1、Chapter 1 discusses the scope of econometrics and raises general issues that resultfrom the application of econometric methods. Section 1.3 examines the kinds ofdata sets that are used in business, economics, and other social sciences. Section1.4 provides an intuitive discussion of the difficulties
2、associated with the inference ofcausality in the social sciences.1.1 WHAT IS ECONOMETRICS?Imagine that you are hired by your state government to evaluate the effectiveness of apublicly funded job training program. Suppose this program teaches workers variousways to use computers in the manufacturing
3、 process. The twenty-week program offerscourses during nonworking hours. Any hourly manufacturing worker may participate,and enrollment in all or part of the program is voluntary. You are to determine what, ifany, effect the training program has on each workers subsequent hourly wage.Now suppose you
4、 work for an investment bank. You are to study the returns on dif-ferent investment strategies involving short-term U.S. treasury bills to decide whetherthey comply with implied economic theories.The task of answering such questions may seem daunting at first. At this point,you may only have a vague
5、 idea of the kind of data you would need to collect. By theend of this introductory econometrics course, you should know how to use econo-metric methods to formally evaluate a job training program or to test a simple eco-nomic theory.Econometrics is based upon the development of statistical methods
6、for estimatingeconomic relationships, testing economic theories, and evaluating and implementinggovernment and business policy. The most common application of econometrics is theforecasting of such important macroeconomic variables as interest rates, inflation rates,and gross domestic product. While
7、 forecasts of economic indicators are highly visibleand are often widely published, econometric methods can be used in economic areasthat have nothing to do with macroeconomic forecasting. For example, we will studythe effects of political campaign expenditures on voting outcomes. We will consider t
8、heeffect of school spending on student performance in the field of education. In addition,we will learn how to use econometric methods for forecasting economic time series.1ChapterOneThe Nature of Econometrics andEconomic Datad 7/14/99 4:34 PM Page 1Econometrics has evolved as a separate discipline
9、from mathematical statisticsbecause the former focuses on the problems inherent in collecting and analyzing nonex-perimental economic data. Nonexperimental data are not accumulated through con-trolled experiments on individuals, firms, or segments of the economy. (Nonexperimentaldata are sometimes c
10、alled observational data to emphasize the fact that the researcheris a passive collector of the data.) Experimental data are often collected in laboratoryenvironments in the natural sciences, but they are much more difficult to obtain in thesocial sciences. While some social experiments can be devis
11、ed, it is often impossible,prohibitively expensive, or morally repugnant to conduct the kinds of controlled experi-ments that would be needed to address economic issues. We give some specific exam-ples of the differences between experimental and nonexperimental data in Section 1.4.Naturally, econome
12、tricians have borrowed from mathematical statisticians when-ever possible. The method of multiple regression analysis is the mainstay in both fields,but its focus and interpretation can differ markedly. In addition, economists havedevised new techniques to deal with the complexities of economic data
13、 and to test thepredictions of economic theories.1.2 STEPS IN EMPIRICAL ECONOMIC ANALYSISEconometric methods are relevant in virtually every branch of applied economics. Theycome into play either when we have an economic theory to test or when we have a rela-tionship in mind that has some importance
14、 for business decisions or policy analysis. Anempirical analysis uses data to test a theory or to estimate a relationship.How does one go about structuring an empirical economic analysis? It may seemobvious, but it is worth emphasizing that the first step in any empirical analysis is thecareful form
15、ulation of the question of interest. The question might deal with testing acertain aspect of an economic theory, or it might pertain to testing the effects of a gov-ernment policy. In principle, econometric methods can be used to answer a wide rangeof questions.In some cases, especially those that i
16、nvolve the testing of economic theories, a for-mal economic model is constructed. An economic model consists of mathematicalequations that describe various relationships. Economists are well-known for theirbuilding of models to describe a vast array of behaviors. For example, in intermediatemicroeco
17、nomics, individual consumption decisions, subject to a budget constraint, aredescribed by mathematical models. The basic premise underlying these models is util-ity maximization. The assumption that individuals make choices to maximize their well-being, subject to resource constraints, gives us a ve
