Public schools in American receive funding in three basic ways: local, state, and federal (Guthrie & Reed, 1991). Local and state funds traditionally have been the most common form of funding for schools, with federal government funding being the most difficult to receive (Guthrie & Reed, 1991). State and federal funds offset local revenue disparities, however, state funds typically are not progressive enough and federal funds are not large enough to compensate local inequalities completely (Wilson, Lambright, & Smeeding, 2004). Because education is constitutionally a state responsibility, funding of schools is best addressed at the state level. In Texas, the state of interest in this study, funding can best be understood by looking at what the Texas Education Code stipulates. Stipulated is that public education is a state responsibility and that the state must provide an educational system financed through state revenue sources. The Texas Education Code directs the Texas Commissioner of Education to determine a cost ratio of administrative to instructional expenses and require districts to be in compliance annually (Kemerer & Walsh, 2000). Each student enrolled in the Texas public school system shall have access to programs and services that should be appropriate to the student's educational needs and should be equal to what any other student receives.

At the federal level, funds are provided by the federal government in many different ways. One important method of funding is through Title I which is intended to assist children in high-poverty schools overcome the disadvantages they experience from being raised in poverty. At its inception 40 years ago, Title I was justified as a relentless assault on the school quality side of the educational achievement challenge (Roza, Miller, & Hill, 2005). The Elementary and Secondary Education Act of 1965 would later be the foundation in which No Child Left Behind Act (NCLB) would build upon with its amendment of the act itself (Walsh et al., 2005). Title I funds are a primary tool used for schools and districts throughout the United States. These funds equal about 2% of total spending nationwide (Wilson et al., 2004). Schools depend on the added funding to assist them in working with students who are economically disadvantaged and migrant students, to name a few. Despite the fact that funding for Title I continue to grow, and the program is now the major funding arm of NCLB Act, funds are not always spent in a way likely to accomplish the purposes of the legislation (Roza, Miller, & Hill, 2005).

That is, schools are permitted a great deal of latitude in how they spend these federal dollars. For example, the Hawkins-Stafford Amendments allow Title I money to be used throughout the campus if 75% of students are at or below the poverty level (Wong, 1999). Monies, as long as certain populations are served, may be difficult to trace at the individual campus level. Title I monies are also permitted to be spent on teacher preparation. Some researchers have suggested that much more can be learned about how money affects public schools by looking closely at the practices of schools and school districts (Monk & Rice, 1999).

Moving back to the state level, the current system of funding Texas schools is called the Minimum Foundation Program (Kemerer & Walsh, 2000). This program has formulas that help determine how much money a district should receive. The formulas ensure that children from all districts, rich or poor, are given an adequate amount of money to help in their education. Set up by this program is a foundation level, which sets the amount a district should get, and sets up a contribution level, which is what a district receives in the form of taxes from property value within the district. The amount left, or the difference between the two levels, is what the state provides and is considered state aid.

Though school revenue comes from the contributions of local, state, and federal governments, local governments contribute the largest percentage of the funding for districts (Guthrie & Reed, 1991). For example, for the 2006-2007 academic school year, one school district selected at large, Cypress-Fairbanks Independent School District received 61.1% from their local government, compared with 36.8% from the state of Texas and 2.1% from the federal government as per their district website (Cypress-Fairbanks, 2007). In obtaining their local monies, school districts adopt two tax rates each year. Districts have a Maintenance and Operations tax rate which charges no more than $1.50 per $100.00 of property value (Carroll ISD, 2006). The other tax rate is the Interest and Sinking Fund, which is used when districts are in debt. No more than $.50 can be taxed for $100.00 property value when using this tax method.

In general terms, the federal government usually contributes about 8% of school funding within a Texas school district and provides the least amount of aid as compared to local and state funding (Jehlen, 2003). For the 2004 fiscal year, the local, state, and federal government provided 43.9%, 47.1 %, and 9.1% of funding respectively (Sable & Hill, 2006). The contributions are split in half. Half of the money (50%) is provided directly to the school district whereas the other 50% is provided to the state or regional education service centers. The money the federal government contributes is usually set aside for either a specific group or for specific programs. With this fact in mind, it appears that the aid that the state contributes is what helps equalize a district so that it can continue to provide services for the children.

