mattnewport comments on Applying utility functions to humans considered harmful - Less Wrong

26 Post author: Kaj_Sotala 03 February 2010 07:22PM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Article Navigation

Comments (114)

Comments (114)

Sort By: Best

You are viewing a single comment's thread.

Comment author: mattnewport 03 February 2010 09:42:17PM 7 points [-]

A model is not terribly useful if it does not do a better job of prediction than alternative models. (Micro)economics does quite a good job of predicting human behaviour based on a very simple model of predictable rationality. It is not clear to me that this model offers a better approach to making meaningful predictions about real world human behaviour. I've only skimmed the article but it appears the tests are limited to rather artificial lab tests. That's better than nothing but I'm skeptical that this model's real world predictive power justifies its complexity.

The 'true' utility function of any particular human is no doubt an intractable beast of a computation but don't be too quick to dismiss the value of assuming a much simpler utility function and assuming that people do a reasonably good job of (boundedly) optimizing for it.

Comment author: bgrah449 03 February 2010 09:54:51PM 1 point [-]

I think it's useful inasmuch as it turns "unknown unknowns" into "known unknowns." Knowing what you're ignoring in your approximation seems valuable.

Comment author: mattnewport 03 February 2010 10:29:17PM 1 point [-]

I think they are claiming that their model more closely matches observed behaviour in certain specific controlled environments. It is a big leap from there to assume that the features of the model map in any useful way to actual features of human reasoning.

Comment author: Kaj_Sotala 04 February 2010 10:40:23AM 0 points [-]

Yes, I admit that it can sometimes be useful to think of humans as having utility functions, and this can be a useful model. I should have said that in the post, now that you mention it. But then one should then always keep in mind that that's just a simplified model that's appropriate for certain situations, not something that can be indiscriminately employed in every case.

Comment author: rortian comment score below threshold [+] (23 children)

Comment author: mattnewport 04 February 2010 02:02:20AM * 3 points [-]

Well, at the simplest and most obvious level, microeconomics predicts how producers and consumers will respond to changing price levels quite well. It predicts that, all else being equal, a rise in the price of a good or service will reduce demand for that good or service. Do you think that prediction is wrong? Since the entire field of microeconomics is largely about how households and firms make decisions to allocate limited resources I find it odd that it is news to you that it makes predictions about human behaviour.

More generally, the impact of a change in incentives can be predicted to a useful level of accuracy by basic microeconomic analysis. By comparison to macroeconomics (admittedly a low bar), microeconomics is quite successful at prediction.

Behavioural economics has produced a lot of valuable insights. I doubt the author of the article I linked would deny that. I think however that behavioural economists sometimes make a similar mistake to the one they criticize traditional economists for making when they extrapolate the results of simple lab experiments to the real world. Just as the naive idea of a perfectly rational agent should not be taken too literally, the results of behavioural economics experiments under artificial conditions in a lab should not be applied too literally to real world decision making.

Comment author: rortian 04 February 2010 02:54:20AM -2 points [-]

I asked about a prediction in human behavior. I am quite well aware of these predictions that are made in general, but this is in an absurdly abstract model with patently false assumptions.

It predicts that, all else being equal, a rise in the price of a good or service will reduce demand for that good or service. Do you think that prediction is wrong?

No, I think it is trivial.

Utility functions in microeconomics are not very useful for predicting human behavior contrary to what you claim. The OP was correct to look for more interesting classes of functions.

Comment author: mattnewport 04 February 2010 04:11:34AM * 10 points [-]

I am quite well aware of these predictions that are made in general, but this is in an absurdly abstract model with patently false assumptions.

By 'this' do you mean the model of humans as perfectly rational agents? That's a caricature of microeconomics and is not necessary to make useful predictions based on microeconomic reasoning.

It predicts that, all else being equal, a rise in the price of a good or service will reduce demand for that good or service. Do you think that prediction is wrong?

No, I think it is trivial.

Maybe the simple fact is trivial but the implications are not. Microeconomic reasoning explains a number of facts that are non-obvious, counter-intuitive and non-trivial to many people, for example:

Price ceilings cause shortages (rent-control, fuel price ceilings in the 70s).
Price floors cause surpluses (agricultural subsidies, minimum wages increasing unemployment).
Purchaser subsidies cause price increases (cheap student loans raise tuition fees, mortgage interest tax deductibility raises house prices).
Low interest rates cause speculative asset bubbles and malinvestment.
Banning the sale and purchase of drugs does not work and makes some criminals very wealthy.

If more politicians and voters understood these 'trivial' economic facts then we might see slightly better policy. Unlikely though, since politicians' behaviour is also best understood as a rational response to incentives.

Comment author: Matt_Simpson 04 February 2010 07:05:33PM 1 point [-]

I wish the parent wasn't downvoted into oblivion so that more people could see this!

Comment author: rortian 08 February 2010 12:45:55AM * 0 points [-]

Sorry to take so long to get back to you on this but I do think this stuff is important.

