Are Choice Schools Worth The Money?

Posted on

School Choice: Future of New Magnet Schools Uncertain1 – a headline in the CT Mirror early this year questions the worth of school choice after the US Department of Education reported that Connecticut’s public school enrollment is projected to decrease by 5% over the next decade. 2 The debate around choice schools, which consist primarily of magnet and charter schools, is often dominated by intellectual considerations of educational equity. Waiting for Superman, a pro-school choice documentary, depicts the failures of the American public education system and glorifies charter schools as the only way out for underprivileged children. Choice proponents advocate school choice as a fundamental right that parents should have to advance their children’s education, while choice opponents argue that choice schools benefit a small proportion of students at the expense of the majority. However, policy decisions are often made from both political and economic standpoints, and it is imperative to address them in the context of school choice.

Looming debt problem

One of the reasons why choice school expansion is up in the air is due to the massive state budget deficit. Using Census Bureau’s 2012 population estimates, the state debt is equal to $16,178 for every resident of a U.S. state. 3 But if we base the state debt in terms of percentage of Gross State Product, Connecticut ranks seventh on the list of most indebted states with 49% as compared to the state average of 33%.

Screen Shot 2015-05-12 at 11.21.51 AM
Source: Eucalitto, Cory

The budget deficit is projected by the Office of Fiscal Analysis to exceed $1.3 billion in 2015-2016 and $1.4 billion in 2016-2017. 4 With this economic landscape in mind, it is worth questioning the decision to increase spending for choice school expansion. Can the state afford to expand choice schools at a time when district schools are in need of more funding? This is a question that the Connecticut General Assembly has to answer, and this piece will offer an economic perspective on choice school spending and deliver proposals for the state legislature to promote educational equity in a fiscally responsible manner.

While policymakers may point out that painting an ugly picture of state budget deficit argues against costly expansion of choice schools and at the same time warrants a reduction in educational spending, I seek to lay out my argument more explicitly by saying that I am not advocating for less spending on education. What I am suggesting is to shift funding away from choice school expansion towards district schools that require more funding per pupil. Since the state budget deficit is substantial, every dollar counts, and every additional dollar should go to the most needy district schools, where the greatest positive impact can be made.

Charter schools cost more

Choice proponents claim that choice schools are able to serve their students at a much lower cost, which might imply that they are more cost-effective in providing the same if not better quality of education than their district counterparts. Based on data available online, I have decided to show graphs for school expenditures in 2008-2009, a period where graphs for magnet, charter, and district schools are readily available. After excluding expenditures for transportation and special education services, Prof. Bruce D. Baker, a professor of education at Rutgers University, made an observation that deserves attention from choice proponents. The following graph compares the expenditure-per-pupil between charter and district schools in Connecticut.

Source: Baker, Bruce

Data points for district schools revolve around $10,000 per pupil, while most data points for charter schools are located above $10,000 per pupil. This shows that charter schools have received more state funding than district schools. Prof. Baker excluded transportation and special education expenditure for a fair comparison between charter and district schools because transportation and special education expenses for charter school students are booked as district expense. In reality, charter schools are even more costly to run if they do not remit transportation and special education expenditure back to the district.

Source: Baker, Bruce

Since district schools incur less expenditure-per-pupil, shifting funding away from choice schools towards district schools is an attractive and economical option for Connecticut. To showcase a more specific comparison between district and charter schools, the graph above illustrates spending between a district school and a charter school in Hartford and Bridgeport. If we look at the bars that exclude transportation and special education services, charter school spending per pupil is actually higher than district school spending per pupil.

Impact of charter schools

Since charter schools do not present a cost benefit for the state, perhaps the state could justify the increase in spending for charter schools through the positive impact on student performance. The National Charter School Study conducted by CREDO 5 revealed that students in charter schools have shown an upward trend in their performance over the past five years. However, the study did not address the fact that charter schools might be cream skimming, a process of selecting more academically able students for enrollment. This selection bias is accounted for in another study of charter schools 6 by the National Center for Education Evaluation, in which students who won the lottery were compared with students who lost the lottery, eliminating the selection bias of intrinsic motivation. They found that on average, charter schools did not have a statistically significant impact on student achievement. The following graph shows the substantial increase in state funding for charter schools over a 14-year period, and every increase in charter school funding represents an opportunity cost for district schools that could have used the funds to serve more disadvantaged students. Given the fact that the impact of charter schools is questionable even when the state has increased funding over time, it does not make economic sense to expand the type of schools that will benefit only a small proportion of students, if any benefit can be found at all.

