Integrated Pell Grant Expansion and Bachelor s Completion Pay for Performance: A Cost-Effectiveness Analysis. Harrison G. Holcomb William T.

1 Integrated Pell Grant Expansion and Bachelor s Completion Pay for Performance: A Cost-Effectiveness Analysis Harrison G. Holcomb William T. Drier School of Public Affairs University of Colorado Denver Author s Note: The authors wish to thank Todd Ely Ph.D. Assistant Professor, Jane Hansberry Ph.D. Scholar-in-Residence, Benoy Jacob, Ph.D. Assistant Professor, Paul Teske, Ph.D. Dean, and Amanda Buchanan, MPA Candidate, for providing us with their invaluable assistance. Our participation in this competition is sponsored by the School of Public Affairs at UC-Denver.

Executive Summary Improving both overall bachelor s degree completion rates and equity in those completion rates between low-income citizens and others should be a policy priority for the United States at the national level. The slow growth in the credential completion rate and the gap in bachelor s degree completion between the wealthiest and poorest Americans threatens both economic growth and the strength of democratic principles in government. This memo evaluates the cost effectiveness of implementing two integrated policy solutions that focus on improving equity in educational attainment as a way to boost bachelor s degree completion in the United States. The first is an expansion of the existing Pell Grant program to cover the costs of concurrent enrollment programs for low-income high school students. Participation in concurrent enrollment has been shown to considerably increase the probability of both attending university and obtaining a degree. The second is the launch of a voluntary pay for performance initiative for four-year institutions to incentivize closing the current degree completion gap between low-income students and their better off peers. These initiatives are designed to work in tandem to improve educational attainment for low-income students from all backgrounds. We project that in the ten year period between 2017 and 2027, these policy solutions can generate a total of up to 2.3 million new bachelor s degrees for a total cost of about $3.45 billion. The rationale of our approach is summarized in the logic model presented in Figure 1 below: Figure 1: Integrated Pell Expansion and Pay for Performance Theory of Change Logic Model 2

Understanding the Challenge of Degree Completion The target population: low income students Low-income students comprise a large and increasing portion of both secondary and postsecondary students in the United States. The Department of Education reports that the percentage of secondary students who qualify for free or reduced price school lunch, a common proxy for poverty, increased from about 38% in 2001 to 51% in 2013. 1 Similarly, the percentage of students qualifying for Pell grants, which are based on family or individual income, have held steady at about 60% of applicants for federal financial aid. 2 Also, the number of students reporting family incomes below $30,000 has increased. 3 This low-income population often faces multiple barriers to success in postsecondary education, including a higher probability of being firstgeneration students, members of disadvantaged minority groups, or dependent care responsibilities. Low-income students also vary in age, with adult learners more common in this population. 4 Research in this population has also indicated higher levels of sensitivity to the costs of higher education, meaning that the presence or absence of financial aid, and how that aid is packaged have a disproportionate impact on their decision-making for further education. 5,6 3 The impact of underachievement As a result of these barriers, low-income students complete postsecondary education at much lower rates than their better off peers. The Pell Institute reported that in 2015, while 77% of dependent students in the top income quintile for the United States completed a bachelor's degree within six years, only 9% of students in the bottom income quintile hit the same benchmark. 7 The Education Trust indicates that at all levels of academic promise, and across all institutional types, students who receive Pell Grants perform worse than their better off peers. 8 Since students fitting the lowincome profile represent a significant portion of current and future postsecondary students, this inequity in completion rates has a significant negative impact on overall completion rates. Impacts on society This poor postsecondary completion rate and its disproportionate impact on low-income students has a negative impact on U.S. economic growth, weakens democratic participation, and can be associated with poorer population health outcomes. Higher levels of education among workers are associated with higher wages and greater worker productivity, leading to increased rates of economic growth. 9,10 Further, low levels of education reduce income for those workers, negatively impacting both the equity and efficiency of the economy. A general level of equity in educational attainment is also important to maintaining democratic principles of governance, since an ability to both access and interpret information about issues facing the state is essential for all citizens. 11 Citizens with considerably lower levels of educational attainment will be less able to engage in this core democratic activity, meaning disparities in education lead to disparities in democratic participation. The importance of the bachelor s degree credential There are a variety of first level postsecondary credentials available to students, including technical certificates, associate s degrees, or a bachelor s degrees. Since the problem is rooted in concerns about democratic participation and economic equity and efficiency, it makes sense to target the bachelor s credential as it is associated with the strongest increases in earnings potential. 12,13

