The success of modern science rests partly on the ability of scientist to separate their views on what does happen from their views on what they would like to happen.
Positive Statements concern what is, was or will be. Positive statements, assertions, or theories may be simple or complex, but they are basically about matters of fact. Positive statements offer various assertions about the world. If it is possible for a positive statement to be proved wrong by empirical evidence, we call it a testable statement.
Normative Statements concern what one believes ought to be. They state, or are based on, judgments about what is good and what is bad (called value judgments). They are thus bound up with philosophical, cultural, and religious systems. Normative statements are not testable. Disagreements over such normative statements as “It is wrong to steal” or “It is immoral to have sexual relations out of wedlock” cannot be settled by an appeal to empirical observations. It is therefore useful to separate normative and positive inquiries. We do this not because we think one is less important than the other but merely because they must be handled in different ways. The question: ‘What government policies will reduce unemployment’ is a positive statement whereas the question: ‘Should we be more concerned about financial instabilities on the liberalized markets’ is a normative one. This distinction can be very important. As an example of the importance of this distinction, consider the question “Has the payment of generous unemployment benefits increased the amount of unemployment?” This positive question can be turned into a testable hypothesis such as “The higher the benefits paid to the unemployed, the higher will be the total amount of unemployed.” If we are not careful, however, attitudes and value judgments may get in the way of the study of this hypothesis. Some people are opposed to all welfare measures and believe in an individualistic self-help ethic. They may hope that the hypothesis is correct because its truth could then be used as an argument against welfare measures in general. Others feel that welfare measures are desirable, reducing misery and contributing to human dignity. That may hope that the hypothesis is wrong because they do not want any welfare measures to come under attack. In spite of different value judgments and social attitudes, however, evidence is accumulating on this particular hypothesis. As a result, we have more knowledge than we had 20 years ago of why and how much unemployment benefits increase unemployment. This evidence could never have been accumulated or accepted if investigators had not been able to distinguish their feelings about how they wanted the answer to turn out from their assessment of evidence on how people actually behaved. The distinction between positive and normative statements allows us to keep our views on how we would like the world to work separate from our views on how the world actually does work. We may be interested in both. It can only obscure the truth, however, if we let our views on what we would like to be bias our investigations of what actually is. For this reason, the separation of positive from normative statements is one of the foundation stones of science. It is also for this reason that scientific inquiry, as it is normally understood, is usually confined to positive questions. However, the classifications positive and normative do not cover all statements that can be made. For example, there is an important class of statements, called analytical statements, whose validity depends only on the rules of logic. Thus the sentence, “if all market economies are efficient and if the X economy is a market economy, then it is an efficient economy” is a valid analytic statement. It means that if two things are true, then a third thing must also be true. Indeed, the sentence ‘all market economies are efficient” is a positive statement that can be refuted. Analytic statements, which proceed by logical analysis, play an important role in scientific work and form the basis of much of our ability to theorize. Moreover, not all positive statements are testable. A positive statement asserts something about some aspect of the universe in which we live. It may be empirically true or false in the sense that what it asserts may or may not be true of the world. Some positive statements are refutable and some others are irrefutable. We have seen that normative questions cannot be settled by a mere appeal to facts. In democracies, normative questions relating to government policies are often settled by voting. So on one hand, we look to observations to shed light on the issue of the extent to which unemployment insurance deters people from working. On the other hand, the political process to decide whether or not, when all the pros and cons are considered, we should have such insurance. Economists need not confine their discussions to positive, testable statements. Economists can usefully hold and discuss value judgments. Indeed, the pursuit of what appears to be a normative statement, such as “unemployment insurance ought to be abolished,” will often turn up positive hypotheses that underlie the normative judgment. In this case, there are probably relatively few people who believe that government provision of unemployment insurance is in itself good or bad. Their advocacy or opposition will be based on beliefs that can be stated as positive rather than normative hypotheses; for example, “Unemployment insurance causes people to remain unemployed when they would otherwise take a job” or “Unemployment insurance increase the chance that workers will locate the jobs to which they are best suited by supporting them while they search for the right job.”
