Economics as a Social Science
Definition
Social science: the systematic study of human behaviour and society using theory and evidence; economics qualifies because it explains how people and societies allocate scarce resources.
Scarce resources: the factors of production (land, labour, capital and enterprise) that are finite relative to unlimited human wants.
- Like the natural sciences, economics builds and tests theories, but its subject is human behaviour, which responds to expectations and is far less predictable than physical matter.
- A whole economy cannot be placed in a laboratory, so economists cannot run controlled experiments; they rely instead on simplified models and observational data.
- This is why its findings are tendencies that hold on average rather than exact laws, which is exactly what makes it a social rather than a natural science.
Building Economic Models
Definition
Model: a simplified representation of reality that captures the key relationships between a few variables while stripping out detail.
Assumption: a simplifying condition, such as rational behaviour, taken as given to make a problem manageable.
Ceteris paribus: Latin for other things equal; the device of holding all other influences constant so the effect of one variable can be isolated.
- Economists begin from assumptions because reality has too many moving parts to analyse at once.
- A good assumption removes noise without distorting the core relationship being studied.
- Ceteris paribus then isolates cause and effect: change one variable, hold the rest constant, and any change in the outcome can be attributed to that variable.
- The model is finally tested by comparing its predictions against real-world data; if the data repeatedly contradict it, it is refined or rejected.
Example
- A fall in the price of Coca-Cola raises the quantity demanded of Coca-Cola, ceteris paribus: income, tastes and the price of a close substitute such as Pepsi are all held constant.
- Coca-Cola and Pepsi are substitutes (XED > 0), so Pepsi's price must be held constant to attribute any change in quantity to Coca-Cola's own price alone.
- Without that assumption, too many things would change at once to reveal any single relationship.
Uses of Models
- A model does not need to be realistic in every detail to make accurate predictions.
- Isolating one variable reveals cause and effect that would otherwise be buried in the data.
- Models give economists a shared framework for debating policy precisely.
Analogy
- An economic model is like a map of the London Underground.
- It leaves out real distances and street layouts, yet it is far more useful for planning a journey than a photograph of the city would be.
Can we trust economic models?
- It holds because a model only needs to predict well, not to be realistic in every detail; the demand model works despite ignoring most of what drives buyers.
- But models built on rational behaviour can mispredict, because real people are swayed by habit, emotion and bias, the insight behind behavioural economics (Theme 3).
- But because controlled experiments are rarely possible, competing theories are hard to falsify, so weak models can survive longer than they should.
- On balance, a model's value depends on whether its assumptions fit the context: used carefully it is the most powerful tool economists have, but used blindly it misleads.
Exam technique
- Say ceteris paribus when you isolate one variable, to show you know other influences are held constant.
- Treat models as tools, and be ready to question their assumptions in evaluation.
- Support claims with real-world evidence where you can, since economics is empirical.
Common Mistake
- Do not dismiss a model simply because it is unrealistic.
- A simplification can still predict well, which is what matters.
- Do not treat ceteris paribus as a claim that nothing else ever changes.
- It is an analytical assumption used to isolate one effect, not a description of the real world.
Self review
- Why is economics classed as a social science?
- What is a model, and why do economists use them?
- What does the ceteris paribus assumption mean?
- Give one reason economics cannot easily run controlled experiments.
- Why can a model be useful even if its assumptions are unrealistic?