18、ry powerful framework for creatingtractable economic models and making clear predictions. In the context of consumptiondecisions, utility maximization leads to a set of demand equations. In a demand equa-tion, the quantity demanded of each commodity depends on the price of the goods, theprice of sub
19、stitute and complementary goods, the consumers income, and the individ-uals characteristics that affect taste. These equations can form the basis of an econo-metric analysis of consumer demand.Economists have used basic economic tools, such as the utility maximization frame-work, to explain behavior
20、s that at first glance may appear to be noneconomic in nature.A classic example is Beckers (1968) economic model of criminal behavior.Chapter 1 The Nature of Econometrics and Economic Data214/99 4:34 PM Page 2EXAMPLE 1.1(Economic Model of Crime)In a seminal article, Nobel prize winner Gary Becker po
21、stulated a utility maximization frame-work to describe an individuals participation in crime. Certain crimes have clear economicrewards, but most criminal behaviors have costs. The opportunity costs of crime prevent thecriminal from participating in other activities such as legal employment. In addi
22、tion, thereare costs associated with the possibility of being caught and then, if convicted, the costsassociated with incarceration. From Beckers perspective, the decision to undertake illegalactivity is one of resource allocation, with the benefits and costs of competing activitiestaken into accoun
23、t.Under general assumptions, we can derive an equation describing the amount of timespent in criminal activity as a function of various factors. We might represent such a func-tion asy H11005 f(x1,x2,x3,x4,x5,x6,x7), (1.1)wherey H11005 hours spent in criminal activitiesx1 H11005 “wage” for an hour s
24、pent in criminal activityx2 H11005 hourly wage in legal employmentx3 H11005 income other than from crime or employmentx4 H11005 probability of getting caughtx5 H11005 probability of being convicted if caughtx6 H11005 expected sentence if convictedx7 H11005 ageOther factors generally affect a persons
25、 decision to participate in crime, but the list aboveis representative of what might result from a formal economic analysis. As is common ineconomic theory, we have not been specific about the function f(H11080) in (1.1). This functiondepends on an underlying utility function, which is rarely known.
26、 Nevertheless, we can useeconomic theoryor introspectionto predict the effect that each variable would have oncriminal activity. This is the basis for an econometric analysis of individual criminal activity.Formal economic modeling is sometimes the starting point for empirical analysis,but it is mor
27、e common to use economic theory less formally, or even to rely entirely onintuition. You may agree that the determinants of criminal behavior appearing in equa-tion (1.1) are reasonable based on common sense; we might arrive at such an equationdirectly, without starting from utility maximization. Th
28、is view has some merit,although there are cases where formal derivations provide insights that intuition canoverlook.Chapter 1 The Nature of Econometrics and Economic Data3d 7/14/99 4:34 PM Page 3Here is an example of an equation that was derived through somewhat informalreasoning.EXAMPLE 1.2(Job Tr
29、aining and Worker Productivity)Consider the problem posed at the beginning of Section 1.1. A labor economist would liketo examine the effects of job training on worker productivity. In this case, there is little needfor formal economic theory. Basic economic understanding is sufficient for realizing
30、 thatfactors such as education, experience, and training affect worker productivity. Also, econ-omists are well aware that workers are paid commensurate with their productivity. This sim-ple reasoning leads to a model such aswage H11005 f(educ,exper,training) (1.2)where wage is hourly wage, educ is
31、years of formal education, exper is years of workforceexperience, and training is weeks spent in job training. Again, other factors generally affectthe wage rate, but (1.2) captures the essence of the problem.After we specify an economic model, we need to turn it into what we call an econo-metric mo
32、del. Since we will deal with econometric models throughout this text, it isimportant to know how an econometric model relates to an economic model. Take equa-tion (1.1) as an example. The form of the function f(H11080) must be specified before we canundertake an econometric analysis. A second issue
33、concerning (1.1) is how to deal withvariables that cannot reasonably be observed. For example, consider the wage that aperson can earn in criminal activity. In principle, such a quantity is well-defined, but itwould be difficult if not impossible to observe this wage for a given individual. Evenvari
