Monitoring post: US debate on possible US Nudge Unit
Rumour has it that White House considers establishing a nudge unit similar to the UK Behavioural Insights Team. Given The Nudging Networks purpose and the work of the iNudgeYou-team this post is updated continuously to follow the US debate as it unfolds. Which issues are raised? How is the debate different from the one in the UK? And how does our work relate to the issues raised?
Subscribe to the blog and we will keep you up to date.
UPDATE 1: Maya Shankar’s request
Not long ago I, along with other scholars, received a request from Maya Shankar to provide input and ideas on how behavioural insights could help the US federal government. The request also asked people interested in working for this unit to forward applications.
In addition, the document mentioned that the federal government is currently creating a new team aimed to build federal capacity to experiment with the same approaches as those used by the UK Behavioural Insights Team. All in all, it was quite the open and sympathetic way of asking science with relevance to society to chip in after the public has paid for the knowledge developed within academia.
Fox News opens the debate (July 30)
It took a while – actually quite some time – before the US media picked up on the story. In a brief piece Gov’t Knows Best? White House creates ‘nudge squad’ to shape behavior (there’s also a video-interview on this link) Fox journalist Maxim Lott broke the story in a reasonably sober article (the only thing that stands out is that he refers to the Behavioural Insights Teams as “UK Nudge Squad”).
In the piece Richard Thaler tells FoxNews.com that the new program sounds good, and is quoted for saying that:
“I don’t know who those people are who would not want such a program, but they must either be misinformed or misguided”.
We have browsed through the debates and comments and have summarized the main objections along with what we believe represents the best responses below.
1) The ‘Gov’t Knows best’ critique?
The piece as well as the video interview mentions some well-known critizisms of nudge. Michael Thomas, an economist at Utah State University, says that
“I am very skeptical of a team promoting nudge policies,… Ultimately, nudging … assumes a small group of people in government know better about choices than the individuals making them.”
Of course, as we argued in the Nudge & The Manipulation of Choice this argument doesn’t just apply to the nudge-approach as such, but is an argument that pertains to all kinds of government activity. Even when government limits itself to supplying information, that information is picked, framed and distributed with certain aims in mind. Every government action needs a purpose, and that purpose is defined by a goal which government has to define as best it can.
However, it has been our experience that unless this is pointed explicitly the the public tends to believe that this point applies solely for nudges and forgets or overlooks that it applies in general to public policy in general.
2) The ‘Unintended consequences’ critique
The same goes for the other critique mentioned in the Maxim Lott interview: that of unintended consequences.
In essence the argument is that the application of nudge inspired policies may lead to unintended consequences, which the policy-maker did not predict.
The philosophers Evan Selinger and Kyle Whyte might disagree with us on this point. But it seems to us that any public policy may have unintended consequences – providing information can lead to apathy or cognitive overload, incentive regulation can shipwreck due to price elasticity and bans lead to black markets and new avenues for crime. History is ripe with examples of policies gone wrong, but naturally it is just as ripe with policies that achieved their aim and made all of us better of. This underscores that behaviour is a complex issue, and in fact, the evidence based trial-and-error approach advocated by the UK Behavioural Insights Team which also guides our work seems much more serious about this fact than most of the traditional policy making.
3) The Nudge-may-turn-to-shove critique
Jerry Ellig, an economist at the Mercatus Center, is quoted for saying that while some “nudges” are reasonable, one should be careful about enabling a slippery slope.
“If you can keep it to a ‘nudge’ maybe it can be beneficial,” he added, “but nudges can turn into shoves pretty quickly.”
Slippery slope arguments are always difficult to engage with because they rely on how the level of imaginary sophistication that the one making them possesses. Still, the “nudges-may-turn-to-shoves” argument seems to rely on a conceptual misunderstanding. A government that uses nudges is not moving towards increased control but away from it. Nudges are softer regulatory instruments than traditional policy measures (excluding the provision of information), since they leave the decision maker free to choose as they please without additional costs. Oddly enough, the public tend to forget that excise taxes, e.g. on cigarettes and gasoline, and bans, e.g. on driving without buckling your seatbelt, are also implemented as means to regulate behaviour. They just do so without providing the decision maker with an option for opting out without paying the additional costs of doing so, as well as being inefficient, expensive to maintain and socially unbalanced (the cost of cigarettes are much higher for the average blue-collar worker than they are for the CEO relative to their respective income).
