Wage gaps across demographic groups in the U.S. labor market are well documented. A key question is the degree to which group-based sorting into high- versus low-paying occupations reflects underlying preferences, versus structural barriers or prior educational experiences. High school Career and Technical Education (CTE) programs offer insight into the preference side of this question, since CTE pathways are largely open-access and allow students to explore a career field without committing to it. We study CTE enrollment patterns across four states and one large metro area to assess if potential pay in students' CTE fields foreshadows longstanding inequities in the labor market. The dominant theme that emerges from parallel multi-state analyses is that women concentrate in fields linked to jobs with 7%–20% lower pay, a range that includes the actual U.S. gender pay gap. We also find disparities in potential pay by race, ethnicity, family income, and disability identification, although these are much smaller and less consistent across locations than the gender gap.
APA
Carruthers, C. K., Dougherty, S., Goldring, T., Kreisman, D., Theobald, R., Urban, C., & Villero, J. (2026). Who takes high-earning CTE pathways? Southern Economic Journal. Advance online publication. https://doi.org/10.1002/soej.70044
BibTeX
@article{carruthers_cte_pathways,
author = {Carruthers, Celeste K. and Dougherty, Shaun and Goldring, Thomas and Kreisman, Daniel and Theobald, Roddy and Urban, Carly and Villero, Jes\'us},
title = {Who Takes High-Earning {CTE} Pathways?},
journal = {Southern Economic Journal},
year = {2026},
doi = {10.1002/soej.70044}
}
We study the effects of an increase in post-secondary educational opportunities on teen fertility by exploiting policy-induced variation from Ser Pilo Paga (SPP), a generous college financial aid program in Colombia that dramatically expanded college opportunities for low-income students. Our preferred empirical approach uses a triple difference design that leverages variation in the share of female students eligible for the program across municipalities and the fact that the introduction of SPP should not affect the education and fertility decisions of older women not targeted by the program. We find that after the introduction of SPP, fertility rates for women aged 15-19 years old decreased in more affected municipalities by about 6 percent relative to less affected municipalities. This effect accounts for approximately one-fourth of the overall decrease in teen fertility observed in the years following the program's announcement. Our results suggest that increasing economic opportunities through expanding college access can contribute to lowering teen fertility rates.
APA
Bloem, M. D., & Villero, J. (2024). College opportunity and teen fertility: Evidence from Ser Pilo Paga in Colombia. Journal of Development Economics, 171, 103321. https://doi.org/10.1016/j.jdeveco.2024.103321
BibTeX
@article{bloem_college_2024,
author = {Bloem, Michael D. and Villero, Jes\'us},
title = {College Opportunity and Teen Fertility: Evidence from \emph{Ser Pilo Paga} in {Colombia}},
journal = {Journal of Development Economics},
volume = {171},
pages = {103321},
year = {2024},
doi = {10.1016/j.jdeveco.2024.103321}
}
We describe alignment between high school career and technical education (CTE) and local labor markets across five states—Massachusetts, Michigan, Montana, Tennessee, and Washington. We find that CTE is partially aligned with local labor markets. A 10-percentage-point higher share of local jobs related to a CTE career cluster is associated with a 3-point higher rate of CTE concentration in that cluster. Women and students from racial or ethnic minority groups are better aligned with local employment than men, in part due to their selection of CTE fields like Education & Training, Health Science, and Hospitality & Tourism, which correspond with a large portion of the workforce in almost every metro area. We find more limited evidence of dynamic, short-term adjustments in CTE after changes in local labor markets. A small degree of realignment lags the labor market by two-to-three years and is only observed following changes in college-level employment.
APA
Carruthers, C. K., Dougherty, S., Goldring, T., Kreisman, D., Theobald, R., Urban, C., & Villero, J. (2024). Career and technical education alignment across five states. AERA Open, 10. https://doi.org/10.1177/23328584241276504
BibTeX
@article{carruthers_cte_alignment_2024,
author = {Carruthers, Celeste K. and Dougherty, Shaun and Goldring, Thomas and Kreisman, Daniel and Theobald, Roddy and Urban, Carly and Villero, Jes\'us},
title = {Career and Technical Education Alignment Across Five States},
journal = {AERA Open},
volume = {10},
year = {2024},
doi = {10.1177/23328584241276504}
}
We document an increase of roughly five percentage points in the employment rate gap between women with a college degree and women without a college degree over the last two decades, mainly driven by a sustained increase in employment among college-educated women with young children. We argue that differential patterns in motherhood child penalties are partially behind these trends, with child penalties in employment decreasing faster for women with college degrees since the mid-2010s. Exploiting cross-state variation in the share of jobs with future telework potential in the early 2000s, we find that child penalties for college-educated women decreased more in states with more employment in teleworkable occupations. We confirm this finding by leveraging individual-level occupational data and calculating child penalties for mothers in teleworkable and non-teleworkable occupations within a given education level. Child penalties decreased by more for women in teleworkable occupations, which are disproportionally held by women with college degrees.
