Spending $42,000 of a $50,000 income is an APC of 0.84
The average propensity to consume is the fraction of income that goes on consumption: total consumption divided by total income over the same period. A household that takes home $50,000 a year and spends $42,000 of it has an APC of 42,000 ÷ 50,000 = 0.84 — it spends 84 cents of every dollar.
APC = C ÷ Y
Economists measure it against disposable income, which is income after taxes and transfers, because that is the income a household can actually choose to spend or save. Using gross income instead makes every APC look lower than it is, since taxes sit in the denominator but can never be spent.
Worked example
C = $42,000 and Y = $50,000
APC = 42,000 ÷ 50,000
APC = 0.84
(so 84 cents of each dollar is spent)
Every dollar is either spent or saved, so APC and APS always add to 1
Income that is not consumed is saved, by definition: Y = C + S. Divide both sides by Y and the identity becomes 1 = C/Y + S/Y, which is APC + APS = 1. An APC of 0.84 therefore fixes the APS at 0.16 without any further information.
Y = C + S → APC + APS = 1
The identity also explains the sign of saving. When consumption exceeds income the APC is above 1 and the APS is negative — the household is dissaving, running down savings or borrowing to cover the gap. The calculator reports saving as a signed figure rather than stopping at zero for that reason.
With a = $8,000 and b = 0.68, APC falls from 1.08 on $20,000 to 0.76 on $100,000
Keynes argued that consumption rises with income, but by less than income does, and that a household spends something even at zero income. In the linear consumption function C = a + bY the average propensity is APC = a/Y + b: the fixed a is spread over more income as income rises, so the APC falls toward the marginal propensity b and never reaches it while a is positive.
APC = a ÷ Y + b
With autonomous consumption of $8,000 and an MPC of 0.68, a household on $20,000 spends $21,600 — an APC of 1.08, paid for from savings or credit. On $50,000 it spends $42,000, an APC of 0.84; on $100,000 it spends $76,000, an APC of 0.76. The APC crosses 1 at the break-even income a ÷ (1 − b) = $25,000, where saving is exactly zero.
One function at three incomes
Y = 20,000: 8,000 ÷ 20,000 + 0.68 = 1.08
Y = 50,000: 8,000 ÷ 50,000 + 0.68 = 0.84
Y = 100,000: 8,000 ÷ 100,000 + 0.68 = 0.76
(APC = 1 where Y = 8,000 ÷ 0.32 = 25,000)
APC averages over all income; the MPC prices only the next dollar
The average and the marginal propensities answer different questions. The APC asks what share of total income is spent; the MPC asks what share of an extra dollar would be spent. In the linear function they differ by exactly a/Y, so the APC sits above the MPC whenever autonomous consumption is positive — 0.84 against 0.68 in the example, a gap of 8,000 ÷ 50,000 = 0.16.
The distinction decides fiscal arithmetic. A stimulus payment is extra income, so its effect runs through the MPC; putting an APC into the multiplier 1 ÷ (1 − MPC) would give 1 ÷ 0.16 = 6.25 instead of 1 ÷ 0.32 = 3.125, doubling the answer. The marginal propensity to consume is estimated from two observations of spending and income, not from one.
US households saved 3.0% of disposable income in July 2026, against 31.8% in April 2020
The US Bureau of Economic Analysis publishes the personal saving rate — personal saving as a share of disposable personal income — every month, and it is the closest official series to an economy-wide APS. It stood at 3.0% in July 2026 and 6.2% in December 2019. Its record, 31.8% in April 2020, came when lockdowns cut spending far faster than income fell; its low, 1.4%, was in July 2005.
The official rate is not exactly 1 − C/Y. BEA subtracts all personal outlays from disposable income, and outlays include interest payments and transfer payments as well as consumption spending, so consumption divided by disposable income comes out a little below one minus the saving rate. For a household budget the gap rarely matters; for comparing a computed APC with the published series it does.
An APC above 1 is dissaving, and it is normal at both ends of a working life
A household spending more than its income finances the difference from past saving, from borrowing, or from selling assets. Franco Modigliani’s life-cycle hypothesis predicts the pattern: young adults and retirees tend to have APCs above 1 while people in their peak earning years save, so an economy’s APC depends on its age structure as well as on its income.
A single year above 1 is not by itself a warning — a year of study, a new child or a temporary loss of income all produce one — but an APC that stays above 1 means net worth is falling, and the calculator’s negative saving figure is the size of that fall each period.
Most wrong APCs come from the inputs, not the division
The arithmetic is one division. The errors are in what goes into it.
- Dividing by gross income. Taxes cannot be spent, so the APC is measured against disposable income; gross income makes every APC look lower.
- Mixing periods. Monthly spending over annual income gives an APC twelve times too small; both figures must cover the same period.
- Treating the APC as the MPC. The average share of income spent is not the share of an extra dollar spent, and only the marginal share belongs in a multiplier.
Sources & References
Figures on this page are checked against primary, authoritative sources. Links open in a new tab.