Chapter 1 Notes: Basic Concepts of Research in Economics

Chapter 1 Basic Concepts of Research in Economics In the first section, we start with the definition of various terms relating to research. Terms to be discussed are ‘research’, ‘research methods’ and ‘research methodology’ and, finally, a brief discussion of various types of research. In the second section, we will discuss what is economics and what economists do. This discussion is presented at the outset to illustrate and highlight various skills needed to carry out economic analysis. What is attempted here is to highlight the importance and interdependence of economic theory and measurement in the study of economics. [Further discussion of ‘theory without measurement’ versus ‘measurement without theory’ comes here]. In the third section, we introduce basic concepts of research. As is well known, the method of research or analysis economists use in carrying out their task is the scientific method, which is used in all of science. Therefore, it is important to discuss science in general and its method, namely scientific method. We should note, however, that there is no such a thing as the scientific method, because there are many variations. Scientific method essentially refers to the general or generalized process called the "scientific approach" to obtaining new and reliable knowledge. What we attempt to do in this section is to discuss the key terms and concepts of the scientific method, before we delve into research procedures in the next chapter. These concepts include theory and model, variables, assumptions, parameters, the hypothesis, and the testing of hypothesis among others. 1.1 Key Terms in Research What is Research? The term "research" is often loosely defined and thus used in a similar way. This unfortunate development results from a misconception about what is research. To properly understand what is research, it is good to start with common misconceptions about research. First, fact transferal is not research. Consider a typical high school research project. The teacher assigns a "research project" on some topic. The students went to the library, checked out several books, and might have copied several pertinent pages from the book. The typical student organized collected information and wrote up the "research report". What these students did is information gathering and organization; it is nothing more or nothing less. No doubt the student went through some motions associated with research. But finding fact and fact transferal alone is not research. Transfer of information from one source, namely books and pertinent pages, to another source, namely the so-called research report, is nothing more than fact transferal, but not research. To my distress I find many college students repeat this same mistake by submitting a ‘research report’ which is nothing but fact transferal from one source to their report. A second misconception about research is that research is related to laboratory research (for example, in chemistry or biology in the natural sciences). When people hear term the "research", they often conjure up this image. But research is not limited to certain fields of study; it is characterized by the methods used. What, then, is research? Webster’s Collegiate Dictionary (1996?) defines research as "studious inquiry or examination; esp: investigation or experimentation aimed at the discovery and interpretation of facts, revision of accepted theories or laws in the light of new facts, or practical application of such new or revised theories or laws". Research Methods and Research Methodology Research methods provide the specific details of how one accomplishes a research task (procedures and methods); It provides specific and detailed procedures of how to initiate, carry out, and complete a research task by mainly focusing on how to do it. Research methodology deals with general approaches or guidelines to conducting research. It provides the principles for organizing, planning, designing, and conducting research, but it cannot tell you in detail how to conduct a specific, individual research. In carrying out an applied and quantitative economic research, there are several necessary backgrounds. The researcher should first have solid training in economic theory, quantitative methods(statistics and econometrics), data analysis techniques, and adequate training in micro-computer technology, as well as some training in research methods. Unfortunately, undergraduate students beginning their research most often do not have these backgrounds. Therefore, they are going to be overwhelmed and intimidated by the lack of necessary skills. Many often give up in frustration, even if they are willing to persevere and to learn these skills. What is sorely need is a practical guide to initiate, conduct, and complete an applied and quantitative economic research. One proven and effective way of learning these various skills which economists use is learning by doing by example. Types of Research Basic vs. Applied Research: The distinction between basic and applied research is largely by the focus of its application. This distinction comes from basic science vs. applied science. Example: physics and engineering. Basic research focuses on determining or establishing the basic or fundamental relationships within a discipline without paying attention to any practical applications to the real world. In contrast, applied research is usually conducted