Almost every professional analysis of the Alternative Minimum Tax (AMT) of which I am aware takes a highly critical view of this misguided tax. Extensive research has revealed that it has undesirable incentive effects, and often accounts for tax liabilities that are unreasonable and even at times indefensible. Until recently, the AMT was something of a footnote in the tax code. However, as my colleague on this panel Len Burman has demonstrated in a series of important papers with various coauthors, taxpayers are affected by the AMT more and more each year.[1]
 
Since I feel safe in presuming that Mr. Burman can more competently convey his research findings than I can, I will focus my remarks on parts of the academic literature that complement his work. Before I do that, however, let me stipulate that Mr. Burman’s work accurately and exhaustively characterizes the state of the AMT problem today. He has carefully documented the economic harm that the AMT causes. His estimates of the costs of various reforms in the updated version of his Journal of Economics Perspectives piece appear to be accurate, as does his estimate of the proportion of taxpayers who have become subject to the AMT by the EGTRRA. The question, I believe, is not whether this tax needs to be reformed, but rather, how best to do it. This is where my testimony will focus most of its attention. 

My first recommendation is that reliance on stopgap measures be abandoned. To date, the AMT problem has been delayed by stopgaps. One measure that has been used to soften the blow is a temporary increase in the AMT exclusion. The use of this stopgap is effective in the short run, but it is likely harmful to future policy as a whole. This is because current long run forecasts include ever more significant amounts of revenue from the AMT that will not be realized if successive temporary stopgaps are adopted. Elected officials will base their decisions on highly misleading revenue forecasts if they rely on them.

Future stopgap measures might continue to reduce actual AMT revenue, but there are no guarantees. The AMT is acquiring such a big footprint that these temporary measures will soon become quite costly, and politically more challenging. This will make uncertainty concerning AMT reform a key factor for tax planners. Mr. Burman has indicated in his work how serious the aggregate problem is becoming. A specific example will help illustrate how close we are to the AMT precipice under current law. For Figure 1 and Figure 2, I used data on state and local income taxes to estimate whether a married couple with an income of $150,000, three children, and normal state and local deductions would be on the AMT. Under current law, with the temporarily high $58,000 AMT exemption, taxpayers in only eight states face the AMT. Next year, when the exemption is reduced to $45,000, this changes dramatically. This family would face the AMT regardless of what state it lived in. Note, however, how uneven the liability is. Individuals in states that are the darkest will bear a much higher AMT burden than individuals in the lighter shaded states. A family in New York would pay the most, $4,058, while the same family in Tennessee would pay the least, $1,243.

This latter point is one key problem with the AMT that has been highlighted in the literature. While it was originally motivated as a tax to ensure social justice, it likely does the opposite. It taxes individuals across states in a hodgepodge way, hitting similar individuals quite differently. Mr. Burman has also demonstrated that it disproportionately harms families with children, and married couples. Such harm is an unintended consequence of bad tax design, having no conceivable philosophical justification.

What economic impact would the AMT tax hike likely have?

This uneven and unjust tax has significant economic consequences. The AMT causes economic harm for at least two reasons. First, it increases tax complexity and compliance costs for the many millions of taxpayers who must calculate their taxes according to two different sets of rules each year. Second, the AMT often increases marginal tax rates for those who fall under it. A recent study of these marginal tax rate increases found that they were very uneven, increasing marginal rates sharply for some taxpayers, and only slightly for others.[2]  The authors found that for 81 percent of taxpayers, wages face a higher marginal tax rate under the AMT, with 18.6 percent of taxpayers experiencing an increase of 10 percent or more. The increase occurs because the lowest AMT bracket rate is higher than the income tax rate of middle income Americans who are swept onto the AMT, especially in future years. On average across all taxpayers, the AMT is projected to increase average marginal income tax rates by 1.5 percentage points by 2010, though the increment to the marginal tax rate is lower in earlier years.[3]
 
These higher marginal tax rates likely will suppress economic activity, and do so unevenly, hitting some states much harder than others. Although the scale of this effect is the subject of intense academic debate, recent research suggests that marginal tax rates have large effects over longer time horizons.[4]
 
