Rational choice theory, the current model of human behavior used in most mainstream economism, is undoubtedly the best first order approximation. It works surprisingly well even when everybody knows it's incredibly crude and inaccurate model.
Next step to improve upon state of the art should be something only slightly more complex. The best candidate seems to be Regret theory. "Regret theory: An alternative theory of rational choice under uncertainty", Economic Journal, 92(4), 805–824, Loomes, G. and Sugden, R.
Rational choice theory assumes that people don't have internal time dependent state in their choices. Regret theory adds the anticipation of regret. It adds just one term into utility function but it covers many aspects of human behavior. There is also empirical support backing it up.
Tversky and Kahneman, who are seminal in the field of "biases" in psychology (for lack of a better term), were explicit in their hope of informing the utility functions that economists use to model decision making and well being.
Kahneman famously won the Nobel in economics as a sort of belated acknowledgement. (Belated because Tversky passed away before they got it).
> (Daniel) Kahneman told The New York Times in an interview soon after receiving the honor: "I feel it is a joint prize. We were twinned for more than a decade."
Blah blah dismal science and all that, but microeconomics is considerably less hand wavy than macroeconomics and at least in the US there are all sorts of policies that pretend that people don't respond to incentives (never mind if they respond as homo economicus or not).
The trouble is not that they don't respond. It is that they tend optimise for the incentive rather than for what the incentive is intended to achieve - which are not always the same thing.
Microeconomics is less hand wavy sure, they have clear cut formulas and solid maths, right? Too bad microeconomics is complete shit, and has far less predictive power than even the most simple macroeconomic models. Microeconomics is what brings us simplistic assumptions about behaviour, and assumes 100% rationality. And then when microeconomics actually attempts to become useful, with things like indifference curves and non-cooperative games, it becomes hand-wavy, because what's more hand-wavy than arbitrary numbers for preferences? Behavioural economics came into being because of how bad microeconomics is.
It's true. The standard economic assumptions (people have rational preferences, are purely selfish, and have perfect information) don't reflect real-world humans. But that's what makes them robust. Human behavior and culture can change drastically, but standard econ tools still work. They can predict and explain outcomes in ancient Rome just as effectively as present-day San Francisco.
Behavioral economics is very useful, and there are still tons of unexplored questions in the field. But importantly: behavioral econ is more speculative than standard econ. Behavioral econ models depend on getting both econ and human psychology right. Even if such models are fleshed-out, human behavior is quite plastic. So behavioral models are less likely to be correct when applied outside of present-day western culture.
Do you honestly believe that? Would you say "psychology is not much better than throwing bones" or "epidemiology is not much better than throwing bones?" Because the demands for statistical rigor in economics are far greater than in those fields. http://qje.oxfordjournals.org/content/by/year
Statistically rigorous astrology is still bullshit, no matter how much curve-fitting you apply because the core premise is false. Most psychology, as we are now learning when people actually try to replicate experiments, is also bullshit. If economics is starting from equally flawed initial premises regarding individual choice and decision-making, as I and many others believe, then no amount of statistical rigor in the process will change the field from being a contived method of throwing darts and making grand claims about how much you understand dartboards and the aerodynamics of darts.
I agree that there are many valid criticisms of neoclassical economics, ranging from assumptions of equilibrium and human rationality, to the emphasis on economic growth clashing with the reality of finite ecological resources on Earth.
However, I'd ask if you typically expect a field of study to be perfect before you'd consider it useful? Do you expect psychology to completely understand the human brain before you'd seek help for substance abuse? Or for physicists to understand super-symmetry before you'd use a cell phone?
> Behavioral econ models depend on getting both econ and human psychology right
This is the important bit. The much criticised "rational expectations revolution" in modern economics was less about popularising a hypothesis that people were on average good at anticipating the future and more a case of arguing that if a proposed policy stopped working if humans started anticipating how it would affect them, it probably wasn't a good policy...
> Raj Chetty has estimated that every $1 of tax subsidy only leads to around 1 cent of additional saving (making it a pretty good place to start if you have a large budget deficit to sort out).
Wow. That's horrific. I suspected it was inefficient, but this is much worse than I thought.
> Think about ‘social trust’. It’s a better predictor of national economic growth rates than human capital, and varies greatly between countries.
Viewed alternately: Mistrust is a drag on economic growth. The people I spend to keep you from ripping me off are people that make no net contribution to the economy as a whole.
And here's my take on what needs to happen to macroeconomics: People want different things. Some want savings. Some want status. Some want to improve their human capital. Some want to have fun. (And, there may be other major categories as well.) In fact, most people want several of those things. At any given time, they seek several different goals. The economy is made up of people who have different amounts of emphasis they place on the different categories, but the economy as a whole might be characterizable by the sum of peoples' emphasis on each of the categories.
A change in circumstances can make them change the emphasis that a person gives to different category. A large, economy-wide change can change a the emphasis of a whole lot of people, and therefore change the behavior of the economy as a whole. Macroeconomics might be able to figure out from, say, the shock of 2008, that many people are going to be in fear, and place much more emphasis on preserving their savings, and from that figure out what the effect on the economy will be going forward (at least until peoples' perception of their circumstances changes).
> The people I spend to keep you from ripping me off are people that make no net contribution to the economy as a whole.
