Permanent Income Hypothesis

The permanent income hypothesis (PIH) is a theory of consumption which holds that households base their spending on their "permanent" income, the average income they expect to receive over the long term, rather than on their current income. Temporary ups and downs, such as a one-off bonus, a short spell of unemployment or a good harvest, should have little effect on spending; they are mostly saved or absorbed by drawing on savings. Household spending on goods and services is what economists call consumption, the largest part of demand in most economies, which is why the theory attracted so much attention.

The theory was set out by Milton Friedman in A Theory of the Consumption Function (1957), work cited when he received the 1976 Nobel prize. It challenged the Keynesian consumption function, in which spending depends mainly on this year's income. It is a close relative of the life-cycle hypothesis of Franco Modigliani and Richard Brumberg; both assume people look ahead and smooth their spending, and the two are often combined in modern models.

Think of a farmer whose harvest is large in some years and poor in others. A sensible farmer does not throw a feast after every good year and go hungry after every bad one. Instead the family eats about the same each year, storing grain in good years and using the store in lean ones. Their spending follows their typical harvest, not this year's.

Friedman argued that most households behave the same way with money. A salesperson with a big commission one month, or a worker who receives a one-off tax refund, will treat it as a lucky extra and spend only a little of it at once. A permanent pay rise, by contrast, changes what they can afford for years, so they adjust their spending fully.

Friedman split measured income into two parts and related spending to the permanent part:

• $Y$ is measured, or current, income.
• $Y^{P}$ is permanent income, the expected long-run average.
• $Y^{T}$ is transitory income, the temporary deviation.
• $C$ is consumption.
• $k$ is the share of permanent income consumed, which depends on interest rates, tastes and wealth.

Suppose $k = 0.9$ and permanent income is 50,000, so the household spends 45,000. If it receives a one-off bonus of 5,000, its permanent income rises only by the interest that sum could earn forever: at a 4 percent rate, $0.04 \times 5{,}000 = 200$. Spending rises by $0.9 \times 200 = 180$, and the rest of the bonus is saved. If instead the household gets a lasting 5,000 pay rise, spending rises by $0.9 \times 5{,}000 = 4{,}500$.

The theory resolved a puzzle in the data of the 1940s. Household surveys showed that richer families saved a larger share of income, which suggested that saving rates should climb as countries grew richer. Yet long-run figures assembled by Simon Kuznets showed the US saving rate roughly stable over decades. Friedman's answer was that in any one year, many high earners are having an unusually good year and save the windfall, while low earners are having a bad year; over the long run, consumption is a stable share of permanent income. The PIH also implies that the marginal propensity to consume out of temporary income is small, so the multiplier from short-lived tax cuts should be weak. Robert Hall showed in 1978 that, with rational expectations, the theory predicts consumption changes should be largely unpredictable, a "random walk".

Empirical work has found that spending responds to temporary income more than the pure theory allows, a finding known as excess sensitivity. Studies of US tax rebates in 2001 and 2008 found households spent a significant part of the money within a few months, and responses are strongest among households with little cash or credit available. Economists explain this by liquidity constraints (people who cannot borrow cannot smooth), precautionary saving and "hand-to-mouth" households who spend what they receive; John Campbell and Gregory Mankiw estimated in 1989 that a large share of income goes to such rule-of-thumb consumers. Behavioral economists add mental accounting and limited foresight. The debate matters for policy: whether cash transfers in a recession boost aggregate demand depends on who receives them. Modern models therefore mix forward-looking households with constrained ones, keeping Friedman's core idea while accepting that many people cannot or do not smooth perfectly. The hypothesis also underpins Ricardian equivalence, which applies the same forward-looking logic to government debt.

• Nobel Prize: Milton Friedman, facts
• National Bureau of Economic Research: A Theory of the Consumption Function