major, w
i
(m) be the latent wages they would have earned
if they had selected major m , and w
i
= w
i
( m
i
) be their observed wage given that
they chose m
i
. The variable T is the treatment major (economics). Let P
m
0
be the
probability of choosing non-economics major m for the barely below-threshold stu-
dents who would have earned economics majors if their EGPA s had been slightly
higher (that is, below-threshold policy compliers), P
m
R
be the probability of a
37
First-year career-intention survey responses (prior to majoring in economics) are smooth across the threshold.
We examine sophomore and junior responses because those students have (likely) already declared the economics
major but have not yet been hired into postgraduate employment. Six 2012 sophomore respondents—economics
majors with 2.7 EGPA s—are omitted from estimation as outliers; see online Appendix FigureA-14.
38
See online Appendix Table A-5, which shows estimated changes for each two-digit NAICS code.
Accounting—in which UCSC Economics offers several courses—is the most-impacted six-digit NAICS code out-
side of FIRE industries.
39
This conclusion is supported by a $15,400 estimated IV wage coefcient in the presence of 6-digit-NAICS
industry xed effects, though that estimate is statistically noisy (SE $8,000). If industries are partitioned into just 3
groups—FIRE, accounting, and all other industries combined—the 2 can explain only a $2,300 (IV) wage increase
at the threshold. Mean industry wages calculated using earlier UCSC cohorts and 2009–2010 wages provide nearly
identical estimates, suggesting that this information could have been partly known by students. NAICS codes with
fewer than ten observed workers are omitted.
16 AMERICAN ECONOMIC JOURNAL: APPLIED ECONOMICS APRIL 2022
student in R selecting m conditional on not selecting economics, and w
–
m
0
and w
–
m
1
be the expected latent wages in major m of UCSC policy compliers just below and
above the GPA threshold. We can then estimate equation(1) in our sample of UCSC
Econ1 and 2 takers either using each student’s observed wage as the dependent
variable or replacing it with the w
̃
m
R
of their chosen major. These regressions yield
estimates, respectively, of
(2) LAT E
RD
(
w
)
= w
–
T
1
−
∑
m≠T
P
m
0
w
–
m
0
,
(3) LAT E
RD
(
w
̃
m
R
)
= w
̃
T
R
−
∑
m≠T
P
m
0
w
̃
m
R
.
These equations show that wage-by-major statistics from R can be used to pre-
dict the treatment effect of earning an economics major for barely above-threshold
UCSC students if they are similar to policy compliers’ latent wages by major near
the GPA threshold.
Figure6 shows the average early-career wages by major for barely above-threshold
economics majors’ 10 most common second-choice majors—led by psychology
(20percent), environmental studies (14percent), and “technology and information
management” (12percent)—and for UCSC’s 3 economics tracks.
40
Average wages
by major ( w
̃
m
R
) are calculated in three ways: by linear regression of UCSC students’
early-career wages on major dummies with and without detailed student controls
and by the median wages of all early-career college graduates in California.
41
The
gure also shows estimates of LAT E
RD
( w
̃
m
R
) for each set of average wage statistics
as the difference between two dashed horizontal lines. These are estimates of equa-
tion(3), which implicitly weights the average wage in each counterfactual major by
the likelihood that a below-threshold policy complier would select it. They are jux-
taposed at the far right with the causally identied return to majoring in economics,
as estimated following equation (2).
42
At UCSC and across the state, economics majors have substantially higher
average wages than college graduates who earned the observed counterfactual
majors.
43
Using either OLS estimates or median wages, the difference between
the average wages of economics majors and the weighted-average wage among the
counterfactual majors underestimates the causally estimated return to majoring in
economics by up to 21percent.
40
Above-threshold policy compliers are more likely to choose the BME track than the average economics
major. The fraction of economics majors on the BME track only increases slightly and statistically insignicantly
across the GPA threshold (10.5percentage points, SE 6.1), suggesting that the large share of policy compliers on
that track largely results from local student demand, not department policy. See online Appendix FigureA-15.
