What this covers
Richard Thaler and Russ Roberts debate the merits of using behavioral economics to improve people's choices through careful design of default rules. Thaler, from the University of Chicago, argues for libertarian paternalism—the idea that policymakers and choice architects can ethically guide decisions without restricting freedom, since someone must set defaults anyway and designing them to serve people's own interests is both unavoidable and desirable. Roberts, the host, accepts that individuals make imperfect choices but challenges whether government has either the knowledge or the incentive to identify and serve citizens' true self-interest better than individuals and markets can on their own.
The conversation moves through concrete examples that illustrate the default effect's power. Thaler presents evidence from Swedish pension privatization, where workers given freedom to choose invested poorly—incurring high fees, chasing past performance, and losing 40% in early years—while defaulting to a single fund produced far better results. He also discusses organ donation systems, where switching from opt-in to opt-out (presumed consent) dramatically raises participation across European countries, and 401(k) auto-enrollment, which increases plan participation by up to 40% despite the trivial cost of filling out a form. Roberts counters that heterogeneity among individuals makes one-size-fits-all defaults potentially harmful to those whose preferences differ, and that relying on government officials to align incentives with citizens' welfare is unrealistic. He argues instead for expanding choice and constraining government's role, characterizing attempts to make government "benevolent" as futile given the institutional incentives officials actually face. The discussion also touches on whether segmented defaults—target-date funds by age or risk-preference tiers—can adequately capture the diversity of individual circumstances.
Thaler argues that because anyone designing default rules must choose some arrangement, designing those defaults to make people better off as judged by themselves (libertarian paternalism) is unavoidable and desirable; Roberts counters that government lacks both the knowledge and incentive to identify citizens' self-interest, so expanding choice and getting government out is preferable.
- Choice architects (cafeteria managers, plan sponsors, governments) cannot avoid picking some default, and there is no neutral arrangement
- Defaults and choice architecture demonstrably change behavior (401k auto-enrollment, Swedish pensions, organ donation opt-out)
- Roberts disputes that third parties—especially government—can know or serve individual self-interest better than individuals and markets
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
When Swedes chose funds on their own rather than taking the default, their portfolios had much higher fees, were more poorly diversified, exhibited home bias (half invested in Swedish stocks despite Sweden being ~1% of world product) and heavy tech concentration chasing past performance, and the average self-directing investor lost 40% in the first three years.
“the investments people made on their own had much higher fees much more poorly diversified half of the money they invested was in Swedish stocks this we call in finance the home bias... the average investor lost 40% of their money in the first three years”
Switching the 401k default from opt-in to automatic enrollment (opt-out) produces large increases in participation—sometimes as much as a 40% increase—contradicting standard economic predictions that the trivial cost of filling out a form would make the default irrelevant and that companies would already have adopted the better arrangement.
“if you switch the default and make the default that... unless you fill out this form we're going to enroll you... the automatic enrollment has huge effect sometimes isn't it much as a 40% increase in enrollments”
There are economies of scale in financial decision-making: most workers know almost nothing about investing—many believe their own company's stock is safer than a mutual fund when virtually every economist regards concentrating in your employer's stock as the worst choice—so a more sophisticated party can offer sensible advice and a well-chosen default fund benefits unsophisticated investors.
“most of the workers know approximately nothing about investing surveys reveal that they think that there's the company they work for stock is safer than a mutual fund where I think virtually every economist on earth thinks that investing in your own company is the dumbest thing you can do”
The heterogeneity objection is not fatal because smart defaults can be segmented: target-date funds keyed to age and lifestyle funds keyed to risk preference (low/medium/high) capture about as much insight as 90-95% of participants have about their own preferences.
“the most common new defaults are what's called time dated so there you pick a fund that's based on your age... lifestyle funds that you pick a fund based on your risk preferences... that's probably as much insight as 90% of the participants have about their own preferences”
Thaler's only real complaint about the Swedish design is that the government ran a large advertising campaign discouraging people from electing the default ('it's your Swedish duty to choose for yourself'); had it instead gently recommended the default, far better outcomes would have resulted, as shown by newer young workers—over 90% of whom now elect the default after the campaign stopped.
“My only real complaint is that they discouraged people from electing the default... over ninety percent are electing the default fund now... the advertising campaign trying to convince people to choose for themselves has stopped”
More investment options cause harm: research shows the more options a plan offers, the longer people take to sign up (paralysis) and the more likely they are to put all their money into a money market account because they are flustered—so the private sector is now retrenching toward fewer options and better defaults.
“there's research that the more options there are the longer it takes for people to get around to signing up because there's some paralysis and the more options there are the more people that elect to put all their money into a money market account because they're so flustered”
Opt-out (presumed consent) organ donation systems in many European countries produce dramatically higher donor consent—over 90%—than the U.S. opt-in system, and this difference is institutional rather than cultural, with Spain being especially effective at harvesting organs.
