Maine Revenue Services

 

Distribution Tables and Frequently Asked Questions

 

Maine Revenue Services (MRS) has produced a set of tax incidence tables that show the distribution of resident tax liability by income before and after LD 1495.  The tables also provide several measures of the change in tax by income group.  Many analysts find that these tables are a helpful tool for evaluating tax reform.  To see the potential impact of the tax reform, see the Maine Revenue Services distribution tables.

 

This document provides answers to frequently asked questions about how these tables are produced and how the various measures included in these tables should be interpreted.  There are three tables showing the distribution of the income tax, sales and use tax, and the income and sales tax combined.  The income tax and combined tables have more columns because there are both taxpayers with a tax cut and taxpayers with a tax increase in those tables.    

 

Additional information and other frequently asked questions are available at the State Planning Office.

 

Frequently Asked Questions

 

(1)   What models were used by MRS to model the tax changes made in LD 1495, and how are they constructed?

 

The MRS individual income tax microsimulation model is primarily based on a direct match of approximately 600,000 state and federal income tax returns from tax year 2000.  The tax return information is supplemented with returns filed through the Maine Resident Property Tax Program (i.e. circuit breaker returns) for 2000.  To account for non-income tax filers and to provide additional demographic and income information not reported on income tax returns a statistical match is made with the Public Use Microdata Sample (PUMS) from the 2000 Census of Population and Housing for Maine. Finally, consumption profiles are imputed to each record using information from the Consumer Expenditure Survey.  At the end of this process the MRS model has income, consumption and demographic information for over 665,000 tax families in the State of Maine.     

 

The sales and excise tax model utilizes detailed input-output tables from the Bureau of Economic Analysis (BEA).  These tables capture the flow of goods and services between business sectors and from businesses to consumers.  These national tables are scaled and adjusted to reflect the Maine economy.  Final consumer sales are split between sales to Maine resident consumers and non-resident consumers using data from the travel industry. 

 

(2)   If the models are based on data from 2000, how are these data used to model tax changes in 2010 and beyond?

        

For tax years 2001 to 2006 tax data is targeted to match actual state and federal income tax returns filed by Maine residents for those tax years. For tax years 2007 to 2013 the tax data is “aged” based on national and state economic projections provided by the Consensus Economic Forecasting Committee.  The sales and excise tax model is similarly extrapolated.  For fiscal years 2001 to 2008 the model is targeted to match actual sales tax receipts. For fiscal years 2009 to 2013 the tax data is “aged” based on national and state economic projections provided by the Consensus Economic Forecasting Committee.

 

(3)   Does the MRS model include the effects of tax reform on all Maine tax families or just those families that file federal or state tax returns?

 

The MRS models are designed to model the impact of tax changes on all Maine tax families regardless of whether they filed a state or federal tax return (See description of models in question (1)).  At the state level this is very important because while low income families may not have an income tax liability, almost every tax family has a sales, excise or property tax liability that needs to be accounted for in any analysis. 

 

(4)   The MRS distribution tables show the impact of tax reform on “tax families”.  What is a tax family?

 

A tax family is an individual, or a married couple who file a tax return jointly, along with dependents of that individual or married couple.  A tax family is different than the Census definition of household, and in some cases different than a federal or state tax return.  A household may consist of unrelated people.  For example, two persons cohabitating would be one household for Census purposes, but two different tax families because they are required to file separately for state and federal tax purposes.  For this reason there are many more tax families than households. 

 

Dependents with certain levels of income are required to file a separate federal and state tax return, but are claimed as a dependent on a parent’s tax return.  The MRS model combines these two tax returns into one tax family. 

 

(5)   What is expanded income?

 

Expanded income is a broader measure of income than is required for state or federal income tax purposes.  The purpose of expanded income is to get a more meaningful measure of a tax family’s ability to pay state and local taxes.  Income such as tax-exempt interest, total social security income, and certain welfare benefits are added back to federal adjusted gross income (FAGI) to achieve this broader measure of income.  For most taxpayers the difference between their FAGI and expanded income is minimal; for a small number of families, the difference can be significant. 

 

(6)   What taxes are included in the MRS incidence tables?

