Book reviews for The Journal of Pension Economics and Finance reflect reviewers' own views and in no way represent the views of the institution(s) with which they are affiliated.
Employer-sponsored retirement programs – pensions in the American use of the word – are important. More than 75% of American households have claims on future pension income or actually receive such income. Pensions account for almost 20% of household wealth at retirement. While smaller than Social Security and individual savings for old age, the famous stool would tip over without this third leg.
This book presents all what you might want to know about pensions in the US and can be measured by the US Health and Retirement Study (HRS). Alan Gustman and Thomas Steinmeier, in this book reinforced by Nahid Tabatabai, are veritable pension buffs. There is no stone in the pension realm that this researcher duo has not turned over and thoroughly investigated. This book is a collection of their beloved stones; each stone turned to the readers' eyes through the limelight emitted by the HRS.
While some may call this a collection of stones a desert – and the material in this book is indeed rather dry for non-buffs – there are, just like in the Anza-Borrego, wonderful cacti with colorful blossoms. To see them, one had better wear sturdy footwear against thorny variable definitions and thickets of tables. The blossoms are hidden, but between many tables and lengthy, sometimes repetitive explanations are fascinating insights into the behavior of homo not so economicus: failures of long-run planning, procrastination, and an immense lack of knowledge about how pensions work.
The flowers grow, thanks to the richness of the HRS. So, quite appropriately, the introduction starts with a description of this wonderful data set, and Chapters 3 and 4 are devoted to a detailed survey of how the HRS is set up and which kind of pension information can be drawn from it. The panel structure of the HRS exhibits the trends that have revolutionized the pension world during the almost 20 years of HRS data collection from 1992 through 2010 and permits disentangling these trends from behavioral changes over each individual respondent's life-cycle and from cohort to cohort. It should be noted that Alan Gustman has been instrumental in setting up and maintaining the pensions' module of the HRS.
Chapter 2 provides a bit of economic theory that might guide the analysis of the HRS data. While the authors give tribute to the life-cycle theory of saving and consumption smoothing, it is quite clear that the extent of long-run planning failures, procrastination, and lack of knowledge have impressed them to be modest about what theory can predict. In essence, therefore, this book is a detailed descriptive analysis of pensions in the US.
The book is clearly structured: Chapter 5 starts with pension plan participation. The following Chapters 6 through 10 describe plan type, eligibility age, value, and pension incentives (for defined benefit (DB) plans). For each dimension, participants' information as collected in the HRS is juxtaposed to external data (such as the Watson Wyatt payroll data) or plan descriptions provided by the employers of HRS respondents.
We learn that pensions are important because participation is high. The HRS permits a household view, thus accounting for coverage through spouses. We also learn that DB plans are still carrying most of the weight in terms of the benefit volume, although the number of defined contribution (DC) plans has long outpaced that of DB plans. What we learn most from these tabulations, however, is the extent of the lack of information among plan participants about their own pension plan arrangements. Chapter 7 goes into the fascinating detail of what can go wrong. Overall, only about 60% of respondents knew their plan type (DB vs. DC) in the early 1980s, based on data from the Survey of Consumer Finances. In the HRS, 10 years later, the percentage declined to 49%, and further 10 years later to 45%.
Not knowing the difference between DB and DC implies that households have no perception of the risk of their portfolio. People who believe that they have a DB plan but actually have a DC plan are most likely not aware that they have to do something to receive a reliable stream of pension income after retirement through old age. And, somewhat ironic for the authors' super-detailed descriptive analyses: since respondents screw up plan type, they are likely to receive the wrong questions in the HRS survey instrument and thus to give meaningless answers. For my taste, Chapter 7 is the most colorful blossom in this pension Anza-Borrego and impresses by its detail: the authors compared respondents' answers over time, compared them to the employer-provided plan descriptions, listened to interview tapes, and matched answers with payroll data from Watson Wyatt. Moreover, this chapter very nicely weaves economic analysis into survey design and vice versa.
Similar misinformation abounds regarding eligibility age, pension value and incentives for early retirement. The reader might conclude that a large fraction of HRS respondents make decisions completely uninformed about their own institutional background. I am not so sure, however, that this is a correct conclusion. We know that Social Security rules are in many ways as arcane as pension rules. Retrieving information from respondents about how Social Security works has been notoriously devastating. Nonetheless, we have fairly solid evidence that people behave quite rationally with respect to the early-retirement incentives exerted by the US Social Security system, as complicated as they are. Information may flow differently than gathered by surveys. People copy other people, learn from their human resource officers what to do when, and then do exactly the right thing while not being able to phrase this in the terms typically asked in survey questionnaires. The book by Gustman, Steinmeier and Tabatabai thus makes us reflect on the human ability to cope with complicated systems, and the challenge for researchers to better understand through which mechanisms this ability emerges.
The book would have profited from some glimpses over the fence. Its focus is exclusively US, even in terms of semantics – pensions in other English-speaking countries include e.g., public pensions. There is a wide variety of occupational pension set-ups in Europe from which more can be learned about the hypotheses in the theory chapter of the book. Looking over the fence, just to take one example, identifies the lack of deferred annuities in the US as a dimension not mentioned in the book but important to understand the trade-offs between DB and DC plans. Even in the US context, it would have been nice to better understand the interactions between Social Security and employer-provided pensions, especially with the wealth of HRS data.
Every pension specialist must read this book, and even more so designers of surveys that aim to measure pensions. Hopefully, its second edition will add a chapter on international comparisons and the interaction among the three legs of the stool.
