# Most “economists” aren’t

I’ve always thought that an economist is someone who understands opportunity cost. If there is one thing a first-year undergraduate economics course should teach, it’s an understanding of this concept. So it’s alarming to discover that most members of a sample drawn from participants in the profession’s most important conference are not, at least by my definition, economists.

Via Harry Clarke, I found this paper by Ferraro and Taylor (guest registration or subscription required). Ferraro and Taylor presented their volunteer subjects with this question.

‘You won a free ticket to see an Eric Clapton concert (which has no resale value). Bob Dylan is performing on the same night and is your next-best alternative. Tickets to see Dylan cost \$40. On any given day, you would be willing to pay up to \$50 to see Dylan. Assume there are no other costs of seeing either performer. Based on this information, what is the opportunity cost of seeing Eric Clapton?

(a) \$0
(b) \$10
(c) \$40
(d) \$50.

Take some time to think before looking over the fold

The correct answer is (b), the net value to you of buying a ticket and attending the Dylan concert, which is your next best alternative. This isn’t the easiest of opportunity cost questions, and, in a first-year undergraduate exam, it would distinguish those who’d really learned the concept from those who’d memorised the definition.

Still it’s not that difficult, and doesn’t involve any tricky thinking about information as in (say) the Monty Hall problem (it’s adapted from the intro textbook of Frank and Bernanke). Allowing for the fact that some people can get jobs in economics with a good technical training but no real understanding of economics[1], I’d have expected maybe 70 per cent of grad students or PhD qualified economists to get it right.

In the Ferraro and Taylor study, the proportion choosing (b) was 21.6 per cent, the least of any answer and worse than if everyone had picked at random. Since presumably some people did choose at random, this suggests that less than 20 per cent of “economists” are actually economists, at least by my definition.

fn1. I blame game theory myself, but then I always blame game theory. And opportunity cost reasoning is just as essential to game theory as to economics, even if its role is not so explicit.

## 39 thoughts on “Most “economists” aren’t”

1. I got it right! But then I’m currently studying economics…

Yes, you always do blame game theory and you should stop it! The trouble with game theory isn’t that it can’t be used to solve problem, but rather that it’s used extremely poorly in situations where it shouldn’t be used!

2. Silly question. One that can only be posed and answered by economists. No sane person gets the same mileage out of seeing the same concert twice.

If John Howard gets a 1pt boost in opinion poll on any day he allows a conscience vote on RU486, whats the opportunity cost to John Howard if he postpones the RU486 vote by (a) one day, (b) one week?

3. Albert Newton says:

I just put this question to my daughter who is studying economics in high school. She got it immediately correct and explained to me the importance of using net cost in the calculation not gross cost.

I’m never studied economics so I wouldn’t really know, but could it be that understanding of economics is inversely related to the number of years spent studying it?

4. James Farrell says:

John

Since the paper is not avaiable, I wonder if you could you quote exactly what they said about who took the test.

5. Katz says:

“You won a free ticket to see an Eric Clapton concert (which has no resale value). Bob Dylan is performing on the same night and is your next-best alternative.”

Hmmm.

\$10 may be the opportunity cost for a baby boomer.

But what is it for a GenXer?

6. edward squire says:

The actual opportunity cost is:

(e) \$10 – W, where W = the monetised unanticipated pain of having to listen to people born in the 1950s complain about how Dylan’s lyrics used to be soooo much better than they are now.

7. Harry Clarke says:

There is quite a discussion of this problem at your cross-posting at Crooked Timber. Also there was an earlier discussion at Marginal Revolution.

I’ve read the comments and still think that the failure to give the correct answer to this task is a problem for economics education. The ability to identify the costs of taking certain actions in somewhat complex situations would seem to be a basic requirement to be an effective economist.

The idea of evaluating situations in terms of what you have to forego net is basic.

When I posted the problem on my blog I was not aware of the discussion at Marginal Revolution (the problem was originally pointed out to me by a colleague, John King). But having now read the discussion there also I again think a real difficulty has been identified by the Ferraro/Taylor paper.

