Income up $10,000, spending up $6,800: an MPC of 0.68
The marginal propensity to consume is the share of an extra dollar of income that is spent. It is a ratio of changes, not of levels: the change in consumption divided by the change in income. A household whose income rises from $40,000 to $50,000 and whose spending rises from $35,200 to $42,000 has an MPC of 6,800 ÷ 10,000 = 0.68.
MPC = ΔC ÷ ΔY
On a graph of consumption against income the MPC is the slope of the line joining the two points. Keynes’s “fundamental psychological law” was that this slope lies between 0 and 1 — people spend part of a raise and save the rest — and the whole expenditure multiplier rests on that one number.
Worked example
ΔC = 42,000 − 35,200 = 6,800
ΔY = 50,000 − 40,000 = 10,000
MPC = 6,800 ÷ 10,000
MPC = 0.68
The 32 cents not spent is the MPS, so MPC + MPS = 1
An extra dollar of disposable income is either spent or saved, so the marginal propensities add to one exactly as the average ones do. An MPC of 0.68 means an MPS of 0.32. In the household above, saving rose from $4,800 to $8,000 — a change of $3,200 on $10,000 of extra income.
MPC + MPS = 1
The average propensities tell a different story from the same data: the APC was 0.88 before the raise and 0.84 after. The APC fell because the household spent less than its average share of the raise, which is exactly what an MPC below the APC means.
An MPC of 0.68 turns $100 of new spending into $312.50 of output
When one person spends an extra dollar it becomes someone else’s income, and they spend the MPC share of it, and so on. The rounds form a geometric series — 1 + 0.68 + 0.68² + … — whose sum is 1 ÷ (1 − 0.68) = 3.125. That is the spending multiplier: $100 of new government purchases or investment raises output by $312.50 in the simplest model.
k = 1 ÷ (1 − MPC); k_T = −MPC ÷ (1 − MPC)
A tax cut works one step later. It adds to disposable income, of which only the MPC share is spent in the first round, so its multiplier is −MPC ÷ (1 − MPC) = −2.125: a $100 tax cut raises output by $212.50, and a $100 tax rise lowers it by the same. The table on this page follows a $100 injection round by round at your MPC.
Multipliers at MPC = 0.68
k = 1 ÷ (1 − 0.68) = 3.125
k_T = −0.68 ÷ 0.32 = −2.125
$100 of spending → $312.50 of output
Households spent 12–30% of the 2008 stimulus payments on non-durables within three months
Measured MPCs depend heavily on what kind of income changes. Parker, Souleles, Johnson and McClelland used the randomised timing of the 2008 US economic stimulus payments to estimate that households spent about 12–30% of the payment on non-durable goods in the three months it arrived, and roughly 50–90% once durable purchases, mainly vehicles, are counted.
Liquidity is the main divider. Households with little cash or credit spend a much larger share of a windfall than households with savings to fall back on, which is why the same stimulus produces different multipliers depending on who receives it.
A raise is spent more readily than a one-off bonus of the same size
Milton Friedman’s permanent-income hypothesis says households base spending on the income they expect over the long run. A permanent raise changes that expectation and is largely spent; a one-off bonus hardly changes it and is largely saved. The MPC measured from a temporary change is therefore lower than the MPC measured from a permanent one, even for the same household.
That is why two observations should be chosen with care. If the second year’s income was lifted by a one-off payment, the MPC computed here describes the response to a windfall, not to a lasting change in income.
An MPC above 1 or below 0 usually means something other than income moved
Within Keynes’s model the MPC lies between 0 and 1. Measured values outside that range are common in household data and nearly always mean the two observations differ in more than income: a car bought in the second year, a change in interest rates, a move, or an income fall that the household did not cut spending to match.
At an MPC of 1 or more the multiplier formula breaks — the series never converges — and the calculator reports that no finite multiplier exists rather than printing a meaningless number.
Sources & References
Figures on this page are checked against primary, authoritative sources. Links open in a new tab.
What the calculation assumes
- Both observations are for the same household or economy, over periods of the same length.
- The change in consumption was caused by the change in income, not by prices, interest rates or wealth.
- The multipliers are the closed-economy, saving-only textbook forms.
Where it stops
- Two observations measure one slope; a noisy year or a one-off purchase can move it a long way.
- The MPC out of a temporary windfall is usually lower than the MPC out of a permanent raise.
- Real multipliers are smaller than these because taxes and imports leak spending too.