Macro by Mark

Unlock Full Macro Model Library with Starter.

This feature is exclusively available to Starter, Research, and Pro. Upgrade when you need this workflow, review pricing, or send a question before changing plans.

Upgrade to Starter View pricing Questions?Already subscribed? Sign in

What you keep on Free

Create and edit one custom board
Use up to 3 widgets on each Free board
Browse indicators and calendar

← Models Overview History Concepts Models Schools

Friedman's Permanent Income Hypothesis: consumption depends on permanent income, not current income. C = k * Y_P, where permanent income Y_P is an adaptive-expectations weighted average of current and past income.

Derivation

Step-by-step mathematical derivation with typeset equations and expandable detail.

Sections

Friedman's income decomposition

Milton Friedman's permanent income hypothesis (PIH), published in 1957, decomposes measured income $Y$ into two components: permanent income $Y_P$ and transitory income $Y_T$ . Permanent income is the stable, long-run average income that a household expects to earn over its lifetime, determined by human capital, wealth, and persistent productivity. Transitory income is a mean-zero random deviation from that trend: an unexpected bonus, a temporary layoff, a one-time tax rebate.

The decomposition is additive: $Y = Y_P + Y_T$ , with $E[Y_T] = 0$ and $\text{Cov}(Y_P, Y_T) = 0$ . Friedman argued that the same decomposition applies to consumption: $C = C_P + C_T$ , where permanent consumption $C_P$ is proportional to permanent income and transitory consumption $C_T$ is a random, mean-zero fluctuation uncorrelated with transitory income. The core claim is that households base their consumption decisions on permanent income, not current measured income.

Y = Y_P + Y_T, \quad E[Y_T] = 0

Measured income equals permanent income plus a mean-zero transitory component.

C = C_P + C_T, \quad E[C_T] = 0

Measured consumption likewise decomposes into a permanent and transitory component.

C_P = k \cdot Y_P

Permanent consumption is proportional to permanent income, where $k$ depends on the interest rate, time preference, and the wealth-to-income ratio.

Marginal propensity to consume out of each component

The PIH's most powerful prediction concerns the marginal propensity to consume (MPC) out of different income types. Because households gear consumption to permanent income, the MPC out of a permanent income increase is $k$ (close to one over the long run). A permanent raise of $\$ 1{,}000 $per year raises annual consumption by roughly$ \ $1{,}000 \cdot k$ . By contrast, the MPC out of transitory income is approximately zero: a one-time windfall (lottery win, tax rebate) is largely saved or used to pay down debt, because it does not alter the household's view of its long-run income.

This resolves the apparent contradiction between cross-sectional and time-series consumption data that puzzled economists before Friedman. In cross-sectional data, high-income households appear to save a larger fraction of income, suggesting a declining average propensity to consume (APC). The PIH explains this: high measured income often reflects positive transitory income, which is saved rather than consumed, making the APC appear low. Over time, as transitory components wash out, the APC stabilizes at $k$ , consistent with the roughly constant long-run APC observed in aggregate data.

\frac{\partial C}{\partial Y_P} = k \approx 1

MPC out of permanent income: close to unity, as households consume nearly all of a lasting income change.

\frac{\partial C}{\partial Y_T} \approx 0

MPC out of transitory income: near zero, as windfall income is saved rather than consumed.

Forward-looking behavior and consumption smoothing

The PIH implies that rational, forward-looking households smooth consumption over time. Rather than letting consumption track the ups and downs of current income, households use saving and borrowing to maintain a stable standard of living anchored to their permanent income. When income is temporarily high, they save the excess. When income is temporarily low, they draw down savings or borrow against future earnings. The result is a consumption path that is much smoother than the income path.

Robert Hall (1978) formalized this insight by showing that under rational expectations and a constant real interest rate equal to the rate of time preference, the PIH implies that consumption follows a random walk: $C_{t+1} = C_t + \varepsilon_{t+1}$ . The best forecast of next period's consumption is this period's consumption, because all available information about permanent income is already incorporated. Only genuinely new information (the innovation $\varepsilon_{t+1}$ ) causes consumption to change. This random-walk result has become one of the most-tested predictions in empirical macroeconomics, and while it does not hold perfectly (due to liquidity constraints, precautionary saving, and habit formation), it remains a central benchmark.

C_t = r \cdot W_t + Y_P

Consumption equals the annuity value of total wealth: interest on accumulated assets plus permanent labor income.

C_{t+1} = C_t + \varepsilon_{t+1}, \quad E_t[\varepsilon_{t+1}] = 0

Hall's random-walk result: consumption changes are unpredictable under rational expectations and the PIH.

Back to Graph Compare scenarios Overview

Theory-Based Models

Loading Theory-Based Models