Bush Savings Incentives

Brief Analyses | Taxes

No. 477
Monday, June 21, 2004
by R. Glenn Hubbard

In the 2004 Budget, President Bush proposed to simplify and expand saving incentives. Instead of choosing from among a bewildering array of narrowly focused accounts, individuals would be able to save much larger amounts in a Lifetime Savings Account usable for any purpose and a Retirement Savings Account designed for retirement. His proposals will stimulate broader debates over tax reform and Social Security reform.

Lifetime Savings. The Lifetime Savings Account (LSA) is substantially simpler than a number of tax-advantaged accounts, such as Education Savings Accounts and Health Savings Accounts. The LSA would allow families to save more easily for a home down-payment, children's education or medical expenses. Although individuals could not deduct LSA deposits from their taxable income, account earnings would accumulate tax free. There would be no income limits on who could establish the accounts or make deposits. With no penalties or taxes on withdrawals, these accounts create stronger incentives to save than traditional IRA and 401(k) accounts - particularly among moderate income households who worry about tying up funds for long time periods, or about penalties for early withdrawals.

Like the president's proposal to eliminate investor-level taxes on dividends, the LSA is consistent with the idea that income should be taxed only once. Indeed, because of the generous contribution limit of $5,000, most households would be able to shelter all the returns on their savings from the income tax. As with tax reform proposals for a consumption tax or a Flat Tax, they would pay taxes when they earned wages or business income, but not on the returns to saving. This is an important step toward fundamental tax reform.

Retirement Savings. The other account President Bush has proposed, Retirement Savings Accounts (RSAs), would replace Individual Retirement Accounts and Roth IRAs. Contributions to traditional IRAs are tax-deductible, but withdrawals are taxed - often at the higher marginal tax rates faced by seniors. Deposits to Roth IRAs are made from after-tax income, but the interest earned is not taxed when the funds are withdrawn. Like Roth IRAs, contributions to RSAs would also be made from after-tax income and there would be no tax on earnings withdrawn at age 58 or later. The annual contribution limit on RSAs would be higher ($5,000) than current IRAs and there would be no income limits on who could participate.

IRAs were very popular in the mid-1980s, when there were no income limits. When Congress imposed income limits, the number of eligible savers was greatly reduced and IRA deposits plummeted. RSAs avoid this problem.

Cost of Tax Incentives for Savings. Opponents will likely argue that these proposals will generate little new saving, and merely reward saving done anyway, particularly by affluent households. If that were the case, the federal government would forgo tax revenues for little economic benefit. Indeed, it is difficult to determine what portion of contributions to currently available tax-advantaged accounts - principally IRAs - is new saving. Most estimates of how much of each dollar contributed is new savings range from zero to 60 percent. The studies are not conclusive, but a conservative estimate of IR A deposits that represent new savings is 26 cents of each dollar.

But that is only part of the question. Each dollar saved is invested, adding to the capital stock that fuels economic growth. The fraction of each dollar contributed to an IRA that is new savings will grow over time as it earns interest. Thus, to determine the cost-effectiveness of these accounts, one should compare the present value of the private capital added to the present value of the tax revenue lost. Thus, for tax-deductible IRAs, Jonathan Skinner of Dartmouth College and I calculated that:

  • If just 26 cents of each dollar contributed to an IRA is new saving, that 26 cents generates $2.21 of new capital for each dollar of tax revenue the government forgoes. [See the figure.]
  • If the tax subsidy is financed by borrowing (thus removing funds from the capital market) we get $1.21 of new capital, on net.
  • If as much as 40 cents of each dollar is new saving, we gain $4.31 of new capital for each $1 of tax subsidy.
Net Increase in Capital Per Dollar of Tax Subsidy

When we include the corporate income taxes paid on profits generated by the higher capital stock - as Harvard University economist Martin Feldstein suggests we should, the results are even more dramatic:

  • If 26 cents of each dollar contributed is new saving, then $4.84 is added to the capital stock for each net dollar of revenue lost.
  • If each dollar contributed contains at least 40 cents of new saving, the tax cut actually pays for itself.

These results would be even more favorable if we did the same calculation for Roth IRAs, and LSAs and RSAs, since taxes have already been paid on the wages deposited to these accounts.

Thus the expanded savings incentives proposed by President Bush are likely to generate substantial gains in capital accumulation per dollar of net revenue cost. By simplifying the current patchwork of savings incentives, the Bush plan would increase participation. By significantly increasing contribution limits, a greater share of contributions should consist of new saving - especially over time.

But these promising calculations don't answer the bigger question. The higher capital stock from saving incentives is not manna from heaven; rather, it is the consequence of households consuming less today to have more resources later, say in retirement. Why should we do this?

Retirement Savings and Social Security Reform. All of these accounts reduce the economic harm from taxing capital. However, if that were our only concern, we could simply exempt all or part of all capital income from taxation, as President Bush recommended in his proposal to eliminate investor-level taxes on dividends. For most households, the LSA's high contribution limits would accomplish this. Therefore, the remaining question is: Why do we need Retirement Savings Accounts?

The usefulness of RSAs can be seen in the context of Social Security reform. Any successful reform that restores Social Security's long-term financial footing is likely to reduce average replacement rates - benefits relative to average wages - for young workers and future generations. Accordingly, private saving will likely need to play a larger role. The administration's RSA proposal would provide a significant vehicle for accomplishing this. It would also mesh well with Social Security Personal Accounts, into which a portion of an individual's Social Security payroll taxes could be deposited.

Conclusion. The administration's proposals for new and improved saving incentives make good sense. Their simplicity alone would be a good reason to champion them. But they also constitute bold new initiatives that are consistent with tax reform and Social Security reform. This makes them worth fighting for.

R. Glenn Hubbard is a former chairman of the President's Council of Economic Advisers and is a professor at Columbia University. A version of this essay appeared in The Wall Street Journal.


Read Article as PDF