Two Kinds of Office

The two quotes above illustrate two ways of thinking about the firm, firm “culture” and employer-employee relations.

Office 1: At one end of the continuum, we can (and many employers do) think about employer-employee relations, and critically (for our purposes) compensation and benefits policy, as (using Hayek’s term) a “macro-cosmic” transaction, as strictly financial and un-affected by “sentimental yearnings.”

There are lots of companies in America that must operate within a very rigid financial calculus. They raise money in competitive capital markets over which they have no control. To survive they must produce a minimum return on investment. They make goods or provide services that are sold in competitive markets over which, again, they have no control. To survive they must price their goods as dictated by those markets.

What happens in between, critically, for our purposes, how much they pay their employees in compensation and benefits, must fit inside that rigid calculation. If they are, e.g., going to increase retirement benefits, the cost of that increase will generally have to come out of the employees’ pocket. There’s no other pocket it can come out of.

Office 2: At the other end of the continuum are companies like the one Michael Scott is imagining, where the values of the micro-cosmos—the family—play a significant part in compensation and benefits policy decisions. Often, these are companies that are in some sense less accountable to the capital markets (e.g., they are very large, and share ownership is widely dispersed) and/or are less accountable to consumers (e.g., they enjoy some formal or informal monopoly power).

For the latter companies, decisions about retirement benefits may involve many factors other than a rigid financial calculation: personal issues, as in “we have a commitment to taking care of our people,” reputational issues and regulatory issues. Increases in retirement benefits may be paid for by investors or consumers. For instance, a public utility may be able to pay for an increase in retirement benefits simply by persuading the public agency that supervises it to allow a rate increase.

It is in the context of these two very different firm-economies—and of all those in between—that we consider some fundamental questions: What are retirement benefits? Who pays for them? And how does tax policy affect the answers to those questions?

What Are the Retirement Benefits?

With respect to DB plans, the answer to this question is somewhat obscure and ambiguous. Indeed, the question itself is a little obscure—what I’m trying to ask is, how do DB plans—as an employer-provided retirement benefit—function in the economy of the firm?

What, Legally, is a Retirement Benefit?

One obvious starting point: A retirement plan is a contract between the employees and the employer. The employees provide services; the employer pays their wages and will also, under the terms of the plan-contract, provide them with a benefit.

The services are provided, and the wages are paid, currently. But the benefit will be paid in the future. This is generally called an executory contract. The law-school example of an executory contract is: I offer you $50 to cross the Brooklyn Bridge. You start walking, but half way across I cancel my offer, or otherwise change the deal. What rights do you have?

That is, in fact and for our purposes, a pretty good analogy. Much of the participant unrest and unhappiness with the change from the DB system to the 401(k) system stems from the disappointment of expectations.

What Exactly is the Employer Promising?

So—what is the extent of this contract, how enforceable is this promise the employer makes? ERISA answers this (it would seem) pretty clearly. Participants have a legal right to their accrued benefit. This concept itself is a little ambiguous. ERISA is very clear about the benefit payable at normal retirement age (e.g., age 65)—a DB plan must specify exactly what that benefit is and how it is accrued. And whatever portion has accrued to date cannot be taken away.

Let’s go back to our example benefit of 1.5% of 3-years pay times service. If the employee currently has 10 years of service and 3-year average pay of $50,000, his accrued benefit is currently an annuity of $7,500 a year beginning at 65. The employer can’t take that benefit away.

But if the employer-sponsor freezes the plan, then it can stop that benefit from getting any bigger. Both service and pay after the freeze may be disregarded.

The Status of Subsidized Benefits

Now, what about, for example, an unreduced benefit that is available if a participant reaches age 55 and has 30 years of service? That’s a valuable benefit because the full benefit starts at 55 rather than 65, with no reduction. Can a sponsor simply take this benefit away “half way across the bridge?” Even for a participant that has, say, 29 years of service and is 54? The employee is still working, earning service under the plan, and is still getting older.

There was a lot of litigation over this issue until, in 1984, Congress passed the Retirement Equity Act, generally prohibiting the elimination of these sorts of “retirement-type subsidies” to the extent they relate to service before the change in the plan.

So, our 54-year-old with 29 years of service, if they stuck around for another year, could get that subsidized benefit, at least with respect to the service and pay they had when the amendment took place.