to solve a particular and concrete problem. Descriptive vs. Analytical Research: The distinction between descriptive and analytical research is based on the question it asks. Descriptive research attempts to determine, describe, or identify what is, while analytical research attempts to establish why it is that way or how it came to be. The descriptive research uses description, classification, measurement, and comparison to describe what phenomena are. The analytical research usually concerns itself with cause-effect relationships. Examples. Examining the fluctuations of U. S. international trade balance during 1974-1995 is an example of descriptive research; while explaining why and how U.S. trade balance move in a particular way over time is an example of analytical research. Another example: Starting from late 1986, the value of U.S. dollar value has steadily increased against the Japanese yen and German Mark. Examining the magnitude of this trend in the value of U.S. dollar is another example of descriptive research; while explaining how and why this surge in the value of the U.S. dollar is occuring. If one attempts to explain how and why this surge in the value of U.S. dollar is going to affect the U.S. economy,as well as the economies of Japan and Germany, this is another example of analytical research. Disciplinary, Subject-Matter, and Problem-Solving Research: By disciplinary research, we mean research "designed to improve a discipline" as Johnson(1986) defines it. It dwells on theories, relationships, and analytical procedures and techniques within the discipline. Examples: Economic research or social research. By subject-matter research, we mean research "on a subject of interest" within a discipline. Example: Research in resource economics or in international economics. By problem-solving research, we mean research "designed to solve a specific problem for a specific decision maker". It is often multidisciplinary. Example: A multidisciplinary study of on the demand for new mass transit involving economics, sociology, and civil engineering. Or a multidisciplinary study of new medical surgery involving medical doctors, engineers, and an economist. 1.2 What is economics and what do economists do? What is economics ? Case and Fair define economics as "the study of how individuals and societies choose to use the scarce resources that nature and previous generations have provided" Since it deals with the behavior of human beings and their interactions, it is a social science. Stiglitz also defines it as a social science, but he adds that "it studies the social problem of choice from a scientific viewpoint, which means that it is built on a systematic exploration of the problem of choice". This systematic exploration involves both the formulation of theories and the examination of data to test the validity of theories. This latter definition of economics by Stiglitz is helpful in identifying two major tasks of economists: They are observing facts and trends in data and explaining and interpreting certain or whole aspect of the economy. From the above definitions of economics, the importance of both theory and measurement is obviously the essential ingredients of meaningful and serious economic research. Theory without measurement is unfruitful and empty; measurement without theory is equally meaningless. A theory consists of a set of assumptions (or hypotheses), and conclusions derived from those assumptions. Theories are logical exercises: if the assumptions hold, then the results follow. Examples. When theory is formally presented, it is called model. Examples of model car, ship, or airplane to understand how economists use models. Bamoul and Bliner’s definition of theory: "A theory is a deliberate simplification of relationships whose purpose is to explain how those relationships work. It is an explanation of the mechanism behind observed phenomena. The relationship is usually couched in terms of the relationship between or among variables. What is variable? A variable is anything which varies over time and space; then the measurement of variable in terms of numbers or facts expressed in quantitative terms are data. The data are then quantitative information on various aspects of economics and business. We need to learn how to describe and analyze data. The description of data is one area of statistics: descriptive statistics. Data analysis involves the analysis of statistical data of one variable (univariate distribution). Data analysis also analyzes the relationship between two variables (bivariate distribution). Data analysis may involve the analysis of among more than two variables (multivariate distribution). The presentation of data can either be tabular and graphical. In this part, both the description and computation, illustration, using any easily available computer softwares are to be presented side by side. What do economists do? In a certain sense, economics is really what economists do. Two major tasks on which economists spend their energies are: First, observing facts and trends in data and, second, explaining the relationship(s) between them in a cause-effect fashion. Observing Facts and Trends in Data: A fact is a verifiable observation or phenomenon. It is an observed fact that retail gasoline prices in California are higher than elsewhere in the U.S. The problem is then to explain why and how this is so. A fact(s) is usually presented in terms of variable(s). A variable is a quantity