Unbundling Risk

Lee Anne Fennell - University of Chicago Law School

Posted in , ,

Lee Anne Fennell

When individuals and households select products, services, or endeavors, they are usually making a bundled choice that comes with a certain level of risk exposure or insurance protection built in. Buying a house? You’re also buying a hefty dose of local housing market risk, for better or worse.1 Shopping for a car or a kitchen appliance? You’ll get an insurance policy along with it, and perhaps a lottery ticket as well, courtesy of the tort system2—but you’ll pay for those perks, and they’re not optional. Deciding whether to invest in a cutting-edge career, live the life of an artist, or take a steady job at a large firm? The choice you make will determine not only how you will spend your working days but also your vulnerability to income fluctuations. Similar examples abound. You can modulate your exposure to risk, but often only by changing what you buy or do. Aside from a few varieties of insurance, free-standing or unbundled risk adjustments can be hard to come by.

The menu of risk-customization opportunities available to individuals and households is not only quite limited, but also exhibits puzzling gaps and inconsistencies. Homeowners can easily insure their homes against fire but cannot readily insure against local housing market fluctuations that carry a similar potential to gut their investments. Parents routinely forgo purchasing life insurance for their minor children but cannot avoid carrying implicit insurance on them through the tort system. It is easy for people to hold onto the right to a risky future income stream but often hard to alienate a share of it in exchange for a sum certain. And so on.

These gaps in risk markets have not escaped notice, and academics and entrepreneurs have long been interested in finding creative ways to fill them. The result has been a rich literature advancing new ideas for rearranging risk within situated contexts, as well as a variety of innovative business models. In Unbundling Risk, I seek neither to invent nor to advocate for new risk-shifting devices. Instead, I explore questions surrounding risk unbundling itself, including the optimal amount of stickiness in society’s default risk allocations, the effects of heterogeneity in risk arrangements, and the implications (cognitive and otherwise) of starting from one risk baseline rather than another. The answers implicate both law and policy, given the government’s role in setting the rules for risk rearrangement among private entities and in directly delivering risk protection. As scholarly and entrepreneurial interest in risk innovations intensifies, such questions will become increasingly pressing. Unbundling Risk provides a framework for answering them.

Suppose that when making a purchase or undertaking an activity, people were routinely asked, in essence, “Would you like risk with that?” If everyone’s answer to this question would be identical in a given context, then it might not seem worth asking; the product or endeavor could simply embed the universally preferred amount of risk or insurance. Indeed, in some cases the question might seem absurd—taking on risk is the entire point (think of casino gambling).3 But in many other cases, activities or goods may be consumed not because of the risk profile they present, but in spite of it. Alternatively, otherwise attractive choices may be avoided by some individuals solely because of the risk or insurance with which they are bundled. Significantly, people seem to vary in their tolerance and taste for risk. There can be gains from shifting risk to those persons or entities who are in a better position to bear it, as through diversification, or who can reduce it by undertaking efforts on dimensions that lie under their control.

People are also likely to be heterogeneous in the degree to which they value monetary compensation for certain kinds of nonpecuniary losses, such as the death of a minor child. Because these losses do not typically increase the need for or marginal utility from money, people might prefer shifting money from the state of the world in which the child has died to the one in which the child is still alive. Yet, as commentators have observed, the implicit insurance provided through the tort system does the opposite by effectively extracting premiums from the family when the child is alive and returning a larger amount when the child has died. On this account, the implicit insurance may be utility reducing. Making insurance for such losses wholly unavailable might also be a mistake, however, given the various ways that people use insurance to allocate utility across states of the world.4 For example, some recent research suggests that, at least where objects of sentimental value are concerned, insurance payouts may serve a “consolation” function.5

Against the advantages of customization must be weighed a number of costs. Some risk rearrangements impose costs on other people or on individuals’ own future selves; thus, the law might prohibit people from undoing various forms of social insurance. Other costs may flow directly from heterogeneity in risk arrangements, as where allowing people with private information to elect coverage produces an adverse-selection dynamic. Where markets fail to offer risk customization in the absence of a legal ban, the explanation might be that a given risk rearrangement is simply unsustainable, whether due to moral hazard, adverse selection, or otherwise. Yet existing patterns of risk-rearrangement opportunities and gaps cannot be easily explained by reference to these considerations, raising the possibility that other factors, such as lack of familiarity or framing, may be implicated.

To see this point, it is helpful to first classify the sorts of risk moves that people might make. From a given individual’s perspective, there are four basic possibilities, which I collectively refer to as risk/expected value exchanges (REVEs). To see them, consider the following colloquy between an individual, Ida, and a talking urn that contains ninety-nine black balls and one red ball.

Urn: C’mon, Ida, take a draw. Maybe you’ll get the red ball!

Ida: And if I do?

Urn: Well, that depends on the consequences that are set for the drawing. For example, today’s red ball comes with the following consequence: “$1 million cash.” But yesterday’s was “one broken arm.” And you don’t want to know about last week. For each drawing, there’s a little ticket attached to the red ball, metaphorically speaking.6

Ida: Who attaches these consequences?

Urn: Oh, the law, or nature, or society.

Ida: Can they be changed?

Urn: Yes and no. You can detach monetary consequences or attach monetary consequences, at the right price—at least in theory.

Ida: So for today’s drawing, you say the red ball gets me $1 million. What if I want to make it $2 million?

Urn: You’d need to buy an “upside event-enhancing REVE”—what you might call a gravy gamble. That would cost you $10,000, plus an administrative cut for whoever is selling these things.

Ida: What if I would rather get $10,000 myself, no matter which ball I draw out?

Urn: Then you’re talking about an “upside event-detracting REVE”—a gravy giveup. You’d sell the red ball ticket to someone who pays you its expected value, less an administrative increment.

