Supply and Demand

In the econometrics context, supply and demand refers not just to the microeconomic concept but specifically to the identification problem that arises when trying to estimate supply or demand elasticities from observational price-quantity data. Both supply and demand equations contain price as a regressor, but price is jointly determined by both curves — it is endogenous in each equation. Regressing quantity on price with OLS does not identify either curve; the scatter of equilibrium points is a mixture of movements along both curves.

The Graddy (1995) Fulton Fish Market study (Lecture 06) is the canonical example: to identify the demand elasticity, weather (which shifts only supply) is used as an instrument for price — stormy-day supply contractions generate price variation that is purely supply-driven, tracing out movement along the demand curve. Similarly, day-of-week dummies (shifting demand) identify the supply curve. This is the classical simultaneous equations identification strategy: a variable that shifts one curve serves as an instrument for the endogenous price in the other curve's equation. See Simultaneous Equations & Time Series (Lecture 06).

When to use

Reach for the simultaneous-equations framing whenever price and quantity are jointly determined. The diagnostic is that price appears on the right-hand side of both the supply and demand equations — OLS on either equation alone yields biased estimates. The Instrumental Variables solution requires finding variables (curve shifters) that are excluded from one equation but directly affect the endogenous variable in the other.