Draft #4: December 23, 1997

Note 1. This draft does not include (1) the Cover Sheet, (2) a forwarding letter from Jim Rier and Wes Bonney to Education Committee and all members of the Legislature, (3) Table of Contents.

(to be written last)

L.D. 1895 (Chapter 61 of the Resolves of 1997) directs the State Board of Education to establish a committee "to study the school funding formula". Two specific duties are identified in Resolve 61.

"Review and make recommendations regarding the report presented by the Commissioner of Education to the Joint Standing Committee on Education and Cultural Affairs related to the current methods used to calculate the income and cost-of-living adjustment factors."

"Review the essential programs and services plan presented by the State Board of Education…for the purpose of developing an adequate and equitable method to fund essential programs and services."

A report on the first of these duties is to be presented to the Joint Standing Committee on Education and Cultural Affairs by January 1, 1988. A report on the second of these duties is to be presented by January 15, 1999.

The following report reflects the work of this Study Group regarding the first of the duties described above.

Recommendation 1. The use of income in the funding formula should continue, as well as the current weighting of 85% property valuation and 15% median household income, in the computation of each community’s ability to pay taxes used in the calculation of Operating Costs subsidies.

Study by prior task forces. This section will be completed later.

Considerations by this Study Committee.

Income is an important measure of ability of a taxpayer to pay taxes so the Study Committee recommends the continuation of the income component of the formula for determining the ability to pay taxes. There are only three ways a tax can be paid. First it can paid from income; second, it can be paid by converting an asset which is one component of wealth to cash; and, third, the funds can be borrowed.

State Valuation of property has been the traditional method for determining ability to pay. The starting point for State Valuation amount is the amount of each municipality’s local assessment of property. The local assessment of property is not required to represent 100% of market value. In comparison, State Valuation of property represents an estimate of 100% market value. The process of determining State Valuation begins by considering, for property that has been recently sold, the local assessment value of this property and its sale price. By comparing these two amounts for appropriate parcels, the Bureau of Taxation is able to calculate an adjustment which can be applied to the local assessed value of property to create the amount of State Valuation.

Property valuation is not the most rational basis on which to assess a tax based on the taxpayer’s ability to pay. It is only one form of wealth; it is one of the most difficult forms of wealth for which to establish a fair and equitable value. Some of the limitations that are inherent in the use of State Valuation are as follows: (1) Local assessment is performed, in some cases, by untrained assessors. (2) The process of determining local assessment amounts is not an absolute science and necessarily involves a certain degree of judgment calls. (3) Because State valuation is based on market value, it represents the income value which the owner will obtain if and when the property is sold. However, unless and until the owner sells this property, state valuation does not necessarily measure the owner’s ability-to-pay current tax bills. (4) For small towns, state valuations, the amount of sales data to calculate State Valuation from local assessed value may be limited or even non-existent. In this case, additional special methods must be employed to determine State Valuation. (5) In many cases there are varying degrees of mortgage debt outstanding. Using the gross value without considering the level of indebtedness aggravates its fair use as a tax on wealth. The Study Committee acknowledges the strengths of the Bureau of Taxation’s outstanding work to make the State Valuation data both valid and useful for school funding, as well as for other purposes. Nevertheless, in consideration of the factors described above, the Study Committee has a concluded that it makes sense to complement the use of State Valuation with some form of income measure, in order to best determine the ability to pay of taxpayers. However, the Study Committee also recognized the limitations regarding the use of income as a measure of ability to pay. Two principle reasons are as follows: (1) The only sources of income are residential income measures. Income for industrial and commercial property owners, homeowners who do not reside in the community, utilities, etc., are not available on a town-by-town basis. (2) Municipalities do not have the authority to impose taxes on income. Only property is taxable by municipalities. The following example illustrates the consequence of this fact. If one of two equal-valued parcels of land is owned by a millionaire and the other is owned by a retiree who lives on a tiny fixed income, the two tax bills will be for the same amount.

In conclusion, because both state valuation and income have strengths and weaknesses, the Study Committee has concluded that if used together, they serve useful and complementary roles in measuring the ability to pay.

David Silvernail of the Maine Educational Policy Research Institute did a correlation study on the Relationship Between Valuation and Income for the Study Committee (see Appendix no ___) which showed that while there is a strong correlation between the total property valuation of a community and the total household income, there was a low correlation between total property valuation and median household income. This study suggests that state valuation and median household income are measuring different constructs. Furthermore, this result is consistent with a conclusion that measuring an entire community’s ability to pay taxes based on property valuation can put undue stress on individual taxpayers while the community as a whole has the ability to pay. Introducing the adjustment of income is clearly appropriate.