Thus, increasing this credential could reasonably be expected to have the most long-term impact on improving equity. Concurrent Enrollment: A Proven Strategy to Improve Educational Outcomes Concurrent enrollment programs, which allow high school students to take college courses jointly with high school classes, have been widely implemented across the United States, with all 50 states documenting some concurrent enrollment programs operating at the local level. 14 47 states have also established state level policies and guidelines. These programs take a variety of forms. However, the predominant model, used by more than 60% of programs, allows high school students to take college-level courses taught by college faculty either on campus or at their high school. 15 Participation in concurrent enrollment has been associated with a variety of benefits for students including: increased high school graduation rates, higher rates of enrollment in postsecondary education, higher postsecondary graduation rates, and reduced time to degree. 16 One analysis conducted on students in the University of Texas system found that participants in concurrent enrollment graduated from college at a rate of 65%, and at a median time of 4.15 years. 17 Another study, conducted at Boise State University also found positive outcomes for students that participated in concurrent enrollment programs including improved academic performance. 18 However, despite widespread implementation, participation in concurrent enrollment remains comparatively low, particularly among low-income students. The US Department of Education reports that in 2013 only 8% of US students participated in concurrent enrollment with local universities. 19 Further, in the state of Colorado, low-income students (defined as those receiving free or reduced price lunch) were considerably underrepresented, participating at about half the rate of their better off peers. 20 These low participation rates and disparities in participation mean that the potential benefits of concurrent enrollment programs are not being realized. One reason for this underuse of concurrent enrollment is the lack of guaranteed funding for tuition, fees, textbooks, and transportation required of participants. Particularly to those who are more sensitive to educational costs than higher income peers, as noted by both Heller and St. John. 21,22 Only 15 states provide any state-level funding for concurrent enrollment programs, with others leaving it to the discretion of local school districts to fund. 23 Among the states that do provide any funding, even fewer cover all costs or target low-income students. This results in more limited program implementation and presents a barrier to low income students. Filling the Concurrent Enrollment Funding Gap: Pell Expansion Program The Pell Grant program was launched for the 1973-4 academic year, with the purpose of providing financial support to low-income students pursuing postsecondary education. The program is administered by the Department of Education, under authority granted to the department by the Higher Education Act. 24 In order to be eligible for a Pell Grant under the current system, a student must be enrolled at least part time at postsecondary school eligible to receive federal aid and complete a Free Application for Federal Student Aid to demonstrate financial need. 25 Pell grants are made on a sliding scale, based on the student s ability to pay, number of credits attempted, and the expected educational costs. 26 Under current policy, students participating in concurrent enrollment programs are not eligible to receive Pell grants or any other form of federal financial 4

aid to cover costs. This is despite the fact that most students participating in concurrent enrollment programs are at schools that meet the institutional eligibility criteria for Pell, specifically the rule that the coursework must count toward some academic program that can lead to a credential, including a bachelor or associate degree. 27 Our proposal would expand the Pell Grant program to include high school seniors participating in concurrent enrollment programs at institutions eligible to receive federal aid dollars. The requirement to complete a FAFSA to demonstrate need and document enrollment would remain the same, however students would be entitled to up to one academic year of additional eligibility for concurrent enrollment which would not be counted toward the six years of postsecondary eligibility that students are currently granted. Other aspects of the program, including award calculation procedures and funding allocations for administration by colleges and universities would remain largely unchanged. The only difference being that concurrent enrollment students would be treated as full time enrollees for the purposes of award calculation. Pell grants are subject to the same caps based on total expected costs as other forms of federal aid, meaning the grant cannot exceed the total educational costs reported by the institution. 28 As a result, Pell grants for high school students are likely to be smaller than for traditional postsecondary students, as they would be taking fewer credits and many states have established tuition discounts at public institutions, or created fee exemptions for concurrently enrolled students. 29 The expansion would specifically target low-income students for participation in concurrent enrollment programs, while also offering a general level of federal support for concurrent enrollment. This proposal represents a scaling up of an existing experimental program launched by the Department of Education in 2015, which provides grant money to postsecondary institutions to implement concurrent enrollment initiatives targeted at Pell eligible students. 30 Benefits of a Pay for Performance Approach to the Pell Completion Gap Adopting a pay for performance model to address the completion gap between Pell recipients and other students offers three advantages in comparison to traditional federal grant making efforts: it increases efficiency in expenditure, encourages innovation at participating institutions, and ensures sustainability for the new programs. These benefits have been documented in other initiatives in the social sector. 31 These advantages are particularly important in attempting to address the challenges of Pell grant recipients, who are a diverse population that often faces additional challenges impacting persistence in addition to low income. These challenges include: firstgeneration student status, membership in a disadvantaged minority group, or responsibility for dependents. 32 A pay for performance model changes the incentives of traditional grant making by paying the recipient only after agreed upon outcomes are achieved, rather than in advance based on a proposed set of outputs. In the case of Pell recipient degree completion, this approach focuses federal spending only on actual improvements in the completion gap, and encourages schools to maximize efficiency. 5

