Your Federal Quarterly Tax Payments are due April 15th Get Help Now >>



									                   STATE INCOME TAX REVENUES IN 2002 AND 2030:

                      Nina Manzi, Joel Michael, and Paul Wilson, Minnesota House of Representatives

                  INTRODUCTION                                 FEDERAL AND STATE TAX BENEFITS FOR THE ELDERLY

        implications for public finance of the baby                                    Federal
        boom generation (roughly individuals                      The federal income tax provides three tax benefits
born between 1946 and 1960). Although much of                  targeted to the elderly: (1) the partial exclusion
the focus has been on federal and state spending               of Social Security benefits from taxation; (2) an
programs for the elderly, there are also implications          additional standard deduction amount, and (3) the
for the revenue side of the budget equation. In part           elderly or disabled tax credit. The partial exclusion
because of special federal and state tax benefits,              of Social Security benefits provides the largest tax
elderly taxpayers face lower average effective                 benefit (OMB, 2005). Most Social Security benefits
federal and state tax rates than the nonelderly. Thus,         are not taxed. For higher income recipients, up to 85
just as the demand to finance public services for               percent of benefits can be taxed under a complicated
the elderly is increasing, federal and state revenues          formula, based on provisional income over dollar
may decline relative to personal income or other               thresholds that vary by filing status (Scott, 2005).
broad measures of economic activity.                           The dollar thresholds are not indexed for inflation and
   Using a stratified random sample of Minnesota                have diminished in real terms since their enactments
tax returns, we simulate the impact of an aging                in 1983 (first threshold) and 1993 (second threshold).
population on Minnesota income tax revenues                    Second, an additional standard deduction applies
between our base year, 2002, and 2030. Since                   to taxpayers who are 65 years or older: $1,200 for
Minnesota tax return data includes the age of the              single taxpayers and $950 for each married taxpayer
taxpayer and spouse, we can simulate effects for               in tax year 2005. Finally, federal law also provides a
different age cohorts within the elderly and non-              limited, nonrefundable elderly or disabled credit.
elderly populations. We use the macroeconomic
forecast by the state’s economic consultant (Global
Insight, Inc. or GII) to adjust the relative shares of                                 State
capital, labor, and retirement income to match the                All states grant at least the equivalent of the
projected shares in 2030. Some tax preferences                 federal tax preferences for the elderly. Most states
for the elderly are not indexed, so we adjust for              provide tax preferences for the elderly that are
the effect of inflation.                                        significantly more generous than under federal
   In general, our analysis finds that for state               law. The state provisions fall into three categories:
tax structures similar to Minnesota’s (i.e., those             (1) more generous exemptions of Social Security
allowing essentially only the federal prefer-                  benefits; (2) partial or full exclusion of various
ences for the elderly), the decline in tax revenues            forms of pension and similar retirement income;
resulting from the aging of the population will be             and (3) age-based general exemptions, deductions,
relatively modest (1.8 percent). While the growth              or credits. These benefits are summarized in the
in the elderly population with its lower effective             Appendix. Many state tax benefits for the elderly
tax rates (ETRs) reduces revenues, the reduction               use fixed-dollar amounts in their calculations. Only
will be partially offset by inflation eroding the               rarely are these additional amounts indexed for
value of the Social Security exclusion, increasing             changes in either prices or real income.1
ETRs for the elderly. States with more typical
elderly tax benefits (e.g., exemption of all Social
Security benefits and partial pension exclusions)                 DATA SOURCES AND MINNESOTA INCOME TAX
will experience much larger drops in revenue (4.4
percent). Exemption of all Social Security benefits                          Description of the Dataset
and all pension income would result in an even                  We use a stratified, random sample of 2002
larger decrease (10 percent).                                  Minnesota income tax returns prepared by the


