Sampling Bias – The Mirror That Lies
Imagine you want to know the average height of students in your school. You stand outside the basketball court and measure everyone who walks out after practice. Your result: the average height is 6 feet. But is that really the average for the whole school? No — you only sampled tall students who play basketball. Your sample is biased toward tall people.
That is sampling bias in a nutshell: the error that occurs when the group you actually study is not representative of the group you intend to study.
The Precise Meaning
In economics, we often cannot survey or measure an entire population — every consumer, every firm, every household. So we take a sample, a smaller subset, and use it to draw conclusions about the whole population. Sampling bias creeps in when the method of selecting that sample systematically favours certain types of individuals over others.
Sampling bias is not about having a small sample. A small but randomly chosen sample can be fine. A large but biased sample is useless — it gives you a precise answer to the wrong question.
Why It Matters in Economics
Economics is full of surveys and data collection: unemployment rates, consumer spending, inflation, poverty estimates. If the sample is biased, every conclusion built on it is suspect.
Consider the Consumer Price Index (CPI). To measure inflation, the government surveys households about their spending patterns. If the survey only reaches urban households (because rural areas are harder to reach), the CPI will reflect urban spending habits — and miss the fact that rural households spend a much larger share of their income on food. The inflation rate you get is biased.
Sampling bias can lead to wrong policy decisions. If the government thinks inflation is low (because the biased CPI says so), it may not act to protect the poor — who are actually facing high food-price inflation.
A Concrete Example: The Employment Survey
India's unemployment data is collected through periodic surveys like the Periodic Labour Force Survey (PLFS). The survey selects households from a list of census villages and urban blocks. Now suppose the survey frame (the list from which you pick) is outdated — it misses newly built slums or migrant labour camps. The sample will underrepresent casual labourers and overrepresent stable households. The unemployment rate you calculate will be lower than the true rate, because you missed the people most likely to be jobless. …