July 11th, 2015

Another excerpt from my book-in-progress, Economics in Two Lessons. To recap, the Two Lessons are

Lesson 1: Market prices reflect and determine opportunity costs faced by consumers and producers.
Lesson 2: Market prices don’t reflect all the opportunity costs we face as a society.

In this section, I’m working on Lesson 1, leading up to the point (my restatement of what’s usually called the First Fundamental Theorem of Welfare Economics) that an ideal competitive equilibrium is one in which there are no unexploited potential gains from technical improvements or mutually beneficial exchange. For reasons I’ve spelt out already I don’t want to use the term “Pareto-optimal” to talk about this. I also want to confine “efficient” to its normal meaning of “technically efficient” and avoid the common economist practice of extending this to cover various definitions of “market efficiency”. So, I’m talking about “free lunches” or, more formally, benefits with no opportunity cost.

In Lesson 2, I’ll be looking, among other things, at the Second Welfare Theorem, which says any outcome with no free lunches corresponds to a particular initial allocation of property rights, broadly defined to include taxation obligations and entitlements of all kinds.

Now please comment, criticise and hopefully enjoy

The acronymic adage TANSTAAFL (There Ain’t No Such Thing As A Free Lunch). The saying was popularized, particularly in libertarian circles, by Milton Friedman’s book of that name and, a little earlier by by Robert Heinlein’s science fiction classic, The Moon is A Harsh Mistress. The acronym is derived from a marketing ploy used in 19th century saloons, whereby a ‘free’ lunch was offered to customers, on the assumption that they would wash it down with beer or other drinks. Naturally, the cost of the lunch was incorporated in the price of the drinks.

The key idea may therefore be restated in terms of the broader point that it is opportunity cost, rather than just monetary cost, that matters when making economic decisions. Although there is no explicit charge for the lunch, patrons can only consume it at the opportunity cost of forgoing cheaper beer.

Libertarians commonly use the TANSTAAFL adage to point out that services provided ‘free’ by governments will, in general, have an opportunity cost. ‘Free’ provision of some service must be funded either by higher taxes or by reductions in other areas of public expenditure. The more general point, that it’s necessary to look at the full opportunity cost of any good or service, and not just the immediate price, is yet another version of Lesson 1.

But there is a contradiction here. Most economists think that improved economic policy could yield better outcomes for everyone, even though they may disagree about which policies would yield this result. Libertarians, who extol the benefits of rolling back the state and giving markets free rein, are no exception to this rule. The same is true of technological advances that allow us to do more with less, for example, by producing goods and services with smaller inputs of labor, energy and capital.

A free lunch is ‘something for nothing’, that is a benefit obtained with no opportunity cost. The TANSTAAFL adage embodies an important truth applicable to many apparent ‘free lunches’, in which the true opportunity cost is carefully hidden.

If TANSTAAFL were literally true, however, humanity could never have risen above a subsistence level of existence. Every technological advance since people first learned how to make flint tools and control fire has provided a potential free lunch, literally and metaphorically, for humanity as a whole. The same is true of improvements in social and economic organization that have allowed larger and larger groups to co-operate in mutually beneficial ways.

TANSTAAFL holds if and only if there are no free lunches left on the table, which in turn will only happen if all options for technological progress have been exhausted and, in addition, the economic system is functioning perfectly1. So, if outcomes can be improved for everyone, the correct statement is TISATAAFL, that is, There Is Such A Thing As A Free Lunch’.

Economists have understood this point ever since Adam Smith wrote The Wealth of Nations, the first serious study of economic growth, in the 18th century. Even the poorest person in a modern developed economy enjoys a range of goods and services that were unavailable to our ancestors, with less effort and toil and, at least potentially, with less use of resources and damage to the environment.

The improvements in living standards generated by a modern economy are, for us, a free lunch. In fact, economics tells us about two kinds of free lunch, technological innovations and improved allocation of resources.

Technological innovations are the most obvious kind of free lunch. Technological innovations that allow us to produce a given output with less of every kind of input, including labour, provide us with the classic example of free lunch. Adopting the new technology allows us to increase output without using any additional resources. So, the opportunity cost of the additional output is zero. To put this point the other way around, additional production entails opportunity costs only if it is technically efficient.

