Chapter VI

State Funding and Tax Effort

Much of the discussion related to fairness of school financing systems focuses on the pattern and amount of education spending. Another aspect that has generated interest is the fairness in the distribution of the tax burden to raise revenue for schools.

A thorough analysis of the tax burden distribution for education would have to include each revenue source (i.e., property, income, and sales) and the level of government (i.e., local, state, and federal) responsible for raising the revenue. This would require a much more extensive inquiry than is covered by the scope of this study.

This chapter examines how tax effort was defined by the court and how the education financing situation has changed. Additional analysis broadens the discussion to alternative ways of examining the tax burden question.

Most of the money funding education in Connecticut derives from local government through its primary revenue source the property tax. The rationale behind this lies in part on the strong historical preference for local control of the educational system. Large variations in each town's capacity to raise revenues (based on property wealth) relative to the educational challenges within the respective school system, however, can require some towns to impose heavier tax burdens on their residents than others to provide a given quality of educational experiences. What education finance litigation has made clear, though, is that it is still the state's constitutional duty to fund the schools.

Although inequities may largely be offset by differences in state aid, some distortions may remain. Many believe the property tax imposes a burden on some taxpayers because it is not directly related to ability to pay. Regressivity can be a concern if higher property tax rates are levied in towns with above average proportions of poor households and lower rates levied in towns with wealthier households.

Ultimately, the question of tax fairness requires an examination of what values underpin the concept of fairness. For example, to some a flat tax appears to be a fair way to raise revenue, while a tax that increases in proportion to income seems fair to others. The answers incorporate specific values expressing preferences of what should be and what should not be and are not susceptible to empirical investigation. These types of wide-ranging policy changes are appropriately decided by the legislature.

State Funding and Per Pupil Spending

In the Horton decision, the Connecticut Supreme Court accepted a number of findings made by the trial court that taken together demonstrated the extent of the disparities that existed in financing elementary and secondary education. These findings along with a comparison between the situation in 1973 and today are provided below.

State's contribution. The court first examined overall funding for elementary and secondary education and found the state's contribution inadequate. A review of the current funding practices produced the following findings.

• The state increased its share of funding for education. In 1973, the court noted 70 percent of education expenses were funded from local sources and about 25 percent came from the state. As indicated in the overview section, in 2001 the state funded about 41 percent of education expenses, while local governments funded about 53 percent.
• The overwhelming majority of state funding for education takes into account each town's ability to pay. Eighty percent of state aid in 1973 was distributed through a flat grant and 20 percent came from 13 grants. None of these aid programs was distributed based upon the town's ability to finance education. In 2000, of the $2.6 billion expended by the state, 54 percent ($1.4 billion) was distributed through the Education Cost Sharing grant, which explicitly takes into account each town's ability to pay. Most of the current categorical grants take into account ability to pay, including the largest grants -- transportation and school construction -- which account for over 62 percent of categorical funding and 17 percent of total state funding.
• The goal of equal state-local funding has not been met at this time. The State Board of Education adopted a goal to achieve equal state-local share of the total cost of public schools by 2005 or shortly thereafter. The board supports increasing state's share of education costs and reducing reliance on local property tax revenues as a way to reduce inequities among districts.

Taxable property and operating expenditures. Because the local property tax was (and continues) to be the principal source of funding used by local governments to pay for their share of public school education, the court asserted the ability of various towns to finance local education is based on the dollar amount of taxable property per pupil in each town. This can be determined by dividing the comparable value of the property in a town by the number of pupils (also called the yield per pupil).

Using this measure, the court reasoned the wide disparities in the amount spent on education by towns result from the wide disparities in taxable wealth. Higher tax rates in property-poor towns cannot generate the revenue that lower rates in property-rich towns do, leaving many taxpayers in property-poor towns to pay higher tax rates and get less for their children. So, regardless of educational needs of any particular town's schoolchildren in 1973, more educational dollars were allotted to children who lived in property-rich towns than to children who lived in property-poor towns.

Table VI-1 shows the towns divided into deciles based on per pupil property wealth and compares average yield per pupil and average per pupil operating expenditures for local share for 1973 and 2000. Operating expenditures include expenditures from all revenue sources, excluding reimbursable regular education transportation, tuition revenue, capital expenditures and debt service. In general, the following issues can be noted.

