Tax reform and retirement savings

On September 15, 2011, House Majority Leader John Boehner called for the “Super Committee” (the committee authorized by the debt ceiling compromise) to “develop principles for broad-based tax reform that will lower rates for individuals and corporations while closing deductions, credits, and special carveouts in our tax code. Tax reform should deal with the whole tax code, both the personal side and the corporate side, and it should result in a code that is simpler and fairer to everyone.”

On the same day, a bi-partisan group of 36 Senators called for that same committee to “[f]ocus on the major parts of the budget and include long-term entitlement reforms and pro-growth tax reform.”

This article reviews the extent to which the tax reform debate directly affects retirement plans. In that regard, it’s worth noting that on the same day that Speaker Boehner and the 36 Senators were calling for broad based tax reform, the Senate Finance Committee held a hearing titled “Tax Reform Options: Promoting Retirement Security.” At that hearing several proposals for drastically revising the current system were discussed.

Significance of tax reform for retirement plans

Tax benefits are a primary reason why employers currently maintain retirement plans. Why is that? Let’s take as a baseline that, absent some reason external to the firm, like the Tax Code, cash compensation is generally easiest both for the employer and the employee. So that “normally” both employer and employee would prefer cash; and the employee can then choose to provide (or not provide) for his or her own retirement. Even in this simple world, however, there will be firms with a particular sort of culture — and long-term employer-employee relationships — that make the provision of a pension a preferred element of compensation.

Private sector retirement plans got their start during World War II as a way around wartime wage controls. This fact highlights one non-tax-based “feature” of, particularly, defined benefit plans — their lack of transparency. That lack of transparency has in the past been a virtue in many circumstances. Classically, it has allowed some companies to understate the value of benefits, for accounting or funding purposes, allowing them to provide $1.05 worth of benefits while only booking $0.95. Or, indeed, to overstate the value of benefits, for tax purposes or, e.g., under a cost-plus contract or as a “base” for utility rate-setting.

But the world-wide drive for transparency — resulting in changes in accounting and funding regimes — has turned this “feature” into a “bug.” And the result is a widespread movement away from less-transparent DB plans and towards more transparent account-based plans. The latter are, in many respects (and often explicitly in their design), indistinguishable from cash compensation. Indeed, they are (as in the 401(k) plan) often funded out of the employees’ cash compensation.

And the primary virtue of these plans is, both for the employer and the employee, that they are a good tax deal. The employer is able to provide $1.20 worth of compensation at a cost of $1, because of the magic of our current tax system. That’s why, currently, most private sector retirement plans exist. (Those numbers — $1.20/$1 — are, by the way, just “made up” for purposes of illustration. The actual value of this benefit will depend on how long saved money remains in the tax exempt trust and how much it earns.)

Who benefits

Under the current system, this tax deal works as follows: money goes in tax free; it accumulates tax free; and it is taxed when distributed. The math here is important. The value of the tax benefit is not the value of the deduction. It is the value of the tax free earnings.

It is axiomatic that under such a system, the employees who benefit the most are those with the highest marginal tax rate. Indeed, for the nearly half of American workers who pay no federal income taxes at all, there is no explicit tax benefit under this system.

The tradeoff — the quid pro quo for this tax benefit for higher paid/high marginal tax rate employees — is that an employer can’t have one of these plans unless low paid (and low marginal tax rate) employees are also covered. In the 401(k) plan this rule is implemented via the actual deferral percentage (ADP) test. And, in 401(k) plans for instance, to assure participation by low-paid employees, employers provide matching contributions, “advertising” and automatic enrollment to drive employee participation. The Tax Code rules that implement this policy are called the “nondiscrimination” rules.

So, to “buy” the tax benefit for high margin taxpayer-employees, the employer, under this system, provides a benefit to low margin taxpayer-employees. We call this the “implicit tax benefit” under the current system.

What is the value of that implicit tax benefit? No one knows. It can be defined as the value of benefits provided to low-paid employees over and above the cash wages the employer would have paid them if there were no plan. (Thus, and ironically, to the extent the employee views the plan benefit as the same as cash wages, there is no implicit benefit.) By the way, this implicit tax benefit is financed by the employer and is not run through the tax system.

