Cobb-Douglas Production Function

The Cobb-Douglas production function, introduced by Charles Cobb and Paul Douglas in 1928, models how capital and labor combine to produce output. It remains one of the most widely used specifications in macroeconomics and microeconomics.

The function:

Y = A x K^alpha x L^(1-alpha)

where Y is output, A is total factor productivity (TFP), K is capital input, L is labor input, and alpha (typically 0.33 for developed economies) is capital’s share of income.

Marginal products:

MPK = alpha x Y / K (marginal product of capital) MPL = (1-alpha) x Y / L (marginal product of labor)

Both marginal products are positive and diminishing. Doubling capital while holding labor fixed less than doubles output.

Returns to scale. The exponents alpha and (1-alpha) sum to exactly 1, which means constant returns to scale: doubling both K and L exactly doubles output. If the exponents sum to more than 1, there are increasing returns; if less, decreasing returns.

Income shares. A key prediction: capital earns fraction alpha of total output (Y x alpha goes to capital owners) and labor earns fraction (1-alpha). That lines up well with observed income distribution data, since capital’s share in the US has been roughly 30-35% historically.

TFP (A) captures everything not explained by capital and labor: technology, institutions, education quality, management practices. Much of the difference in output between rich and poor countries is attributed to TFP differences, not just capital or labor.

The general two-parameter form. Textbooks often write Cobb-Douglas as Y = A · K^α · L^β with α and β as independent exponents. When α + β = 1 that collapses to the constant-returns version above. When α + β > 1 the function shows increasing returns to scale, meaning doubling both inputs more than doubles output, which is typical of industries with strong network effects or scale economies. When α + β < 1 there are decreasing returns. Empirical estimates for the US private sector come out close to α + β ≈ 1, which is why the constant-returns version is the usual starting point.

The labor exponent box below is optional and is what switches between the two. Leave it blank and the calculator sets β = 1 - α, giving constant returns and the classic textbook answers. Fill it in and the calculator uses your two exponents as written, reports the sum, and says which regime you are in.

Watch what that does to the doubling lines in the result. Under constant returns, doubling both K and L always gives exactly 2Y, no matter what numbers you put in. That is not the calculator being clever, it is a property of exponents that sum to 1, and it is the single most-quoted fact about this function. Set α = 0.4 and β = 0.8 and the same doubling gives 2^1.2 = 2.30 times output instead. The scale factor is always 2^(α+β), which is worth checking by hand once.

Income shares also stop adding to 100% once α + β ≠ 1, and that is not a bug either. Under increasing returns, paying capital its marginal product and labor its marginal product costs more than the firm earns. A perfectly competitive firm with increasing returns to scale cannot break even, which is why models with scale economies have to abandon perfect competition. The calculator flags the gap rather than hiding it.