Minnesota Department of Revenue (DOR). This                puting the tax base, and then applies a progressive
sample includes most items from the Minnesota tax          rate structure with three rates: 5.35, 7.05, and 7.85
return, including taxpayer and spouse date of birth        percent. Realized capital gains are taxed in full as
and most of the relevant items on sources of income        ordinary income.
and deductions from the federal return that must              Unlike most states, Minnesota does not provide
be filed as part of the Minnesota return. Identifiers        significant income tax benefits for the elderly.
were removed and masking procedures used to                Minnesota taxes both Social Security benefits and
insure confidential information was not disclosed           pensions on the same basis as the federal tax and
outside DOR. The original dataset consisted of             does not provide broad additional exclusions for
18,287 returns.                                            taxpayers over a certain age. Minnesota allows a
   The House Income Tax Simulation (HITS)                  largely inconsequential exclusion for low-income
model is a parameterized microsimulation model.            elderly taxpayers.4
HITS inflates data items from each return from
the base year to the year being simulated, using
45 growth factors. HITS recalculates liability for                        2002 BASE SIMULATION
each sample return and multiplies the result by the           Since we seek to isolate the effect of the changed
weight of each return to yield an estimate for the         age distribution in 2030 on state income tax rev-
entire population.                                         enues, we need to control for tax law changes. To
   We dropped nonresident and part-year resident           make the 2002 estimates consistent with those for
returns and returns for which age variables were not       2030, we assume that 2010 federal and state law
available.2 This reduced the number of records used        was in effect in 2002 (referred to throughout the
to 16,913. Ages were available for 93.7 percent            rest of this paper as “2002 base”) and will be in
of returns. Table 1 shows the percent of returns           effect in 2030 (referred to as “2030 projected”).
that reported age, and the percent of the various          That is, we assume that features of the federal
components of federal adjusted gross income                tax that are scheduled to expire in 2011 will be
(FAGI) and net Minnesota liability reported on             extended through 2030.
those returns. Returns not reporting age account              Because of data limitations or modeling difficul-
for a disproportionately high share of Schedule            ties, we do not simulate the effects of two minor
E income, but a roughly proportional share of tax          features of the Minnesota income tax, the alter-
liability (6.5 percent of liability came from the 6.3      native minimum tax (AMT)5 and the Minnesota
percent of returns that did not report age).3              subtraction for certain dependent K-12 education
                                                           expenses.6 These features involve relatively small
       The Minnesota Individual Income Tax                 and offsetting amounts of revenue (less than
   The Minnesota individual income tax uses fed-           0.6 percent). We also exclude Minnesota’s three
eral taxable income (FTI) as the starting point in         refundable credits for low-income taxpayers: the
computing its tax base. It requires a small number         state version of the earned income tax credit, the
of additions to and subtractions from FTI in com-          refundable and income-limited dependent care

                                          Table 1
   Proportion of FAGI Components and Net Liability Reported on Returns Not Reporting Age
Percent of:                          Returns reporting age                      Returns not reporting age
All returns                                  93.7%                                         6.3%
Earned income                                94.9%                                         5.1%
Capital income                               90.6%                                         9.4%
Retirement income                            95.2%                                         4.8%
Schedule E income                            86.7%                                        13.3%
FAGI                                         94.2%                                         5.8%
Net MN liability before credits              93.7%                                         6.3%
Earned income consists of wages, Schedule C (sole proprietor and partnership) income, and Schedule F (farm)
income. Capital income consists of dividends, capital gains, interest, and other gains or losses. Retirement income
consists of taxable and nontaxable Social Security benefits, pension income, and IRA distributions.

                                                                     NATIONAL TAX ASSOCIATION PROCEEDINGS

credit, and the K-12 education credit. All three of       elderly have more nontaxable income, older age
these credits could be viewed as social transfer          groups face lower ETRs than their younger coun-
programs that could be operated independently of          terparts, even with Minnesota’s minimal elderly
the income tax system.                                    preferences. The average ETR for taxpayers age 65
                                                          and older (2.90 percent) is only 68 percent of the
                    2002 Results                          average for those under age 65 (4.26 percent).
   Figure 1 shows the composition of total income
by age for the 2002 base.7 Throughout the paper,
total income is defined as FAGI plus nontax-                               2030 PROJECTION
able pensions, nontaxable Social Security ben-               Minnesota’s income tax revenue will change
efits, nontaxable IRA distributions, and federally         between 2002 and 2030 for many reasons. To iso-
exempt interest.8 As expected, earned income              late aging’s impact, we hold total population and
makes up over 50 percent of total income for all          real income constant at 2002 levels, but allow the
groups through age 64, while retirement income            age distribution and the mix of incomes (earned
makes up over one-half of income for returns              income, capital income, and retirement income)
reporting ages of 65 or older.9                           to vary. We compare the tax paid in the 2002 base
   Figure 2 shows the portion of total income that is     with the tax that would have been paid in 2002 if
nontaxable increasing as age increases, exceeding         both the population and income mix matched what
10 percent after age 60 and 30 percent after age 80.      is projected for 2030.
If each age cohort’s share of FAGI and nontaxable
income remained constant, all other things equal,                    Adjusting Population Shares
a growing share of total income would become                 Population shares by age group were estimated
nontaxable as the baby boom generation ages.              for both 2002 and 2030, using projections prepared
   Figure 3 shows average FAGI per taxpayer10             by the Minnesota Planning (2002). Because we hold
climbing sharply as age increases, topping $40,000        population constant, the number of young and mid-
for taxpayers age 40 to 44 and remaining over             dle-aged tax filers falls while the number of older
$40,000 until age reaches 60. The average drops to        filers rises. The share of taxpayers age 65 and over
less than $30,000 for taxpayers age 65 and older.         rises from 14.9 percent to 24.8 percent. As Figure 5
Figure 4 shows ETRs by age group. The ETRs are            shows, population falls for all age groups under age
computed relative to total income. Because the            55, and rises for all groups age 55 and older.

