This collection of papers by Thomas Schelling follows by a year his selection as winner of the Nobel Prize. Its range is similar to his 1983 collection Choice and Consequence, but most papers were written since 1983. Five of the 18 essays were first published in the 1980s, nine were published in the 1990s, and three were published since 2000. Four focus on ecological issues, four on national defense and warfare, four on the issue of "self-command," three on methodological issues, and two on the valuation of human life. An early essay, dating from 1971 and one of the best in the volume, focuses on the dynamics of residential segregation.
A collection of Schelling's works covering so many years and so many subjects makes more apparent than ever the qualities that unite his writings. His economy and clarity of presentation, and his ability to offer memorable examples, are hard to equal. Economists have a deserved reputation for writing too abstractly, with good examples an occasional event at best. In contrast, for Schelling, colorful examples are practically non-stop and support the view that more attention to examples would go a long way toward lessening the divide between economists and the rest of the world.
Nowhere is his power with words clearer than in the essay "Prices as Regulatory Instruments." Over a space of 40 pages appears a streamlined account of the role of the price mechanism in correcting for spillovers. No graph is to be found or equation confronted, and yet one is left with a nuanced sense of spillovers. I found myself second-guessing the extent of my own reliance on graphical demonstrations of these ideas. I also felt convinced that no better introduction to correction for spillovers at the intermediate undergraduate level or above can be found.
Three other essays stand out as particularly powerful. "What do Economists Know" (first published in 1995) presents the most compelling defense of mathematical identities that I've seen; Schelling's main point here is that what is necessarily true is too often equated with what is obviously true. "Social Mechanisms and Social Dynamics" (first published in 1998) more clearly specifies what distinguishes social mechanisms from theories and laws. This is one of only two essays that relies on some mathematics (mainly relegated to an appendix), and Schelling's reader-friendly way of fitting the mathematics into the discussion would do Alfred Marshall (the mathematical minimalist) proud. "Dynamic Models of Segregation" is the second oldest in the collection, dating back to 1971. It reveals, more than any other essay in this volume, Schelling's mind in action, as he manages to derive variations of residential housing patterns that follow from different patterns of preferences that people have about the racial mix of their neighborhoods. This essay, likely more than any other, inspired the title of Schelling's 1978 book Micromotives and Macrobehavior. Pronounced segregation is demonstrated to be consistent with agents not seeking to avoid other races but simply seeking to avoid being themselves a minority group in their neighborhood. Hearing Schelling describe how he carried out his projects, pre-computer, is a testimony to the power of his mind.
With all this said, there are three broad shortcomings to note: one having to do with the translation of essays into book form, one with the extension of commitment strategies from the social to the intra-individual level, and one with a curious gap in the essays.
A more reader-friendly approach, especially for the historically inclined reader, would have included the original date of each essay on its opening page. Instead, one has to dig into the credits at the end of the book to find where the essay first appeared. This format was used in Choices and Consequences, and that book's success might justify such a layout. Nonetheless, I despair of the less careful reader not realizing the age of what is being read. It might be said that classics transcend time, but in Schelling's case time is important. Indeed, Schelling stands out as an economist who has managed to successfully combine highly abstract ideas with historically specific events. To ride the Schelling train of thought without recognizing the particulars that inspired him, and the events to which he saw his work applying, is to be shortchanged.
There are at least a couple of occasions where slight changes made to "update" the original essay have a negative effect. In the essay dealing with the termination of life support, Terry Schiavo's name is added to Karen Ann Quinlan's in the spirit of connecting with the younger reader who may have heard of the former but not the latter. The fact that this article first appeared in 1988 — well before Schiavo's death — can make this misleading. Similarly, the very first sentence in "The Legacy of Hiroshima" announces that "It is more than sixty years since the first — and last — nuclear weapons were used in warfare." It was more than 50, rather than 60 when this article first appeared in 2000. As Schelling's works are historically situated, there are losses in packaging past essays as present essays. For the Hiroshima essay, in particular, this is unfortunate. It was written before recent attempts by Iran and North Korea to go nuclear, and the insightfulness of what appears in this essay only stands out if the reader is aware that it predates these recent events. A republished Gettysburg Address would never be changed to open with "11 score and 11 years ago" even though that is how long it currently is since the event that Lincoln refers to occurred. Knowing when this address was written is important. Out of a similar respect for the historical importance of Schelling's writings, it is easy to make the case that messing with the impression of when they were written should be avoided.
The second shortcoming I perceive has to do with Schelling's writings on internal conflict in essays 5 through 8. Commitment strategies between agents make sense, but to speak of conflicts among partial selves within a greater self (while an interesting metaphor) runs the risk of confusing the matter. As one who has sought to promote the notion of second-order preferences, I have argued [George 1998] that Schelling's way of confronting internal conflict is not able to make normative assessments of the market's influence on tastes. The unreality of the assumption that a person consists of "warring selves" may not matter to a follower of Friedman's famous methodological claim that "assumptions don't matter," but it is particularly odd that one as interested as Schelling in capturing the "lived experience" has also been a promoter of this way of modeling internal conflict.
The third and final notable shortcoming has to do with an omission. Schelling's essays, both those appearing in this book and those appearing elsewhere, carry with them much relevance for controversies over how societies should be organized. Schelling has served as an advisor on the political front throughout his career, but when it comes to raising questions about the spread of market forces he has nothing to say. This is especially hard to understand in his essays on "Commitment as Self Command" (the title he gives to the essays on self-control). The examples of conflicts within the self that he chooses to focus on are dramatic but unrelated to what sort of economic system prevails. The currently sober self having to pre-commit to make impotent the future drunk who aspires to drive, and the pregnant woman who takes measures to disempower her future self who might refuse a natural birth procedure are powerful examples. But I would like to have seen more attention given to the conflicted overeater and the conflicted undersaver, who are becoming more familiar as market forces spread.
These criticisms must not be taken to suggest that this is a book to be avoided. Thomas Schelling manages to enlighten while holding the reader's attention at the same time, a greatly needed gift in our profession. Strategies of Commitment was a pleasure to read, a particularly nice perk for a reviewer.