A Legitimate Expectation That the Employer Would Continue the Plan

What about the participant we discussed with respect to cash balance plan conversions—who, early in their career, received not-very-valuable accruals under a plan that “promised” a big payoff for full-career, older, longer service participants. And then had that big payoff taken away—in the cash balance conversion—just when it was about to become valuable?

Should that participant get some sort of right to that future benefit? This also was a major issue in the cash balance controversy.

So, one really deep issue with respect to DB plans was, employers and employees did not have a clear, or the same, idea of exactly what the deal was.

What, Economically, Is a Retirement Benefit?

With respect to DB plans, this also is a slippery question.

You would think—as many economists do—that because they are (as we just discussed) compensation for services, DB retirement benefits should be considered part of employee compensation. That is, as it were, the Office 1 version (presented above).

But, go back to Chapter 6. There we considered two identical employee-participants with the same pay and the same service and the same age-65 benefit—$15,000 per year. One is 40 and the other is 55. The 40-year-old’s benefit is worth (on a present value basis) around $67,000. The 55-year-old’s benefit is worth around $129,000. That difference generally has nothing to do with the value of the services the two participants are providing to the employer.

It sounds like there’s something else besides a strictly financial transaction in play here—something like the Office 2 version of the firm (above).

Obviously, In Some Cases, Retirement Benefits Aren’t Compensation

In Michael Scott’s “Office 2” world, retirement benefits are more about taking care of a “family member” than about paying an employee compensation for what their services are worth.

In that sort of corporate world, benefits function as something more than compensation: they function (at least in part) as a way of recognizing and reinforcing the sense of the microcosm—that this firm that I work at, and the people that I work with, are uniquely part of one microcosm, a “family” of sorts. Thus, health benefits are more valuable to employees with families, to certain demographics (women of child-bearing age, older employees) and even to certain individuals (e.g., those with chronic diseases). Because “we” take care of each other. In these micro-cosmic contexts, retirement benefits function as a way to “take care” of loyal, full-career employees.

IBM may actually have been a company like that, in the 1980s say. But by the 1990s it was having to compete with a very different type of company—new economy technology companies that view their employees as something like somewhat temporary “partners,” rather than as family members.

Are all companies going to adopt the corporate culture of new economy technology companies? That seems pretty unlikely. Not everybody wants to work for a company like that, and not everybody can.

DB Plans, a Verdict

Given all of the foregoing, the best that can be said for DB plans—as a way of providing retirement benefits—is that they work well in a limited set of conditions.

Both the legal ambiguity of the status of employees’ expectation rights and the differential treatment of identical (except for age) employees are manifestations of the DB plan’s lack of transparency. Because—going back to the beginning—DB plan design does not start with the compensation question, what is the employee worth? Instead it starts with the question, what does the employee need?

And in that regard, DB plans produce a one-size-fits-all answer: 70%–80% of pre-retirement pay, not considering any of the issues we raised in Chapter 5 about differing needs in retirement.

Nevertheless, the inherent inflexibility of the DB plan design may at some firms be a feature, not a bug—it guarantees that long-term employees will be taken care of. And while old economy technology companies may have had to abandon their traditional DB plans to survive in a new economy, lots of other companies—in the early 2000s—did not have to and did not do so.

Even the sponsor’s exposure to interest rate risk, was, as we’ve seen, tameable. If all those companies who wanted to stay in the DB system had, say, in 2000, bought LDI overlays and hedged out their interest rate risk, most of the interest rate losses sponsors have suffered could have been avoided. But that would have required something like clairvoyance.

In the end, what killed DB plans was not interest rate risk as such. It was that interest rates are now so low that providing a DB plan benefit is, for most companies, simply too expensive.

We will return to the theme of interest rates and retirement savings in Part III.

Finally, we would note that these problems either do not exist or are substantially less problematic with respect to cash balance plan designs, which, in our current 401(k)-dominated retirement savings system, remain a viable plan design option for many sponsors.

Who Pays for Retirement Benefits?

In the Office 1 case (at least), I find the answer to this question (also) to be difficult and obscure.

