A regression of salary on education may combine the effect of schooling with characteristics related to both schooling and earnings. Leaving out a relevant variable does not always bias the slope: the omitted variable must affect the outcome and correlate with the included regressor.

Two models, two slopes

Suppose the data-generating mechanism is:

Yi=β0+β1X1i+β2X2i+εi,E[εi∣X1i,X2i]=0.Y_i=\beta_0+\beta_1X_{1i}+\beta_2X_{2i}+\varepsilon_i, \qquad E[\varepsilon_i\mid X_{1i},X_{2i}]=0.

X1X_1 represents education and X2X_2 an invented latent ability variable. If we fit only Yi=α0+α1X1i+viY_i=\alpha_0+\alpha_1X_{1i}+v_i, the population projection slope is:

α1=β1+β2Cov⁡(X1,X2)Var⁡(X1)⏟Omission bias.\alpha_1=\beta_1+\underbrace{\beta_2\frac{\operatorname{Cov}(X_1,X_2)}{\operatorname{Var}(X_1)}}_{\text{Omission bias}}.

For positive β2\beta_2 and positive correlation, the simple regression attributes part of the effect of X2X_2 to education, inflating the slope. Setting correlation to zero removes this term; negative correlation reverses its sign.

The slope changes when a variable is missing90 synthetic observations
Low ability (X₂)Medium ability (X₂)High ability (X₂)
Simple slope α̂₁
—
Controlled slope β̂₁
—
β̂₂
—
Cov(X₁, X₂)
—
Var(X₁)
—
Bias = α̂₁ − β̂₁
—

Simulation: Y = 10 + 2X₁ + 8X₂ + ε. Ability is an invented standardized variable, not an observed biological trait. All three controlled lines are conditional fits from one multiple regression.

What controlling means

The controlled regression estimates both coefficients simultaneously. The three lines are conditional predictions at three values of X2X_2: they share an education slope but have different intercepts. They are not three independent regressions.

The simulation fixes β1=2\beta_1=2 and β2=8\beta_2=8 for 90 observations. Marginal dispersions remain constant as correlation changes. Disturbances are constructed orthogonal to the regressors so the decomposition also holds exactly in this sample; real datasets add sampling variation. The panel computes the simple and multiple fits rather than merely drawing preset slopes.

Methodological conclusion. Controlling for a covariate does not automatically establish causality. Measurement, causal structure and remaining omitted variables all matter. Ability here is a teaching device, not an empirical measure of innate differences or a definitive account of wage inequality.