By Dr. Blake Haselton and Michael Davis(*)
From Foresight, Vol. 11, No. 1
published 2004
In a 2003 Foresight article (“Do Resources Produce Results?”), University of Kentucky political science professor Phillip Roeder questioned whether increased spending on education since the enactment of the Kentucky Education Reform Act of 1990 (KERA) has led to increased student achievement in Kentucky. Here we examine this important public policy issue in more detail.
Prior to KERA, Kentucky’s educational system was among the worst in the nation. In 1988 Kentucky ranked last in adult literacy and high school completion with rates of 69 percent and 53 percent, respectively, and at 11 percent, nearly last (49th) in the percent of adults with a four-year college degree. The state ranked 48th in K-12 per-pupil spending, 41st in pupil-teacher ratio, and 38th in average teacher salary.(1)
In June 1989, in response to a lawsuit brought by the Council for Better Education Inc., the Kentucky Supreme Court issued what many have termed a landmark decision, declaring the entire system of common schools in Kentucky unconstitutional. The Court concluded that the state’s system of common schools did not comply with the constitutional requirement that the General Assembly provide an efficient system of common schools throughout the state. The Court ruled that school systems must be funded adequately to achieve their goals and become substantially uniform throughout the state so “every child is provided with an equal opportunity to have an adequate education.”(2) As envisioned by the Court, an adequate system of education as mandated by the state constitution must provide every child with at least seven capacities, including communication skills sufficient to function in society, satisfactory preparation to pursue a career, and skills necessary to compete favorably with students from other states.
The General Assembly’s response to the court decision was swift and impressive. Eleven months after the court determined the entire system of education unconstitutional and directed the legislature to recreate it, the governor signed House Bill 940, the Kentucky Education Reform Act (KERA), into law.
Many of the questions raised by Roeder concerning education funding are similar to those articulated before the Kentucky Supreme Court in Rose v. Council. KERA required significant increases in educational funding and caused many people to question whether these increased funds would have any impact on student performance. Consequently, when evaluating whether additional funding for educational resources have an impact on educational outcomes, the first question should be, “Has progress been made in educating Kentucky’s children over the last decade?” Or, to put it more bluntly, “Is KERA working?”
Regardless of one’s feelings about KERA, most would admit that the increased education funding resulting from KERA has led to significant improvement in Kentucky’s education system. Since our education was realigned and made our state’s highest priority, student achievement has been positively affected. Kentucky has since soared on national achievement tests. For example, from 1992 to 2002, Kentucky increased six points in reading on the 4th Grade National Assessment of Educational Progress (NAEP) with only five other states showing larger gains.(3) From 1998 to 2002, Kentucky increased three points on 8th Grade NAEP reading, which exceeded the national average increase, and only nine other states had larger achievement gains. Furthermore, results from Kentucky’s 2002 CTBS/5 assessment showed sustained progress in reading, language arts, and mathematics, with Kentucky now above the national average at all three testing stages: grades 3, 6, and 9.(4)
Increasingly, graduates of Kentucky’s high schools are leading productive and successful lives. In 2000, Kentucky led the nation in the percentage increase in the high school completion rate for 25- to 34- year olds.(5) Statistics from the Kentucky Department of Education demonstrate that college enrollment has increased; 10.3 percent more graduating seniors enrolled in college in 2001 than in 1993.(6) Moreover, in 2000, 95.4 percent of our graduating seniors were pursuing postsecondary education, employment, or military service.
Where Kentucky once had a system of education that no state would want, today Kentucky’s is a model educational system that other states are emulating. It has consistently placed in the top five states for testing and accountability in Education Week’s “Quality Counts” report. Where once our educational system was in a state of crisis, today Kentucky is recognized as an innovator and a leader in education. Our education system is no longer a disgrace. Instead, the improvement in our educational system is something that Kentuckians can view with great pride. However, because Kentucky began this long-range endeavor at or near the national bottom, our relatively rapid upward movement still leaves us with much ground to gain.
Another way to evaluate whether additional funding for educational resources has an impact on educational outcomes is to examine the research on this subject. Contrary to Roeder’s assertion that, “research that systematically and empirically links education resources to results is in short supply,”(7) a rich and bountiful supply of research exists directly linking expenditures for educational resources to increases in student achievement. Indeed, the Kentucky Supreme Court in Rose v. Council reached the same conclusion: “Uniform testimony of expert witnesses at trial, corroborated by data, showed a definite correlation between the money spent per child on education and the quality of the education received.”(8)
A number of methodologically sound studies have shown that expenditures for certain programs and services do positively influence student achievement. For instance, the model experiment for class size reduction is Tennessee’s Student-Teacher Achievement Ratio (STAR) initiative.(9) The format of the experiment was championed as both “systematic and empirical,” with renowned education professor Bruce J. Biddle noting, “The Star Project was arguably the largest, best-designed field experiment that has ever appeared for education.”(10)
The results of this initiative show that education resources for class-size reduction are strongly correlated with higher achievement in student performance. Kindergarten students in small classes were found to be three quarters of a month ahead of the students in standard-sized classes.(11) The small-class-size advantage for students after the first grade was nearly two months, and after the end of the fifth grade, students in smaller classes were five months ahead of their peers in standard-sized classes. Because the results of the STAR Project were both substantial and academically rigorous, more than 30 states have instituted some type of class-size reduction effort.
