Back to Tax Revision Commission main page — Back to list of research papers
|
Columns DCWatch
Archives Elections Government and People Budget issues Organizations |
The District of Columbia's Individual Income Tax A Research Report to the District of Columbia Tax Revision Commission January 6, 1998 Robert P. Strauss The author is Professor of Economics and Public Policy at the Heinz School, Carnegie-Mellon University, Pittsburgh, Pennsylvania, 15213-3890. Email: RS9F@Andrew.CMU.EDU; Home Page: http://www.heinz.cmu.edu/~rs9f/ . The research reported in this paper was conducted under a signed confidentiality agreement with the District of Columbia which enabled the author to analyze District and federal tax return information for tax administration purposes. The author wishes to thank Dr. Julia Friedman, Dr. Modibo Coulibaly, and Dr. Mohammed Hassan for their assistance in facilitating the reported research. The author also wishes to thank Ms. Emily Gross and Ms. Beth Kils of the Statistics of Income Division of the Internal Revenue Service for providing a variety of summary information. Responsibility for the analysis and opinions in this paper and any errors rests solely with the author.
1.1 Some Public Finance Principles and Personal Income Taxation State and local governments have generally been enabled to employ a broad arsenal of revenue instruments to finance state and local services. The US Constitution imposes few impediments to their use of property, sales, gross receipts, excise, franchise and income taxes. Given the multiple and sometimes conflicting objectives of any tax system to:
a question arises about how a small geographic area such as the District of Columbia should employ a tax as visible as a personal income tax. When raising public revenues to finance public services, there is merit is determining at the outset if the service is general or narrow in impact. If the latter, then a case can be made for supporting its budgetary cost through benefit related taxes or even fees. At the local level, resident use of police and fire services may thought to be measured by the value of protection to the real estate owned by the resident; a local property tax applied at a proportional rate can be viewed as a benefit tax. Other services such as public education or health are provided to achieve distributional objectives. Here, one typically favors financing such activities through ability to pay taxes such as an income or broad based consumption tax. When the geographic area of a government is small, one typically expects a government which covers a larger geographic area to assist in financing income redistribution activities. Otherwise, one asks poor areas of a state to finance their contributions to income maintenance and the provision of services which are income redistribution in kind. As the District has no higher level of government other than the federal government to look towards for assistance in financing services, there is likely to be a constant tension between the service needs of the community, and the willingness of its residents to finance it.(1) Tax administration at the federal, state and local level depends critically on voluntary compliance. It is useful to remember that when the tax paying public believes that neighbors are not adequately contributing through taxes to pay for public services, and/or the quality of public services provided is not perceived to be commensurate with the taxes levied, then voluntary compliance can suffer. It should also be remembered that when elected officials at any level of government argue for tax cuts to encourage taxpayers to relocate to an area, they are also, at least in the short run, reducing budgetary revenues which must be associated with service reductions. Whether such revenue reductions reduce educational services or public safety services, the attraction of an area can be reduced when taxes are cut in the short run. Unlike most states, the finances of the District are impacted by two other important phenomena:
While the District is not enabled, as a condition of its Home Rule Act of 1971, to impose a personal income tax on commuters, the renegotiation of the federal payment and reshuffling of financing and service responsibilities which have occurred in 1997 make the analysis of a commuter tax of some interest to those seeking to reform the District's tax system. 1.2 Personal Income Taxes among the States State taxation of personal income dates back to colonial America when property and income taxes were combined in the form of "faculty taxes.'' Pennsylvania levied a 1% tax on salaries in 1840, and income taxes were prominent in the North and South to finance the costs of the Civil War.(3) The first modern state personal income tax was adopted by Hawaii in 1901. Wisconsin became the first continental state to adopt a personal income tax in 1911; it had a top marginal tax rate of 6%. By the close of that decade, eight other states(4) had adopted personal income taxes, and another six(5) the following decade. The Depression witnessed an additional sixteen states adopting a personal income tax(6). The District of Columbia adopted its income tax in 1947, and Alaska adopted its in 1949. From 1961 through the mid-1970's 11 additional states added personal income taxes(7).(8) 1.3 History and Importance of DC's Personal Income Tax The District of Columbia's individual income tax dates back to 1939, although the broad based version was adopted in 1947. It has amounted to about 25% of the District's annual own source tax collections over the past decade. With regard to the District's neighboring states, Virginia's state and local governments relied on the individual income tax for 26% of its tax revenues, while Maryland's state and local income taxes accounted for 38% of total Maryland state and local taxes.(9) Nationally, state and local governments relied on the individual income tax for one out of every five tax dollars raised(10). Thus, the District is above average in its reliance on the personal income tax to finance services to District residents, but in the region relies less than its major competitor (Maryland) for District residents seeking to move out of the District(11) Nationally, state and local personal income taxes averaged $878 per capita, and $1,141 in the District of Columbia in 1993/4. In Maryland the percapita figure was $980 and in Virginia it was $581. Nationally, the average state and local personal income tax burden was 2.4% of personal income as measured by the National Income and Product Accounts; for the District it was 3.8% of resident personal income as measured by BEA, while for Maryland it was 4.1% of resident personal income, and 2.7% of resident personal income in Virginia.(12) Table 1 compares 1993/4 personal income tax burdens to those in 1975. While the District's burden rose, it was due primarily to personal income rising more slowly than taxes. DC's 1993 personal income tax collections were 4.5 times that of 1975, but it's personal income in 1993 was only 3.4 times that of 1975. In Maryland, both personal tax collections in 1993 were about 4.9 times that of 1975.(13)
In 1989, 310,000 DC individual income taxpayers had a personal income tax liability of $534.9 million, while in 1995, 251,000 taxpayers had a net personal income tax liability of $556.0 million. Had there been as many taxpayers in 1995 as in 1989, and had they paid the average per return tax in 1995 of those with similar adjusted gross incomes, 1995 income tax revenues would have been $39.1 million or 7% higher.(14) The District's decline in resident population and decline in resident tax paying population, and the resulting sluggish revenues will be a constant focus of the research reported below. This paper seeks to address four broad questions related to the District's personal income tax:
To address these questions the paper is organized as follows: Section 2 examines the major components of the DC personal income tax in terms of structural as well as empirical characteristics, and compares it and its tax burdens to Maryland and Virginia. The next several sections examine the issue of in and out migration of DC taxpayers. Section 3 examines over the period 1969-95 resident population and income of residents and non-residents as measured by the Bureau of Economic Analysis; Section 4 examines aggregate migration into and out of the District based on federal tax returns; Section 5 examines migration into and out of the District through the analysis of 1989 and 1995 DC tax returns, and examines if changes in the level of crime in the District are associated with movements of taxpayers by DC zip code; The next two sections examine policy options for the DC personal income tax. Section 6 examines policy options that would enhance the revenue productivity and ease of administration of the DC individual income tax through simplification and greater conformity with the Internal Revenue Code; Section 7 examines the arguments for and against DC taxation of commuters, empirical aspects of them, and possible rates, revenues and administrative issues associated with a commuter tax. Section 8 concludes. 2 Structural Features of the DC Individual Income Tax The District of Columbia is one of 37 states to impose an individual income tax based initially on an income concept derived from the Internal Revenue Code. Of the states with an individual income tax, only five(15) are uncoupled from the IRC. The District's individual income tax follows 25 others in beginning its definition of income subject to tax by reference to federal Adjusted Gross Income. Five states begin with federal taxable income, and three (North Dakota, Rhode Island, and Vermont) base their individual income tax on federal liability.(16) Over the last 30 years, changes in tax rates and bracket have been enacted. In 1965 the top marginal tax rate was 5% of taxable income in excess of $25,000, and was increased to 11% in 1976. The top marginal rate was lowered to 10% for tax year 1987, and 9.5% for tax year 1988. In 1970 DC conformed its taxation of capital gains to the federal definition, and in 1975 conformed personal exemption and child care deductions.(17) Since 1988, the rate of tax has been 6% for taxable incomes under $10,000, 8% for taxable incomes between $10,00 and $20,000, and 9.5% for taxable incomes in excess of $20,000 of taxable income. Over time the value of the personal exemption has been increased: $885 for tax year 1987; $1,025 for tax year 1988; $1,160 for tax year 1989; $1,270 for tax year 1990, and $1,370 for tax year 1991 to the present.(18) A key element of the taxation of household income is the determination of whether or not the income is subject to a jurisdiction's authority to tax, and whether or not taxes imposed at the place of work are recognized, through credit or deduction, at the place of residence. Given the generally high level of mobility of individuals and households(19) and the fact, as noted above, the District's rate of population mobility is about twice the national figure of 6%, the determination of residence has significant revenue implications as well as administrative implications for the District's personal income tax administration. Four concepts of "resident'' may be found among state personal income taxes:(20)