41
National wage-by-major medians display a similar pattern; see online Appendix TableA-6. California and
US statistics are from the ACS (Ruggles etal. 2020). See online Appendix Table A-7 for a UCSC-ACS major
crosswalk.
42
The imputed wage estimates partition students by their set of majors to calculate averages, whereas the
major-specic estimates assign multi-major students to their higher-earning major; see online Appendix FigureA-16.
Estimates of below- and above-threshold UCSC policy compliers’ imputed and actual wages follow Abadie (2002).
43
BME majors have somewhat higher average wages than other economics majors at UCSC, but not else-
where. UCSC’s high-wage technology and information management major includes the economics department’s
core course sequence as required courses.
VOL. 14 NO. 2 17
BLEEMER AND MEHTA: WILL STUDYING ECONOMICS MAKE YOU RICH?
Why might wage-by-major estimates differ from the treatment effect of majoring in
economics? To see the possible sources of bias, note that linear regression of observed
wages on treatment in population R estimates β
OLS
R
(w) ≡ w
̃
T
R
−
∑
m≠T
P
m
R
w
̃
m
R
and that it is generically true in a Rubin causal model that
(4) β
OLS
R
(
w
)
=
E
(
w
i
(
T
)
| m
i
= T
)
− E
(
w
i
(
∼T
)
| m
i
= T
)
]
Average Treatment Effect on Treated inR(To T
R
)
+
[
E
(
w
i
(
∼T
)
| m
i
= T
)
− E
(
w
i
(
∼T
)
| m
i
≠ T
)
]
Selection Bias
.
Equation(4) shows that OLS overestimates economics majors’ true wage gains
if those selecting economics would have earned more in non-economics majors
than those who did not select economics—due to, e.g., stronger prior quantitative
F6. A W D E C M
Notes: This gure shows average early-career 2017–2018 wages by major of UCSC students (estimated by OLS,
with and without control variables) and all California college graduates (ACS medians) for UCSC’s three eco-
nomics tracks and for the ten most common counterfactual majors earned by below-threshold UCSC policy com-
pliers, juxtaposed with the causally identied local average treatment effect on early-career wages for below- and
above-threshold UCSC policy compliers (following Abadie 2002). The black dotted lines show the average wages
of the majors chosen by below- and above-threshold policy compliers, calculated by assigning each 2008–2012
UCSC student to their corresponding majors’ average wage— leave-one-out in the UCSC no-controls sample—
and using the linear RD IV model on the resulting imputed wages. Counterfactual major shares are estimated by
the linear RD IV model predicting an indicator for earning that major; the shares sum to over 100percent because
below-threshold policy compliers earn more multiple majors. Bar widths are proportional to the major shares.
UCSC statistics from 2008–2012 UCSC students matched to 2017–2018 wages; California statistics calculated
from age 23–28 2017–2018 ACS respondents. OLS coefcients from regressions of wages on major indicators
with or without covariates ( gender-ethnicity, SAT score, zip code average adjusted gross income, cohort year, and
high school xed effects), partitioning students by their highest-earning major. See online Appendix FigureA-7 for
UCSC-ACS major mapping. Wages and wage-by-major averages are CPI adjusted to 2018 and winsorized at 2per-
cent above and below.
Sources: UC-CHP student database, CA Employment Development Department, and ACS (Ruggles etal. 2020)
60,000
40,000
20,000
0
$19,247
$17,461
$19,293
$22,123
Counterfactual majors and shares
Psychology: 20%
Env. studies: 14%
Tech./info. mgmt.: 12%
Sociology: 10%
Film and dg. med.: 8%
Legal studies: 8%
Mathematics: 7%
Lat. Amer. stud.: 5%
Art: 4%
Anthropology: 4%
Other majors: 19%
Econ. majors
Bus. mgmt. econ.: 90%
Global econ.: 6%
Economics: 4%
Est. average wages ($)
UCSC OLS,
no controls
UCSC OLS,
with controls
CA median wages
(ACS)
Local average
treat. effect
18 AMERICAN ECONOMIC JOURNAL: APPLIED ECONOMICS APRIL 2022
training or stronger preferences for high wages. Combining equations(2), (3),
and(4) yields
(5) LAT E
RD
(
w
̃
m
R
)
− LAT E
RD
(
w
)
=
[
LAT E
RD
(
w
̃
m
R
)
− β
OLS
R
(
w
)
]
Counterfactual Major Correction
+
[
To T
R
− LAT E
RD
(
w
)
]
Treatment Effect Heterogeneity
+
[
Selection Bias
]
.