Source: Thomas, Jacqueline

Impact of magnet schools

Magnet schools are the state’s go to solution to comply with the Sheff mandate of at least 25% white enrollment. Casey D. Cobb, a professor of educational leadership at the University of Connecticut, Robert Bifulco, Associate Professor of Public Administration at Syracuse University, and Courtney Bell, an associate research scientist at the Educational Testing Service, conducted research 7 on the improvement in academic performance of students in Connecticut magnet schools in 2009 by comparing test scores for city students who attended magnet schools and those who applied but did not get in. Even though the report concludes that interdistrict magnet high schools have positive effects on mathematics and reading achievements of urban students, I find it difficult to draw reliable conclusions due to the small sample size and the potential for unobservable selection bias. If similar results were obtained after using a larger sample size, we need to ask ourselves whether millions of state dollars spent on magnet schools could have been used by district schools to produce similar if not better results.

Screen Shot 2015-05-03 at 8.20.57 PM
Source: Thomas, Jacqueline

Magnet schools are expensive 

State funding for magnet schools has increased drastically over time, and with features such as specialized sciences or arts programs, magnet schools also cost more to run. Using Connecticut Department of Education data, the CT Mirror reported that magnet schools in Hartford region incur a higher expenditure-per-pupil at an average of $12,845 per pupil, which is about $2,500 more than the average expenditure-per-pupil for Connecticut district schools.

Source: Frahm, Robert A.

Since magnet schools cost more than charter schools and district schools, in order to justify an increase in funding from the state, the rate of return of magnet schools has to be substantially greater than suggested in the Bifulco paper. More complete research involving a larger sample size of schools should be conducted to evaluate the effectiveness of magnet schools and the costs associated to run them.

Choice schools under-serve needy communities

With the questionable impact of choice schools in mind, more funding should be given to cash-strapped district schools given the fact that they enroll more English language learners (ELL) and special education students who need access to more costly resources. Robert Cotto Jr., the Director of Urban Educational Initiatives, and Kenneth Feder, a policy analyst at Connecticut Voices for Children, reported 8 that magnet and charter schools have a lower percentage of ELL and special education students.

Screen Shot 2015-05-03 at 8.25.40 PM
Source: Cotto & Feder

The lower percentage of ELL could result from cream skimming practices or the lack of conducive environments for disadvantaged students. In a CT Mirror article 9, Jenifer Colon, an eighth grader from Puerto Rico said that magnet schools “are not really focused on English language learning students,” which explains why it made sense for some students like her to attend a district school to learn the language. Cotto’s report also found that charter schools are hyper-segregated by race, serving more than 90% minority students.

Screen Shot 2015-05-03 at 8.27.48 PM
Source: Cotto & Feder

In all four towns on the graph above, charter schools have the highest percentage of student body that identified as students of color. Since it is not clear whether charter schools improve student performance, the lack of racial integration should be a strong signal to the state that perhaps charter schools are not worth further investments.

Expand Open Choice program

Despite the fact that Connecticut could ill afford to spend more money on choice school expansion at a time when limited conclusions can be drawn about choice school impact on educational outcomes, the proposal to shift funding from choice schools towards urban district schools may still be unpopular in the Connecticut state legislature, which is dominated by suburban lawmakers who represent constituents less inclined to want their taxes spent on more urban children. My first proposal is to increase financial incentives for suburban district schools to enroll more urban students through the Open Choice program. Suburban districts get an average of $2000-$6000 from the state for every urban student that they enroll, which is substantially lower than the average expenditure per pupil for magnet schools in meeting the court-mandated Sheff requirement. In light of the fact that mandatory participation in the program is “not on the table right now” 10, the state should increase funding for Open Choice by another $3000 or more per pupil, which is still more cost effective than choice school spending, to improve educational resources for current district schools. This would also promote racial integration and allow a higher proportion of urban students to benefit from richer educational resources that suburban district schools possess. Since suburban schools would be incentivized to enroll more urban kids, the increase in state funding would benefit both suburban and urban students and dispel some resistance from suburban legislators. Educationists could then study educational models of successful suburban district schools and replicate them in the city to help improve urban student performance. I should also note that Open Choice might incur higher transportation expenses that may not appear in the district schools’ budgets, since the state bears those costs. However, this caveat applies to choice schools as well as they require transportation expenditure that would be incurred by districts.

