Background on Potential Budget Gimmick in Tax Reconciliation Conference
March 29, 2006
Lawmakers currently trying to work out differences between the versions of the tax cut reconciliation legislation that the House and Senate passed last year reportedly are considering using a gimmick -- which entails claiming that one tax cut "pays for" another tax cut -- to evade an important Senate rule.
That Senate rule prohibits use of reconciliation legislation to increase the deficit in any year outside of the period to which the reconciliation process applies, which is from fiscal year 2006 through fiscal year 2010 for the pending tax-cut reconciliation measure. That rule is the only current procedural barrier against enactment of reconciliation tax or entitlement legislation that would increase long-term deficits.
Deficit Reduction and the Reconciliation Process
The reconciliation process provides for consideration of tax and entitlement legislation under special rules in the House and Senate, including a prohibition against a filibuster in the Senate. In the 1980s and 1990s, the reconciliation process was used to reduce deficits. In recent years, it has also been used to enact legislation that increased the deficit. Although some members of Congress and outside budget watchers have questioned the propriety of using the reconciliation process to increase the deficit within the budget planning period, such increases have been allowed under House and Senate rules. But a key Senate rule has kept the reconciliation process at least from being used to increase the deficit permanently.
This Senate rule prohibits a reconciliation bill from increasing the deficit outside the budget planning period (which generally is a five- or ten-year period, depending on the number of years covered by the budget resolution). That is, any reduction in taxes or increase in entitlement spending in any year outside the budget planning period can be included in a reconciliation bill only if it is fully offset in every year outside the planning period by other provisions that raise revenues or reduce entitlement expenditures by equivalent amounts. This effectively means that legislation that increases the deficit inside the budget planning period can be passed in the Senate with a bare majority vote (50 Senators voting in favor plus a vote by the Vice President to break the tie), but legislation increasing the long-term deficit (the deficit outside the budget planning period) can be passed only with the support of 60 Senators.
The immediate challenge to this rule has been precipitated by a provision included in the House version of the tax reconciliation legislation that would extend for two years the reduced rates on capital gains and dividend income that were enacted in 2003 and are scheduled to expire at the end of 2008. According to the Joint Committee on Taxation, this provision would reduce revenues by $51 billion over ten years, with $31 billion of the revenue loss coming in 2011 through 2015, which is outside the five-year period that reconciliation covers. Without offsets, this provision clearly would violate the Senate rule. The Senate-passed reconciliation bill contained some provisions that would increase revenues in years after 2010 and could be used to offset part of the costs of extending the capital gains and dividend cuts, but House conferees reportedly are unwilling to agree to those measures.
To include the capital gains and dividend cuts in reconciliation without revenue-raising measures to offset the costs in 2011-2015, conferees are reportedly considering a gimmick that would allow them to evade the rule and claim that the costs of the capital gains and dividend cuts are offset not by provisions that would raise revenues, but by another tax cut.
Lifting the Income Limits on Roth IRAs
To "pay for" the costs of the capital gains and dividend cuts, the conferees are considering a proposal to remove the income limits on who can convert traditional Individual Retirement Accounts (IRAs) to Roth IRAs (most likely for a temporary period). Based on past Joint Committee on Taxation estimates of similar proposals, lifting the income limits on Roth IRAs would spur a large number of high-income households to convert their traditional IRAs to Roth IRAs in order to take advantage of the long-term Roth IRA tax breaks. This would lead to an increase in revenues over the 2011-2015 period, because people converting a traditional IRA to a Roth IRA would pay taxes in that period on the amount being converted.
But it also would reduce revenues in years beyond 2015, because withdrawals in retirement from the new Roth IRAs would be tax free. This is essentially a timing shift that accelerates into the 2011-2015 period revenues that otherwise would be collected in subsequent years.
Moreover, over the long run, the proposal would result in a net reduction in tax revenues. People would elect to convert their traditional IRAs to Roth IRAs only if doing so would be to their advantage because it would lower their tax bills; that is, they would elect to pay some taxes now only if they expected it would reduce their tax bills by a larger amount in the future. Anyone who expects that such a transaction would increase his or her tax bill simply would not make the transaction.
Budget Gimmick Would Violate Senate Rule
The loss of revenues in years after 2015 that would result from this proposal -- a loss confirmed by a recent Congressional Research Service analysis -- would clearly violate the Senate rule, assuming there were no offsetting tax increases. However, proponents of this plan apparently hope to convince the Senate Parliamentarian that a Joint Committee on Taxation cost estimate that covers only years through 2015 and shows no net loss of revenues in 2011 through 2015 is the only evidence he should take into account in determining whether there is an increase in the deficit after 2010. In other words, the advocates would claim implicitly that the fact that the Joint Committee typically does not issue cost estimates beyond 10 years should be taken as evidence that there is no cost, even though the cost is patently obvious.