Artful science in economics


This lecture will try to answer the following questions: What is a science?  How is it different from other disciplines, such as theology, the law?  Is auto mechanics a science?  What about folk medicine? How is science different from religion?

A discipline is a science if it tries to make sense of observable phenomena.  From the dictionary, we know that “phenomena” is not the name of a plant (like rhododendra).  The word refers to “facts perceptible by the senses.”  In this sense, a science is necessarily based on observation, and in practice most scientists make observations through careful experiment.  The methods of a science are usually also labeled “the scientific method,” and as we will find out, an important element of the scientific method is reliance on experiment or observation.  In other words, something scientific is necessarily empirical, i.e. based on the “real world.”


Dictionary definitions:

Science: Knowledge gained through experience; Observation, identification, description, experimental investigation, and theoretical explanation of phenomena. (Root:  Latin, scientia, knowledge.)

Art: Discipline that does not rely on the scientific method; Creative or imaginative activity.

Phenomenon: A fact perceptible by the senses.

Experiment: A test made to demonstrate a known truth.

Empirical:  Based on observation or experiment.

Religion: A cause or activity pursued with zeal or devotion.


Generally, science also essentially involves theorizing about causality:  “If A, then B” means that A “causes” B.  Example from economics:  If we raise the minimum wage, there will be more unemployment.  This is a statement made by an economist as a scientist.  Such a scientific statement is also classified as a positive statement, because it “posits” a theory or truth that can be verified empirically.

It is a positive statement, because it is also merely descriptive, making a claim about how the real world operates.  It would be a normative statement if the economist changed its form and said he favored not raising the minimum wage, because he likes more rather than less employment. A normative statement is a statement about “what should be.”  Normative statements are also known as prescriptive statements, because they “prescribe.” 


Positive statements can be contradicted by data. They can be verified empirically. Normative statements cannot, but such statements gain or lose acceptance based on our underlying fundamental values or religious/ethical beliefs. 

It is arguable that some normative statements can be rephrased as positive statements.  For example, a particular religion might state that if you sin, you go to hell. That looks like a positive statement, but it seems that no normal human being has come back from hell to tell us about it, and to verify that he did go there because he committed sin!  Thus, it appears that a key difference between religion and science is that the former cannot be subjected to experiment, whereas experiment is an essential feature of science. I note also that the idea of experiments in science is one of repetition: we keep doing and re-doing the same experiment, “just to be sure” it still comes out “the same”!

To summarize the difference between positive and normative statements:

          Positive is value-free.  It simply states that certain things lead to certain other things.  Example:  If you smoke, you are likely to get lung cancer.

          Normative is based on an underlying assumption that something is good or preferred, and we should do something to bring that good or preference to reality.  Example:  You should not smoke because you will get lung cancer, which is a bad thing, because we value the quality of our lives.

Going back to positive statements, one author states that the task of economic science is to discover those positive statements that are consistent with real world data (Parkin).  This is only partly correct.  Usually we have two or more positive statements that are consistent with the data, but these statements may point to different “theories.”

Example:  A is put on a lie detector.  The lie detector works on the theory that lying causes the body to produce signals such as higher blood pressure, pulse rate, etc.  But there could be an alternative theory:  A is a nervous person whose blood pressure and pulse rate rise in the presence of police; in other words, he may be honest, but the lie detector will say he is lying.  How can we tell the truth?  One way of deciding is to use the ceteris paribus condition:  We might accept the lie detector test if all the examinees of the test had the same level of nervousness.  In other words, the causality theory of the lie detector is limited to instances of “other things being equal” (the literal meaning of ceteris paribus).

Since science is about theorizing on causality, the choice of theory is really a choice on what we think is the causative element in the observable data. If we did not know that A was a nervous person, we might conclude that he is lying, when in fact he is honest.

I would also note that there is unending debate, at the philosophical level, as to whether causality can really be proved.  This is because we can find theories that “fit” the data, but this does not necessarily mean that they are the correct ones.  They are correct only until we find data that disprove some or all of them.  Then we have to start all over again!

