An Evaluation of the Kentucky Education Reform Act

William H. Hoyt

The Kentucky Education Reform Act (KERA) caused drastic changes in the way education is provided in Kentucky. Because KERA attempted to equalize funding and dramatically alter the curriculum and governance of schools, it is appropriate to see how effective the reforms have been in affecting spending and educational outcomes. Comparisons of measures before and after KERA show that it has equalized funding among school districts, with some districts gaining substantially in their funding of students. Teacher salaries and student/teacher ratios have also improved relative to national averages since the introduction of KERA. Educational outcomes have not improved, however. There has been no appreciable gain in test scores or graduation rates in Kentucky relative to other states, and there is no evidence that KERA has increased college attendance rates of Kentucky high school graduates.

Introduction

Among the many states that had their public school financing systems invalidated by court rulings, the changes required of the Kentucky system by the courts as a result of Rose v. Council (1989) were among the most radical. Not only did the Kentucky Supreme Court find the financing of schools in the state to be unconstitutional, but it also ruled that the "entire system of common schools is unconstitutional" and required the legislature to "recreate, re-establish" the entire system of public education. In a remarkably brief interlude, the Kentucky General Assembly, in response to the court decision, passed the Kentucky Education Reform Act (KERA) of 1990.

KERA has attracted a great deal of attention in the popular press and public discussions in both Kentucky and throughout the United States. It has been criticized and praised and played a major role in numerous state elections, including the most recent gubernatorial election. The vast attention focused on KERA, in contrast to finance reform acts in other states, must in part be attributable to the broad scope of the legislation. It was not simply striking down a finance system that was not providing enough resources for poorer districts — the court ruling emphasized equality in funding, not simply additional funding. Nor was the court ruling solely a statement about the funding. It was also about the outcomes of education. Again, the court emphasized the need for education to be reformed to equalize not only spending but also performance.

Our objective is to provide an evaluation of KERA by comparing educational funding and, when possible, performance before and after KERA. We evaluate it in light of the emphasis of the original court case on equity in both financing and performance. In addition, to give more perspective on the impact of KERA, we contrast the changes in primary and secondary education in Kentucky to changes throughout the nation. In the next section, we begin by providing a brief discussion of the legislation that created KERA and some of the institutional features of KERA. In Section 3 we examine the impacts of KERA on funding of education emphasizing both changes in the distribution of funding within Kentucky and how the level of spending and equality in spending in Kentucky compares to the U.S. average and some of our neighboring states. In Section 4 we focus on measures of educational performance and outcomes and as with funding, our emphasis is both on the changes in the equality and level of performance in Kentucky schools. Again, to give perspective for at least some measures, such as dropout rates, we compare Kentucky to the rest of the nation.

KERA and Education in Kentucky

The Kentucky Education Reform Act of 1990 (KERA, House Bill 940) was not simply a change in the financing of primary and secondary education, though it radically altered the financing of education in Kentucky. KERA is unique in its attempts to integrate school finance reform with reforms in the curriculum and school governance. Instead of simply providing additional funds to local school districts, KERA provided a dramatically expanded role for the Commonwealth in mandating curricula, evaluating school performance, and placing restrictions on district employment and compensation. Broadly, then, we might classify reforms in three categories: curriculum, governance, and finance.

Curriculum

The major innovations related to curriculum are the restructuring of primary schools (particularly kindergarten to third grade); the implementation of standardized student assessment throughout the Commonwealth; the encouragement of increased use of technology; and the introduction of school-based decision making, which provides a formal structure for involving parent representatives in decisions regarding curriculum, personnel, and instruction.

The General Assembly in 1992 and again in 1994 expanded on KERA’s requirement of replacing grades kindergarten to third with a "primary school program." The 1992 General Assembly listed "critical attributes" of primary schools that include "multi-age and multi-ability classrooms," "professional teamwork," and "qualitative reporting methods." Essentially, in the General Assembly’s interpretation of KERA, the traditional separation of students by grade (age) is eliminated (or at least reduced) with students of both different ages and abilities found in the same classroom. The traditional classroom with a single teacher is to be replaced (at times) by team teaching. Finally, and perhaps most controversially, qualitative assessment (written reports assessing a student’s performance) replaces the traditional letter grade system.

