Wednesday, May 26, 2010

Book Review: Nudge

 Nudge: Improving Decisions About Health, Wealth, and Happiness (Revised and Expanded Edition)
- Thaler and Sunstein, 2008, 2009. -- 271 pages

Nudge is a favorite target for many conservatives.  It has gained popularity and notoriety as a widely read book in the Obama White House.  In some conservative circles, the book is portrayed as a "how-to" manual on furtherance of the statist/liberal/Euro-socialist agenda by subtly herding the population into sheep-like policy compliance.  While the conservative criticism is slightly overstated, the authors are dangerously naive about the Pandora's box they have opened.  And let's face it, the idea of being nudged by the government is at least a little creepy prima facie.



Despite my conservative/libertarian heebie-jeebies, the basic premise of the book does make some sense.  Thaler and Sunstein define a "nudge" as a paternalistic suggestion or default-option that does not prevent individuals from exploring self-directed choices ("libertarian paternalism").  Opt-outs or changes from the default must be low-cost in order to qualify as nudges.  Nudges are said to be useful when individuals are facing choices that are rare, complicated, consequential, and removed from immediate feedback .  Furthermore, there are unavoidable circumstances in which there must be some "choice architecture" and default options, so it would be reasonable to carefully consider how the choices and defaults are presented.

The authors begin by cataloging a number of cognitive errors and biases routinely affecting human decision making.  Most of these are in some sense "classic" and were previously known to me, though some I had not heard before.  To get an idea of the kinds of mistakes people make, watch Dan Ariely's presentation on YouTube.

Having established that people are not reliably rational and sometimes need help, the book explores several real-world examples where government or society has established poor "default" options and the authors suggest improvements.  They examine savings, retirement planning, credit, privatization of social security (using the Swedish example), the byzantine Medicare-D structure, and school choice.  They also explore more sensitive topics such as organ donation, environmentalism, medical liability, and marriage (specifically, the "privatization" thereof).

I have divided the topics somewhat differently that the authors did.  Nudge makes its strongest case when examining financial decisions such as saving and investing, and when improving existing programs such as Medicare-D or public schooling.  They have a much weaker case when they stray into the fringe of environmentalism and when they are discussing marriage.

The chapter on "Saving the Planet" is where things start to get a little creepy.  The authors are big proponents of "Cap and Trade" style regulation, particularly as compared to "Command and Control", but also in preference to a straight carbon tax.  Sorry fellas, I'm not buying it.  Carbon Cap and Trade does not have a great history, and the version that passed the House this term was so filled with exemptions and tweaks as to effectively be the "command and control" the authors wish to avoid.  Should we actually determine that carbon is a serious environmental externality (-a currently suspect premise), the fairest way to manage it is with a straight tax on fossil fuels.  Some of their proposed nudges are benign, such as greenhouse gas disclosure for large emitters, and others are a little more Orwellian, like a voluntary Facebook widget that monitors one's greenness and broadcasts it to the world.  They suggest that the Toyota Prius sells better than the Camry hybrid because the Prius is only available as a hybrid, making the Prius logo a much brighter peacock plumage for the greeniacs -- a nudge for others to be green.  I'm more inclined to agree with the South Park flavored interpretation of many Prius drivers.

Nudge's prescription for marriage is "privatization", argued -- at least at first -- on libertarian grounds.  They argue for civil unions for all couples (same and opposite sex), and leaving the marriage business to the religious domain. Had they stopped with the libertarian philosophical argument their case might have been stronger.  Instead, the authors wade into arguments about increasing divorce rates (largely coincident with the adoption of no-fault divorce laws and the availability of birth control), changing norms about sexual conduct and procreation, the social value of state-sanctioned "marriage" per se, and historical state discrimination against interracial couples.  In these latter arguments, the authors are in way over their heads, and unwittingly trip virtually every conservative talking point about decaying social norms and the value of marriage.  After reading Nudge, I am actually less inclined to support marriage "privatization" than I had been previously.  Eliminating state-sanctioned marriage is likely to be another nudge (or anti-nudge) towards suboptimal outcomes of transient serial-monogamy and ever more pervasive single parenthood.  Seemingly as an afterthought, the authors offer suggested "nudges" of streamlining divorce proceedings through a set of legal defaults akin to criminal sentencing guidelines, though this has virtually nothing to do with their central point about civil unions.

To the extent that certain "choice architecture" is unavoidable, it certainly does behoove us to construct the choices and defaults in a logical and helpful manner.  "Neutrality" is often not possible.  Conservatives and hard-libertarians would do well to accept this, if perhaps not embrace it as gleefully as liberals have.  But this is where conservatives should get off the bus.  If it may sometimes be true that conservatives and libertarians err on the side of trusting private interactions and institutions, Thaler and Sunstein fall into the normal liberal trap of trusting the government too much.  Nudge is a toybox to liberals and Pandora's box to conservatives.  And like Pandora's box, the contents have escaped permanently, so we'd better get used to the idea.