The relationship between resources and student performance is still not very clear (Pan, Rudo, Schneider, & Smith-Hansen, 2003). Tajalli and Opheim (2004) showed improved performance at the elementary level when observing expenditures in instructional leadership. They revealed that money used to pay higher teacher salaries in combination with more experience improve the performance of students in poorer schools. In a study by Barth, Haycock, Jackson, Mora, Ruiz, Robinson, and Wilkins (1999), most high poverty schools used a larger amount of district and state funding to support increased professional development thus becoming high performing schools. These schools were able to produce gains in learning that schools in similar positions were unable to accomplish. In a study conducted by Pan et al. (2003), a strong relationship was observed between district resources and student success. High-performing districts showed different resource allocation patterns in specific fiscal and staffing categories than did the low-performing counterparts. Similar results were obvious when districts with strong improvement gains were studied. Improving school districts were comparable to high-performing school districts in resource allocation practices by spending more per pupil instruction and other instructional related areas. Practices such as reallocating administrative and other non-instructional funds to instructional areas were observed throughout the improvement school districts.

Wilson et al. (2004) adjusted expenditures for difference in costs of education and student needs. The authors used a data set for individuals and families instead of using data sets that compare school districts. Descriptive statistics indicated that school expenditures varied based on income, race, and urbanization. Income was a statistically significant predictor of school expenditures across all of their tested specifications except the cost adjustment. Urban area children had lower expenditures per student than suburban or rural area children.

Wenglinsky (1997) stated that fourth and eighth graders math achievement was positively associated with expenditures on instruction and school district administration. Not significantly related were expenditures on facilities, recruitment of highly educated teachers, or school level administration. A number of achievement variables are significantly related to local per pupil expenditures and provide support for the idea that student achievement is related to a school district’s financial standing (Napier, 1997). Napier reported that 20 of the 23 achievement variables investigated were significantly correlated with local per pupil expenditures.

Bray (2003) disclosed findings by the Southwest Educational Development Laboratory in which independent school districts in Arkansas, Louisiana, New Mexico, and Texas were studied. Five years of data were used to analyze operating expenditures by function, teacher and administrative staffing, and student test scores. In all four states, patterns existed between the spending of funds on instruction as a share of total expenditures. High performing school districts spent more monies on instruction per pupil and employed more teachers than did their low performing counterparts. Increased math scores in New York were attributed to a greater proportion of total spending dedicated to paraprofessionals (Iatarola, Stiefel, & Schwartz, 2002).

Jacques and Brorsen (2002) stated an additional $1000 per student in instructional expenditures would almost certainly add nearly a point increase in test scores. Such expenditures would focus on teachers, textbooks, and supplies as they were seen to be more productive. In 2004-2005, the average operating expenditure per student in the United States was $8, 208 whereas in Texas it was $7, 335 (Friends of Texas Public Schools, 2008a). Positive yet insignificant factors were workshops, seminars, and computers which is classified as instructional support. Instructional support could potentially aid in teacher productivity, but has been found to be statistically insignificant factors. Schools which spent more money on student support and school administration faired lower on test scores whereas those schools that spent more on instruction had higher scores.

When studying failing schools in Michigan, Anderson and Cotton (2001) found that students in a district that failed to have at least 75% of their high school students meet the basic standards of Michigan were much more likely to be in a high spending district rather than a low spending district. This information highlights that funding is not a determining cause greater student performance. As noted before, funding would seem to have greater significance if the relationship between instruction and non-instructional uses were studied.