KERA created a new system of assessing students and schools: the Kentucky Instructional Results Information System (KIRIS). Based on KIRIS, a "performance-based" approach, the progress of schools is monitored. Schools that are succeeding, that is, schools that have had increases in average KIRIS scores, are financially rewarded, along with their teachers. KIRIS is not a test in the tradition of multiple choice objective tests. Instead, students create a "portfolio" of writing and mathematics work in the 4th (writing only), 8th, and 12th grades. Thus, students may include lengthy essays in their portfolios, although the content of these essays is based on a list of topics emphasized by the Kentucky Department of Education.

School councils were created by KERA as an attempt to decentralize policies and plans for schools. These councils consist of six people: the school’s principal, three teachers, and two parents with teachers and parents being elected representatives. These councils are to make decisions on curriculum, instruction and instructional materials, discipline, extracurricular programs, and the school budget. In addition, the council is charged with hiring principals if that position vacant and consult with the principal on the hiring of teachers.

Increased use of technology, including computers, was provided special funds in KERA, and acceptable use of these funds is outlined in the Kentucky Education Technology System (KETS). To obtain matching funds from KETS, districts must have prepared and submitted a plan for use of technology in their district.

Governance

In addition to the school councils discussed earlier, KERA restructured the organization of primary and secondary education at the state level, creating a Board of Elementary and Secondary Education and a Commissioner of Education. At the district level, the legislation was designed to reduce employment based on nepotism or political involvement. In addition, the local superintendent’s power vis--vis the local school board was increased with the Commissioner of Education having veto power over any dismissal of the superintendent by the local board.

Finance

Kentucky, through foundation grants and power equalization programs, had extensive state funding assistance of local education prior to KERA. Under the previous funding system, however, local support varied dramatically. For example, the Kentucky Office of Education Accountability reported that in 1989-90, property wealth per pupil varied from $39,138 to $341,707; local equivalent mill rates varied from $0.229 to $1.119; and local revenue per student varied from $80 to $3,716. While state aid also varied from $1,750 to $2,753 per student, average per pupil spending for instruction still varied from $1,499 to $3,709.1

The focus of KERA was to reduce these disparities in educational funding. To do so, a new funding program called Support Educational Excellence in Kentucky (SEEK) was established. Under this program, districts are to receive a guaranteed level of revenue per student. In 1990-91 this base was $2,305 per pupil and gradually increased to $2,570 per pupil in 1995-96. This base is adjusted (increased) for at-risk children (measured by participation in federal school lunch programs), transportation, and exceptional students. While the state guarantees this amount of revenue, the district must share in the financing by providing a minimum level of effort equivalent to $0.30 per $100 of property value. This funding, however, need not be obtained through the property tax. Then the state funding, the adjusted base guarantee equals the base plus additional funding for at-risk children, transportation, and exceptional student, minus the local effort.

In addition to this uniform funding base, each district can increase its funding by up to 15 percent of the base while receiving state funding if its property value per student is less than 150 percent of the state average. If its property tax base is below this amount, the state provides state funding to guarantee revenue equal to the amount collected on this property value. This is referred to as Tier I funding. Through Tier II funding, the district can collect up to an additional 30 percent beyond the base and Tier I funding, but it will receive no matching state funds. This funding must also be approved by a vote of the electorate in the district. Finally, the state also provided a guaranteed annual minimum increase in state funds (8 percent in 1991-92 and 5 percent in 1991-92) and an annual maximum increase (25 percent).

How Kentucky Compares Before and After KERA

While KERA has led to many changes in both the financing and the actual delivery of educational services, the major increases in educational spending were not new for Kentucky. The increase in real spending per student from 1989-90 to 1995-96 in Kentucky, 30 percent, was the highest among the 50 states, which had an average increase in real spending per student of 3.8 percent. As Table 1 indicates, Kentucky moved from ranking 45th in spending per student to 31st during this period. While this was a major increase in spending, it is not unprecedented for Kentucky. During the five-year period from 1984-85 to 1989-90, real spending per student also increased by 22 percent, well above the national average of 16.8 percent. In fact, from 1969-70 to 1995-96, the 148.7 percent increase in real spending per student in Kentucky was the highest among all 50 states and significantly above the national average increase of 84.2 percent. Thus, while KERA was a major increase in spending and an increase that came very quickly (10 percent in one year), even prior to KERA, Kentucky had seen major increases in educational spending.