This model also allows schools greater flexibility in choosing which program investments it feels will most efficiently achieve outcomes. This will encourage both innovation and diversity in interventions. Since these interventions are likely to be tailored to the particular characteristics of an institution s population, they are also likely to have benefits for non-pell recipients who may share some of the additional barriers faced by many Pell students. In addition, a pay for performance approach will yield more sustainable projects, since institutions must rely on existing funds to implement their program activities. Programs that rely on internally allocated funds would be more likely to continue even when pay for performance funds are reduced or eliminated. Further, the pay for performance funds are not earmarked for particular activities. Therefore, there is even less of a contingent relationship between the availability of federal dollars and institutional activities. Even if incentive funds are received, they need not be spent on the programs that generated them. Pay for performance represents an efficient, innovative, and sustainable approach for the federal government to take in addressing the bachelor s degree completion gap between Pell recipients and other students. Implementing Pay for Performance The degree completion gap is a serious issue the Pell Grant program faces when measuring its success. A study conducted by the Education Trust found a 14% gap between Pell and non-pell students national college completion rate (51% to 65%). 33 The Department of Education found a similar rate which falls between 15-20%. 34 However, since institutions that receive Pell Grants are not required to disclose information on graduation rates for Pell recipients, more specific data is not available. We propose implementing a pay for performance program for Title IV bachelor s degree granting institutions. The program would have two goals, first to move Pell students toward parity in college graduation with their higher-income peers, and second to better understand what interventions are likely to achieve this goal at the institutional level. The objective of this program is to spur innovative and situation-specific solutions to addressing this gap, specifically for the group of students who are not able to benefit from concurrent enrollment credits via the Pell Expansion Program. Thus, schools would not be compensated for any degrees earned by Pell recipient students who also received a Pell Expansion Program grant. This would both control program costs, and ensure a more equitable allocation of resources among different groups of lowincome students. In order to address the gap, our Pay for Performance program would invite all bachelor s degree granting Figure 2. Percent of students at institutions by Pell/non-Pell completion gap Gap is > 3% 70% Gap is 3% 30% 6

Title IV institutions to participate. According to the Educational Trust this represents approximately 3000 institutions. 35 Approximately 35% of schools have completion gaps of 3% or less, and thus would be unlikely to participate. 36 However, the remaining 2100 institutions likely to participate serve 70% of the undergraduate student population. 37 The reach of the program is summarized in the graph to the right. Participating schools would sign performance agreements that would offer to compensate the schools at a rate of $4,000 per additional degree earned by a Pell grant recipient, excluding any Pell recipient students who also received a Pell Expansion Program grant. Payments begin at an improvement of 1% and up to parity in their Pell vs. non- Pell student six-year graduation rate. These performance dollars will be paid out at the close of the six-year performance cycle. This is designed to give schools the opportunity to have at least one cohort of students reap the full benefits of any interventions for the full period of Pell grant eligibility. Following the close of the initial six-year program period, an evaluation of outcomes and current status of the completion gap between Pell and non-pell students will be conducted using the dramatically increased volume and quality of data on Pell student graduation rates in comparison to the current information. This would provide an opportunity to evaluate how to better target institutions in future pay for performance arrangements. This period of evaluation would also allow for better assessment of the role of the Pell Expansion Program in improving completion rates and achieving parity between Pell and non-pell students. Analyzing the Cost-Effectiveness of Integrated Pell Expansion and Pay for Performance We conducted a cost-effectiveness analysis to determine the efficiency of the interventions. Costeffectiveness analysis uses projections of potential costs as well as possible benefits to determine the cost per outcome of an intervention. In this case, the desired outcome is additional college degrees, and the proposed interventions are an expansion of the Pell Grant program and a pay for performance program for four-year colleges. The costs are those associated with making additional grants available to high school students participating in concurrent enrollment through the Pell program and the payment of performance incentives for universities that narrow the completion gap. In developing our forecasts for costs and effects of each intervention, we relied on a variety of assumptions and parameters about demographic trends, participation in higher education, and costs which are summarized in two tables below: 7