Regardless, it’s an empirical claim whether nudges leads to a slippery slope (in either direction) and we still haven’t seen any evidence indicating that a slippery slope exists. Feel free to provide an example if you know of one.
Debate over the ‘nudge’ initiative (July 31)
A seemingly well intentioned government initiative is raising some serious questions about Big Brother.
Thus the discussion was opened on Fox News on Saturday, July 31. It seems that ‘nudging’ is a familiar concept to Fox News journalists, but to be fair journalist Shannon Bream does avoids reducing the debate to mere rhetorics quite well.
However, the interview was framed in the traditional “oppositions” model with Kirsten Powers, Political (Democratic) Columnist on the pro-side and Monica Crowley, Fox Contributor on the con-side.
The “oppositions” model often leaves a false impression of two distinct sides in a conflict and tends to polarize a debate far beyond reason, so the interview should be viewed with some caution if is interested in how Americans really feel about the issue. Yet, if one was forced to pick sides, we would definitely stay away from Monica Crowley who seems to know nothing about the issue at hand what she is discussing. We briefly summarizes the main points from the interview below.
4) “Oh, it’s nothing” defense
In general, Kirsten Powers leaned towards a “oh, it’s nothing” defense and downplayed the positive as well as possible negative outcomes of adopting a nudge strategy somewhat.
4.1 It’s just about fine tuning
On the question of whether this isn’t a case of “government knows best” (echoing the original Fox article from Maxim Lott), she answers that this argument never really worked for her and that it is just “efficiencies”.
That may actually not be true, as even efficiency may easily mix with values such as it does in the proposition to adopt an opt-out system for organ donation. Nudges are tools to achieve certain aims, and if the tool is wrong for the job, then that is a valid concern in and of itself and deserves to discussed as such.
Also, as Bryan Caplan mentions in this comment, one could – as he does – prefer a more radical role for a nudge unit. We’ve earlier pointed to the necessity of keeping an eye on issues of countability, and as we wrote earlier in When is a Nudge Enough the nudge-approach is in principle compatible with both libertarian-paternalism, as well as paternalism and libertarianism.
4.2 It’s just about information
Later in Powers adds that nudges are just about “… providing people with information, so that they can make informed decisions”. However, this would obviously be somewhat a mischaracterization. Nudging behaviour isn’t just about providing information, but also about defaults, prompted choice, salience, working with loss aversion and hyperbolic discounting, and much, much more.
4.3 It’s not even started yet
Finally, Powers points to the fact that a nudge-unit has not even been established or started its work yet. Again this doesn’t seem to be the best defense. It’s always important to discuss these issues at an early point in order to avoid running into unsolvable problems later on.
5) Rambling rhetorics
However, compared to her ‘opponent’ Monica Crowley, Powers resembled a textbook case of rationality.
On many philosophy courses in critical thinking and logic there is an exercise called “Bullshit Bingo”. The idea is that you play a video- or podcast clip and ask the students to identify as many fallacies in reasoning and argumentation as possible… and oh boy!
Here’s the first rambling rhetorics from Crowley:
“It seems to me to be really creepy. I’m completely creeped out by this. I think it has an Orwellian ring to it. In fact, the name of it is “Behavioral Insights Team” which comes straight out of George Orwell…. And yet we know that in the best case scenarios some of those programs and bureaucracies has ended up as economic disasters and at worst they’ve been abused. This is partly what the IRS scandal is about. The abuse of power there. So I think there is a very slippery slope going from one thing that sound pretty good on paper to actual to active practice where the potential for abuse and control over the American people really could be exploited. I mean, look, you either believe that we are a nation based on individual freedom, where people have the right to make mistakes and make their own choices, or you don’t. And We’re so down that road now, where something like this seems acceptable.”