Tuition Effects of IDR Plans: Evidence from the Introduction of the PAYE Repayment Plan
We study the effects of an increase in the generosity of income-driven repayment (IDR) plans on net tuition (tuition less school-provided financial aid) using policy-induced variation from the introduction of the Pay As You Earn (PAYE) repayment plan in 2012. We estimate future wages, student loan borrowing, and the present value of loan repayment savings from opting into PAYE at the student level based on the student's SAT score, college attended, field of study, gender, race, parental income, and other attributes. Using a triple difference framework, we find that elite colleges (all private non-profit universities and highly selective public universities) increase their annual net tuition by $3,324 to capture about $21 for every $100 in potential loan repayment savings; this effect is statistically insignificant and negligible for non-elite, non-selective colleges. Our estimates are robust to accounting for the potential endogeneity of applying for federal student aid post-PAYE, alternative definitions of elite and non-elite institutions, and variations in the nature of our treatment variable (discrete/continuous). As an application, we estimate that President Biden's proposed SAVE plan would effectively transfer about $12 billion to selective colleges over the next 10-year budget window.
By 2059, uncapping STEM green cards would increase effective labor input by 1.7 percent, total factor productivity by 1.3 percent, capital by 3.9 percent, and output by 4.0 percent relative to current law.
Average labor income would rise 2.9 percent by 2059. Low-education workers would gain the most, while incumbent foreign-born STEM workers would be the only group with persistently lower wages.
Federal revenues would rise 3.8 percent by 2059 while outlays would rise 0.9 percent. The primary deficit would be 41.7 percent lower and debt held by the public 5.5 percent lower than under current law.
Every native-born group would experience lifetime welfare gains across ages and income levels, with the largest gains accruing to younger, low-education households.
The $100,000 fee applies to H-1B petitions for workers abroad, an estimated 60 percent of FY 2024 registrations. Compared with fee-exempt workers, fee-subject workers are less likely to hold a master's degree or higher (11.5 versus 63.8 percent), more likely to be Indian nationals (82.2 versus 51.2 percent), and about five years older on average.
We estimate that employers would pay the fee for 30 to 56 percent of fee-subject registrations, depending on expected payroll savings. This screening would reduce the fee-subject share of the lottery pool from about 60 percent to between 31 and 46 percent.
At current prevailing wages, the fee and wage-level-weighted lottery would raise mean selected pay by $7,551 to $18,799 (6.7 to 16.7 percent) over the pre-2026 random lottery. Under the higher 2026 NPRM wages, the gain would rise to $16,542 to $28,686 (14.7 to 25.5 percent). The fee alone changes mean pay by −$3,606 to +$7,642.
Relative to the random lottery, the combined policies would reduce India's selection share from 66.8 percent to between 55.7 and 63.5 percent and raise the doctorate share from 5.7 percent to as much as 9.5 percent. The share with no prior U.S. visa would fall from 44.5 percent to between 26.1 and 33.9 percent, while the former F-1 share would rise from 43.3 percent to between 48.5 and 50.8 percent.
Under the proposed experience-benchmarking alternative, 56 percent of H-1B registrations would fall below the new Level I wage floor and receive no lottery entries, compared with 21 percent under the NPRM's primary rule.
Combined with the wage-level-weighted lottery, experience benchmarking would raise mean compensation by $27,686 (24.7 percent) over the prior random lottery. This gain is $7,076 larger than the $20,610 increase under the NPRM's primary rule.
Relative to the random lottery, selected workers would be 2.3 years younger on average and the former F-1 share would rise by 9.1 percentage points. These shifts run opposite to the primary rule because experience benchmarking rewards wages that are high relative to a worker's credentials.
Strategic occupational reclassification could offset about 12 percent of the compensation gain, compared with 18.9 percent under the primary rule, reducing the estimated increase from 24.7 percent to 21.6 percent.
Under the proposed prevailing-wage thresholds, 21 percent of current registrations would fall below the new Level I floor and receive no lottery entries, while only 7 percent would reach Level IV, down from 17 percent.
Combined with the wage-level-weighted lottery, the higher thresholds would raise mean compensation by $20,611 (18.4 percent) over the pre-2026 random lottery, nearly twice the $11,159 (10 percent) gain from wage weighting alone.
Relative to the random lottery, India's selection share would fall by 5.2 percentage points, the doctorate share would rise by 3.3 points, and Computer and Mathematical occupations would lose 2.8 points.
Strategic occupational reclassification could offset about 19 percent of the compensation gain, reducing the estimated increase from 18.4 percent to 14.9 percent.