It is significant to note that there are three components to the school funding formula (Operating Costs, Program Costs, and Debt Service), and that State Valuation of Property continues to be the only basis for determining ability to pay in all but the Operating Costs component. Furthermore, it is also important to understand that even for the Operating Costs component, the power of the income variable is limited by its 15% weight. State Valuation of property continues to be the dominant factor in determining ability to pay, even for this Operating Costs component.

The Study Committee is proposing this recommendation in recognition of the fact that the criticisms voiced regarding income and regarding State Valuation have, in general, not been critical of the relevance of income and property data in the school funding formula. Rather, most concerns have centered on whether high quality data is being used in the funding formula.

Recommendation 2. Cost-of-living should be eliminated from the school funding formula. The concept of cost-of-living is clearly relevant as an adjustment to income as used in the school funding formula. However, the Study Committee was concerned about unresolved technical issues regarding the validity of measuring this concept. Therefore, Study Committee recommends removing cost-of-living from the school funding formula, until and unless a resolution of these technical issues can be identified.

Study by prior task forces. This section will be completed later.

Considerations by this Study Committee. The concept of using a COLA adjustment for the cost of living in different parts of the state is clearly a reasonable

concept. It is clear that there are differences in the cost of living in various parts of the state. The problem comes in obtaining valid data from each individual community in the state. Sampling techniques are used and because some of the communities are small it is difficult to get valid and useable data. The methodology used is one developed by the National Chamber of Commerce. The data used in the current adjustment are collected under contract by Planning Decisions. They attempt to price a market basket of goods including housing costs and a market basket of other goods typically purchased by households. Collecting data with demonstrable validity is difficult, for four reasons collect.. One reason is the difficulty of establishing costs for the housing component if there is no recent sales data. A second reason isis determining the price of the food component in the market basket because many towns do not have markets and there is not always a clear pattern of where the residents of each town shop. A third reasonis the composition of the market basket. It is quite clear that what a family buys in an urban area can be quite different from what a rural family may buy. Finally, a fourth reason occurs because the consequence of owning property has conflicting consequences. Property, as a part of State Valuation, results in decreased school funding. In comparison, property, as a part of the cost-of-living adjustment, results in increased school funding. Collection of cost-of-living data by regions, rather than by individual municipalities, can mitigate the first three of these issues. However, unless a regional approach to cost-of-living is taken, and unless the fourth of these issues is resolved, and as long as the taxes for the local costs of education are raised on an individual community basis, the attempt to adjust incomes for a cost of living index should be abandoned.

Study by prior task forces. This section will be completed later.

Considerations by this Study Committee. Adopting this recommendation would reduce the rate of increase in ___ communities thereby eliminating the extreme cases. The current method used to update median household income from the 1990 U.S. Census numbers is done by a firm named Claritas. The method is proprietary and uses economic and demographic trends at the national, state and county levels to update the numbers for all communities in the state. The results of this process create some rates of increase for a few communities, particularly some smaller ones, that do not meet the test of reasonableness and consequently cast doubt on the appropriateness of all the data. (The term "outliers" in the above recommendation refers to these suspect increases. Several alternative methods of updating the median household income were investigated but none of those alternatives produced results that encouraged continuation of attempts to update the data. (These alternative methods are described in the discussion of Recommendation.) However it was felt that several methods could be tested with validation of the results against the data that will be gathered for the census data for the year 2000. This suggestion will be discussed in more detail in recommendation # 4 following.

Study by prior task forces. This section will be completed later.

Considerations by this Study Committee. Apart from the lack of any annual updating, the Study Committee considers the U.S. Census income data to be the most sound basis for any income calculations. It has two primary virtues. First, almost all sources of income, such as are wages and salaries, self-employment income, interest and dividends, social security, public assistance and welfare, etc., are includedExamples of income that is not included are ________. Second, Most types of residents are considered. Notable exclusions are individuals who are 14 years or younger, and residents of nursing homes, prisons, etc.

Regarding Part (a) of the recommendation. Currently, Median Household Income is the statistic used in the funding formula computations. A critical weakness of this statistic is that household size varies widely among communities. The following example illustrates the consequence of this fact. Suppose two communities, A and B, had the same Median Household Income: $20,000. However, the typical household in community A has 4 members, while the typical household in community B has 3 members. Although the median household income amounts are identical, this ignores the fact that household income in community A must serve a larger household than the household income in community B., The Study Committee considered several alternatives to resolving this issue. It finally considered that the simplest and most effective method would be to divide the Median Household Income of each community by the average number of persons in a household in that community.