Evidence of being domiciled often includes the registration of a motor vehicle, maintenance of bank accounts, state of driver's license, and state for voting purposes. The District taxes a resident's income, but can not, under Paul S. Davis v. D.C. tax the income earned outside of the District of an individual prior to his becoming a resident. A resident is defined for DC personal income tax purposes to include an individual who is either:
The District also taxes non-resident's income earned in the District from unincorporated business sources at graduated rates of 6.0 to 9.5% plus a surcharge of 2.5% less a $5,000 exemption. Non-resident personal service income, where capital is not a material income-producing factor, is exempted from the non-resident tax.(21) Thus, non-resident income earned by the legal profession is exempt, but partnership income earned from the rental of apartment buildings owned by non-resident partners is taxable. The District individual income tax form instructions add an obligation to file a return if:
Members of Congress who maintain a place of abode in the District in relation to their attending sessions are not taxable, nor are:
In addition, foreign embassy personnel are generally exempt from DC individual income taxation. A person for Maryland personal income tax purposes is a resident if he:
Maryland goes on to define a resident to include, for the part of the taxable year that an individual resides in Maryland, an individual who:
Virginia defines a resident as follows:
Moreover, Virginia defines a part-year resident as: A. Any person who, during the taxable year, becomes a resident of Virginia, whether domiciliary or actual, for purposes of income taxation, by moving to the Commonwealth from without during such taxable year, shall be taxable as a resident for only that portion of the taxable year during which he was a resident of the Commonwealth and his personal exemptions shall be reduced to an amount which bears the same ratio to the full exemptions as the number of days during which he was a resident of the Commonwealth bears to 365 days. No person to whom the preceding sentence applies shall be entitled to any credit on his income tax payable to Virginia for any income tax paid to the state or other jurisdiction of his former domicile or actual residence for that part of the taxable year during which he was a domiciliary or actual resident of such other state or jurisdiction, notwithstanding the provisions of Sec. 58.1-332. B. Any person who, on or before the last day of the taxable year, changes his place of abode to a place without the Commonwealth with the bona fide intention of continuing actually to abide permanently without Virginia shall be taxable as a resident for only that portion of the taxable year during which he was a resident of Virginia and his personal exemptions shall be reduced to an amount which bears the same ratio to the full exemptions as the number of days during which he was a resident of this Commonwealth bears to 365 days. The fact that a person who has changed his place of abode, within six months from so doing abides again in the Commonwealth, shall be prima facie evidence that he did not intend permanently to have his place of abode without Virginia. The fact that a person has removed his abode to a place without the Commonwealth is not conclusive evidence of a change of domicile. C. Any person who is taxable as a resident of the Commonwealth for only a portion of a taxable year because he moved to this Commonwealth from without Virginia during the taxable year as set out in subsection A, or because he changed his place of abode during the taxable year to a place without Virginia as set out in subsection B, and who, as a nonresident of Virginia for any other part of the taxable year derived income from any property owned or from any business, trade, profession or occupation carried on in Virginia shall be taxable as a nonresident with respect to such income as provided in Sec. 58.1-325. Disputes over DC residency rules have been prominent for a long period of time, and there is significant case law dealing with: the precise nature of having a DC domicile in comparison to that of another state of residence, the determination of intent to return to another place of domicile, and the domiciliary treatment of Foreign Service officers and others appointed by the President to positions in the Executive Branch of government. Disputes in Maryland and Virginia over whether their tax statutes allow Maryland and Virginia residents to take credits for DC taxes paid against their Maryland and Virginia resident income tax liability have also been prominent. While the Maryland statute permits residents a credit for "income tax'' paid to "another State upon such part of his net income,'' the Maryland Supreme Court in 1957 denied a credit paid for the DC tax on unincorporated business, because the tax was determined by the Court to be a privilege tax rather than an income tax. The Virginia Supreme Court reached the opposite conclusion about the nature of the unincorporated business income tax, and allowed its residents a credit against the Virginia individual income tax on the same set of facts in 1990.(22) Recently,(23) the Virginia tax department's denial of a Virginia resident's claim for a credit for DC non-resident income tax on unincorporated business income paid was upheld, because the DC non-resident income tax was found by the Virginia Supreme Court(24) to be a commuter tax in violation of the DC Home Rule Act. Virginia currently provides a credit only for other state's taxes which are legal and authorized under other state's laws. The DC personal income tax recognizes five major filing statuses:
As is well known, the number of DC individual income tax filers has dropped over time. Table 2 and Table 3 display for 1989 and 1995 the number individual income tax filers by AGI class and type of filing unit. Note that the AGI classification is based on the sum of each person's federal adjusted gross income shown on Line 1 of the D-40 and is as close an approximation to household economic income as is possible. Several things are immediately evident from an inspection of these two tables. First, the number of tax returns dropped overall by 20% during this 6 year period.(25) Second, the fall in returns occurred in the under $45,000 AGI classes; there was some growth in the highest brackets overall. Note that the number of taxpayers in the $100,000-500,000 income class grew by 30%. Second, while there was modest overall growth in the number of tax filers in the AGI classes above $45,000, this was not the case for Married Combined Filing Separately; their numbers dropped in all brackets except the $100,000 to $500,000 AGI class. Overall, this group of taxpayers showed the largest percentage reduction overall among all filers: their numbers fell by 31%. Table 4 displays the results for 1995 of matching DC taxpayers to their federal tax returns.(26) Overall, there were 246,399 returns which could be matched by Social Security number. Note there were 46,190DC tax filers who did not file federal tax returns; this likely reflects the fact that the DC tax entry point is well below the federal. Thus, 29,432 DC single taxpayers filed for DC tax purposes but not for federal tax purposes. The relationship between DC and Federal filing status is generally quite strong. All but 2,179 out of 135,231 overall DC single taxpayers filed as single taxpayers for federal purposes; all but 1,208 out of 55,206 DC Head of Household filers filed as head of household. Table 5 displays the amount of DC tax paid by different filing units in 1995 scaled in $1,000s. Single taxpayers were the largest filing group in 1995: overall they paid $246.4 million out of the $556 million in 1995 tax liability of matched DC and federal tax returns. This was 44% of total tax liability although they were 54.9% of total 1995 DC tax filers. The next largest group in terms of tax liability in 1995 was Married Combined filing Separately; they had $149.2 million in 1995 liability or 26.8% of the total; however, as noted above in Table 4, such taxpayers numbered only 19,693 or 8% of total DC tax return filers. Also, note that $32 million of DC Married Combined Separately liability was attributable to those for whom no federal returns could be matched. The third most important filing status in terms of tax liability was Married Filing Jointly; their tax liability was 15.1% of the total, while they constituted 7.7% of tax returns filed for DC tax purposes. The Head of Household filers were 22.7% of total returns but only 9.9% of tax liability. Single parents evidently have significantly lower taxable income than couples filing joint or combined separate returns. Derivation of taxable income for DC tax laws begins with federal adjusted gross income. The DC tax instructions indicate that the taxpayer must first complete his federal return before completing the DC tax return. All federal adjustments are provided to DC taxpayers,(27) although there are a series of add-backs as well. Excluded from DC gross income are:
In tax year 1995, DC taxpayers reported on their DC tax returns $10.191 billion in federal adjusted gross income, additions of $23.6 million, and subtractions of $1.015 billion; DC adjusted gross income in 1995 was $9.179 billion.(28) For those DC tax returns for whom federal tax returns could be matched, DC federal AGI was 98.7% of AGI on the IRS transactions file.(29) 2.5 Personal Exemption, Deductions, Tax Rates DC adjusted gross income is reduced to taxable income by reductions for the larger of the standard or itemized deductions, and personal exemptions. A standard deduction of $2,000 is available for Single, Head of Household, and Married filing jointly filers, and $1,000 each for Married filing separately or Married filing combined separately. Alternatively, the taxpayer may take itemized deductions shown on their federal return (Schedule A) with addbacks for deductions taken during periods of non-residence, and deductions for DC taxes. As noted earlier, the DC income tax has three income brackets, and three marginal tax rates (see Table 6).