The educational movement in Texas for the past two decades has emphasized school accountability. In the early 1980s, Governor Mark White appointed Ross Perot as the chair for the Select Committee on Public Education, thus beginning the political movement of accountability in Texas. Schools, elementary and secondary alike, all over Texas have felt the impact that such accountability brings with standards-based testing. Taking high-stakes tests, such as the TAKS (Texas Assessment of Academic Skills) or its predecessor, the Texas Assessment of Knowledge and Skills or the TAKS test (TEA, 2006a), has been essential in determining student learning. With the TAKS test came the following school and school district ratings: Academically Unacceptable, Academically Acceptable, Recognized and Exemplary (TEA, 2006b). School districts in Texas attempt to obtain the highest rating possible, not only to be granted more federal money, but also for the status a school district receives as a member of a successful campus. Schools are held accountable not only for the test scores of their students, in aggregate and by subgroup, but also for factors such as student attendance (TEA, 2005). To be considered Exemplary, a school must have no more than 0.2% dropout rate and 90% of the students and subgroups passing TAKS (TEA, 2005).

In each of the research questions, the relationships between school finance with school district accountability rating were investigated. Because student demographic characteristics influence how school monies are spent, all statistical analyses were controlled for student demographics. Thus, for all research questions, analysis of covariance (ANCOVA) procedures were performed. The independent variable in each of the questions was the school accountability rating, which consists of four groups and the dependent variable was a specific school finance variable. The ANCOVA method of statistical analysis was selected because it is capable of removing the obscuring effects of pre-existing individual differences among participants. An ANCOVA benefits when a comparison of mean achievement gain scores of different groups because it permits one to attribute gains to the effect of a technique rather than to initial scores (Gall, Gall, & Borg, 2003). Given that the percentage of economically disadvantaged students differs by school district and that the achievement of economically disadvantaged students tend to be lower than the achievement of students who are not economically disadvantaged, controlling for this variable should assist in removing any differences in how monies are spent as a function of economically disadvantaged student percentage. Before statistical procedures were conducted, data were examined to make sure that all data were congruent with assumptions under ANCOVA.

For the 2003–2004 academic year, a statistically significant difference was present, F (3, 1119) = 53.17, MSE = 15.15, p < .001, in percent expenditure by central administration among the school district accountability ratings. The effect size was small, 13% (Cohen, 1988). The Academically Unacceptable school districts had the largest adjusted mean (M = 15.18) and the Academically Acceptable districts had the lowest adjusted mean with (M = 5.56).

For the 2004–2005 academic year, a statistically significant difference was present, F (3, 1124) = 36.14, MSE = 15.32, p < .01. This result also supported that the percent expenditure by central administration varied as a function of school district accountability rating. The accountability factor accounted for 9% of the variance of the dependent variable, a small effect size (Cohen, 1988). The Exemplary school districts had the largest adjusted mean (M = 12.29) and the Academically Acceptable districts had the lowest adjusted mean with (M = 6.32).

For the 2005–2006 academic year, an ANCOVA revealed that a statistically significant difference was present, F (3, 1124) = 31.95, MSE = 16.13, p < .01. A difference was present in percent expenditure by central administration as a function of the level of the accountability rating received by the school district. The strength of relationship between the accountability rating and percent expenditure by central administration was small, .08, as assessed by partial eta square (n²). The Exemplary school districts had the largest adjusted mean (M = 11.04) and the Academically Acceptable districts had the lowest adjusted mean with (M = 6.0).

Results were congruent across the three years of the AEIS database. That is, statistically significant differences of the same magnitude were present among the four levels of district accountability ratings during the three academic years. It is important to note that in all three academic years that were reviewed, Academically Acceptable school districts had the lowest adjusted mean when compared to the other school districts. Recognized and Academically Acceptable districts were the lowest all three years while Exemplary school districts had the highest mean during the last two years.

For the 2003–2004 academic year, a statistically significant difference was present, F (3, 1119) = 14.87, MSE = .76, p < .001, in the percent expenditure by instructional leadership as a function of the level of the accountability rating received by the school district. The strength of relationship between the accountability rating and percent expenditure by function-instructional leadership was small, n², at .04 (Cohen, 1988). Academically Acceptable school districts had the largest adjusted mean (M = 0.85) and the Exemplary districts had the lowest adjusted mean with (M = 0.43).

