The permanent income hypothesis explains consumption as a response to expected sustainable resources, with temporary and persistent income changes treated differently.
The permanent income hypothesis (PIH) is an economic model in which households base consumption mainly on the income or resources they expect to sustain over time, rather than reacting one-for-one to current income. A temporary bonus and a lasting pay increase can therefore produce different spending responses even when the first-year dollar amount is the same.
“Permanent” does not mean guaranteed forever. It means the persistent component of expected resources under the model. Households can misjudge persistence, face borrowing limits, or revise expectations as new information arrives.
A common teaching representation separates measured income into two components:
Where:
A simplified consumption relation is:
Here, (k) summarizes factors such as interest rates, preferences, age, horizon, and expected income growth. This compact equation is not a complete household budget. Modern applications usually consider wealth, uncertainty, borrowing constraints, taxes, family composition, and other information.
The difficult step is not the algebra. It is deciding which part of an income change is likely to persist. A one-time tax refund is usually more transitory than a credible promotion, but even a new salary can disappear through unemployment, business failure, illness, or inflation.
Suppose a household receives news about future resources:
| Income event | Likely interpretation | Basic PIH prediction |
|---|---|---|
| One-time bonus | Mostly transitory | Smaller immediate consumption response |
| Credible permanent raise | More persistent | Larger consumption response spread over time |
| Temporary unemployment | Negative transitory shock | Use saving or credit to limit the consumption decline, if feasible |
| Unexpected long-term disability | Persistent negative shock | Larger downward revision to planned consumption |
| Anticipated annual payment | Already reflected in expectations | Little response when the payment actually arrives |
The last row matters. Consumption can change when information becomes known, not only when cash enters an account. If a payment was fully expected, receiving it may confirm an existing plan rather than create a new spending decision.
Assume two households each receive 10,000 of additional after-tax income. For illustration only:
2,000 of it during the measured period.8,000 during the measured period.Their measured marginal propensities to consume are:
The numbers are assumptions, not universal estimates. The example isolates the model’s central idea: the perceived persistence of the income change can affect how much consumption changes. Liquidity, debt, confidence, timing, and household needs could reverse or weaken the contrast.
Current income is observable after it is received. Permanent income is an expectation and cannot be read directly from a pay stub. Economists infer it using income histories, anticipated changes, household surveys, asset positions, and models.
That distinction creates a measurement problem. If consumption barely reacts to a payment, the household may view it as temporary, may have anticipated it, may be paying debt, or may simply delay spending. One observation does not identify the reason.
The model also does not imply that households always save temporary gains in a bank account. Saving can include paying down debt, acquiring assets, or leaving cash unspent under a particular accounting definition.
The two frameworks share consumption-smoothing logic but organize it differently.
| Feature | Permanent-income hypothesis | Life-cycle hypothesis |
|---|---|---|
| Primary emphasis | Persistent versus transitory resources | Resources and needs over a finite lifetime |
| Typical question | Is this income change expected to last? | How should resources be allocated across work and retirement years? |
| Treatment of age | Can affect expectations and (k), but is not the headline distinction | Explicitly central to earnings, horizon, and asset use |
| Common empirical use | Analyze responses to income shocks or policy payments | Analyze age profiles of saving, wealth, and consumption |
They can operate together. A household near retirement may revise lifetime resources after a persistent earnings shock and also adjust saving because its remaining work horizon is short.
Fiscal policy: The consumption effect of a transfer or tax change depends partly on whether recipients expect it to persist and whether they can smooth consumption. A single multiplier assumption can miss important differences among households.
Earnings analysis: Temporary overtime, a cyclical bonus, and a recurring salary change are economically different even if they produce the same current cash flow.
Credit analysis: A borrower may support current payments using volatile or temporary income. Underwriters and analysts need to assess recurrence rather than treating every recent inflow as sustainable.
Economic forecasting: Consumption may remain stable during a short income interruption or change before income does when households receive credible information about future resources.
Borrowing constraints: A household with little liquidity and no credit may need to cut consumption when current income falls, even if the loss is expected to be temporary.
Precautionary saving: Uncertainty can make households save more than the basic certainty model predicts. The response depends on risk tolerance, insurance, and the distribution of possible outcomes.
Expectation errors: A supposedly permanent raise may disappear, while a temporary shock may last. Plans adjust as beliefs change.
Habits and durable goods: Consumption may respond slowly because households dislike abrupt changes or because large purchases are lumpy and easy to postpone.
Household heterogeneity: Wealth, debt, age, family obligations, and access to credit affect marginal spending responses. An aggregate average can hide large differences.
Measurement: Income, consumption, and saving are defined differently across surveys and national accounts. Timing differences can make a measured response look larger or smaller than the underlying decision.