Economics 314 Coursebook, 2009 Jeffrey Parker 16 THEORIES OF CONSUMPTION AND SAVING Chapter 16 Contents A. Topics and Tools............................................................................. 1 B. The Kuznets Paradox ....................................................................... 3 C. Relative-Income Hypothesis .............................................................. 5 D. Life-Cycle Model and Permanent-Income Hypothesis .............................. 7 Life-cycle model ............................................................................................................ 8 Permanent-income hypothesis ........................................................................................ 9 Permanent vs. temporary changes in income ................................................................. 11 The dangers of ignoring macroeconomists: A policy faux pas .......................................... 12 E. Understanding Romer=s Chapter 7 ....................................................... 13 Empirical consumption functions ................................................................................. 13 Uncertainty, rational expectations, and consumption ..................................................... 14 Quadratic utility ........................................................................................................ 16 F. Empirical Tests of the Random-Walk Hypothesis .................................... 16 Testing the random-walk hypothesis............................................................................. 18 Later tests and excess sensitivity of consumption to income .............................................. 20 G. Suggestions for Further Reading ........................................................ 23 Classic studies of consumption ..................................................................................... 23 Some recent advances .................................................................................................. 23 H. Works Cited in Text ....................................................................... 23 A. Topics and Tools There is no topic in macroeconomics that has a longer, deeper, or more prominent literature than households’ choice of how much of their income to consume and save. As we saw earlier in the course, the theory of consumption is central to the model of16 – 2 Keynes’s General Theory, which is often considered to be the origin of macroeconomics. Since then it has been the subject of countless theoretical and empirical studies. Keynes treated consumption on a very “common sense” level. Like most other economists of his day, his methodology included neither abstract, mathematical theory nor detailed econometrics. Rather he relied almost entirely on intuition, as he demon-strates when he introduces the central principle of his consumption theory in Chapter 8: The fundamental psychological law, upon which we are entitled to de-pend with great confidence both a priori from our knowledge of human nature and from the detailed facts of experience, is that men are dis-posed, as a rule and on the average, to increase their consumption as their income increases, but not by as much as the increase in their in-come. Keynes (1936) Keynes gives no basis for his theory in terms of utility maximization nor indeed gives any consideration of why a consumer would behave in the way he assumes. In place of rational-choice theory, Keynes relies on his “knowledge of human nature.” Nor does he give any support using numerical data, but instead claims to glean support from “de-tailed facts of experience.” How much economics has changed in seventy years! While Keynes placed consumption theory at the center of the macroeconomic stage, he left it for future generations of economists to work out the microeconomic basis for his theory and competing theories. Keynes also inspired pioneers in the emerging field of econometrics to swarm over the newly invented national income and product statistics looking for verification or refutation of his model. Keynes’s basic model of consumption was that current consumption expenditures are determined mainly by current disposable income. The Keynesian consumption func-tion is usually written in linear form: Ct = a + bYt. The coefficient b, which Keynes called the “marginal propensity to consume” or MPC and which we would define concisely as ∂C/∂Y, was to vie for the title of “most estimated coefficient” for several decades. Initial linear econometric consumption functions estimated by ordinary least squares produced results that conformed to Keynes’s theory: consumption seemed to be closely related to current disposable income and the MPC seemed to be positive and less than one. How-ever, Nobel-laureate Trygve Haavelmo used the consumption example prominently in pointing out the bias that is present in OLS estimation when the variable on the right-hand side of the equation (income, in this case) is correlated with the error term. When the “Haavelmo problem” was accounted for, the corrected estimates of the MPC turned out to be considerably lower than OLS estimates. At about the same time, Simon Kuznets (another Nobel winner) refined national-account measures of income and consumption and pointed out a paradox that could not be explained by the simple linear consumption function. The Kuznets paradox was that the percentage of disposable income that is consumed is remarkably constant in the long16 – 3 run, which suggests a proportional consumption function, i.e., that the intercept term a is equal to zero. However, estimates across individual households or using short-run ag-gregate time-series fluctuations in income and consumption consistently produce esti-mates implying that a > 0, which means that the share of income consumed declines as income rises. Explaining the Kuznets paradox became a primary goal of consumption theorists in the 1950s. One early approach was the “relative-income hypothesis,” which asserted that a household’s consumption depends not only on its current disposable income, but also on current income relative to past levels and relative to the income of other households. This hypothesis enjoyed considerable popularity in the 1950s, but is not discussed much anymore. Two other theories pioneered by Nobel laureates, the life-cycle model associated with Franco Modigliani and the