The same statistical procedure, an ANCOVA, was repeated for the 2004 – 2005 school year. For the 2004–2005 academic year, a statistically significant difference was present, F (3, 1124) = 5.5, MSE = 1.16, p < .01. This result also reflected that the percent expenditure by instructional leadership varied as a function of school district accountability rating, albeit a very small effect size, .01 (Cohen, 1988). Academically Acceptable school districts had the largest adjusted mean (M = 1.00) and the Exemplary districts had the lowest adjusted mean with (M = 0.42).

Concerning the 2005–2006 academic year, the ANCOVA revealed that a statistically significant difference was present, F (3, 1124) = 15.53, MSE = 1.16, p < .001, n²= .04, in the percent expenditure by instructional leadership as a function of the level of the accountability rating received by the school district. The strength of relationship between the accountability rating and percent expenditure by function-instructional leadership was small (Cohen, 1988). Academically Acceptable school districts had the largest adjusted mean (M = 1.11) and the Exemplary districts had the lowest adjusted mean with (M = 0.17).

Similar to the first research question, the results were commensurate across the three years of the AEIS database. Statistically significant differences were present among the four levels of district accountability ratings during the three academic years. Exemplary school districts had the lowest mean every year whereas Academically Acceptable districts had the highest mean. Thus, Exemplary school districts were more likely to spend a lower amount of funds on instructional leadership than were Academically Acceptable districts.

Concerning the 2003–2004 academic year, a statistically significant difference was present, F (3, 1119) = 13.28, MSE = 2.47, p < .001, n² = .03, in the percent expenditure by school leadership as a function of the level of the accountability rating received by the school district. The strength of relationship between the accountability rating and percent expenditure by function-school leadership was small (Cohen, 1988). Academically Unacceptable school districts had the largest adjusted mean (M = 7.13) and the Recognized districts had the lowest adjusted mean with (M = 4.87).

The same statistical procedure, an ANCOVA, was repeated for the 2004 – 2005 school year. For the 2004–2005 academic year, a statistically significant difference was present, F (3, 1124) = 15.98, MSE = 3.47, p < .01, n² = .04. This result also reflected that the percent expenditure by school leadership varied as a function of school district accountability rating. Academically Unacceptable school districts had the largest adjusted mean (M = 7.61) and the Exemplary districts had the lowest adjusted mean with (M = 4.20).

Regarding the 2005–2006 academic year, an ANCOVA revealed that a statistically significant difference was present, F (3, 1124) = 13.12, MSE = 3.23, p < .01, n² = .03, in the percent expenditure by school leadership as a function of the level of the accountability rating received by the school district. The strength of relationship between the accountability rating and percent expenditure by function-school leadership was small (Cohen, 1988). Academically Unacceptable school districts had the largest adjusted mean (M = 7.19) and the Exemplary districts had the lowest adjusted mean with (M = 5.09).

Results were quite similar across the three years of the AEIS database. Statistically significant differences were present among the four levels of district accountability ratings during the three academic years. Interestingly, Academically Unacceptable school districts had the highest mean showing that more money was used in school leadership than in the other districts. During the last two years studied, Exemplary and Recognized school districts had the lowest mean with Exemplary being the lowest in both years. Thus, fewer dollars were spent on school leadership in Exemplary and Recognized districts when compared to the other rated districts.

Regarding the 2004–2005 academic year, a statistically significant difference was present, F (3, 1124) = 25.64, MSE = 2.71, p < .01, n² = .06. This result also reflected that the percent expenditure by co-curricular/extracurricular activities varied as a function of school district accountability rating. Recognized school districts had the largest adjusted mean (M = 3.57) and the Exemplary districts had the lowest adjusted mean (M = 1.32).

For the 2005–2006 academic year, an ANCOVA revealed that a statistically significant difference was present, F (3, 1124) = 12.32, MSE = 2.81, p < .001, n² = .03, in the percent expenditure by co-curricular/extracurricular activities among accountability ratings. The strength of relationship between the accountability rating and percent expenditure by function-co-curricular/extracurricular activities was small (Cohen, 1988). Recognized school districts had the largest adjusted mean (M = 3.63) and the Exemplary districts had the lowest adjusted mean with (M = 1.94).