Economics programs often do attempt to (i) teach too much and (ii) become bogged down in technicalities. Teaching too much occurs because we often do emphasise ‘teaching rather than ‘learning’. Many first and second-year microeconomics texts contain enough material to cover a three-year curriculum. Technique is important but so too is learning the basics. Sometimes business schools do the ‘basics’ better than economics departments.

One thing I notice is that some of the best Australian economists tended in the past to come from agricultural economics backgrounds where lots of emphasis was placed on partial equilibrium analysis of particular markets. Less weight was placed on technique and general equilibrium analysis. I wonder if the emphasis on basics here is the reason for their comparative success.

Unfortunately agricultural economics is fast disappearing and being replaced by programs based on whats new in the US, examples include mechanism design and game theory. Good stuff – you can certainly display vast intellectual skills with it – but perhaps not as central as the core microeconomic principals used by the profession for decades.

It is dangerous to be too prescriptive about how economics should be taught but something seems to me to be not working well.

8. As a layperson, I always thought the opportunity cost was the full cost of the thing you have given up (in this case \$40.) Can you expand on why the net cost is used?

9. still working it out says:

If you had a question as basic as that in Physics, Chemistry or Engineering I doubt it would get such a poor response from a similarly qualified group of people. But I don’t think this is due to a difference in the quality of education.

The difference is that if Phys/Chem/Engineering student does not understand a basic concept thoroughly the first time that student will keep paying for it throughout their academic life. For example, if you do not understand Thermodynamics properly in first year engineering then you are going to struggle with second year thermodynamics. And by your fourth year you may be required to design a basic heat engine which will simply be impossible. At some point you will have to learn it properly or you will fail to complete the degree.

Maybe the problem is that Opportunity Costs is only taught at the start of a degree and that knowledge is not built on in later years.

10. Voodoo Supply-Sider says:

Helen, it’s better to think of it as the value of what you’ve given up. Suppose the maximum price you would have paid for Dylan was coincidentally \$40. You would have been ‘indifferent between’ going and not going to Dylan, so the value forgone through going to Clapton would be zero (no opportunity cost at all). But if the value you get from Dylan is actually \$50, you have forgone a surplus of \$10.

11. Paul Norton says:

I got it right, after studying two subjects of undergraduate environmental economics and one subject at Honours level in the same field.

edward, I calculate that the opportunity cost is still \$10 after taking W into account. The cost of attending the Dylan concert on the night of the Clapton concert is \$40 + W. The hypothetical subject is prepared to bear a monetised cost of \$50 + W to attend a Dylan concert on another night. QED opportunity cost is \$10.

It would still be \$10 after factoring in X, which is the monetised unanticipated pain of having to listen to people born in the 1920s, 1930s and 1940s complain about how much better Dylan’s lyrics and music were before he went electric in 1965.

12. Katz says:

“The actual opportunity cost is:

(e) \$10 – W, where W = the monetised unanticipated pain of having to listen to people born in the 1950s complain about how Dylanâ€™s lyrics used to be soooo much better than they are now.”

So “W” would be time-dependent?

13. StephenL says:

I’ve never set foot in an economics class room (except when it was being used to teach English Literature) and I got it right. It didn’t even seem that hard to me, which makes me wonder whether the more economics you study the less you are able to understand basic things like this.

What really interests me is how many parliamentarians would get this questions right.

14. cs says:

Bah! This is a totally preposterous question, based as it is on the entirely ludicrous assumption that a ticket to Eric Clapton has no resale value!

15. Aidan says:

Hmmm. So 50 – 40 = 10, riiight. But, um Sir, I don’t quite seee the question. What does “next-best alternative” mean? I can’t sell my Clapton ticket, but are you saying I can’t swap it for a Dylon ticket either? And do I want to see Clapton or Dylon or both or what?

16. What if we also knew that we would have been prepared to pay up to \$50 (or some other amount) for the Clapton ticket? After all, we entered the competition because we wanted the prize.