We estimate that 61 percent of H-1B registrations could reach a higher wage level by switching to one of five closely related occupations, gaining 1.3 lottery entries on average among those that reclassify. With only the closest alternative available, 41 percent could reach a higher level.
Under the benchmark reclassification strategy, Architecture and Engineering occupations would gain 1.78 percentage points of selection share, while Computer and Mathematical occupations would lose an additional 1.40 points beyond the DHS rule.
Reclassification would raise the Level I selection share from 13 percent under the DHS rule to 21 percent, compared with 28 percent under the random lottery. The Level IV share would fall from 31 percent to 23 percent, compared with 18 percent under the random lottery.
Reclassification could offset 42 percent of the rule's expected compensation increase, reducing the gain in mean selected pay from 10 percent to 5.8 percent relative to the random lottery.
Across five allocation methods, the new DHS rule would raise mean selected compensation by $9,554 (8.5 percent), from $112,309 under the random lottery to $121,863. One alternative produces a similar mean of $122,504, while two raise it to $149,682 and $166,778.
The DHS rule would reduce Level I's selection share from 27 percent to 14 percent and raise Level IV's share from 15.5 percent to 26 percent. The alternative rules would allocate between 51.6 and 67.2 percent of selections to Level IV.
Relative to the random lottery, the doctorate share would rise by 1.5 percentage points under the DHS rule, compared with more than 5 points under the ranking alternatives. India's share would fall by 2.1 points to 65.5 percent, while women's share would fall from 32.5 percent to 31.9 percent.
Across the rules considered, estimated five-year wage effects for U.S.-born workers range from −0.04 to +0.01 percentage points. The DHS rule produces modest occupation and industry shifts, while the ranking alternatives shift more selections away from computer-related work.
About 1.3 million people held Temporary Protected Status as of March 2025, with nearly half living in Florida. Recipients from Venezuela, Haiti, El Salvador, Ukraine, and Honduras account for 97 percent of the total.
TPS workers provide 8 to 10 percent of hours worked in some occupations and major metropolitan areas. Relative to U.S.-born workers, they are 5.4 times as likely to work in building and grounds cleaning, 3.2 times as likely to work in construction, and twice as likely to work in transportation.
TPS workers generated an estimated $35.9 billion in GDP in 2023, including $10.7 billion in Florida, $4.3 billion in Texas, $3.6 billion in California, and $2.8 billion in New York.
Black borrowers are estimated to receive 43 percent of SAVE benefits among current borrowers with outstanding debt, compared with about 25 percent among future borrower cohorts.
Female borrowers would receive 71 percent of SAVE benefits among current borrowers and 68 percent among future cohorts, or roughly 70 percent across the two groups.
Borrowers in the top income quintile would receive an estimated 24 percent of subsidies among current borrowers, compared with 18 percent among future cohorts.
Borrowers with graduate loans would receive an average subsidy of $6,500 to $8,800 per beneficiary because their balances are larger, despite the plan's focus on undergraduate debt.
President Biden announced five student loan relief provisions. Some overlap with the SAVE plan, while forgiveness of longer-term debt is incremental and extends eligibility to higher-income households.
We estimate that the new provisions would cost $84 billion in addition to SAVE's estimated $475 billion, bringing the combined cost to about $559 billion.
Long-term debt relief would reach about 750,000 households with average income above $312,000, largely because SAVE already provides comparable relief to lower-income households.
The employment rate of prime-age women reached a record 75.3 percent in 2023, nearly 1 percentage point above its earlier peaks and 6 points above the rate a decade before.
The share of prime-age women with a college degree rose from below 30 percent to more than 45 percent over two decades. Because college graduates are more likely to work, this shift added 2.7 percentage points to women's employment since 2000.
Among college graduates with a child under age 10, the share working rose by 10 percentage points from the early 2000s to 2023, while the share out of work for family-care reasons fell by the same amount. The changes were largest for mothers of infants and toddlers and were only briefly interrupted by the pandemic.
The employment child penalty fell for all women in the 1980s and 1990s. After 2000, it continued to shrink for mothers with a college degree but stalled for mothers who did not attend college.
Using data from four large metro-Atlanta school districts, the study compares students just above and below the passing thresholds for Career and Technical Education (CTE) technical assessments. Passing an assessment has no detectable effect on college enrollment, either across all assessments or among the most frequently taken tests. Although the estimates cannot rule out modest positive or negative effects, the results suggest that earning a passing score does not alter the postsecondary plans of students near the threshold.
The report examines Career, Technical, and Agricultural Education (CTAE) participation across four metro-Atlanta school districts, focusing on differences by race and ethnicity, economic status, and gender. Differences by race and economic status are modest and largely reflect variation in course offerings and participation across schools. Gender differences are more pronounced: meaningful separation in both overall participation and the types of CTAE courses students take remains even among students attending the same school.
