I know that the Constant elasticity of substitution production function is given by:

$$Q=\gamma\left[\delta L^{-\alpha}+\left(1-\delta\right)K^{-\alpha}\right]^{\frac{-1}{\alpha}}$$

where $\gamma$ is the efficiency parameter, $\delta$ is the distribution parameter and $\alpha$ is the substitution parameter.

I can also show that the $$CES = \frac{1}{1+\alpha}$$.

But does all this mean that the "production function represents the average of two inputs $L$ and $K$ for different values of $\alpha$, given that $\gamma>0$ and $0<\delta <1$"?

I could not imagine this to be true graphically since the shape of the Isoquant depends on the parameter $\alpha$. For instance, when $\alpha=1$ it takes convex shape, $\alpha=-1$ it takes downward sloping straight line shape, and when $\alpha=\infty$ it takes L shape.

Edit 1: The source of the question is the Indian Economic Services General Economics Paper 1, 2019. This is an exam conducted by the Government of India to hire economists.

  • 1
    $\begingroup$ This production function is decreasing in $L$ and $K$ for all $\alpha\neq 0$. It does not look like a sensible production function. $\endgroup$
    – Amit
    Commented May 27 at 11:01
  • $\begingroup$ @Amit Sir Thank you for the response. There was a typo. I have corrected it. Can you provide your insights about it now? $\endgroup$ Commented May 27 at 11:30
  • 2
    $\begingroup$ What do you mean by average of two inputs $L$ and $K$ for different values of $\alpha$? $\endgroup$
    – Amit
    Commented May 27 at 14:25
  • $\begingroup$ I think here it means the generalized weighted mean taking specific forms as I also mentioned in the question. On considering your comment I found this. $\endgroup$ Commented May 27 at 15:11

1 Answer 1


Yes, all this does mean that the production function represents the average of two inputs $L$ and $K$ for different values of $\alpha$, given that $0<\gamma<1$. The key thing to consider here is the broader definition of a mean. The initial equation itself is a weighted generalized mean. And later depending on the value of $\alpha$ it gets transformed to other forms of mean like weighted mean, and geometric mean.

It is has been shown in detail in a previous answer on Stack Exchange here.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.