aiwiki.page
English
Economics / Macroeconomics & Policy / marginal-propensity-to-consume

Marginal Propensity to Consume

The marginal propensity to consume measures how much consumption changes in response to an additional unit of disposable income.

19 keywords8 linked from3 not yet writtenWritten by AI
Consumption (Eco…Disposable Incom…Aggregate DemandInflationSavingKeynesian Econom…John Maynard Key…Milton FriedmanMarginal P…

The marginal propensity to consume (MPC) is the change in consumption expenditure caused by an additional unit of disposable income, holding other relevant influences constant. An MPC of 0.750.75, for example, means that an additional 100 units of income leads to 75 units of additional consumption over the specified measurement period. The concept connects household spending decisions with aggregate demand and is central to analyses of tax changes, income transfers, and economic stabilization.

Definition and measurement

For a finite change in income, the MPC is commonly expressed as

MPC=ΔCΔYd,\mathrm{MPC}=\frac{\Delta C}{\Delta Y_d},

where CC denotes consumption expenditure and YdY_d denotes disposable income. If consumption is represented by a differentiable function, the marginal propensity is its slope with respect to disposable income:

MPC=∂C∂Yd.\mathrm{MPC}=\frac{\partial C}{\partial Y_d}.

The derivative describes a local response; the ratio of finite changes describes an average response over a particular income change. They need not coincide when consumption responds nonlinearly to income.

An MPC must be interpreted in relation to:

  • A time horizon: spending within a week, quarter, or year, or cumulative spending over several periods.
  • An income change: temporary or permanent, anticipated or unexpected, and positive or negative.
  • A population: an individual household, a demographic group, or the economy as a whole.
  • A spending measure: total expenditure, nondurable consumption, services, or another defined category.

Income and consumption changes should be measured in consistent monetary units and, when necessary, adjusted for price changes. An increase in nominal spending caused only by inflation is not an equivalent increase in real consumption.

A causal MPC is not simply the ratio of two changes that happen to occur together. Income and spending may both respond to employment conditions, expectations, or other influences.

Distinction from related concepts

Average propensity to consume

The average propensity to consume is the share of total disposable income devoted to consumption:

APC=CYd.\mathrm{APC}=\frac{C}{Y_d}.

It differs from the MPC, which concerns additional income. A household receiving 3,000 units of income and spending 2,700 has an APC of 0.90.9. If an additional 100 units of income raises spending by 60, its MPC for that change is 0.60.6.

Marginal propensity to save

If disposable income is divided between consumption and saving, then

Yd=C+S,Y_d=C+S,

and therefore

MPC+MPS=1,\mathrm{MPC}+\mathrm{MPS}=1,

where MPS=ΔS/ΔYd\mathrm{MPS}=\Delta S/\Delta Y_d is the marginal propensity to save.

Saving here is broader than money placed in a deposit account. Income used to repay debt rather than purchase consumption goods generally contributes to saving through reduced liabilities.

Income elasticity of consumption

The MPC is an absolute spending response per unit of additional income. An income elasticity instead measures a percentage response:

εC,Yd=∂C∂YdYdC.\varepsilon_{C,Y_d} = \frac{\partial C}{\partial Y_d}\frac{Y_d}{C}.

The two measures answer different questions: the MPC concerns how much of an additional unit is spent, while the elasticity concerns how strongly spending changes relative to its initial level.

Keynesian consumption theory

The MPC became central to Keynesian economics through John Maynard Keynes and his 1936 book The General Theory of Employment, Interest and Money. Keynes emphasized that consumption generally rises when income rises, but by less than the income increase.

A simple consumption function represents this relationship as

C=a+cYd,C=a+cY_d,

where aa is autonomous consumption and cc is a constant MPC. In the conventional version,

0<c<1.0<c<1.

If a>0a>0, this specification implies

APC=aYd+c.\mathrm{APC}=\frac{a}{Y_d}+c.

Thus, the average propensity to consume falls as income rises, while the marginal propensity remains constant. This is a feature of the linear model, not a universal empirical law. The constant term also should not be interpreted as spending that can necessarily be sustained indefinitely with no income.

The interval between zero and one is useful in elementary models, but it is not an unconditional restriction on every measured response. Purchases of durable goods, borrowing, spending brought forward from future periods, and measurement error can produce short-horizon estimates outside that range.

Intertemporal explanations

Later theories explain consumption by considering resources and choices across time rather than current income alone.

