Consumption Based on Gross Domestic Product: a Keynesian Consumption Function

College of Business
Western Carolina University


Consumption expenditures are forecast to grow approximately 1.5% in 2001 and 2002.The consumption forecast is based on the fact that consumption is a variable of GDP.This approach is based on the Keynesian Consumption Function.GDP is the sum of consumption, investment, government spending, and net exports.Consumption spending is a function of real income, measured as real GDP.(JEL: E21, E120)

Part 1.Introduction

This paper forecasts personal consumption in the years 2001 and 2002.U.S. Gross domestic product (GDP) is the explanatory variable that will predict consumption in the future.Consumption is one of several variables that affect GDP.This approach is based on the Keynesian aggregate expenditure model. This model will be used because it is useful for macroeconomic forecasting and policy analysis, as well as its correspondence to the National Income and Product Accounts (NIPA). Nominal aggregate expenditure equals GDP over any time period, which is the sum of consumption, investment, government spending, and net exports.

The rest of this paper is organized as follows:Part 2. presents the data used to forecast consumption; part 3. explains the theoretical basis for the approach adopted in forecasting consumption; Part 4. presents forecasts of consumption for 2001 and 2002; Part 5. evaluates the importance of the forecast for the economy; and Part 6. discusses conclusions for economic policy.

Part 2. Data

The first variable is consumption.This data is retrieved from the Federal Reserve Bank of St. Louis Federal Reserve Economic Data (FRED).The FRED descriptor is PCECC96.The figures used for the sample period are in billions of chained 1996 U.S. dollars and is observed quarterly.The explanatory variable is real GDP.The FRED descriptor for GDP is GDPC96.The figures used for the sample period are seasonally adjusted in chained 1996 U.S. dollars and is observed quarterly.This information is gathered from FRED because it is very reliable and readily available for all persons. 

Chart 1

Part 3.The Keynesian Consumption Function

Consumption consists of autonomous consumption and induced consumption.Autonomous consumption does not depend on real GDP or disposable income.Induced consumption is dependent of both real GDP and disposable income.The intercept of the consumption function is autonomous consumption, and the slope of the consumption function is the marginal propensity to consume (MPC).Marginal propensity to consume relates to the thought of the more income people have, the more they are willing to spend on consumption (Keynes is a believer in this statement).“Keynes suggested that there was a ‘psychological law’ that any increase in income would result in and increase in consumption, but that the increase in consumption would be in less that a one-to-one proportion” (Income and Consumption).Consumption will increase by a small proportion of income when income increases by a dollar. This fraction is the marginal propensity to consume.The consumption function can be written as:

C = Co + MPC x Y

Where Co is autonomous consumption, (MPC * Y) is induced consumption, and Y is real GDP.Lagging the right hand side of the consumption function by two years provides the following forecasting equation:

C = Co + C1 x Yt-2

Part 4. Empirical Results

The r-square is .9957.This means that 99.57% of variation in consumption is related to GDP.The t-statistics of the intercept is low, which shows low rejection of the null hypotheses.On the other hand, the t-statistics for GDP is very high, showing high rejection.Table 1 shows the regression summary.

Table 1

Regression Estimate of Consumption Forecasting Equation:1991.1-2000.4

Explanatory Variable
Estimated coefficient
R Square = .9995673
Multiple R = .997834
The negative coefficient of the intercept tells us that consumption spending will decrease a year after GDP goes up.The t-ratio for GDP is very high, which means that it is not statistically significant.These values may not seem very valid due to the fact that consumption and GDP’s annual percentage growth is very low.Their low growth rate could be due to inflation causing consumers to not spend as rapidly as they have in the past.
Part 5.  Forecast Implications

This forecasts shows that consumption will decrease in the first quarter of 2000, but then begin increasing again.I believe the forecasted data resembles data from the past.This could be due to consumers saving their money due to a scare of inflation.Induced consumption depends on disposable income, so if disposable income decreases, consumption will also decrease.This is assuming that consumption expenditures are depending on total real income and could be a theory of what will happen in the future.Table 2 gives the forecast results.

Table 2
Forecast of Consumption for 2001 and 2002
Forecast Consumption
% Change
Part 6. Conclusions on Policy

This forecasts shows that there will be very little change in consumption based on GDP, except for in the first quarter of 2001.If there is an increase in consumption spending, there could also be an increase in output price, causing inflation.  Inflation can occur from consumption increasing, and unemployment could fall under the natural rate of employment.The more that consumers are spending, the less they are saving.The government appreciates the lack of saving because the more consumers spend, the more the government can earn on tax dollars.When there is a lack of saving, there is an increase in consumption.

Federal Reserve Bank of St. Louis, Federal Reserve Economic Data (FRED), (2-08-01) (2-15-01)
“Income and Consumption,” (2-25-01)