One of the core pieces of modern macroeconomic theory, handed down to us by the great Milton Friedman, probably missed the mark.

And now it might be on the way out.

And this shift has big implications for how we think about economic policy and finance.

The idea is called the permanent income hypothesis (PIH).

Friedman first put it on paper in 1957, and it still holds enormous sway in the econ profession.

The PIH says that people's consumption doesn't depend on how much they earn today, but on how much they expect to earn over their lifetime.

If a one-time windfall of money drops into your lap, says Friedman's theory, you won't rush out and spend it all -- you'll stick it in the bank, because you know the episode won't be repeated.

But if you get a raise, you might start spending more every month, because the raise was a signal that your earning power has increased for the long term.

That assumption about human behavior has huge implications for policy.

If true, the PIH means that the effectiveness of a fiscal stimulus is likely to be a lot lower than economists thought in the 1960s.

If the government tries to goose spending by mailing people checks, people will just deposit the money in the bank, instead of going out and consuming.

It's also important for finance.

Lots of academic theories are based on the PIH.

Friedman's idea says that consumers want to smooth out their consumption -- they don't like dips.

So in theory people will spend a lot for financial assets that pay off during recessions, allowing them to avoid tightening their belts.

PIH is so dominant that almost all modern macroeconomic theories are based on it.

They enshrine the idea with a formula called a "consumption Euler equation," which has appeared in the vast majority of academic macro models during the past few decades.

Those are the models that many central bankers use to set interest rates.

So it's not much of an exaggeration to say that Friedman's PIH is the cornerstone of modern macroeconomic theory.

Unfortunately, there's just one small problem -- it's almost certainly wrong.

Not completely wrong, mind you, just somewhat wrong.

There probably are a lot of consumers out there who do behave just the way that Friedman imagined.

But the problem is, there are a lot of others who act very differently.

Slowly, economists have been building up evidence that the latter group is important and sizable.

Early tests showed support for the PIH.

But in 1990, economists John Campbell and Greg Mankiw estimated that only about half of consumers obeyed Friedman's principle.

The rest, they said, probably consume hand-to-mouth -- if they get a bonus at work, or a big tax refund, or a government stimulus check, they go out and eat at nice restaurants, or buy more home furnishings, or just spend more.

A blow to the mathematical version of the theory came in 2006, when Georgetown economists Matthew Canzoneri, Robert Cumby and Behzad Diba wrote a paper testing the consumption Euler equation directly against real financial data -- something that, for reasons that escape me, no economist seems to have actually tried before.

The equation says that when interest rates are high, people save more and consume less -- this is the way they smooth their consumption, as Friedman predicted.

But Canzoneri et al. found that the opposite is true -- for whatever reason, the fact is that people tend to consume more when rates are high.

Oops.

Anyway, research continues to come out that's at odds with the PIH.

A recent study by Northwestern's Luenz Kueng found that when Alaskans got a sudden payout of money from the Alaska Permanent Fund (the state's wealth fund, which collects money from natural resource industries), they went and consumed the windfall immediately instead of sticking it in the bank like Friedman would predict.

That kind of random natural experiment, which is becoming more and more popular in the econ profession, is powerful and simple evidence.

The mounting evidence against the PIH -- the papers I cited are only a small sampling -- is causing economists to cast around for an alternative.

My Bloomberg View colleague Narayana Kocherlakota recently blogged that:

The choices made 25-40 years ago - made then for a number of excellent reasons - should not be treated as written in stone or even in pen.

By doing so, we are choking off paths for understanding the macroeconomy.

Kocherlakota thinks macroeconomists should set aside their big, complex formal models of the economy, since these elaborate constructions are built on a foundation that probably doesn't describe reality all that well.

He recommends that economists go back to the drawing board, and look around for new, more accurate kernels of insight with which to build the theories of tomorrow.

In the meantime, we should all recognize that Milton Friedman's ideas might have been too influential.

His impact on economics was deep and lasting, but this theory, at least, hasn't stood the test of time.