Additional personal exemptions are accorded for those over age 65, and blind. While DC's personal exemption was increased several times since the late 1980'sit is instructive to compare its gross tax entry point to the federal poverty line for different size households, as well as to the tax entry points for DC's neighboring states' personal income taxes. Table 7 displays these data for households of size one to nine for 1995. It is evident that the federal income tax generally does not tax persons or households below the poverty line. Table 8 displays the gross tax entry points as a percent of the 1995 federal poverty line. DC, Maryland, and Virginia each has gross tax entry points at below 1/2 of the federal poverty line. For a District couple of three (husband, wife, and one child) with earnings at the above poverty threshold of $12,267, the couples' taxable income would be $6,157 ($12,267- the standard deduction of $2,000 and - $4,110 for the three exemptions (3 x $1,370). Gross tax due would be $369 (.06 * taxable income of $5,157); however, the Low Income Credit for 1996 for this family (joint return with three exemptions) is $494 and thus eliminates taxation of the family at the poverty line. If the husband worked full time to earn the $12,267, he would be subject to DC withholding(30) and have to file a return to obtain the refund due to the Low Income Credit.(31) Maryland alleviates the potential problem of taxing low wage workers below the poverty line in two ways. First, Maryland has a $11,800 minimum federal AGI filing requirement of $11,800 for joint returns, ($6,550 for Single returns, and $8,450 for Head of Household). Second, the Maryland personal income tax creates a reduction in federal adjusted gross income of earned income which is less than the poverty line (the so-called poverty income deduction). For the above family of three in 1996, earnings below the poverty income level of $12,980 are subtracted from federal adjusted gross income before the standard deduction and exemptions are deducted from federal AGI. Thus, in Maryland the family of three would be tax-free since income reduced by earnings would be further reduced by the standard deduction and the value of personal exemptions. Finally, since the local county income tax is a percentage of the state (60% in Montgomery and Prince George's counties), the household would not pay a local income tax. Virginia partially addresses the problem of taxing households with income at or below the poverty line through a rate schedule with four brackets of: 2% for taxable incomes under $3,000, 3% for taxable incomes between $3,000 and $5,000, 5% for taxable incomes of $5,000 to $17,000, and 5.75% for incomes in excess of $17,000.Thus for family of three (husband, wife, one child), in Virginia earning the poverty line of $12,267, the first $7,400 of earnings is tax-free (adjusted gross income is reduced by the $5,000 standard deduction plus the exemptions of $2,400 (3 x $800) ) which subjects the taxable income of $4,867 to a tax of $116.01 (2% of the first $3,000 or $60 plus 3% of the remaining $1,867 or $56.01). As already noted, DC tax law accords low income earners a Low Income Tax Credit. DC provides tax credits for:
2.7 Progressivity of DC Income Tax: 1989 vs. 1995 While the primary purpose of any revenue source is to finance needed public services, the individual income tax also is used by many jurisdictions to achieve agreed upon distributional objectives. The DC individual income tax does not contain any refundable features. Tax progression is usually defined to mean that the rate of taxation should rise with ability to pay. There are a variety of ways to measure progression, perhaps the most intuitive is to examine across income classes the ratio of net taxes to as broad a measure of income as possible. Due to the limitations of data available to this study, federal adjusted gross income as reported to DC is the measure of ability to pay. Table 9 shows for 1989 and 1995 the number of taxpayers in each year (N89 and N95 respectively) and three representative taxpayers in each income interval. For example, in the fourth income interval ($5,000-$15,000), there were 23,757 DC taxpayers (aggregated to one filing unit). The column denoted Q195 displays the effective tax rate of the 25th percentile of the distribution of all 23,757 taxpayers, ordered from lowest to highest effective tax rate. That tax return showed an effective tax rate of 1.6% in 1995, and 2.3% in 1989. Column 6 and 7 display the median effective tax rate in 1989 and 1996 for the same income interval: it fell from 3.5% in 1989 to 3.0% in 1995. If one examines a given part of the distribution of effective tax rates, one finds that they rise in both 1989 and 1995 for all points in the distribution except for the highest income group ($500,000+). There were 364 1995 taxpayers with federal adjusted gross income of $500,000 or more and effective tax rates of less than 5%. Inspection of this high income, low effective tax rate group indicates that some of the taxpayers were in the District less than a year, many had very large aggregate subtractions and itemized deductions. Unfortunately, the District does not key enter the underlying detail of subtractions and itemized deductions to ascertain the economic reality of the reductions in taxable income. As is well known, the opportunity to deduct DC taxes against one's federal individual income taxes softens the impact of the DC tax. Since federal marginal tax rates rise with income, it follows that federal deductibility is more valuable to higher income taxpayers. It also follows that federal deductibility makes the overall impact of the DC income tax less progressive.(32) Table 10 shows the pattern of DC effective tax rates for taxpayer who itemized in 1995 and for whom federal income tax information was matched. The marginal federal tax rate is approximated by the ratio of federal income taxes to federal adjusted gross income. Column 2 shows the number of taxpayers in each income class, and the subsequent columns show the first quartile effective tax rate without and with federal offset. If we examine the $35,000-$45,000 income interval with 9,593 taxpayers, we observe that the 1st quartile effective tax rate is 3.8%, but after federal offset it is 3.5%. Moving across we not that the median DC federal effective tax rate with federal offset is lower than not taking federal deductibility into account. (One can also note, by comparing Table 9 with Table 10, that itemizers in 1995 often had lower effective tax rates than all taxpayers.(33) 2.8 Comparisons of Tax Burdens with Maryland and Virginia Each year the Department of Tax and Finance publishes a comparison of tax rates and burdens across all major revenue sources with the major neighboring Maryland and Virginia Counties.(34) The methodology of the analysis is to construct four hypothetical couples at incomes of $25,000, $50,000, $75,000 and $100,000. The composition of income between husband and wife, the nature of their transportation methods, and housing choices are chosen to represent realistic demographic groups. Table 11 displays the analysis at 1996 tax law. The first portion of the table shows the estimated income tax burdens in dollars, and the second panel restates the DC tax burden as a percentage of the suburban tax burdens. For households with $25,0000 in income, the DC tax burden is 126% of the burden in Montgomery County and Prince George's counties, and 189% of the burden in various Virginia counties. The high relative tax burdens reappear for the $100,000 couples. According to the analysis, DC tax burdens are reasonably competitive with suburban Maryland for incomes in the $50,000 and $75,000 levels. 3 The General Economic and Demographic Situation 3.1 Census Data on Resident Population: 1969-94 As is well known, the resident population of the District has declined for a long period of time. Since 1969, the District's population has dropped from roughly 760,000 to about 580,000 in 26 years or a rate of decline of about -1% per year. Over the same period of time, Arlington has been stagnant in population, Prince George's county has grown modestly after a slight decline in the mid 1970's, and Montgomery has grown steadily from about 500,000 to 800,000. The Fairfax-Falls Church area has grown from about 480,000 to 900,000, and been the most spectacular in the DC metropolitan area (see Figure 1). If we compare 1994 population as a percentage of 1969 population we find:
