How Important Is Instructional Spending to College Students’ Future Earnings?

Upon high school graduation, students must decide whether to go to college and, if applying, what type of college to attend. While college students tend to earn more than those who forgo college altogether, their earnings typically depend on their major, academic ranking and the type of college they attended.

In this blog post, we examine the earnings gap between students attending the most competitive and the least competitive colleges. We ask whether this earnings gap can be explained by the corresponding gap in instructional expenditures—an input we view as important for students’ skill formation. More precisely, our hypothesis is that schools that invest more in their faculty, libraries and science labs offer a higher-quality education that translates into higher earnings for their students.

Students’ Future Earnings Differ by College Type

We utilized college-level data sourced from the Integrated Postsecondary Education Data System (IPEDS) and students’ median earnings 10 years after starting college from College Scorecard, an online tool from the U.S. Department of Education. We limited our sample to four-year colleges in the United States, split into five groups on the Barron’s Selectivity Index (we excluded a sixth Barron’s category, noncompetitive colleges, as well as special colleges). Ranging from the “most competitive” to the “less competitive” schools, these categories are determined by factors including acceptance rate, median freshman SAT/ACT scores and the average ranking of students in their high school class.See Barron’s Profiles of American Colleges: 2007 (27th edition).

The figure below displays colleges’ median student earnings in 2011—again, 10 years after those students entered school—averaged across institutions in each Barron’s category.College Scorecard reports median earnings for students 10 years after starting college regardless of their graduation outcomes. Higher median earnings for a given college likely indicate greater persistence toward graduation and better early career outcomes for its students. Each school only reports student earnings to College Scorecard as a median, which we averaged for institutions in a Barron’s category. Averages are calculated using student enrollment as weights. Students attending the most competitive schools in the early 2000s earned a median salary of $65,366 in 2011, which was 26% higher than the earnings of students in the second most competitive category and 75% higher than those in the lowest (less competitive) category.

Median Student Earnings by Barron’s College Types, 2011

SOURCES: College Scorecard, Integrated Postsecondary Education Data System and authors’ calculations.

NOTES: Colleges only reported a median for student earnings, which is then averaged for the schools in the Barron’s category. Averages are calculated using student enrollment as weights.

What are plausible reasons for these large earnings gaps? We began by investigating how these schools differed along other dimensions. The table below displays average instructional expenditures per student, the shares of students majoring in the more lucrative STEM and business fields, and average SAT scores.

Instructional Spending, STEM Majors Are Significantly Higher at More Competitive Colleges

Our analysis shows that college categories differ dramatically in terms of instructional expenditures per student (first row in the table above). The most competitive schools spent nearly 2.5 times the amount spent by schools in the second (highly competitive) category and 5.2 times the amount spent by those in the lowest (less competitive) category.

Colleges in the most competitive category also had a larger share of students majoring in STEM-related fields: The share was 27% for the most competitive category versus 17% for the less competitive category. STEM major composition partially explains the earnings differential across the college categories. Interestingly, the same cannot be said about the business major composition: The less competitive schools had larger percentages of students majoring in business (third row in the table).

The more selective colleges also tend to enroll students who are better prepared academically. The average SAT score in the most competitive college category was 150 points above the average score in the second most competitive category and 383 points above that in the lowest category. Higher achieving students are more likely to graduate and do better in the labor market regardless of the college they attend. Therefore, student selection by itself may be why student median earnings are higher among the more selective schools.

Which College Characteristics Explain the Earnings Gap?

By looking at college characteristics, we can conclude that instructional expenditures, the share of STEM majors and student selection all help explain the earnings differentials among school categories. But how quantitatively significant is each characteristic?

To answer this question, we need to measure the effect of each college characteristic on student earnings. To this end, we employed college-level data to estimate a linear regression model of the natural logarithm of median student earnings on the characteristics described above. That is, we built a model to see how important each characteristic is in explaining student earnings.

The model does a good job of predicting median student earnings. For any given pair of college categories, the model predicts an earnings gap that is close to the observed earnings gap. The predicted earnings gap can then be decomposed into contributions of differential college characteristics reported in the table above.

The table below shows the decomposition of the earnings gap between the two highest Barron’s categories, the most competitive and the highly competitive colleges.

The results above show the gap in instructional expenditures explains 31% of the total predicted earnings gap between the most competitive and highly competitive schools. This suggests it is an important explanation of the observed gap between these two categories. The gap in instructional expenditures is a more significant explanation than STEM major composition, as the two highest college tiers educate a similar share of STEM students (27% and 26%, respectively). However, it is a less significant characteristic than student selection, as the gap in average student SAT score explains 44% of the predicted earnings gap.

Conclusion

Large differences in student earnings exist between colleges of different competitiveness level. Focusing on the first two categories of most competitive and highly competitive, we showed that differences in college-level instructional expenditures per student explained an important part—31%—of the earnings gap. Its contribution was similar or slightly smaller when comparing the most competitive category with the remaining three categories of schools (very competitive, competitive, and less competitive).

Notes

1. See Barron’s Profiles of American Colleges: 2007 (27th edition).
2. College Scorecard reports median earnings for students 10 years after starting college regardless of their graduation outcomes. Higher median earnings for a given college likely indicate greater persistence toward graduation and better early career outcomes for its students. Each school only reports student earnings to College Scorecard as a median, which we averaged for institutions in a Barron’s category. Averages are calculated using student enrollment as weights.