Saw one estimate that high inequality countries spend about 25% of their economic output on 'security' Think six able bodied men 'working' AKA sitting all day outside a bank with machine guns. One problem with economists are they refuse to see that as a problem.
One of the barriers to integrating more realistic psychological ideas into economics is that the ideas about human nature in political ideologies such as conservatism and liberalism include mistaken assumptions.
As someone with a PhD in economics, and a believer in the what the mainstream econ departments teach, I still found this article very reasonable. I hope no one will take this article to be saying that orthodox economics is fundamentally flawed because at no point is that argued. Instead they argue that behavioral biases exist and are well enough understood that we can in some cases craft policy based on them, instead of the standard economic assumptions.
My favorite points were
- Unpack ‘animal spirits’. These arise in the minds of economic actors. We can unpack the biases behind them and do something about them – an indeed BIT is currently working on such a project.
The debate on "animal spirits"[0] is ongoing, but the potential payoff from understanding them is huge. They deserve more research using more modern ideas and methods.
- Promote rainy-day saving. Recent work suggests that the behavioural and economic effects of having even small amounts of saving are even larger than previously thought.
There are two contradictory arguments often made in regard to poverty and rational behavior: (A) poor people are completely constrained by their poverty and unable to make the "right" decision to save and (B) poor people are unable to make rational decisions because poverty affects their decision making ability. I am mostly in the (B) camp and as such, I think that helping poor people to behave more rationally is a great idea. However it does require honestly stating that poor people do make bad decisions, even if anyone else in that situation would have done the same.
The weakest point IMO was
- Trial light-touch ways to improve estimates by key economic actors. Encourage the use of estimation frames.
This is based on a claim that is much more general, and therefore requires more evidence. I would like to see a lot more studies on improving estimation accuracy in general, before I accepted that there is some general formula to make people's estimates more accurate.
[0] A term of trade for Keynes' original, but discarded by the mainstream, idea that markets were moved by irrational exuberance/pessimism by society as a whole. The main advocate of this theory today is Robert Schiller.
Yes please and lets start with finding ways to include technology and technological progress not as an externality but as an integral part of the equation.
I've thought China's Sesame Credit system could be interesting in this direction. I know there are some dystopian perspectives on the program. At the same time, consider how much they can learn about consumer behaviors from the aggregated data. And, they can even control this behavior to some degree through gaming mechanics (compulsion loops / social impact).
I don't know much about economics, but it seems this kind of environment (a national-scale, economic video game) could provide a mass of data to inform the 10 suggestions from the author. "practice based on plausible myths" - the opposite of this is practice based on meaningful, real-time data about economic actors, right? Would be very curious of an economists opinion of the role of such a system..
Economic policy needs to understand human behavior more so the author suggests QR codes on dollar bills so people can track their spending.... Ummm what ?
When the author says "add QR codes on bills so that consumers can easily extract their consumption data and make better choices", I think they mean "bill" as in a printed or written statement of the money owed for goods or services, rather than "bill" as in US banknote.
To be honest, it's still a bit of a WTF in this particular article given that there's plenty of behavioural evidence that (i) people really don't like scanning receipts into tracking systems, even when they're being paid to do so by getting their expenses refunded (ii) most consumers don't know or care what a QR code is
The problem with rational choice is a paradox: Not only can the individual respond to incentives, the individual can also prioritize incentives based entirely on subjective rationality.
You can know the decision an actor makes, but not the motivation. You can know the motivation an actor has, but not any future decision.
>> Shift focus from products to ‘choice engines’. Get Price Comparison sites to do the work – but make sure they stay clean.
HAHAHA! That is honestly one of the most laughable statements I've seen in years. Yes, "get the internet to fix it". The idea of useful price comparison sites free of bias and toxic reviews magically arising from the chum is lunacy. We all know how useful Amazon reviews are, or how travelocity has produced a generation of purely rational travel consumers.
Coming from the UK bank personnel, I think the authors are thinking more about gocompare and moneysupermarket, which you can think of as centralised reverse-auction sites.
And from personal experience, at least for general home utilities and insurance providers ... they work reasonably well. They would be really good if the quotes also included the small print features in comparable dimensions. At the moment you'll need to go over all those yourself, and doing that side-by-side even just for the best options takes a lot of time.
But due to financial nature of the services, there is plenty of space for dirty kickbacks. Hence, trying to reduce collusion within and between the choice engines is a laudable goal.
I've been seeing more and more 'compensated' reviews on Amazon, or even worse, lots of suspiciously short 5 star reviews. Some product groups are worse than others but it only seems to be getting worse.
Things can be imprecise, and can be modeled as such. Things can be chaotic, and can be modeled as such. Technologists tend to think in a binary between things that can be represented by a single number and those that can't, but that doesn't reflect reality but rather the common (limited) tools of our trade. Sometimes you've got to do a dot product of some distributions for possible values of future vectors, and you get a vague -- but still very real -- answer.
Next step to improve upon state of the art should be something only slightly more complex. The best candidate seems to be Regret theory. "Regret theory: An alternative theory of rational choice under uncertainty", Economic Journal, 92(4), 805–824, Loomes, G. and Sugden, R.
Rational choice theory assumes that people don't have internal time dependent state in their choices. Regret theory adds the anticipation of regret. It adds just one term into utility function but it covers many aspects of human behavior. There is also empirical support backing it up.
http://people.few.eur.nl/wakker/pdfspubld/15.2regret_history...