Summary

In this study, we examine the value-added earnings outcomes of 935,767 students who enrolled in 86 public institutions in Texas from 2008-09 through 2018-19 and who sought bachelor’s degrees, associate’s degrees, or certificates.

We conducted this study in Texas because of its high-quality longitudinal data system. We are grateful to the Texas Higher Education Coordinating Board and the Texas Education Research Center for granting us access to Texas’ exceptionally high-quality data system.

We find that students who entered public institutions of higher education in Texas from 2008-09 through 2018-19 on average experienced positive cumulative net value-added earnings (VAE), as follows:
 

• Bachelor’s degree-seeking students who entered public institutions in Texas in 2008-09 experienced cumulative net VAE of $86,806 in year 15 after entry. Bachelor’s degree-seeking students who entered public institutions in Texas from 2009-10 through 2013-14 and have completed 10-14 years of the full 15-year follow-up period exhibited similar results.
   
• Associate’s degree-seeking students who entered public institutions in Texas from 2008-09 through 2013-14 experienced cumulative net VAE of $25,338 in year 10 after entry.
   
• Certificate-seeking students who entered public institutions in Texas from 2008-09 through 2018-19 experienced cumulative net VAE of $3,818 in year 5 after entry.
   

These students’ cumulative net VAE reached a low point after entry when their combined losses from paying tuition and fees and from foregone earnings peaked. This financial low-point occurred in year 5 after entry for bachelor’s degree-seeking students (-$33,925), in year 4 after entry for associate’s degree-seeking students (-$10,282), and in year 2 after entry for certificate-seeking students (-$3,461).

These students reached a financial break-even point several years after entry when their cumulative net VAE turned positive for the first time. This financial break-even point occurred in year 10 after entry for bachelor’s degree-seeking students, in year 7 after entry for associate’s degree-seeking students, and in year 4 after entry for certificate-seeking students.

Exhibit S1. Cumulative Net VAE for Students Seeking Bachelor’s Degrees, Associate’s Degrees, and Certificates

Image

Exhibit Note: Cumulative net VAE values in this exhibit are averages, stated in 2023 dollars, for cohorts of entering students that include a mix of eventual completers and non-completers. This exhibit pertains to 28,614 bachelor’s degree-seeking students who entered 29 public institutions in Texas in 2008-09, 559,068 associate’s degree-seeking students who entered 57 public institutions in Texas from 2008-09 through 2013-14, and 67,486 certificate-seeking students who entered 57 public institutions from 2008-09 through 2018-19.
 

Importantly, within the large cohorts of entering students depicted in the exhibit above, which pool students across institutions and across multiple years, we find that cumulative net VAE varies significantly across — and sometimes turns negative for — smaller cohorts of entering students who are organized by institution, program, or demographic group.

Exhibit S2. Variation in Cumulative Net VAE by Institution, Program, and Demographic Group

Exhibit Note: Cumulative net VAE values in this exhibit are averages, stated in 2023 dollars, for cohorts of entering students that include a mix of eventual completers and non-completers. This exhibit pertains to 28,614 bachelor’s degree-seeking students who entered 29 public institutions in Texas in 2008-09, 559,068 associate’s degree-seeking students who entered 57 public institutions in Texas from 2008-09 through 2013-14, and 67,486 certificate-seeking students who entered 57 public institutions 2008-09 through 2018-19.

In our study, we analyze in various ways the variation that we observe in cumulative net VAE across different types of entering cohorts. Two important findings from this analysis are:
 

• Program Choice Explains More Than Institution Choice. Students’ choice of program (holding constant their institution) explained more of the variation in their cumulative net VAE than did their choice of institution (holding constant their choice of program). This finding held for bachelor’s degree-seeking students, associate’s degree-seeking students, and certificate-seeking students. It is evidence that program design by institutions and program choice by students are crucial to students’ cumulative net VAE.
   
• Institution Choice Explains More than Household Income Level. Students’ household income level (holding constant their choice of institution) explained substantially less of the variation in their cumulative net VAE than did their choice of institution (holding constant their household income level). This finding held for bachelor’s degree-seeking students, associate’s degree-seeking students, and certificate-seeking students. It is evidence that cumulative net VAE of students – regardless of their household income level – is closely related to attributes of the institutions in which they enroll.