Regarding Part (b) of the recommendation. When the new data is available three methods of updating the income data should be tested retrospectively from current data to the new U.S. Census data for 2000 to determine their relative validity as means of updating data. The three updating methods are: 1. Maine income tax data, 2. the Bureau of Economic Analysis income data and 3. the Claritas data. The proposed research would compare predictions of the U.S. Census’s 2000 income data, based on each updating method, with the actual U.S. Census data for 2000. Criteria for acceptable predictions must also be formulated. If any of these three updating methods are validated as being close to the new census data they should be used for updating. Otherwise the 2000 census data should be used until the 2010 data is available without any effort at updating. Each of the above methods of updating Decennial income data was investigated. Without having new census data against which to validate each method of updating it was not possible to feel comfortable that one method was superior to the others. If a reasonable method of updating the data that produces results that are believable and fair can be found, it makes sense to update the data. Testing the change by communities in the U.S. Census data from 1980 to 1990 demonstrated clearly that the rate of change in incomes varied significantly among the many communities in Maine. That fact argues for making an attempt to find an appropriate method of updating, if possible. On the other hand the eleven other states that use income in their school funding formulas do not annually update their income numbers. Six of the states use the latest U.S. Census Decennial data and five states use state income tax filing data. The advantage that the U.S.Census data has over Maine income tax filings is that it includes all kinds of income as well as incomes for residents of the community. The Maine tax data does not include tax exempt income and there is no information on the incomes of those who are not required to file state tax returns. Also, it is significant to note that large percentage of Maine residents are not required to file an income tax form, and a large increase in the number of exemptions is expected. It is recognized that use of the Maine Income Tax data provides some benefits. For example, there is no sampling involved, and the methodology is well documented,. A method that would seem to be appropriate is to use the U.S. Census data for the latest census and then to update that data based on the rate of change that is computed through the annual state income tax data. It is important that this alternative be tested to insure its appropriateness. To implement this suggestion the Department of Taxation should be asked to require all state income tax returns to indicate the community of residence.

Regarding Part (c) of the recommendation. Any data source for the annual updating may occasionally produce results for a few communities which are of doubtful accuracy. Even the State Valuation process occasionally produces suspect results; fortunately, there is an appeal process which can be employed to resolve such concerns. However, because these income updates are provided by a private sector firm whose policies is not subject to Maine laws, there are limits to the use of any appeal process, and other mechanisms are therefore necessary. The Study Committee believes that the most efficient resolution is to use a three-year rolling average. Each annual update would be calculated as the average percentage of increase of the three most recent years of income data. In addition, outliers would be identified and adjusted, as described in Recommendation 4.

Recommendation 5. A continued and expanded taxpayer equity analysis should be performed, to provide additional information to the Legislature as it considers these recommendations. This analysis would study the effect of the income adjustment on taxpayer equity. The analysis would continue and expand a pilot study that had been completed by the Department of Education and presented to the Study Committee. The conclusions of that pilot study were tentative in nature, because of the small sample used. However, the methodology is considered promising, because it offers an external statistical criteria which can be used to assess the merits of any proposed change in Maine’s school funding formula.

Considerations by this Study Committee. The Study Committee attempted to define taxpayer equity in order to devise some way to measure it. By doing so it was felt that the impact of the income and COLA adjustments could be measured in a statistical way and a decision could be made concerning their use in the formula based on fact rather than emotion or intuition. Four definitions of taxpayer equity were identified:

(to be completed later)

(to be completed later)

(to be completed later)

(to be completed later)

None of the measures of taxpayer equity are completely satisfactory by themselves but the committee felt that taken together they could give some helpful insights. The Department of Education conducted a study which identified six communities that were high receivers of state subsidy funds, six that were low receivers of funds and six that were average receivers of funds and tested four different definitions of taxpayer equity for the eighteen communities. A description of four taxpayer equity concepts that were examined is provided in Appendix B. The methodology used in the study is provided in Appendix C. The results of this study were intriguing in that three of the four measures of taxpayer equity showed less variance with the use of both the income and COLA adjustments than with property valuations only. This indicates that with the income and COLA adjustments there was a higher degree of taxpayer equity. There were some concerns over the study. These concerns were as follows. (1) The small sample size (18 school units) should be increased, to improve the validity of the results? (2) The statistical significance of the equity indicators should be stated. (3) were the results of the test of the definition that didn’t show a smaller degree of variance of enough significance to offset the impact of the other three definitions? (4) The pilot study focused on school units, rather than municipalities, and did not reflect the impact of a district’s cost-sharing formula. Unfortunately this study was presented at the December 15, 1997 meeting of the committee and with a January 1, 1998 deadline for the report to be submitted to the State Board of Education and to the Joint Committee on Education and Cultural Resources not enough time was left to do a more complete study. It was felt that a continuation of this study could help members of the Legislature and other policy makers make more informed policy decisions in the future.

