What statistics does, and its limits

Statistics helps an economist in several ways:

• It makes facts exact. 'The train was 47 minutes late' tells you far more than 'the train was very late'. Exact figures are more convincing than vague words.
• It condenses a large mass of data into a few summary measures. You could never remember the incomes of a million families, but you can remember one average income.
• It finds and tests relationships between economic factors: does demand fall when price rises? Does consumption rise with income? Does more government spending push up prices?
• It helps predict. From past data, economists forecast future consumption, the future demand for oil, or how today's investment will affect income in later years. If demand for a product has risen every festival season for ten years, a firm can plan its supplies.
• It helps make and test policies. Should the government spend more on family planning? How much oil should India import? Statistics provides the evidence.

But statistics has limits. It is a tool, and it can mislead if used without thought. It is no substitute for common sense. The book's river story makes the point: a decision based on an average alone can go badly wrong, because an average hides the variation within the data.

So statistics makes facts exact and gives policies a basis, but good judgement is still needed to use it well.