Why the Consumption Function Matters
Households do not generally spend every dollar of income, nor do they spend the same dollar amount at every income level. The consumption function is a simple model of how planned consumption varies with disposable income. Its central parameter, the marginal propensity to consume, describes how much consumption changes when disposable income changes by one unit.
The function is useful for organizing calculations and interpreting household behavior. It is a model of a relationship, not a claim that every household has the same spending pattern or that income is the only influence on consumption. Keep the units and the meaning of each income measure clear: the relevant input is disposable income, not automatically gross or pre-tax income.
Interpreting the Parameters
In \(C=a+cY_d\), \(a\) is the intercept: the model's predicted consumption when disposable income is zero. It is called autonomous consumption because, within this simple model, it does not vary with current disposable income. If \(a\) is positive, the model predicts positive consumption even at zero disposable income. The household would have to finance that consumption from some other source, such as borrowing or using past savings. The intercept is not a guarantee that every household can or will spend that amount.
The coefficient \(c\) is the slope of the consumption function. It measures the change in consumption per unit change in disposable income. In the standard linear model, \(0\leq c\leq1\): an additional dollar of disposable income leads to between zero and one additional dollar of consumption. The portion not consumed is saved, under the simple budget identity. Thus the marginal propensity to save is \(1-c\).
The model is most naturally read with all dollar amounts measured in the same currency and time period. If income is reported in thousands of dollars per year, consumption must also be in thousands of dollars per year. A function estimated for one range of income or one period need not describe behavior outside that range or after prices, preferences, credit access, or expectations change.
Worked Example: Estimating the Intercept and MPC
A household survey summarizes consumption and disposable income in thousands of dollars per year. At disposable income of 40, consumption is 35; at disposable income of 70, consumption is 59. Assume these observations lie on a linear consumption function. Find the function, predict consumption at disposable income 85, and calculate saving there.
Method. Use the change in consumption divided by the change in disposable income to find the slope \(c\). Then substitute one observation to solve for the intercept \(a\).
The change in disposable income is \(70-40=30\) thousand dollars. The change in consumption is \(59-35=24\) thousand dollars. Therefore,
Substitute the first observation into \(C=a+cY_d\):
The estimated function is \(C=3+0.8Y_d\), with both \(C\) and \(Y_d\) measured in thousands of dollars per year. At \(Y_d=85\),
Saving is disposable income minus consumption, so \(S=85-71=14\) thousand dollars per year. As a check, the model's MPS is \(1-0.8=0.2\), and \(0.2(85)-3=17-3=14\), the same saving amount.
Common mistake avoided: Dividing consumption by income at the first observation gives \(35/40=0.875\), but that is an average consumption-to-income ratio at that point, not the MPC. The MPC is the change ratio, \(24/30=0.8\).
From Consumption to Saving
The relation between the MPC and MPS follows directly from the household budget identity. Saving is the part of disposable income not consumed. Substituting the consumption function into that identity gives the saving function. The derivation also shows why the MPS is the complement of the MPC in this model.
Disposable income is either consumed or saved: \(Y_d=C+S\), so \(S=Y_d-C\).
Using \(C=a+cY_d\), obtain \(S=Y_d-(a+cY_d)\).
\(S=-a+(1-c)Y_d\). Therefore the saving-function intercept is \(-a\), and its slope is \(1-c\).
For a change in disposable income, \(\Delta S=(1-c)\Delta Y_d\). Hence the MPS, \(\Delta S/\Delta Y_d\), is \(1-c\), and the MPC plus MPS equals one.
This result depends on the simple allocation identity and on the same change in disposable income being divided between consumption and saving. It does not mean that the level of saving must be positive. If autonomous consumption is positive, saving can be negative at low disposable income, even while the MPS is positive.
Worked Example: Finding Saving and Break-Even Income
A household's annual consumption function, in dollars, is \(C=12{,}000+0.75Y_d\). Find its saving function, calculate consumption and saving at disposable income of 40,000 and 64,000 dollars, and determine the disposable income at which saving is zero.
Method. Subtract the consumption function from disposable income to obtain saving. Then set saving equal to zero to find the break-even income.
First, the MPS is \(1-0.75=0.25\). Therefore,
At \(Y_d=40{,}000\), consumption is \(12{,}000+0.75(40{,}000)=12{,}000+30{,}000=42{,}000\) dollars. Saving is \(40{,}000-42{,}000=-2{,}000\) dollars. The negative value means consumption exceeds disposable income by 2,000 dollars in this model.
At \(Y_d=64{,}000\), consumption is \(12{,}000+0.75(64{,}000)=12{,}000+48{,}000=60{,}000\) dollars. Saving is \(64{,}000-60{,}000=4{,}000\) dollars.
For break-even income, set \(S=0\):
At that income, consumption is \(12{,}000+0.75(48{,}000)=48{,}000\) dollars, so saving is indeed zero. Common mistake avoided: A positive MPS does not imply positive saving at every income. The intercept \(-12{,}000\) makes saving negative below the break-even income.
