Milton Friedman’s pet theory finally shown to be wrong

28 Jan, 2017 at 15:36 | Posted in Economics | 7 Comments

150514006_4Milton Friedman’s Permanent Income Hypothesis (PIH) says that people’s consumption isn’t affected by short-term fluctuations in incomes since people only spend more money when they think that their life-time incomes change. Believing Friedman is right, mainstream economists have for decades argued that Keynesian fiscal policies therefore are ineffectual.

As shown over and over again for the last three decades, empirical facts totally disconfirm Friedman’s hypothesis. The final nail in the coffin is new research from Harvard:

Unemployment is a particularly good setting for testing alternative models of consumption because it causes such a large change in family income. A literature starting with Akerlof and Yellen (1985), Mankiw (1985) and Cochrane (1989) has argued that because ignoring small price changes has a second-order impact on utility, a rule of thumb such as setting spending changes equal to income changes may be “near-rational.” More recently, many researchers have documented evidence of an immediate increase in spending in response to tax rebates and similar one-time payments …

We compare the path of spending during unemployment in the data to three benchmark models and find that the buffer stock model fits better than a permanent income model or a hand-to-mouth model …

To summarize, we find that families do relatively little self-insurance when unemployed as spending is quite sensitive to current monthly income. We built a new dataset to study the spending of unemployed families using anonymized bank account records from JPMCI. Using rich category-level expenditure data, we find that work-related expenses explain only a modest portion of the spending drop during unemployment. The overall path of spending for a seven-month unemployment spell is consistent with a buffer stock model where agents hold assets equal to less than one month of income at the onset of unemployment. Because unemployment is such a large shock to income, our finding that spending is highly sensitive to income overcomes the near-rationality critique applied to prior work. Finally, we document a puzzling drop in spending of 11% in the month UI benefits exhaust, suggesting that families do not prepare for benefit exhaustion.

Peter Ganong & Pascal Noel

So — now we know that consumer behaviour is influenced by short-term fluctuations in incomes and that this is true even if consumers know that their situation may well change in the future.

Since almost all modern mainstream macroeconomic theories are based on PIH –standardly used in formulating the consumption Euler equations that make up a vital part of ‘modern’ New Classical and New Keynesian macro models — these devastating findings are extremely problematic.main-qimg-1b106c1df117b1c788bd8f4089d394e3-c

In many modern macroeonomics textbooks one explicitly adapt a ‘New Keynesian’ framework, adding price rigidities and a financial system to the usual neoclassical macroeconomic set-up. Elaborating on these macromodels one soon arrives at specifying the demand side with the help of the Friedmanian Permanent Income Hypothesis and its Euler equations.

But if people — not the representative agent — at least sometimes can’t help being off their labour supply curve — as in the real world — then what are these hordes of Euler equations that you find ad nauseam in these ‘New Keynesian’ macromodels gonna help us?

My doubts regarding macro economic modelers’ obsession with Euler equations is basically that, as with so many other assumptions in ‘modern’ macroeconomics, Euler equations, and the PIH that they build on, don’t fit reality.

In the standard neoclassical consumption model people are basically portrayed as treating time as a dichotomous phenomenon today and the future — when contemplating making decisions and acting. How much should one consume today and how much in the future? The Euler equation used implies that the representative agent (consumer) is indifferent between consuming one more unit today or instead consuming it tomorrow. Further, in the Euler equation we only have one interest rate, equated to the money market rate as set by the central bank. The crux is, however, that — given almost any specification of the utility function – the two rates are actually often found to be strongly negatively correlated in the empirical literature!

From a methodological pespective yours truly has to conclude that these kind microfounded macroeconomic models are a rather unimpressive attempt at legitimizing using fictitious idealizations — such as PIH and Euler equations — for reasons more to do with model tractability than with a genuine interest of understanding and explaining features of real economies. Mainstream economists usually do not want to get hung up on the assumptions that their models build on. But it is still an undeniable fact that theoretical models building on piles of known to be false assumptions — such as PIH and the Euler equations that build on it — in no way even get close to being scientific explanations. On the contrary. They are untestable and hence totally worthless from the point of view of scientific relevance.

Ganong’s and Noel’s research finally shows that mainstream macroeconomics, building on the standard neoclassical consumption model with its Permanent Income Hypothesis and Euler equations, has to be replaced with something else. Preferably with something that is both real and relevant, and not only chosen for reasons of mathematical tractability  or for more or less openly market fundamentalist ideological reasons.