17. what Homer said on Harry’s blog. No person in his right mind would pass up the chance to see Dylan just because of a free ticket for Clapton

18. Terje Petersen says:

I’m happy to say I got it right (my ego is very relieved and insists that I tell everybody).

Like John I find it disappointing if actual qualified economists can’t figure out the correct answer to such a basic problem. Disappointing but not that surprising given that qualified economists seem to pedal all kind of crap that seems to be something that they just memorised rather than understood.

It’s obviously \$10 because that is how much you would save by not seeing Eric Clapton today. In real life however there may in fact be other options not mentioned. Perhaps rather than waste time at a concert you could be having an evening in bed with with that special person in your life. Of course the question made it clear that this was not an option because it made explicit your “next-best alternative”. Also it is unlikely that an economist would have a love life.

19. rog says:

Yes, I put it at \$0 because I would have missed the opportunity to see Dylan.

The real answer is \$50 (on any day)-\$40 (on that night only). By not attending Dylan I missed the once only discount of \$10.

20. Craig says:

I’m not sure that the opportunity cost is \$10 at all, given that the ticket has been won, not purchased. A ticket won would perhaps have some (although not all) of the attributes of a ticket received as a gift.

Gifts have sub-optimal utility, in the sense that when asked to value them recipients will usually assign a lower monetary value to a gift than the giver was obliged to outlay for it on purchase.

Certainly the winner has gone to the trouble of entering a contest or raffle of some sort offering a concert ticket as a prize, suggesting that he/she attaches some value to it, but not \$40 or else he/she would have paid that much for it.

I might pay \$2 for a raffle ticket offering a chance to win an HSV Commodore V8 at the shopping centre, but I would never pay \$60,000 for one. If I win it, but am not allowed to sell, what is it worth to me? Who knows, without being able to enter into the process of higgling for an acceptable price with a potential buyer how can this be established?

On a more serious note, if even economists cannot get an opportunity cost calculation right, then what does that say about the value of a discipline that assumes the under-pinning usefulness of the concept, supposedly universal in the minds of economic actors, as an engine of efficient resource allocation?

21. “suggesting that he/she attaches some value to it, but not \$40 or else he/she would have paid that much for it.”

Not necessarily. They might have been willing to pay \$40 for it, but decided to wait to see if they won it in the competition.

22. Kevin K says:

I’m a layman, so it’s quite possible I’m absolutely wrong.

However..

It seem to me the question is ill-defined or insufficiently defined. The questioner is assuming as certain that the music fan in question will have other opportunities to see Dylan, but only at the highest possible price the music fan is willing to pay. If Dylan were either

1) never playing again, or the fan will never have the opportunity to see him again, (the question did not rule this out)

or

2) playing again where you have an opportunity see him but also again at \$40 (again, the question as stated says nothing about future opportunities for seeing Dylan)

then would the answer not be for 1) the full cost of seeing Dylan, and for 2) \$0?

23. Kevin K says:

To add, if the above is correct, then the question has in fact no one definitive correct answer, as the correct answer depends on information that the question does not provide.

24. James Farrell says:

Kevin, those considerations are factored in to the \$50. That ‘reservation price’ takes into account how much our protaginist likes Dylan, how urgently she wants to hear him in concert, what she estimates is the probability that other opportunities will arise, what travel costs are entailed by those other opportunities… in short, everything.

25. Jimmythespiv says:

the key phrase is “next best alternative”, which implies a preference for Clapton over Dylan, which means if you see Clapton, you give up the \$10 difference in utility between the two.

26. Dogz says:

I’ve never studied economics, although I did get the official answer. However, I think \$0 is more correct than the official answer.

The problem is with the statement “On any given day, you would be willing to pay up to \$50 to see Dylan”. It doesn’t correspond to any real-world human preference. Some days I would pay \$10 to see Dylan, some days you couldn’t pay me to see him – eg if I am holed up in bed with a terrible flu, or if I had recently seen him. [you’d have to pay me to see Dylan at any time but that’s just me – never have understood the penchant some people have for 60-something has-been rockstars].