The second kind of free lunch, the core concern of economics, arises from improved allocation of resources. Lesson 1 leads us to think about improvements that can be generated by allowing markets to work. Lesson 2 shows how public policy can yield improved resource allocation when markets fail to match prices and social opportunity costs.

In this section we will look at Lesson 1, and the gains from exchange discussed earlier. Exchange through trade and markets can generate benefits for everyone, compared to a situation where everyone relies on themselves. When Crusoe trades fish for Friday’s goat, each obtains a meal that would have had a higher opportunity cost in the absence of trade. The improvement is a (partly) free lunch, or maybe a free dinner.

By contrast, the saloon story underlying TANSTAAFL, in which an apparent bargain turns out to be nothing of the kind, stands in stark opposition to the economic idea of exchange as a bargain in which both parties benefit. It is in line with the pre-modern view of trade as a zero-sum game, in which any gain to one part is a loss for the other.

With the correct economic analysis, the saloon story illustrates TISATAAFL. Suppose that the customer would be willing to pay the saloon’s price for the beer alone. Then the price must less than the opportunity cost of obtaining the beer some other way, for example, through home brewing. On the other hand, assuming the saloon is not operating at a loss, its price must cover the saloon’s opportunity cost of providing both the beer and the lunch. In these circumstances, compared to the situation in the absence of exchange, the lunch really is free.

Under ideal conditions, the market outcome will ensure that there are no free lunches left on the table. More precisely, there are no potential benefits that can be obtained unless an opportunity cost is borne by someone. These are the conditions of perfect competitive equilibrium, the subject of our next section.

1 More precisely, ‘functioning perfectly, given the initial allocation of wealth’. We will look at this point later in the book.

Categories: Economics in Two Lessons Tags:
  1. Ikonoclast
    July 11th, 2015 at 13:10 | #1

    I have to be honest. This seems like a fairly safe, unadventurous and orthodox topic. I seem to recall you said your publisher thought it would be saleable and thus a worthwhile venture. Correct me if I am wrong on that.

    My own feeling is that if a modern economics or political economy author is not going to tackle what is fundamentally and systemically wrong with our economy then he/she is not really writing anything of real or lasting value. This presumes of course that there is something fundamentally and systemically wrong with our economy which contention I do make.

    Writing and thinking in the same tired old BAU paradigm (with just a little twist here or there) is not enough. Without addressing the questions of ownership and control of our economy, we are going to get nowhere. Neoliberalism or market liberalism will continue to dominate until it decimates. It’s already decimating in Greece.

  2. John Quiggin
    July 11th, 2015 at 14:19 | #2


    As I tried to hint, the problems of ownership and control will be dealt with in Lesson 2.

  3. Ikonoclast
    July 11th, 2015 at 15:37 | #3

    @John Quiggin

    Fair enough. One issue which will have to be dealt with is the tendency to inequality under the current (broad) ownership system. This is something which Piketty has highlighted without being ideological in “Capital in the 21st Century”. He has examined in detail which conditions lead to more inequality and to less inequality in the current system (broadly speaking “really existing capitalism”) sans direct government redistribution.

    When r>g there is a strong and very definite tendency to greater inequality. When inflation is low there there is also a tendency to greater inequality. These are two points I remember from Piketty. The ways to fight inequality in the current system are at least, increase the growth rate and/or increase government welfare redistribution and/or permit or encourage increased inflation. But it still begs the question. Why tolerate a system with a systemic tendency to greater inequality which needs constant high-effort, high-cost correction when one accepts that less inequality is both better socially and even economically?

    Rather, could a system be engineered with a greater tendency to equality? The latter does not necessarily mean a tendency to complete equality. The “equilibrium point” sought would have to have some allowance for inequality. I suspect a stable, totally equal society would also take excess effort and cost to maintain. I assume there would be diminishing and even negative returns as one attempted to approach absolute equality.

  4. hc
    July 11th, 2015 at 18:22 | #4

    I try to make the point in class above about opportunity cost only mattering if things are efficient by drawing a production possibility curve and showing that shifts along the curve must involve tradeoffs. Shifts in the curve outward provide free lunch opportunities – not an inevitable need for tradeoffs. I think the basic message was presented in this way in Samuelson’s text on Economics from the 1960s.

