Thursday, January 14, 2010

New Paternalism on the Slippery Slopes, Part 8: Hyperbolic Discounting in Public Policy

As discussed in a previous post in this series, the new paternalists often use the concept of hyperbolic discounting (roughly, excessive impatience) to show that people make systematic errors that could, in principle, be corrected by government intervention. But what if policymakers, too, are prone to hyperbolic discounting? That is the question raised in the next section of the paper (p. 724-725):
Policymakers can have short time horizons for various reasons. They might no longer hold office when future costs and benefits of their policies occur. Insofar as voters have imperfect memories, they might fail to fault policymakers for the ill effects (or credit them with the good effects) of policies they supported. Both of these effects give fully rational policymakers an incentive to discount future consequences.

If policymakers are hyperbolic discounters, there is yet another reason they will tend to discount the future: because they apply especially high rates of discount when some costs or benefits are in the present (or near future).
If so, then just as regular people may succumb to temptations like desserts and cigarettes that promise short-term pleasures, we should expect policymakers to succumb to “policy temptations” that generate short-term political gains. For instance, they might be tempted in election years to adopt policies, such as fiscal stimulus bills and trade restrictions, that will improve their electoral chances while pushing costs into the future.

How does this worsen slippery-slope risks? Slippery-slope events are necessarily sequences that play out over time: policy A’s adoption now leads to policy B’s adoption later, leading to policy C’s adoption yet further in the future. Hyperbolic discounting implies that when policymakers are faced with a policy proposal that is appealing in the present, but which creates a danger of bad policies being adopted further down the line, they will be inclined to focus on the former at the expense of the latter. In short, they will be less cognizant of slippery-slope risks.

For instance, policymakers might be tempted to create a small fat tax on grounds that it will induce marginally “better” eating decisions. Opponents might argue that adopting a small fat tax will create a danger of a larger fat tax in the future, as future policymakers—having already incurred the costs of creating a tax collection mechanism—see the opportunity to increase their tax revenues and fund special-interest constituencies. If they are hyperbolic discounters, the policymakers will not take this risk seriously enough, even if they recognize it as real.

Like hyperbolic discounters in the private sector, policymakers should be expected to exhibit time inconsistency: the tendency to make commitments and promises and then break them when the moment of choice arrives. They might, for instance, repeatedly express a willingness to take measures to fight budget deficits in the future, while nevertheless passing bloated budgets and incurring large debts in the present. Note that critics of slippery-slope arguments will sometimes claim to be able to resist the urge to adopt bad policies in the future. The idea is that we can do the right thing today and resist doing the wrong thing tomorrow. They might, for instance, promise to keep fat taxes relatively low (and linked to scientific evidence about the extent of present-bias). The existence of time inconsistency bears directly on the plausibility of promises to do the right thing in the future even in the face of temptation.
So how can the ill effects of hyperbolic discounting in government be resisted? (p. 725-726):
We have also argued that people afflicted by excessive impatience have various self-debiasing mechanisms at their disposal, such as imposing internal rewards and punishments, structuring their external environment, and enlisting the help of third parties (like families and support groups). Policymakers may have access to similar devices. We suggest that the analogous devices in the policy arena usually take the form of institutional constraints, such as judicial review and constitutional limitations on what areas can be regulated by government. The greater need for external restraints follows from the fact that bad self-governance by a single person primarily affects that person, whereas bad governance by policymakers affects all of those governed. Thus, the individual has a rational incentive to rein in his own irrational impulses, whereas a policymaker’s incentive to do so is attenuated. In other words, policymakers are more likely to exhibit “rational irrationality” about matters of personal choice than are the private citizens who make those choices.
In short, if the new paternalists take the problem of hyperbolic discounting seriously, we ought to hear them arguing for stronger limits on the power of government. Thus far, I haven’t heard them doing so.

(As usual, full citations are available in the full paper. Cross-posted at ThinkMarkets.)


Doc Merlin said...

No, No No!
It annoys me to no end when people say that hyperbolic discounting is irrational. If you allow for liquidity risk, hyperbolic discounting is very rational.

Glen Whitman said...

Doc Merlin -- You may be correct. But here, we are taking the paternalists' position on hyperbolic discounting at face value. They say it's a problem for private citizens. We're saying, "Okay, suppose you're right. What about policymakers, then?"

Anonymous said...

No more rational than a free market theorist... or not allowing judges to friend lawyers on facebook:

what is the world coming to?

Anonymous said...

Why is abortion never mentioned during a discussion of paternalism? Forcing a person to carry a zygote to full term seems about the most paternalistic thing a society can do.

How do we stop a pregnant woman from smoking, drinking, and doing drugs doing inevitable harm to her offspring? Isn't this a proper place for a paternalistic intervention by society/government to protect the un/future born?

Where is the proper outrage?

Glen Whitman said...

Anon -- yes, you're right. People just aren't outraged enough about the abortion issue. ;)

But seriously, the answer is pretty simple. Paternalism means trying to change people's behavior for their own good. People who oppose abortion do so because they believe the fetus is a person with rights. Thus, they are not making a paternalistic argument; they are making an argument about harm to another person.

However, it's true that occasionally pro-lifers will bolster their position with a paternalistic argument -- specifically, the claim that women who have abortions suffer bad consequences (like depression) later. That argument would be subject to some of my rebuttals.