of something which varies and the researcher is interested in. There are two types of variables: discrete (or discontinuous) and continuous variables. When a variable(s) is continuous, the researcher can identify a trend in it. Examples of trend are a linear trend, parabolic or quadratic trend. There are several publications containing fact and trends of economic variables on the overall U.S. economic conditions. The first is the Statistical Abstract of US and Economic Report of the President, the annual publication by the Council of Economic Advisors. Two other well-kown publications on more current economic situations are Economic Indicators, a monthly publication prepared by the Council of Economic Advisors and Economic Trends, another monthly publication by the Federal Reserve Bank of Cleaveland. To obtain specific information on some specific aspect of the economy, say health care or energy, one needs to consult specialized publications. Some examples of interesting facts and trends in the US economy during the last two decades are: • • • Heath care spending in the U.S. has increased faster than other spending categories than before. The Budget deficit and national debt have increased faster than before. Foreign exchange rates have fluctuated more widely than before. When a fact or phenomenon is measured (quantified), organized, and presented in a desirable fashion, we call them data. Data provide factual information. They are usually regarded as reliable subject to the implicit understanding that there may be some sampling or measurement errors in the process of collection. To learn how to work with macroeconomics data on the U.S. economy, consider the following set of major economics data on the U. S. economy for the period from 1960 through 1995. Home Work Assignment. Spread sheet data & How to Use Excel for Data Analysis [Ed’s work] Measurement of variables in level, first difference, index numbers, percentage changes, per capital measures, converting nominal to real variables, fixed weight vs. chain weight GDP measure, adjustments to data (seasonal adjustments, various moving averages, normalized variables etc. Use of spread sheet (Excel) to manipulate data and graph. When we examine facts and trend in the variable (s), we often present information in a table or in a graph. When we examine facts and trend, we need to be mindful of the distinction between correlationvs. causation between the two variables. There are a number of important issues relating to measurement of variables, and these issues will be taken up later chapter on data and data analysis. Explaining and Predicting the Relationship: Even though fact or phenomena provides useful factual knowledge, fact alone is not interesting. Even though facts are interesting, they are not very useful by themselves. As C.S. Peirce expressed so clearly, "a person can stare stupidly at phenomena; but in the absence of imagination they will not connect themselves together in any rational way"* More interesting and useful are raising analytical questions regarding facts. With the first fact on rapid increase in health-care spending, one can ask why and how health-care spending increased faster than spending in other sectors? Relating to this question, one can also ask why and how the price of health-care increased faster than other prices? Observing another fact, say the large fluctuations in foreign exchange rates, one may ask whether or not, and why and how a change in the international monetary arrangement might affect the pattern of change in exchange rate. To answer this question, we need to have a measure for various exchange rates. They are bilateral versus multilateral, or the effective exchange rate, and the nominal versus real exchange rate. Given the first fact, the problem is to explain why and how retail gasoline prices have increased during this particular period and to explain why and how retail gasoline prices in California are substantially higher than in the United States. Explaining the relationship between the share of health care expenditure out of GDP requires the use of the scientific method. Therefore, it is necessary for us to be familiar with concepts of research , such as hypothesis, theory, and model. We also have to learn operational terms of scientific method, such as variables, assumptions, parameters, and functional form. Explaining and predicting requires the formulation of economic theories as to specific issue under consideration and testing the validity of theories. **H.R.’s earlier writing on what is a theory and how to do it goes here. 1.3 Scientific Method Scientific vs. Non-scientific Methods of Gaining Knowledge Having finished a brief discussion of the meaning of research and related terms, we are now ready to take up the discussion of the scientific method. Let us start by stating that there are essentially two methods of obtaining knowledge: scientific and non-scientific methods. Let us start with non-scientific method. Non-Scientific Method: The first method of gaining knowledge is through senses, experience, intuition, and revelation, all of these may be classified as non-scientific methods. Some gain knowledge through physical senses - sight, sound, touch, taste, and smell -, and experience. Senses and Experiences. Some knowledge is obtained by senses and experiences. When one gets too close to a fire and gets burned once, he or she gains the knowledge that it is dangerous to be too close to the