• The difference in yield per pupil increased -- meaning the value of the property tax base in property-rich towns increased at a greater rate than that in property poor towns. In 1973, the court noted "wide disparities" in the effective yield per pupil ranging from $20,000 per pupil to $170,000 per pupil for a high to low ratio of 8.5. By 2000, the range was $209,420 to $2,492,153 per pupil for a high to low ratio of 11.9.
• The difference in average per pupil operating expenditures between the top decile towns and the bottom decile towns has been reduced. In 1973, the average per pupil operating expenses of the tenth decile towns were 35 percent less than the average of the first decile towns. In 2000, the average per pupil operating expenses of the tenth decile towns were 14 percent less than that of the first decile towns.
• The statistical relationship between operating expenditures and town wealth has been reduced. Many experts argue that no relationship should exist between educational spending and local property wealth per pupil - often referred to as fiscal neutrality. Statistical correlations were used to measure this relationship. Correlations are ways to measure the degree to which two variables are linearly related. While a strong correlation does not prove a causal relationship, it often suggests one exists. The correlations over the years show weakening relationships from 1996 through 1999.¹

Average total spending and state aid. Table VI-2 presents average total spending per pupil by wealth decile, revenue source, and total average spending with and without construction expenses for FY 2000. In this case, total spending includes operating expenses as well as special education and transportation expenditures. The table demonstrates the effect of state aid and the role it plays in bringing the per pupil spending of property-poor and property-rich towns closer together. The table shows:

• Average total per pupil spending is greatest at the first (wealthiest) decile. Total per pupil spending at the first decile is ($10,643), followed by the second decile ($9,336) and then the tenth decile ($8,889). The lowest are the sixth ($8,357), seventh ($8,213), and eighth deciles ($8,128).
• The greatest difference in per pupil spending is between the first and eighth decile. The difference between the top (first decile) and lowest (eighth decile) per pupil expenditure is $2,515 (31 percent lower), and the difference between the first decile and the tenth decile is $1,754 per pupil (20 percent lower).
• Property-poor towns (tenth decile) receive the most state aid. As property wealth increases, the state share of education decreases - the first decile towns receive an average of $239 per pupil (not including construction grants), while the tenth decile towns receive $5,519 per pupil (or 62 percent of the average total spent). Conversely, local share increases as property wealth increases - the first decile towns fund an average of $10,202 per pupil from local sources, while the tenth decile towns fund an average of $2,796 per pupil.

Education Mill Rates and Local Share

Below is an analysis of tax effort measured in several different ways. First, an examination of the tax rate for the local share of education expenses based on an equalized net grand list is presented. Secondly, a comparison is provided of the tax rates for local share based on the equalized net grand list adjusted for income. Finally, an assessment of local share as a portion of household income and as a percentage of total town income is offered.

Mill rate based on ENGL. The court stated tax effort of a town to finance education may be measured by determining the net school mill rate which is that part of the mill rate a town spends on education. An equalized property wealth measure was developed to compare the effort among the towns because unadjusted town mill rates are based on different assessment ratios and schedules for valuation. The equalized net grand list (ENGL) calculated by the Office of Policy and Management is the value of real and personal property at 100 percent fair market value. A mill is equal to $1.00 of tax for each $1,000 of assessment. Because the mill rate for education presented below is calculated based on the ENGL, it is an indicator of the difference between towns not the actual amount that is paid.

The local share of all current education expenditures was obtained for each town and divided by each town's ENGL to determine each town's education mill rate. Figure VI-1 contains education mill rates by wealth decile for the local share of education expenses in 1973 and 2000. The figure shows a number of distinctions among the towns.

• The difference between the education mill rate for the property-rich towns and the property-poor towns for the total local share of education expenses, while still significant, has decreased. In 1973, the difference between the first decile's average mill rate and the tenth decile's was 137 percent, while in 2000 this difference was reduced to 49 percent. The greatest difference is now between the first decile and the fifth and sixth decile (63 and 65 percent);
• Another way of considering the difference is that in 1973, the average net school mill rate of the tenth decile towns was almost 2½ times the first decile towns. In 2000, the average education mill rate of the tenth decile towns declined to 1½ times the first decile.
• The equalized education mill rate among the towns does not meet the principle of a uniform mill rate. Ideally, according to education finance literature, similarly situated homeowners should be paying a similar rate for education in each town. To use a simplified example of inequity, a taxpayer in the first decile towns with property valued at $200,000, would pay about $1,500 to support the local share of education, while the tenth decile taxpayer would pay about $2,300 on an equalized basis.

MER and major expenses. The equalized tax rates were also calculated for the top and bottom 10 percent of towns for the 1) minimum education requirement expenses (MER), which consist of all regular public elementary and secondary educational expenditures except those related to special education, state and federal grants (except ECS and federal impact aid), transportation, most construction and debt service expenditures, and adult education²; and 2) MER plus the local share special education and transportation, which represents the major mandated education programs.

Figure VI-2 shows the results for these two categories of expenditures. The average rate for the bottom 10 percent of towns (property-poor) based on ENGL is greater for both categories of educational expenditure than the top 10 percent. When considered together with the previous finding regarding the tax rate for the total local share, property-poor towns have greater equalized rates for education expenses than property-rich towns.

Income adjusted mill rate. A measure was also calculated to compare the relative tax effort for education among the towns based on income adjusted equalized tax rates derived from the local share of education expenditures. The equalized net grand list (ENGL) produced by the Office of Policy and Management, adjusted for both per capita and median household income, was used to create a comparative measure.

The income adjustment was made because a town's taxing capacity is affected by the income of its residents and is consistent with the wealth factor used in the ECS formula.³ This adjustment effectively reduces of the value of the taxable property in each town by comparing each town's per capita and median household income to the town with the highest per capita and household income. The adjusted equalized tax rates were calculated for the MER, major expenses, and all local expenditures.

