Dr. Barry Haworth
University of Louisville
Department of Economics
Economics 202

The Aggregate Expenditure Model

The Aggregate Expenditure Model
We’ll define Aggregate Expenditure (AE) as the sum of expenditures on all final goods and services at a given price level. That is, when the price level is specified at a certain level, AE is the total amount of money people will spend on final goods and services at different levels of income.

There are several different expenditure categories we can consider. AE is the sum of all (domestic) consumption expenditure, investment expenditure, government expenditure, and expenditure on exports, less the expenditure on imports. Let’s consider each expenditure category more directly, beginning with consumption (data on the actual amounts of each expenditure can be located at a variety of websites, including the Bureau of Economic Affairs website).

Consumption includes the purchase of final goods and services like food, clothes, personal care items, etc., but doesn’t include things like the purchase of a home (which is considered part of investment).

Our goal is to create an equation that describes (mathematically) how much a given individual or group of individuals will spend on consumption. To do this, we must ask about variables that might change consumption expenditure. An obvious candidate is disposable income. That is, we know that changes in disposable income lead to subsequent changes in consumption spending. While there are probably other variables which also affect consumption expenditure, we'll keep things simple here and assume that income is the primary one. Based on the idea that disposable income is the primary cause of changes in consumption expenditure, let's divide consumption expenditure into two categories:

  • variable consumption expenditure - that part of consumption that changes with one’s current disposable income
  • non-variable consumption expenditure - which is consumption expenditure that’s independent of disposable income (e.g. a subsistence level of consumption, or consumption that depends on the income of previous periods). This latter type we call autonomous or fixed consumption expenditure, where autonomous means independent of income.

    Before writing a consumption equation, we want to consider variable consumption expenditure a little more closely. More specifically, we want to know the rate by which this category varies. Let’s assume that consumption varies at a constant rate, which we’ll call the marginal propensity to consume. Note that consumption probably does not vary at a constant rate in reality, but this is a good approximation for smaller ranges of disposable income. This marginal propensity to consume will represent the change in consumption that results from a change in disposable income (note that connection between the word marginal and the use of the words "change in"). We also know that this value cannot be negative or exceed one.

    With all this information in hand, we can take a stab at writing a consumption expenditure equation. That equation will be:

    C = m(DI) + b


    m = marginal propensity to consume (mpc), measured as DC/DDI
    DI = disposable income
    b = autonomous consumption (expenditure)

    We'll assume that for every dollar increase in disposable income, consumption expenditures rise by 75 cents. That is, let’s assume that the mpc = 0.75. Similarly, we can assume that $100 is the amount of autonomous consumption. Our consumption equation becomes C = 0.75(DI) + 100. Note that we could have chosen any values to replace "m" and "b" in the consumption equation that make sense, not necessarily the values we chose here. If we draw a graph of this equation, using DI as the independent variable (x-axis) and C as the dependent variable (y-axis), then we have something that looks like:

    Investment expenditure falls into two categories as well. The first category involves capital expenditures. This includes expenditure on residential housing, buildings, equipment, etc. The second category involves changes in inventories. Note that inventories can either rise or fall, which implies that this category can be positive or negative. No other expenditure has this characteristic.

    In reality, investment expenditure varies with changes in interest rates and possibly incomes, but here we will assume this does not occur. This is to keep our model simple. Note that by doing so, we are unable to determine the full impact of certain types of macroeconomic change. Our assumption is that investors determine their level of investment in advance of the current period. We call this pre-planned investment by another name, autonomous investment expenditure. It’s possible for autonomous investment to change during a given period, but when this occurs, it is not because of a change in income.

    Government Spending
    Government expenditure will include all expenditure by each level of government. The primary way that governments make expenditure is by providing public goods like national defense, fire and police protective services, parks, etc. Just as with investment, we will assume that the government makes all of its expenditure decisions in advance. That is, government expenditure is pre-planned and we’ll call it autonomous government expenditure. One obvious problem with this assumption is that we’re saying governments will spend the same amount of money, no matter whether the economy is in a recession or not. In reality, this is obviously untrue. Nevertheless, we will retain this assumption to keep the model simple enough to work with.

    Exports and Imports
    Export and import expenditures are very straightforward. Exports are added to aggregate expenditure, while imports are deducted. We’ll assume again that export and import expenditures are pre-planned, making them autonomous export expenditures and autonomous import expenditures. This is probably the case for exports, but with imports this is less realistic. Our desire to import goods is related to changes in income. As we receive more income, we desire more imported goods. Again, to keep the model simple, however, we’ll ignore this relationship.

    Back to the overall AE model
    Recall that AE is the sum of all expenditure (less imports). In equation form that implies:

    AE = C + I + G + X – M


    C = total consumption expenditure
    I = investment expenditure
    G = government expenditure
    X = export expenditure
    M = import expenditure

    Let’s assume we’re looking at a specific country. We’ll use the consumption equation from the graph to represent consumption expenditures, and assume that investment, government spending, exports and imports take on the following (autonomous) values:

    I = 1000
    G = 1000
    X = 500
    M = 500

    If we substitute the C equation above as well as these values into the AE equation, we get:

    AE = [0.75(DI) + 100] + 1000 + 1000 + (500 - 500)

    Simplifying this equation, we have:

    AE = 0.75(DI) + 2100

    Note that DI is disposable income. Disposable income was assumed to affect consumption, but we know that there is something else, controlled by government, that can directly affect disposable income. That something is taxes. If we define disposable income as income after taxes, then we can write an equation for DI. DI = Y - T, where Y is income (or GDP) and T represents total tax revenues. Making one final assumption about T, let’s assume that T is autonomous as well (i.e. there are no income taxes) and that T = 1000. If we substitute this into the AE equation, then we have:

    AE = 0.75(Y - 1000) + 2100

    Simplifying this, we end up with:

    AE = 0.75Y + 1350

    It is this equation that we’ll use to represent aggregate expenditures. On a graph, we have something that looks very similar to the consumption equation graph above. The AE graph appears as follows:

    We want to make note of a couple things. First, the mpc is the slope of the consumption function (equation) and the slope of the AE equation. This is true because we have assumed that so many of our expenditure categories are autonomous. One impact of moving closer toward reality (i.e. by not assuming so many variables are autonomous) is that the slope of this graph changes. The mpc remains the slope of the consumption function, but becomes only a part of the slope of the AE equation.

    A second thing to note is that even though we have an equation for AE, we don’t know the exactly how much will be spent. That is, we don’t know the equilibrium values of AE and Y. This requires doing some further work, something we’ll save for another handout.