Reform options

The complexity, high marginal rates, and uneven reach of the AMT make repeal the best reform option. The key factor justifying this recommendation in my mind is the absence of unambiguous logical justification for the current view that it is “just” to punish taxpayers for taking advantage of tax preferences that are perfectly legal. For example, capital market equilibrium may equalize after- tax returns of assets with different tax attributes. This means that if the tax rate were 50 percent, one would expect to observe, to use bonds as an example, that tax free bonds would pay 2.5 percentage points interest if taxable bonds paid 5 percentage points. Now, one millionaire might purchase taxable bonds and pay healthy tax on that interest, while another might buy tax free bonds and not do so, but each of them derives the same welfare. If an unexpected penalty is imposed on the individual purchasing the tax free bonds because they failed to pay tax, then one in fact diminishes social justice by increasing after tax inequality. If the tax penalty is anticipated, it will simply change the interest rate on the “tax free” instrument in a manner that leaves both investors with the same after tax income. Other tax preferences are folded into our complex economy in a similar way. Thus, there are two possible outcomes. The penalty accomplishes a bad thing, or nothing at all. It is hard to imagine why a rational individual would want to keep it, or even amend it to make it “better”.

If the AMT were repealed, a natural question, given the current difficult fiscal situation, would be whether the repeal should be revenue neutral. 

The current forecasts assume that significant AMT revenue will be realized. Tax revenues rise as a share of GDP to a level significantly above the historical average (Figure 3) over the next 10 years. Outside of the budget window, the revenue gains associated with the AMT are even more impressive. If policymakers do not modify the AMT or any other provisions of current tax law, revenues relative to GDP will rise to virtually unprecedented levels. By 2050, tax receipts would be 24.7 percent of GDP, compared to 18.4 percent if they remained at their historical average.[5]
 
If we abstract from the problem with entitlements, the AMT could easily yield enough revenue to push the budget from deficit back into surplus, assuming that the revenues are not spent. Since the government should, in present value, try to align tax revenues and spending, this could be viewed as a positive step, at least until revenues start to skyrocket in the out years. It may be a stretch however, to assume that the new revenues from such an approach would be used for deficit reduction since higher revenues are sometimes found to lead to higher spending.[6]
 
If these tax receipts are spent, it would have a significant and negative impact on growth forecasts. Robert Barro (1991)[7] and others following his influential work have used cross-country regressions to measure the effect of public sector size on economic growth. Barro (1991) studied the effect of government spending (excluding spending on education and national defense, which can be viewed as investment) on economic growth for a sample of 98 countries, and found that every percentage point of government expenditure per GDP is associated with a slower annual growth rate of between 0.1 and 0.18 percent for the period of 1960 to 1985. Engen and Skinner (1992)[8] had similar results and found that a percentage point increase in government expenditure per GDP is associated with an annual growth rate that is 0.108 percentage points lower for their sample of 107 countries over the period from 1970 to 1985.

Over the very long run, projected AMT revenues increase so much that revenue neutrality seems inadvisable. Strict adherence to revenue neutrality would push revenues and spending as a share of GDP to unprecedented heights, knocking perhaps as much as a percent a year off of expected long run growth. Achieving long-run revenue neutrality would require income tax rates to rise to very high levels as well.

In the nearer term, however, this Committee may wish to consider different offsets to reduce the fiscal impact of AMT repeal. Some may argue that the current fiscal situation is the result of reduced revenue associated with the tax cuts from 2001 through 2004, and hence, that any offset to AMT repeal should come from a reversal of these tax reductions. However, analysis of the sources of change in the long run fiscal outlook does not support this view unambiguously, even before one considers the possibility that the tax reductions provided a stimulus to economic growth. In particular, conclusions about the relative importance of spending and revenue revisions in explaining the change in the overall fiscal outlook depend on what year one chooses to construct the base case comparison.

Figure 4 and 5 help illustrate this point. They compare current baseline spending and revenue to what was projected to occur in the 2000 CBO budget outlook. Figure 4 documents that the difference in projected revenue is $180 billion in 2006, shrinking to only $65 billion by 2009. As can be seen in Figure 5, the difference in spending is much larger, a striking $425 billion in 2006, climbing to $535 billion in 2009. From this comparison, it appears that spending increases that were unanticipated in 2000, more than lower tax revenues, explain the transition from projected budget surpluses to budget deficits.

