Marginal propensity to save measures how much saving changes relative to an incremental change in disposable income over a stated period.
The marginal propensity to save (MPS) measures the change in saving associated with an incremental change in disposable income. If disposable income rises by 1,000 and saving rises by 300 over the same period, the measured MPS is 0.30.
MPS is not the household’s savings rate. A savings rate compares total saving with total income, while MPS compares changes. It is also not a fixed behavioral score: liquidity, debt, expectations, wealth, and whether the income change is temporary can all affect the measured response.
Where:
The measures must use compatible definitions and periods. A quarterly change in saving divided by an annual change in gross income is not a meaningful MPS. Analysts should also state whether figures are nominal or adjusted for inflation.
Assume a household’s monthly disposable income increases from 4,500 to 5,500. Its monthly consumption rises from 4,000 to 4,700.
Before the income change, measured saving is:
After the change, measured saving is:
Therefore:
Consumption rose by 700, so MPC is 0.70. Under the same measurement boundary:
The household’s savings rate after the change is different:
MPS describes the treatment of the incremental income. The savings rate describes saving relative to the whole measured income level.
| Measure | Formula | Question answered |
|---|---|---|
| Marginal propensity to save | (\Delta S / \Delta Y_d) | How did saving change relative to an income change? |
| Savings rate | (S / Y_d) | What share of current disposable income was saved? |
| Marginal propensity to consume | (\Delta C / \Delta Y_d) | How did consumption change relative to an income change? |
A household can have a high savings rate but a low MPS during a particular period, or the reverse. For example, a high-saving household may spend most of a temporary income increase on a deferred repair while still saving a large share of total income.
In a simple household budget, saving is disposable income not used for current consumption. That can show up as:
Transfers between two asset accounts are not new saving by themselves. Buying securities with cash already held changes the asset mix but does not necessarily change current-period saving. Definitions in household surveys, financial accounts, tax records, and national accounts can differ, so the source should be stated.
The standard result follows from:
If the entire change in disposable income is either consumed or saved under the same definitions, then:
The identity does not require MPC and MPS each to stay between zero and one in every observation. A household can increase consumption by more than the income gain and reduce saving, producing MPC above one and MPS below zero. The two can still sum to one if the accounting boundary is consistent.
Temporary versus persistent income: A temporary payment may be saved at a higher margin than an income increase expected to continue, as emphasized by the permanent-income hypothesis.
Precautionary motives: Job, health, business, or market uncertainty can increase desired liquid reserves.
Debt and liquidity: A household may use additional income to reduce expensive debt. Economically, principal reduction can increase net worth even though it does not look like money placed in a savings account.
Credit access: Households unable to borrow may save less of additional income because they have postponed essential or high-priority consumption.
Life stage: Education costs, home purchases, dependents, and retirement can change both resources and desired saving.
Timing: A household may initially save a tax refund and spend it later. The estimated MPS depends on whether the measurement window is a week, quarter, or year.
Economic analysis: MPS helps describe how an income change is split between current consumption and saving. It is one input in simplified expenditure models.
Fiscal analysis: Transfers and tax changes can have different near-term consumption effects across households. A high assumed MPS implies a smaller immediate consumption response, but actual policy analysis requires many additional channels.
Financial stability: Changes in household saving can affect liquidity buffers and debt reduction, though aggregate flows do not reveal the condition of every household.
Capital markets: Aggregate saving supplies funds to the financial system, but higher household saving does not translate mechanically or immediately into productive business investment. Banks, government borrowing, international flows, asset prices, and firms’ investment demand all intervene.
Calling MPS a savings rate: One uses changes; the other uses levels.
Treating all asset purchases as new saving: Reallocating existing wealth is not the same as saving current income.
Assuming MPS is always between zero and one: That range is a model assumption, not an inviolable empirical rule for every interval.
Ignoring income expectations: A one-time refund and a durable raise may produce different responses.
Inferring causation from two observations: Saving may change because of a planned purchase, debt refinancing, or uncertainty rather than because income changed.
Applying an aggregate average to a person: Household responses differ materially by wealth, debt, age, and credit access.
This page is educational. MPS does not determine how an individual should allocate additional income or whether a specific debt, deposit, or investment action is suitable.