Figure 1:   Composition of total income by age group, tax year 2002


Figure 2:   FAGI and nontaxable income as shares of total income, tax year 2002

Figure 3:   Average FAGI by age group, tax year 2002

                                                         NATIONAL TAX ASSOCIATION PROCEEDINGS

Figure 4:   Effective tax rate by age group, tax year 2002

Figure 5:   Minnesota population by age group, 2002 base and 2030 projected


      Adjusting Income Shares and Inflation              from 2002 to 2030. Although population is held
   Our analysis projects income growth from 2002        constant, the change in the age mix increases the
to 2007 based on the February 2005 state forecast.      number of taxpayers (joint returns counted as two
Growth projected from 2007 to 2030 is generally         taxpayers) by 4.1 percent. Much of the decline in
based on GII’s long-term trend projection (GII          the younger population is among those under age
2005). Because GII’s projections do not include         20, few of whom file a tax return. Their absence
capital gains or pension income, we projected their     does little to reduce the number of taxpayers. The
growth from other sources. Capital gains reported       aggregate population of the remaining age groups
on federal tax returns are assumed to grow from         rises. Not only does the number of tax returns
2.28 percent of GDP in 2002 to 3.25 percent of          increase (by 2.5 percent), but a higher proportion
GDP in 2030—the ratio assumed for 2015 by the           are joint returns.12
U.S. Congressional Budget Office (CBO, 2005). As            Real income is held constant, but earned
a result, capital gains are assumed to grow 43 per-     income falls by 12.7 percent. Capital income rises
cent faster than GII’s projection of GDP between        by 34.1 percent, and retirement income rises by
2002 and 2030. Pension and IRA income reported          51.6 percent. Exempt income rises by 9.6 percent
on federal tax returns is assumed to grow from 4.0      while FAGI falls slightly (by 0.7 percent). Total
percent of GDP in 2002 to 7.4 percent of GDP in         tax liability falls by 1.8 percent.
2030, based on projections by CBO (2004).11 As             These changes combine the impact of changes
a result, pension income (taxable and nontaxable)       in population shares, income shares, and the CPI.
is assumed to grow 85 percent faster than GII’s         If we had changed only the population shares,
projection of GDP between 2002 and 2030.                earned income would have declined less (by 6.4
   Real total income is projected to roughly double     percent rather than 12.7 percent). Capital income
between 2002 and 2030. To hold real income con-         would have risen by 29.7 percent (rather than
stant at its 2002 level, all components are scaled      34.1 percent), retirement income by 50 percent
back proportionally; as a result, some taxpayers        (rather than 51.6 percent), and exempt income by
have lower incomes (particularly those with mostly      43 percent (rather than 9.6 percent). FAGI would
wage income) and others have higher incomes             have risen by 1.9 percent (rather than falling by 0.7
(particularly those whose income is mostly capital      percent), and total income by over 5 percent. Tax
or retirement income).                                  revenue would have risen by 1.7 percent, rather
   GII projects that Consumer Price Index (CPI)         than falling by 1.8 percent. Clearly, adjusting only
will roughly double between 2002 and 2030. This         the sample weights (while assuming that each
estimate deflates the nonindexed portions of tax law     sample taxpayer’s income and tax liability remain
by roughly one-half in the 2030 projection.             at 2002 levels) presents an incomplete picture of
                                                        the future.
  Projected Changes in Income, Tax Revenue, and            The relatively small reduction in tax revenue—
                Effective Tax Rates                     1.8 percent—is largely due to the lack of complete
   Table 2 summarizes the projected change in           indexing of tax parameters.13 In 2002, only 32
the number of taxpayers, income, and tax revenue        percent of Social Security benefits reported on
                                                        Minnesota tax returns were included in FAGI and
                  Table 2                               subject to tax. Under current law, that percentage
  Changes from 2002 Base to 2030 Projected              is projected to rise to 63 percent in 2030. If those
                                                        thresholds were fully indexed for inflation, the
                                2002 base to
                               2030 projected           share of Social Security subject to tax would rise
                                                        only to 36 percent.
Tax returns                          2.5%
                                                           Figure 6 shows the ETRs in 2002 base and 2030
Taxpayers (joint=2)                  4.1%
Earned income                      -12.7%               projected. The ETR rises by almost 20 percent
Capital income                      34.1%               for taxpayers age 65 and over (from 2.90 percent
Retirement income                   51.6%               to 3.43 percent of income). For those age 75 and
FAGI                                -0.7%               over, the ETR increases by 24 percent (from 2.47
Exempt income                        9.6%               percent to 3.05 percent of income). Roughly half
Total income                         0.0%
                                                        of each change is due to the failure to index the
Tax                                 -1.8%
                                                        Social Security thresholds and the cap on net capital

