Maine’s Tax Burden – History and Projections

• Maine’s state and local tax burden expressed as a percent of personal income peaked in fiscal year 1998 at 13.8%, driven largely by the growth of taxable income from capital gains, but it is projected to decline to 12.3% by fiscal year 2009 (see Graph A – State and Local Taxes as a % of Personal Income).
• Maine’s state and local taxes have grown at an average annual nominal rate of 6.1% from fiscal years 1986 to 2005 or 3.0% in that same period adjusting for inflation.
• Maine’s state and local taxes are projected to increase by 3.4% average growth annually from fiscal years 2005 to 2009 or 0.4% on an inflation adjusted basis.
• Maine’s state and local taxes are projected to increase on a per capita basis by $471 from fiscal year 2005 to 2009 (see Graph B – State and Local Taxes Per Capita), or after adjusting for inflation, taxes per capita decline by $7 over the same period (see Graph C – State and Local Taxes Per Capita (Adjusted for Inflation)).
• Projections of Maine’s reliance on local property taxes in its tax mix remain high compared to other states at slightly less than 36% of total taxes and fees (see Graph D – State and Local Tax Mix).

BETR and Circuit Breaker Changes
Maine’s effort to get “credit” in the national rankings for its 2 major property tax reimbursement programs by changing their budgetary treatment contributes to the projected reduction of Maine’s tax burden. The treatment of these tax programs as reductions of individual income tax rather than expenditures is projected to lower Maine’s tax burden by approximately 0.2% of personal income and by nearly $100 per capita. However, this reduction does not appreciably affect Maine’s rank nationally.

LD 1 (PL 2005, c. 2) – Property Tax Reform and Spending Limits
In January 2005, the Maine Legislature adopted statutory limits on spending and property tax increases. It is still too early to judge the full effect of these reforms. During its 1st partial year of implementation, the annual growth of local property tax commitments slowed to 1.6% after having an annual average growth rate over the last 20 years of 6.6%. At this point, it is extremely difficult to gather data and analyze local tax and spending patterns. Part of the tracking effort created by LD 1 requires additional local data collection. Through this effort, future analyses and projections may be significantly improved.

Maine’s Tax Burden – Risks from Structural Gaps
This report projects Maine’s tax burden through fiscal year 2009 assuming no changes in current law. Next January, the Governor and the Legislature will be making decisions on how to balance the General Fund and Highway Fund structural gaps for the 2008-2009 biennium. The General Fund and Highway Fund structural gaps are currently estimated to be $425 to $450 million and $80 to $90 million, respectively. If these imbalances are addressed entirely through changes that increase revenue in a revenue source captured by Census definitions, the projected reduction of Maine’s tax burden would be eliminated.

I. Introduction

Maine’s tax burden has received much attention due to the national comparisons of state and local tax burden based on data and estimates compiled by the US Department of Commerce, Bureau of Census (Census). Based on 2002 data, the most recent available, Maine ranks 2nd highest in state and local tax burden expressed as a percentage of personal income at 12.9% and 9th highest on a per capita basis at $3,562 per capita (compilation by Federation of Tax Administrators). While it appears to the Office of Fiscal and Program Review (OFPR) that Census estimates overstate Maine’s local tax burden probably as a result of double counting of some local property tax revenue, adjusting for this overestimate (approximately $200 million in fiscal year 2002 – see Appendix B) does not significantly change Maine’s rankings. Based on data collected by Maine Revenue Services, OFPR’s calculation of tax burden for fiscal year 2002 lowers Maine’s tax burden to 12.3% of personal income (3rd highest) and $3,383 per capita (14th highest).

This particular method of measuring tax burden, dividing revenue from taxes and fees by a state’s total personal income, is widely used. This measure is now a benchmark within Maine’s recently enacted state and local government spending limitations law. Under the new law, when Maine’s tax burden rank as measured as a percentage of total personal income falls to the middle third of the states, the allowable growth of appropriations and local taxes increases.