Equation (5) decomposes the difference between the observational difference in
average wages by major in population R and our estimated treatment effect of
majoring in economics at UCSC. The counterfactual major correction is positive
whenever the majors selected by below-threshold UCSC policy compliers are sys-
tematically higher-earning than those selected by non-economics majors in R —as is
clear from comparing the denition of β
OLS
R
(
w
)
to equation(3). The treatment effect
heterogeneity term is positive whenever economics majors in R have larger latent
treatment effects than those of policy compliers near the GPA threshold. Selection
bias is positive when economics majors in R would have earned higher wages in
non-economics majors than nonmajors in R .
The left-hand side of equation(5) is negative and small when R consists of all
UCSC graduates, and the counterfactual major correction is very small. This implies
that the treatment effect heterogeneity and selection bias terms must roughly can-
cel each other out.
44
Figure6 shows this clearly. Above-threshold policy compliers
have lower average earnings than the average UCSC students on their economics
tracks, but their wages would have been even lower—to an even greater degree
than the difference in average wages by major—if they’d earned their second-choice
majors instead.
45
,
46
Combined with the fact that selection bias resulting from
observable characteristics is positive ( $19, 247 − $17, 461 > 0 ), this suggests
that To T
UCSC
< β
OLS
UCSC
< LAT E
RD
(w) : the average economics major earned a
return smaller than the observational wage difference, while students who were
barely unable to declare the economics major may have earned a return larger than
the observational wage difference.
Together, these results suggest that OLS and wage-by-major medians well
approximate, and in fact slightly underestimate, the causal effect of majoring in
economics identied by our instrumental variable design.
44
With all UCSC graduates as R , we estimate LAT E
RD
( w
̃
m
R
) = $19, 427 (Figure6), LAT E
RD
(w) = $22, 123
(Figure6), and β
OLS
R
(w) = $20, 039 (online Appendix TableA-6). The LHS is then −$2,876, the counterfactual
major correction is −$792, and the heterogeneity and selection terms sum to −$2,084, which is less than 10 percent
of the estimated treatment effect by magnitude.
45
This is consistent with students having comparative advantage in their preferred major (Kirkeboen, Leuven
andMogstad 2016), one dimension of treatment effect heterogeneity.
46
Using the CPI-adjusted 2009–2010 wage-by-major medians of earlier UCSC cohorts to impute the 2008–2012
cohorts’ wages yields LAT E
RD
( w
̃
m
R
) estimates that are strikingly similar to the true local average treatment effect
(online Appendix FigureA-17), suggesting that those effects are relatively stable over time.
VOL. 14 NO. 2 19
BLEEMER AND MEHTA: WILL STUDYING ECONOMICS MAKE YOU RICH?
VII. Conclusion
The UCSC Economics Department’s 2008–2012 binding major restriction policy
provides an unusual opportunity to transparently identify the personal early-career
wage return to earning an economics major in college. We show that the wage return
to economic education is very high relative to education in students’ second-choice
social science disciplines, causing a 46 percent increase in mid-twenties earnings
despite no change in educational investment or degree attainment. About half of the
observed effect can be attributed to economics majors’ specialization in particular
high-wage industries, in part reecting changes in students’ reported preferences
across professions. Mirroring a similar nding from studies of the return to addi-
tional years of education (Card 1999), we show that major-specic OLS estimates
and differences in median wages by major both slightly underestimate the observed
wage return to economics. For reference, a comparison between the national median
wages of college graduates with economics degrees and those of graduates with
degrees in UCSC economics students’ second-choice majors suggests that major-
ing in economics raises the net present value of a student’s college education by
$536,000, with the early-career annual wage difference widening over time.