8 Table 1. Pay for Performance Program Assumptions Assumptions Source # of students enrolled in postsecondary education (PSE) (2017) 11,022,000 National Center for Education Statistics (NCES) 65% of institutions represent 70% of students 80% Education Trust College graduation rate of Pell students 51% Education Trust College graduation rate of non-pell students 65% Education Trust National education gap 12% Education Trust FAFSA application rate 85% Office of Federal Student Aid (OFSA) Pell eligible rate 60% OFSA Pell accepting rate 80% OFSA Table 2. Pell Expansion Program Assumptions Assumptions Concurrent enrollment participation growth rate- no intervention 0.5% % high school (HS) grads to PSE 70% # of HS Grads 3,328,740 Concurrent enrollment participation growth rate- intervention 4% Source National Center for Education Statistics (NCES) Bureau of Labor Statistics (BLS) NCES projection through 2023, ~0.2% average growth rate 2024-2027 Jobs For the Future (JFF) report on Colorado CE goals Colorado Department of Higher Education Concurrent enrollment success rate 93% Concurrent enrollment graduation rates 65% ACT Non-concurrent enrollment graduation rates 59% NCES HS to PSE rate 66% BLS # eligible HS seniors 4,058,000 US Census Bureau Adoption rate Minimum Pell award $599.82 Discount rate 0.9% 4% of student population adopts per year Apply for FAFSA 85% Pell eligible rate 60% OFSA Pell accepting rate 80% OFSA JFF report on Colorado CE goals Office of Federal Student Aid, 1% growth rate per year Office of Management and Budget Office of Federal Student Aid (OFSA)

Intervention Costs: Grants and Incentives The primary costs associated with our interventions are the additional Pell grants made to high school students through the Pell Expansion Program, and the potential payments made to schools through the pay for performance initiative. Pell Expansion Program The baseline cost for Pell Grants is set at the minimum potential award for each year, which begins at $599, and has consistently increased by about 1% per year in the past five years. 38 Our justification is that students eligible for the Pell Grant will be earning credits in a part-time capacity they will not have to worry about cost of living, and credits obtained through concurrent enrollment are usually discounted for high school students within most current local level program support. In fact, the Department of Education reports that more than half of institutions that have concurrent enrollment students offer some discounting. 39 For the purposes of this analysis, we also assume no major change in state and local level funding for concurrent enrollment programs. 40 Cost for degrees = Eligible 12th graders CE adoption rate Pell accepting rate Pell award amount Pay for Performance Program The $4000 paid per new degree is based on a median value from other equity achievement grants provided to postsecondary institutions by the federal Fund for the Improvement of Post Secondary Education. 41 In the table below we project total program costs based on various levels of organizational performance, ranging from 5% to a maximum of 30%, which would approximately reach parity in college completion. We anticipate performance outcomes at about 5% on average across all institutions, based on analysis of the general trends and variance in bachelor s degree completion rates. Table 3. Pay for Performance Sensitivity Analysis Gap Closure % Degrees Cost 30% 453,300 $1,913,300,000 15% 226,600 $956,460,000 5% 75,550 $319,000,000 Cost for degrees = Pell student population (graduation rate improvement previous graduation rate) $4000 Approach to discounting costs In evaluating both interventions we use a discount rate of 0.9% which is recommended by the Office of Management and Budget (OMB) for cost-effectiveness analysis. 42 This is considerably more conservative than the 3.0-3.5% rate used in academic cost-effectiveness studies of education policy interventions. 43,44 Potential variance in the actual discount rate, which is typically adjusted annually by OMB, would have an impact on the actual cost of the program. 9