Results, across the three years of the AEIS database, were commensurate. Statistically significant differences were present among the four levels of district accountability ratings during the three academic years. The lowest adjusted mean the last two years was found within Exemplary school districts. It is interesting to note that Recognized districts had the highest adjusted mean in all three years being studied.

Regarding the 2003–2004 academic year, a statistically significant difference was present, F (3, 1119) = 6.44, MSE = 34.51, p < .01, n² = .02, in the percent expenditure by instruction as a function of the level of the accountability rating received by the school district. The strength of relationship between the accountability rating and expenditure by instruction was small (Cohen, 1988). Academically Acceptable school districts had the largest adjusted mean (M = 51.27) and the Academically Unacceptable districts had the lowest adjusted mean (M = 45.46).

Concerning the 2004–2005 academic year, a statistically significant difference was present, F (3, 1124) = 19.75, MSE = 23.33, p < .01, n² = .05. This result also reflected that the percent expenditure by instruction varied as a function of school district accountability rating. Exemplary school districts had the largest adjusted mean (M = 57.06) and Academically Unacceptable districts had the lowest adjusted mean (M = 50.28).

For the 2005–2006 academic year, an ANCOVA revealed that a statistically significant difference was present, F (3, 1124) = 11.54, MSE = 24.18, p < .01, n² = .03. The strength of relationship between the accountability rating and percent expenditure by instruction was small. Academically Acceptable school districts had the largest adjusted mean (M = 56.39) and the Academically Unacceptable districts had the lowest adjusted mean (M = 52.23).

Results were commensurate across the three years of the AEIS database. Statistically significant difference was present among the four levels of district accountability ratings during the three academic years. In all three years being studied, Academically Unacceptable school districts had the lowest adjusted mean of all four rated groups. This finding signifies that as a group, less money was utilized for instruction expenditures in Academically Unacceptable districts than in the other three rated groups.

For the 2003–2004 academic year, a statistically significant difference was present, F(3, 1119) = 4.86, MSE = 1.60, p < .01, n² = .01, in the percent expenditure by instructional related services as a function of the level of the accountability rating received by the school district. The strength of this relationship was small (Cohen, 1988). Exemplary school districts had the largest adjusted mean (M = 3.07) and Academically Unacceptable districts had the lowest adjusted mean (M = 1.61).

Regarding the 2004–2005 academic year, no statistically significant difference was present, F(3, 1124) = 1.09, MSE = 2.67, p = .35. This finding, unlike the previous year, indicated that the percent expenditure by instructional related services did not vary as a function of school district accountability rating. Exemplary school districts had the largest adjusted mean (M = 3.34) and the Academically Unacceptable districts had the lowest adjusted mean (M = 2.42).

Concerning the 2005–2006 academic year, the ANCOVA revealed that a statistically significant difference was present, F(3, 1124) = 3.90, MSE = 2.65, p < .01, n² = .01. The strength of relationship between the accountability rating and percent expenditure by instructional related services was small (Cohen, 1988). Academically Acceptable school districts had the largest adjusted mean (M = 2.77) and the Academically Unacceptable districts had the lowest adjusted mean with (M = 2.03).

Results were not as consistent for this research question as for the previous five questions. One of the three years of data analyzed for this research question did not yield a statistically significant result. The 2003 – 2004 and 2005 – 2006 academic years showed consistent levels of statistical significance unlike the 2004 – 2005 academic year. In all three years being studied, Academically Unacceptable school districts had the lowest adjusted mean of all four rated groups. This finding is interpreted that, as a group, less money is utilized for instructional related services expenditures in Academically Unacceptable districts than in the other three rated groups.

In this investigation, six research questions were addressed and repeated for three years of data. The three years of data were collected from the Texas Education Agency Academic Excellence Indicator System website. For all of the research questions, with one exception for one year of analysis, differences were yielded in the ways in which school districts spent their monies for central administration, instructional leadership, school leadership, co-curricular/extracurricular activities, instruction, and instructional related services among the four accountability ratings. For each of the research questions, the magnitude, sign, and effect size of the statistically significant differences were similar across the years of data analyzed.

