A Word on Economic ‘Fallacies’

I’ve developed an aversion to the use of the word ‘fallacy’ in economics, as it seems to be little more than a tool people beat others over the head with when they don’t want to engage in critical thinking. Often, the so called ‘fallacies’ that are trotted out in economics are used inappropriately and the stories told to justify them require exploration.

Here I present the three worst offenders – note that I don’t disagree with all of them entirely, but just wish to highlight that the story is often not a simple as it seems and cannot be captured by simply shouting down your opponent with the word ‘fallacy’.

The Broken Window

In the essay ‘What Is Seen And What Is Not Seen‘, a 19th Century economist named Frederic Bastiat  wrote a story about a boy who breaks a shopkeeper’s window. In replacing it, the shopkeeper gives money to the glassman, and the town observes that the broken window provided a boost to the local economy. However, Bastiat emphasises that this fallacy ignores the unseen fact that, had the window not been broken, the shopkeeper would have bought a new pair of shoes. Hence, there is no net gain for the economy.

Don’t get me wrong, it’s an important essay with an important point: if you look at only the benefits of government programs, you miss the hidden costs – where the tax money would otherwise have been spent, money unspent due to tariffs, and so forth.

However, Bastiat makes two hidden assumptions:

(1) All money that is spent would have been spent elsewhere – call this Say’s Law.

(2) That the replacement for the proverbial broken window is not better in any way.

The problem with (1) can be demonstrated by supposing that the shopkeeper was, in fact, not going to spend his money at all – in that case there would have been a boost to the economy. Not the best way to boost income, perhaps, but an income boost nonetheless. If the economy is not at full employment then spending more money does not require that you displace existing spending – to argue the opposite is to argue that private sector spending cannot increase employment either.

The problem with (2) can be illustrated by supposing that the shopkeeper’s window had been in a poor state to start with. In that case, replacing the window would have had a degree of benefit to the shopkeeper greater than in the more simple version of the story. After all, proponents the broken window fallacy often speak approvingly of creative destruction- replacing old capital and ideas with new, better capital and ideas. A similar logic applies – again, going around breaking things that seem worn out isn’t a suggested strategy for development, but it’s not as clear cut as it first seems. (‘Broken windows’ can also become a rationalisation for renovation – ‘it’s about time we redid the shop front anyway’, etc.)

Lump of Labour

This ‘fallacy’, as with the rest of labour economics, was born as a reaction to working class political movements in the late 19th century. It is basically an argument against shorter working hours – the claim is that you can’t simply ‘split up’ existing working hours, as the productivity of one sector of the economy has an impact elsewhere. Therefore, working people don’t understand econ101, etc.

The vacuousness of the fallacy can be seen in defenses of it, which seem to insist that higher productivity implies more income and more work. Take, for example Paul Krugman’s hot dog story:

But wait–what entitles me to assume that consumer demand will rise enough to absorb all the additional production? One good answer is: Why not?

Great. I’ve got a better answer: Why? Why not reduce working hours? What about limited natural resources? What if people don’t want twice as many hot dogs?

The blogger named ‘Sandwichman‘ appears to have made it his task to demolish this supposed ‘fallacy’, and has a website devoted solely to this cause. Sandwichman’s main point is that Lump of Labour proponents often mischaracterise their opponents as assuming there is a ‘fixed’ amount of work to be done in an economy, when of course they do no such thing. Discussions over the amount of work to be done is irrelevant to discussions of how that work should be distributed.

Correlation-Causation/Post Hoc Ergo Propter Hoc

Confusing correlation for causation, of course, is a fallacy, but this does not justify the mirror image delusion that correlation is meaningless. Often an observed correlation in the data has an implicit, intuitive causal link, such as alcoholism and recessions, corporate savings and unemployment, or, in the case of austerity, spending cuts/tax increases and a stagnating economy (I have seen some on the right shout down criticisms of austerity producing low growth as post hoc ergo propter hoc. Of course this is ridiculous – we know exactly why austerity produces low growth) .

If, as in the case of Steve Keen’s work on private debt, you have a clear theoretical link between two things, strong correlation, and the numbers changing in the right order (a decline in private debt acceleration portends lower growth), these things cannot be dismissed on grounds of correlation causation and post hoc ergo propter hoc.

Of course, there are many more – even the well established logical fallacies are prone to misuse and misconception (if you think about it, appeal to authority is historically quite a large component of scientific progress). Specifically, economic fallacies are generally an attempt to look for easy answers in a complex field, when the real story is often far more nuanced. It is important not to fall into this trap of lazy argumentation that often pervades the internet.


, , ,

  1. #1 by Paul Rosenberg (@PaulHRosenberg) on April 14, 2012 - 2:51 pm

    One usefull overall observation is that almost all logical fallacies are both fallacious and appealing at least in part because they represent valid hueristics that should be the starting point, not the end point of reasoning. Your third point can be read as a specific example of fleshing out this more general point.