There are many different measures of tax burden. For instance, the Maine Economic Growth Council, which formerly used the Census data as a benchmark, now uses a similar but slightly more complicated measure produced by the Tax Foundation as one of its benchmarks. In its most recent report, this benchmark of moving Maine’s tax burden toward the New England average received a “red flag” from the council, having missed this benchmark for the third year in a row. Even the most simple tax burden measures deserves a cautionary note in terms of its ability to adequately compare the tax structures and tax incidence across states with their unique mix of taxes and other revenue sources and their unique economic and demographic situations. The methodology section of this report discusses some of the many factors that should be considered when analyzing tax burden as well as referencing some of the other measures of tax burden provided by different organizations. Much of the discussion is based on presentations by Michael Allen, Director of Econometric Research, Maine Revenue Services.

With the cautions noted above and in the methodology section, this report then describes the methods and assumptions employed by OFPR in presenting this historical compilation and projections of Maine’s tax burden. This is the first time that the OFPR has provided this level of supporting detail and narrative to its update of Maine’s tax burden, which OFPR has been tracking and updating for a number of years using the Census methodology. OFPR would like to thank Maine Revenue Services, the State Planning Office, Kennebec County and the Maine Municipal Association for their input and data collection efforts that have helped produce and improve OFPR’s tax burden update. OFPR always appreciates comments and feedback that might help to improve this analysis for future updates.

A. Tax Burden Analysis Factors

Many factors must be considered when using or evaluating tax burden analysis. Some of the major considerations are:

1. Revenue. What sorts of revenue should be included? Taxes may seem obvious, but should taxes paid by nonresidents be included when income of nonresidents is not part of the calculation? Should revenues be reduced by tax refund payments paid from appropriations? How should fees be treated?

2. Interjurisdictional comparisons. Analysts who want to compare different jurisdictions must stick to readily available data that is consistently applied across jurisdictions. Census data is the most readily available but may be subject to some inconsistencies in treatment of different types of revenue. Census data on local revenues may be based largely on projections of revenue rather than regularly gathered data.

3. Applications of tax burden analysis. Many advocates use tax burden analysis for the purpose of encouraging particular policy choices. Academic analysis of the impact of tax burden on taxpayer behavior is widely varied. All claims of the role of tax burden in taxpayer behavior should take into consideration the complexity of all factors involved.

B. Methods of Tax Burden Analysis

Many organizations publish annual rankings of state and local tax burdens in an attempt to assist decision makers in comparing the impact of state and local policy choices. The most common rankings are based on one or more of the following methods of analysis. Each method of analysis has its strengths and weaknesses.

1. Aggregate Tax Estimates. These methods of analysis are the easiest to calculate and the most commonly cited. They involve few variables and data is readily available from Census. Comparisons are usually expressed in one of the following formats.

• Taxes per $1,000 of personal income. This method calculates a total amount of revenue for a jurisdiction and divides it by the total personal income for the jurisdiction, yielding a tax burden expressed in dollars of revenue per $1,000 of personal income.
• Taxes per capita. This method is similar to the previous method except that the denominator in the calculation is total population for the jurisdiction. This analysis yields a tax burden expressed in dollars of taxes per capita for the jurisdiction.

2. Aggregate Economic Incidence Estimates. This method is similar to the previous method but calculates taxes as a percent of per capita income.

3. Micro-Simulation Economic Incidence Estimates. This method calculates tax burden in a more complex way by analyzing the distribution of taxes among different income groups. It provides information about the progressivity of a tax system.

4. Representative Household or Firm Estimates. This method calculates tax burden by modeling different categories of households or firms and calculating the application of a jurisdiction’s tax structure to representative taxpayers.

C. Census Methodology vs. Maine’s Tax Incidence Analysis

As noted above, the most commonly used measure is the aggregate tax estimate using Census data because of its relative simplicity and comparability across states. This method is also used by OFPR because it is the easiest to update, avoids a lot of complicated assumptions and is easy to compare with other states. This calculation operates outside of the political question of whether or not a proposal is a “tax increase” or “tax decrease.” Any legislative change or revenue forecast revision that increases tax and fee revenue as defined by Census (see Appendix C for Census’ definition of “taxes”) increases estimates of tax burden and vice versa for tax and fee revenue decreases. OFPR tracks these legislative changes that affect revenue classified under Census definition of taxes and compiles these changes at the end of each session. The primary difference between Census and OFPR calculation is the collection methods for the state and local tax data.

