CHAPTER 2 Tools of Positive Analysis McGraw-Hill/Irwin Copyright © 2008 by The McGraw-Hill Companies, Inc. All rights reserved. The Role of Theory Economic models: Theory plays a crucial role for empirical researchers by helping them to isolate a set of variables that may influence a particular kind of economic behavior. Empirical work then tests whether the theory is consistent with real-world phenomena. The most widespread method of empirical research in economics is econometric analysis, because economists tend to be most comfortable with results based on data from real-world environments. However, honest econometrician can come to very different conclusions, because the data and statistical techniques are imperfect. 2-2 The Role of Theory The economic theory is set by both the politicians and the economists. There is one fact which is that there is lack of definitive answers in economy because the economists are generally unable to perform carefully controlled experiments. Therefore, we never know for certain the consequences of various policy changes. 2-3 The Role of Theory Because economists generally cannot perform controlled experiments with the economy, the effects of economic policy are difficult to determine. Economic theory helps specify the factors that might affect a given kind of behavior. However, theory alone cannot say how important any particular factor is. 2-4 The Role of Theory Empirical analysis: When doing an empirical analysis we have to consider the following: 1. virtue of simplicity 2. judging a model 3. limitations of models 2-5 Empirical research Empirical research attempts to measure both the direction and size of the effect of government policy changes on behavior. Common types of empirical studies are: 1. Interview studies 2. Social and laboratory experiments 3. Econometric analysis 2-6 1. Interview Studies Interview studies consists of directly asking people how various policies affect their behavior. Pitfalls: 1. Interpreting interview surveys results requires quotation. 2. People may not actually react to policies in the way they say they do. 2-7 2. Experimental studies Experimental studies are based on random assignment to control and treatment groups. Experimental studies are divided into: 1. Social experiments. 2. Laboratory experiments. 2-8 2. Social Experimental studies There is a problem to do empirical work in economics because of the inability to do controlled experiments with the economy. Social experiments subject one group of people to some policy and compare their behavior with that of a control group. Example: Public versus private schools. 2-9 Pitfalls of Social Experimental Studies 1. 2. 3. 4. 5. 6. Ethical issues: People are People. Technical problems: Sample has to be representative of the population. Response bias: People are aware they are under experiment. Impact of limited duration of experiment: Life and personal circumstances always change. Generalization of results to other populations, settings, and related treatments. High costs. 2-10 2. Laboratory Experimental studies Laboratory experimental studies are used to some types of economic decisions. Pitfalls of laboratory experiments include: 1. They occur in the artificial atmosphere where persons may not replicate real-world behavior. 2. Undergraduate students are unlikely to be representatives of the population. 2-11 3. Econometrics Studies Observational study – empirical study relying on observed data not obtained from experimental study. Sources of observational data Surveys Administrative records Governmental data Econometrics: It is the statistical analysis of economic data. In econometrics, the effects of various policies are inferred from observed behavior. Regression analysis: It is used to pick the “best” parameters for an economic model. Knowing the parameters allows one to predict the effects of policy changes. 2-12 Effects of Public Finance Policies Substitution effect: With these high taxes, it’s really not worthwhile for me to work as much as I used to. Income effect: With these high taxes, I have to work more to maintain my standard of living. Both of the two effects can reflect perfectly rational behavior. 2-13 Types of Data Cross-sectional data Time-series data Panel data 2-14 Causation vs. Correlation Statistical analysis Correlation Control group Treatment group Conditions required for government action X to cause societal effect Y X must precede Y X and Y must be correlated Other explanations for any observed correlation must be eliminated 2-15 Conducting an Observational Study L = α0 + α1wn + α2X1 + … + αnXn + ε Dependent variable Independent variables L Parameters Stochastic error term A: dividends and interest rate, X1: age, X2: number of children, etc… Regression analysis Regression line Standard error Intercept is α0 Slope is α1 α0 wn 2-16 Pitfalls of Econometric Studies Misleading results occur if: 1. Data from greatly dissimilar groups are combined. E.g. unmarried and married women, old and young. 2. Wrong mathematical form is adopted. Parameters changes over time. 3. Variables are incorrectly measured. 4. Um-measurable variables are used including perceptions and attitudes. 5. There is simultaneous causation between variables. 2-17 Quasi-Experimental Studies Quasi-experimental study (= natural experiment) – observational study relying on circumstances outside researcher’s control to mimic random assignment 2-18 Pitfalls of Quasi-Experimental Studies Assignment to control and treatment groups may not be random Not applicable to all research questions Generalization of results to other settings and treatments 2-19