These ndings imply that students’ major choices could have nancial implica-
tions roughly as large as their decision to enroll in college (Autor 2014), highlighting
the centrality of heterogeneity in the private returns to higher education. They also
point to students’ college major choice as a key decision point where policymakers
can intervene to substantially impact youths’ long-run labor market outcomes.
47
Finally, these ndings illuminate the relationship between major-specic returns
and industrial composition, suggesting an important role for preferences and
industry-specic human capital acquisition in postsecondary education.
These ndings come with four caveats. First, our results are estimated for stu-
dents at a moderately selective public university—at the sixtieth percentile of the
university average SAT distribution—where nearly all students eventually earn a
bachelor’s degree (at UCSC or elsewhere); the ndings may not be representative
of the average university student. Second, our analysis is restricted to students who
already choose to take introductory economics courses and may not extend to other
students. Third, there are many US states (unlike California) where economics
majors do not earn above-average early-career wages, suggesting an important role
for local labor demand in shaping major-specic returns.
48
Finally, higher educa-
tion’s broad public and nonpecuniary returns imply that wage returns are insufcient
in themselves for drawing conclusions about the efciency of educational policies
(e.g., see McMahon 2009).
47
Indeed, Bleemer and Mehta (2021) show that GPA-based major restrictions regressively shape students’
major choices, tending to decrease disadvantaged students’ access to universities’ high-demand majors.
48
For example, in the 15 states where industries’ employment shares among college graduates are least simi-
lar to California’s, 2017–2018 ACS statistics show that economics majors do not have higher median wages than
other college graduates and earn lower wages than nonmajors in most two-digit industries. See online Appendix
FigureA-18.
20 AMERICAN ECONOMIC JOURNAL: APPLIED ECONOMICS APRIL 2022
S A
We analyze students’ responses to two UCUES survey questions. The rst ques-
tion asks, “How many hours: -Studying and other academic activities outside of
class,” and respondents are provided 8 radio-button alternatives: “0; 1–5; 6–10;
11–15; 16–20; 21–25; 26–30; More than 30.” We code each range to its mean and
code “More than 30” to 35.
The second question asks, “Career hope to eventually have after education com-
plete.” Students’ available responses are “Agricultural/agribusiness; Artistic, creative
professions; Business, nance-related professions; Civil service/government;
Education; Engineering, computer programming; Law; Medicine, health-related
professions; Military; Psychology, helping professions; Researcher, scientist; I have
no idea whatsoever; Other.” Our analysis uses an indicator for whether the student
selected the third response, “Business, nance-related professions.”
REFERENCES
Abadie, Alberto. 2002. “Bootstrap Tests for Distributional Treatment Effects in Instrumental Variable
Models.” Journal of the American Statistical Association 97 (457): 284–92.
Ahn, Thomas, Peter Arcidiacono, Amy Hopson, and James R. Thomas. 2019. “Equilibrium Grade
Ination with Implications for Female Interest in STEM Majors.” NBER Working Paper 26556.
Altonji, Joseph G., and Seth D. Zimmerman. 2019. “The Costs of and Net Returns to College Major.”
In Productivity in Higher Education, edited by Caroline M. Hoxby and Kevin Stange, 133–76. Chi-
cago, IL: University of Chicago Press.
Altonji, Joseph G., Erica Blom, and Costas Meghir. 2012. “Heterogeneity in Human Capital Investments:
High School Curriculum, College Major, and Careers.” Annual Review of Economics 4(1): 185–223.
Altonji, Joseph G., Lisa B. Kahn, and Jamin D. Speer. 2014. “Trends in Earnings Differentials across
College Majors and the Changing Task Composition of Jobs.” American Economic Review 104(5):
387–93.
Altonji, Joseph G., Lisa B. Kahn, and Jamin D. Speer. 2016. “Cashier or Consultant? Entry Labor
Market Conditions, Field of Study, and Career Success.” Journal of Labor Economics 34 (S1):
S361– S401.
Altonji, Joseph G., Peter Arcidiacono, and Arnaud Maurel. 2016. “The Analysis of Field Choice in
College and Graduate School: Determinants and Wage Effects.” In Handbook of the Economics of
Education, Vol. 5, edited by Eric A. Hanushek, Stephen Machin, and Ludger Woessmann, 305–96.