Figure VI-3 compares the average statewide equalized education tax rate with the top and bottom 10 percent of towns (in terms of wealth) for the three expenditure categories described above. Overall:

• the wealthier towns' (top 10 percent) tax burden is less for each of the expenditure categories;
• for MER related expenses, the mean statewide is 15.38; top towns 5.88, bottom towns 23.94;
• for major mandated expenses, the mean statewide is 21.43, top towns 8.45, bottom towns 34.89; and
• for all current expenditures, the mean statewide is 28.47, top towns 13.07, bottom towns 40.28.

Local Share and Income

Although the court measured tax effort by comparing equalized mill rates among the towns, tax effort can also be related to income. Two alternative measures of tax effort were developed. The first measure represents the residential portion of the local share of education expenses as an average percent of median household income by wealth decile. Because median household income is related to residents, this measure is calculated by separating the residential and non-residential portions of the tax base and assigning a portion of the local share of education expenses to each part of the base. The average percentage of median household income is then derived for each wealth decile.

The second measure examines local share as a percent of total town income by wealth decile. Figure VI-4 presents the results.

• Generally, the wealthier the town the greater the proportion of median household income dedicated to local share of education expenses. The property-poor towns have the lowest percentage of residential income dedicated to local share. Although the percentage dips a bit for the third decile towns, the proportion gradually increases from 2.11 percent from the property poor towns to 4.67 percent for the property rich towns.
• Except for the two wealthiest deciles, the wealthier the town the greater the proportion of total town income dedicated to local share of education. Local education expenses on average represent about 2.89 percent of total town income for the first decile property-poor towns. This percentage peaks at about 4.39 percent for the third decile towns and declines to about 3.58 percent for the wealthiest first decile towns.

In summary, the program review committee reports the following findings.

State Funding and Per Pupil Spending

• Since the Horton decision (1973), the state increased its share of funding for education and the overwhelming majority of state funding for education takes into account each town's ability to pay, but the goal of equal state-local funding has not been met.
• The difference in average per pupil operating expenditures between the wealthiest towns and the poorest towns has been reduced.
• Property-poor towns (tenth decile) receive the most state aid - an average of $5,519 per pupil or 62 percent of the average total spent in FY 2000.
• The statistical relationship between property wealth and the amount of per pupil operating expenses has been reduced since the mid-1990s.
• Still, the average total per pupil spending is about 20 percent greater in the wealthiest towns compared to the poorest.

Education Mill Rates and Local Share for Education

• The difference in the mill rate for the local share of education expenses between the property-rich towns and the property-poor towns, while still significant (49 percent), has decreased.
• The equalized education mill rate among the towns does not meet the principle of a uniform mill rate - that is similarly situated taxpayers paying the same mill rate for education.
• When adjusted for income, mill rates indicate wealthier towns' (top 10 percent) tax burden is less for the local share of education expenditures.

Local Share For Education and Income

• Generally, the wealthier the town the greater the proportion of median household income dedicated to local share of education expenses.
• Except for the two wealthiest deciles, the wealthier the town the greater the proportion of total town income dedicated to local share of education.

The committee did not recommend any changes to the formula with regard to tax effort. The current elementary and secondary education finance system appears to be consistent with the directives of the Connecticut Supreme Court. The ECS aid formula and categorical grants are by and large distributed according to the ability of towns to pay for education. The state's contribution to education and the degree of targeting of educational aid has increased considerably. Clearly, the gaps in spending and in tax effort between rich and poor communities, while not eliminated, have been reduced through the efforts of the state distribution scheme.

Two areas of concern are worth highlighting -- lack of uniformity in equalized mill rates for local share of education and the 20 percent differential between property-rich and property-poor towns. While significant progress has been made in both these areas, further improvement requires the state do one or more of the following: provide additional funding for education, reduce funding to wealthy communities, cap the spending of wealthy towns, or require additional payments from wealthy communities to aid property-poor towns.

Given the place local control over education has within our political system and the general disdain for additional property taxes, it is unlikely caps or additional payments would be feasible or desirable. However, additional funding from the state derived largely from the more progressive income and sales tax may be the more palatable alternative.

¹ The coefficient of determination (r²) is a number between zero and one, which describes the portion of variation in one variable which can be attributed to variation in the other. In 1996, the r² value for the correlation between wealth and operating expenditures was 0.41 and in 1999 it was 0.32.

² ECS grant has been subtracted from the MER amount for each town to isolate local expenses. Federal impact aid is funding given directly to towns to offset costs for students who reside on federal tax exempt land.

³ Formula for this adjustment is same as ECS i.e., Property Wealth = Equalized Net Grand List x Income Adjustment Factor (IAF); IAF = (((PCI/HPCI) + (MHI/HMHI))/2) where HPCI = Highest Town per Capita Income, HMHI = Highest Town Median Household Income, PCI = Per Capita Income, HPCI = Highest Town Per Capita Income

Return to Table of Contents