  1. The empirical proof of something that we already knew is, after all, welcome! Now, perhaps mainstream will claim that reality must change to safe their models…

  2. There is also the question of a learning process complicating the spending dichotomy issue where a sudden large drop of income associated with immediate unemployment (companies going bankrupt in a down turn) Individuals will learn to adjust their spending pattern to much reduced income over time especially where there are benefits granted. However I would have thought there would be data available in sufficient volume over time to enable ‘laws’ to be inferred?

  3. Friedman and the cluelessness of fake scientists
    Comment on Lars Syll on ‘Milton Friedman’s pet theory finally shown to be wrong’
    Lars Syll comments on the refutation of the permanent income hypothesis (PIH): “My doubts regarding macro economic modelers’ obsession with Euler equations is basically that, as with so many other assumptions in ‘modern’ macroeconomics, Euler equations, and the PIH that they build on, don’t fit reality. … But it is still an undeniable fact that theoretical models building on piles of known to be false assumptions — such as PIH and the Euler equations that build on it — in no way even get close to being scientific explanations. On the contrary. They are untestable and hence totally worthless from the point of view of scientific relevance.”
    This is true but not new. Economics is built since 200+ years on false assumptions. In other words, economics is what Feynman famously called a cargo cult science. Friedman is only one fake scientist in the long line that stretches from the storyteller Adam Smith to the loudspeaker Paul Krugman.
    It is of utmost importance to distinguish between political and theoretical economics. The main differences are: (i) The goal of political economics is to successfully push an agenda, the goal of theoretical economics is to successfully explain how the actual economy works. (ii) In political economics anything goes; in theoretical economics the scientific standards of material and formal consistency are observed.
    Theoretical economics has to be judged according to the criteria true/false and NOTHING else. Scientific truth is well-defined by material and formal consistency. But theoretical economics had been hijacked from the very beginning by the agenda pushers of political economics. Can there be the slightest doubt that Smith, Ricardo, Malthus, Marx, Keynes, Hayek, Friedman, Krugman, Syll and almost everybody in-between falls into the category of political economist or fake scientist?
    Political economics has produced NOTHING of scientific value in the last 200+ years. Since the founding fathers economists claim to do science but they have never risen above the level of opinion, belief, wish-wash, storytelling, soap box propaganda, and sitcom gossip. During his lifetime Milton Friedman produced plain scientific rubbish and never realized that his axiomatic foundations were false.
    Friedman’s policy prescriptions are regarded as outdated but the representative economist still applies Friedman’s false premises. Orthodox economics is built since Jevons/Walras/Menger upon this set of foundational propositions, a.k.a. axioms: “HC1 economic agents have preferences over outcomes; HC2 agents individually optimize subject to constraints; HC3 agent choice is manifest in interrelated markets; HC4 agents have full relevant knowledge; HC5 observable outcomes are coordinated, and must be discussed with reference to equilibrium states.” (Weintraub)
    Methodologically, these premises are forever unacceptable. The microfoundations approach has already been dead in the cradle. The axiom set contains THREE NONENTITIES: (a) constrained optimization (HC2), (b) rational expectations (HC4), (c) equilibrium (HC5). Every theory/model that contains a nonentity is A PRIORI false. So, not only Friedmanian economics but orthodox economics from Jevons/Walras/Menger to DSGE/RBC/New Keynesianism is axiomatically false. There is no need to go into the details of the analytical superstructure, axiomatically false is the death sentence for a paradigm.
    There is not much use to criticize a political clown like Friedman much longer. To get out of failed economic theory requires nothing less than a full-blown paradigm shift from false Walrasian microfoundations and false Keynesian macrofoundations to entirely NEW macrofoundations. Everything else is a continuation of fake science.
    Egmont Kakarot-Handtke

  4. I’m not sure that you understand this criticism. What the authors suggest works is “the buffer stock model”, which is a PIH model augmented with idiosyncratic risk. There are as many Euler equations there as in the PIH.

    • Yes, and did I quote and comment on the Ganong/Noel alternative …
      If authors show A to be false, you don’t necessarily have to share their views on alternative theory B.

      • If the same evidence that rejects A supports B then it’s hard to ignore it …

        • That’s a big IF …

Sorry, the comment form is closed at this time.

Blog at
Entries and comments feeds.