4 ms·

I think it is worth noting that some of the more surprising results in the behavioural economics literature have failed to replicate. A quick search will find m

I think it is worth noting that some of the more surprising results in the behavioural economics literature have failed to replicate. A quick search will find many articles discussing this. I believe there is definitely a strong core in behavioural economics, but the especially "pop science" experiments (e.g. people walk more slowly when primed to think about old age, which IIRC is in "Thinking Fast and Slow") should be taken with a grain of salt.

2T1Qka0rEiPr 6y ago
This is fair criticism from what I've seen too, though for passers by I'd just say that it shouldn't necessarily entirely discredit the field. What particularly interests me, is the "nudge" applications which have been adopted by govts. over the past number of years. Whilst most of their plans miss, the ones that stick are incredible, such as the setting of defaults to drastically increase organ donors and improve pensions and investments.
- Kosirich 6y ago
  Was about to make similar comment about connection with behavioral economics and how we should be wary of certain claims. Also there is some strong criticism of applying nudge theories on complex systems. That being said, as you point out as well, some nudge techniques do lead to beneficial results (for all) and I think it's important to recognize (and econs. point out) that if you design a UX, process, building, gadget ect. you are by definition what they call a choice architect. Side note: I was fascinated how much design/choice/nudge thinking is considered here in Denmark. It's very noticeable in design of public buildings and govt. webservices
- nostromo 6y ago
  I’m not sure I want the government opting me in to programs any more than I want Google opting me in to things that may not be in my best interest. If dark patterns start working their way into government, god help us all.
  - okintheory 6y ago
    Just FYI, govt. dark patterns in the US have been around for a very long time. For example, you could read about the "literacy" tests for voting in Southern states. Robert Caro has a nice description in Master of the Senate, vol. 2 of his LBJ biography. EDIT: (From Caro) Questions on these tests posed to Black people in Alabama have included: "Give the names of all sixty-seven county judges in the state", "State the date when Oklahoma was admitted to the Union", and "How many bubbles there are in a bar of soap".
  - reallydontask 6y ago
    There are positives to the default being opt in, e.g. organ donor. There could be negatives of course, there is no substitute for an informed citizenry and indepedent media to challenge such instances.
  - sokoloff 6y ago
    In setting defaults, you’re always making a choice. “By default, all drivers will not be marked as organ donors.” vs “By default, all drivers will be marked as organ donors.” “By default, all employees at companies with 401(k)s will save X% of their eligible salary in 401(k)s.” You’re making a choice for them in any event.
    - ghaff 6y ago
      I agree with you but we also need to be careful given how powerful defaults are. Even Thaler has, I believe, argued against organ donation being opt out; rather I think he's argued for requiring a yes/no choice when driver's licenses are renewed. (The argument is opt out would lead to next of kin/families arguing that the deceased never made a choice to donate.)
      - function_seven 6y ago
        Ad companies argued this when Microsoft updated IE (or was it Edge?) to set the DNT flag to “1” by default. I think they were right to argue as such, and wonder if MS knew that defaulting DNT to on would poison the well like that. It’s a good argument. When people argue over what a default should be, they often don’t realize that it’s possible to have no default at all, and force an explicit choice.
kashyapc 6y ago
Yes, but more specifically on Thinking, Fast and Slow, as noted elsewhere[1][2]: It is only the priming-related chapter (called 'The Associative Machine' in the book) that put "too much faith in under-powered studies", as Kahneman put it[3]. Not the entire book! The book is a synthesis of forty years of Kahneman's research and his collaboration with his late colleague, Tversky. A wide range of topics are covered; and it still absolutely merits reading. Patiently dive into the book and make up your mind. (But yes, a v2 of this book definitely is worth it, given the "authority" of the Nobel Memorial Prize.) [1] https://news.ycombinator.com/item?id=22054603 https://news.ycombinator.com/item?id=22054603 [2] https://news.ycombinator.com/item?id=21757524 https://news.ycombinator.com/item?id=21757524 [3] https://retractionwatch.com/2017/02/20/placed-much-faith-underpowered-studies-nobel-prize-winner-admits-mistakes/ https://retractionwatch.com/2017/02/20/placed-much-faith-und...
nojito 6y ago
This always gets mentioned and upvoted to the top but there has really only been one priming experiment that has been discredited.