    Technological improvements relax the need to sacrifice guns if you want more butter. You can have more of both. Indeed the astonishing thing to me has always been that this is true even if it is only the dairy farmers who learn to tug on teats more efficiently.

  5. Geoff Edwards
    July 12th, 2015 at 06:26 | #5

    Prof John

    Three cautiously critical comments. First, your model with its the objective of reaching “ideal competitive equilibrium” hasn’t overcome the familiar problems: what is the starting point (there is nothing normative about the present condition which market forces are said to improve); can producers and consumers have perfect foresight; are there no externalities including free-riders or feedback as participants change their minds.

    Second, your intro refers to opportunity cost for consumers and producers but these have available to them only the opportunities that are presented by companies and governments. Governments have an unlimited range of opportunities, for example. They are perfectly capable of changing the bounds of the “ideal competitive market”. Perhaps your essay could add a section showing how even with frequent changes to the boundary settings, markets can still minimise opportunity cost. However, so what? I’m not sure what practical outcomes this dictum would have, or how somebody would apply it other than through normal satisficing behaviour.

    Third, in thermodynamics there certainly isn’t any such thing as a free lunch. Energy dissipates and is lost forever. There is no such thing as a perpetual motion machine. I think your analysis would be strengthened if you could explain the link between the biophysical world and the world of exchange. Your hypothesis could be that markets allow the amount of energy and materials wasted to be kept to a minimum by pricing opportunity costs. The problem is that in practice they don’t. Markets are careless of the waste of biophysical resources unless they exert price-pain.

  6. July 12th, 2015 at 07:01 | #6

    This is a digression. In a real economy, the state of the art is not followed uniformly. The best and most profitable firms are close to best practice, the worst ones are far behind. Contrast any well-run firm like Ikea with its slowest competitors. There is therefore always a substantial space of attainable Pareto Quiggin improvements. Technical progress can be divided into (a) the advance of the state of the art by R & D, (b) the lagged dissemination of the state of the art. Process (b), which is most of economic growth, consists of free lunches.

  7. Geoff Edwards
    July 12th, 2015 at 07:02 | #7

    I have just written an article on the oil shale industry in central Queensland. I have a section on opportunity cost and have argued that markets are incapable of shedding light on opportunity cost.

    First, to the company running a project, it has only a single project and can’t take up other opportunities to produce energy solutions. Even its parent has only a limited number of opportunities because of its expertise, personnel and culture.

    It is not even true that the parent will pursue this project if it has the best (lowest) opportunity cost. There are other strategic considerations that it may take in to account.

    Second, to the government, the price at which a plant can produce fuel is a minor consideration compared with others: international geopolitics, emissions policy, current account deficit, regional jobs, path dependence and the desirability of transitioning to a different say electrical system of transport. Government has a more or less unlimited range of opportunities and there is no market in which these can be evaluated against each other. Expressed in other words, the cost at the present time to one commercial firm of producing one optional product is or ought to be of little relevance.

  8. Julie Thomas
    July 12th, 2015 at 07:12 | #8

    hc says:

    “Technological improvements relax the need to sacrifice guns if you want more butter. You can have more of both. Indeed the astonishing thing to me has always been that this is true even if it is only the dairy farmers who learn to tug on teats more efficiently.”

    What about the cow? Talking to a former dairy farmer who is now a silage expert, the new opportunity for increased efficiency in milk production that is already taking place so that China can have milk will cost the cows a lot in terms of health and longevity.

    And China is peopled with humans who are not adapted to milk consumption so the costs of increased milk drinking in that country will be an increase in health problems in that country.

    There is no such thing as a free lunch; it only looks like it is free because neo-liberal economists don’t know or care about any of the people who pay the costs and they do not value the other things that suffer as they force their value system on everyone and everything.

  9. Ikonoclast
    July 12th, 2015 at 08:21 | #9

    A few points.

    1. John, I am not saying that you are seeking to develop a theory of everything from market prices, but it does occur to me that you might be seeking to develop a theory of everything from opportunity costs. Doesn’t imperfect information remain the problem? How can we know the true opportunity cost of many decisions in view of imperfect information?

    2. In my view the issue of markets is different from the issue of ownership although the two do interact and condition each other. It is assumed at the vulgarised level of economics that describing our economy as a “free market” entirely describes our economy. Yet the “free market” only co-ordinates (some) relations and operations in the acts of consumption or transfer. It does not co-ordinate relations and operations in many other acts especially the acts of production.