                                                                    NATIONAL TAX ASSOCIATION PROCEEDINGS

Figure 6:   Effective tax rates by age group, 2002 base and 2030 projected

losses (which both affect seniors more than oth-         Social Security benefits in 2030 were 20 percent
ers). The other half is due to the change in income      less, tax revenue would fall by 1.4 percent rather
shares (less earned income and more capital and          than 1.8 percent.
retirement income).14                                       Slower growth in capital gains income would
   As both the number of older taxpayers and their       have the opposite effect. Minnesota taxes capital
ETRs increase, the share of total tax paid by those      gains income at regular rates and most capital
age 65 and older more than doubles (from 10.6            gains are realized by higher income taxpayers, so
percent to 21.6 percent of total tax).                   the average tax rate on capital gains exceeds that
   Seniors will face higher ETRs in 2030—82              on most other forms of income. The 2002 base
percent of the average ETR for nonsenior tax-            had an unusually low level of capital gains. To
payers—than they do in 2002, when the average            test the sensitivity of our results, we recalculated
ETR for seniors is only 68 percent of the average        2002 tax assuming that capital gains income was
for nonseniors. The gradual increase in seniors’         30 percent higher—equal to the share of GDP pro-
ETRs in the coming years could increase pressure         jected for 2030. Other income types were reduced
on policymakers to provide exclusions to seniors,        proportionately to hold total income constant. Tax
particularly in states like Minnesota that offer few     liability in 2002 rose by 0.5 percent, so the drop
such benefits today.                                      between 2002 to 2030 would be 2.3 percent, rather
                                                         than by 1.8 percent. Use of a more typical base
          Sensitivity Analysis and Effect                year might therefore have made our results slightly
       on Projected Changes in Tax Revenue               more pessimistic.15
   Our results depend on how income shares                  To see whether Minnesota’s progressive rate
shift between 2002 and 2030. Because we hold             structure affected our results, we repeated the
total income constant (and retirement income is          analysis assuming a revenue-neutral flat tax on
taxed at a lower average rate than other income),        Minnesota’s tax base. Tax revenue fell by slightly
slower growth in retirement income would have            less (1.5 percent rather than 1.8 percent).
a favorable revenue impact. If total pension and            Our results also depend on how we define total
IRA distributions in 2030 were 20 percent less,          income, because we hold 2030 total income at the
we would show a smaller drop in tax revenue              2002 level. For example, including fringe benefits
(0.8 percent rather than 1.8 percent). Similarly, if     in our measure of income would have increased