Ida: So let’s say it was “broken arm day” instead.

Urn: You can’t void the broken arm ticket. But you could add a ticket that would pay your medical costs and let you buy things that would take your mind off your broken arm.

Ida: Insurance.

Urn: If you say so. I’d call it a “downside event-enhancing REVE,” or ditch coverage. It will cost you the expected value of the payment you’d receive, plus an administrative increment. Or, if you’d prefer ditch exposure, you could go with a “downside event-detracting REVE.”

Ida: And what would ditch exposure mean on broken arm day? Another broken arm?

Urn: Well, that’s hard to arrange. But you could get a significant fine to go with your broken arm!

Ida: That sounds unpleasant.

Urn: Well, that’s why you’d get paid for attaching that consequence. You might enjoy the money more when your arm is in good shape than when it is broken, right?

Ida: Maybe so. For now, I need an “upside event-detracting REVE” for that million dollar drawing.

Urn: Yeah, good luck with that.

As the urn’s sarcasm suggests, event-detracting REVEs tend to be rare, even though certain kinds of event-enhancing REVEs—insurance and lotteries—are readily available. It is not possible to explain this difference by saying that society does not want people to be exposed to unremediated losses or “ditches,” because the law often allows people to refrain from purchasing insurance in contexts where the result could be an unremediated loss. Likewise, as the taxonomy above suggests, some event-detracting REVEs simply truncate possible gains, which would not seem especially normatively troubling. Of course, there are many other reasons that particular REVEs might not emerge, or might be the subject of legal bans. But it is at least worth investigating whether the particular mix of risk-rearrangement opportunities that currently exists is more a function of framing and tradition than a reflection of real normative differences.

My point here is about consistency, not about the overall level of risk-rearrangement opportunities that may be optimal. Nonetheless, thinking carefully about the prospects for unbundling risk may reveal places where minor tweaks could alleviate concerns that may have made certain risk transactions seem unworkable. For example, one of the most discussed event-detracting REVEs is the sale of unmatured tort claims—potential future claims by people who have not yet sustained an injury.7 Here, one concern is that people will myopically choose an upfront payment in exchange for their future legal claims when it is not in their best interests to do so.8 But if risk customization is the goal, payment arrangements could be deferred or otherwise structured to present expected value equivalents that do not raise these cognitive risks.

Important in this discussion is the role of government as a risk manager that may also be seeking to accomplish a variety of other goals, from protecting people from cognitive pitfalls to redistributing resources among people who enter the risk pool with different attributes. Recognizing the full complement of possible risk moves may open up new policy opportunities by offering innovative combinations of risk defaults and levels of risk stickiness. If nothing else, awareness of the available risk moves can help to make the background risk conditions more transparent and to reveal them for what they are—socially constructed arrangements that could be otherwise.

In The Lottery in Babylon, Jorge Luis Borges conjures a pervasive societal system of chance-taking orchestrated by an ominous “Company” that delivers favorable and unfavorable outcomes in kind to all the inhabitants of the city.9 The arrangements seem alien, even monstrous. Yet at the end of the story, the reader learns that some inhabitants claim “the Company has never existed and never will,” while others assert that “it makes no difference whether one affirms or denies the reality of the shadowy corporation, because Babylon is nothing but an infinite game of chance.”10 This revelation brings a shock of recognition, prompting the reader to consider the ways in which variance is embedded in life, as well as the role of background institutional and social arrangements in determining how individuals will confront it. In a very different way, my Article hopes to invite a similar set of inquiries into risk baselines and the choices society affords for moving away from them.


Copyright © 2011 Duke Law Journal
Lee Anne Fennell is a Professor at the University of Chicago Law School

This Legal Workshop Editorial is based on the following article:
Unbundling Risk, 60 Duke L.J. 1285 (2011).

  1. See William A. Fischel, Why Are There NIMBYs?, 77 LAND ECON. 144, 146 (2001) (analogizing the purchase of a home to the purchase of undiversified stock in the local housing market).
  2. See Richard Craswell, Deterrence and Damages: The Multiplier Principle and Its Alternatives, 97 MICH. L. REV. 2185, 2230 (1999) (analogizing a punitive damages multiplier to a lottery ticket).
  3. Barbara H. Fried, Ex Ante/Ex Post, 13 J. CONTEMP. LEGAL ISSUES 123, 140, 144 (2003).
  4. See, e.g., Steven P. Croley & Jon D. Hanson, The Nonpecuniary Costs of Accidents: Pain-and-Suffering Damages in Tort Law, 108 HARV. L. REV. 1785, 1832 (1995).
  5. See Christopher K. Hsee & Howard C. Kunreuther, The Affection Effect in Insurance Decisions, 20 J. RISK & UNCERTAINTY 141, 145–46 (2000).
  6. HOWARD KUNREUTHER, RALPH GINSBERG, LOUIS MILLER, PHILIP SAGI, PAUL SLOVIC, BRADLEY BORKAN & NORMAN KATZ, DISASTER INSURANCE PROTECTION: PUBLIC POLICY LESSONS 47 (1978) (describing insurance and similar devices as “tickets that can be cashed in for money if certain states of nature occur”).
  7. See, e.g., Robert Cooter, Towards a Market in Unmatured Tort Claims, 75 VA. L. REV. 383, 383–87 (1989).
  8. Alan Schwartz, Commentary on “Towards a Market in Unmatured Tort Claims”: A Long Way Yet to Go, 75 VA. L. REV. 423, 425 (1989).
  9. Jorge Luis Borges, The Lottery in Babylon, SUR, Jan. 1941, reprinted in COLLECTED FICTIONS 101 (Andrew Hurley trans., 1998).
  10. Id. at 106.

Post a Comment (all fields are required)

You must be logged in to post a comment.