    It’s worth reading Krugman’s short article “The Corporate Archipelago” in this regard. (May 25, 2015 – New York Times. A substantial quote:

    “… what I found myself thinking and talking about is actually the present of capitalism — and in particular about the peculiar delusion that we live in a world of individual competition in freewheeling markets.

    I mean, some of us do — by and large, men gathered at street corners in the early morning, waiting to be picked up by vans that carry them off to various forms of day labor. But most people work in organizations. Even in the private sector, two-thirds of workers are employed by firms with more than a hundred workers, half by firms with more than 500, a quarter by firms with more than 10,000 employees. We may live in a market sea, but most of us live on pretty big command-and-control islands, some of them very big indeed. Some of us may spend our workdays like yeoman farmers or self-employed artisans, but most of us are living in the world of Dilbert.” – Krugman.

    I used to have a link to essay, by Krugman I think, about how businesses and corporations internally are command and control driven. Their intenral operations are not market driven. This kind of puts into perspective the fact that markets are important but they are far from everything.

    Focusing on markets and opportunity costs is too narrow. That is not the way to develop any comprehensive theory of political economy.

  10. Fran Barlow
    July 12th, 2015 at 08:31 | #10

    While not ruling out the possibility of there being ‘free lunches’ the technological improvements most of us think of when we compare ourselves with the humans of the period before settled agriculture relate mostly to the improved ability for harnessing/extracting ecosystem services and reconfiguring them to the satisfaction of human wants and needs.

    That’s certainly a ‘lunch’, but it’s by no means a free one if the resource is finite or damaged by human hands. Until comparatively recently in history, very few grasped that contemporary humans were eating the figurative lunches of future humans or even eating those of rival humans that were alive at the time.

    The size of the human footprint on the ecosystem of the planet is stupendously large, compared with what it was 7000BP. Those humans weren’t getting a free lunch either, but their lunch was so small that its opportunity was trivial and to them, nearly invisible, and thus easy to miss, or it would have been had they bothered to reflect on it.

    Today, we are more aware that chowing down on ecosystem services has consequences both in waste/pollution and the permanent destruction of valuable commodities — fossil hydrocarbons and oxygen producing and carbon-storing vegetation for example. Our contemporary attempts to approach a free lunch turn on energy efficiency and low-footprint technologies. We know that there’s no such thing as a perpetual motion machine but we fancy that the closer we get to that ideal using resources that will survive us, the closer to that elusive free lunch we will be.

    It seems to me that the free lunch paradigm relies greatly on dissonance — or on zero-rating some opportunity cost. In the traditional ‘guns v butter’ argument, people point to how many people good be educated, fed, clothed or given access to high grade water for the cost of a day’s military expenditure, which is nice because most of us see this latter as almost entirely antithetic to human welfare and serving the ill-deserving, corrupt and malign. Surely here there’s a free lunch to be had?

    Well, not quite. Removing benefits from the ill-deserving, corrupt and malign who are harming humanity is very appealing, especially if one uses the resources to enrich the lives of the marginalised, but even allowing that that is exactly what happens that’s still an opportunity cost for people employed in ‘defence’ — at least until they find new ways to meet their needs. And since economics is not generally in the business of making explicit ethical judgements, losses to rich spivs and criminals are still losses. That most of us would cheer and find our warm inner glows at such news is beside the point.

    And when the swords turn into ploughshares we are still mining the Earth for nutrient, polluting the air, water and land and encroaching on natural habitats that are, or would else be our legacy to future humanity.

  11. John Quiggin
    July 12th, 2015 at 09:12 | #11

    Geoff @5 As I said to Ikonoklast, these are the issues to be addressed in Lesson 2.

  12. Donald Oats
    July 12th, 2015 at 14:01 | #12

    I am looking forward to you expanding on the “society” in the second point, especially on political costs inflicted. Australia today is a wealth of examples on this very point.

    The free lunch at the salon is classic in that there is a statistical aspect to it: the salon proprietor is calculating that even if some of the population indulge in the free lunch and don’t consume any beer (free-riders), the rest of the population will more than make up for it, producing a neat little profit from the exercise. Both the proprietor and the drinking patrons would have a mutual interest in turfing out the “total abstainers” 🙂 (to coin an expression from Jock McCannon character (actor Graham Crowden, RIP) in “A Very Peculiar Practice”, BBC).