The authors are overly dismissive of the "Slippery Slope" objection.  "Slippery-slope arguments are most convincing when it is not possible to distinguish the proposed course of action from abhorrent, unacceptable, or scary courses of action."  Sounds like the original Obama/Pelosi health care "public option" to me, specifically designed to fail-unsafe into single payer.  The authors cling to their principles of not eliminating choices, but here's a real-life proposal (not discussed in the book, as it emerged post-publication) that would harness market forces to do just that without explicitly eliminating options.  Hey, you can keep your plan... so long as your employer continues to offer it...

In fighting against inaction biases, Nudge seems almost contemptuous of social norms and the status quo.

"Those who make [the slippery slope argument] sometimes speak as if government can be absent -- as if the default terms that set the background come from nature or from the sky.  This is a big mistake.  To be sure, the default terms that now apply in any particular context might be best, in the sense that they promote people's interests overall or on balance.  But that view must be defended, not assumed." --pp. 241.

Notice the shifted burden of proof.  Given the often dismal performance of large scale social experimentation, I proudly and bitterly cling to my null theses, and expect any proponent of a new thesis to meet the burden of proof.  With some empirical historical justification, I embrace the bias of inaction with eyes wide open.  Nudge briefly examines a Burkean traditionalist view (in a footnote), but dismisses the concern.

However, my biggest gripe is with Nudge's response to the idea of bad or even evil nudges.  It is here where the authors' misplaced faith in Our Betters shines through. Consider the following:

"On the face of it, it is odd to say that the public architects are always more dangerous than the private ones. After all, managers in the public sector have to answer to voters, and managers in the private sector have as their mandate the job of maximizing profits and share prices, not consumer welfare." -- pp. 242.

This borders on the absurd.  There are certainly instances where private industry promotes self-serving nudges rather than benevolent ones.  But private companies are not only subject to market forces, but to regulation and oversight by government as well if they step too far out of line.  "Managers in the public sector" are often career bureaucrats who are rarely fired in cases other than extreme negligence or malfeasance.  As to elected officials, I hold up as recent counterexamples two individuals indicted by the Pennsylvania Attorney General in a massive multi-year investigation into misappropriation of state funds, John Perzel (R) and Bill DeWeese (D) who were recently re-elected by their parties to stand in the general election.  You just can't get these guys out of office that easily.  Jeff Skilling went to jail.  Barney Frank is still a Congressman.  Congress has done much more financial damage to the United States than Enron ever could have.

Nudge, pp 251:

"If the Nudgers are incompetent, then they could easily do more harm than good by directing people's choices.  And if the risk of self-dealing is high, then it is right to be wary of attempts to nudge. [...] But for those with less pessimistic views about government, who think politicians and bureaucrats are just Humans, not much more likely to be stupid or dishonest than (say) business executives, lawyers, or economists, we can ask whether a situation contains special risk of self-dealing.

It's striking to me that Nudge focuses on self-dealing so much while downplaying incompetence and totally ignoring ideological fervor.  Thanks to Penn and Teller, we know that recycling is bullshit, yet we are still encouraged to recycle.  The EPA Energy Star nudge has been proven to be problematic.  The Compact Fluorescent Bulb nudge is suspect given the toxic materials used in their construction, quality control issues shortening lifespan, and non-universal applicability, never mind the eventual (non-nudge, slippery slope) ban of conventional technology.  In the nudge to get people to eat less animal fat, even more harmful trans-fat substances like margarine were promoted.  The food pyramid, "helpfully" staring at us from our cereal boxes, promotes a questionable dietary reliance on grains that may be making us fatter.  The authors applaud a nudge of high-schoolers towards college, yet many would argue that too many people attend college as it is.  And don't forget the multi-decade, multi-faceted nudges (and non-nudge policies) towards home ownership that played a roll in inflating the housing bubble of recent years.

Where is the check on these evidently failed nudges?  The ballot box?  Get real.

Thaler and Sunstein suggest their libertarian escape hatch as the check on bad nudges:

"And by arguing for a libertarian check on bad plans, we hope to crate a strong safeguard against ill-considered or ill-motivated plans.  To the extent that individual self-interest is a healthy check on planners, freedom of choice is an important corrective." -- pp. 243-4.

While I am grateful for the libertarian grace afforded here, this cop-out is contradicted by most of the rest of their book.  They spent two and a half hundred pages arguing that people are persuadable by social norms, frequently accept defaults, and are biased towards the status quo.  Why should we expect a bad nudge to be significantly less effective than a good one?

Nudges are here to stay, but it worries me that they have been welcomed so readily by some of Our Betters who in too many cases can not be trusted to know what a good nudge is, nor are they capable of stopping themselves at mere nudges.

2 comments:

Samay said...

There are plenty of conservative "nudges" out there too - taboos on drug and alcohol use are the big ones, especially the way the age 21 to drink law is implemented.

I think a nudge isn't perfect, but it's definitely preferable to a mandate.

JoeCollins said...

To be sure, nudges are preferable to mandates. And take your examples of drugs and alcohol -- policy makers went beyond nudges. Tying federal high way funds to the drinking age would't qualify under the authors' definition of a nudge because the opt-out cost isn't insignificant.

Yeah, nudges are here -- perhaps even necessary, but I'm not overly excited about the idea.