If causality cannot be proved, there nonetheless remains an “artistic” criterion on how to choose theories.  This criterion is known as Occam’s Razor, and is a philosophical principle that marks a study as scientific.  The fact that science relies on Occam’s Razor means that there is “art” in science!


It is a principle, attributed to the medieval philosopher William of Occam, which states that we should prefer the theory that relies on the fewest assumptions and/or information requirements.  It is like saying the theory that is best is the one that can explain more phenomena than any other.  Intellectually, good science will give you greater bang for your theoretical buck. 

Here’s an example of how we might use Occam’s Razor.  Suppose that we have the following data.  B is late 5 times in 5 meetings.  At each meeting, when asked why he’s late, he gives 5 different answers: traffic, he got lost, the pedicab was slow, his driver didn’t show up, and his dog ate his homework.  Never mind that the last answer doesn’t make sense!  You can formulate 5 different theories on why B is late, if you believe him, and it would look like they all make sense.  They explain all the 5 occurrences of his tardiness.  But there is another theory:  B is habitually late because by being late he gets attention, and as they say, he is a KSP person.  Suppose also that it is true that B is KSP, and has been known to be habitually late.  Occam’s Razor says you will support the theory that the cause of B’s tardiness is his personality, and not any of the 5 excuses he gives.  Why?  Because, tomorrow, B will be late again, but you will be correct only 20% of the time that, for example, traffic is the “cause.” But his KSP will be a good and consistent predictor.


The preceding discussion allows us to list the essential characteristics of the scientific method:

          It is empirical – we look at the world, i.e., at phenomena and try to make sense out of it.

          It is theoretical, because we propose theories that explain the empirical data.

          We test the theories, and we can discard some using the method of logic.  If the theory predicts that B (the empirical data) follows from A (the theory), and then we find that when A happens, B does not happen, what do we conclude?  The theory is obviously wrong.  But if A happens, and B follows, is the theory correct?  It may be correct, but it has to compete with other theories that also “explain” B. 

Theories start from postulates or assumptions, and using logic, continues with certain theoretical results.  The theoretical results may then themselves result in “testable hypotheses” (which are testable positive statements).  Example:  In economics, we theorize or assume that humans prefer to eat more rice than less, but that they also like to eat meat (also more is better than less).  We also assume that their incomes are fixed or limited, and predict that the higher the price of rice (or meat), the less of each good they will consume.  The last clause is a testable statement.  If we see that as the price of rice increases, people eat more of it, then obviously the underlying theory is wrong, and as scientists we would have to begin again and look for a theory that works!

We choose from alternative theories using Occam’s Razor and our professional inclinations on what seems to be the best theory.  That’s why there are disputes and varying schools of thought within the profession.  For example:  Monetarists think inflation is monetary, caused by the central bank printing money.  Cost-push theorists think inflation is based on what drives up costs, such as the cartel decisions of oil-exporting countries.  Both theories explain why there was inflation in the 1970s when the price of oil rose dramatically.  Which theory is right is not clear. The cost-push school will say that the inflation was there to begin with and was reinforced by “accommodating” monetary policy: the central banks didn’t want to induce global recession so they printed more money.  The monetarists would say that if the central banks had not printed the money, the cartel’s power would not be effective.  Occam’s Razor seems to favor the monetary theory on this question.

AL GORE AND GLOBAL WARMING – A cautionary note on how Al Gore may be misleading all of us in the global warming debate

Science can also be misused.  Some economists consider that the recent alarm raised by environmentalists about global warming is in essence a “religious” matter (religion being defined in terms of an excess of zeal on behalf of a “cause”).  This is because it is also possible that global warming has “good” effects:  some cold places would become warmer, and certainly, if more ice melted, there would be less land and more sea, which would be welcomed by those who prefer seascapes to landscapes!  In other words, the fact that the earth is getting warmer, a scientific “fact” on its own, does not necessarily mean that we should do everything to get the earth to “cool down.”  The debate is really about the rights of those whom Al Gore and other environmentalists would want to suffer now or in the near future for the benefit of certain others in the farther future! 

(Revised June 12, 2008)


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s