“many more people agree to have their organs harvested under an opt-out system than an opt-in system it's a dramatic difference... more than 90 percent of people agree under an opt-out system”
Any policymaker or choice architect who sets rules and operating procedures must choose some arrangement; there is no neutral or choice-free option, so the cafeteria director who arranges food cannot avoid influencing what people eat.
“anyone who has to just determine rules and operating procedures have to choose something and there's no way around that”
Because most people never rebalance and rarely change their plan, a small initial time investment in a financial decision can lock in choices with profound effects—hundreds of thousands of dollars in cost—so the amount of attention at the initial decision point matters more than later attention.
“we know that the typical person makes... most people make no changes to their plan rebalancing is extremely rare and so the fact that they made not much of a time investment initially can have profound effects hundreds of thousands of dollars in cost”
A survey of Cornell faculty found the typical person spent less than half an hour choosing their asset allocation for a pension plan worth hundreds of thousands of dollars, and the most frequently consulted adviser was the clerk receiving the enrollment forms—evidence that despite incentives, people do not seek expert financial advice.
“the typical person spent less than half an hour the most frequent person they asked advice from turned out to be the clerk who was in charge of receiving the forms”
A provision in the new pension bill provides 'safe harbor' protecting companies from lawsuits if they default employees into investments that subsequently lose value, removing the legal disincentive that had previously made firms afraid to offer strong default options.
“it has made it easier for companies to offer good default plans by giving safe harbor so that they don't get sued by somebody if they default them into something that goes down”
Thaler argues that asking 'what should government do?' is legitimate because libertarian paternalism is non-coercive and there is no alternative to it except 'inept neglect'—choosing rules at random or in ways that make people worse off—so given that government must set rules anyway, it should think systematically about which rules make society better off.
“what we argue at the end of our paper is that we see no alternative to libertarian paternalism unless it's what we call... inept neglect be the random or whatever”
Libertarian paternalism is not an oxymoron: paternalism means attempting to make people better off as judged by themselves, and libertarian means a policy that does not restrict anyone's freedom to choose, so a non-coercive policy aimed at people's own welfare combines both.
“by paternalism all we mean is attempting to make people better off as judged by themselves... libertarian... we just mean a policy that does not restrict anyone's freedom to choose”
For organ donation Thaler favors 'required choice'—forcing each person to check one of two boxes when getting a driver's license—over opt-out, because the issue is emotionally charged and required choice would be more politically acceptable while still protecting everyone's rights and likely producing more organs.
“the proposal I would favor is when you get a driver's license you must check one of those two boxes... because this is a emotionally charged issue and I think that that system might be politically more acceptable”
The Swedish Social Security privatization (year 2000) was a pro-choice, free-market design—a 2.5% payroll tax directed into individual accounts, any fund meeting fiduciary standards allowed in, funds free to charge and advertise—and grew to 456 (later 600+) funds, demonstrating an extreme version of choice that a free-market advocate would approve.
“they adopted an approach that I would think you would greatly approve of because it was really a pro-choice free-market approach people were allowed to direct their own portfolios and any fund that met certain fiduciary responsibilities was allowed in”
Economic nationalism fears are historically self-contradictory and usually a disguise for special interests: in the early 1990s Americans feared foreign (Japanese) firms would relegate them to low-end call-center jobs, while today Americans fear losing those same call-center jobs to foreign firms, showing the rhetoric shifts to whatever scares people; in general it is good for assets and people to be employed in their highest-valued use.
“in the early 1990s Americans were worried about getting stuck with call center jobs because of foreign owned companies and today Americans are worried about losing call center jobs to foreign based companies... economic nationalism is usually an excuse for advancing special interest cloaked in the disguise of patriotism”
Roberts' objection is not the non-coercion of libertarian paternalism but that he lacks faith both in government officials' interest in serving citizens' self-interest and in their knowledge of that self-interest, so he sees no reason to push the idea of empowering government to choose defaults.
“since I don't have much faith in their interest in serving my self-interest so what and I don't have much faith in their knowledge of my self-interest why would I want to push that idea”
Because people are heterogeneous, a single one-size-fits-all default that is good for one type of person may be bad for another, and individuals have far more information about their own situation than any HR manager, government employee, or agency head could have.
“people are very heterogeneous... the default that's good for one type of person might not be the ideal default for another people have a lot more information about their own situation than any corporate manager you know HR manager or government employee”
Roberts argues it is futile to 'proselytize government' to act benevolently because acting in citizens' interest is not in government's choice set or toolkit—'like encouraging pigs to fly'—since officials face constraints and incentives that are not aligned with citizens' welfare; instead one should proselytize fellow citizens to make government smaller.