Amsterdam: Elsevier.
Andrews, Rodney J., and Kevin M. Stange. 2019. “Price Regulation, Price Discrimination, and Equal-
ity of Opportunity in Higher Education: Evidence from Texas.” American Economic Journal: Eco-
nomic Policy 11 (4): 31–65.
Andrews, Rodney J., Scott A. Imberman, and Michael F. Lovenheim. 2017. “Risky Business? The
Effect of Majoring in Business on Earnings and Educational Attainment.” NBER Working Paper
Series23575.
Angrist, Joshua D., and Alan B. Keueger. 1991. “Does Compulsory School Attendance Affect School-
ing and Earnings?” Quarterly Journal of Economics 106 (4): 979–1014.
Angrist, Joshua D., and Victor Lavy. 1999. “Using Maimonides’ Rule to Estimate the Effect of Class
Size on Scholastic Achievement.” Quarterly Journal of Economics 114 (2): 533–75.
Arcidiacono, Peter. 2004. “Ability Sorting and the Returns to College Major.” Journal of Economet-
rics 121 (1–2): 343–75.
Arcidiacono, Peter, Esteban Aucejo, and V. Joseph Hotz. 2016. “University Differences in the Grad-
uation of Minorities in STEM Fields: Evidence from California.” American Economic Review
106(3): 525–62.
Arteaga, Carolina. 2018. “The Effect of Human Capital on Earnings: Evidence from a Reform at
Colombia’s Top University.” Journal of Public Economics 157: 212–25.
Autor, David. 2014. “Skills, Education, and the Rise of Earnings Inequality among the ‘Other 99 Per-
cent’.” Science 344 (6186): 843–51.
VOL. 14 NO. 2 21
BLEEMER AND MEHTA: WILL STUDYING ECONOMICS MAKE YOU RICH?
Beffy, Magali, Denis Fougère, and Arnaud Maurel. 2012. “Choosing the Field of Study in Postsecond-
ary Education: Do Expected Earnings Matter?” Review of Economics and Statistics 94 (1): 334–47.
Berger, Mark C. 1988. “Predicted Future Earnings and Choice of College Major.” ILR Review 41 (3):
418–29.
Black, Dan A., Seth Sanders, and Lowell Taylor. 2003. “The Economic Reward for Studying Econom-
ics.” Economic Inquiry 41 (3): 365–77.
Bleemer, Zachary. 2018. “The UC ClioMetric History Project and Formatted Optical Character Rec-
ognition.” UC Berkeley Center for Studies in Higher Education Research and Occasional Paper
Series CSHE.3.18.
Bleemer, Zachary. 2021. “Top Percent Policies and the Return to Postsecondary Selectivity.” UC Berke-
ley Center for Studies in Higher Education Research and Occasional Paper Series CSHE.1.21.
Bleemer, Zachary. 2022. “Afrmative Action, Mismatch, and Economic Mobility after California’s
Proposition 209.’’ Quarterly Journal of Economics 137 (1): 115-160.
Bleemer, Zachary, and Aashish Mehta. 2020. Major Restrictions, Socioeconomic Stratication, and
Student Outcomes. Oakland, CA: UCOP Institutional Research and Academic Planning.
Bleemer, Zachary, and Aashish Mehta. 2021. College Major Restrictions and Student Stratication. UC
Berkeley Center for Studies in Higher Education Research and Occasional Paper Series CSHE.14.21.
Bleemer, Zachary, and Aashish Mehta. 2022.“Replication data for:Will Studying Economics Make
You Rich? A Regression Discontinuity Analysis of the Returns to College Major.”American Eco-
nomic Association [publisher], Inter-university Consortium for Political and Social Research [dis-
tributor].https://doi.org/10.38886/E126941V1.
California Employment Development Department (EDD). 2019. “California Labor Market Informa-
tion Customized Data.” State of California. Accessed April 2019.
Canaan, Serena, and Pierre Mouganie. 2018. “Returns to Education Quality for Low-Skilled Students:
Evidence from a Discontinuity.” Journal of Labor Economics 36 (2): 395–436.