Percent expenditure by instructional related services was the only variable on which inconsistent results were obtained. The 2004 – 2005 school year financial data indicated that the school districts did not differ in how they spent their instructional related services monies, however, differences were present for the other two years of data. Of particular interest is our finding that Academically Unacceptable school districts spent significantly less of their monies on instruction than did the other three groups of school districts.

School finance variables influence school districts daily and are two critical components that focus on important aspects of education. It is critical to identify financial factors that help promote successful learning and produces student achievement in multiple areas. School districts with different accountability ratings appear to spend their monies in different ways. The intent of these researchers was to determine if, when student demographic characteristics were controlled for in statistical analyses, school districts with better accountability ratings spent their monies in different ways than school districts with poorer accountability ratings. As such, this information could be useful in determining if the ways in which school monies were spent could aid in the success of a school when working with high populations of at-risk, low-income, minority children.

This study has contributed a unique method of research in that data were analyzed of Texas school districts for three years (2003-2004, 2004-2005, and 2005-2006) when studying the relationship between accountability ratings and a variety of staff and finance factors. This study built on current knowledge and practice by school districts, but unlike other studies, a study of the three consecutive years mentioned was not available in regard to the factors being studied. Most of the data analyzed supported a much more deserving need to analyze staff and finance variables when planning for success.

The findings presented allow further insight into finance factors and the impact they have on district accountability in the State of Texas. The study was intended to provide information that could be useful to school and district administrators, school districts, policy makers within a school district and federal level and universities. The findings specifically targeted school district administrators and campus level administrators.

Support may be needed for school district administrators and campus administrator leaders working in districts with large concentrations of economically disadvantaged students among other populations. Another recommendation would be to spend less money on school leadership and make districts less top heavy or less inflated at the higher echelon of school district administration. Findings reported that Academically Unacceptable school districts spent more money on average on school leadership than did school districts in the other three categories. Redirecting funds to other areas such as instruction may reveal added benefits to student performance. Another important finding is that the Academically Unacceptable school district category on average spent less funding on instruction than did the other three accountability groups. These issues must be addressed and reviewed at the district and campus levels. Such information could possibly help Academically Unacceptable school districts with the issue of higher student performance and the methods used to achieve student success.

Lest readers overgeneralize from these findings, several caveats are in order. The research findings present in this study are pertinent to the State of Texas for the three years of data analyzed. The extent to which these findings generalize to other states is unknown. Second, no clear picture exists regarding the relationship between the manner in which school dollars are spent and student achievement. Though statistically significant differences were yielded in how these school districts spent their monies, no cause-and-effect relationship is present. That is, though causal-comparative research studies such as this one, no cause-and-effect relationship can be determined. Thus, though a relationship was demonstrated between school district accountability rating and the ways in which these school districts spent their monies, more research is needed to determine any causative factors.

Clearly, more research studies are needed in this area. One such study we suggest is an in-depth qualitative study of the 10 most successful school districts and the 10 least successful school districts, at each individual state, to determine the ways in which they spent their monies and the rationale behind the allocation of these dollars. A similar study of the top school districts with high percentages of economically disadvantaged students and of the poorest school districts with similar percentages of economically disadvantaged students could be conducted. What is needed is a deeper understanding of the ways in which dollars are allocated and spent. Until such time as studies such as this one are replicated, readers are urged to be cautious in their reliance upon our findings.

Anderson, P. L., & Cotton, C. S. (2001). Failing schools in Michigan: The surprising scale. Anderson Economic Group, Lansing, Michigan.

Barth, P., Haycock, K., Jackson, H., Mora, K., Ruiz, P., Robinson, S., & Wilkins, A. (1999). Dispelling the myth: High poverty schools exceeding expectations. Washington, DC: Education Trust.

Bray, J (2003). Investing in instruction for higher student achievement. Insights on Educational Policy, Practice, and Research. 15, 3-14.