These population changes — decline in central city population and increase in suburban ring population — are not unique: Baltimore City's population fell to about 75% of its 1969 level, while suburban Baltimore county grew to about 115%; Harford county grew about 170%, Anne Arundel grew about 150 %, and Howard county grew by 350 % of their 1969 population (see Figure 4). If we move up the East Cost to the Philadelphia metropolitan area, we observe the same general population pattern: the City of Philadelphia's population fell to 80% of its 1969 level while the affluent suburban counties grew. A relatively poor, neighboring county to Philadelphia, Delaware county, fell to 90% of its 1969 population. Bucks, Chester, and Montgomery counties grew substantially by contrast (see Figure 4). 3.2 Recent Evidence from Federal Tax Returns: 1990-5 A second way to examine demographic trends is to examine the number of federal tax returns filed annually in the District of Columbia, and those filed in neighboring states (Maryland and Virginia). These data, from the Statistics of Income Division of IRS annual publication Statistics of Income Bulletin, are tabulations of the IRS individual master file. The residence concept is the mailing address of the federal individual income taxpayer, and while the address may reflect the location of an accountant or tax attorney, over time the numbers are indicative of trends. Figure 5 indicates that, overall, Maryland and Virginia experienced small increases in the number of federal tax returns filed in 1990-5 (increases of under 5%) while DC's 1995 count of federal tax returns was about 86% of its 1990 level. 3.3 Evidence from Employment and Security Administration Tabulations of Wages by Place of Work and Place of Residence While resident population and federal tax returns show DC experiencing a consistent decline in residents and taxpayers, total wages paid in the District, or measured by the place of work of the employee receiving the wages, have been rising steadily. Figure 6 shows that from 1969 to 1994, total wages paid in the District grew from about $5 billion to about $27 billion. Of course, the observed population decline means that DC resident wages were far less than $27 billion. Figure 7 displays the ratio of BEA's estimate of resident earnings (wages and salaries and self-employment earnings, but not including capital income such as rent, royalties, and dividends and interest). DC's residents' share of earnings from employment fell from about 30% in 1969 to a little over 25% of total BEA earnings by place of work in 1994. While DC experienced growth in earnings by place of work, it did not do as well as any of its suburban neighbors. If we include capital income (which includes cash dividends, interest, rents, royalties received, as well as imputations for build up in pension plans and imputed rent on owner-occupied housing which became significant in the late 1980's and 1990's when depreciation was recalibrated in the GNP accounts) to the above earnings concept, we see that the economic position of the suburbs grew far faster on an absolute (see Figure 8) and relative basis (see Figure 11) than did their population. The Fairfax suburban area's 1994 resident wage and capital income was better than 1200% of the 1969 level, while the District's resident wage and capital income level was slightly over 400% of its 1969 level. Figure 10 shows the composition of the DC resident income base, using BEA concepts, and it is evident that capital income has become somewhat more important and net earnings somewhat less. Transfers (social security, afdc, medicare, medicaid, food stamps, public assistance etc.) are shown as a percentage of non-transfer income, and they have grown from 15% to about 28%. In one sense this suggests a combination of greater generosity in transfer payments as well as a greater tax burden on the non-poor and the federal government to finance these transfers. In terms of the starting point of calculating DC taxable income, the DC personal income tax base, as reflected on federal tax returns, has been stuck at the $10 to $10.5 billion level for the past seven years. Figure 11 shows the total amount of AGI in Virginia and Maryland returns compared to DC with 1990 set at 100%. While DC's total AGI rose less than 5% in the 1990-95 period, Maryland's total AGI grew by almost 20%, and Virginia's AGI grew by 25%. 3.4 Employment and Wages of DC Residents and Commuters: 1980-1990 In 1990, 724,412 individuals were employed in the District, or 11.8% more than the 650,137 employed in 1980, and 43% more than 504,611 employed in 1970. As noted above, this employment growth contrasts markedly with the resident population decline of the District. In 1980, the federal government was the single largest employer of District residents; however, in 1990, the Services industry was the single largest industry employing District residents. During that decade, federal employment of DC residents dropped from 70,775 to 48,342 or 32% (see Table 12). Federal employment actually grew by 5.6% of Montgomery County residents and fell slightly of Prince George's County residents. Not only did Services industry employment grow strongly for DC residents (20% across the decade), it grew by 49% for Prince George's County residents, and 59% for Montgomery County residents. District residents have generally held lower paying federal jobs than their commuting counterparts; this was also true in the Services industry in both 1980 and 1990. 3.5 Distribution of Federal AGI over Time: 1989-95 While DC's aggregate federal AGI has been stagnant, which implies stagnant DC individual income tax revenues to the extent the two bases move together, an examination of the composition of DC's federal income tax base shows that the number of returns in the $0-$50,000 AGI classes have been declining substantially, while those in the $50,000 or above AGI brackets have been growing (see Figure 12). The amount of AGI attributable to the two lowest income classes has also fallen; see Figure 13. If we compare the number of returns in the lower AGI groupings in DC to those of Maryland and Virginia, we see that the number of DC's federal returns fell more dramatically than those of Maryland and Virginia. Also, DC's returns in the $50-75,000 bracket grew more slowly than in Maryland or Virginia (see Figures 14, 15, and 16). 4 Migration in and out of the District of Columbia: Evidence from Federal Tax Returns As demographers are aware, the population level of an area is due to the net of natural increase, immigration, and out migration. Each year the IRS Statistics of Income Division analyzes the location of federal taxpayers in terms of their prior year's mailing address, and compares the number of returns to the prior year Also, this data permits the identification of the origin and destination, in terms of county, of such movers.(35) Figure 17 shows that by the middle of the 1980's the number of federal taxpayers moving out of the District grew dramatically: from about 21,000 per year to about 28,000. At the same time, the number of immigrants of federal taxpayers into the District fell off from a high of 24,500 in 1980/1 to less than 21,000 in 1990/1 and slightly over 20,000 in 1995/6. The difference between these two lines is the net out migration. Immigrants generally are smaller filing units; since 1985/6 they have fallen to around 1.4 to 1.5 exemptions per return, while the number of exemptions per return of out migrants has risen to as high as 1.75 and closed at 1.7 in 1995. Singles and small couples move in, larger households move out. (see Figure 18). These large flows into and out of the District have personal income tax administration implications. With about 26,000 federal tax returns departing with 1.7 exemptions/return, there are about 44,200 annual outmigrants. If we add to this the number of immigrants of 20,000*1.4=28,000, there are about 72,200 out of 550,500 or a gross turnover rate of 13% which is well over the US average of 6% for fraction of the population changing counties each year. Undoubtedly this imposes a significant administrative challenge to the District. Figure 19 displays the net of total exemptions on returns migrating in and migrating out. Since 1985/6 the flow has been a net out migration and has been stable in the 1990's at about 15,000 per year. On a five year basis that means that 75,000 persons reflected on federal tax returns have left, on net, the District. Interestingly, if we examine the AGI per tax return of migrants, we see that out migrants have lower average AGI than immigrants. The different in 1995/6 was on the order of $5,000 (see Figure 20). 