Always thought the 'nudges' often referred to in these experiments would only work if participants were unaware of their existence. If participants become aware of the fact that they're being 'nudged' one way or the other, then their reaction will depend on whether or not they have issues with authority, whether they value their freedom enough to fight for their right to have an independent opinion free from manipulation by other actors, and probably a few other factors.
Many of the successful nudges are things that people don't _want_ to resist, it's just taking frictions away, e.g., organ donation, saving in your 401k, registering to vote, not eating garbage, etc. Stuff people would generally express interest in doing.
I think there might be varying levels of nudges. But yes you're right, people are nudged to do the things they want to as well. Maybe the resentment comes from people being nudged to do what they dont want to e.g. give your credit card information for free trials is almost a universally hated nudge
Most normal people can be told they are being manipulated and it will still work. The percentage that will actually hold the thought that they should resist in their mind without losing focus is very small, in my experience.
The problem with the piranha analogy is you can apply the large number of possible causes and effects which occur in different directions to pretty much any science, social or otherwise. There are at any one time a very large number of pathogens competing for the attention of the immune system all associated with overlapping symptoms and thus many competing hypotheses about the cause of a particular ailment, but this doesn't make the study of medicine futile.
Even if you've got competing parties actively trying to nudge people in a particular direction, it doesn't mean there is no discernible effect; increases in political polarisation and emotional attachment to brands can be observed even as people consciously attempt to nudge other people in broadly opposite directions.
This doesn't mean the social psychology and behavioural economics experiments hypotheses are necessarily good ones or the experiments well constructed. But complex systems and competing hypotheses are present in all science and are what controlled experiments and statistical controls are for...
I guess the point is that if there are a large number of big, variable, competing effects with interactions etc, you can't just look at each of them in isolation and expect them to predict a large fraction of the variance. So you need a more sophisticated (possily impossibly complex) model or study, as you said.
In the nudging example, this appears to imply something like "conditioned on a large effect having been measured using a sufficient amount of data (!), you have to be extremely confident in the study design and model to actually believe it"?
There's a related criticism made by some economists about psychology: the 1-800 critique.[1] The pattern is that psychologists claim that effect X exists, e.g. "hostile media bias" or "social identity" or "priming". But they don't say when it will apply, i.e. in what circumstances. So the critique was, these guys need to put a 1-800 number on their papers, so you can call them and ask, hey will your effect apply in my situation?
The ridiculous thing is the researchers not only don't know but they don't care because they never controlled for those factors and never even bothered to replicate their own experiment in different situations to explore the scope of applicability. Nor even replicate it in the same situation. It's as if they're trying to hide the truth from themselves because really they know it's a fraud and they'll only end up exposing the nothingness of their finding if they try too hard.
Somebody's reaction to the color blue is probably socially constructed, like the author suggests. There's several generations of popular culture in the US that reinforce the idea that 'blue equals sad.'
But someone's reaction to a smile is on a more instinctual level. I'm pretty sure that smiles carry a universal meaning for people going as young as infancy. So, it doesn't seem misguided to look for these effects that happen outside of a particular cultural context.
> Somebody's reaction to the color blue is probably socially constructed, like the author suggests.
On some level this is obviously true. The important question is on what level? Blue being male coded, socially construed. Blue being perceived as colder probably not, which leans me towards sadness association being real. Something to test.
On a related note I listened to an Agnes Callard podcast we here she used small girls as examples of experts, on prettiness, as that’s the thing they care about disproportionately much. Thus the fact that they agree pink and purple are the prettiest colors is evidence this is in some real way true. I thought that was interesting as my son’s favorite colors at the time were purple and mauve. He’s now moved on to pink. I’m persuaded that something like an objective ranking of color pleasantness exists. Obviously you can self-modify to like things, like architects self-modify to find Brutalism and other crimes a against beauty good, or how people grow to like coffee.
> On some level this is obviously true. The important question is on what level?
I don't disagree. In my post I was primarily thinking about the author's assertion that these instinctual reactions are not verifiable, which I disagree with. We probably have more in common here than you realize.