Card, David. 1999. “The Causal Effect of Education on Earnings.” In Handbook of Labor Economics,
Vol. 3A, edited by Orley C. Ashenfelter and David Card, 1801–63. Amsterdam: Elsevier.
Carnevale, Anthony P., Ban Cheah, and Andrew R. Hanson. 2015. The Economic Value of College
Majors. Washington, DC: Georgetown University Center on Education and the Workforce.
Castleman, Benjamin L., Bridget Terry Long, and Zachary Mabel. 2018. “Can Financial Aid Help to
Address the Growing Need for STEM Education? The Effects of Need-Based Grants on the Com-
pletion of Science, Technology, Engineering, and Math Courses and Degrees.” Journal of Policy
Analysis and Management 37 (1): 136–66.
Cohodes, Sarah R., and Joshua S. Goodman. 2014. “Merit Aid, College Quality, and College Com-
pletion: Massachusetts’ Adams Scholarship as an In-Kind Subsidy.” American Economic Journal:
Applied Economics 6 (4): 251–85.
Daly, Moira, and Daniel Le Maire. 2021. “University Admission and Preferred Field of Study.”
http://lemaire.dk/University_Admission_June_6.pdf.
Deming, David J., and Kadeem Noray. 2020. “Earnings Dynamics, Changing Job Skills, and STEM
Careers.” Quarterly Journal of Economics 135 (4): 1965–2005.
Denning, Jeffrey T., and Patrick Turley. 2017. “Was That SMART? Institutional Financial Incentives
and Field of Study.” Journal of Human Resources 52 (1): 152–86.
Fairlie, Robert W., Florian Hoffmann, and Philip Oreopoulos. 2014. “A Community College Instructor
Like Me: Race and Ethnicity Interactions in the Classroom.” American Economic Review 104(8):
2567–91.
Hahn, Jinyong, Petra Todd, and Wilbert Van der Klaauw. 2001. “Identication and Estimation of
Treatment Effects with a Regression-Discontinuity Design.” Econometrica 69 (1): 201–09.
Hamermesh, Daniel S., and Stephen G. Donald. 2008. “The Effect of College Curriculum on Earnings:
An Afnity Identier for Non-ignorable Non-response Bias.” Journal of Econometrics 144(2):
479–91.
Hastings, Justine S., Christopher Z. Neilson, and Seth D. Zimmerman. 2014. “Are Some Degrees Worth
More than Others? Evidence from College Admission Cutoffs in Chile.” NBER Working Paper19241.
Hastings, Justine, Christopher A. Neilson, and Seth D. Zimmerman. 2015. “The Effects of Earnings
Disclosure on College Enrollment Decisions.” NBER Working Paper 21300.
Hoekstra, Mark. 2009. “The Effect of Attending the Flagship State University on Earnings: A Discon-
tinuity-Based Approach.” Review of Economics and Statistics 91 (4): 717–24.
James, Estelle, Nabeel Alsalam, Joseph C. Conaty, and Duc-Le To. 1989. “College Quality and Future
Earnings: Where Should You Send Your Child to College?” American Economic Review 79 (2):
247–52.
22 AMERICAN ECONOMIC JOURNAL: APPLIED ECONOMICS APRIL 2022
Kinsler, Josh, and Ronni Pavan. 2015. “The Specicity of General Human Capital: Evidence from
College Major Choice.” Journal of Labor Economics 33 (4): 933–72.
Kirkeboen, Lars, Edwin Leuven, and Magne Mogstad. 2016. “Field of Study, Earnings, and Self-Se-
lection.” Quarterly Journal of Economics 131 (3): 1057–1111.
Kolesár, Michal, and Christoph Rothe. 2018. “Inference in Regression Discontinuity Designs with a
Discrete Running Variable.” American Economic Review 108 (8): 2277–2304.
Lee, David S., and David Card. 2008. “Regression Discontinuity Inference with Specication Error.”
Journal of Econometrics 142: 655–74.
McMahon, Walter W. 2009. Higher Learning, Greater Good. Baltimore, MD: Johns Hopkins Univer-
sity Press.
