On this page
Two senators debate the minimum wage. One says: "Raising the minimum wage will cost 500,000 jobs." The other says: "We should raise the minimum wage because workers deserve a living wage." These are fundamentally different kinds of claims. The first can be evaluated with data — economists can study employment in states that raised their minimum wages and compare outcomes to similar states that didn't. The second cannot be settled by evidence alone — it depends on values: how much weight to give worker income against potential job losses, and whether the distribution matters more than the aggregate. The first is positive economics; the second is normative.
The quick distinction
Positive economics makes descriptive claims about how the world works — cause and effect, if-then relationships, empirical regularities. These are in principle testable against evidence.
Normative economics makes prescriptive claims about how the world should be — what policies are good, what outcomes are desirable, what trade-offs are worth making. These depend on values and priorities that evidence alone cannot resolve.
| Positive | Normative | |
|---|---|---|
| Nature | Descriptive | Prescriptive |
| Example | "A tariff on steel raises domestic steel prices" | "We should protect the steel industry" |
| Testable? | Yes — with data | No — requires value judgment |
| Resolved by | Evidence | Democratic deliberation |
Positive economics, explained
Positive statements are the domain of empirical economics. The Congressional Budget Office's analyses are almost entirely positive: CBO estimates the cost of a bill, the employment effect of a tax change, the revenue impact of a rate adjustment. These estimates can be right or wrong — but they are evaluated against evidence, not values. The CBO explicitly does not take positions on whether a policy is good; it only models what it will do.
The Bureau of Labor Statistics employment data is a repository of positive economic facts: unemployment is 4.1 percent, wages grew 3.8 percent year-over-year, manufacturing employment fell by 12,000 last month. These are observations about what is happening, independent of what anyone thinks should happen.
Normative economics, explained
Normative statements enter whenever a claim uses "should," "ought," "fair," "best," or any other evaluative language. "The government should prioritize low unemployment over low inflation" is normative — it asserts a value hierarchy that cannot be settled by data. Two economists who agree completely on all the positive facts (how unemployment and inflation relate, how each responds to policy) can still disagree on which to prioritize, because the disagreement is about values.
This is not a weakness of economics — it is an honest recognition of the limits of empirical analysis. The Federal Reserve's dual mandate — maximum employment and price stability — is itself a normative choice made by Congress about what the central bank should pursue. The implementation of that mandate is then a positive question: what policy rate achieves those goals?
How to keep them straight
Ask: "Could this claim be proven false by data?" If yes — positive. If the claim would survive any empirical result because it rests on a value judgment — normative. Most heated economic policy debates conflate both: the positive dispute (what will this policy do?) and the normative dispute (what should we want it to do?) need to be separated before either can be resolved.