4.2 County Origins and Destinations of Migrants: Evidence from Federal Tax Returns The same federal tax return information from IRS allows more detailed geographic identification of where immigrants come from, and where DC out migrants go to. Focusing just on the number of returns, we see that Prince George's County has historically been the largest source of movers into the District, on the order of 4,000 to 3,500 annually in the past 5 years, while Montgomery County lost about 2,000 movers annually into the District each year. Movement from suburban Virginia into the District was on the order of 1,000 per year (see Figures 21 and 22). For the last 8 years, DC has lost annually about 7,000 federal tax filers to Prince George's County, and 4,000 to 3,000 to Montgomery County. Migration from DC to suburban Virginia was no more than 2,000 per year to Arlington (which has been rising), and about 1,600 per year to Fairfax. This implies that the substantial population growth in Fairfax is not due to DC out migration but to other factors such as the autonomous growth in the Fairfax regional economy (see Figure 24). 5 Movers and Stayers: Evidence from DC Tax Returns Another way to examine the issue of the stability of the DC economy and its tax base is to examine the same taxpayers over time. As a consequence of having access to two years (1989 and 1995) of DC tax return data, it is possible to focus more completely on the nature of DC in and out migrations, and in particular the particular areas which seem to have experienced the greatest turnover over in DC taxpayers. 5.1 Characteristics from DC Tax Returns In 1989 about 310,000 DC tax returns were filed, and about 251,000 DC tax returns were filed in 1995. Of the 251,000 DC tax returns filed in 1995, 105,337 also filed DC returns in 1989. Thus, about 42% of 1995's taxpayers were "stayers'', or about 34% of 1989's taxpayers were "stayers''. Half of these stayers were single persons, and the remainder were head of household (22%) or married (26.5%). Of the 310,000 DC filers in 1989, 200,441 did not file DC 1995 returns or 67%. These " movers'' were 59% single taxpayers, 21.4% head of household, and 19.4% were married filing jointly. Finally, 145,607 DC filers in 1995 did not file returns in 1989. These are the migrants into the District, composed of: 56% singles, 22.4% Head of Household, and 26% married (see Table 13). If one examines the DC net taxable incomes of stayers vs. movers into DC in 1995, we find that the stayers tend to generally have higher income. The median net taxable income of singles present in 1989 and 1995 was $20,585 compared to $13,158 of those moving into the District.(36) This pattern is true across all filing types. Note also that the mean incomes are considerably greater than the medians, indicating that there are some very high income filers (see Table 14). 5.2 Moving and Staying within the District: Analysis at the Zip Code Level We may analyze movement within the District by examining tax filers who were present in 1989 and not present in 1995 (deemed to be outmigrants), and tax filers present in 1995 but not present in 1989 (deemed to be immigrants). If we take the ratio of those leaving by zip code to the total number of returns by zip code, and the ratio of those coming into the District by zip code compared to the number of returns by zip code in 1989, we can get some idea of the spatial variability in migration patterns. Some zip codes show dramatically higher rates of turnover than others, and rather different income characteristics. For example, Zip Code 20014 experienced a 94.7% out migration between 1989 and 1995, but only a 5.3% immigration. Similarly, Zip Code 20023 experienced a 95.8% out migration and only a 4.2% immigration. Other areas experienced comparable fractions of taxpayers leaving an area and moving into an area; Zip Code 20042 experienced by a 25% out migration ratio and an identical immigration ratio (see Table 15). 5.3 Changes in the Level of Crime and Moving vs. Staying in the District A question arises about what might be the cause of the extreme variability observed in Table 15 above. The study of household location decisions typically focuses on housing demand, educational services, proximity to work and shopping, and crime, and the price of foregone private consumption imposed by taxation. Recently, Nechyba and Strauss (1997) showed with New Jersey data that location decisions, holding constant housing demand, and proximity to work and shopping were quite sensitive to educational service quality as well as the level of crime in neighboring municipalities around Trenton, New Jersey. With regard to educational services, a one percent increase in educational services was associated with a 1.65 to 3.1 percent increase in the probability of moving to that jurisdiction. The elasticity of locating in a municipality given differential violent crime rates varied from -.1 to -.4; a one percent increase in violent crime is associated with a .1 to .4 percent reduction in the probability of moving to that jurisdiction.(37) Cullen and Levitt(1996) found with a sample of 80 cities and 1980 Census data that each additional reported crime is associated with a one person decline in city residents. They find that "...almost all of the crime-related population decline is attributable to increased out-migration rather than a decrease in new arrivals to a city.''(38) To examine if differential crime in the District is associated with differential out migration and immigration over time, annual crime data were obtained from the DC Police Department. That data, which are different types of crimes reported to the Police Department were aggregated from Census tracts to zip codes, and then matched to the number of tax returns by zip codes.(39) Since population changes are of particular interest, the statistical model estimated was of the following form:
For the zip codes for which crime in 1989 and 1995 could be matched to zip codes with tax returns in 1989, and 1995, the average decline in number of taxpayers per zip code was 3,256. Since the number of returns filed in 1989 generally exceeded the number of returns filed in 1995, we expect that as the number of crimes increased in 1995 over 1989, in a given zip code, that should be negative, e.g., more crimes in an area compared to 1989 will "push'' out taxpayers in 1995 compared to 1989 if out migration is systematically related to crime. Table 16 shows the statistical analysis results of different crimes. Column 2 shows the mean increase in crimes reported to the DC police, 1995 compared to 1989, by zip code. Overall, there were 275.9 more crimes per zip code reported in 1995 than in 1989; there were 5.7 more rapes per zip code reported and so forth. Column 3 displays the estimated effect of one more reported crime on the number of taxpayers in a zip code, 1995 compared to 1989. Thus, one more overall crime reported is associated with -2.7 fewer taxpayers in an average zip code in 1995 vs. 1989. Column 4 shows the probability that the associated effect is due to randomness; a value of .05 or smaller is extremely reliable. Finally, Column 5 shows the percentage of variation in the decline in taxpayers explained by the crime variable in question. While there were not that many more (5.7) rapes reported in an average zip code in 1995 compared to 1989, the drop off in taxpayers associated with one more rape is very large: 256.9. Note also that 79% of the variation in the decline in number of taxpayers is explained by the increased number of rapes. A second way to examine the effect of crime on the number of taxpayers is to take advantage of the fact that we know in 1995, compared to 1989, who left the District. That is, we examine all 1989 taxpayers who were no longer present in DC in 1995. Equation 2 states the relationship between the growth in crime by zip code on the number of taxpayers per zip code absent in 1995. That is, we estimate the effect of the growth in crime per zip code on the number of taxpayers who actually left the District in an average zip code. Table 17 displays the statistical analysis results. Because we are now predicting that more crime causes more taxpayers to leave, we expect that the effect of crime will in this model be positive. We find, again, that increases in crime operate in the expected direction, and in highly statistically significant ways. Overall, an increase in one crime per zip code over the period 1989-95 is associated with 5.3 taxpayers no longer in the District. This is about one departure per year or remarkably close to the results of Cullen and Levitt (1996). Again, one more rape is associated with a very large number of taxpayer departures: 416.3, although the explanatory power is now lower than in the previous model (compare 79% above with 53% here).