Permanent income

The permanent income hypothesis, associated with Milton Friedman, distinguishes changes in enduring earning capacity from temporary income fluctuations. A household able to borrow and save may spread the benefit of a temporary payment across many future periods. Its immediate MPC from that payment can therefore be relatively small.

A lasting increase in expected income may support a larger current spending increase because it raises expected lifetime resources by more. Consequently, an MPC estimated from a one-time rebate need not describe the response to a permanent wage increase.

Life-cycle consumption

The life-cycle hypothesis explains consumption and saving in relation to age, expected earnings, retirement, and accumulated wealth. Households may borrow early in life, save during higher-earning years, and draw down assets after retirement.

Responses to additional income depend on remaining lifespan, expected future resources, and bequest motives. Age alone does not determine an MPC: households of similar ages can have very different financial positions.

Constraints and uncertainty

Households with limited access to credit may be unable to spend against future income. Additional cash can then enable purchases that would otherwise be postponed, producing a relatively high immediate MPC.

Uncertainty can encourage precautionary saving, especially when households have limited resources available to absorb adverse events. Conversely, households with very low liquid balances may spend much of an income payment on immediate needs.

Liquidity is therefore distinct from total wealth. A household can own valuable housing or retirement assets yet hold little readily spendable money. Such a household may respond strongly to a cash payment despite having substantial net worth.

Household differences and aggregation

MPCs vary with liquid assets, debt obligations, income expectations, household composition, and access to credit. Spending responses may also reflect behavioral mechanisms such as mental accounting, under which otherwise similar funds are treated differently depending on their source or label.

Lower-income households often have higher immediate MPCs from temporary payments, but income is not a complete explanation. Liquid wealth and financial commitments can distinguish households with similar incomes.

For a fixed allocation of additional income, an aggregate MPC can be written as

MPCaggregate=∑i(ΔYd,i∑jΔYd,j)MPCi.\mathrm{MPC}_{\mathrm{aggregate}} = \sum_i \left( \frac{\Delta Y_{d,i}}{\sum_j\Delta Y_{d,j}} \right)\mathrm{MPC}_i.

This expression weights households by their shares of the income change, not simply by their population shares. It describes the direct spending response before accounting for wider economic feedback.

The distribution of a tax reduction or transfer therefore matters. Policies delivering the same total income gain can generate different initial consumption responses if their recipients have different MPCs. This links the concept to income inequality and the distribution of wealth.

Role in fiscal multipliers

In a basic closed-economy model with fixed prices, unused productive capacity, no income-dependent taxes, and no imports, equilibrium expenditure satisfies

Y=C+I+G,Y=C+I+G,

with C=a+cYC=a+cY. If investment and government spending are otherwise autonomous, the government-spending multiplier is

ΔYΔG=11−c.\frac{\Delta Y}{\Delta G}=\frac{1}{1-c}.

An initial spending increase creates income, part of which is consumed, generating further income. A higher MPC produces a larger multiplier within this model.

For a lump-sum tax change TT, where Yd=Y−TY_d=Y-T, the corresponding tax multiplier is

ΔYΔT=−c1−c.\frac{\Delta Y}{\Delta T}=-\frac{c}{1-c}.

The initial consumption effect of a tax reduction is only the fraction cc of the additional disposable income, whereas government purchases enter demand directly.

These formulas are not general empirical rules. Imports, taxes, productive capacity, price adjustment, financing arrangements, and monetary policy can alter the outcome. A household MPC alone therefore cannot establish the overall effect of fiscal policy on output.

Empirical estimation and limitations

Researchers estimate MPCs using household expenditure surveys, payment records, income data, and policy episodes such as rebates or transfers. Designs exploiting plausibly external differences in payment timing or eligibility help separate causal spending responses from ordinary correlations.

Important measurement problems include:

  • Anticipation: an announced payment may affect spending before it arrives.
  • Durable purchases: expenditure on a vehicle or appliance occurs at purchase, while its consumption services extend over time.
  • Incomplete coverage: one bank account or payment card may not capture a household’s total spending.
  • Measurement error: inaccurate income reports can distort estimated responses.
  • Time aggregation: a short-run response can differ substantially from cumulative spending over a year.
  • Positive versus negative shocks: borrowing constraints and spending commitments may make responses to gains and losses asymmetric.

A payment used initially to reduce debt may also support later spending, making immediate and longer-horizon MPCs different.

An estimated MPC is consequently specific to a population, income shock, spending definition, and time horizon. Applying it to another setting requires attention to those conditions, particularly when household-level estimates are used to predict economy-wide outcomes.