This line from wikipedia on the minimum wage really captures what I would like to say about a lot of this stuff:

Michael Anyadike-Danes and Wyne Godley [21] argue, based on simulation results, that little of the empirical work done with the textbook model constitutes a potentially falsifying test, and, consequently, empirical evidence hardly exists for that model.

The minimum wage stuff is Econ 101/ideological claim that doesn't take into account a lot of factors. There is a lot to say about this one issue but the lack of empirical evidence is an important thing to keep in mind about economics.

I think the point about rent control is right, but I think you are leaving out positive externalities: middle class workers can live near their place of employment.

The fuel price ceilings in the 70s sounds ridiculous to me though. It turns out the price spikes were largely due to OPEC. There is much less harm if you are dealing with a monopolist/oligopolist with low cost in setting a low price.

In another comment I addressed farm subsidies. They are pernicious and many people wish they would die.

Purchaser subsidies...I went to a school with low tuition that was subsidized by the state. Student loans are not exactly the biggest factor in tuition rates so I don't think there is much of a point to be had there. Mortgage interest tax deductions are an absurd tax break for the rich and lets get rid of them.

Low interest rates cause speculative asset bubbles and malinvestment.

That is not a result of economics but it is a favored talking point of the right who would prefer not to talk about over-leveraged-formerly-seen-as-perfectly-rational financial institutions. Low interest rates leading to inflation during times of high utilization in the short to medium term is a result of macro know as the Phillips Curve.

Bubbles almost always are the result of herd behavior which is not a prediction of economics; nevertheless they are very real.

Banning the sale and purchase of drugs does not work and makes some criminals very wealthy.

This is a complex issue. Personally I'd like to see most substances removed from the criminal justice system.

Let me first return to your first point:

That's a caricature of microeconomics and is not necessary to make useful predictions based on microeconomic reasoning.

No it isn't. I am quite familiar with microeconomics mathematical models and you really need some simple frameworks/powerful differentiation techniques to get far. If you are willing to throw out all talk about consumer surplus then you might get a start with a workable framework. However, this is what so many people that love hard-right econ love to tout.

Perfect rationality/information assumptions are made because they are mathematically easy. I'd love to see some bounded rationality models, but trust me that it opens up all sorts of mathematical thickets.

I used to be very sympathetic to the right-wing investor/business community and I am very familiar with their arguments. Often when they say they are arguing with economics on their side, they are arguing from a highly ideological point of view and very few people will admit/realize that.

Comment author: Matt_Simpson 04 February 2010 07:07:42PM 1 point [-]

I asked about a prediction in human behavior. I am quite well aware of these predictions that are made in general, but this is in an absurdly abstract model with patently false assumptions.

As is any model of human behavior that is tractable. All we're really going for is prediction anyway, so who cares?

Comment author: rortian 05 February 2010 01:55:36AM 0 points [-]

Look if econ had little influence outside its field, I would agree and say who cares. However this is hardly case.

I would agree with something you suggested though. We would do well do just discuss the end results and remember that the models are trash.

Comment author: mattnewport 05 February 2010 02:12:45AM 1 point [-]

The real damage comes from macroeconomics and I'd agree that most of the models used there are crap - because they fail at prediction.

Comment author: Matt_Simpson 05 February 2010 05:03:48AM 0 points [-]

Look if econ had little influence outside its field, I would agree and say who cares. However this is hardly case.

I'm not sure I understand what you mean. Why does it matter whether econ influences other fields? Are you suggesting that people in other fields end up taking economic models too seriously? Or something else?

Comment author: rortian 05 February 2010 07:50:48AM 0 points [-]

Good question. I would say that does happen. Dan Drezner comes mind on this front.

I meant to say that if it had little influence outside of the academy.

Comment author: Matt_Simpson 05 February 2010 09:38:42AM 1 point [-]

Well the models influencing the academy and what influences public policy are definitely not the same, if that's what you mean. I still don't see what you're getting at. If anything, I think basic economic models are under appreciated. Consider mattnewport's post for examples.

Comment author: rortian 05 February 2010 02:56:59PM * 2 points [-]

I haven't gotten back to matt's post, but I will. This sort of amazes me:

Well the models influencing the academy and what influences public policy are definitely not the same

Economist have a huge amount of influence in public policy and US jurisprudence. I would be shocked to hear about another set of models simply for political and judicial consumption. Often they are leaning on economists and not the original work, but they would still be using the same model in this case.

Were it not for the structure of the Senate, we wouldn't have farm subsidies. Everybody but those from largely flat and empty states want them gone.

continue this thread »

mattnewport comments on Applying utility functions to humans considered harmful - Less Wrong

Article Navigation

Comments (114)

Nearest Meetups

Virtual Study Room

Recent Comments

Recent Posts

Recent Rationality Quotes

Recent Wiki Edits

Top Contributors, 30 Days

Recent Karma Awards

Comments (114)

Top Contributors, 30 Days

You'll need to login or register to do that

(Don't worry, it only takes a few seconds)

Create

Login