Glaeser on Paternalism
Tyler instructs all good men to shout Ed Glaeser’s new NBER working paper on “Paternalism and Psychology” from the rooftops. I am more than happy to oblige. At a glance, it is clear that this is an important paper. Glaeser stresses a point that Vernon Smith often makes. Sure, folks ain’t hyper-rational homo economicus. We make systematic errors in reasoning. But in very many cases, if the price of error goes up, the incidence of error goes down. (We should be less obsessed with the fact that people do make errors and more obsessed with designing market institutions that generate the best kinds of error-reducing feedback.) Because consumers face the costs of their decisions in a way government regulators never do, we should not expect goverment officials, who are subject to the same errors in reasoning, to do an especially good job of guiding our behavior through paternalistic policy.
Additionally, and very importantly, Glaeser notes that expertise in soft paternalistic regulation just is skill at manipulating the beliefs and behaviors of the citizens they are supposed to serve. But the purely altruistic regulator is a chimera. As I emphasize at length in my paper on Social Security as a “noble lie”, the US government has in fact used techniques of manipulation precisely to achieve ideological objectives that could not be achieved through free and open democratic deliberation. There is no reason to believe that nouveau paternalists, armed with the latest behaviorialist research, will resist the temptation to manipulate the public for their own ideological purposes. That may be dangerous. But even if it is not, such programs of manipulation amount to an abandonment of republican ideals of self-government, and undercut precisely those aspects of liberal democracy that are thought to confer legitimacy on government action.
Here is what Glaeser says:
Assume that soft paternalism involved a public education campaign to induce people to think more about the future and make people aware that their own rosy scenarios will not necessarily occur. As Benjamin and Laibson suggest, from the point of view of fighting self-control problems, such a campaign might indeed have beneficial results.
But this public education campaign also offers many degrees of freedom that can be used in other, less benign ways. Perhaps the soft paternalism campaign would warn of inflation, and might suggest that other, less careful political leaders (that is, the opposition party) might print money and devalue nominal dollars. Perhaps the soft paternalism campaign might suggest that the stock market might fall, especially if non-business friendly leaders were elected. Perhaps the government might suggest that investing abroad is particularly perilous, given the unreliability of other countries (especially, say, France). All of these messages might be justifiable, but would also be pernicious.
. . . The commotion surrounding this expenditure [paying Armstrong Williams for pro-NCLB columns] should remind us that the ability of incumbents to ensure victory through the powers of office, which include the bully pulpit, is a constant risk in democracy. Advocating soft paternalism is akin to advocating an increased role of the incumbent government as an agent of persuasion. Given how attractive it is to use persuasion for political advantage, an increased investment in soft paternalism seems to carry great risks.
It is increasingly clear to me that paternalism just can’t be squared with a cosmopolitan liberalism. The kind of soft-paternalism Glaeser discusses depends on the assumption that there is a conception of value that is so widely shared that government manipulation meant to make people better off according to that standard would not be suspect, even if government manipulation was OK. But there is no such standard. So, in practice, soft paternalism requires the imposition of a substantive conception of value that some reasonable citizens reject. Additionally, it assumes that there is some way by which government agents could be legitimately authorized to manipulate their fellow citizens.
I think paternalist often have the Fed in the back of their minds–an “independent” bureaucracy manned by “experts.” Public health folks would like to see an “independent” public health bureacracy manned by experts that is authorized to set health policy. The difficulty comes in identifying “experts” according to any kind of neutral metric. The crazy “true cost economics” people, I’m sure, don’t think Ben Bernanke’s really an expert in economic truth. Public health is deeply moralized domain, turning on substantive questions about the value of health relative to competing values. I think we’d all see the problem if someone proposed to establish an “independent” panel of moral experts to set policy for all of us on matters of moral hygiene. (”Experts on an independent government agency today reported that a ‘culture of life’ is necessary for any sustainable moral order . . .”)
I think you can detect in some (surely not all or even most) behavioralist work a drive to establish that there is a need for manipulation, that there is a science that tells us how to do it, and that there are experts who are available to do it for us. Glaeser does us an important service by showing us why, on their own intellectual grounds, we should not trust them to manage our lives for us.