Carroll Independent School District. (2006). Carroll Independent School District. Retrieved May 4, 2007, from School Finance Web site: http://www.southlakecarroll.edu/budget.htm

Cypress-Fairbanks ISD. (2007). Cypress-Fairbanks Independent School District. Retrieved May 4, 2007, from Finance Web site: http://www.cfisd.net/dept2/finance/state.htm

Friends of Texas Public Schools. (2008b). TAKS passing rates hit record highs. Retrieved February 10, 2008, from http://www.fotps.org/media/taksrecordhighs.pdf

Gall, M. D., Gall, J. P., & Borg, W. R. (2003). Educational research: An introduction (7th ed.). Boston: Allyn & Bacon.

Guthrie, J., & Reed, R. (1991). Educational administration and policy: Effective leadership for American education. Needham Heights, MA: Allyn & Bacon.

Iatarola, P., Stiefel, L., & Schwartz, A. E. (2002). School performance and resource use: The role of districts in New York City. Working Paper, Wagner School of Public Service, New York University, New York.

Jacques, C., & Brorsen, W. B. (2002). Relationship between school district expenditures in various categories and student performance. Department of Agricultural Economics, Oklahoma State University.

Jehlen, A. (2003, March). High-stakes questions. NEA Today, 21(6), 8-11.

Kemerer, F., & Walsh, J. (2000). The educator’s guide to Texas school law (5th ed.). Austin, TX: University of Texas Press.

Monk, D. H., & Rice, J.K. (1999). Modern education productivity research: Emerging implications for the financing of education. In W.J. Fowler, Jr. (Ed.), Selected papers in school finance, 1997-99, (pp.111-139). Washington, DC: National Center for Education Statistics.

Napier, L. (1997). Funding Disparities between Public School Districts in Mississippi. Memphis, TN: Jackson State University.

Pan, D., Rudo, Z., Schneider, C. L., & Smith-Hansen, L. (2003). Examination of resource allocation in education: Connecting spending to student performance. Austin, TX: Southwest Educational Development Laboratory.

Roza, M., Miller, L., & Hill, P. (2005). Strengthening Title I to help high-poverty schools. Seattle, WA: Center on Reinventing Public Education.

Sable, J., & Hill, J. (2006). Overview of public elementary and secondary students, staff, schools, school districts, revenues, and expenditures: School year 2004-05 and fiscal year 2004 (NCES 2007-309). U.S. Department of Education. Washington, DC: National Center for Education Statistics.

Tajalli, H., & Opheim, C. (2004). Strategies for closing the gap: Predicting student performance in economically disadvantaged schools. Educational Research Quarterly,28(4), 44-54.

Texas Education Agency. (2003b). Glossary for the Academic Excellence Indicator System 2003-2004. Retrieved July 7, 2009, from http://ritter.tea.state.tx.us/perfreport/aeis/2004/glossary.html

Texas Education Agency. (2005). Requirements for each rating category. Retrieved July 7, 2009, from http://ritter.tea.state.tx.us/perfreport/account/2005/manual/index.html

Texas Education Agency. (2006a). Students Assessment Division: About the Students Assessment Program. Retrieved July 7, 2009, from http://ritter.tea.state.tx.us/index.html

Texas Education Agency. (2006b). The basics: Determining a rating. Retrieved July 7, 2009, from http://ritter.tea.state.tx.us/perfreport/account/2006/manual/index.html

Texas Education Agency. (2006d). 2006 accountability system state summary. Retrieved July 7, 2009, from

http://ritter.tea.state.tx.us/perfreport/account/2006/index.html

Wenglinsky, H. (1997). When money matters. Princeton, NJ: Educational Testing Service.

Wilson, K., Lambright, K., & Smeeding, T.M. (2004). School finance, equivalent educational expenditure, and income distribution: Equal dollars or equal chances for success? Syracuse, NY: Working Papers 62, Center for Policy and Research, Maxwell School, Syracuse University.

Wong, K. (1999). Funding public schools: Politics and policies. University of Kansas Press.