Why We Need a Pension Revolution

Abstract

A broad consensus exists that workplace pension arrangements around the world are sick and in need of strong medicine. Rather than resurrect the traditional defined-benefit (DB) plan or broaden defined-contribution (DC) plan coverage, this article argues that we move from an “either/or” to an “and/and” mindset to improve global workplace pension coverage, adequacy, and certainty. Pension arrangements can combine the best of DB and DC plans and minimize the impact of their less-attractive features. However, we must also redesign the institutions through which workplace pensions are delivered. The ideal pension-delivery institution is expert, has scale, and acts solely in the best interests of plan participants.

A broad consensus acknowledges that workplace pension arrangements around the world are sick and in need of strong medicine. Pension coverage and adequacy are low, and pension uncertainty is high. The prescription of some pension experts is to resurrect the traditional defined-benefit (DB) plan. Others say broad defined-contribution (DC) plan coverage is the cure. This article argues that we have to move from an “either/or” to an “and/and” mindset if we want to seriously improve global workplace pension coverage, adequacy, and certainty. Integrative thinking about these issues leads to pension arrangements that combine the best of traditional DB and DC plans and minimize the impact of their less-attractive features.

The kernel of “the optimal pension system” (TOPS) lies all the way back in Robert Merton’s seminal 1971 article that laid out the optimal consumption and portfolio rules in a continuous-time model. The basic idea is that informed, rational people strive for smooth and adequate lifetime consumption. Unfortunately, we have learned that the “informed, rational” part of the assumption is inoperative in the real world. So, we need to design a series of “autopilot” mechanisms for enrollment, target pension, implied contribution rate, age-based investment policy, and capital-to-annuity conversion processes. Although these redesign ideas might have sounded radical just a few years ago, they are now being widely discussed as the way to move forward.

However, redesigning the pension formula is only half the cure. We must also redesign the institutional arrangements through which workplace pensions are delivered. The ideal pension-delivery institution is expert, has scale, and acts solely in the best interests of plan participants. Far too few pension funds in the world today can pass this triple test. Consider each in turn: Expert means that the pension institution would have enough internal expertise in pension finance, investments, and administration to be managed expertly “from the inside out” rather than by outside agents. That does not necessarily mean that all functions would be carried out 100 percent internally. It does mean that outsourcing decisions would be based on expert judgment that outsourcing is the cost-effective alternative. A related issue is that organization oversight (or governance) would be carried out by a board with sufficient expertise and experience to expertly perform this critical task.

This kind of expertise cannot be assembled and retained without scale. Pension-delivery institutions must be large enough to operate at low unit costs. So, ideally, the institutions would manage assets in at least the tens of billions of dollars and would serve hundreds of thousands of plan memberships.

At the same time, the institutions would have to pass the governance test: in the best interests of plan participants. This test requires that the ideal pension-delivery institution has an arms-length, co-op legal structure. Peter Drucker observed 30 years ago that such institutions—acting as motivated, expert owners—are ideally equipped to own capitalism’s means of production. Although a few such institutions already exist, many more are needed. Indeed, we need a literal pension revolution.

Editor’s Note: This article is adapted from the introductory chapter to the author’s book Pension Revolution (John Wiley & Sons), which is scheduled for release in January 2007.

Notes

¹ See, for example, the article in this issue by Waring and Siegel (2007), in which the authors argue that DB plans are more efficient at providing retirement income than defined-contribution plans. A recent position paper by the Association of Canadian Pension Management (ACPM 2005) also defends DB plans.

² Strathern’s (2001) Dr. Strangelove’s Game provides easy access to the game theory–based decision-making frameworks developed by, first, Von Neumann and Morgenstern (1944) and, then, Nash.

³ IBM ended up in court defending the formula it proposed using to convert pension credits earned under its DB plan into employee cash balances in its proposed cash-balance plan. The company lost its case in a lower court, although litigation continues through a series of appeals. A recent dispute over the ownership of a pension plan surplus between Monsanto Canada and a group of former employees went all the way to the Supreme Court of Canada. The company lost its case.

⁴ Lead articles in the business sections of the 13 August 2004 and the 31 July 2005 issues of the New York Times provide colorful details of the UAL pensions saga. The corporate finance framework within which this type of behavior becomes perfectly clear has been with us for at least three decades; see Treynor, Regan, and Priest (1976).

⁵ These decisions were made in in camera negotiating sessions between officials of the Ontario Teachers’ Federation and the Government of Ontario. The Ontario Teachers’ Pension Plan (OTPP) itself has been quite transparent in setting out the intergenerational implications of the funding and benefit policy choices facing the two “partners.” For example, plan managers forced the federation and government to deal with the deficit issue by lowering the liability discount rate, thus making the asset shortfall transparent to all. For further detail, see the OTPP 2005 Annual Report (http://www.otpp.com/web/website.nsf/web/ar05_index).