One can just start listing the fallacies here. We have appeal to emotion, false attribution, Ignoratio Elenchi… In fact, I will invite readers to identify as many mistakes as you can in the comments.
The basic points are that Crowley somehow confuses a potential American “Behavioral Insights Team” with government in general, points to non-existing examples of inefficiency and seems to have come down with a severe case of ‘bathmophobia’ – an abnormal and persistent fear of steep (slippery) slopes.
In conclusion, we hope that the quality in the public US debate on a possible future US nudge unit increases in the following weeks, and promise to keep updating the blog as the debate evolves.
UPDATE 2: Here come the bloggers (August 1)
There is usually little more depressing – from the point of view of reason – than when the bloggers enters. And to be sure, the wake of Fox News has produced some waves of copy-paste-commentary from people who knows (and who wants to know) little or nothing about the issue (examples here, here and former-superior-court-judge-here).
However, there are also some great exceptions. One such is David Berreby’s piece How to Think About the Federal “Nudge Squad” on the blog Mind Matters. In this David Berreby first gives a nice and balanced criticism of 4) and 5) above, with references to several key players in the discussion. He then goes on to give his version of the “Who knows best critique”, number 1) above, thereby echoing Maxim Lott’s first stab, though much detailed and precise. However, as we also pointed out above, this is actually not a criticism of the nudge-approach as such, but rather of government as such.
6) Democratic accountability
He then provides a critical comment concerning the problem of ‘democratic accountability – and, we believe, this is actually an important point to discuss (and we actually discussed it here: Nudge – when knowledge becomes a Democractic challenge).
Nudges, by definition, are indirect, and often are supposed to be unnoticed. Being threatened with a penalty tells you that the government is out to get you to pay your taxes on time. Being told the neighbors paid theirs may get you to comply sooner, but it doesn’t provide a moment of clarity, in which you can say to yourself, oh, those people, over there, in government, want me to do that.
This is not easy to square with the ideal of democratic accountability. When I am threatened or cajoled, I can see who is cajoling and why—which means I know whom to punish at the ballot box if I don’t like the policy. When I am nudged, I don’t know whom to complain to. I might not even know I have been nudged.
Now, as we point about in the Nudge and the Manipulation of Choice it is neither true in principle or in practice that nudges are often supposed to be unnoticed. This assumption seems more to be connected with our tendency to focusing on the fascinating examples – and what could be more fascinating than influences that go unnoticed. Yet, Berreby is completely right the criticism of democratic accountability is one that should be discussed (and feel free in the comments).
7) How do we know one choice is better?
A final balanced criticism raised by Berreby is actually an epistemic variation of the political “who knows best” criticism. Thus, this criticism asks us how we can ever know that one choice is better than another. Berreby himself points out the limitations of this criticism. Again we add, that this seems to be a general problem of public policy, rather than something peculiar to the nudge-approach.
Is the discussion fading out? (August 2-)
The last couple of days has seen little addition to the discussion of a possible US nudge unit. A few right wing bloggers throw around really weird thoughts about left wing conspiracies, brainwash and a communist takeover. Quite amusing actually.
Thus, the first test of the idea of a US nudge unit seems to have run fairly well. On the negative side a real discussion didn’t occur. Behavioural scientist and other academics have been awfully quiet adding almost nothing of value (Berreby is an exception of course). On the positive side, the lack of discussion is better than a debate conducted on false premises by polemics like Monica Crowley. Perhaps the US is just set and ready for a US nudge-unit. They ought to.
UPDATE 3: Reason chimes in
Disregard the last paragraph, the debate has apparently not slowed down but simply shifted towards a more nuanced and thoughtful level, which, as we know, takes more time than hot-headed blogging and dramatized interviews.
A recent contribution to the “nudge-squad” debate is by the political pundit and soft-core conservative David Brooks who defends the use of nudges in public policy in his New York Times Op-Ed: The Nudge Debate. Another contribution is a commentary in the USNews by research fellow Sherzod Abdukadirow called: When ‘Nudge’ Comes to Shove. Both pieces mirror some of the earlier arguments, mainly the risk of creating a slippery-slope towards more forceful paternalist policies, but they also bring new arguments to the field, both pro and contra, that we believe should be addressed.
