Working Paper: NBER ID: w1391
Authors: Robert E. Hall
Abstract: Consumption and income tend to move together; the correlation of their first differences is about 0.14. In most accounts, the correlation is attributed to the upward slope of the consumption function. When the publicis better off, they consume more. But in the microeconomic theory of the household, income is a variable chosen by the household. Choosing to workmore, and therefore to consume less time away from work, is a sign of diminished well being.The structural relation between earnings and consumption should have a negative slope.The explanation of the observed positive correlation of consumption and income must rest on shifts of the consumption-income relation, not movements along it. An examination of data for the U.S. in the twentieth century shows that the slope of the consumption-income relation has been approximately zero. Shifts in consumer behavior explain the positive observed correlation; they are an important, but not dominant, source of overall fluctuations in the aggregate economy.
Keywords: Consumption; Economic Fluctuations; Military Spending
JEL Codes: E21; E32
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| military spending (H56) | GNP (E10) |
| military spending (H56) | consumption (E21) |
| shifts in consumption schedule (D15) | GNP (E10) |
| consumption (E21) | GNP (E10) |
| other spending (H59) | GNP (E10) |