    (Indeed, if one were to be a totally strict positivist, no amount of observations EVER prove anything more than correlation. Which is why I’m not a totally strict positivist. It’s just too silly.)

    • #2 by Unlearningecon on April 14, 2012 - 3:04 pm

      Good comment.

      I believe Mises argued something similar about positivism and correlation – basically, that historical relationships do not matter and a priori observations are the only way to go. Obviously that’s a slightly impractical premise.

  2. #3 by originalsandwichman on April 15, 2012 - 7:13 pm

    Yup. There’s also this from Pigou:

    “If it were a good ground for rejecting an opinion that many persons entertain it for bad reasons, there would, alas, be few current beliefs left standing!”

  3. #4 by Will on April 16, 2012 - 1:13 am

    That Krugman quotation comes from “The Accidental Theorist,” no? I have admired and respected Krugman for a long time, but those 90s essays attacking critics of neoliberalism are not the brightest spots in his work. I do think that, ironically, it is to those pieces that he owes his prominence. The Times wouldn’t have given a column to someone as unreservedly leftish as Krugman now is — especially back in 2000, when error reigned supreme.

    • #5 by Unlearningecon on April 16, 2012 - 2:42 pm

      I read somewhere (possibly on Sandwichman’s blog, but I can’t find it) that the book he was attacking – ‘One World, Ready or Not: The Manic Logic of Global Capitalism’ basically proved extremely prescient in its predictions for global capitalism. And yes, Krugman’s smug ‘economists know better’ attitude in that book reminds me of Bryan Caplan. In fairness, he has recently relaxed his position on free trade and would probably advocate shorter working hours.

  4. #6 by Nick on April 16, 2012 - 1:53 pm

    As a non-economist peering in I marvel at the fantasy land that economists live in. They can pretty much assume anything away and thus prove their point.

    Bastiat does not assume the shopkeeper’s money will not be spent. He says, “In brief, he would have put his six francs to some use or other for which he will not now have them.” He could spend it or he could save it. If he had his six francs already saved then his six francs are already in the economic system as loans or investments which would have to be liquidated to pay the glazier. The only one who wins from the broken glass is the glazier. The economic system sees no net benefit because for every winner we have an equivalent loser…somewhere.

    But it gets worse. The glazier is duped into thinking the market for glaziers is about to take off. He now hire an apprentice glazier to help him with this new work. Excited by his new job the apprentice tells all of his friends about it, who then rush to sign up at the local trade organization to also become glazier apprentices.

    After a month though the apprentice is let go because the master glazier can no longer afford him because he made a business miscalculation. The glazier students drop out. Misery everywhere.

    As Bastiat suggest, look at the unseen. Money moving around changing hands might look like real activity, but perhaps it is not.

    • #7 by Unlearningecon on April 16, 2012 - 2:46 pm

      He could spend it or he could save it

      It’s a common misconception that the identity ‘savings=investment’ means that savings create investment (put it down to confusion over the exact meaning of the word ‘savings’). If the glazier simply keeps the money under his mattress it will do no good.

      Your story about the glazier’s apprentice is a possibility, but really it’s just speculation. I could say the same thing about the shoemaker – Austrians or ‘free market’ proponents have a tendency to assume that everything that takes place in the ‘market’ is real and sustainable, but there is no neutral policy. From how property rights are defined to luck to what constitutes fraud, to immigration restrictions and limited liability laws, everything has an impact on the workings of a market economy.

  5. #8 by originalsandwichman on April 17, 2012 - 12:17 am

    And why would the shopkeeper be buying a new pair of shoes? Because his old ones wore out! So we are supposed to believe that wearing out shoes is good for the economy but breaking windows isn’t?

    I dunno. If the shopkeeper’s shoes hadn’t worn out (and the window hadn’t been broken), maybe he could have taken the rest of the day off and gone fishing. Surely THAT would have been good for the economy? No? It wouldn’t have added any value to the GDP? Sacre bleu! Incroyable! So the economy is only good when things wear out but not when they get broken or when they don’t wear out. Hmmmm. It is all too complex for me to understand. So much to not be seen!

  6. #9 by Justin Oliver on April 29, 2012 - 6:00 pm

    I think you are right that Bastiat’s lesson is simply to emphasize that the hidden and unintended costs of a policy should also be used to calculate the impact of that policy.

    I think I have a have different premise as far as what is the proper measure of the status of an economy and what would be considered “a boost to the economy,” as you called it. I would regard wealth as a better indicator of the overall economy than economic activity. So while GDP growth is typically an indicator of a prospering economy, I think it would depend on the underlying cause of that activity.

    If wealth is the proper indicator of an economy’s condition, that presents some measurement challenges to policy planner, and that is why I think they are more comfortable with indirect and less ambiguous measurements of the economy, like GDP, which are then unfortunately treated as ends in-and-of themselves.

    I would be interested in any feedback you can offer.

    Thanks for the interesting post and opportunity for discussion!

  1. Broken Windows and Idle Money | Economic Thought