“I wouldn't want to proselytize my friends to encourage government to be kind to us or good to us because that's like encouraging pigs to fly it's not in their choice set it's not in their toolkit”
Roberts argues there is no unified 'interest of society' or 'we' because people are all different, so claims that government should act 'benevolently' or 'in our self-interest' rest on an undefinable collective good—this being where he and Thaler fundamentally part company.
“the problem is better off is hard to define there is no interest of society we are all different... when we say things like government should act benevolently... there's no we”
The order in which food is placed in a cafeteria changes consumption: placing ice cream first leads people to take ice cream, while placing fruit first leads them to take fruit, even though regulars know all options are available.
“if you put the ice cream first and then the fruit people are more likely to take ice cream... whereas if they put the fruit first... people are more likely to take the fruit”
A good pension plan should have a default investment fund chosen to approximate what employees would pick if they were sophisticated (had a PhD in finance), helping the unsophisticated; whether this is called paternalism or self-interest is merely arguing about words since both aim at the same outcome.
“a good pension plan will have a default investment fund that is chosen to help the unsophisticated investors... to pick a fund that the employees... is our best guess as to what they would pick if they went out and got a PhD in finance”
Half of our term is libertarian: the goal is not government making decisions for people but helping unsophisticated people make better decisions without restricting the freedom of those who are informed (or mistakenly believe they are informed) to choose for themselves.
“half of our term is libertarian... we're interested in having the government help unsophisticated people make better decisions without restricting the freedom of people who are informed or who are mistakenly think they're informed to choose for themselves”
Standard economic explanations for the default effect—procrastination and transaction costs—are unconvincing because the dollar amounts at stake are very large while the transaction costs of enrolling are relatively small.
“the standard arguments that economists would give for that difference emerging of procrastination transaction costs are not very convincing because as you point out these are very large amounts of money in the transactions costs are relatively small”
In a private-sector setting companies have a natural market incentive to arrange cafeterias, benefit plans, and work environments to attract and please employees, so a market solution would already produce employee-favorable defaults and corrupt managers who take supplier bribes would be punished by the marketplace.
“companies set up their cafeterias in a way to attract employees so they have a natural incentive to array their food choices... so why wouldn't that be the default that would emerge from a market solution”
Antitrust policy's benefits do not necessarily follow from competition being good: while competition is good, calling something 'antitrust' does not mean it increases competition, and like all government power it can be captured by the politically powerful to protect themselves from competition.
“just because it's called antitrust doesn't mean it increases competition it can often be used and all government power is prone to this it can often be used by the politically powerful to protect themselves from competition”
The Medicare prescription drug program was a design disaster because it maximized choice—40 to 60 options per state varying by state—and assigned people who failed to choose to a plan randomly rather than via sensible guided defaults, offering no help to seniors trying to identify the best plan.
“the most important thing they were concerned with is maximizing choice so there are forty to sixty options in each state... if you didn't choose for yourself they would put you into what... randomly”
The Swedish default fund was well-designed: low fees of only 16 basis points, well diversified, roughly 90%+ equity (an intentional policy choice because the larger defined-benefit portion served as fixed income), invested via index funds with small hedge fund and venture capital exposure—a fund Thaler personally would have been happy to invest in.
“the fund had very low fees it was well diversified the fees and this fund were only 16 basis points which is so that's 16 percent of one percent and altogether it looked terrific I personally would have been happy to invest in that fund”
Roberts holds that the best organ-allocation system would be a market in which people can buy and sell organs, but acknowledges people are squeamish about this for emotional and philosophical reasons even if convinced it would yield more kidneys and save more lives.
“what I think is the best system which is a market solution where people can buy and sell organs people are squeamish about that for emotional and philosophical reasons even if you could convince them that it'll lead to more kidneys and more lives saved”
In the real world people who lack expertise typically seek advice from doctors, mechanics, friends, relatives, Consumer Reports, and Money magazine, choosing whom to ask for help, so the existence of unsophistication does not establish that a third party must decide for them.
“most of us go out and try to find advice from people who we think at least have a chance of having our self-interested heart those are friends relatives so-called unbiased sources”
There are $20 bills on the sidewalk for a while before somebody figures out how to pick them up.
“there are $20 bills on the sidewalk for a while before somebody figures out how to pick them up”
It is hard to find anyone outside an economics department willing to call themselves libertarian, and there is no greater crime than calling yourself a paternalist, making both terms particularly unpopular.
“it's hard to find anyone outside of an economics department that's willing to call themselves libertarian and there is no greater crime than calling yourself a paternalist”