National Student Clearinghouse (NSC). 2019. “StudentTracker Database.” National Student Clearing-
house. Accessed January 2019.
Patnaik, Arpita, Joanna Venator, Matthew Wiswall, and Basit Zafar. 2020. “The Role of Heteroge-
neous Risk Preferences, Discount Rates, and Earnings Expectations in College Major Choice.”
NBER Working Paper 26785.
Porter, Catherine, and Danila Serra. 2020. “Gender Differences in the Choice of Major: The
Importance of Female Role Models.” American Economic Journal: Applied Economics 12 (3):
226–54.
Ruggles, Steven, Sarah Flood, Ronald Goeken, Josiah Grover, Erin Meyer, Jose Pacas, and Matthew
Sobek.
2020. “Integrated Public Use Microdata Series USA: Version 10.0.” Minneapolis, MN:
Minnesota Population Center, IPUMS. https://doi.org/10.18128/D010.V10.0 (accessed March
2020).
Rumberger, Russell W., and Scott L. Thomas. 1993. “The Economic Returns to College Major, Qual-
ity and Performance: A Multilevel Analysis of Recent Graduates.” Economics of Education Review
12(1): 1–19.
Sacerdote, Bruce. 2001. “Peer Effects with Random Assignment: Results for Dartmouth Roommates.”
Quarterly Journal of Economics 116 (2): 681–704.
Sekhri, Sheetal. 2020. “Prestige Matters: Wage Premium and Value Addition in Elite Colleges.” Amer-
ican Economic Journal: Applied Economics 12 (3): 207–25.
Sjoquist, David L., and John V. Winters. 2015b. “State Merit Aid Programs and College Major: A
Focus on STEM.” Journal of Labor Economics 33 (4): 973–1006.
Stange, Kevin. 2015. “Differential Pricing in Undergraduate Education: Effects on Degree Production
by Field.” Journal of Policy Analysis and Management 34 (1): 107–35.
Student Experience in the Research University (SERU). 2019. “University of California Undergrad-
uate Experience Survey (UCUES).” Santa Cruz, CA: University of California Santa Cruz. https://
www.universityofcalifornia.edu/infocenter/ucues-data-tables-2018 (accessed December 2019).
University of California ClioMetric History Project (UC-CHP). 2020. “University of California, Santa
Cruz Campus Transcript Database.” University of California ClioMetric History Project. Accessed
February 2020.
US Bureau of Labor Statistics (BLS). 2019. “All Urban Consumers (Current Series) Consumer Price
Index - CPI.” Washington, DC: U.S. Bureau of Labor Statistics. https://www.bls.gov/cpi/data.htm.
Accessed December 2019.
US Census Bureau. 2017. “2017 NAICS Structure with Change Indicator.” Washington, DC: U.S. Cen-
sus Bureau. https://www.census.gov/eos/www/naics/downloadables/downloadables.html (accessed
December 2019).
US Department of Education. 2019. Technical Documentation: College Scorecard Data by Field of
Study. 2019 ed., Washington, DC: U.S. Department of Education.
US Internal Revenue Service (IRS). 2018. “SOI Tax Stats - Individual Income Tax Statistics – 2018 ZIP
Code Data (SOI).” Washington, DC: U.S. Internal Revenue Service. https://www.irs.gov/statistics/
soi-tax-stats-individual-income-tax-statistics-2018-zip-code-data-soi (accessed January 2018).
Wiswall, Matthew, and Basit Zafar. 2015. “Determinants of College Major Choice: Identication
Using an Information Experiment.” Review of Economic Studies 82 (2): 791–824.
Wiswall, Matthew, and Basit Zafar. 2018. “Preference for the Workplace, Investment in Human Capi-
tal, and Gender.” Quarterly Journal of Economics 133 (1): 457–507.
Zafar, Basit. 2013. “College Major Choice and the Gender Gap.” Journal of Human Resources 48 (3):
545–95.
Zimmerman, Seth D. 2014. “The Returns to College Admission for Academically Marginal Students.”
Journal of Labor Economics 32 (4): 711–54.