So a better model would be to have a probability distribution over possible scenarios under which you’d get to see Dylan, and attach a dollar value to each scenario (actually, you really should model this as a stochastic process since the value of seeing Dylan today depends upon whether I saw him yesterday, but to avoid making things too complicated let’s ignore that).

One could then interpret “On any given day, you would be willing to pay up to \$50 to see Dylan” as an assertion that the maximum you’d ever be willing to pay to see Dylan under any scenario is \$50 [strictly, we also need to discount future values back to today, but that can be handled easily enough by assuming all values are expressed in today’s dollars].

You also need to model the ticket cost of seeing Dylan under each scenario. Presumably, there’s some likelihood of getting a free ticket (after all, we got one for Clapton), and the price is also likely to decrease in real terms over time (he ain’t getting any more entertaining as he gets older). So it is fair to assume that at least on average, in real terms, the price of a Dylan ticket will be less than it is today (although to some extent the reduction in price due to his decreasing entertainment value will simply reflect the decreasing value to you of seeing him, so the difference between price and value will not necessarily increase).

Under these assumptions (which seem perfectly valid given the problem statement) the answer “\$10” is only correct if you assume _every_ possible future Dylan-seeing scenario has a less favourable net cost than seeing Dylan today.

However, if, as seems reasonable, it is certain that one day you’ll get to see Dylan under more favourable circumstances than today, then your opportunity cost of attending Clapton today is \$0.

27. jquiggin says:

Couldn’t walk away, Dogz? Don’t worry, the first hit is free, as is the first answer !

You’re quite right that, if you expect future opportunities to see Dylan, and you only want to see him once, the opportunity cost is not \$10, but the value of the next-best alternative, which might be playing your old Clapton records at home for free.

28. Dogz says:

Couldnâ€™t walk away, Dogz?

I know. I feel pathetic. Saw you on TV last night. You looked younger than your picture here – makeup? [BTW, the arrogance of the ex-pollies – Carr, Egan, Greiner – was just disgusting. Very much “f*ck what the public think”. Thought the treatment of the Tunnel CEO was pretty rough though – he’s just a hired gun of the owners and has clearly been told not to concede anything on pain of being sacked]

Even if you want to see Dylan more than once, and you expect future opportunities, the opportunity cost is not \$10. Depends on how much impact not seeing him today has on the expected number of times you’ll see him, and how much you value seeing him any given number of times.

I think the wording of the problem is quite misleading if \$10 is meant to be the correct answer:

â€œOn any given day, you would be willing to pay up to \$50 to see Dylanâ€?.

Qualifying with “On any given day” ensures the reader is automatically thinking about future opportunities, not just the opportunity presented today.

29. Waratah says:

I thought the correct answer was MINUS \$40 but that wasn’t an option so I decided upon \$0 (nothing ventured, nothing gained).

30. jquiggin says:

Makeup? – No, just friendly lighting.

31. edward squire says:

Paul Norton Says (February 20th, 2006 at 9:28 am):
I calculate that the opportunity cost is still \$10 after taking W into account. The cost of attending the Dylan concert on the night of the Clapton concert is \$40 + W. The hypothetical subject is prepared to bear a monetised cost of \$50 + W to attend a Dylan concert on another night. QED opportunity cost is \$10.

My post was a joke – and yes, a lame one. Freud’s greatest disservice to humanity was his “Jokes and Their Relation to the Unconscious” – the least funny book on jokes ever written. Freud manages to destroy every joke he touches by analysing it to death. I can’t hope to emulate the master, but this post is in that spirit.

The post/joke hangs on the word “unanticipated”. The joke implicitly assumes the consumer is (relatively speaking) a dill whereas the narrator of the joke is omnipotent; that is, the former possesses imperfect knowledge and the latter, perfect knowledge. The implicit inference from this is that the reader makes a distinction between the anticipated consumer surplus and the unanticipated (and what would be “the actual” consumer surplus if the consumer had, in a parallel universe, actually attended the Dylan concert). The economic crux of the joke (as opposed to the ‘making fun of some baby-boomers’ part of the joke) is that the reader should be initially thinking strictly in terms of the standard definition of the consumer’s willingness to pay (see James Farrell’s post above), which incorporates only anticipated outcomes (the agent consciously deals in probabilities, be they equal to or less than 1) – and then on reading my post is supposed to, by an imaginative leap, shift to a different concept: the consumer’s willingness to pay is an “actual” one based on the “actual” experience of a Dylan concert which includes features (whiners) completely unanticipated by the consumer (because the consumer is a bit of a dill).