8) We are always being nudged
Brooks write that:
The pro-paternalists counter that government is inevitably setting contexts and default positions anyway, so they might as well be aligned with individual and social goals. There’s very little historical evidence that there is an inevitable slippery slope leading from soft paternalism to hard paternalism…
I’d say the anti-paternalists win the debate in theory but the libertarian paternalists win it empirically. In theory, it is possible that gentle nudges will turn into intrusive diktats and the nanny state will drain individual responsibility.
But, in practice, it is hard to feel that my decision-making powers have been weakened because when I got my driver’s license enrolling in organ donation was the default option.
Never mind that the conclusion doesn’t really follow from the premise, the notion that nudges are inevitable because government has to apply settings and contexts to people’s lives that will then influence their behavior and choices, is still an interesting one and deserves, not one, but two replies.
The first reply is that we have to differentiate between intentional and unintentional nudges and how they relate differently to responsibility. If we design a choice architecture with a specific purpose, then we are responsible, at least in part, for the consequences that takes place within that specific context. If we design a policy form with unintended consequences, ie. that we could not have been aware of, then we are responsible (at least in a legal sense) for those consequences. This adds additional responsibility to the choice architect relative to the standard public policy maker who unwittingly and unintentionally nudges us towards poorer outcomes.
Imagine that you had to send a letter to late tax payers. The traditional policy maker is responsible for the legal accuracy of the content, but (s)he cannot be held responsible for a 5pct. reply rate. The choice architect is, on the other hand, responsible for both the content and the response rate because (s)he has the means and insights to increase it.
This can be summed up by stating that nudges are always made intentionally and with a specific aim in mind. Naturally, we can influence others unintentionally but that is a whole different story best left for the philosophers.
The second reply is that even if nudges were inevitable, which they aren’t, that does not a priori mean that they should be used for enhancing welfare. Nudges and the notion that they should be used to enhance welfare typically stems from a getting nudges and libertarian paternalism mixed up. We could use nudges for other things, for instance law professor David Mitchell suggests that they could be used to enhance freedom, not welfare.
9) Irrational Choice Architects
Abdukadirow is more critical towards nudge and a US nudge-unit than Brooks, and points out that policy makers are just as human as those they seek to influence:
If individuals making private choices are irrational, so are the policymakers trying to do the nudging. Policymakers tend to overestimate their ability predict policy outcomes, in what is known as hindsight bias. They may also myopically choose expedient policies over those that provide long-term benefits
This argument is not entirely new, but it is important because it intuitively makes sense that we don’t want irrational people calling the shots for us.
We should recognize that everyone is somewhat equally irrational (or prone to decision errors). You, your neighbor, wife and/or husband, your political representative and the guy doing your taxes. Keith Stanovich, who is a prominent cognitive psychologist, argues that there are individual differences and that they are related to IQ. But, even if this is the case, it still means that politicians and policymakers are more than likely placed in the average IQ pool with the rest of us.
We argue that although this may seem scary, it doesn’t really matter all that much, because decision errors and irrationality are context dependent phenomena that occur when our less-than-perfect cognitive system encounters certain circumstances. There is a world of difference between wanting to loose weight before and after dinner, just as your economic behavior is largely dependent on the availability of tempting offers.
Policymakers make mistakes just as the rest of us. But they are designing and implementing policies in contexts that are vastly different from the ones they are meant to take effect in, and have the option of getting expert advice to help them avoid the worst of their own errors.
Finally, the insight that policy are just as prone to irrationality as the rest of us, is really something that affects all types of policies, whether they are behaviorally informed or not.