Table 1
Real Expenditures per Student in theUnited States, 1988-89 and 1995-96

Table 1 (for frames enabled browsers)                   Table 1 (no frames)

As we emphasized earlier, KERA was the legislative response to a court order about the equality of educational spending and performance, not the level of spending. While educational spending had been increasing in Kentucky prior to the legislation, KERA changed the distribution of these increases and dramatically changed the relationship between educational spending and revenues and district property wealth and income.

We can see how KERA changed the distribution of educational funds in several ways. We first compare the distribution of funding among districts in Kentucky to some neighboring states. To do this we use a measure of inequality, the Gini coefficient. The Gini coefficient is a measure of inequality (generally income) that ranges between 0 and 1. If the Gini coefficient equals 0 there is no difference in distribution across groups — everyone receives the same funding. On the other hand, a coefficient of 1 means one group (individual) has all the resources.

In our case, we examine the distribution of educational spending per student among school districts in the state and compare that to four states bordering Kentucky. As Table 2 shows, the Gini coefficient in 1989-90 ranks Kentucky third among the five states but by 1994-95, Kentucky is second only to West Virginia in indicating equality of funding and decreased significantly between the two periods.

TABLE 2

Gini Coefficients for Kentucky and Selected Adjacent States,
1989 - 1990 and 1994 - 1995
 
  Gini Coefficient   Gini Coefficient  
 

1989 - 1990

Rank

1994 - 1995

Rank
Kentucky 0.1001 3 0.0700 2
 
Indiana 0.0908 2 0.0944 4
Ohio 0.1642 5 0.1262 5
Tennessee 0.1012 4 0.0922 3
West Virginia 0.0343 1 0.0290 1

In Table 3, expenditures per student are given for the academic years 1994-95, 1989-90, and 1985-1986. Inspection of Table 3 suggests that while generally the largest (percentage) increases in funding went to the districts spending less both between 1985 and 1989 and between 1989 and 1994, the variation in increases was much more pronounced between 1989 and 1994. For example, from 1985-86 to 1989-90 real spending per student in Kentucky increased by 18.7 percent while it increased by only 13.2 percent between 1989-90 and 1994-95, despite the fact that real increases in educational spending across the state were much greater during the later period. One notable exception appears to be the Jefferson County school district, which saw only a 6.7 percent increase in real spending in both periods. At the other end, while those districts that had large increases in spending after KERA tended to have larger than average increases prior to KERA, these increases were not nearly as large. For example, Somerset Independent school district had the greatest increase in spending from 1989-90 to 1994-95 at 71.6 percent but only had a modest 15.9 percent increase in spending between 1985-86 and 1989-90.

Table 3
Expenditures per Student in Kentucky Public School Districts, 1985 - 1995

Table 3 (for frames enabled browsers)                   Table 3 (no frames)

KERA has continued and expanded on a trend in redistribution of resources among the school districts and, in a very real sense, eliminated an inequity in educational spending, if not educational outcomes. We can see the impact of KERA by examining both the relationship between revenues per student and property value per student. In Figure 1a we depict the relationship between total revenue per student and assessed property value per student (in $1,000) for the 1988-89 academic year. As the line in the figure shows, we find a strong positive relationship between revenue per student and property value per student in a district. Based on the relationship we estimate, we would predict a difference in spending of $1,336 (in 1994 dollars) per student based on differences in property value per student between the district with the highest property value per student (Harlan County) and the lowest (Elliot County).

a3f1a99.jpg (20753 bytes)

Figure 1b shows the same relationship between revenues per student and property value per student for the 1994-95 academic year. As the figure shows, during this year there is no apparent relationship between revenue per student and property value per student. While the slope of the line in Figure 1b is slightly positive, it is not statistically different from zero, and, if we use it to predict the difference in spending between the districts with the highest (Frankfort Independent) and lowest property values (Franklin County), this difference is only $20.

a3f1b99.jpg (22848 bytes)