                                               Table 3
                       Sensitivity of Tax Revenue to Alternative Assumptions
                                                                                         2002 base to
                                                                                        2030 projected
Baseline result                                                                             -1.8%
Assume pension income is 20% less in 2030                                                   -0.8%
Assume Social Security benefits are 20% less in 2030                                         -1.4%
Assume capital gains were 30% higher in 2002                                                -2.3%
Revenue-neutral single tax rate on Minnesota taxable income                                 -1.5%

the share of exempt income in 2030, and our               Security thresholds. If the thresholds had been
analysis would have shown a larger reduction in           indexed since 2002, projected 2030 tax revenues
tax revenue. On the other hand, if we had excluded        would have fallen by 1.9 percent. Aging would
nontaxable pension and IRA distributions, tax             have reduced revenues by 3.7 percent between
revenue would have risen between 2002 and 2030,           2002 and 2030, rather than by 1.8 percent. ETRs
rather than fallen.                                       for seniors would have risen only half as much as
   Finally, if Minnesota’s refundable tax credits had     under current law.
been included, revenues might have dropped less
(or risen instead). The credits reduced tax liability                     Pension Exclusions
by 3.2 percent in 2002, and 93 percent of the bene-          We simulate the effect of a $10,000 pension
fits went to taxpayers under age 50. Though largely        exclusion, both indexed and not indexed, as well
indexed for inflation, the changing age distribution       as an unlimited exemption for pension income.
would likely shrink their relative size.                  The $10,000 limit (assumed to apply equally to
                                                          married and single filers) is typical of the dol-
                                                          lar-limited pension exclusions allowed in many
       SIMULATION OF COMMON ELDERLY                       states. An unindexed $10,000 exclusion would
    TAX BENEFITS PROVIDED BY OTHER STATES                 reduce Minnesota revenue by 2.7 percent in both
   Aging’s impact on Minnesota’s revenue would            the 2002 base and the 2030 projection. Increases in
be significantly larger if Minnesota provided the          the number of beneficiaries and the average level of
additional elderly tax preferences common in other        pension income are offset by inflation eroding the
states. Table 4 shows how our results would change        real dollar value of the $10,000 limitation. Because
if Minnesota had provided such benefits both in            revenue falls by the same proportion in both years,
2002 and 2030.                                            aging’s impact would be the same as under current
                                                          law (-1.8 percent). ETRs for seniors, though lower
     Full Exemption of Social Security Benefits            in both years, would rise more rapidly than under
   If Minnesota fully exempted Social Security            current law.
benefits (as 28 states do), tax revenues would fall           If the $10,000 exemption were indexed, the cost
by 2.3 percent in the 2002 base and by 5.0 percent        in 2030 would increase from 2.7 percent of revenue
in the 2030 projection. Aging more than doubles           in 2002 to 4.6 percent of revenue in 2030. Aging
the relative cost of a full exemption for Social          would reduce revenue by 3.8 percent, rather than
Security benefits; a larger share of the population        1.8 percent. Seniors would see a smaller increase
receives benefits, the average real benefit rises, and      in their ETRs.
a higher portion of benefits are subject to tax under         A full pension exclusion would reduce revenue
current law. If Minnesota fully exempted Social           by 6.4 percent in the 2002 base and 12.6 percent
Security benefits, aging would reduce revenue by           in the 2030 projection. With such an exclusion,
4.5 percent, rather than 1.8 percent. The increase        aging would reduce revenue by 8.3 percent
in ETRs for seniors would be smaller (though still        between 2002 and 2030.
                                                               Full Social Security and Pension Exclusion
        Indexing Social Security Parameters                 The most generous senior preferences (avail-
  The President’s Advisory Panel on Federal               able in five states) include full Social Security and
Tax Reform (2005) proposed indexing the Social            pension exclusions. This would reduce Minnesota

                                                                              Table 4
                                                               Revenue Effects of Elderly Tax Benefits
                                           Revenue impact
                                             as percent of                                         ETR for taxpayers       ETR for seniors as % of
                                                                         Total change
                                                                                                   age 65 and older         ETR for nonseniors
                                      2002                 2030          2002 to 2030
                                      base              projected       with preference        2002              2030          2002        2030
                                    revenues            revenues         in both years         base            projected       base      projected
      Current law                      NA                  NA                -1.8%             2.90%              3.43%        68%          82%
      Full exemption of               -2.3%               -5.0%              -4.5%             2.38%              2.74%        56%          66%
      Social Security

      Indexing of Social Security      NA                 -1.9%              -3.7%             2.90%              3.16%        68%          76%
      benefits, 2002 to 2030

      $10,000 pension exclusion,      -2.7%               -2.7%              -1.8%             2.50%              3.14%        59%          76%

      not indexed

      $10,000 pension exclusion,      -2.7%               -4.6%              -3.8%             2.50%              2.93%        59%          72%

      Full pension exclusion          -6.4%               -12.6%             -8.3%             1.89%              2.04%        46%          51%