Marginal Propensities and Average Ratios
A marginal propensity describes a change; an average ratio describes a level at a particular income. At a positive value of disposable income, the average propensity to consume is \(C/Y_d\), and the average propensity to save is \(S/Y_d\). Since \(C+S=Y_d\), these average ratios sum to one. They need not equal the MPC and MPS.
For the linear consumption function, the average propensity to consume is
When \(a>0\), the term \(a/Y_d\) changes with income. Consequently, the average propensity to consume generally differs across income levels even though the model holds the MPC \(c\) constant. A common exam error is to infer the MPC from one consumption-to-income ratio rather than from a slope or a change.
Worked Example: Distinguishing a Marginal from an Average Measure
Suppose a linear consumption function passes through two annual observations, both in thousands of dollars: at \(Y_d=20\), consumption is 26; at \(Y_d=50\), consumption is 47. Find the function, the MPC and MPS, and consumption and the average propensity to consume at \(Y_d=80\).
Method. Estimate the slope from the two changes, recover the intercept, and then calculate the requested level and ratio.
The income change is \(50-20=30\), and the consumption change is \(47-26=21\). Thus \(c=21/30=0.7\), so the MPS is \(1-0.7=0.3\). Use the first observation to find \(a\):
The function is \(C=12+0.7Y_d\). At \(Y_d=80\), consumption is \(12+0.7(80)=12+56=68\) thousand dollars. The average propensity to consume is \(68/80=0.85\). This is not the MPC: the MPC is 0.7, while 0.85 compares the total consumption level with the total disposable income at this point.
Check saving: \(S=80-68=12\) thousand dollars. The saving function is \(S=-12+0.3Y_d\), which gives \(-12+0.3(80)=-12+24=12\). The consumption and saving amounts add to disposable income, \(68+12=80\). Common mistake avoided: Reporting 0.85 as the MPC confuses an average ratio with the slope of the consumption function.
Assumptions, Uses, and Limits
The linear form is a compact way to describe planned consumption over a specified range. It assumes a stable intercept and a constant slope over that range, with disposable income measured on a consistent basis. It holds other influences on consumption fixed for the calculation. In reality, wealth, credit conditions, expectations, household needs, and the distribution of income can also matter. A fitted line summarizes a relationship under particular conditions; it does not by itself establish that changing income alone caused the observed change in consumption.
For a calculation, identify the income measure, its units, and the period first. If the problem gives gross income along with taxes and transfers, determine disposable income before applying the consumption function. If it gives \(Y_d\) directly, do not subtract taxes again. Then distinguish whether the question asks for a level, such as consumption at a specified income, or for a marginal response, such as the MPC.
- Autonomous consumption \(a\): the intercept, or predicted consumption at \(Y_d=0\).
- MPC \(c\): the slope, measuring the change in consumption per unit change in disposable income.
- MPS \(1-c\): the share of an additional unit of disposable income allocated to saving in this model.
- Saving function: \(S=-a+(1-c)Y_d\), obtained from \(S=Y_d-C\).
These distinctions support the fiscal-policy analysis aligned with the SOA VEE Economics (Validation by Educational Experience) syllabus: they clarify how changes in disposable income relate to household consumption and saving. The consumption function alone does not determine total output or the eventual effect of a policy change; those require additional parts of the expenditure model.
Check Your Understanding
For each problem, keep the income measure and units consistent, and distinguish changes from levels.
- A consumption function is \(C=8+0.6Y_d\), with amounts in thousands of dollars per year. At \(Y_d=35\), calculate consumption and saving, and state the MPS. (Answer: \(C=29\), \(S=6\), and MPS \(=0.4\), all monetary amounts in thousands of dollars per year.)
- In thousands of dollars per year, consumption is 28 when disposable income is 30 and 56 when disposable income is 70. Assuming a linear function, find \(a\), the MPC, consumption at \(Y_d=90\), and saving there. (Answer: \(a=7\), MPC \(=0.7\), \(C=70\), and \(S=20\), with monetary amounts in thousands of dollars per year.)
- A household has annual gross income of 25,000 dollars, pays 4,000 dollars in personal taxes, and receives a 1,000-dollar transfer. Its function is \(C=5{,}000+0.8Y_d\). Calculate disposable income, consumption, and saving. (Answer: \(Y_d=22{,}000\), \(C=22{,}600\), and \(S=-600\) dollars per year.)
- A consumption function is \(C=18+0.65Y_d\), with amounts in thousands of dollars per year. Find the saving function, saving at \(Y_d=40\), and the break-even disposable income. (Answer: \(S=-18+0.35Y_d\), saving \(=-4\) thousand dollars per year at \(Y_d=40\), and break-even income \(=51.43\) thousand dollars per year, rounded to two decimal places.)
- Disposable income rises by 12,000 dollars and consumption rises by 9,000 dollars. Under the linear model, find the MPC, MPS, and change in saving. (Answer: MPC \(=0.75\), MPS \(=0.25\), and saving rises by 3,000 dollars.)