
Last December, a panel of five economists, headed by Michael Boskin, Chairman of the Council of Economic Advisers (CEA) during the Bush Administration, released its report on the Consumer Price Index (CPI). The Boskin Commission report, Toward a More Accurate Measure of the Cost of Living, analyzes technical issues regarding the CPI and makes recommendations intended to lead to a more accurate measure of changes in the cost of living. This report also calls for legislative action to adjust indexing provisions.
The Commission found that the current CPI may overstate annual change in the cost of living from 0.8 to 1.6 percentage points. The Commission also concluded that the most plausible point estimate of this overstatement is 1.1 percentage points per year. Although there is considerable agreement among economists that the CPI probably overstates price inflation to some degree, there is great uncertainty over the extent of this overstatement.
The Commission's report has proved controversial because a variety of Federal entitlement programs, including Social Security and military retirement, are indexed using the CPI. This paper will focus on how a reduction in annual CPI adjustments would affect the Federal income tax. A previous Joint Economic Committee (JEC) report[1] found that income tax increases, falling primarily on middle class taxpayers, would comprise about 40 percent of the direct budget effects of a CPI revision. This paper takes no position on the policy issues related to adjusting the CPI.
Under the provisions of the Economic Recovery Tax Act (ERTA) of 1981, certain features of the individual income tax were indexed to the CPI starting in 1985. These features include the personal exemption, standard deduction, and tax bracket boundaries. The effect of indexing is to expose a smaller proportion of income to taxation and to tax a portion of income at lower as opposed to higher tax rates. Conversely, a legislated cutback in annual tax indexing means that a higher proportion of personal income would be taxable, and some of it would be taxable at higher tax rates. Over time, the cumulative effects of curtailing tax indexing are very significant.
According to the available estimates, a 1.1 percentage point reduction in tax indexing would lead to a tax increase of about $322 billion over the next 12 years. Though the tax increases in the early years are not very large, the cumulative effects of de-indexing mount rapidly after the turn of the century. By 2008, the final year projected in the Boskin Commission report, the annual tax increase grows to about $56 billion. Thus, a reduction in tax indexing would lead to a major structural change in revenues in relation to other components of the budget. Figure 1 displays the amount of annual tax increases over the next 12 fiscal years.
Any attempt to calculate the effects of this proposal on individual taxpayers is very difficult because of the different tax situations of taxpayers. The number of personal exemptions, use or non-use of the standard deduction, and the proximity of taxable income to tax bracket thresholds are some of the variables involved. Nonetheless, the aggregate revenue numbers can be used conservatively to estimate the average tax increase per taxpayer resulting from reducing the CPI adjustment. By dividing the annual aggregate tax increase by the number of tax returns, the average impact per taxpayer can be approximated.
The projected number of individual tax returns for the next decade by tax year is available from the Internal Revenue Service. Use of tax filer data for this purpose is a conservative approach to determining the average tax increase per taxpayer because more than 15 percent of tax filers do not actually incur income tax liability. However, the erosion of tax indexing would force many low income filers currently without tax liability to become subject to the income tax.
The data show the significant effects caused by the erosion of tax indexing. By the year 2003, the average tax increase per taxpayer would total $208 annually. By the last year of the Boskin Commission projection, 2008, the average tax increase per taxpayer would amount to $405 annually. Over the entire 12-year period, the average tax increase would amount to about $2,424. Figure 2 displays the cumulative effects of this proposal.
Up until now, the conventional assumption has been that the tax increases and benefit savings from a CPI revision would be devoted to deficit reduction. However, this assumption is open to question as there is no assurance these resources could not be rededicated to spending increases in discretionary programs or certain entitlement programs. If history is any guide, the revenue from this tax increase will likely stimulate more spending, not deficit reduction. According to a 1991 JEC study, the Federal government has spent $1.59 for every dollar of tax increases during most of the post-World War period[2]. If this pattern were repeated with the tax increases resulting from a CPI revision, not only would the entire tax increase be expended, but the additional increase in Federal spending would erase much of the entitlement savings as well.
A legislated reduction in the CPI adjustment to the Federal income tax would result in large and growing annual tax increases within several years. By the end of the period reviewed by the Boskin Commission, these tax increases would average more than $400 per family each year. These tax increases would fall primarily on middle class taxpayers. Moreover, the conventional assumption that these tax increases would necessarily result in deficit reduction rather than additional spending cannot be substantiated.