      Full exemption for Social       -4.9%               -7.4%              -4.4%             2.00%              2.48%        48%          61%
      Security and $10,000
      pension exclusion,
      not indexed

      Full exemption of Social        -8.1%              -15.8%            -10.0%              1.51%              1.60%        37%          41%
      Security and pension
                                                                                                                                                     NATIONAL TAX ASSOCIATION PROCEEDINGS

revenue by 8.1 percent in the 2002 base and 15.8                        Four records for which age was unavailable accounted
percent in the 2030 projection. Aging would reduce                      for nearly one-quarter of this Schedule E income.
income tax revenue by 10 percent between 2002                           The exclusion follows the federal credit for elderly or dis-
and 2030, rather than by 1.8 percent. The average                       abled. It benefits about 11,000 taxpayers per year, at an
                                                                        annual cost of about $1 million, (Minnesota Department
ETR for seniors would be halved in 2002 (and
                                                                        of Revenue 2004) and is not indexed for inflation.
by even more in 2030), and would rise by only a                     5
                                                                        Because the Minnesota AMT exemption is not indexed
small amount during that period. The average ETR                        for inflation, over a 25-year period it would become
for seniors in 2030 would fall from 3.43 percent                        a major factor if the law were not modified. This
to 1.60 percent, to only 41 percent of the average                      seems unlikely, though, and would cloud our efforts
nonsenior ETR.                                                          to isolate the effects of an aging population and elderly
                                                                        preferences on regular tax liability. Taxpayers age 65
                                                                        and older constitute 15 percent of taxpayers but only
                                                                        7 percent of those with state AMT liability.
                      CONCLUSION                                    6
                                                                        We lack information to model long-run growth of this de-
   These results suggest that a state (like Min-                        duction, which reduces revenue by about 0.2 percent.
nesota) that follows federal law in taxing Social                   7
                                                                        We use taxpayer age to assign joint returns to an age
Security benefits and pension income may not face                        group.
large declines in tax revenue due to the aging of                   8
                                                                        This definition is narrower than the definition of
the population. The relative growth in retirement                       comprehensive income used by Edwards and Wallace
income and capital income (much of it received                          (2004). It excludes unrealized capital gains (not includ-
by older taxpayers), combined with the lack of                          ed in the Minnesota sample), retirement contributions,
indexing of the Social Security tax thresholds, will                    and self-employed health insurance deductions.
raise the ETR of older taxpayers. As a result, tax                      “Other” includes business income reported on Sched-
revenues projected for 2030 are only 1.8 percent                        ule E, state income tax refunds, alimony received,
                                                                        unemployment compensation, other income and
lower than in the 2002 base. States with more
                                                                        losses, reduced by “above-the-line” subtractions on
generous elderly preferences may face a different                       lines 23 to 33 of the 2002 federal Form 1040. See Note
future, however. Our results suggest that revenue                       in Table 1 for definitions of other terms.
decreases could range from 4.4 percent to 10.0                  10
                                                                        Married joint returns were counted as two taxpayers.
percent depending on the preferences offered.                           Over one-half of returns from taxpayers age 35 to 79
   Future work should include more complete                             are from married joint filers, while most returns from
modeling of future income changes, particularly for                     younger and older taxpayers are from single or head
pension and IRA distributions, especially in light of                   of household filers.
recent policy proposals by the administration and                       Our analysis assumes identical 2002 to 2030 growth
                                                                        rates for taxable pensions, nontaxable pensions, tax-
the Tax Reform Advisory Panel. Exploration of in-
                                                                        able IRA distributions, and nontaxable IRA distribu-
creased labor supply among older taxpayers may also                     tions. It would be preferable to model these separately
yield valuable insights. Updating the analysis using                    to account for changes in retirement savings. CBO
the 2003 income tax sample and updated version                          notes its estimates are most sensitive to the assump-
of the HITS model (available in early December)                         tions about “the degree to which taxpayers switch to
would help identify any bias from using 2002, which                     Roth IRAs and ‘back-loaded’ 401(k) accounts” (CBO
appears to be an atypical year with respect to capital                  (2004), p. 1). Given the fairly recent availability of
gains realizations, as the base year.                                   these types of plans – tax year 1998 for Roth IRAs
                                                                        and 2006 for back-loaded 401(k)s – the growth rates
                                                                        for nontaxable distributions should not be assumed to
Notes                                                                   be the same as taxable pensions.
                                                                        Because we hold income constant, the 4.1 percent
    California, Idaho, Kentucky, Maryland, Michigan,                    increase in the number of taxpayers reduces aver-
    and Montana index various preferences for the elderly               age income per taxpayer by about 4 percent. With
    (Manzi et al., 2006).                                               a progressive tax, spreading the same income over
    Nonresidents’ returns often reflect unusual circum-                  a larger population reduces tax revenue. If we had
    stances of little relevance to elderly tax benefits. For             held the number of taxpayers constant (rather than
    example, they often are high-income investors who                   population), tax revenue would have fallen only 0.2
    own real estate or business interests in Minnesota.                 percent between 2002 and 2030, rather than the 1.8
    We dropped part-year residents, since we could not                  percent shown in Table 2. The 1.6 percent difference is
    distinguish them from nonresidents.                                 explained by noting that the average marginal tax rate