One thing it sounds like Callard didn't cover in her podcast was the movement toward instinctual and social reactions. It may be possible that pink has some intrinsic beauty, but how many of the subjects only thought to provide that opinion after being influenced in some almost imperceptible way by their peers?
One thing the pandemic showed us is that people can relatively easily be made to believe something without sufficient evidence if that belief gives some benefit for them personally (i.e. anti-vaxxers). Given that most news today is consumed only through the headlines of news articles, its stands to reason that even briefer pieces of information could have just as large an influence. This is mostly why I take issue with what seems to be the overall goal of the article.
> Obviously you can self-modify to like things, like architects self-modify to find Brutalism and other crimes a against beauty good, or how people grow to like coffee.
This seems a little biased, particularly with beauty and flavor being given as objective properties. Are you sure something subtle on your screen didn't elicit an emotional response before you posted this?
> It may be possible that pink has some intrinsic beauty, but how many of the subjects only thought to provide that opinion after being influenced in some almost imperceptible way by their peers?
No earthly idea. My strong update to it being true is based mostly on a single small child whose close contacts weren’t attempting to push any color preferences.
>> Obviously you can self-modify to like things, like architects self-modify to find Brutalism and other crimes a against beauty good, or how people grow to like coffee.
> This seems a little biased, particularly with beauty and flavor being given as objective properties. Are you sure something subtle on your screen didn't elicit an emotional response before you posted this?
If, the first time a human tastes alcohol or straight black coffee they like them there is something wrong with their taste buds or elsewhere. Alcohol is poisonous. Coffee is bitter, it’s aversive. I’m slightly less confident in “things that are fruit or flower colored” are pleasant to look at, but not much. Brutalism being more pleasant to look at than biophilic, ornamented buildings I consider self-evident, on the same level as “A park is more pleasant to look at than Park Avenue.”
"'small “nudges', often the sorts of things that we might not think would affect us at all, can have big effects on behavior. Thus the claims that elections are decided by college football games and shark attacks..."
In Richard Thaler and Cass Sunstein's book "Nudge," where I believe the term was coined or at least popularized, nudges are almost exclusively presented as the results of "choice architecture," that is the way decisions are presented to people. Of note is that this makes the nudge directly related to the decision itself, and bears resemblance to the butterfly effect only in that both make claims about cause and effect.
To say that behavioral economics is obsessed with trying to hunt down unexpected correlations and publish papers is at best misleading or just plain wrong. I'm sure there are obscure researchers invloved in such pursuits, but it's crazy to characterize the entire discipline this way. He has the caveat "at least how it is presented in the news media." What is that? He should have talked to someone or read something (like Nudge, the book) before ranting. Did he see John Stossel do a segment or something and now he's an expert?
Also of note is that the author is a statistician and political scientist, but his "piranha argument" is a crude attempt at colloquializing analysis that economists (including behavioral economists!) do all the time. Close behind in ubiquity of the economic mantra that "there is no such thing as a free lunch" is that "people respond to incentives." The law of unintended consequences was a term that was popularized by a sociologist, but its roots can be found in the writings of Joseph Schumpeter.
The author indicates at the beginning that the word "nudge" is any contextual influence, irrespective of the choice made. In fact, the "choice architecture" itself could be a "nudge" in this sense.
> To say that behavioral economics is obsessed with ...
That's not what he's saying. It's more like: in defense of some theory, researchers publish findings with an effect that could (according to the literature) be attributed to a gazillion other causes, but they never bother to address that. The interpretation of the data is thus (extremely) limited, or dishonest, and that's disregarding other aspects (such as generalization). They just present the result as support for their own theory. Hence there's a myriad of theories, all of which are almost completely false.
> now he's an expert?
To me, the author pretty much seems to be an expert. "Andrew Gelman is a professor of statistics and political science at Columbia University." Perhaps not an expert in behavioral economics per se, but an grave methodological error in one discipline is also one in any other discipline.
