How Economists Approach the World

Economics as a social science

Economics studies human behaviour, which makes it a social science rather than a natural one. That has a practical consequence: economists cannot usually run controlled experiments. They cannot hold a whole economy constant, change one variable and observe the result.

▶

Instead they build models. Use historical data, and reason carefully about what would happen if one thing changed. Because the evidence is rarely decisive, economists genuinely disagree, not only about values but about how the economy works. The Keynesian and new-classical views of the long-run aggregate supply curve are a live example that runs through the whole macro syllabus.

Recognising this is itself a syllabus point, and acknowledging competing viewpoints is a mark of a strong answer rather than a weakness in it.

Models and ceteris paribus

A model is a simplified representation of reality, built to isolate the relationships that matter.

Simplification is the point, not a flaw. A map that showed every detail of the terrain would be useless; a model that included every influence on demand would be impossible to reason with. The question to ask of a model is not "is it realistic?" but "is it useful for this purpose?"

Ceteris paribus means "all other things being equal". It allows an economist to examine the effect of one variable while holding the others constant.

The demand curve is drawn ceteris paribus: it shows how quantity demanded responds to price assuming income, tastes, and the prices of other goods are unchanged.

This is exactly why a change in the good's own price causes a movement along the curve, while a change in anything else, one of the things being held constant, shifts it. The distinction that dominates Unit 2 comes directly from this assumption.

The limitation: in reality several things change at once. That is why real-world outcomes often differ from a model's prediction, and pointing this out is legitimate evaluation rather than an objection to using models at all.

Correlation is not causation

Because economists work with observational data rather than experiments, this distinction does most of the work that a control group would do in a natural science, and it is a legitimate evaluative move on any data-response question.

Two variables moving together can arise in four different ways:

• A causes B: the claim usually being made.
• B causes A: reverse causality. Countries with more police have more crime; the crime is what caused the police to be hired.
• A third factor causes both: a confounding variable. Ice cream sales and drowning rise together because both follow hot weather.
• Coincidence, particularly in short data series.

The practical consequence for an exam: when a stimulus shows two series moving together. Say what else could explain it before accepting the causal story. "Countries with higher education spending have higher GDP per capita" is compatible with education driving growth, with rich countries simply being able to afford more education, or with both.

This connects directly to ceteris paribus. In a model, other things are held constant by assumption; in data; they are not held constant at all, which is precisely why moving from a correlation in the data to a conclusion about cause requires care.

Positive and normative statements

Two points examiners look for:

• A positive statement can be wrong. "Raising interest rates causes inflation" is positive; it is testable, and also false. Positive means testable, not true.
• Most real economic debate mixes both. A policy argument usually combines a positive claim about consequences with a normative judgement about whether those consequences are desirable. Separating them is what allows an argument to be assessed properly.

Rational decision-making and the margin

The standard model assumes economic agents are rational: they act to maximise their own objective given their constraints.

• Consumers maximise utility, the satisfaction from consumption.
• Firms maximise profit.
• Governments are usually assumed to maximise social welfare.

Rationality here means consistent pursuit of an objective, not that people are calculating or selfish in an everyday sense.

Thinking at the margin

Economists analyse decisions at the margin, asking about the effect of one more unit rather than about totals.

A rational agent continues an activity while the marginal benefit exceeds the marginal cost, and stops where MB = MC.

This single rule underlies an enormous amount of the syllabus: a firm produces where MR = MC, a consumer buys until marginal utility per dollar is equal across goods, and the socially optimal output is where MSB = MSC. Recognising that these are the same idea in different settings makes the course substantially easier.

Marginal thinking also explains why sunk costs are irrelevant to a rational decision: money already spent cannot be recovered, so it should not affect the choice about the next unit.

Limitations of the rationality assumption

The assumption is a simplification, and the syllabus expects awareness of its limits, which is the bridge to behavioural economics (2.4):

• People have bounded rationality, limited information, limited time and limited computational capacity, so they use rules of thumb rather than optimising.
• People show bounded self-control, acting against their own long-term interests.
• People show bounded selfishness, caring about fairness and about others.
• Framing affects choices: the same option is chosen or rejected depending on how it is presented.

None of this means the model is useless. It means the model is a starting point whose predictions should be checked against how people actually behave.

Worked example

Consider the statement: "The government should cut income tax because it will increase employment."

Separating the two components:

• "It will increase employment" is a positive claim. It asserts a causal relationship that can be tested against evidence, and it may turn out to be false, or true only under certain conditions.
• "The government should cut income tax" is a normative claim. It rests on value judgements about the relative importance of employment, public services funded by that revenue, and the distribution of the tax cut.

Why the distinction matters practically. Someone might accept the positive claim entirely and still reject the recommendation, because they weigh the lost public spending more heavily. Disagreements that look factual are often really about values, and vice versa, and identifying which is which is the first step in evaluating any policy argument.

Common exam mistakes

• Thinking a positive statement must be true. It must be testable.
• Classifying a statement by its topic rather than its wording, look for "should", "ought", "fair".
• Criticising a model for being unrealistic without asking whether it is useful.
• Treating ceteris paribus as a claim that other things really are constant, rather than an analytical device.
• Interpreting "rational" as "selfish" or "always correct".
• Comparing totals when a question requires marginal analysis.

Exam technique

When asked to classify statements, quote the wording that settles it, "the word should makes this normative", rather than asserting the classification.

Where a question invites evaluation of a model or policy. Use ceteris paribus explicitly: state what the model holds constant, then ask what happens when that assumption fails in the real case.

Marginal analysis is a habit worth building early, because MB = MC recurs throughout the syllabus in different disguises.

Quick revision

• Economics is a social science: controlled experiments are rarely possible, so economists disagree.
• Models simplify deliberately; judge them by usefulness, not realism.
• Ceteris paribus holds other variables constant, the source of the movement/shift distinction.
• Positive = testable (and possibly false). Normative = value-based, containing "should" or "ought".
• Rational agents maximise: consumers utility, firms profit.
• Marginal thinking: continue while MB > MC; stop where MB = MC. Sunk costs are irrelevant.
• Rationality is bounded, the bridge to behavioural economics.

Not the topic you were looking for? Describe what you are stuck on in your own words and we will take you to the notes that answer it.