  13. Megan
    July 12th, 2015 at 14:50 | #13

    In Athens you can get a free lunch:

    Kostas lost his job as a marketing specialist in a big firm in Athens early on when the financial crisis broke. Gradually, he had no choice but to move back in with his mother and was forced to share her ever more meagre pension. He must have been close to 45 at the time, he’s 50 now.

    Then one day in 2011, as he tells it, he saw two kids fighting over some food they found in a dumpster (yes, that is Athens, even back then). The next day, he decided to go to a farmer’s market and ask the stall keepers for leftover food. Right then and there, he started to cook a meal with what they gave him, and to give it away to anyone who wanted to eat, to share.

    His main motto still is: Free Food For All! He will not tolerate discrimination of any kind. And he always eats along with the people he feeds. We share!

    These days, he feeds over 300 people every day. Athens is the city of homeless people. And they’re not winos or people with mental issues, as we know from North American and European cities, though some of them inevitably are.

    In Athens, they’re the people who not long ago had good jobs and good prospects, and often families to raise, and who now find themselves with nothing left. Many many people have moved (back) in with parents or family or friends, but not all have that choice. And even if they do, there is no future anywhere to be seen.

  14. J-D
    July 12th, 2015 at 15:17 | #14

    @Donald Oats

    ‘Coin an expression’ does not mean ‘use an expression’ or ‘borrow an expression’; it means ‘originate an expression’, which is what you avowedly aren’t doing. (The expression ‘total abstainer’ does not originate with the character of Jock McCannon in A Very Peculiar Practice; it’s much older than that.)

  15. rog
    July 12th, 2015 at 19:21 | #15

    @Geoff Edwards “Third, in thermodynamics there certainly isn’t any such thing as a free lunch. Energy dissipates and is lost forever.”

    My reading of that law is that energy cannot be created or lost, it dissipates then reforms.

    The free lunch analogy seems weak in that there exists plenty of examples where we don’t know the parameters defining lunch, free etc (ref Rumsfeld known unknowns, unknown unknowns etc)

    Globally we seem to be dealing with the consequences of past actions, apparently made with the best of information.

  16. ZM
    July 13th, 2015 at 01:49 | #16

    In terms of the “no such thing as a free lunch” idea you might want to include contemporary examples like the Lentil As Anything restaurants and their successes and problems:


  17. Geoff Edwards
    July 13th, 2015 at 06:12 | #17

    Afterthought on free lunch in economics vs biophysical world

    The best known advocate of the free lunch in terms of resource consumption is Amory Lovins with “Natural Capitalism” and locally, Charlie Hargroves with “The Natural Advantage of Nations”. These authors have predicted an optimistic future as we design our equipment and systems smarter. To my mind, Hargroves reads as a string of anecdotes of how individual enterprises have downsized resource use and saved money, but I see no theory that allows individual success stories to be scaled upwards to apply to a whole economy. As Julie above has explained, one person’s savings gained is another person’s job lost.

  18. Ikonoclast
    July 13th, 2015 at 10:51 | #18

    @Geoff Edwards

    Technically, there is no free lunch in physics terms. On earth, while the sun shines we get external energy (sunshine) coming into the earth system. This is new energy into our open system.

    Technically earth is an open thermodynamic system as matter and energy can pass in and out. We lose atmosphere (slowly), meteors do come in and energy enters and leaves so we are an open system.

    However, there will still be limits to quantitative growth on earth (built infrastructure and material production). There will also eventually (far in the future maybe) be limits on qualitative growth. The quantitative limits will be do with flows and circulations in the biosphere, and in and out of it, and this will include bioservices effects.

    I agree there are free lunches in a way but there will be an ultimate limit to these free lunches in the biophysical and economic senses.

  19. Nevil Kingston-Brown
    July 13th, 2015 at 11:48 | #19

    This assumes that R&D and investment in new and relatively unproven technologies have no opportunity cost. It’s certainly true in the long term that R&D pays off for society as a whole, but it’s not necessarily true for individual scientists and entrepreneurs.

