In 1995, Texas began roll out of an incentive program to increase the offering of AP classes in high schools across the state. AP curriculum is standardized across schools and allows high-achieving students to demonstrate college readiness and earn credit for college classes.
The state policy provided subsidies for teachers to become trained in AP course materials, financial incentives for schools whose students passed AP exams, grants for schools to invest in start-up equipment, and subsidies of students’ exam fees. This led to the percentage of high schools offering AP classes in Texas to expand from 20% in 1995 to 70% in 2006.
Zhu and Huynh (2025) utilizes a difference-in-differences framework comparing the outcomes of students at schools that eventually adopted APs before and after adoption, to the outcomes of students in the same cohorts at schools that never adopted APs. The paper relies on longitudinal data covering the full population of Texas public school students to link high school enrollments with Texas public university enrollment and graduation, as well as labor market outcomes. The paper finds that AP implementation led to increased likelihood of enrolling in a four-year university as well as an earnings increase of three percent at age 27.
Pays for Itself
The paper estimates the net cost of Texas AP program implementation as the various direct costs as well as the fiscal externality from increased earnings.
The direct costs include teacher training reimbursements, grants for start-up equipment to schools implementing AP programs, subsidization of students’ AP exams, student-performance bonuses for schools, and schools’ operating costs for AP courses.
Teachers of AP courses were eligible to attend an AP workshop and receive a $450 stipend in a school-year that their school offered a new AP course (at most once every two years). The paper assumes that each teacher of an AP course only applies for the stipend once. Summing the number of teachers eligible in each year from 1995 to 2006, multiplying by $450, leads to a total cost of $18,828,704 for teacher stipends.
In addition, between 1999 and 2002 the Texas government provided one-time grants of $3,000 to schools implementing a new AP program. Summing the number of schools implementing AP programs in these years leads to a cost for equipment grants of $1,662,297.
A key component of the Texas AP program is $50 stipends for each student’s AP exam. Summing the total number of AP exams taken from 1995 to 2006 leads to a total cost of $43,302,883.
Since 2000, schools have received performance bonuses of $100 for each student scoring 3 or higher on at least one AP exam. Summing the number of students who meet this criterion from 2000 to 2006 leads to a total cost of $74,304,456.
The paper assumes the operating cost of an AP course per semester is $5,000. Summing across the total number of courses operated between 1995 and 2006 leads to a total cost of operating AP courses of $794,865,032 million. Separately, the paper finds a start-up cost of implementing AP courses of $3,857,358.
Thus, the direct costs for implementing AP program in Texas high schools is $936,881,957.
The paper also estimates the effect of the AP program on labor market outcomes and incorporates the additional taxes recuperated as a fiscal externality. Using the difference-in-difference analysis, the paper estimates the treatment effect of attending a high school that offers AP classes on earnings for years one through nine post high-school graduation. They then apply the year 9 percentage gain in earnings to the average wages observed for Texas-born US citizens ages 27-65 in the American Community Survey (2017-2019) to estimate a lifetime earnings effect. This leads to an estimate of additional lifetime earnings of $91,979 per student. Assuming an effective tax rate of 18.7%, leads to increased tax revenue of $17,200.20 per person. Aggregating this by the total number of affected students leads to $34,602,330,564 in increased tax revenue.
Thus, the net government cost for implementing AP in the 1995 through 2006 high school cohorts is $936,881,957-$34,602,330,564 = -$33,665,448,607. All components have been discounted back to 1995 at a 3% rate and adjusted for inflation to 2024 dollars.
The paper estimates the willingness-to-pay as the increased after-tax earnings due to the AP program implementation. As described above, the paper finds increased pre-tax earnings of $91,979 per person and assumes an 18.7% tax rate which leads to increased post-tax earnings of $74,779.5 per person. Aggregating up to the population of students at schools with AP classes leads to a total willingness-to-pay of $150,436,870,312.
With a positive willingness-to-pay of $150,436,870,312 and a government cost of -$33,665,448,607 the estimated MVPF for the Texas AP Incentive Program is infinite.
The paper notes that even if only estimated additional tax revenue from earnings in the first nine years post high-school graduation is included, the estimated net cost is still negative, yielding an infinite MVPF.
The paper finds that the Texas Incentive Program boosts college attendance, which would increase government spending on public universities. While this is not accounted for in the estimated net costs, the increase in spending would need to be over $33 billion for the estimated MVPF to not be infinite.
Zhu, Huihuang and Johnny Huynh (2025). “On the Right Track? Evaluating the Achievement and Inequality Effects of Academic Tracking.” Working Paper. https://www.dropbox.com/scl/fi/00qudn432zgpyq8u6fed3/AP_Huihuang.pdf?rlkey=1iu2ylhq4bcm2lrvjowf4kjrr&e=1&dl=0.