It is worth noting that as one uses more recent CBO projections, the relative importance of spending diminishes (but does not disappear) as an explanation for deficits. This is because the CBO incorrectly (in retrospect) ratcheted up revenue and growth forecasts just after the forecasts included in these charts. Even with the recession, the 2000 forecast underestimated the level of GDP that would occur today by a significant amount, which reduced its revenue forecast as well. Forecasts that were made later erred in the opposite direction on both GDP and revenue. While it is difficult to choose which error should be the basis of comparison, it is interesting to note that we would be in a dramatically better fiscal situation today if the 2000 forecast for spending had been correct, even with today’s revenues. If a CBO analyst from late 1999 were transported in a time machine to today, she would find the level of spending to be far more surprising than the level of revenues. It is simply incorrect to assert that the lion’s share of the fiscal deterioration is unambiguously attributable to EGTRRA, even before one allows for revenue feedbacks from lower tax rates.

If these pictures suggest that spending is an important additional source of the fiscal deterioration, then one might wish to explore what changes to spending would be required if one wanted to repeal the AMT and leave the deficit unaffected in the near term.

The CBO estimates that, relative to baseline budget projections, repealing the AMT would cause tax revenues to decline by approximately $600 billion over the next ten years. If the EGTRRA tax provisions are made permanent, the projected revenue reduction jumps to about $900 billion.[9]
 
Because spending is projected to increase over the next ten years, the cost of eliminating the AMT could plausibly be offset with reductions in spending growth, without actually reducing spending itself. The current projected annual growth in total outlays from 2005 to 2014 is 4.5%; if that growth rate were decreased by 0.37 percentage points, the resulting savings would offset the decline in revenues from repealing the AMT. Focusing on discretionary spending only, the projected 1.9% annual growth rate would have to be decreased by 1.15 percentage points. To pay for AMT repeal if EGTRRA is made permanent, those figures would have to change to 0.55 and 1.75 percentage points, respectively.

While these cuts may seem politically impossible, it is worth noting that spending would be extraordinarily higher than it was forecast to be a little more than five years ago, even after these reductions.

I would seek to reduce spending while repealing the AMT. I understand that others would prefer alternative approaches. The AMT is such a terrible tax, however, that I hope strategies for achieving AMT reform in a responsible manner will receive the careful scrutiny of this Committee, and that disagreement concerning the merits of the different approaches does not become an obstacle to a reform that is necessary and urgently needed.

Kevin A. Hassett is the director of economic policy studies at AEI.
 
Notes

[1] Burman, Leonard E., William G. Gale, and Jeffrey Rohaly (2003) “The AMT: Projections and Problems,” Tax Notes, Vol. 100, No. 1, pp. 105-117. Burman, Leonard E., William G. Gale, and Jeffrey Rohaly. (2003) “The Expanding Reach of the Alternative Minimum Tax.” Journal of Economic Perspectives, Vol. 17, No. 2, pp. 173-86

[2] Feenberg, Daniel R. and James M Poterba. (2003) “The Alternative Minimum Tax and Effective Marginal Tax Rates.” MIT Department of Economics Working Paper No. 03-37, pp. 28

[3] Ibid., page 16

[4] See Prescott, Edward (2002) “Prosperity and Depression,” American Economic Review, Vol. 92, No. 2, pp. 1-15. Prescott, Edward (2003) “Why Do Americans Work So Much More than Europeans?” Federal Reserve Bank of Minneapolis Staff Report 321. Davis, Steven J.,and Magnus Henrekson. (2004) “Tax Effects on Work Activity, Industry Mix and Shadow Economy Size: Evidence from Rich-Country Comparisons.” NBER Working Paper 10509

[5] Congressional Budget Office, The Long-Term Budget Outlook, December 2003.

[6] Calomiris, Charles W. and Kevin A. Hassett. (2002) “Marginal Tax Rate Cuts and the Public Tax Debate.” National Tax Journal, Vol. 55, No. 1, pp. 119-131.

[7] Barro, Robert J. (1991) “Economic Growth in a Cross Section of Countries.” The Quarterly Journal of Economics. Vol. 106, No. 2, pp. 407-443.

[8] Engen, Eric and Jonathan Skinner. (1992) “Fiscal Policy and Economic Growth.” NBER Working Paper No. 4223.

[9] “Revenue and Tax Policy Brief: The Alternative Minimum Tax”, Congressional Budget Office, April 15, 2004. As budget windows change, these estimates increase.