10 Effect: Additional Bachelor s Degrees Pell Expansion Program The chart below summarizes the number of additional degrees anticipated as a result of the Pell expansion program. We determined the number of degrees by calculating the difference in total degrees per year with enhanced adoption of concurrent enrollment supported through Pell expansion, and the existing trend in completion rates with no intervention. The additional degrees are visualized in the following graph: 1700000 1600000 1500000 1400000 1300000 1200000 1100000 Figure 3: New Degrees from Pell Expansion 1000000 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 Expected Degrees PEP Degrees The formula is provided below: Degrees earned = (Entering PSE population CE participation rate CE success rate CE graduation rate) + (Entering PSE population (NonCE participation rate + CE fail rate) NonCE graduation rate) We further conducted sensitivity analysis to determine the potential impact of variance in the adoption rate for concurrent enrollment on the Table 4. Pell Expansion Program Sensitivity number of additional degrees earned, using Analysis adoption rate data from Colorado as a proxy. 45 Adoption Rate Degrees Cost per Degree Colorado was selected as it has implemented a 2% 684,500 $ 4,600 funding model similar to Pell at the state level. 46 We 4% 2,258,700 $ 1,400 find that 4%, which is what Colorado projects to be 6% 2,803,700 $ 1,100 the most likely outcome, as the state has consistently seen increases above 3%. 47 Pay for Performance During its six-year cycle, we calculated the additional degrees generated by the Pay for Performance program based on national estimates of the total volume of improvement in the completion gap across all participating institutions. Using data from the Department of Education,

we estimate that the overall improvement will be a 5% reduction in the completion at participating schools, generating an additional 75,500 degrees over six years. Degrees earned = # students enrolled 2017 70% of eligilbe students Pell acceptance rate Graduation rate of Pell students Gap closure rate Results In the ten-year period between 2017 and 2027, we anticipate that our combined interventions will yield a total of 2.3 million degrees for a total cost of about $ 3.45 billion. These additional degrees represent a 14% increase, compared to expected number of degrees with no intervention. The full results are summarized in the table below: 11 Table 5. Summarized Results % New Cost Intervention Cost 2017-2027 # Degrees 2017-2027 Degrees Degree Pell Expansion $3,100,000,000 2,250,000 13% $1,460 Pay for Performance $345,000,000 75,500 1% $4,000 Integrated $3,445,000,000 2,300,000 14% $1,500 per Conclusion The combination of the Pell Expansion Program for concurrent enrollment with the Pay for Performance program represent an evidence-based and cost-effective approach to the dual challenges of low postsecondary completion rates and income based disparities in college completion. These problems limit economic growth and erode the democratic foundations of the US government. This proposal leverages the existing successes of concurrent enrollment programs and pay for performance programs to improve equity in college graduation, and thus fuels robust growth in overall college completion. It is particularly effective since it manages to broadly target low-income students of all ages, as well as creating incentives for the expansion of concurrent enrollment and improved support services for all students. Thus, this policy can cost-effectively help the United States meet the need for a globally competitive workforce, thriving economy, and robust democracy.

