No Such Thing as a Free Lunch: Principles of Economics (Part 3)

The first lesson of economics is scarcity: there is never enough of anything to fully satisfy all those who want it. The first lesson of politics is to disregard the first lesson of economics.

Thomas Sowell

The Scope and Method of Economics

(Part C)

by

Charles Lamson

Theories and Models

In many disciplines, including physics, chemistry, meteorology, political science, and economics, theories build formal models of behavior. A model is a formal statement of a theory. It is usually a mathematical statement of a presumed relationship between two or more variables.

A variable is a measure that can change from time to time or from observation to observation. Income is a variable---it has different values for different people, and different values for the same person at different times. There are countless other examples.

Because all models simplify reality by stripping part of it away, they are abstractions. Critics of economics often point to abstraction as a weakness. Most economists, however, see abstraction as a real strength.

The easiest way to see how abstraction can be helpful is to think of a map. A map is a representation of reality that is simplified and abstract. A city or state appears on a piece of paper as a series of lines and colors. The amount of reality that the mapmaker can strip away before the map loses something essential depends on what the map will be used for. If I want to drive from St Louis to Phoenix, I need to know only the major interstate highways and roads. I lose absolutely nothing and gain clarity by cutting out the local streets and roads. However, if I need to get around in Phoenix, I may need to see every street and alley.

Most maps are two-dimensional representations of a three-dimensional world; they show where roads and highways go but do not show hills and valleys along the way. Trail maps for hikers, however, have "contour lines" that represent changes in elevation. When you are in a car, changes in elevation matter very little; they would make a map needlessly complex and much more difficult to read. However, if you are on foot carrying a 50-pound pack, a knowledge of elevation is crucial.

Like maps, economic models are abstractions that strip away detail to expose only those aspects of behavior that are important to the question being asked.

Be careful---although abstraction is a potential tool for exposing and analyzing specific aspects of behavior, it is possible to oversimplify. Economic models often strip away a great deal of social and political reality to get at underlying concepts. When an economic theory is used to help formulate actual government or institutional policy, political and social reality must often be reintroduced if policy is to have a chance of working.

The key here is that the appropriate amount of simplification and abstraction depends on the use to which the model will be put. To return to the map example: you do not want to walk around San Francisco with a map made for drivers---there are too many steep hills.

All Else Equal: Ceteris Paribus It is almost always true that whatever you want to explain with a model depends on more than one factor. Suppose, for example, that you want to explain the total number of miles driven by automobile owners in the United States. The number of miles driven will change from year to year or month to month; it is a variable. The issue, if we want to understand and explain changes that occur, is what factors caused those changes.

Obviously, many things might affect total miles driven. First, more or fewer people may be driving. This number, in turn, can be affected by changes in the driving age, by population growth, or by changes in state laws. Other factors might include the price of gasoline, the household's income, the number and age of children in the household, the distance from home to work, the location of shopping facilities, and the availability and quality of public transport. When any of these variables change, the members of the household may drive more or less. If changes in any of these variables affect large numbers of households across the country, the total number of miles driven will change.

Very often we need to isolate or separate these effects. For example, suppose we want to know the impact on driving of a higher tax on gasoline. This change would raise the price of gasoline at the pump, but would not (at least in the short run) affect income, workplace location, number of children, and so forth.

To isolate the impact of one single factor, we use the device of ceteris paribus, or all else equal. We ask: What is the impact of a change in gasoline price on driving behavior, ceteris paribus, or assuming that nothing else changes? If gasoline prices rise by 10 percent, how much less driving will there be, assuming no simultaneous change in anything else---that is, assuming that income, number of children, population, laws, and so on all remain constant?

Using the device of ceteris paribus is one part of the process of abstraction. In formulating economic theory, the concept helps us simplify reality to focus on the relationships that interest us.

Conditions and Pitfalls In formulating theories and models, it is especially important to avoid two pitfalls: the post hoc fallacy and the fallacy of composition.

The Post Hoc Fallacy Theories often make statements, or sets of statements, about cause and effect. It can be quite tempting to look at two events that happened in sequence and assume that the first caused the second to happen. This is not always the case. This common error is called the post hoc, ergo propter hoc (or "after this, therefore because of this") fallacy.

There are thousands of examples. The Colorado Rockies have won seven games in a row. Last night, I went to the game and they lost. I must have "jinxed" them. They lost because I went to the game.

Stock market analysts indulge in what is perhaps the most striking example of post-hoc fallacy in action. Every day the stock market goes up and down, and every day some analyst on some national news program singles out one or two of the day's events as the cause of sudden change in the market: "Today the Dow Jones industrial average rose five points on heavy trading; analysts say that the increase was due to progress in talks between Israel and Syria." Research has shown that daily changes in stock market averages are very largely random. While major news events clearly have a direct influence on certain stock prices, most daily changes cannot be directly linked to specific news stories.

Very closely related to the post-hoc fallacy is the often erroneous link between correlation and causation. Two variables are said to be correlated if one variable changes when the other variable changes. However, correlation does not imply causation. Cities that have high crime rates also have lots of automobiles, so there is a very high degree of correlation between number of cars and crime rates. Can we argue, then, that cars cause crime? No. The reason for the correlation may have nothing to do with cause and effect. Big cities have lots of people, lots of people have lots of cars, and therefore big cities have lots of cars. Big cities also have high crime rates for many reasons---crowding, poverty, anonymity, unequal distribution of wealth, and readily available drugs, to mention only a few. However, the presence of cars is probably not one of them.

The Fallacy of Composition To conclude that what is true for a part is necessarily true for the whole is to fall into the fallacy of composition. In short: Theories that seem to work well when applied to individuals or households often break down which when they are applied to the whole.

Testing Theories and Models: Empirical Economics In science, a theory is rejected when it fails to explain what is observed or when another theory better explains what is observed.

Economic theories are also confronted with new and often conflicting data from time to time. The collection and use of data to test economic theories is called empirical economics.

Scientific research often seeks to isolate and measure the responsiveness of one variable to a change in another variable, ceteris paribus. Physical scientists, such as physicists and geologists, can often impose the condition of ceteris paribus by conducting controlled experiments. They can, for example, measure the effect of one chemical on another while literally holding all else constant in an environment that they control completely. Social scientists, who study people, rarely have this luxury.

*MAIN SOURCE: CASE & FAIR, 2004, PRINCIPLES OF ECONOMICS, 7TH ED., PP. 9-12*

end