fire. Other gain knowledge by intuition or revelation. Intuition. Some rely on intuition as a source of knowledge. Intuition is the strong hunch or feeling that what one perceives to be the case is indeed true. If one strongly believes what one perceives is real and true, knowledge thus is obtained. While there is no reason to doubt the truthfulness of the knowledge obtained by intuition, like knowledge obtained from senses and experiences, it is subjective. Revelation. Some knowledge are obtained by revelation. Revelation is the presentation of the truth from a supernatural source, such as deity. Knowledge acquired via experience, intuition, revelation, and even measurement remain as private knowledge. The validity of knowledge obtained through nonscientific methods cannot be subject to objective testing. Scientific Method: The second method of obtaining knowledge is the scientific method. This method of gaining knowledge is learning by reasoning. It is considered today to be the most reliable method of gaining knowledge. In contrast with the scientific method, the validity of knowledge obtained by scientific method can be subject to testing. In all science, research proceeds within the framework of the scientific method. According to Lastrucci, "science may be defined as an objective, logical, and systematic method of analysis of phenomena devised to permit the accumulation of reliable knowledge".* His definition of science contains all essential elements of scientific methods. First it defines science as the method of analysis. [for details, see page 7 "Method" Social Research by Sedlack and Stanley]. Second, it highlights three major characteristics of the scientific methods as objective, logical and systematic. Let’s explore the meaning of these three characteristics. First, scientific method is objective, not in the sense of being value free. But it is objective in the sense that the analysts are not biased or prejudiced or subject to personal whims. Second, scientific method is logical in the sense that science follows logical reasoning. Logical reasoning is thinking in reasonable fashion. It is sufficient to point out two types of reasoning process, namely deductive and inductive logic. A full discussion on the logical process will be presented below. Third and finally, the search for truth in science is systematic. This means that researchers follow a systematic set of procedures through which knowledge is gained. Deductive and Inductive Logic Let us now turn our attention to the logical process in scientific method. First, note that there are two types of reasoning process, namely deductive and inductive logic. Deductive logic is the process of reasoning from general conditions or premises using assumptions to specific conclusions. Economic theory rests largely on deductive logic. We establish a set of assumptions about conditions and behaviors to arrive at conclusions through logical process. Examples: utility maximization in the consumer behavior; profit maximization in the producer behavior. Inductive logic is reasoning form the specific outcomes to a generalized conclusions. This is usually done by observing many individual experiences and cases to formulate a general conclusion. The inductive logic of reasoning is followed in most empirical economic research. Hypothesis, Theory, and Law We deduce from facts and assumptions to a hypothesis. "A hypothesis is a tentative assertion of a relationship between factors or events that is subject to verification or rejection." In short, a hypothesis is a testable proposition of the relationship between or among variables. How is hypothesis related to theory? And theory to law? Hypothesis is a single statement that attempts to explain a single interesting or puzzling phenomenon. In other words, a hypothesis is a testable proposition on an interesting or puzzling phenomenon. It usually takes the form of an educated guess or conjecture. Usually the hypothesis is based on facts and assumptions. Theory is a whole system of thought (or systematic explanation) that refers to many phenomena and whose parts are related to one another in deductive, logical form. Law: A theory that has been subjected to extensive testing over time and across space, and that has won virtually universal acceptance, is called a law. For instance, the law of supply and demand refers to the commonly observed phenomena that, in a free market, the forces of supply and demand generally push the price toward its equilibrium level, the price at which the quantity supplied and quantity demanded are equal. Another example is the law of diminishing marginal returns. Language of scientific method In the previous section, we first learned that the essential element of science lies in its method of analysis, and that the three main characteristics of the scientific method are that it is objective, logical, and systematic. Then we learned about the process of scientific method beginning from, say, deductive logic to hypothesis, from hypothesis to theory, and from theory to law. Now we need to learn the operational aspects of the scientific method by focusing on terms and concepts of research. In particular, we need to learn basic concepts such as assumptions, variables, parameters, and functional forms. Finally, we need to learn something about ceteris paribus. Assumption In our study of gasoline price hike, we first assume that consumers are "rational" in that their decision-making in the purchase