Funding disparity between districts

We need to be aware of the funding disparity between school districts, because Open Choice is likely to increase funding for rich districts more so than for poor districts. This may be a bitter pill that policymakers have to swallow, as suburban lawmakers have to be incentivized to pass a bill that would benefit district schools in both suburbs and cities. The following graph illustrates the most advantaged school districts and the most disadvantaged school districts in Connecticut.

Source: Baker, Bruce

Ironically, two of the most disadvantaged school districts, Bridgeport and New Britain, have higher percentages of students who are ELL, who receive FRPM, or identify as Black or Hispanic, but the nominal current expenditure per pupil and average teacher salary is relatively lower than their suburban counterparts.

Increase funding for under-resourced district schools

Ultimately, the main driver behind educational inequality in Connecticut still lies with urban district schools that require the most state funding. My second proposal, albeit less popular than the first one if proposed as a bill in the state legislature, is to directly allocate more funds to urban district schools. Based on the previous graph, the expenditure per pupil in the most disadvantaged districts, which happened to be urban districts, is significantly lower than advantaged districts. Even though channeling more funds into district schools should improve educational resources, improving student performance is not a direct result of simply spending more money. Drawing on a national longitudinal dataset collected over 12 years, W. Norton Grubb, the author of Money Myth 11, makes a crucial distinction between “simple” resources such as higher teacher salaries and “abstract” resources such as school culture that cannot be readily bought. While an inefficient use of state funds that improve only simple resources will not necessarily improve performance, it does not detract from the fact that on average, expenditure per pupil is positively correlated with improved student outcomes. 12 While the effect of increased financial resources may be larger for some students than others, “money matters, resources that cost money matter, and more equitable distribution of school funding can improve outcomes”.

Ethical implications

Open Choice Program could potentially draw more bright and motivated students away from urban schools towards suburban schools, and worsen the average student performance in urban schools. Moreover, the burden of busing through the Open Choice Program falls primarily on minority urban students as they have to experience long bus rides every morning to get to suburban schools. Many white suburban students do not have great incentives to travel to urban schools for their education and could comfortably go to suburban schools in their neighborhoods, escaping the burden of busing. Magnet schools in suburbs and cities distributed the burden of busing more evenly, as suburban students would bear the burden of long bus rides to get to urban magnet schools. However, if funding for urban district schools significantly exceeded funding for suburban district schools, more suburban parents would be incentivized to send their children to enroll in urban schools to take advantage of greater educational resources that arise from higher expenditure.

Choice school expansion may not be the way forward to achieve educational equity for many children in Connecticut. Increasing funding for choice schools reveals a dark implication that we no longer believe in district schools as the hallmark of American public education. It reveals our preference for abandoning poor district schools and leaving children trapped in those schools to fend for themselves while a small group of students well-poised to take advantage of school choice leave their struggling classmates behind. We need to start questioning the effectiveness of choice schools in providing quality education to our children and to pay more attention to district schools that have been pushed to the sidelines. Policies should aim to achieve educational equity for as many students as possible rather than to concentrate on improving academic outcomes for a small subset of the student population at the expense of the majority. Are choice schools really worth our money? The answer to this question contains implications for efforts that could reduce the income gap in Connecticut. After all, if policymakers were able to reform struggling district schools into exemplary providers of first-rate public education, Davis Guggenheim, director of Waiting for Superman, would have less of a sob story to tell.


  1. Thomas, Jacqueline Rabe. “School Choice: Future of New Magnet Schools Uncertain | The CT Mirror.” January 6, 2015.
  2. Hussar, W.J., and Bailey, T.M. (2013). Projections of Education Statistics to 2022 (NCES 2014-051). U.S. Department of Education, National Center for Education Statistics.
  3. Eucalitto, Cory. “State Budget Solutions’ Fourth Annual State Debt Report > Publications > State Budget Solutions.” January 8, 2014.
  4. The Day. “The Day – Budget Cuts Costs, Breaks Promises.” February 19, 2015.
  5. Center for Research on Education Outcomes (CREDO). 2013. National charter school study 2013. Stanford, CA: CREDO.
  6. Gleason, Philip, Melissa Clark, Christina Clark Tuttle, and Emily Dwoyer. The Evaluation of Charter School Impacts: Final Report. NCEE 2010-4029. National Center for Education Evaluation and Regional Assistance, 2010.
  7. Bifulco, Robert, Casey D. Cobb, and Courtney Bell. “Can Interdistrict Choice Boost Student Achievement? The Case of Connecticut’s Interdistrict Magnet School Program.” Educational Evaluation and Policy Analysis 31, no. 4 (December 1, 2009): 323–45. doi:10.3102/0162373709340917.
  8. Cotto, Robert, and Kenneth Feder. “Choice Watch: Diversity and Access in Connecticut’s School Choice Programs | Connecticut Voices for Children,” April 2014.
  9. Thomas, Jacqueline Rabe. “Despite Robust Options, Thousands Pass on School-Choice Lottery | The CT Mirror,” December 10, 2014.
  10. Thomas, Jacqueline Rabe. “Efforts to Desegregate, Expand School Choice a Challenge with Looming Deficits | The CT Mirror,” November 2, 2012.
  11. W. Norton Grubb, The Money Myth : School Resources, Outcomes, and Equity (New York: Russell Sage Foundation, 2009).
  12. Baker, Bruce. “Does Money Matter in Education? | Shanker Institute.” Accessed May 3, 2015.