                                                                               NATIONAL TAX ASSOCIATION PROCEEDINGS

     (roughly 5.6 percent) is roughly 1.4 times the average        Global Insight, Inc. The U.S. Economy: 25-Year Focus,
     tax rate (about 4.0 percent). Holding the number of              Waltham, MA, 2005.
     taxpayers constant would raise the tax rate on 4 percent      Manzi, Nina, Joel Michael, and Paul Wilson. State Income
     of total income by 40 percent (from 4.0 to 5.6 percent),         Tax Revenues in 2002 and 2030: Impact of the Retire-
     thereby raising total revenue by 1.6 percent.                    ment of the Baby Boom. State Tax Notes 39 (January
     The principal unindexed feature (other than the Social           2006): 215-240.
     Security thresholds) is the $3,000 cap on capital losses.     Minnesota Department of Revenue. Tax Expenditure
     Estimated inflation between 2002 and 2030 reduces the             Budget, Fiscal Years 2004 to 2007. St. Paul, MN,
     $3,000 cap to $1,489 in real terms. In a typical year,           2004.
     this would increase reported capital gains realizations       Minnesota Planning, State Demographic Center. Min-
     by about 2.6 percent. Because 2002 was atypical, with            nesota Population Projections 2000 – 2030. St. Paul,
     a large number of filers reporting the maximum net                MN, 2002.
     loss, we limited the impact to 2.6 percent.                   National Conference of State Legislatures. State Personal
     The average ETR drops by 2 percent for taxpay-                   Income Taxes on Pensions and Retirement Income:
     ers under age 65 (from 4.16 to 4.06 percent). With               Tax Year 2004. Denver, CO, Nov. 3, 2004.
     income held constant, average incomes fall for younger        Scott, Christine. Social Security: Calculation and His-
     taxpayers, reducing ETRs under a progressive tax.                tory of Taxing Benefits, CRS Report for Congress
     One-half of the income drop for younger taxpayers is             (Jan. 14, 2005).
     due to the declining share of earned income. The other        The President’s Advisory Panel on Federal Tax Reform.
     half is due to the projected increase in the number of           Simple, Fair, and Pro-Growth: Proposals to Fix
     taxpayers, as described in note 12, which spreads total          America’s Tax System. Washington, D.C., 2005.
     income over more filers.                                       U.S. Congressional Budget Office.
     Retirement income was a relatively high share of                 Tax Deferred Retirement Savings in Long-Term
     income in 2002, though, which may somewhat offset                    Revenue Projections. Washington, D.C., 2004.
     this effect because retirement income is taxed at                The Budget and Economic Outlook: Fiscal Years 2006
     relatively low rates.                                                to 2015. Washington, D.C., 2005.
                                                                   U.S. Office of Management and Budget. Analytical Per-
                                                                      spectives, Budget of the United States Government
References                                                            Fiscal Year 2006. Washington, D.C., 2005.
Edwards, Barbara and Sally Wallace. State Income Tax               Wisconsin Legislative Fiscal Bureau. Individual Income
   Treatment of the Elderly. Public Budgeting and                     Tax Provisions in the States, Informational Paper #4,
   Finance 24 (June 2004): 1-20.                                      Madison, WI, 2005.


           State Income Tax Preferences for the Elderly (single filers; tax year 2004)

                                                 NATIONAL TAX ASSOCIATION PROCEEDINGS

                        Appendix (continued)
State Income Tax Preferences for the Elderly (single filers; tax year 2004)


To top