Recommendation 6. As much as possible, income in the school funding formula should be used primarily to adjust the portion of State Valuation which is residential homes. State Valuation of property is a measure of the ability to pay of all property taxpayers. However, the only usable income data reflects only the ability to pay of a subset of property taxpayers: resident homeowners.

Study by prior task forces. This section will be completed later.

Considerations by this Study Committee. State Valuation includes, in addition to residential property, other categories, such as industrial, commercial, utilities, etc. However, the only income data which is available on a municipality by municipality basis is the income of residential homeowners. This fact is important because the percentage of State Valuation that is represented by residential property varies widely across the state. The ideal use of income should reflect both of these considerations. The Department and the State Planning Office are continuing to work on a method which would consider the percentage of State Valuation which is owned by residential homeowners and are confident that a resolution is feasible.

Recommendation 7. The Governor and the Maine Legislature should carefully assess current tax policy as it relates to the funding of education.. Because of current state tax policies, two trends are evident. First, there is an increasing property tax burden. Second, there is an increasing inequity of the burden among property taxpayers. The trend is especially unfortunate because of the considerable range in property values behind each pupil.

Study by prior task forces. This section will be completed later.

Considerations by this Study Committee. In an attempt to develop better pupil equity and taxpayer equity, basic tax policy particularly as it relates to the reliance on property taxes for such a relatively high proportion of education funding should be reassessed. There were several issues concerning educational funding that were introduced into the work of the committee that were not directly related to the issues the Legislature had asked the committee to consider. However these issues were so closely related to its work, the committee felt it would be remiss if it did not comment on the issues. Those issues are:

1. The level of state funding has been neither adequate nor predictable enough to meet the education needs of all Maine kids. (A discussion of the actual numbers and the delay in communities knowing what they could expect of receive as a subsidy should be placed here.)

2. The issue of the significant variance in the amount of funding per student across the state makes the goal of equitable educational opportunity for all Maine children a difficult one to attain. There is incredible variance in the state property valuation per student in the various communities across the state. For fiscal 1997/98 the property valuation per pupil ranged from a high $5,942,857 to a low of $3,415. with the state average at $304,007. There were twenty-eight communities with $1 million or more per pupil and seventy-four with less than $200,000 per pupil. While the distribution formula the state uses for the state subsidy tries to compensate for this wide variation it is virtually impossible to end up with an equitable amount of funding for all students. This is particularly true in recent years as a higher portion of general program assistance education funding has been provided by local communities and has been raised by the property tax. Two major contributors to this wide variance are the very different mix of residential and non-residential real estate in various communities. Some communities such as Wiscasset have significant non-residential real estate i.e. Maine Yankee Power Plant and some of the small rural communities have significant taxable forests and woodlands while other communities have very little non-residential taxable property. The other significant factor in such a variance is the number of pupils. There is not necessarily a correlation between number of pupils and property valuation. An effort should be made to either reduce the relative reliance on property taxes or to find some way to equalize the incredible variance in property tax per community.

LD1895

July 1^st Memorandum of Understanding from Peggy & Shirley to Wes

Membership (Do we include memberships in BOTH Task Forces?)

Selected handouts considered by the Task Force (which ones?)

Selected printouts provided to the Task Force (which ones?)

This appendix will quote relevant parts of the current funding formula that are relevant to the work of this Study Group.

20-A MRSA, §15603, sub-§ 11-A states, in part, that the state valuation calculation for Program Costs and Debt Service is "the average of the state valuation amounts for the two most recent years prior to the year of funding".

20-A MRSA, §15652, sub-§ 6 states, in part, that the state valuation calculation for Operating Costs is the "the lesser of the average of the state valuation amounts for the 2 most recent years prior to the year of funding or the state valuation amount for the most recent year."

20-A MRSA, §15657, sub-§ 2 states, in part, that "the Department shall use local median household income data, updated annually from the Federal Decennial Census. The Department shall contract for the acquisition of annual normalized regional cost adjustment data, based on the National Chamber of Commerce model."

20-A MRSA, §15657, sub-§ 1 stipulates that each school administrative unit’s per-pupil state valuation for Operating Costs, as described above, is "multiplied by a property weight of 0.85".

20-A MRSA, §15657, sub-§ 2 stipulates that each school administrative unit’s median household income, divided by the statewide average median household income and then divided by the unit’s COLA, "is multiplied by an income weight of 0.15".

This appendix will summarize the methodology used in the pilot study of 18 units,

The glossary will define key school funding terms which are used in the report, two forms of wealth (stock & flow), tax base, ability to pay, tax effort.