The Term 'Marginal' Has Opposite Meanings in Economics and Statistics

In statistics, a marginal distribution refers to the probability distribution of a variable irrespective of other variables. Marginal probability is obtained by adding (discrete case) or integrating (continuous case) the joint distribution over other variables.

In economics, ‘marginal’ refers to the incremental change in a variable corresponding to a unit change in another variable. The calculation involves taking the derivative of the relevant function. Examples: marginal cost (change in total cost when one more unit of a good is produced), marginal product (change in the total production for one additional unit of input), marginal utility, marginal tax rate, etc.

Thus, ‘marginal’ refers to opposite ideas in statistics and economics. For the statistical context, think of how totals in a contingency table are written on the margins. For the economics context, think of how the aggregate value of a variable changes ‘at the margin’ for a unit change in another variable.