⁶ For a pensions-related perspective from the field of behavioral finance, see Mitchell and Utkus (2004). In this book’s introduction, the authors list six behavioral challenges standing between people and sound pension planning: lack of self-control, lack of firm preferences, inertia/procrastination, choice overload, improper inferences/overconfidence, and loss aversion.

⁷ Jack Bogle, founder of the Vanguard Group, has been making these agency-based conflict-of-interest arguments for many years (see Bogle 2005). In “Beyond Portfolio Theory: The Next Frontier” (Ambachtsheer 2005), I bolstered Bogle’s arguments by invoking the concept of informationally asymmetrical markets championed by Nobel Laureate George A. Akerlof. The bottom line is that the gap in the long-term net return between a well-managed, low-cost pension fund and the pension assets of typical individuals trying to make their way through the mutual fund maze on their own has been, and will likely continue to be, in the area of 5 percentage points a year. Bogle calculated a historical average mutual fund investor return “shortfall” of 6.1 percentage points. This shortfall arises from a combination of the much higher costs incurred by most mutual fund investors and from their tendency to actively move out of funds and markets that have performed poorly (i.e., sell low) and into funds and markets that have performed well (i.e., buy high). In short, if the net return on a well-managed, low-cost pension fund over the next 30 years is 6 percent, a dollar invested in such a fund today will grow to $5.42. With a 5 percentage point a year shortfall, the same dollar in the hands of the typical individual investor will grow to only $1.33, a 75 percent shortfall after 30 years of missed compounding. Investment theory is finally catching up with these very real agency and behavioral issues; see, for example, Cornell and Roll (2005).

⁸ Cui et al. (2005) found that, although the standard DB arrangement “beat” the standard DC arrangement (as measured by a welfare/utility-gain metric they developed), a TOPS-type individual-life-cycle pension arrangement, in turn, “beat” the standard DB arrangement. The intellectual foundations of individual-life-cycle pension arrangements were laid in the 1970s by Merton (1971) and developed by Bodie (2003) in the 1980s and 1990s.

⁹ Governance effectiveness in the pension fund sector has been a research focus for me for decades. See, for example, Ambachtsheer (1994); Ambachtsheer, Capelle, and Scheibelhut (1998); Ambachtsheer and Ezra (1998). These published works, plus yet-to-be published research findings, all support the hypothesis that pension funds that combine economies of scale with good governance will, on average, produce measurable value for their stakeholders. Funds with these characteristics are also more likely to be genuine long-horizon investors and hence the knowledgeable owners of the means of production that Drucker envisioned.

¹⁰ Drucker used TIAA-CREF as a positive role model throughout The Unseen Revolution (1976). In a personal interview three months before his death last year, I asked him whether he had changed his favorable opinion of the organization. He replied, “Faithfully, every month, Doris and I receive our pension check. We are very happy with them.”

¹¹ The Dutch have a long history of being strong savers and rightly consider their sophisticated workplace pension system a source of national pride and comparative economic advantage. Even the Dutch Association of the Self-Employed has a serious pension plan for its members! All pension plans in the Netherlands will soon be regulated on the same basis as banks and insurance companies, with liabilities marked to market and with a “risk-buffer” requirement against the risk of balance sheet mismatch. This requirement is forcing Dutch DB plans to make pension payments vary on the basis of the strength of the pension balance sheet. The next logical step is to bifurcate collective pension balance sheets into individual annuity and “at-risk” account components. These developments are steadily moving Dutch pension plans in the direction of TOPS-type arrangements.

The Australians legislated mandatory participation of the entire workforce in workplace-based pension arrangements in 1992. Employers are required to contribute to plans a minimum 9 percent of an individual’s pay. Over time, the Australian system has evolved into an at-risk, individual, capital-accumulation model that is largely managed by arms-length, expert pension co-ops. Today, 14 years after the start of this pension regime, individual pension accounts are starting to sport six-figure balances. What the Australians have not done thus far is to create autopilot arrangements to move these balances (or at least a part of them) into deferred life annuities. Recognition is growing that this feature is now a matter of urgency. Thus, although they started in a different place from the Dutch, the Australians also are steadily moving in the direction of TOPS-type pension arrangements.

¹² To people with hammers, all problems look like nails. To pension people, apparently all problems can be solved with either DB or DC plans. Those who want to move from an either/or to an and/and mindset should read Kelly (2006). Chapter 10 on using innovation to create sustainable solutions is especially relevant.

Some of these creative strategies were successfully used in a recent workshop, Redesigning the Investment Function, conducted by the Rotman International Centre for Pension Management in June 2006 (see http://www.rotman.utoronto.ca/icpm/workshop06062006.htmfor a summary of the workshop proceedings).

¹³ The U.K. government’s white paper on pension reform, “Security in Retirement: Towards a New Pensions System,” can be accessed at http://www.dwp.gov.uk/pensionsreform/whitepaper.asp.