of gasoline is consistent with maximization of consumer satisfaction. Likewise, suppliers of gasoline (retail gasoline merchant) are also assumed to be rational in that their decision-making in pricing, inventory, etc. is consistent with profit maximization. We theorize that the retail gasoline price is jointly determined by the forces of demand and supply of gasoline. On the demand side, the quantity of gasoline demanded depends inversely on the retail price of gasoline, positively consumer income, and a host of other factors. On the supply side, the quantity of gasoline supplied depends positively on the price of gasoline price, cost of production (including price of crude oil and refinery cost), and a host of other factors, including the gasoline tax and environmental cost. Once we introduce the role of assumption in the research, it appropriate to further introduce related basic concepts of research: variables, functional relationship and parameters. Variables in General A variable is a quantity of something which varies and you are interested in. Price of gasoline is a variable to an economic analyst studying the recent gasoline price increase in 1996 but not to most motorists and not even to an economic analyst studying the relation between stock and bond prices. The researcher chooses his or her variables. Choosing variables correctly is one of the first essential step of carrying out research. Therefore, choose variables with extreme care. To choose variable correctly, one has to know the two types of variables: the dependent variable and independent variable. The Dependent Variable The dependent variable is that quantity whose change the researcher wants to find out, explain, or predict. In the cause-effect relationship, the effect variable is the dependent variable. In the study of the demand for gasoline, the quantity of gasoline demanded is the dependent variable, because the quantity of gasoline demanded changes in response to changes in gasoline price, consumer income, and other demand side factors. This researcher is ultimately interested in measuring the impacts of these changes on gasoline price In some cases, however, the dependent variable is not quantity, but represents a qualitative choice. An example of the latter is decision to buy or not; or marry him (or her) or not. The Independent Variable The independent variable is a variable whose effect upon the dependent variable one is trying to understand, explain, and predict. In the cause-effect relationship, it is the cause variable. Using the study of the demand for gasoline, the independent variables are the price of gasoline, consumer income, and other variables such as fuel efficiency and population characteristics. The Parameter The concept of parameter is a bit more tricky to define. A parameter is a quantity measuring the response of the dependent variable to change in the independent variable, and is usually assumed to remain constant during the period of study. To illustrate, Figure 2.1 shows the hypothetical relationship between the quantity of gasoline demanded and gasoline price over time. On the vertical axis, the quantity of gasoline demanded measured in gallons. On the horizontal axis, the price of gasoline is measured in U.S. dollars per gallon The a scatter plot of cross represents the combination of the quantity of gasoline demanded and corresponding gasoline price in each year over the period. The line is drawn through the scatter plot of the quantity of gasoline demanded and gasoline price as closely as possible. The line represents a best guess of the average relationship between the observed quantity of gasoline demanded and corresponding gasoline price during the period of study. The Functional Form The algebraic relationship Qdg = a + bPg is linear relationship between the two variables Qd and P. It is linear both in the parameters and variables. Parameters like a2 or 1/a. are non-linear; variables like Q2 and 1/Q are examples of non-linear. When there is a relationship between two variables, we say that " y(dependent variable) is a function of x(independent variable)" in the mathematical language. By this we mean to say that the magnitude of y depends on the magnitude of x. But when the dependent variable y is determined by not only by x but also by other independent variables, we say that y is a function of x1, x2, and x3. Or in short y = f(x1, x2, x3). In the gasoline demand example, the quantity of gasoline demanded (Qdg) depends upon ( or is a linear function of) price of gasoline (Pg) and consumer income (Y). Algebraically, we write: Qdg = f (Pg, Y). Linear functional form of this relationship can be written as Qd = a + bPg + cY. Since parameters a, b, and c enter the equation are linear, the relationship is linear. But the relationship may be other than linear functional forms. They include the log-linear model, semi-log models, and reciprocal models. Most - but not all- economic relationship is the causal relationship between the dependent variable and one or more independent variable. The value of the dependent variable y that is observed for subject in the study depends on three elements: First, it depends on the level of the independent variable; second on the influence of the independent variable(s) on the dependent variable, and finally on the influences of all other variables that may act upon the value of the subject but you are not interested in and thus excluded in the specification of the functional