Although progress had been made in Kentucky, adequate funding to assure class-size reductions has not been. As recently as the 2001-02 school year, elementary class sizes in Kentucky averaged more than 25 students.(12)
This evidence demonstrates that methodologically sound, peer-reviewed research has linked educational resources to student achievement. And class-size reduction studies are not the exception; similar benefits can be seen for other educational expenditures, such as full-day kindergarten, but all require adequate resources.
Given the evidence we have presented thus far, how did Roeder reach the conclusion that a weak link exists between funding and student outcomes? We believe that his analysis of the education data is incomplete. Roeder did not consider many factors that influence the relationship between per pupil expenditures and district performance results. For instance, are the populations in the highest-performing districts similar to the populations in the lowest-performing districts? Was poverty, or population migration into or out of the district, or the size of the district, considered? Obviously, assuming similar populations, larger districts will require less per-pupil funding due to sheer economy of scale. Were student demographics considered? Students in poverty, students with learning disabilities, and students learning English as a second language require significantly more resources to achieve proficiency in educational outcomes than students without these characteristics.
As an illustration, the data from which Roeder produced his Table 1 can be reanalyzed with respect to poverty, measured in this case by the percentage of students eligible for free lunch under federal guidelines. The eligibility rate for Kentucky students statewide is approximately 38.8 percent. All but two of the highest-performing school districts fall below this rate, whereas all of the lowest-performing school districts exceed it. In fact, a student living in one of the lowest-performing districts is 178 percent more likely to be in poverty than a student in one of the highest-performing districts. Students in poverty have been shown to require more services and more help, and therefore more funding per pupil in order to achieve similar outcomes. Regrettably, Roeder does not account for the socioeconomic status of these students. By not accounting for the district-level percentage of students eligible for free lunch, a reliable proxy for poverty, Roeder’s results present the counterintuitive impression that higher levels of education funding lead to lower student achievement.
Table 1: Comparison of Highest- and Lowest-Performing School Districts, 1999-2000
The Kentucky constitutional delegates of 1891 made a radical decision: they concluded that education should assume priority over all other public goals for the Commonwealth. Considering the age, one in which the skyscraper, the washing machine, and the escalator were novelties, this was truly a visionary concept. Today, this argument is no longer radical but so mainstream as to be nearly universally accepted. Today, we take for granted that a person without an adequate education—without the ability to read, write, and communicate effectively, without the ability to solve problems and analyze alternatives, without a basic knowledge of our political, social, and economic systems—has little chance to find meaningful employment, lead a productive life, or participate fully as a contributing citizen. Today, we hold that an adequate education is essential for everyone. This is not merely an accepted public policy aspiration, it is a legally binding requirement under Rose v. Council: “All children in Kentucky have a constitutional right to an adequate education.”(13) Just as we recognize that individual success requires an adequate education, we also must recognize that our state’s very prosperity, in the fullest sense of the word, depends upon the adequacy of the education we provide each and every child in the Commonwealth.
* Dr. Haselton is Superintendent of the Oldham County schools and Vice President of the Council for Better Education. Michael Davis is a senior at Vanderbilt University and served as a summer Administrative Intern in Policy Studies at the Oldham County Board of Education. Return to text.
1 Prichard Committee for Academic Excellence, Gaining Ground: Hard Work and High Expectations for Kentucky Schools, (Author: Lexington, 1999). Return to text.
2 Rose v. Council for Better Education, Inc., 790 S.W. 2d 186 (1989). Return to text.
3 National Assessment of Educational Progress (NAEP) website http://nces.ed.gov/nationsreportcard/sitemap.asp. Return to text.
4 Kentucky Department of Education (KDE) website http://www.kde.state.ky.us. Return to text.
5 U.S. Bureau of the Census, 2000 Decennial Census (SF3, PCT25). Return to text.
6 KDE, “Non-Academic Data: 1993 – 2001, Region and State Totals,” July 2002. Return to text.
7 Phillip W. Roeder, “Do Resources Produce Results?,” Foresight 10.1 (2003): 6-9. Return to text.
8 Rose v. Council 198. Return to text.
9 Attempting to demonstrate the effectiveness of class size reduction before committing additional funding, Tennessee legislators authorized a four-year study in which kindergarten, first, second, and third grade students and teachers were randomly assigned to: (a) classrooms of 13 to 17 pupils, (b) classrooms of 22 to 25 pupils, or (c) classrooms of 22 to 25 pupils and the teacher was assisted by an aide. Curriculum and standardized tests were issued to assess student performance in reading, mathematics, and basic study skills for 6,500 pupils at approximately 80 schools. Return to text.
10 B. Biddle and D. Berliner, “What Research Says About Small Classes and Their Effects,” WestEd (PP-02-01) Winter 2002: 1-24. Return to text.
11 J. Finn and C. M. Achilles, “Tennessee’s Class Size Study: Findings, Implications, and Misconceptions,” Education Evaluation and Policy Analysis 21.2 (1999): 97-110. Return to text.
12 KDE, “2001-2002 School Profile” http://170.180.8.163/district%20profile/mainpage.cfm. Return to text.
13 Rose v. Council 213. Return to text.