Figure 2  presents the same type of comparison between revenue per student and per capita income in the school district. Again, as shown in Figure 2a, in the 1988-89 school year there was a strong relationship between per capita income and revenue per student, with an additional $1,000 in per capita income increasing revenue per student by approximately $61. The predicted difference between the district with the highest income per capita (Oldham County) and the lowest per capita income (Owsley County) is $852. From Figure 2b it is apparent that there is not a strong relationship between district per capita income and revenue per student in 1994-95. In fact, there is a weak (statistically insignificant) negative relationship between per capita income and revenue per student.

a3f2a99.jpg (18937 bytes)

a3f2b99.jpg (22851 bytes)

While KERA has led to significant increases in educational spending, what is most significant about its impact is not the magnitude of these increases but which districts received them. KERA, in a very real sense, has eliminated the inequity in educational spending in the state. While there are still differences in spending throughout the state, these differences are of a smaller magnitude than they were prior to KERA. Further, these differences are unrelated to either income or property wealth — poor districts, in either income or property wealth, can be expected to receive as much funding as wealthy districts.

How Has the Money Been Spent?

While much of the media attention and discussion that preceded KERA was focused on the unequal funding of schools and corrupt assessment practices, attention was also paid to how districts used their resources. Concerns about the appropriate use of educational monies are also apparent in the KERA legislation itself. In particular, KERA has increased standardization of curricula, attempted to increase parent and teacher input in governance, and placed restrictions on employment practices in attempts to eliminate nepotism and politically motivated appointments.

While it is beyond the scope of this report to provide a thorough examination of how successful KERA has been in ensuring that funds are used productively, we can offer some general information on how the additional funds KERA has provided have been spent.

Comparison between Kentucky and the Rest of the U.S.

Table 4 provides information regarding personnel and employment in primary and secondary education aggregated to the state level. Again, we make comparisons among the states both before and after the implementation of KERA. The legislation did reduce class sizes in Kentucky, from 17.7 students per teacher in 1988-89 to 16.9 students per teacher in 1994-95. However, during a period when per pupil spending increased by 30 percent, the highest among all the states, this was only a 4.5 percent reduction in the student-teacher ratio, ranking 13th among the states.

Table 4
Student/Teacher Ratio, Teacher Earnings, and Teachers as a Percentage of Total Staff, 1988 - 1995

Table 4 (for frames enabled browsers)                   Table 4 (no frames)

At the same time, teacher salaries in Kentucky did increase from an average of $32,688 (in 1996 dollars) in 1988-89 to $33,950 in 1996-97. This 3.86% increase in salaries was the 11th highest among the states and increased Kentucky’s rank from 38th to 30th in the ranking of average teacher salaries. Again, while this was a significant increase in teacher salaries relative to other states, only a small share of the increase in educational spending from KERA can be attributed to increased teacher salaries.

While Kentucky has increased its number of teachers relative to students during a time when the majority of states have seen fewer teachers per student, teachers as a percentage of total public school staff have decreased significantly, as Table 4 illustrates. Kentucky has traditionally been a state with a low percentage of its staff employed as teachers, ranking 45th in 1988-89 with about 50 percent of its staff employed as teachers. By 1995-96, however, the state ranked last among the states with only about 46 percent of its staff being teachers.

While Kentucky has increased its number of teachers relative to students during a time when the majority of states have seen fewer teachers per student, teachers as a percentage of total public school staff have decreased significantly, as Table 4 illustrates. Kentucky has traditionally been a state with a low percentage of its staff employed as teachers, ranking 45th in 1988-89 with about 50 percent of its staff employed as teachers. By 1995-96, however, the state ranked last among the states with only about 46 percent of its staff being teachers.

What are the Results?

Perhaps the one feature of KERA that distinguishes it more from earlier educational finance reform acts in other states is the emphasis on equality in student performance and achievement, and not simply equality in funding.

Attempts to achieve equality in performance required a means of monitoring and measuring the performance of students. This led to the development of the Kentucky Instructional Results Information System (KIRIS). KIRIS is an ambitious testing system, relying not simply on objective questions but also requiring extensive writing and even group projects.

The results of KIRIS have been criticized for a number of reasons, however. Evidence of fraud has surfaced in some districts2 , possibly as a result of the bonuses provided to schools and teachers based on performance on the KIRIS. A recent study by the Rand Corporation argues that the increase in KIRIS scores is an overestimate of the increase in performance by Kentucky students. The study supports this conclusion by noting that the increases in KIRIS scores are much higher than the increase in scores for Kentucky students on the National Assessment in Educational Progress (NAEP) during the same period, as well as the gains in scores received by Kentucky students taking the American College Test (ACT), a standard test required for college entry. The study suggests that much of the gain might be related to teaching directed at questions that are reused over time.