Summarizing: Employers want to convert some of the wages they would otherwise pay to high margin taxpayer-employees into tax favored “retirement benefit” dollars. To these employees they provide (under the current tax system) $1.20 of value at a cost of $1. That’s the explicit tax benefit. To accomplish this they must provide a benefit to non-high margin taxpayer-employees — a broad-based plan, matching contributions, etc. That is the implicit tax benefit.

Current tax reform proposals — limit cuts

There are, essentially, two different sorts of tax reform proposals that would affect the current retirement savings tax incentive system.

First, there are proposals to reduce the tax benefit for retirement savings, by cutting the upper limit on contributions. The “cut” proposal getting the most press is the one included in the Final Report of President Obama’s National Commission on Fiscal Responsibility and Reform (NCFRR — also known as the Simpson-Bowles deficit commission), the so-called 20/20 proposal. That proposal would “Consolidate retirement accounts; cap tax-preferred contributions to lower of $20,000 or 20% of income, expand saver’s credit.” It’s not entirely clear how this would work, but it is generally being interpreted as a reduction in tax benefits for contributions to account based plans (e.g., 401(k) and 403(b) plans). Defined contribution plan limits are now $49,000/100% of income. So these cuts would be substantial — more than 50%.

Cut in limits = a cut in savings?

Will that reduce retirement savings? Axiomatically, yeah. But it is fair to ask (as advocates of this proposal no doubt will) whether that’s really a problem. More money will, presumably, be paid in cash to employees. After taxes are taken out, will that money be saved? There are certainly lots of policymakers who think that right now it would be better if it were spent — improving the economy — than saved. But it is certainly at least as likely that high margin taxpayers will look around for some other tax favored investment (life insurance? capital gains? municipal bonds?) or simply save this “extra” cash on a taxable basis. On the other hand, it is likely that the retirement savings of non-high margin taxpayer-employees will go down — these employees are generally cash hungry.

Moreover, it is fair for some policymakers to ask whether, e.g., 401(k) savings are really retirement savings anyway, given the prevalence of loans, distributions for home buying, etc. (aka “leakage”).

Who loses?

Tax “expenditures” would be cut under this proposal, and somebody is going to lose. Who? Both the high margin taxpayer-employees and non-high margin taxpayer-employees. The high margin taxpayer-employees no longer get tax-favored compensation; for some portion of his or her compensation he or she will now only get $1 when he or she used to get (after taking account of tax effects) $1.20. The low margin taxpayer-employees will no longer get the corresponding implicit tax benefit — some portion of the contributions in support of a broad based plan.

Quantifying all of this is impossible, since there is no way of determining what the implicit tax benefit is worth. But it is likely, as limits go down, that (relatively) more implicit tax benefits (that is, benefits for low paid employees) will be lost than explicit tax benefits (that is, benefits for high paid employees). The analysis here is very technical; basically, the problem is that the last dollar contributed for a low-paid employee is more likely to contain more of the implicit tax benefit.

So long as all employers have to play by the same rules, employers will not lose. In fact, their life may become simpler, because compensation will be both simpler and more transparent.

Interestingly, if tax benefits for “retirement compensation” are cut, then at the margin employees will be paid less (the employer is not pocketing this money in any meaningful sense). That is, employees used to get their pay + an (explicit or implicit) tax benefit. Now they will only get their pay. You could even argue that these sorts of cuts are “job killers” — the alternative to cutting pay for all employees is to fire some of them and increase pay for the rest.

Finally, it’s worth asking at what point does this whole process tip? Whatever retirement savings tax incentives are left will have to be sufficiently more valuable than what is available via IRAs — sufficient to outweigh all the costs of administration, litigation risk, management time. At some point, if you cut the tax benefits enough, employers will simply exit the system.

Current tax reform proposals — tax credits

The second set of proposals are those to convert the current system of deductions to one of tax credits. One such proposal was put forward by William G. Gale of the Brookings Institution at the September 15, 2011 Senate hearing. Under this proposal the current system of deductions would be converted to a flat-rate refundable credit — in effect, a matching contribution — of 18% of a taxpayer’s retirement saving contributions. Everything else — contributions and earnings — would be taxed at regular rates.

There are other proposals like Gale’s. Professor Theresa Ghilarducci has a widely publicized tax-credit proposal; and the Administration has proposed an expansion of the Saver’s Credit that could have a similar effect. All of these tax credit proposals shift the tax incentive/benefit from high margin taxpayer-employees to “all” employees.