34、ables such as the probability of being arrested cannot realistically be obtained for agiven individual, but at least we can observe relevant arrest statistics and derive a vari-able that approximates the probability of arrest. Many other factors affect criminalbehavior that we cannot even list, let
35、alone observe, but we must somehow account forthem.The ambiguities inherent in the economic model of crime are resolved by specify-ing a particular econometric model:crime H11005H92520 + H92521wagem + H92522othinc H11001 H92523freqarr H11001 H92524freqconvH11001 H92525avgsen H11001 H92526age H11001
36、u,(1.3)where crime is some measure of the frequency of criminal activity, wagem is the wagethat can be earned in legal employment, othinc is the income from other sources (assets,inheritance, etc.), freqarr is the frequency of arrests for prior infractions (to approxi-mate the probability of arrest)
37、, freqconv is the frequency of conviction, and avgsen isthe average sentence length after conviction. The choice of these variables is deter-mined by the economic theory as well as data considerations. The term u contains unob-Chapter 1 The Nature of Econometrics and Economic Data414/99 4:34 PM Page
38、 4served factors, such as the wage for criminal activity, moral character, family back-ground, and errors in measuring things like criminal activity and the probability ofarrest. We could add family background variables to the model, such as number of sib-lings, parents education, and so on, but we
39、can never eliminate u entirely. In fact, deal-ing with this error term or disturbance term is perhaps the most important componentof any econometric analysis.The constants H92520, H92521,H92526are the parameters of the econometric model, and theydescribe the directions and strengths of the relations
40、hip between crime and the factorsused to determine crime in the model.A complete econometric model for Example 1.2 might bewage H11005 H92520 H11001 H92521educ H11001 H92522exper H11001 H92523training H11001 u, (1.4)where the term u contains factors such as “innate ability,” quality of education, fa
41、milybackground, and the myriad other factors that can influence a persons wage. If we are specifically concerned about the effects of job training, then H92523 is the parameter ofinterest.For the most part, econometric analysis begins by specifying an econometric model,without consideration of the d
42、etails of the models creation. We generally follow thisapproach, largely because careful derivation of something like the economic model ofcrime is time consuming and can take us into some specialized and often difficult areasof economic theory. Economic reasoning will play a role in our examples, a
43、nd we willmerge any underlying economic theory into the econometric model specification. In theeconomic model of crime example, we would start with an econometric model such as(1.3) and use economic reasoning and common sense as guides for choosing the vari-ables. While this approach loses some of t
44、he richness of economic analysis, it is com-monly and effectively applied by careful researchers.Once an econometric model such as (1.3) or (1.4) has been specified, varioushypotheses of interest can be stated in terms of the unknown parameters. For example,in equation (1.3) we might hypothesize tha
45、t wagem, the wage that can be earned in legalemployment, has no effect on criminal behavior. In the context of this particular econo-metric model, the hypothesis is equivalent to H92521 H11005 0.An empirical analysis, by definition, requires data. After data on the relevant vari-ables have been coll
46、ected, econometric methods are used to estimate the parameters inthe econometric model and to formally test hypotheses of interest. In some cases, theeconometric model is used to make predictions in either the testing of a theory or thestudy of a policys impact.Because data collection is so importan
47、t in empirical work, Section 1.3 will describethe kinds of data that we are likely to encounter.1.3 THE STRUCTURE OF ECONOMIC DATAEconomic data sets come in a variety of types. While some econometric methods canbe applied with little or no modification to many different kinds of data sets, the spe-c
48、ial features of some data sets must be accounted for or should be exploited. We nextdescribe the most important data structures encountered in applied work.Chapter 1 The Nature of Econometrics and Economic Data5d 7/14/99 4:34 PM Page 5Cross-Sectional DataA cross-sectional data set consists of a samp
49、le of individuals, households, firms, cities,states, countries, or a variety of other units, taken at a given point in time. Sometimesthe data on all units do not correspond to precisely the same time period. For example,several families may be surveyed during different weeks within a year. In a pure crosssection analysis we would ignore any minor timing differences in collecting the data. Ifa set of families was surveyed during different weeks of the same year, we would stillview this as a cross-sectional data set.An important feature of cross-sectional data is that we can ofte