6 Policy Alternatives for the DC Individual Income Tax The sluggish individual income tax revenues noted at the outset, and the District's financial crises warrant a critical review of policy alternatives. However, "reform'' of the District's personal income tax rests on an definition of the "problem'' which may not be universally accepted. Moving towards any of three of the goals of a good tax system outlined above (revenue adequacy, simplicity and administratability, and distributional fairness) can readily imply different policies for the personal income tax, and, depending on one's behavioral assumptions, again different policies. Moreover, without an agreed-upon revenue target for the District's overall own-source taxes, and a well defined role for the personal income tax, reform can take on an overall tax cut or tax increase flavor. We shall proceed to discuss how one might change the DC personal income tax to achieve, separately, each of the three good tax system objectives. 6.2 Enhancing the Revenue Productivity of the DC Individual Income Tax 6.2.1 Supply Side Approaches Were the District to engage in significant tax reduction, as has been suggested by a variety of policy makers, as a strategy to make the District a more attractive place to live in comparison to surrounding areas, it is unlikely that greater economic growth in resident taxable incomes would occur quickly. The earlier analysis of BEA estimates of earnings by place of work showed that the District wages and self-employment income have been growing, albeit at a slower pace than in most suburban areas. Conversion of some of those commuter earnings to resident earnings through tax cuts that would cause relocation decisions would take time as commuters would need to be convinced that the tax reductions were likely to remain in place. In the meanwhile, questions would arise about personal income tax revenue reductions which would occur for DC residents who would also benefit from low tax rates. Unless unusual improvements in service productivity were to be obtained, the overall discipline of the budget would require that services be further curtailed. Whether this would be acceptable to current residents as well as prospective residents and create an environment conduce to converting commuters into residents is difficult to speculate about. It is likely, however, to be a difficult matter to sell. Should educational services deteriorate and police protection deteriorate as well, through higher crime rates, it is likely that the tax reduction would not be offset through greater migration into the city. Indeed, it might be that during the initial adjustment period, matters could compound to the disadvantage of the District. 6.2.2 DC and Federal Tax Filers: A Puzzle and Revenue Opportunity When the District obtains tax return information from the IRS under its information exchange agreement, it receives all federal tax returns filed with a DC mailing address. Analysis above with data files with DC tax return information indicates that about 50,000 DC tax filers did not file federal tax returns (with DC mailing addresses). A question arises whether or not there are federal tax filers with DC mailing addresses who did not file DC tax returns. Table 18 displays the four logical possibilities of federal and DC filing; the entries are the result of merging the 1995 DC and federal files by SSN. Also noted are whether or not these returns were present in DC in 1989; those which were not present in 1989 are deemed "movers'' and those who filed in DC are deemed "stayers.'' The logical possibilities for DC and federal filing are: Case (1): DC and Federal tax return both present when merged by SSN Case (2): DC tax return present but federal tax return not present Case (3): DC tax return NOT present but federal tax return present Case (4): DC tax return NOT present, federal tax return NOT present: unobservable Case 1 and Case 2 show $556.0 million in 1995 tax liability on their returns. Case 3 is extremely interesting. There were 51,404 federal tax returns in 1995 with DC addresses for which there is no DC tax return. In addition, another 14,034 very late federal filers (they filed 1995 federal returns late in 1996), fall into this category. While DC tax information is not available for the Case 3 federal taxpayers, and the very late DC taxpayers, there is enough federal tax information to construct a simple DC tax calculator based on federal filing status. Application of the calculator shows:
Thus, if all of the 65,438 Case 3 taxpayers are indeed subject to DC tax, individual income tax liabilities for 1995 would be $134 million or 24% higher! Whether or not these taxpayers are legally subject to DC income tax is beyond the scope of this study. As noted above in Section 2.2, there are a variety of circumstances in which those with DC mailing addresses to be legally exempt from the DC personal income tax. These federal taxpayers may elect to have their federal tax returns mailed to their place of work (but they may be residents of Maryland or Virginia), or mailed directly to their tax advisors. Some of these taxpayers may be Congressional or embassy employees who are exempt. On the other hand, 65,438 is a particularly large number, and the computed potential liability of $134 million to be the outer bound of what might be collected under audit and compliance work. Presumably systematic investigation of these tax returns for DC collection activity is worthy of consideration. 6.3
Achieving Greater Conformity with the Federal Individual Income Tax: In October, 1972, the Congress enacted optional federal collection of state individual income taxes.(40) Upon prospectively entering into an agreement with the Secretary of the Treasury, a "qualified'' state would have its income taxes collected by the federal government in such a manner as if the taxes were imposed by the federal government. It was hoped by its supporters that such a voluntary system would substantially simplify the efforts of taxpayers who were required under federal, state, and sometimes local law to maintain separate records because of the differences in information required by different income tax provisions. As a corollary, employers in electing piggyback states would have to make only one deposit of federal and state withheld taxes. It was expected that taxpayers in electing states would have to fill out only one federal form 1040 on which a few lines would reconcile state withholding and final state income tax liability. Electing states would further benefit through prompter turnover of withholding, and the application of the IRS for primary audit and compliance efforts. Two types of piggybacking on resident incomes were available to an electing state under the 1972 legislation:
An electing state would thus decide which form of resident income tax it wished the federal government to collect, and notify the Secretary of the Treasury what the rate of tax would be. The federal piggybacking provisions also provided for a tax on wage and business income derived within a piggybacking state by a non-resident individual. Both forms of the resident income tax would require some modifications in the federal base to eliminate state taxation of federal bond interest, and permit the state through policy to exempt its own and other states' interest on bonded indebtedness. Also, the federal provisions permitted a state to allow a credit, which the IRS would administer, against its own income tax for income taxes paid to another state or political subdivision of another state. Thus, the legislation contemplated that the federal government would make adjustments for qualifying states for commuter taxes. An important feature of the piggybacking legislation was the replacement of the state by the IRS in the protection of state interests on most matters involving tax disputes. The federal government would thus deal with taxpayers and appear in court on behalf of any state in a dispute over income taxes that were collected by the IRS. While the IRS would represent the state interest in civil and criminal matters, the legislation provided that the State would represent itself in state court involving state constitutional issues and with respect to proceedings involving the relationship between the state and the federal government. The 1972 legislation required five states to trigger the piggyback system nationally; the Tax Reform Act of 1976 eliminated this requirement, and clarified that the federal government could not charge any state for collecting its income tax. Despite some obvious attractions, no state ever triggered the system, and in 1990, the provisions were eliminated from the Internal Revenue Code as part of an omnibus budget package which eliminated so-called "deadwood'' provisions of the IRC. In January, 1997 President Clinton proposed in his State of the Union message, as part of his proposed replacement of the $660 million annual "federal payment'' to DC with federally provided services, to have the Internal Revenue Service take over responsibility for collecting and administering DC's individual income and payroll taxes. Initially, the Administration proposal provided for $15 million in FY1998 start up costs, and $25 million/year in annual operating expenses for FY 1999 and thereafter. Since the piggybacking provisions in the Internal Revenue Code no longer were operative, actual implementation of the proposal would have required federal tax legislation. Subsequently, the proposal to have the DC income tax piggybacked as contemplated in the 1972 legislation was dropped. Instead, the working relationship between the IRS and District, which operates under the federal-state exchange agreement, has been strengthened. Also, the federal budget passed in the Summer of 1997 enabled the Chief Financial Officer of the District to enter into contracts with a private entity for the administration and collection of personal income tax. While actual federal collection no longer is being actively discussed, moving the District's income tax to greater conformity with the federal income tax has been recommended for a considerable period of time. The DC tax already uses much of the Internal Revenue Code. The starting point of any taxpayer's calculations is federal adjusted gross income as reported on his federal return,(42) and many of the modifications rely on federal concepts. We explore here several different adjustments to the DC income tax:
This past summer, Carol O'Cleireacain (43) suggested replacing the District's individual income tax with a 28% surcharge on DC liability along the lines of North Dakota, Rhode Island, and Vermont. Table \ref{pigrevenue} displays the results of simulating the proposed 28% piggyback tax as a surcharge on federal liability, and a revenue neutral surcharge proposal. Several points about these simulations are important to keep in mind. First, they are based on matched 1995 DC and federal returns whose total DC 1995 liability was $446 million, well below total liability of $556 million. As long as those not in this portion of the universe of DC returns are no different than the others whose federal returns were not available, using $446 million as a benchmark is accurate. Second, because actual tax return data are available, the total revenue effects of any scenario are inherently more accurate than those estimated from published Statistics of Income Tables because they necessarily aggregate across filing units and different marginal tax rates.With these caveats in mind, note that the 28% surcharge on 1995 federal liability would yield only $339 million from the same taxpayers or a 23.9% tax reduction. O'Cleireacain (1997) estimates with 1994 data that overall the 28% proposal would entail a 30% tax reduction, and that all taxpayers would receive a tax cut.(44) Examination of the distributional effects of these two proposals with micro-data indicates that the 28% and 36.7% surcharges would lower the vast majority of DC taxpayer's taxes compared to what they actually paid in 1995, high income taxpayers would experience significant increases in their effective tax rates (and therefore actual tax payments since the comparison is being made to current law). Under the 28% surcharge, the 587 taxpayers for whom federal returns were available for surcharge analysis, the median tax rate in 1995 (see Col (6) of Table 20) was 8.0%. This is below the statutory rate of 9.5% because the value of exemptions and deductions reduce the liability below this theoretical maximum. Imposition instead of a 28% surcharge on high income taxpayers would create a higher median effective tax rate of 8.7%, or a 7.9% increase in effective tax rate (and actual tax dollars). Taxpayers whose effective tax rate was in the 75th percentile (see Cols (9)-(11) of Table 20) would experience an increase in effective tax rate from 8.7% to 9.3%. Table 21 displays the distributional effects of the 36.7% surcharge on federal tax liabilities which is revenue neutral across those taxpayers for whom both 1995 DC and federal tax returns were available. It is evident that the taxpayers in the highest income class will experience very substantial tax increases as a result of this form of piggybacking. Note that the highest income taxpayers, whose effective tax rate was in the 75th percentile, would experience an effective tax rate of 12%, or well above the current-law 9.5% top marginal tax rate. (see Cols (9)-(11) of Table 20). 6.4 Administrative Simplification of the District's Individual Income Tax A second approach to tying the District's personal income tax more closely to the federal personal income tax is to eliminate the various subtractions accorded under DC law (which amount to tax base reductions of $1 billion in 1995), and require DC taxpayers to file on the same basis as they do for federal returns. This would eliminate the Married Combined filing separate class, and obligate them to file as Married Joint returns. This would impose a marriage penalty as their current splitting arrangement is designed to overcome this. Revenues would predictably rise. The simulation model indicates this proposal would raise $658.4 million of tax before credits (and $640.8 million in tax after credits) compared to the base case of $556 million from the same taxpayers at 1995 levels. The second simplification proposal of using federal instead of DC standard deduction and exemption amounts, but continuing DC brackets and tax rates, would create $572.9 million in revenues compared to the base case of $556 million. Table 22 displays the effective tax rates by federal AGI class of the first simplification proposal:
Several things are immediately evident from this distributional analysis. First, the elimination of itemized deductions tightens up the distribution of effective tax rates among higher income households. The first quartile of effective tax rates is no longer 3.3% (see Col 3 of Table 22), but 7.9%. Second, the median and third quartile effective tax rates for the highest income households moves noticeably towards 9.5%, the theoretical maximum effective tax rate. If the additional revenue which this proposal is too large, given the revenue goals of the District, then the top marginal tax rate of 9.5% could be lowered and/or the brackets widened to account for the 8% inflation since their inception. Table 23 displays the second simplification proposal which:
This simplification proposal is analogous to basing the DC personal income tax on taxable federal income with no itemization. It would achieve tax entry points of the federal income tax (see Table 7, Col. (3)). As with the above simplification proposal, the distribution of effective tax rates "tightens'' up considerably, and the lower and moderate family incomes are more favorably treated than current law. The effective tax rates in federal AGI classes up to $45,000 in Table 23 are lower than the effective tax rates under current law. Yet, overall, tax revenues are roughly comparable ($572 million) to the base case of $556 million. 7 Dimensions of Taxing DC Commuters The financial and political positions of capital cities are often precarious and controversial. When Margaret Thatcher became the Tory Prime Minister of Great Britain, she disbanded the Greater London Council which had been dominated by the Labor Party for many years, and distributed its activities to 33 consolidated boroughs throughout Metropolitan London. To make her point lasting, she sold the historic London County Hall to a Japanese investor who turned it into a luxury hotel. Most recently, Tony Blair, recently elected Labor Prime Minister, has created a London City Council with, for the first time, an elected Mayor of London. The Canadian solution to financing their national seat of government has been to create a two-tier system of local government, and to empower the first tier of government to finance its provision of municipal services through a local property tax. The local tier includes narrowly defined Ottawa, and the second tier includes neighboring suburbs, all within Ontario. Canadian law obligates the national government to make payments to the City of Ottawa (tier one) in lieu of property taxes, although the measure of the payment is the application of tier one millages to the negotiated value of all federal properties. Conflict over municipal finances is thus dealt with during negotiations over the market value of federal property. Historically, the District of Columbia has had one of the highest ratio of non-resident workers to residents of any city in the US. As of 1990, As of 1990, 493,716 nonresidents worked in the District; DC residents who worked numbered 304,426 (including 67,694 who work outside the District).(47) There are several different philosophic perspectives one might adopt about how a nation's capital should finance its services; however, it seems likely that most would agree that commuters who use municipal services at their place of work should, to the extent practical, pay for these services. Moreover, is seems likely that most would agree that the place of resident should recognize these tax payments through tax credits against resident income taxes. As already discussed, the District, Maryland and Virginia's income taxes already recognize for their residents the wisdom of this. Also, some would argue (as the Virginia Supreme Court has) that the District already imposes a limited commuter tax on certain (but not all) non-resident unincorporated business income. Were the Home Rule Charter to be amended, within the more general context of rationalizing the District's finances, to provide for a commuter tax on non-resident wages, what sort of dimensions might reason lead it to have? Logic might suggest, for example, that the purpose of the non-resident wage tax be to defray the cost of municipal (as contrasted with educational or income redistribution) services used by non-residents. Income redistribution among the states and the District is already accomplished by the federal government through the federal income tax, and payments to the states and District based on a redistributive formula that compares the ability to pay of each area to the national average. Recall that District residents receive AFDC, Medicaid benefits, and the District, like each of the states, is reimbursed for these income maintenance programs based on the federal reimbursement formula.(48) Logic might also suggest that whatever tax on commuter wages and self-employment income be levied, it should be limited in rate to some fraction of the resident rate in recognition that commuters spend only a portion of their day at their place of work. Simplicity argues for the fraction to be 8/24 or 1/3, and simplicity further argues for a flat rate commuter tax. Given that the District's income tax rate varies from 6 to 9.5%, we have an upper bound, therefore, of between 2.0% and 3.2%. These two points suggest a variety of ways to devise a commuter tax from first principles. One might go through the District's budget and identify those services which are municipal, and determine the share to be borne by commuters. Let us suppose that $750 million are the total services which are enjoyed by residents and non-resident commuters.(49) The $750 million rough figure would include indirect costs associated with these activities. Given that commuters were 62% of those working and enjoying DC municipal services, the total gross cost to be financed by commuters would be $465 million. The Bureau of Economic Analysis reported that DC non-resident earnings were $21.9 billion in 1994; O'Cleireacain (1997), p 108. estimates non-resident wages to be $17.3 billion. These figures suggest a range of commuter tax rates of 2.1% to 2.7% ($465 million ÷ $21.9 billion and $465 million ÷ $17.3 billion); multiplying these rates in recognition of the 8 hour work day then creates commuter tax rates of between .6% to .9% (2.1% ÷ 3 and 2.7% ÷ 3) and revenues of $465 ÷ 3 = $155 million or roughly a quarter of 1995 resident income tax liabilities. Undoubtedly one can work with these assumptions and figures to reach different rates of commuter taxes; however, the above arithmetic indicates that a commuter tax between 1/2 to 1% on commuter wages and self-employment income can be derived in a logical way. How this relates to the existing tax on unincorporated business income and the exemption accorded to various professional services is an interesting series of questions, and worthy of further research. The purpose of this paper has been to review the District's individual income tax in terms of structure and empirical characteristics in relation to observed population shifts. Also, the paper has sought to address how the District's income tax might be more strongly related to the federal personal income tax to aid tax administration and compliance. The research has led to the following conclusions:
Were DC to have retained those who left between 1989 and 1995, and they paid average per return taxes for their household income class, 1995 DC individual income tax liabilities would have been $39 million higher.
