Friday Finance 101: What Can we Learn about Education Costs & Efficiency by Studying Existing Public Schools?

Posted on June 29, 2012



One pervasive reformy argument is that our entire education system may be instantly transformed to be more productive and efficient by instantly adopting untested reformy policies and/or untested solutions of sectors other than education. Further, that we must take these bold leaps of faith because the public education system itself is too corrupt, too bloated, too inefficient to provide any useful lessons! Perhaps the whole system can be replaced with you-tube videos. Or perhaps we can just fire all of the teachers with more than 10 years experience and pay the rest based on the test scores they produce! Or perhaps some other lessons of industry can cure the (unsubstantiated) ills of American public schooling!

Kevin Welner and I addressed this issue in our critique of materials provided on the U.S. Department of Education’s website on improving educational productivity.  Specifically, Marguerite Roza and Paul Hill in one working paper titled Curing Baumol’s Disease argue that the entire public schooling system suffers from a disease of inefficiency and thus any lessons for improving educational productivity must be sought outside of the current system.

Similar arguments have been used by those who claim that state legislatures and state courts should never rely on cost analyses based on current practices of existing educational systems in order to either guide the design of state school finance systems through reform legislation, or to evaluate whether state school finance systems are equitable or adequate.

Researchers and/or policy analysts tend to use either of two general approaches to study education costs, identifying spending levels that should generally be sufficient for achieving desired outcomes and identifying how education costs vary from one location to another across districts within a state and how those costs vary by the needs of varied student populations. One approach involves gathering focus groups of informed constituents to specify the inputs to schooling they believe are needed to get the job done. These professional judgment panels are essentially proposing a hypothesis of the programs and services needed under varied conditions and for varied student populations to achieve desired outcomes. The alternative is to construct statistical models which estimate the relationship between current district spending levels and current student outcomes, with consideration for various factors that affect the cost of achieving desired outcomes (student characteristics, district characteristics, labor market pressures) and with consideration for factors that influence whether districts are more or less likely to spend inefficiently.

This approach, called education cost function modeling has been used extensively in peer-reviewed studies of education costs and cost variation.[1]  As Tom Downes, an economist from Tufts University explained back in 2004: “Given the econometric advances of the last decade, the cost-function approach is the most likely to give accurate estimates of the within-state variation in the spending needed to attain the state’s chosen standard, if the data are available and of a high quality” (p. 9).[2]

But, because these methods are sometimes used beyond academic journals, and in the highly political context of estimating not only how much money is possibly needed to achieve certain outcomes, but also how that money should be distributed across districts and children, they are not without controversy. These methods become the subject of more heated debate when they are introduced as evidence to assist judges in their evaluation of the constitutionality of state school finance systems. Heck, as explained below, a few authors have gone to great lengths to try to explain/argue how such information should never be used to either guide policy development or evaluate the rationality of current policies. Those assertions are completely unjustified.

The goal of education cost modeling – or any form of cost analysis – whether applied for evaluating equal educational opportunity or for producing adequacy cost estimates is to establish “reasonable marks” to provide guidance in developing more rationale state school finance systems. Only with reasonable marks in hand can one make informed judgments as to whether existing policies are wide of those reasonable marks.

Historically, funding levels for state school finance systems have largely been determined by taking the total revenue generated for schooling as a function of statewide tastes for statewide taxation and dividing that funding by the number of students in the system. That is, the budget constraint – or total available revenue – and total student enrollment have been the key determinants of the foundation level, or basic allotment. To some degree, this will always be true. But reasonable estimates of the “cost” of producing desired outcomes, given current technologies of production (the range of practices actually used/tested), may influence the taste for additional taxes by revealing that the preferences regarding taxation and the preferences regarding desired quality of public education are misaligned, meaning that one or the other should be adjusted. That is, if we find out that higher outcomes are going to cost us more, we can then have a more reasonable discussion of whether we are willing to pay that amount more for the expected gain in quality, or whether to lower our expectations. Alternatively, we can simply fly blind!

It’s rather like the individual who wishes to buy a Cadillac Escalade but expects only to spend about $25,000. After a little research, he finds that he can either buy a Ford F-150 for $25,000 or an Escalade for $65,000. That’s where that little bit of research comes in handy – identifying the gap between uniformed assumptions and reasonably informed ones, albeit with greater precision (actual prices) in this example than in cost estimation in education. Heck, if one wants to get really crazy with this, one could fit a statistical model relating prices with various features of existing makes and models of “comparable” vehicles.

Reasonable estimates of cost may also assist courts in determining whether current funding levels and distributions are wide of a reasonable mark, or substantially misaligned with constitutional standards. Cost model estimates are not meant to be exact predictions of what student outcomes will necessarily occur next year if we suddenly adopt a state school finance system based on the cost model estimates. Cost models provide guidance regarding the general levels (predictions with error ranges) of funding increases that would be required to produce measured outcomes at a certain level, assuming that districts are able to absorb the additional resources without efficiency loss.