We provide a few different measures of performance to compare Kentucky to other states, including the scores from the Math NAEP, graduation rates, and ACT scores. Table 5 provides NAEP math scores for both 4th and 8th grades for 1990 and 1996 as well as measures of the changes in these scores. Kentucky’s score on the 8th grade Math NAEP has risen relative to the national average, ranking 5th in percentage increase among the 28 states in which the test was administered during both years. In 1996, Kentucky ranked 28th in scores of the 43 states administering the exam. The 4th grade test began in 1992, and scores in Kentucky increased by 2.8 percent from 1992 to 1996, compared with an average U.S. increase of 3.2 percent. Again, Kentucky ranked 28th among the states administering the exam.

Table 5
National Assessment in Education Progress (NAEP) Scores and High School Completion Rate, 1990 - 1996

Table 5 (for frames enabled browsers)                   Table 5 (no frames)

The ACT scores in 1995 and 1998 give a similar story, with Kentucky having a slight increase in scores from 20.1 to 20.2, while the national average increased from 20.8 to 21.0. While Kentucky’s scores are lower than the national average, some caution in interpreting these results is warranted since a much higher percentage of Kentucky students (67 percent in 1998) take the exam than the national average (37 percent). The higher the percentage of students taking the exam, the lower the scores we should expect.

Also in Table 5 are figures on the percentage of 18- to 24-year-olds who have completed high school for the 1989-91 period. We see that during this period, and in the 1994-96 and 1996-98 periods, the figure for Kentucky is below the national average for three three-year periods: 1989-1991, 1994-96, and 1996-98.3 Kentucky’s high school completion rate is below the national average during all three periods, and there have only been modest gains in Kentucky since KERA. Of course, when considering these figures, it is important to remember that some of the individuals represented by these graduation rates may not have been in school under KERA for more than a few years or even at all.

Overall, and not too surprisingly, these measures suggest that at least thus far, KERA has had little impact on the performance of Kentucky students and schools compared to the rest of the United States.

Comparisons Among Districts

Given the criticisms of KIRIS and limited information at the district level on test scores, we focus on two other important measures of school performance: the dropout rate and the percentage of graduates planning to attend college. Given the strong evidence of the impact of high school completion on earnings and the substantial difference in earnings between high school graduates and college graduates, decreases in dropout rates and increases in college attendance should signal increases in future earnings.

Table 6 gives the percentage of 12th graders planning to attend college for 1993 and 1997 as well as the change in the percentage between the two years.4  As the table shows, there has been only a slight increase in planned college attendance for the state between 1993 and 1997. Some districts, however, have had significant increases in planned college attendance. Seven of the ten districts with the greatest increases in the percentage planning to attend college were districts who ranked 155 or below (out of 173 districts in the state) in terms of the percentage of students planning to attend college in 1993.

Table 6
College Attendance and Dropout Rate in Kentucky Public School Districts,
1991 - 1997

Table 6 (for frames enabled browsers)                   Table 6 (no frames)

The dropout rate for 1991 and 1995 for the 7th through 12th grade is also included in Table 6. The overall dropout rate for Kentucky has actually increased for this period by almost one percent. Of course, educational quality is not the only reason that dropout rates may change. For example, changes in economic conditions and alternatives to schooling will also have a significant impact on the dropout rate. Again, while the dropout rate has increased overall, a number of districts have seen significant decreases. In fact, eight of the ten districts with the greatest reduction in dropout rates were districts that were in the top ten in terms of highest dropout rates in 1991-92. While it is more likely that those districts that have the greatest dropout rates could decrease them the most, this need not happen, and the fact that it has should be considered encouraging.

The relationship between spending on education and educational performance and outcomes is a tenuous one, with many studies finding little or no evidence of the link. While it is beyond the scope of this study to rigorously and formally analyze the impact of KERA funding on educational performance in Kentucky schools, we did attempt to determine whether the additional spending from 1998-89 to 1994-95 had, without accounting for other factors, any significant impact on changes in planned college enrollment or dropout rates. For college attendance, we found no evidence of any impact of additional spending per student. While we did find there was a negative correlation between the change in dropout rate and change in spending per student, once we controlled for 1991-92 dropout rate, this correlation disappeared. Thus, for these two measures, at least for this short time period, we have found little evidence of any significant impact of the spending on educational outcomes.

