Cobb-Douglas Production Function

Output as a constant-returns combination of capital and labor.

Developed by Charles Cobb & Paul DouglasOrigin 1928Intro
SO

Built and reviewed by Stephen Omukoko Okoth

Mathematical Economist · ex-Morgan Stanley FI · Equilar

Theory

What the model says, and why

Cobb and Douglas were trying to fit US manufacturing data and noticed that capital and labor each appeared to receive a roughly constant share of total output regardless of the level. The functional form that delivers that property is a power function of capital and labor with exponents that sum to one:

Y = A · K^α · L^(1−α)

A is total factor productivity — how much output you get from any combination of inputs, capturing everything from technology to institutions to organizational know-how. α is capital’s share of output. 1 − α is labor’s share. They sum to one by construction, which is what makes this constant-returns-to-scale: double both inputs and you exactly double output.

The marginal products are:

MPK = α · A · K^(α−1) · L^(1−α)
MPL = (1−α) · A · K^α · L^(−α)

Both are positive and diminishing — adding more capital (or more labor) raises output, but each additional unit adds less. This is the property that drives most of growth theory.

Empirically, α has tended to be ~0.3-0.4 across many economies and time periods. That isn’t a law of nature — it’s an observation that the model uses to motivate its functional form. Recent decades have seen capital share rise in many advanced economies, which is itself a real research topic.

Interactive playground

Move the parameters, watch the equilibrium move

Inputs

Parameters & factors

Output

Y = 100.00

Output (Y)

100.00

MPK

0.330

Marginal product of capital

MPL

0.670

Marginal product of labor

Capital share

33.0%

MPK · K / Y

Output rises with capital; the marginal product (slope) falls. Diminishing returns visible at a glance.

In the classroom

How to teach it well

Why this functional form, exactly? Three properties make it teachable. (1) Constant returns to scale — doubling K and L doubles Y, which makes per-capita analysis trivial. (2) Diminishing marginal products — both MPK and MPL fall as the corresponding factor rises. (3) Constant factor shares — α and 1−α are the income shares of capital and labor, regardless of the level.

Common misconceptions. Students often think A is a constant — emphasize that A is where institutions, technology, education, and culture live in the model. The reason rich countries are rich isn’t mostly that they have more K or L; it’s that their A is higher.

Pair this with Solow. Cobb-Douglas without growth dynamics is incomplete. Pair it with the Solow model to give students a complete picture of why countries converge to a steady state and what determines that level.