  20. Ikonoclast
    July 13th, 2015 at 12:17 | #20

    @Nevil Kingston-Brown

    A lot of R&D that gave us many current technologies was done by government not private enterprise. People forget that. Command decisions do pay off. And don’t forget every big business and corporation is a command enterprise internally. They (firms) largely don’t have internal markets. And attempts to do set up internal competition in firms have been terrible failures. Look up “The man who ruined Sears Holding Corp” and “Sears’ Collapse Is Now ‘Inevitable'”.

  21. Donald Oats
    July 13th, 2015 at 15:10 | #21

    To be fair, I agree with you actually, in that the literal meaning of “to coin a phrase” is to create a new phrase; however, the modern use of the idiom “to coin a phrase” or “to coin an expression” is to (ironically) introduce a cliche.

    It would appear there are two schools of thought on this.

  22. paul walter
    July 13th, 2015 at 15:34 | #22

    You have to establish your groundings and his two lessons are where it starts. 99% of the stuff that turns up here follows from the basic propositions, to do with what constitutes social and cultural opportunity costs and and (usually misconceived) individual opportunist self interest.

  23. paul walter
    July 13th, 2015 at 15:41 | #23


    A nice little bok Review as summary tht expands a little on what the thread could be about.

  24. Ivor
    July 13th, 2015 at 17:02 | #24

    Although there is no explicit charge for the lunch, patrons can only consume it at the opportunity cost of forgoing cheaper beer.

    I don’t follow. Wouldn’t this be more expensive beer?

    Cheaper beer compared to what?

  25. Tim Macknay
    July 13th, 2015 at 18:14 | #25

    I think the point is that the price of the beers is inflated, to cover the cost of the ‘free’ lunch. So the patrons are paying more for beer than they would have if the ‘free’ lunch had not been provided.

  26. Megan
    July 13th, 2015 at 18:34 | #26

    @Tim Macknay

    That was how I took it.

    But as a side point: normally something like that would be a “loss leader” – so the drinks prices would be the same (neither increased nor decreased from ‘normal’) but with the ‘free lunch’ the publican hopes to sell enough extra beer to turn an increased profit.

    Something like: ‘normal’ day = 5 lunchtime drinkers makes $x profit. ‘Free Lunch’ day = 50 lunchtime drinkers makes >$x profit.

  27. Nevil Kingston-Brown
    July 14th, 2015 at 13:10 | #27

    No disagreement there, but the government R&D wasn’t free either.

  28. Ernestine Gross
    July 14th, 2015 at 21:30 | #28

    “What about the cow?”, Good one, Julie Thomas.

  29. Donald Oats
    July 16th, 2015 at 19:28 | #29

    Actually, the price of the beer at the pub would probably be the same as elsewhere, because one thing that beer drinkers do, and that is remember the price of beers here to the price of beers there. Get a reputation for expensive beer and that free lunch won’t cut it any more.

    Free lunch + Reputational Damage – extra charge on beer = overall cost to the pub, -ve if it is revenue. The reputational damage is the cost in sales lost due to drop in customer visits. Since the cost of the free lunch won’t be covered by the extra charge on the beer, and because the extra charge on beer implies some reputational damage, it seems like a bad proposition to offer free lunches which are subsidised by an extra charge on the beer.

    In fact, I suspect that having a free lunch and same price beer, or even a discounted beer, say first beer with the meal is free, would be more likely to bolster reputation, attracting custom during the times when there is no free lunch. Putting the price up on beer isn’t likely to increase customer visits outside of the free lunch period, and could cost you customer visits. Am I wrong in this?

  30. Tim Macknay
    July 16th, 2015 at 22:45 | #30

    @Donald Oats
    I believe that was Megan’s point.

  31. O6
    July 17th, 2015 at 13:29 | #31

    There’s evidence that advertised free drinks or cheaper drinks lead to increased violence, so those seeking a peaceful lunch, free or not, will go elsewhere (quite rightly, IMHO).

  32. Julie Thomas
    July 17th, 2015 at 14:39 | #32

    But eating while drinking can moderate the negative effects of alcohol on the brain and body and eating while drinking encourages ‘social’ drinking rather than drinking to get drunk and that could change the clientèle and result in a more peaceful venue.

    Maybe if they got some live music the musician could get a free lunch?

Comments are closed.