The NAEP results can not be directly used to compare scores in districts in Kentucky because they are derived from a relatively small sample of students who take the test. But the NAEP does report results based on type of location for each state with results given for central cities, urban fringe/large town, and rural/small town. For the results in mathematics, for both the 4th and 8th grades, the scores in rural/small towns were significantly lower than in central cities or the urban fringe/large towns in 1992. While scores did not significantly change between 1992 and 1996 for central cities and the urban fringe/large towns, however, they did change for the rural/small towns. By 1996, there was no significant difference in mathematics scores between the regions. For reading scores in both 1992 and 1994, there were no significant differences between regions, nor were there any significant differences between the two years.6

Conclusion

The reforms instituted through KERA have dramatically changed the financing of education in Kentucky. Prior to KERA, Kentucky had seen large increases in educational spending and significant state involvement in attempts to equalize spending with spending increases of the order of those in KERA. Perhaps the most dramatic impact of KERA has been its effect on equalizing spending throughout the state. While spending does still vary among districts, this variation no longer bears any relationship to the traditional determinants of educational spending with decentralized educational financing, local income and property wealth. Now districts with low incomes or little property value per student are just as likely to have high educational spending as are wealthy districts.

What is less clear has been how the increased spending from KERA has been spent and how effective this spending has been thus far. Clearly, technology has been a major emphasis on spending. What has not been emphasized has been spending on teachers. Kentucky has increased teacher salaries in real terms and at a faster rate than the national average, and the student-teacher ratio also has fallen. However, the share of current spending allocated to teacher salaries has fallen significantly since the advent of KERA, and Kentucky ranks last in the percentage of its public school staff who are classroom teachers. Given evidence suggesting that teacher quality has significant impacts on student performance, more attention to and funding for improvements in teacher quality might be warranted.

KERA was designed to not only equalize funding but also to equalize educational outcomes. On this score, KERA appears less successful. Kentucky seems to have gained little, if at all, when compared to other states based on test scores or graduation rates. Within Kentucky, while some districts have had significant increases in college attendance by graduates and reductions in dropout rates, there is no evidence suggesting that these gains are related to KERA, or at least not increased funding from KERA.

While the lack of any measurable increase in achievement or performance since KERA might appear discouraging, it should not be surprising. Even if the reforms and spending from KERA will lead to better schooling and performance by students, these changes may take a considerable amount of time. Increased salaries will have only limited impacts on teacher quality if they are received by teachers employed prior to KERA. Only when districts have the opportunity to hire new teachers can we expect to see any benefits from increased spending on teachers. In addition, for some districts the increase in resources has been substantial and now these districts must learn to manage these increased funds. Given that these districts have now had eight years since the initial infusion of KERA funding and the apparent limited spending on teachers, it might now behoove the state to evaluate how this money has been spent.

Endnotes

1. This discussion is adapted from Adams (1993), p. 331.
2. Discussion of fraud in KIRIS testing has even made the front page of the Wall Street Journal (see "Apple Polishing: Kentucky Teachers Get Bonuses, But Some Are Caught Cheating," Wall Street Journal, 9/27/September 27, 1997, p. 1 "Apple Polishing: Kentucky Teachers Get Bonuses, but Some Are Caught Cheating").
3. A more standard measure is the dropout rate. However, only recently have such statistics been compiled for all the states at the National Center for Educational Statistics. For this reason, we use the customary alternative measure of graduation rates that can be compiled from surveys of individuals rather than from information from educational agencies.
4. From information available at the Kentucky Department of Education website, http://www.kde.state.ky.us.
5. See Hanushek (1986), Journal of Economic Literature 24 (September), pp. 1141-77.
6. These results are from The NAEP 1996 State Assessment in Mathematics: Kentucky and The NAEP 1994 State Assessment in Reading: Kentucky, available at http://nces.ed.gov/naep/index.html.

This document is a of CBER, Center for Business and Economic
Research, located at the University of Kentucky, Gatton College of Business and Economics.