1. This tension is not unique to the District. Many states with large central cities constantly must deal with political fights in their state capitols over how much assistance to provide for the dominant municipality and its school district vis a vis neighboring suburbs and rural areas. In New York the friction is between New York City, its suburbs, and UpState New York. In Pennsylvania, it is between Philadephia, neighboring suburbs, and the rest of the State. In Michigan, it is between Detroit, its suburbs and the rest of the state, and so forth. 2. US Bureau of the Census, Journey to Work. 3. Hellerstein and Hellerstein(1992), II, p. 20-2, 20-3. 4. Mississippi, Oklahoma, Massachusetts,Virginia, Delaware, Missouri, New York and North Dakota. 5. North Carolina, South Carolina, New Hampshire, Arkansas, Georgia, and Oregon 6. Idaho, Tennessee(on capital income), Utah, Vermont, Alabama, Arizona, Kansas, Minnesota, Montana, New Mexico, Iowa, Louisiana, California, Kentucky, Colorado, and Maryland 7. West Virginia, Indiana, Michigan, Nebraska, Connecticut(just capital income until 1991 when broadened to labor income), Illinois, Maine, Ohio, Pennsylvania, Rhode Island and New Jersey 9. See US Census Bureau(1997). 10. US Bureau of the Census(1996), Table 472; US Bureau of the Census (1997). 12. All figures refer to 1993/4 and are from US Census (1997). 13. 1975 figures due to Sunley and Wilensky (1978). 14. Author's calculations with DC tax data files. 15. Alabama, Arkansas, Mississippi, New Jersey, and Pennsylvania 16. Source: Federation of Tax Administrators, FTP site, January, 1997. 17. Sunley and Wilensky (1978). 18. CCH, District of Columbia State Tax Reporter, ¶ 425. 19. In the 1990's, 16% of the US population changes its house each year; 11% of the US population move within the same county, 6% of the US population move to a different county but remain in the same state, and 3% move to a different state. Table 33, 1996 Statistical Abstract of the US. 20. Hellerstein and Hellerstein(1992), II, p. 20-03. 21. Other characteristics of non-resident income which qualify as exempt service income are: 1) any trade or business which by law, customs or ethnics cannot be incorporated, 2) any trade, business or profession which can be incorporated only under the DC Professional Corporation Act of 1971, a trade or business engaged in by a blind person. 22. Hellerstein and Hellerstein(1992), 20-73. 23. Virginia Supreme Court, No. 961290, April 18, 1997, and U. S. Supreme Court, Docket No. 97-412, November 10, 1997. 24. The US Supreme Court chose not to hear the case. 25. See Table 3, column 10, Total row. 26. Federal tax filers or the District were defined to be any taxpayer with a DC mailing address. Federal tax information was obtained by the District Department of Tax and Revenue through its Exchange Agreement with the IRS, and provided to the author under an IRS approved confidentiality agreement. 27. Deductions for IRA's, uncompensated moving expenses, self-employed health insurance deduction, contributions to Keoh and self-employed pension plans, and deduction for alimony paid. 28. The aggregate figure is somewhat different than the sum of the individual figures since negative incomes across taxpayers are not allowed to reduce the sum of DC agi. 29. Given that the IRT file is a transaction file rather than a file of filed tax returns, the correspondence suggests good reporting by DC taxpayers. 30. See CCH District of Columbia State Tax Reporter, ¶ 16-765, p. 1692. 31. Given a 2,000 hour year, this would mean working at $6.13/hour or slightly above the minimum wage. 32. If t is the DC effective tax rate, and tfed the effective federal income tax rate, then deductibility means that the DC effective tax rate is t*(1-tfed). Since tfed rises with income, the DC offset tax rate will be lower than otherwise. Were the DC rate fixed at one tax rate, the offset effective tax rate would actually decline with income and thus be regressive overall. 33. Note that the effective tax rates in Table 9 include itemizers, so the comparison is not completely distinct. 34. See A Comparison of Tax Rates and Burdens in the Washington Metropolitan Area. 35. The author wishes to thank the Division for providing this unpublished data to this project. 36. Examination of movers who were only part-year residents does not alter this conclusion. 37. Nechyba and Strauss (1997), Table 6. 39. The relationship between 1980 Census tracts and zip codes was obtained from the University of Missouri MABLE ftp site which maintains such data for the entire US. 40. See Joint Committee on Taxation(1973). 41. As noted above, three states' income taxes are a percentage of federal liability: Rhode Island (27.5%) and Vermont (25%) income taxes are a flat percentage of federal liability, and in North Dakota taxpayers may elect to pay 14% of adjusted federal income tax liability. 42. While the DC tax law references the taxpayer's federal return the taxpayer is filing for DC tax purposes, it never defines the federal return to be the actual signed return filed by the taxpayer in compliance with the filing requirements of the Internal Revenue Code, nor does it state that the tax year should be the same. The DC taxpayer is obligated on the form (and in DC tax statutes) to utilize the same form of deduction, either standard or itemized, as used for federal tax purposes, and in the case of married combined filing separately and married filing separately, both DC taxpayers must use the same form of deduction. While there may be a presumption that DC taxpayers report from their bona fide federal tax return of the same year, it is not transparently stated. Moreover, if one reviews the DC tax return, DC form D-40, it is evident that it does not reference specific line numbers of the (bona fide) federal return from which the taxpayer is to transfer the information. Whether or not taxpayers faithfully report from their bona fide federal return can not be determined from current DC tax administration databases, because only the first page of the DC form D-40 is put into machine readable form. Thus, supporting schedules are not available to cross-check with federal income tax data sources. 43. O'Cleireacain (1997), p. 101 44. O'Cleireacain (1997), p. 101 45. In the case of adding back income taxes, use of this rather than the simulated proposal for add back is inaccurate; however, the data available does not break out this particular figure. 46. In the case of adding back income taxes, use of this rather than the simulated proposal for add back is inaccurate; however, the data available does not break out this particular figure. 47. O'Cleireacain (1997), p. 106. 48. I thus do not find persuasive the argument sometimes made that suburban residents have a responsibility to finance poverty programs in a central city they are next to beyond the one each state legislature finances through state taxation of all state residents. Otherwise, one would create additional local incentives for local forum shopping. In my view, income redistribution should be financed by a government whose geographic reach is sufficiently large that most will not move to avoid taxes to finance income-redistribution. 49. Table 500 of The Statistical Abstract of the US, 1996 reports that the District made direct expenditures of $496 million for Highways, Fire Protection, and Police protection in 1993. Advisory Commission on Intergovernmental Relations (ACIR) (1993). Significant Features of Fiscal Federalism. Volume 1. Budget Processes and Tax Systems. (US GPO, February, 1993. Cullen, Julie B. and Steven D. Levitt, "Crime, Urban Flight, and the Consequences for Cities,'' NBER Working Paper 5737, September, 1996. D.C. Department of Tax and Finance (1996). A Comparison of Tax Rates and Burdens in the Washington Metropolitan Area. (June, 1996). Gold, Stephen D. (1995), "The Income Elasticity of State Tax Systems: New Evidence,'' State Tax Notes, 8, 18 (May 1, 1995), 1849-1856. Goodman, Leonard (1995), "Conforming Federal and State Individual Income Taxation,''State Tax Notes, 8,25, (June 19, 1995), 2471-2482. Hellerstein, Jerome and Walter Hellerstein (1992). State Taxation. II. Sales and Use, Personal Income and Death and Gift Taxes. (Warren Gorham Lamont). Joint Committee on Internal Revenue Taxation (1973). General Explanation of the State and Local Fiscal Assistance Act and the Federal-State Tax Collection Act of 1972. (US Government Printing Office, February 12, 1973, JCS-73). Kies, Kenneth J. (1996), "Written Testimony of the Staff of the Joint Committee on Taxation Regarding H.R. 3244, the District of Columbia Economic Recovery Act,'' (July 31, 1996) JCX-45-96. Nechyba, Thomas and Robert P. Strauss (1997)"Community Choice and Local Public Services: A Discrete Choice Approach,'' with Thomas Nechyba, (forthcoming), Regional Science and Urban Economics. National Bureau of Economic Research Working Paper 5966. Noto, Nonna A. (1997), "District of Columbia Reorganization: Clinton Administration Proposal for the IRS to Collect DC Income Taxes,'' Congressional Research Service Report, 97-343 E. (March 20, 1997). O'Cleireacain, Carol (1997 ). The Orphaned Capital: Adopting the Right Revenues for the District of Columbia. The Brookings Institution. Schenk, Deborah H.(1989), "Simplification for Individual Taxpayers: Problems and Proposals,'' Tax Law Review 45, 1 (Fall, 1989), 121-175. Emil Sunley and Gail Wilensky (1978). The Personal Income Tax. in Technical Aspects of the District's Tax System. House of Representatives, 95th Congress, 2nd Session. Studies and Papers Prepared for the District of Columbia Tax Revision Commission. (December 1, 1978). Serial No. S-11. US Bureau of the Census (1996). Statistical Abstract of the United States for 1996. (US GPO and CD-ROM), October, 1996. US Bureau of the Census, Governments Division (1997). FTP Site. State and Local Government Finances. http://www.census.gov.com/ftp/pub/govs/estimate/94stlus.txt |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Send mail with questions or comments to webmaster@dcwatch.com
Web site copyright ©DCWatch (ISSN 1546-4296)