On the other end of the spectrum, Maine Revenue Services biennially updates the most complicated of the tax burden measures, its tax incidence analysis, Maine Tax Incidence Study: A Distributional Analysis of Maine’s State and Local Taxes. This methodology provides the most thorough analysis of the economic incidence of Maine’s taxes as opposed to the statutory incidence of Maine’s taxes. It uses economic and tax micro-simulation models to analyze the exportation of taxes to non-residents, the amount of business taxes passed through to Maine residents and other “tax shifting” to determine the amount of taxes paid by resident individuals or the “effective tax rate.” It also provides analyses of the effective tax rate as a percent of income for different income groups, the progressivity of Maine’s tax structure. However, Maine Revenue Services’ definition of “taxes” included in its analysis is much more narrow than the Census definition employed by OPFR and excludes license fees, user fees and the health care provider taxes used in “tax and match” initiatives. For the 2002 tax year, Maine Revenue Services calculated an effective tax rate of 11.5% of income and $2,465 per capita.

D. Estimating State Taxes and Fees

When estimating future state revenue, OFPR uses the most recent General Fund and Highway Fund revenue forecasts of the Revenue Forecasting Committee (RFC) established in Title 5 MRSA, chapter 151-B. The most recent RFC estimates are based on current law and reflect all legislative changes to taxes and fees enacted through the 122nd Legislature’s 2nd Special Session. The estimates of state taxes included in this report are based on the March 2006 revenue forecast.

At the end of every major session, OFPR compiles a listing of tax and fees changes, which details all the legislative changes that increase or decrease revenue estimates that are classified as a tax or fee in the Bureau of Census analysis. Those compilations reflect the revenue effect as recorded through the fiscal note process at the time of enactment. After enactment, the RFC may adjust those estimates of the revenue effects in subsequent forecasts. This report factors in all legislative changes and adjusts those estimates for new information and changes since the original estimate. For the General Fund, Highway Fund and Fund for a Healthy Maine any legislative changes are reviewed and adjusted in each subsequent RFC forecast. This most recent revenue forecast, in particular, included several substantial downward adjustments to the original budgeted estimates of legislative revenue that were included in the supplemental budget bill enacted very late in the 1st Special Session of 2005, PL 2005, c. 457,.

Taxes and fees collected as dedicated revenue are estimated looking at recent trends and incorporate the effect of any increases or decreases enacted through the 122nd Legislature’s 2nd Special Session. OFPR analysts consult with department representatives to factor in these legislative changes with the trend analysis. In its most recent forecasts, the RFC has begun to review the dedicated revenue from various health care provider taxes and has provided revenue forecasts for these taxes. This is due to their size and substantial effect on the expenditure side of the General Fund budget. The estimates for these taxes were not revised in the March 2006 forecast, so the amounts included in this analysis for these taxes reflect the December 2005 estimates.

E. Liquor Store Revenue

In past analyses of tax burden, OFPR included net profit from liquor store operation in its calculation of tax burden. Although it is clear in the Bureau of Census classification materials that net liquor store revenue is not included as a “tax” for tax burden calculations (see Appendix C), this revenue was mistakenly included in past Census analyses of Maine’s tax revenue because of the manner in which it was reported. The analysis in this report makes adjustments to the historical data to separate out the revenue representing the net profit from liquor store operations.

F. Change of Budgetary Treatment of BETR and Circuit Breaker

Recent legislative changes to the budgetary treatment of the Business Equipment Tax Reimbursement (BETR) program and the Maine Resident Property Tax Program (Circuit Breaker program) were designed to get “credit” in the tax burden rankings for 2 major tax relief programs. Prior to fiscal year 2005, these programs were funded by General Fund appropriations and were not included as reductions to tax burden under the Census methodology, because tax burden is determined using only the total revenue from taxes and fees. The change of the treatment of these expenditures as reductions and transfers from individual income tax revenue reduces individual income tax on a budgetary basis. This analysis includes the effect of the budgetary treatment of these tax reimbursement programs and reduces tax burden, accordingly. It is assumed that the next time Census updates its tax data for fiscal years 2005 and thereafter, these changes will be reflected as reductions to Maine’s tax burden. The effect of the restructuring of these 2 tax programs reduces Maine’s tax burden calculation by nearly 0.3% of personal income and by $88 per capita in fiscal year 2006. Presented below is a table detailing the effect of these budgetary changes on tax burden. The graphs in this analysis also depict the separate effect of these changes.