 

LD 1495 makes adjustments to the individual income tax, the sales and service provider taxes and the Maine Residents Property Tax Program.  The attached tables only reflect the impact of changes to the individual income tax and sales and service provider taxes on Maine residents.  MRS is capable of including all the major state and local taxes imposed in Maine, but does not include most of them in the attached tables because they were not affected by LD 1495.  

 

(7)   What incidence assumptions are used in the MRS incidence tables?

 

An incidence analysis tries to estimate who ultimately bears the burden of a tax regardless of who it is originally levied upon.  In estimating the incidence of LD 1495, MRS assumed that changes to the individual income tax were directly borne by the tax family.  Sales taxes on consumer sales are assumed to be borne directly by the consumer; this estimate is then broken out by resident and non-resident consumers.  Sales taxes levied on businesses are assumed to be split between the owners of capital, employees of the business and consumers of the businesses’ good or service.  In general, MRS assumes that 50% of the sales tax paid by business is borne by the owner, 25% by employees of the business and 25% by consumers of the good or service.  Adjustments are made to this assumption based on characteristics of the market the business is operating within.  These assumptions are consistent with other groups (Congressional Budget Office, U.S. Treasury, other state governments and the Tax Policy Center) that develop incidence tables.

 

(8)   Do the MRS incidence tables include behavioral responses to tax changes?

 

Yes.  The MRS sales tax model assumes that the price of goods and services taxed for the first time under LD 1495 or that have a change in the tax rate will increase and therefore reduce the demand for those goods and services. 

 

           

(9) What are the commonly used measures in these incidence tables?  What information do these measures convey?

 

A) Percentage change in after-tax income (Column 9)

 

Many analysts believe this is the most useful measure of the impact of a tax change on the well-being of taxpayers.[1]  This measure shows the proportional increase in economic resources for each income group.  In the attached table that shows the combined impact of LD 1495 on Maine resident tax families, the bottom 20% of families have an estimated 0.9% increase in their after-tax income while families in the upper deciles only have a 0.1% increase in their after-tax income.

 

 

B) Percent of taxpayers with a tax increase (Column 13) and percentage of taxpayers with a decrease (Column 10)

 

These figures show the percentage of taxpayers in each income group who pay less in taxes and more in taxes under tax reform.  Of course, these figures say nothing about the amount of gains or losses.  MRS estimates that 87% of Maine tax families will receive a net tax cut from LD 1495.  Most of the families with a tax increase fall into the upper deciles.

 

C) Average tax change (Column 7)

 

This column represents the average tax change for all families in each income group, including those with a tax cut and those with a tax increase.  It is not normalized by either original tax liability or by income.

 

D) Average tax decrease (column 12) and average tax increase (column 15)

 

These columns describe the average tax change, but only for families with a tax cut (column 12) or a tax increase (column 15).

 

E) Share of tax relief (Column 8)

 

This measure, unlike all the other measures, focuses on the aggregate reduction in net taxes.   Even if aggregate resident taxes were unchanged, tax reform would impact the distribution of resident tax liability and all the preceding measures would exist and have the same interpretation.  However, the share of tax relief would not even be defined if the aggregate resident tax change was zero. 

 

The share of tax relief is perhaps most easily interpreted when compared to the original distribution of taxes by income (column 2).  Consider a tax cut for which each income group received an equal percentage tax cut (e.g. column 6 equaled -2.9% for each income group in the combined table).  In this case, the share of tax relief (column 8) would equal the original distribution of taxes by income (column 2).  When column 8 exceeds column 2, the income group receives a disproportionately large tax cut, and its share of total taxes falls. 

 

(10) The share of tax relief in column 8 is different than the share that has been reported in several newspaper columns opposing tax reform.  Why is there a difference?

 

Column 8 equals column 5 divided by the aggregate net change in tax liability (the last row of column 5).  It takes the net tax change and shows what percentage goes to each income group.  The individual elements of column 8 sum to 100%.

 

In several instance the share of tax relief going to an income group has been incorrectly calculated as the net tax change for only families with a tax cut (column 11) divided by the aggregate net change in tax liability for all families (the last row of column 5).  The logic for why that calculation is wrong is equivalent to the logic of why it is impossible to split a pie four ways with each person receiving 40% of the pie.  The sum of the individual elements of column 11 are much larger (in absolute value) than the aggregate net tax change for all families.   If we calculated the percentage of tax relief this way for each income group and then summed over all income groups, the sum would exceed 100%.