Retirement savings tax incentives and the data

Most agree that at some point in the relatively near future Congress will “do something” about the budget deficit, entitlements and tax reform. While there is much partisan disagreement about what should be done, there is widespread, bipartisan agreement that something must be done, and that these three issues — the deficit, entitlements and taxes — are critical.

Many, including, e.g., President Obama’s National Commission on Fiscal Responsibility and Reform, have suggested that tax reform should involve a reduction of marginal rates and a broadening the tax base. The idea is that you can increase revenues by lowering rates while closing “loopholes.” The big “loophole” targets are the Child Tax Credit, the Earned Income Tax Credit, the mortgage deduction, health care tax benefits and retirement savings tax benefits.

In this article we review some of the basic data on just how big the tax benefits for retirement savings are.

Focus on DC plans

In this discussion we are going to focus on defined contribution plan tax benefits. Changes to the tax treatment of DC plans has generally been the focus of most policymakers.

There are several reasons for this. Many policymakers “like” DB plans, believing them to be both “fair” (that is, they provide benefits to lower paid employees) and “adequate” (that is, they provide a significant retirement income). Others see DB funding as a “wind-up” of a “legacy” obligation, with many companies simply funding benefits that have already accrued. So, politically, the possibility that Congress will, for instance, take away or significantly reduce the deduction for funding DB plans is (with some very limited exceptions) simply not on the table.

On the other hand, some policymakers have questioned both the fairness and adequacy of DC plans, and particularly 401(k) plans, and much of the “action” — new plan formation and new contributions – takes place in DC plans.

The retirement savings “tax expenditure”

“Tax expenditure” is a concept used by the Joint Committee on Taxation (a committee made up of both Senators and Congressmen). Section 3(3) of the Budget Act of 1974 defines tax expenditures as:

[R]evenue losses attributable to provisions of the Federal tax laws which allow a special exclusion, exemption, or deduction from gross income or which provide a special credit, a preferential rate of tax, or a deferral of tax liability ….

Tax expenditures are, for many policymakers, a starting point for decisions about tax policy. For example, when thinking about “where to cut” tax benefits, many policymakers are likely to look at tax expenditure numbers to determine where the biggest “loopholes” are and how much revenue can be gained by making changes.

For 2012, the Joint Committee estimates the tax expenditure for DC plans to be $60.8 billion; the tax expenditure for DB plans is $46.3 billion. “DC” plans for this purpose does not only include 401(k) plans. Other corporate plans, e.g., profit sharing and money purchase plans, are included, as are non-corporate DC plans, such as the Federal Thrift Savings Plan and 403(b) plans.

How the tax expenditure number is calculated

Tax expenditures for DC plans are calculated on a “cash flow” basis. The Joint Committee estimates include: (1) deductions/exclusions for contributions (revenue lost); (2) untaxed trust earnings (revenue lost); and (3) taxes paid on distributions (revenue gained). Estimates are made for the current year and over the budget window. So, the net tax expenditure for DC plans is $60.8 billion for fiscal 2012; for the period 2011-2015 it is $375.9 billion.

This cash flow approach may work well for certain tax expenditures, but many have criticized its application to retirement savings tax incentives. Assuming constant contributions and distributions, and constant tax rates, it is possible that a cash flow approach will capture, over time, the “true cost” of retirement savings tax incentives. But given what has actually happened over the past 20 years, cash flow numbers are likely to exaggerate the number. Because 401(k) plans are relatively new, and baby boomers have only just begun to retire, deductions/exclusions for contributions are relatively high and taxes paid on distributions are relatively low. A longer budget “window” (longer than the 5- or 10-year window used for tax expenditure estimates) would capture (relatively) more distributions and lower the estimated net cost to the Treasury of retirement savings tax incentives.

Revenue estimates

While tax expenditure numbers may be a useful place for policymakers to start in determining where to “cut,” revenue estimates — “scoring” the cost/savings to the tax system of particular legislation — are what ultimately affects budget numbers. And while revenue estimates are determined using a cash flow methodology similar to that used for tax expenditures, a critical difference is that revenue estimates take into account certain taxpayer behavioral responses to the proposed change. With respect to retirement savings tax incentives, for instance, a revenue estimate is likely to at least consider whether taxpayers whose ability to save on a tax preferred basis in a 401(k) plan is reduced by a cut in contribution limits might seek other tax favored investments, e.g., municipal bonds.

NPV estimates

An alternative way to look at the “cost”/tax expenditure for retirement savings tax incentives would be to estimate total tax benefits over time with respect to, e.g., $1 of contributions in the current year, and then discount the value of those benefits back to the current year.

Assuming constant tax rates (that is, assuming the taxpayer pays the same marginal tax at the date of contribution and at the date of distribution), the “value” of the tax deduction nets fully against the “cost” of the ultimate taxes paid. The “real” tax benefit is the deferral of taxation on trust earnings. All policymakers and analysts of whatever persuasion accept this premise. Given this premise, the “value” of retirement savings tax incentives is generally understood to be:

(1) The value of the deferral of taxation on trust earnings, plus,

(2) if (as is common) the taxpayer pays a lower marginal tax rate in retirement, the value of, effectively, paying a lower tax rate on the contribution.

In a December, 2011 article in the National Tax Journal that has received wide circulation, two officials of the Department of Treasury have re-estimated the value retirement savings tax incentives on a NPV basis. (Lurie and Ramnath, Long-Run Changes In Tax Expenditures On 401(K)-Type Retirement Plans.)

NPV estimates are very assumption sensitive. For instance, the value of the deferral of taxation of trust earnings is dependent on your assumption of the amount of trust earnings; high earnings = higher value. What discount rate you use also has a significant effect.

To get an idea of the effect, on an NPV basis, of changes in DC contribution limits, the authors estimated the change in the NPV of tax benefits resulting from limiting contributions to $10,000. By their analysis, such a limit would “mainly [impact] taxpayers with earnings above $150,000, as 80 percent of such taxpayers reduce their contributions, on average losing $3,200 in tax benefits from their 2008 contributions.”

The bottom line: in a low interest rate environment (like the present) a reduction of contribution limits to $10,000 — which may well be in line with what policymakers are considering — produces only $12,543 billion in additional revenues (using a 2% discount rate/2% earnings assumption), when calculated on an NPV basis.

The author’s conclusion:

Notwithstanding the large effect on high earners [of the reduction in 401(k)-type limits to $10,000], we find that the $10,000 limit imposed on contributions to 401(k)-type plans, even at high rates of return, decreases the NPV cost of the tax expenditure by at most $33 billion, or about 2 percent of the current $1.5 trillion deficit. Hence, if policymakers hope to raise substantial revenue from reforming 401(k)-type plans, a much deeper reduction in the contributions limit or other reforms would be needed.

* * *

There are very real questions about value of the tax benefit for retirement savings, how that value should be calculated, and how much revenue reductions in contribution limits will produce. These questions go to the heart of the Congressional budget process. How policymakers view these issues will be a critical element in decisions about whether to cut DC contribution limits and by how much.

This is the first article in a series on basic issues underlying consideration of reform of tax incentives for retirement savings. In our next article we will discuss debates over, and data with respect to, the “fairness” of the current system.