November 28th, 2005 13:43
I was misinterpteted when I talked about bell curves in an earlier discussion. This question, to what degree a recognitino that “bounded rationality” should shape public policy comes much closer to my earlier meaning.
There is a strong correlation between IQ and “Lives in poverty” or “Chronic welfare recipient (mothers)” and most worryingly “Ever incarcerated (men)”.
http://en.wikipedia.org/wiki/Iq#Practical_validity
I gather this paper is about how we should help unfortunates, and whether a “paternalistic” model is appropriate. OK, fine.
But I think the rubber meets the road when we talk about what is the “most workable” way to identify people who need help, and to deliver it.
November 28th, 2005 14:54
Do you consider government providing information or requiring that companies provide it to be paternalistic? Even if Glaeser is 100% right, that would not contradict the view that companies have incentives to mislead or underreport adverse information that government should insist be supplied.
What does paternalism really mean to you? There are key, though quite fluid, distinctions to be made between (1) limiting the range of choices and (2) providing incentives to effect choice. And both of these things can be and are done by government, companies, papa, whoever.
Its superficial to portray the choice as between (a) individuals with good incentives and (b) government that lacks incentives. Paternalistic considerations limit choice at myriad levels, sometimes to the good, sometimes not.
November 29th, 2005 13:57
What about the arguments that current shortsighted defenses of Social Security fail to meet NECESSARY conditions for paternalism precisely because:
(1) current policy is motivated by Congress’s inability to give up Treasury revenues rather than to serve the long-term interest of Americans and,
(2) because SS is (as CATO claims, supposing they’re right) not really in the interest of the shorter-living, higher-proportion-of-payroll-tax-paying, low incomers that its supposed to serve?
In other words, are these not NECESSARY condition for paternalisms:
First, that the policy be motivated by a desire to promote the interest of those over whom paternalism is exercised and,
Second, that the policy actually does serve those interests?
Will, and the article he praises, seem to think that the second condition is not necessary. But (even if that does represent normal usage correctly, which I doubt) the really hard cases of “paternalism” are the ones where it actually DOES benefit the recipients (in ways besides respecting their autonomy).
Will does not seem to think that Democrat defenses of Social Security are intended to benefit those who would not otherwise save for their retirement. Nor does he seem to think that that these policies actually benefit the intended recipients.
Fine. I am sympathetic in my cynicism. But that means the policies non-paternalistic, no?
November 29th, 2005 14:24
You make, as always, a persuasive case, and Glaeser has introduced much-needed caveats — but Thaler Sunstein et al still have the germ of a decent point: there is a panoply of government programs that necessarily set defaults. Perhaps there is an excellent case for abolishing said programs, but given that the programs exist and likely will continue to exist, setting the defaults is a salient question for policymakers. To set them in a way that predictably yields consequences broadly recognized as beneficial, with the opt-outs available at any time, strikes me as a reasonable course of action.
Is empowering the government to provide “retirement security,” etc., fraught with peril? Absolutely. Would I much prefer a Shillerian future, in which economic risks are managed individuals participating in free financial markets? Definitely.
But defaulting workers into 401(k)s under the so-called “SMART” plan is, all told, probably better than not doing so. Disbursing food stamps on a semi-monthly rather than monthly basis has been found to reduce self-reported hunger (because recipients often fail to smooth food consumption over the longer stretch of time). This is about setting a default that yields the best consequences. One way or another, the government will set the default — so why not try to set the best possible default?
Now, I see the risk here — the soft paternalism argument is an entering wedge for the notion that government can in fact be both benevolent and efficacious across a wide range of policy domains.
But does that mean we, for example, use a crappy means of tax collection that generates way more deadweight loss just because it’s less likely to raise a lot of revenue (the Mulligan thesis)? Or do we choose our battles, and try to make the government we have work as well as it can? That’s my gut instinct.
But I have been wrong in the past.
November 29th, 2005 20:20
Wow! Still going on about Social Security!
Coming here is like discovering one of those Japanese soldiers who were still in hiding long after WWII ended.
The Social Security battle is over…the good guys won. Accept it.
As for the paper…it seems like academics want to take over the government’s role as daddy of our paternalistic society, not do away with paternalism.
I doubt the average American could name a single living social scientist. If you guys want to take over the family, come up with something other than poorly researched papers that you share amongst yourselves.