Models, evidence and value judgements
| English | Français |
|---|---|
| ceteris paribus/ˈsetərɪs ˈpærɪbəs/ | ceteris paribus |
| normative statement/ˈnɔːmətɪv ˈsteɪtmənt/ | énoncé normatif |
A decision you can investigate
- A researcher predicts that a higher bus fare reduces journeys. During the same month, fuel costs and school attendance also change.
- A before-and-after observation alone does not isolate the fare effect.
Build the explanation
- An economic model simplifies behaviour to explain a relationship. Ceteris paribus 其他条件不变 holds other relevant factors constant. Economics is a social science: whole economies cannot normally be controlled like a laboratory experiment, so data comparisons need careful assumptions.
- A positive statement is testable against evidence; it may still be false. A normative statement 规范性陈述 includes a value judgement about what should happen. Policy choices combine evidence about effects with values about equity, freedom or welfare.
Work through the evidence
- A model predicts Q = 200−5P. At fare 10, predicted journeys are 150; at 12, they are 140. The isolated model effect is −10 journeys.
- If actual journeys fall from 150 to 125 while attendance also falls, the observed −25 cannot all be assigned to the fare. “A fare rise lowers journeys, other factors unchanged” is positive. “The council should keep fares affordable even at a subsidy cost” is normative.
What does the model predict at fare 12?
Q = 200−5 × 12 = 140.
Test the limits
- Positive does not mean desirable or proven. An estimated model can omit variables or apply poorly to another setting. Simple models are useful only when their assumptions suit the question.
- Evidence can challenge the predicted effect; it cannot by itself decide how much weight to place on affordability versus public cost. Distinguish the causal estimate from the ethical choice.
Which statement is normative?
“Should” expresses a policy value judgement here.
A model result automatically proves the same causal effect in every observed economy.
External validity and other changing factors must be examined.
Apply and explain your answer
- Why does the observed fall of 25 not identify a fare-only effect?
- Other relevant factors changed; ceteris paribus did not hold in the observed comparison.
Can a positive claim be incorrect?
Evidence can reject a testable claim.
Use the terms precisely
- ceteris paribus: Holding other relevant factors unchanged when analysing a relationship.
- normative statement: A claim incorporating a value judgement about what should happen.
Match the terms to their meanings.
Use each term for its stated economic relationship.
A model predicts Q = 200−5P. At fare 10, predicted journeys are 150; at 12, they are 140. The isolated model effect is −10 journeys. If actual journeys fall from 150 to 125 while attendance also falls, the observed −25 cannot all be assigned to the fare. “A fare rise lowers journeys, other factors unchanged” is positive. “The council should keep fares affordable even at a subsidy cost” is normative.
Positive does not mean desirable or proven. An estimated model can omit variables or apply poorly to another setting. Simple models are useful only when their assumptions suit the question. Evidence can challenge the predicted effect; it cannot by itself decide how much weight to place on affordability versus public cost. Distinguish the causal estimate from the ethical choice.
An economic model simplifies behaviour to explain a relationship. Ceteris paribus holds other relevant factors constant. Economics is a social science: whole economies cannot normally be controlled like a laboratory experiment, so data comparisons need careful assumptions.