So, one is supposed to think of it as follows: P=price, R=reservation price (willing-to-pay price), A=anticipated surplus (surplus agent believes will be experienced), W=unanticipated pain, U=unanticipated surplus (“actual” surplus that agent would have experienced).

Standard notion: A = R – P
Joke notion: U = R – W – P, or better, U = A – W, where W>0

Yeah, I know: ab-sol-utely hilarious.

Katz Says (February 20th, 2006 at 9:32 am):
So â€œWâ€? would be time-dependent?

Not really. It’s knowledge (or ignorance) dependent. In the joke the consumer is supposed to have absolutely no idea that they would have in fact experienced painful whining had they attended the Dylan concert.

32. Ernestine Gross says:

Most “economists” aren’t

Sample:

(1) Business 26 subjects, 4 (15.4%) correct
(2) Applied Micro 95 subjects, 24 (25.3%)
(3) Micro Theory 7 subjects, 3 (42.9%)
(4) Macro/International 49 subjects, 7 (14.3%)
(5) Methods 17 subjects, 4 (23.5%)

(1): examples: Accounting, Finance, Organisational design
(2); ‘labour’, ‘health’, ‘public finance’, ‘economics of education’
(3): game theory
(4): monetary, international trade, economic growth
(5): econometrics, experimental economics, statistics.

The question was changed. The sentence “Based on this information, which is the minimum amount (in dollars) you would have to value seeing Eric Clampton for you to choose his concert?” was added.

The amended question was administered to 34 people (1/2 from conference, 1/2 from econ departments). Result: 15 out of 34 answered (b); change in result is statistically significant.

Source: the publication listed on this thread.

The authors of the publication argue that the notion of ‘opportunity cost’ is important for decision making. So, why pick first a question such that getting the correct answer might make you feel good for having scored on remembering a calculation while possibly forgetting why you want to do the calculation in the first place?

And, what does a PhD in Economics mean these days?

33. James Farrell says:

Someone might fail a test question on ‘present value’, but that doesn’t mean that, when it came to the crunch, they would choose \$1001 in a year’s time over \$1000 now.

I wonder if the low pass rate on opportunity cost reflects a real lacuna in understanding, or just an unwillingess to invest time on what students regard as superfluous abstract concepts.

To find out, it would be interesting to ask a second group of subjects ‘what is the most you would be prepared to pay for the Clapton ticket?’ rather than ‘what is the opportunity cost?’ (Of course, the ‘no resale’ condition would still need to be stressed.)

If this second group performed much better, we’d need to ask ouselves whether the opportunity cost concept is so indispensible, at least for non economics majors.

I’ve found that, no matter how lucidly, ingeniously and comprehensively I explain it, even the better students tend to come away with the idea that opportunity cost is a synonym for implicit cost.

34. James Farrell says:

By the way, John, did you by any chance delete a comment in which someone requested Edward Squire to supply a detailed Freudian analysis of his joke?

35. edward squire says:

James Farrell Says
To find out, it would be interesting to ask a second group of subjects â€˜what is the most you would be prepared to pay for the Clapton ticket?â€™ rather than â€˜what is the opportunity cost?â€™

Another question might be: Do you even care whether you get this right?

Iâ€™ve found that, no matter how lucidly, ingeniously and comprehensively I explain it, even the better students tend to come away with the idea that opportunity cost is a synonym for implicit cost.

Interesting. I’ve found exactly the same thing. It may be due to the way the division between explicit and implicit costs of the firm is explained by introductory textbooks; they sometimes refer to opportunity cost when speaking of implicit costs but almost never refer to it with respect to explicit costs.

36. Kim…

Looks like your page was heavily hit by spam…