3 thoughts on “Are Choice Schools Worth The Money?”

  1. Are they worth the money indeed Hans? You’re taking on one of the most important questions in education policy right now, and I give you a lot of credit for that. I think your thesis is that “no, they are not worth the money.” If I’m correct about your thesis, you make strong quantitative arguments to support this point. I’m particularly glad to see you drawing on Bruce Baker’s research. I also think you’re making good use of the web-essay format, because your figures clearly support your argument. Right now, however, your argument is a little hard to follow. I think this is first and foremost because you don’t use any subheadings, which are particularly important in a web-essay where the reader just keeps scrolling and scrolling and it’s easy to get lost. I would recommend breaking your essay up into an Introduction, Conclusion, and distinct subsections for each your central points – a) charters and magnets are surprisingly expensive, and CT needs to be cost conscious because of our debt problems, b) charters and magnets serve fewer disadvantaged students, and c) opening any new schools when enrollment is declining doesn’t make sense – as well distinct subsections for each of your two recommendations – a) expand Open Choice instead and b) increase funding for under-resourced local public schools. Right now, all of this runs together. Additionally, through your essay, several times you say “I find…” e.g. “I found a surprising observation…” which implies you did the data analysis yourself. In fact you are citing the data analysis of others, which is fine, but you need to make the attribution more clear in text, whether it is to Bruce Baker or Robert Frahm or Robert Cotto (and me). Finally, I want to challenge two of your statistical arguments; right now they are strong, but I think consideration of the following points could make them stronger. First, per-capita debt is a misleading measure – a better measure is debt as a percentage of Gross State Product (GSP) or as a share of personal income; you might consider using that instead. Second, the reason Baker excludes transportation and special education expenses is because these expenses are remitted back to the district by charter and magnet schools, so expenses booked as district expense may actually be spent on charter and magnet schools; excluding special-ed and transportation allows for apples to apples comparison, and I recommend you make this more clear. Good work and good luck!

  2. Hans – I’m glad to see all of your revisions suggested by your guest evaluator and our discussion about your first draft (which you explained is no longer available online). In particular I was impressed with how you expanded your economic arguments and added newer ethical considerations. But there still are some political or sociological obstacles to overcome. For example, you wrote: “if funding for urban district schools significantly exceeded funding for suburban district schools, more suburban parents would be incentivized to send their children to enroll in urban schools to take advantage of greater educational resources that arise from higher expenditure.” Unfortunately, some White families fled to the suburbs to avoid contact with people of color in the cities, and economic rationalism does not necessarily displace racism. Nevertheless, your dollar-driven analysis gives us much to think about.

    1. Thanks for your comment, Jack.

      While some White families fled to the suburbs for racist reasons, I do believe that most White parents ultimately want to send their kids to the best schools. If substantial amount of funding is given to and used wisely by urban district schools, the quality of education that they provide will inherently improve over time to a level that is greater than the improvement of suburban schools.

      For many White parents who may not be able to afford private schools, heavily-funded urban district schools with great resources (greater teaching training, college-prep curriculum, partnerships or scholarship programs with companies in the cities etc) will be an attractive and economical option that might displace some of the racist effects. I think the fact that urban district schools are located in the city where many corporations and businesses are based is a value that should be explored further.

      After all, those companies want to look good with their CSR programs, and district schools should take full advantage of that along with the greater funding from the state to be the best schools they can possibly be. As the educational quality rises, I believe that more White parents will be incentivized to send their kids to urban schools.

Comments are closed.