relationship. These excluded variables are assumed to remain the same during the period of study. Two Types of Relationship: Correlation and Causation When variables are related in certain fashion, there are two kinds of relationships. When one variable is related with another variable, we say that there exists correlation between the two variables. More precisely, correlation or more precisely simple correlation measures the degree of a linear association between the variables. When one measures correlation among more than two variables, one measures partial correlation. When analysts are interested in measuring the degree of correlation. If they move together in the same direction, there is a positive correlation; if they move in opposite directions, there is negative correlation. The share of health care spending in GDP and the relative price of health care are positively correlated. More examples of correlation. When there exists a cause and effect relationship between two or more variables, we say that causation runs from cause variable to effect variable. Causation then means changes in one variable brings about changes in another variable. The cause-effect relationship is often called the dependence relationship. Using the example of share of health-care spending out of GDP and relative price of health-care, health care share is the effect or dependent variable; and the relative price of health care is cause and independent variable. Does that mean that the health care share is always the dependent variable and the relative price of healthcare is always the independent variable? In other words, is the dependence or causality inherent in the variable themselves? Obviously, answer is no. But then how does a researcher decide on which variable depends on what variable? It is deliberately chosen and imposed by researchers, usually from theory or existing studies on the subject. Importance of determining the causal-relationship in research: Art of specifying causal relationship? Experience and knowledge? In the study of gasoline demand, the quantity of gasoline demanded is assumed to depend on the price of gasoline. There are two reasons for the direction of causality running from the gasoline price to the quantity of gasoline demanded, but not the other way around. When we study the individual demand for gasoline, the price of gasoline is determined by aggregate decisions of all consumers and suppliers. When there are many consumers in the market, a reasonable assumption, the role played by an individual consumer is small enough to be ignored in the determination of retail gasoline price. Analysts then are interested in studying how an individual consumer respond to changes in gasoline price. When economists are interested in studying the feedback effect from aggregate quantity of gasoline demanded to gasoline price, they build simultaneous equation model where the forces of aggregate demand and supply are allowed to interact in determining the gasoline price. -two stage method of estimating gasoline price. Regression analysis to study the dependence relationship. Most economists consider economics as a science because they follow the scientific method. Nobel prize in natural science (physics, chemistry, and medicine). Some economists argue that their discipline does not qualify as science because See Tregarten, p. 15 for specific reasons and ensuing controversy Regardless of the merit or demerit of whether economics should be viewed as science or not, it is true that economic research uses the scientific method. It is good for us to learn the language and process of the scientific method. Modeling Causal Relationship: An Example John Taylor’s Econ 1A book example of US health care study in circular flow: Gasoline price in circular flow: consumers, retail & wholesale gasoline market, refinery, and crude oil producers with gasoline tax and environmental cost. Assumptions about people’s behavior , institutions, and their environment affecting decision making [Jullion Simon, p. 37] Deduction: We deduce Hypothesis is a single statement that attempts to explain or predict Theory is an entire system of thought that refers to many phenomena and whose parts can be related to one another in deductive logical form. Ceteris Paribus Theory and Model Graphical Illustration of Causal Relationship: Relationships among the key concepts of research, such as variables, parameters, and functional forms can be easily illustrated using algebraic relationship with graphical illustration. In algebraic terms, the formula for a straight line is Qdg equals a plus bPg. More concisely, Qdg = a + bPg where Qdg is the price of gasoline and Pg is the price of gasoline and where a and b are parameters, representing the intercept and slope of the algebraic relationship. The intercept term tells us what the demand for gasoline would be if the price of gasoline is zero. But this is a mechanical interpretation of the intercept, and it is reasonable to interpret the intercept as the net influence of other variables in the relationship, namely consumer income and other variables in the case of gasoline demand. The slope coefficient "b" tells us how much the quantity of gasoline demanded would change as consumer income change on the average. The parameter are usually assumed to be constant. But they may change over time and across space. In another time period or in another country, the parameter values may not be the same as those above.