10) Democratic accountability 2.0
Abdukadirow has another interesting objection to the use of nudge:
By giving the most vocal opponents of a policy a chance to opt-out, nudges reduce the opponents’ incentives to track the policy and to express their views. Thus, nudges may face less scrutiny than other policies
This is an interesting variant of the “nudges work in the shadow” argument that deserves a closer empirically look. Here and now it suffices to say that choice architects should seek a mandate for implementing nudge policies from the ones they seek to influence. In Nudge and the Manipulation of Choice, we argue that such a mandate is different depending on the transparency and type of the nudge – transparent reflective nudges require few mandates because they are, well, transparent and because the influenced can reconstruct the intention behind the nudge itself. For less transparent and more behavior based nudges, the mandate needs to be stronger. Seeking a mandate for a nudge policy is a critical part of determining what people understand as better outcomes “as judged by themselves”, and is something that is critical for choice architects if they want to stay on the right side of the ethical fence.
We end this update with a little tale of caution when it comes to using empirical evidence as part of your (mainly) moral argument.
There’s very little historical evidence that there is an inevitable slippery slope leading from soft paternalism to hard paternalism.
Finally, nudges can easily turn to shoves.
We leave the reader to decide which of the two seems more to their taste
We will of course update this post, when more happens.
UPDATE 4: The big guns
This update has less to do with arguments on either side of the US “Nudge-Squad” divide but has points to some interesting commentaries from people central to the development of behavior based policy, namely Richard Thaler and Daniel Kahneman.
11) Richard Thaler
In his NYT column ‘Public Policies, Made to Fit People’ , Richard Thaler unsurprisingly applauds the initiative, and points to some of the complicated social issues which the team is said to have a future interest in. These include improvement in the support for victims of domestic violence, improving compliance with medical treatments and making it easier for low-income families to over come the 30 mio. word gap. In his closing statement, Thaler attempts to engage with opponents on the one issue they seem almost intent on getting wrong.
All of these examples show that the role of behavioral science in policy isn’t for the government to tell people how to think or act. It is to help them achieve their own goals. Parents want their children to excel, callers to a victims’ hot line want help, and sick people want to get well. Offering aids is like providing an alarm clock: it may help people get to an appointment on time, but no one is forcing them to use it.
This statement really sums up one of the central tennets of Nudge, but in order to fully appriciate it, one needs to understand the difference between “what to do” and “how to do it”. Choice architects don’t (or should not) deal in “whats” – that is – they do not and should not waste their time speculating about which goals people should achieve. These things are best left to those with the appropriate mandate, which for better or for worse often means politicians of some sort. They do however, deal in “hows”. Every what naturally translates into a how: ‘How do we help those who want to loose weight achieve their goal?’, ‘how do we ensure that those who want a comfortable retirement also gets one’ and ‘how do we make it easy for those who want to cut back on carbon emissions to do so?’.
Confusing the “how” with the “what” has been one of the trademarks of this debate, especially for the opponents of a US behavioral Insights Team. Sadly it is a problem that affects not only low quality news pundits, but reasoned and respected academics as well. Hopefully Thaler’s continued emphasis on the confusion can minimize its impact on the future debate in the US and elsewhere.
12) Daniel Kahneman
Another noteworthy contribution comes from Nobel Laureate Daniel Kahneman, who at a White House conference specifically dealt with how to incorporate behavioral empiricism into the development of sound policy . One of the problems has to do with academic traditions and political habits. Historically, Economics has been the primary contributor to policy development for at least the last hundred years (before that it used to be theology) which has kept public demand for empirical orientated researchers low. But, the supply side has had its problems as well. Those with the right skills, mainly research psychologists has up until now generally avoided venturing into the political world. Kahneman notes that:
“Part of the problem is the nomenclature,” he said. “Psychology students are not attracted to something called ‘behavioral economics.’ ‘Applied behavioral science’ might be a better term.”
There is a rising consciousness about the fact that documenting effects means more than simply pointing to economic models based on assumptions about what hyper-rational individuals would do given x and y. And it seems like psychologists and experimental economists are finally getting recognized as valuable contributors to the ongoing discussion on how to reach our goals most effectively. Hopefully, researchers from these academics fields are ready to take up the torch and put their skills into helping improve decisions about health, wealth and happiness.
By subscribing to the blog you’ll keep yourself up to date