Goals of Study

We strive for a methodology for measuring students’ value-added earnings outcomes that rivals the rigor and accuracy of academic studies and that can also be implemented at scale.

We hope that our methodology is of use to federal and state policymakers who are interested in implementing — and setting policy based on — rigorous, accurate, and scalable approaches to measuring economic outcomes in higher education. 

We hope that the findings in our study inform decision makers in higher education, including institutions (in their internal improvement work), students (in their choices about pursuing postsecondary education), and policymakers (in their general understanding of outcomes in higher education).
 

Strengths of Study Design
 

• Value-added Earnings. We measure students’ value-added earnings, not their absolute earnings.
   
• All Entrants. We measure outcomes for all entrants to institutions and programs, including both eventual completers and non-completers. We do not merely measure outcomes for completers, for first-time/full-time students, or for students who receive federal financial aid.
   
• Comparison Group Earnings. We generate sophisticated comparison group earnings estimates or “counterfactual earnings” that adjust extensively for attributes of students in our treatment cohorts. We produced comparison group earnings estimates by assembling comparison groups with individuals who closely resemble students in our treatment cohorts (minus the choice to enroll in postsecondary education during the follow-up period). We generate unique comparison group earnings estimates for each of the ~8,500 entering cohorts that we examine in this study.
   
• Earnings, Net Costs, and Foregone Earnings. We generate high-quality earnings observations and net cost estimates because of our access to person-level data in Texas’ data system. We capture the opportunity cost to students of foregone earnings during enrollment because we estimate students’ value-added earnings from their point of entry.
   
• Scale and Repeatability. Our study is large (we examine approximately one million students who entered Texas higher education from 2008-09 through 2018-19) and repeatable (it is software-based and can be implemented in any data system with the requisite person-level data and data linkages).
   

Technical Features of Study Design
 

• Cumulative Net VAE. We estimate cumulative net value-added earnings (VAE) for cohorts of entering students over an appropriate follow-up period after entry. Students’ value-added earnings (VAE) — in a single year or cumulatively over a follow-up period after entry — is the difference between students’ actual earnings and their earnings in a counterfactual scenario in which they do not enroll in the institution in question or pursue postsecondary education elsewhere during the relevant follow-up period. 

Cumulative Net VAE = 
(Cumulative Earnings - Cumulative Comparison Group Earnings) - (Cumulative Net Costs)
 

• Comparison Group Earnings. To estimate comparison group earnings — also referred to as “counterfactual earnings” — we assemble, and observe the earnings of, comparison groups of individuals who closely resemble students in our entering cohorts, minus the choice to enroll in the institution in question or pursue postsecondary education elsewhere during the relevant follow-up period. We match individuals in comparison groups to students in treatment cohorts based on a wide range of attributes to maximize the accuracy of our comparison group earnings estimates, including (but not limited to): 
   
  • Prior Earnings 
  • Household Income 
  • High School Test Scores 
  • Prior Postsecondary Education 
  • Age 
  • Gender 
  • Race and Ethnicity
  • County of High School Attended
     
• Net Cost Estimates. Our cost estimates include tuition and fees, net of federal and state grants, exemptions, and waivers. We estimate costs for each student in each semester, based on their enrollment intensity in each semester.
   
• Foregone Earnings. We capture the cost to students of foregone wages while enrolled because we estimate value-added earnings from entry.
   
• Follow-up Periods. Our follow-up periods for estimating cumulative net VAE are: 
   
  • 15 years from entry for bachelor’s degree-seeking students 
  • 10 years from entry for associate’s degree-seeking students 
  • 5 years from entry for certificate-seeking students
     
• Cohort Types. We analyze three types of entering cohorts, in an institution in a particular academic year: 
   
  • Institutional cohorts consist of students who enroll in the institution for the first time in the same academic year.
  • Programmatic cohorts consist of students in the same institutional cohort who also enroll in the same type of program.
  • Demographic cohorts consist of students in the same institutional cohort who also share a demographic trait (i.e., household income level, high school math achievement, or age at entry).
     
• Data Sources and Sample Construction. We use data from Texas’ longitudinal data system to assemble a large baseline sample of individuals who graduated from a Texas high school, and we enrich this sample with person-level high school, postsecondary, and earnings data. This baseline sample provides extensive, common information on the students in our treatment cohorts and on individuals in our comparison groups.