2T1Qka0rEiPr 6y ago

This is fair criticism from what I've seen too, though for passers by I'd just say that it shouldn't necessarily entirely discredit the field. What particularly interests me, is the "nudge" applications which have been adopted by govts. over the past number of years. Whilst most of their plans miss, the ones that stick are incredible, such as the setting of defaults to drastically increase organ donors and improve pensions and investments.

Kosirich 6y ago
Was about to make similar comment about connection with behavioral economics and how we should be wary of certain claims. Also there is some strong criticism of applying nudge theories on complex systems. That being said, as you point out as well, some nudge techniques do lead to beneficial results (for all) and I think it's important to recognize (and econs. point out) that if you design a UX, process, building, gadget ect. you are by definition what they call a choice architect. Side note: I was fascinated how much design/choice/nudge thinking is considered here in Denmark. It's very noticeable in design of public buildings and govt. webservices
nostromo 6y ago
I’m not sure I want the government opting me in to programs any more than I want Google opting me in to things that may not be in my best interest. If dark patterns start working their way into government, god help us all.
- okintheory 6y ago
  Just FYI, govt. dark patterns in the US have been around for a very long time. For example, you could read about the "literacy" tests for voting in Southern states. Robert Caro has a nice description in Master of the Senate, vol. 2 of his LBJ biography. EDIT: (From Caro) Questions on these tests posed to Black people in Alabama have included: "Give the names of all sixty-seven county judges in the state", "State the date when Oklahoma was admitted to the Union", and "How many bubbles there are in a bar of soap".
- reallydontask 6y ago
  There are positives to the default being opt in, e.g. organ donor. There could be negatives of course, there is no substitute for an informed citizenry and indepedent media to challenge such instances.
- sokoloff 6y ago
  In setting defaults, you’re always making a choice. “By default, all drivers will not be marked as organ donors.” vs “By default, all drivers will be marked as organ donors.” “By default, all employees at companies with 401(k)s will save X% of their eligible salary in 401(k)s.” You’re making a choice for them in any event.
  - ghaff 6y ago
    I agree with you but we also need to be careful given how powerful defaults are. Even Thaler has, I believe, argued against organ donation being opt out; rather I think he's argued for requiring a yes/no choice when driver's licenses are renewed. (The argument is opt out would lead to next of kin/families arguing that the deceased never made a choice to donate.)
    - function_seven 6y ago
      Ad companies argued this when Microsoft updated IE (or was it Edge?) to set the DNT flag to “1” by default. I think they were right to argue as such, and wonder if MS knew that defaulting DNT to on would poison the well like that. It’s a good argument. When people argue over what a default should be, they often don’t realize that it’s possible to have no default at all, and force an explicit choice.

kashyapc 6y ago

Yes, but more specifically on Thinking, Fast and Slow, as noted elsewhere[1][2]: It is only the priming-related chapter (called 'The Associative Machine' in the book) that put "too much faith in under-powered studies", as Kahneman put it[3]. Not the entire book! The book is a synthesis of forty years of Kahneman's research and his collaboration with his late colleague, Tversky. A wide range of topics are covered; and it still absolutely merits reading. Patiently dive into the book and make up your mind. (But yes, a v2 of this book definitely is worth it, given the "authority" of the Nobel Memorial Prize.) [1] https://news.ycombinator.com/item?id=22054603 https://news.ycombinator.com/item?id=22054603 [2] https://news.ycombinator.com/item?id=21757524 https://news.ycombinator.com/item?id=21757524 [3] https://retractionwatch.com/2017/02/20/placed-much-faith-underpowered-studies-nobel-prize-winner-admits-mistakes/ https://retractionwatch.com/2017/02/20/placed-much-faith-und...

nojito 6y ago

This always gets mentioned and upvoted to the top but there has really only been one priming experiment that has been discredited.