Studies of state school finance reform also suggest that the key to successful school finance reforms is that they are both substantive and sustained. If additional dollars to high need districts are best leveraged toward high quality preschool programs and/or early grades class size reduction, we are unlikely to see changes to college readiness outcomes the following year (or following five years). If the additional dollars are best leveraged toward increasing teacher salaries for teachers in their optimal years of experience, allowing districts to recruit and retain “better” teachers over time, we are also unlikely to see immediate returns in student test scores.

Importantly, cost model estimates are estimates based on the actual production technologies of schooling. They are based on the outcomes schools and/or districts produce under different circumstances, for different children – the actual children they serve, based on the actual assessments given, and based on the real conditions under which children attend school.

Some critics of education cost analysis in general, and cost function modeling in particular assert that all local public school districts are simply inefficient, mainly because they pay their personnel based on parameters not associated with improved student outcomes.[3] Therefore, they assert that it is useless to consider the spending practices of current districts when trying to determine how much needs to be spent to achieve desired outcomes. A common version of this argument goes that if schools/districts paid teachers based on test scores they produce and if schools/districts systematically excessed ineffective teachers, productivity would increase dramatically and spending would decline. Thus, educational adequacy could be achieved at much lower cost, and therefore, estimating costs based on current conditions/practices is a meaningless endeavor.[4]

The most significant problem with this logic is that there exists absolutely no empirical evidence to support it. It is entirely speculative, frequently based on the assertions that teacher workforce quality can be improved with no increase to average wages, simply by firing the bottom 5% each year and paying the rest based on the student test scores they produce.  To return to the car purchasing analogy above, this is like assuming that somewhere out there is a car/truck with all the features of the Escalade, but the price of the F-150 – specifically, a version of the Escalade itself produced by a new, yet to be discovered technology with materials not yet invented that allow that vehicle to be sold at less than1/3 its original price.

In fact, the logical way to test these very assertions would be to permit or encourage some schools/districts to experiment with alternative compensation strategies, and other “reforms,” and to include these schools and districts among those employing other strategies (production technologies) in a cost model, and see where they land along the curve. That is, do schools/districts that adopt these strategies land in a different location along the curve? In fact, some schools and districts do experiment with different strategies and those schools carry their relevant share of weight in any statewide cost model. Thus far, what we seem to be seeing is that the more productive experimental approaches being used a) aren’t that bold and b) cost quite a bit!

Pure speculation that some alternative educational delivery system would produce better outcomes at much lower expense is certainly no basis for making a judicial determination regarding constitutionality of existing funding, and is an unlikely (though not unheard of) basis for informing statewide mandates or legislation.  Cost model estimates, as well as recommendations of professional judgment and expert panels can serve to provide useful, meaningful information to guide the formulation of more rational, more equitable and more adequate state school finance systems.


[1] Duncombe, W., Yinger, J. (2008) Measurement of Cost Differentials In H.F. Ladd & E. Fiske (eds) pp. 203-221. Handbook of Research in Education Finance and Policy. New York: Routledge.  Duncombe, W., Yinger, J. (2005) How Much more Does a Disadvantaged Student Cost? Economics of Education Review 24 (5) 513-532. Duncombe, W.D. and Yinger, J.M. (2000).  Financing Higher Performance Standards: The Case of New York State. Economics of Education Review, 19 (3), 363-86. Duncombe, W., Yinger, J. (1999). Performance Standards and Education Cost Indexes: You Can’t Have One Without the Other. In H.F. Ladd, R. Chalk, and J.S. Hansen (Eds.), Equity and Adequacy in Education Finance: Issues and Perspectives (pp.260-97). Washington, DC: National Academy Press. Duncombe, W., Yinger, J. (1998) “School Finance Reforms: Aid Formulas and Equity Objectives.” National Tax Journal 51, (2): 239-63. Duncombe, W., Yinger, J. (1997). Why Is It So Hard to Help Central City Schools? Journal of Policy Analysis and Management, 16, (1), 85-113. Imazeki, J., Reschovsky, A. (2004b) Is No Child Left Beyond an Un (or under)funded Federal Mandate? Evidence from Texas. National Tax Journal 57 (3) 571-588.

[2] Downes (2004) What is Adequate? Operationalizing the Concept of Adequacy for New York State. http://www.albany.edu/edfin/Downes%20EFRC%20Symp%2004%20Single.pdf

[3]Hanushek, E. (2005, October). The alchemy of ‘costing out’ and adequate education. Paper presented at the Adequacy Lawsuits: Their Growing Impact on American Education conference, Cambridge, MA. Costrell, R., Hanushek, E., & Loeb, S. (2008). What do cost functions tell us about the cost of an adequate education? Peabody Journal of Education, 83, 198–223.

[4] For elaboration on this argument, see: Costrell, R., Hanushek, E., & Loeb, S. (2008). What do cost functions tell us about the cost of an adequate education? Peabody Journal of Education, 83, 198–223

Posted in: Uncategorized