An Econometric Analysis of Consumer Sentiment and ... - NABE

More documents

Recommendations

Info

An Econometric Analysis of Consumer Sentiment and Confidence Indexes:Pulling Trigger EffectsWe offer an overview of IHS Global Insight's econometric research on the determinants of consumer mood, usingtraditional explanatory variables such as gasoline prices, initial unemployment insurance claims, the unemploymentrate, and equity prices.In this special report, we offer an overview of IHS Global Insight’s research on the determinants of consumer attitudesusing traditional explanatory variables as well as psychological “trigger points” from the field of behavioral economics.In addition, we give a brief overview of behavioral economics. In subsequent articles, we will focus on thedeterminants of inflationary expectations and Bloomberg's weekly comfort index. Since this report is somewhattechnical and lengthy, we have also provided a relatively brief summary as a perspective article: Consumer Feelingsand Economic Reality, by Chris G. Christopher, Jr. (June 6, 2011).The Conference Board's consumer confidence and Reuters/University of Michigan’s consumer sentiment indices aretwo widely used measures of consumer attitudes and the factors that drive consumer spending. Personalconsumption represents approximately 70% of GDP and is a key component in our understanding of business cycles.IHS Global Insight’s large-scale quarterly U.S. econometric model forecasts and utilizes the Reuters/University ofMichigan's consumer sentiment headline and short-term inflation expectations indices. Sentiment is a driver forvarious inflation-adjusted consumer expenditure components. Inflation expectations affect interest rates, wage rates,and the rental cost of capital.First, we give a brief overview of the current state of behavioral economics in the overall context of “mainstream"economics. We then discuss the econometric models and data used, and outline our key findings. Finally, we offersome brief conclusions.The Homo Economicus Assumption and Behavioral EconomicsMost modern mainstream economics analysis assumes that individuals behave as Homo Economicus (EconomicHuman). In this analytical framework, individuals behave under a narrowly defined “rationality” with purely selfinterestedmotives, where economic agents are strict “utility” maximizers with not much consideration of social,cognitive, or emotional factors in the understanding of human actions. Mainstream economists (Neo- or New-Keynesian, Classical, Neo-Classical, etc.) assume Homo economicus as the starting point in developing theirrespective theories and, in most cases, do not consider it necessary to challenge or empirically test this vital andhighly restrictive assumption. Most other social sciences, however, such as anthropology, sociology, psychology,behavioral sciences, and nonmainstream economics, have rejected the concept of Homo economicus. There isconsiderable agreement that human actions are rational in many circumstances, but not to the extreme of Homoeconomicus—there has to be room for emotions, and spontaneous innate urges.The two major nonmainstream economic schools of thought—Keynesian/Post-Keynesian (M. Keynes, HymanMinsky) and Austrian (Karl Menger, Ludwig Von Misses, F. A. Hayek)—diverge on economic policy prescriptions, butagree that the assumption of Homo economicus is erroneous and lead to ineffective policy prescriptions. According toJohn Maynard Keynes' The General Theory of Employment, Interest and Money, "[...] of our decisions to dosomething positive, and the full consequences of which will be drawn out over many days to come, can only be takenas a result of animal spirits—of a spontaneous urge to action rather than inaction, and not as the outcome ofweighted average of quantitative benefits multiplied by quantitative probabilities."[2].1 IHS Global Insight Published 7/1/11
The relatively new and developing fields of behavioral and experimental economics, evolutionary psychology, andneural sciences offer valuable insights and scientific evidence as to the behavior of individuals in the context ofhuman action. While these fields of research often disagree with each other on the reasoning and conclusionspertaining to human behavior, they find common ground in their rejection of Homo economicus and provideconfirmation to the analytical framework developed by the classical “liberal” tradition of the 19th century andexemplified in the works of Hayek. The field of behavioral and experimental economics has made significant inroadsinto the mainstream economics establishment, while several pioneers in the field have been awarded the NobelMemorial Prize in Economic Sciences. Behavioral economists attempt to empirically define the limits of humanrationality as being bounded and not without limits.The Reuters/University of Michigan consumer sentiment index was designed after World War II at the University ofMichigan by the eminent psychologist George Katona, whose pioneering efforts focused on integrating economicpsychology with macroeconomic theory, modeling, and forecasting. Katona maintained that there was a sharpdifference between income (the ability to buy) and the willingness to buy (consumer psychology). His argument wasthat if people feel better off, they will spend more, while if they feel worse off, they will spend less.Econometric Model, Data, Priors, and MethodologyThis section is a non-technical overview of the statistical results and terminology. We then describe in more detailsome key econometric results, and finally offer some broad insights and conclusions.We estimate monthly consumer sentiment and consumer confidence regressions in a structured time seriesframework. Due to data limitations, the time span is January 1994 through December 2010. The Conference Boardhas employed a different indexing methodology since 2011 that makes yearly comparisons difficult to interpret. Thus,we leave the sample endpoint for consumer sentiment at December 2010 as well, to allow for comparisons with theconsumer confidence regression results.To develop a practical model of the determinants of consumer attitudes, we expand on the traditionally usedexplanatory variables—such as inflation-adjusted gasoline prices, unemployment, stock market performance, pastconsumer optimism, seasonal controls, and recession controls—with variables from behavioral psychologyeconomics.The formal econometric representations of the consumer sentiment and consumer confidence regression equationsareCS(t) = Intercept + Seasonal Dummies+ b11*Relative Price of Gasoline(t)+ b12*SP500 Composite Growth Rate(t)+ b13*Initial Claims per Capita(t) + b14*Unemployment Rate(t)+ c11*Gasoline Price Trigger(t) + c12*DJIA Value Trigger(t)+ d11*2001 Recession(t) + d12*Great Recession(t)2 IHS Global Insight Published 7/1/11
Page 1: An Econometric Analysis ofConsumer
Page 5 and 6: Our expectations on the coefficient
Page 7 and 8: Consumer Sentiment RegressionSample
Page 9 and 10: Consumer Confidence RegressionSampl
Page 11 and 12: Percentage ChangeConsumer Sentiment
Page 13: previous three downturns. It was dr

An Econometric Analysis of Consumer Sentiment and ... - NABE

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?