Tax Burden Reductions Associated with Circuit Breaker and BETR Budgeting Change
2005 2006 Est. 2007 Est. 2008 Est. 2009 Est.
Circuit Breaker Adjustment $26,030,227 $44,328,964 $46,095,820 $48,440,013 $51,389,085

BETR Adjustment $0 $71,463,191 $68,146,508 $77,707,000 $79,646,983
Total - BETR and CB Adjustments $26,030,227 $115,792,155 $114,242,328 $126,147,013 $131,036,068
Adjustment - % of Personal Income 0.06% 0.27% 0.26% 0.27% 0.27%
Adjustment - $ per Capita $20 $88 $86 $94 $97

G. Compiling and Estimating Local Tax and Fee Data

Appendix B provides a comparison of fiscal year 2002 state and local tax data collected by Census to the data compiled by OFPR. The difference for state tax data is relatively small, OFPR data is approximately $15.1 million more than Census data. Census uses similar sources for its state data collections. However, the local tax data comparison shows a much wider discrepancy; Census data is $204.5 million higher than OFPR data in aggregate. There is no single reporting source for all local tax and fee data as there is for the state tax data. The Census methodology for sampling and estimating local tax data produces some wide discrepancies between the local tax data compiled by Maine Revenue Services. Looking at the detail of the Census data, OFPR is assuming that Census may double-count property tax revenue collected by municipalities for counties. Attempts to clarify this discrepancy have not been successful.

OFPR’s local tax estimates rely primarily on data collected by Maine Revenue Services’ Property Tax Division, which publishes an annual data report called the Municipal Valuation Return Statistical Summary (MVR). Municipalities are required to provide the data by law annually. It is published by Maine Revenue Services typically in the summer of each year. From that report, OFPR compiles the total local property tax commitment for each property tax year based on the status of property on April 1st. Although municipal revenues do not have consistent due dates, the total commitment is assumed to be collected in the next state fiscal year. Two hundred and eleven municipalities operate on a July-June fiscal year consistent with the State. Most of the remainder, or roughly half, operate on a calendar year basis. A hand-full operate using some other fiscal year. This, along with various different collection dates/methods (annual, semi-annual or quarterly), may reduce the accuracy of this assumption.

Motor vehicle and watercraft excise taxes are also reported annually on the MVR but are reported for the prior year’s collections. Because the prior year may have been a calendar year or the state fiscal year, an assumption is made to apply this data consistently over time to the prior state fiscal year.

The county share of the Real Estate Transfer Tax is calculated based on the total state share. Due to a lag in the timing of the payments to the state, actual county revenue from this tax may vary from the amounts reported in this report. This small component of county taxes was added for the first time to OFPR’s tax burden analysis. Previous OFPR estimates of local taxes excluded this relatively small portion of local taxes (approximately $2.0 million in fiscal year 2002). This update of Maine’s tax burden also includes for the first time an estimate of some of the other non-property tax local taxes and fees. This category of “taxes” is based on survey data collected by the Maine Municipal Association for fiscal year 2004. For this analysis, the single data point is used and estimates for other fiscal years are extrapolated back historically and projected forward from fiscal year 2004 using the Consumer Price Index. Non-property tax revenue collected by counties is not included in this analysis. OFPR will be working on refining these estimates for future updates and continue to work with Kennebec County (for county data) and the Maine Municipal Association for the municipal data. The data collection efforts that are required by the recently enacted state and local government spending limitations law should help with future updates.

H. Other Data Estimates

Projections of economic variables such as personal income and inflation as measured by the Consumer Price Index, are based on the Consensus Economic Forecasting Commission’s (CEFC) latest forecast, February 2006. Population data and projections are based on Census data tracked and provided by the State Planning Office. The historical data for personal income is from the US Department of Commerce, Bureau of Economic Analysis (BEA) and the historical population data is from Census.