12 1 Institution of Education Sciences (2014). Digest of Education Statistics US Department of Education Washington, DC. Retrieved at: https://nces.ed.gov/programs/digest/d14/tables/dt14_204. 10.asp?current=yes 2 Office of Federal Student Aid (2015). Data Center: Student Application Volume [Data Set]. Retrieved at: https://studentaid.ed.gov/sa/about/data-center/student/application-volume/fafsa-school-state 3 Davison, C.J. (2015). Examining Zero Expected Family Contribution as a New Criterion for Low Income Comparing the Impact on Student Persistence at Two- and Four-Year Institutions. Community College Journal of Research and Practice 39:5 pp. 442-460 4 ibid note 3. 5 Heller, D.E. (1997) Student Price Response in Higher Education: An Update to Leslie and Brinkman. Journal of Higher Education 68:6 pp.624,61 6 St. John, E.P.(1990) Price Response in Persistence decisions: an analysis of the high school and beyond senior cohort. Research in Higher Education 31:4 pp. 387-403 7 The Pell Institute (2015). Indicators of higher education equity in the united states:45 year trend report (2015 Ed.). Pell Institute Publishing. Retrieved at:http://www.pellinstitute.org/downloads/ publications_indicators_of_higher_education_equity_in_the_us_45_year_trend_report.pdf 8 Education Trust, The. (2015). The Pell Partnership: Ensuring a Shared Responsibility for Low-Income Student Success. Retrieved from: https://edtrust.org/wp content/uploads/2014/09/ ThePellPartnership_EdTrust_2015.pdf 9 Berger N., Fisher P. (2013). A well-educated workforce is key to state prosperity. Economic analysis and research analysis network: Washington, DC. 10 Aghion P., Boustan L., Hoxby C., Vandenbussche J. (2009) The causal impact of education on economic growth: evidence from us. Harvard University, Boston MA Retrieved at: http://scholar.harvard.edu/files/aghion/files/causal_impact_of_education.pdf 11 Dahl, R.A. (2000). On democracy. New Haven, CT Yale University Press. 12 Carnevale A.P.,Rose S.J., Cheah B. (2011) The College Payoff: Education, Occupations, and Lifetime Earnings. The Center for Education and the Workforce: Georgetown University Washington, DC 13 Tamborini C.R.,Kim C.H.,Sakamoto A.(2015) Education and Lifetime Earnings in the United States. Demography 52 pp.1383-1407 14 Zinth, J.A. (2015). Individual state profiles: dual enrollment. Education Commission of the States: Denver, CO. Retrieved at: http://ecs.force.com/mbdata/mbprofallrt?rep=de14a 15 ibid, note 14. 16 Cassidy L., Keating K., Young V. (2012). Dual enrollment: lessons learned on school level lessons learned on school-level implementation. EDJ Associates: Herndon, VA. 17 Radunzel J., Noble J., Wheeler S. (2014). Dual credit/dual enrollment coursework and long-term college success in Texas. ACT: Iowa City, IA. 18 Belcheir, M. J. (2008). What happens to concurrently enrolled high school students who attended boise state university during high school? Office of Institutional Research. Boise, ID. Retrieved at: http://ir.boisestate.edu/wp-content/uploads/2011/07/rr200803concurrentenrollmentstudy.pdf 19 United States Department of Education (2013). Dual enrollment programs and courses for high school students at postsecondary institutions: first look. Institute for Education Statistics Washington, DC 20 Colorado Department of Education, Colorado Department of Higher Education (2015). Annual report on concurrent enrollment 2013-2014 school year. Division of Central Services: Denver, CO. 21 ibid note 5 22 ibid note 6

13 23 ibid, note 14. 24 20 U.S.C. Ch. 28 SubCh, IV. Part A 1070 25 30 C.F.R. 690 26 Office of Federal Student Aid. (2015). 2015-2015 Federal Student Aid Handbook (Vols. 1, 3). Washington, DC: US Department of Education. 27 ibid, note 14. 28 ibid, note 26. 29 ibid, note 14. 30 80 F.R. 67734 31 The Non-Profit Finance Fund (2012). Pay for Success: Investing in what works. Washington, DC Retrieved at: http://permanent.access.gpo.gov/gpo23617/pay_for_success_report_2012.pdf 32 Ibid note 3. 33 ibid note 8. 34 U.S. Department of Education, National Center for Education Statistics, 2003-04 Beginning Postsecondary Students Longitudinal Study, Second Follow-up (BPS:04/09). 35 ibid note 8. 36 ibid note 8. 37 ibid, note 8. 38 ibid, note 26. 39 ibid, note 19. 40 ibid, note 16. 41 United States Department of Education (2015). Fund for Innovation in Postsecondary Education Database [Data Set]. Retrieved at: http://fipsedatabase.ed.gov/fipse/search.cfm 42 Office of Management and Budget. (2014). Circular A-94 Appendix C. Retrieved at: https://www.whitehouse.gov/omb/circulars_a094/a94_appx-c/ 43 Levin, H., Belfield, C., Muennig, P., Rouse, C. (2006). The cost and benefits of an excellent education for all of America s children. Columbia University. New York, NY. Retrieved at: http://www3.nd.edu/~jwarlick/documents/levin_belfield_muennig_rouse.pdf 44 Hollands, F.M., Hanisch-Cerda, B., Levin, H. M., Belfield, C.R., Menon, A., Shand, R., Pan, Y., Bakir, I., & Cheng, H. (2015). CostOut - the CBCSE Cost Tool Kit. Center for Benefit-Cost Studies of Education, Teachers College, Columbia University. Retrieved from: www.cbcsecosttoolkit.org 45 Ibid note 20. 46 Ibid note 20. 47 Ibid note 20.