At the strategic level, this sort of proposal increases the progressivity of the Tax Code. But, let’s set aside the issue of whether refundable tax credits or tax deductions are “fairer.” How would such a proposal affect private retirement savings?

Employer role in a tax credit driven system

It’s unclear under these proposals what the ongoing employer role would be. Under the current system, higher contributions/deductions (that is, higher than available under, e.g., an IRA) are only available under an employer-sponsored qualified plan. That plan must provide benefits for a broad cross section of employees. The “employer provided benefit for a broad cross-section” — what we have called the implicit tax benefit — is the trade off for the tax benefits high margin taxpayer-employees get from this system. If you flatten the explicit tax benefit and, thus, eliminate the need for an employer-provided implicit tax benefit, why do you still need the employer?

Clearly, Mr. Gale’s proposal assumes there will be continued employer involvement, but why? Are employers to be given the “privilege” of offering a plan that generates higher tax credits than would be available under, say, an IRA? Why? Why shouldn’t contributors to IRAs get the same tax benefit? And if they do, why should the employer continue to maintain a plan?

Conceivably, the employer could make the contributions that generate the tax credit. Think about it, though: such contributions would be (from a tax standpoint) indistinguishable from cash wages but would be subject to withdrawal restrictions, etc. Why wouldn’t the employer just pay all employees cash and let them decide whether to save? No doubt there are employers that would make contributions (in effect, forced savings minus any tax break), but we would guess that there are not many.

Does it make sense to subtract the employer from the retirement savings process? Doing so seems to discount employer “value add” in marketing the plan and obtaining (at larger firms at least) institutional discounts for administration and investment. The issue of institutional vs. retail is not insignificant, but the employer may not be the only “solution” to the retail problem — unions, trade associations or buyers co-ops might also fill this role. One wonders, however, whether tearing down the installed employer infrastructure for this — scaled administration and investment — might cost more than it saves.

Who manages the money?

In this regard, a really big question is — if there is no employer intermediation (except, perhaps, in payroll deduction and transmitting funds) — who is going to manage this money? When different automatic IRA proposals were drafted last year, some fairly unusual approaches — including an investment manager randomly assigned from a “carousel” — were proposed. With the shrinking of high-margin taxpayer savings, the accounts generated by tax credits could be fairly small, and investment manager interest may be limited. One answer to these questions is provided under Professor Ghilarducci’s proposal: savings would be “administered by the Social Security Administration” and credited with 3% earnings.

Would people continue to save for retirement?

It’s also fair to ask how well such a system would work in real life at producing retirement savings. High paid/high margin tax rate employees can safely be assumed to have a higher preference for savings-over-cash than non-high paid/high margin tax rate employees. Indeed, higher paid employees are generally older, and certainly older employees have such a preference. So this proposal may, in effect, provide tax benefits where they are not needed, distorting behavior, and, in the end, everybody — high- and low-paid employees — may decide that the whole thing just isn’t worth it.

Does a matching contribution approach make sense?

And there’s another issue, with respect to Gale’s proposal at least: why a match? Indeed, some arguing for greater progessivity-as-fairness suggest that benefits under Social Security for lower paid employees should be enhanced. That is a much more direct way of providing retirement benefits to low-paid employees. And it doesn’t leave out those low paid employees (certainly there are a lot of these) who cannot save anything and therefore will not get a match under the refundable-tax-credit-match proposals most commonly under consideration.


Given the widespread discussion of a need for comprehensive tax reform, it’s likely that something is going to happen. Probably after next year’s election, but conceivably before. Much of the debate over future retirement savings tax policy is, obviously, affected by “meta” arguments over how progressive the Tax Code should be. Cuts in defined contribution plan contribution limits are likely to be part of the discussion; and, as long as there are Democrats at the table, the issue of converting some deductions to tax credits is also likely to be considered.

The details of these proposals seem, at this point, significantly undefined and under-analyzed. It’s not clear that the consequences of limit cuts have been (1) fully thought through (it seems as if the NCFRR simply thought “if we reduce this limit to $20,000, we will generate a lot of revenue”) or (2) exposed to the opinions of significant interests, e.g., employers, administrators and the investment manager community. Likewise, if tax reformers are going to take tax credit proposals seriously, employers, administrators and the investment manager community are likely to have very strong opinions on how a tax credit system should be implemented.

All of which is to say that something is likely to happen in the relatively near future, but that it (that something) is likely to go through a lot of permutations before it is passed into law.