The simple answer (and the one economists often give) is that the employee, himself, pays for the retirement benefits he receives, by reducing his cash compensation. ABC Company proposes to a prospective employee: “We are going to pay you $25 an hour. $20 of that will be in cash, and $5 of it will be in the form of a retirement benefit.” And that may be literally true—ABC Company’s cost of this (prospective) employee’s retirement benefit might literally be $5 per hour.

This answer is, however, much too simple. Not all employees value retirement savings the same. That is a really critical point. Some employees—typically low-paid—have a high preference for cash. They are living paycheck to paycheck. They need cash to buy food and pay rent. It’s important to understand that this is not an abstract issue. It’s literally and practically true. Given the above proposition, our prospective employee might simply say: “No thank you. I am going to work for XYZ Company. They’re paying $23 an hour, all cash.”

On the other hand, some employees put a premium on the tax benefits tax qualified retirement savings plans provide. We could imagine another employee who responds to ABC Company’s proposition by saying: “Fabulous. Because $5 in retirement benefits is, given my tax bracket, worth $6 in cash.”

Further complicating this analysis, we need to consider the value of the federal tax subsidy itself: some portion of tax qualified retirement benefits are in effect “paid for” out of (lost) tax revenues. What the Congressional Budget Office calls “tax expenditures.”

There is some research that supports the proposition that high-paid employees are prepared to reduce their total pay to get tax-subsidized retirement benefits, and that that reduction may then be used to finance the (under-appreciated) retirement benefits provided low-paid employees.

In Do Low-Income Workers Benefit from 401(k) Plans? (The Urban Institute, 2011) Eric Toder and Karen E. Smith studied this issue. Quoting from their paper: “Among those who would otherwise consume additional wages, the relative value of employer contributions [to a tax-qualified retirement plan] depends on the size of their benefit from tax-free saving and their degree of preference for present over future consumption.”

Their conclusion: “The findings imply that both low- and high-income workers benefit from employer DC contributions. Low-income workers benefit because their total compensation rises.16 High-income workers benefit because the increased access to tax-advantaged saving more than offsets their loss of money wages, even though their total compensation is about the same. This suggests that conventional approaches may overstate the share of benefits from tax-preferred retirement saving plans with employer participation that go to high-income employees by assuming that contributions reduce wages equally for all employees.”

We will come back to this issue—fundamentally, who gets what and who pays for it—again when we examine tax policy in detail in Part II. But we want to, here at the conclusion of Part I, briefly consider the significance of tax policy for the firm dynamics we’ve just been discussing.

Retirement Savings Tax Policy—Two Views

Two different policy objectives drive much of retirement savings tax policy:

Policy Objective 1: Some policymakers are, in principle, opposed to taxing savings. This is why there is a tax deduction for retirement savings. In a perfect world, at the “extreme,” supporters of this policy goal would want the same deduction for any savings that is currently available for retirement savings; in other words, a consumption tax.

Policy Objective 2: Others are committed to using tax policy to provide retirement income to low-paid workers when they can’t work any longer. This policy, at least implicitly, involves some income redistribution. In a perfect world, at the “extreme,” supporters of this policy goal would want a European style old-age social insurance system.

Interestingly, many policymakers are somewhere in the middle, which can make the process of improving the system more productive, in that there are at least some proposals that have bipartisan support.

And it all Comes Back to the Employer

We also need to observe that a critical focus of retirement savings tax policy, mediating the compromise between these two different policy objectives, is the employer. Under our current system (1) the employer owns the privilege to hand out tax benefits for saving via a retirement plan (Policy Objective 1),17 and (2) the employer must provide a plan that covers a broad cross section of employees (Policy Objective 2).

The second requirement—that whatever plan the employer maintains must benefit a broad group of employees—is responsible for an incredibly complex set of rules in the Tax Code. If all the Tax Code did was provide a tax benefit for savings, you wouldn’t need the employer at all. Those who wanted to save could do so and claim the tax benefit. Those who didn’t, wouldn’t.

The problem (at least for those concerned about Policy Objective 2) is that most of the people who would save in such a simple system would be high-paid employees with extra income available to save, who are in high tax brackets and therefore get a bigger tax benefit from tax qualified saving.

Hence the rule that, if you’re going to have one of these plans, it should benefit a broad cross section of employees.

This compromise—an employer-mediated tradeoff of tax benefits for higher paid employees in exchange for benefits for lower paid employees—is a fundamental principle underlying US retirement savings tax policy.