Ph.D. Management

I received my Ph.D. in Management from the University of Massachusetts at Amherst in 2005.

Dissertation: Thesis

The success of the Massachusetts Education Reform Act of 1993 (“MERA”) can be evaluated by testing the following propositions:

  1. Educational opportunity became more equal as a consequence of MERA.
  2. The degree to which socio-economic status is a determinant of educational outcomes in Massachusetts has decreased as a consequence of MERA.
  3. Education standards have been raised and educational outcomes have improved as a consequence of MERA.

using data for the years from 1988 to 2002 and for a representative sample of Massachusetts’ school districts and using a set of linear programming models collectively known as Data Envelopment Analysis (“DEA”).

The results were inconclusive...

In hindsight...

While I was very frustrated by the inconclusive results of my research I still believe in equal access to education and the objectives of "MERA".

Unfortunately when allocating billions to any project politicians must measure the outcomes: "Accountability".

As in physics: measurement changes the process, providing incentives to governments, towns, schools, teachers, parents and students to either cheat or game the system.

"High Stakes Testing" only adds to the problems and leads to a culture which teaches to the test and teaches how to take tests, and in the process loses sight of learning and the retention of knowledge or other skills.

Dissertation: Abstract

The Charnes, Cooper, Rhodes ratio DEA model (“CCR”) is used, with panel data from 1988 to 2002 from a large sample of Massachusetts school districts, to test three propositions concerning the consequences of the Massachusetts Education Reform Act of 1993 (”MERA”). First, did the degree of positive correlation between Socio-Economic Status (“SES”) and educational outcomes decrease, secondly did educational opportunity become more equal among towns in Massachusetts, and finally were education standards raised and educational outcomes improved overall?

The CCR model is a Linear Programming method that estimates a convex production function using Koopmans’ (1951) definition of technical efficiency and produces the radial or ratio measurements of efficiency proposed by Farrell (1957). It has been widely used in Education Production Function research in which costs and quantities of inputs to the education process are evaluated against outcomes such as standardized test scores – as have econometric and statistical techniques.

The pursuit, through state and federal courts, of equitable funding, allied to the belief that smaller class sizes improve outcomes, has made the provision of K-12 education expensive. The belief that educational outcomes are in constant decline has led to calls for “Accountability” and to “standards” reform in which statewide curricula are established as the standards against which students are regularly tested.

Standards reform was combined, in MERA, with reform of state aid formulas and additional state funding for education, to ensure a minimum basic level of education pursuant to the decision of the Massachusetts Supreme Court in McDuffy v. Robertson.

The one certain relationship revealed by decades of education production function research is a high level of positive correlation between SES and educational outcomes. If MERA succeeded in ensuring a higher basic level of education than that provided before, then the positive correlation between SES and outcomes should have weakened over time as the education of less well SES-endowed children improved. The CCR model was used to measure the correlation between multiple variables on one side and multiple variables on the other side in a single measure. The analysis showed that a strong positive correlation exists and that, in so far as this relationship changed, it appeared to strengthen rather than weaken after MERA. The CCR model was also used to determine if there were changes in the distribution of per pupil expenditures after MERA and, with a time shift between test scores as inputs and test scores as outputs, to determine whether outcomes improved between 1988 and 2002. The analysis suggested that the distribution of expenditures changed for the better and that educational outcomes deteriorated after MERA. This deterioration seems to be closely related to the changes in the proportion of all students, in a grade, actually taking the tests.

In summary, there is little evidence that MERA achieved anything in particular and no basis upon which to argue that it achieved nothing.

Dissertation Chapter downloads (pdf files)