I am trying to better understand the CES function:

$$Y_{t}=C\left[\pi\left(A_{t}^{K} K_{t}\right)^{\frac{\sigma-1}{\sigma}}+(1-\pi)\left(A_{t}^{L} L_{t}\right)^{\frac{\sigma-1}{\sigma}}\right]^{\frac{\sigma}{\sigma-1}}$$

I understand the special cases where $\sigma=1$, where $\sigma=0$ and where $\sigma\rightarrow0$.

I am, for some reason, having a really hard time developing intuition for other cases. Don't the exponents cancel for all other values of $\sigma$ and the expression reduces to:

$$Y_{t}=C(\pi(A_{t}^{K} K_{t}) +(1-\pi)(A_{t}^{L} L_{t}))$$

This, of course, can't be right as $Y_t$ should vary for different values of $\sigma$, no?


The exponents do not cancel each other out. E.g., $$ (a^2+b^2)^{1/2} \neq a + b $$ because $$ (a^2+b^2)^{1/2} \neq (a^2 + 2ab + b^2)^{1/2}. $$

| improve this answer | |
  • $\begingroup$ Ah, yes, I feel embarrassed about the silliness of my question. May I ask a simple follow-up? In the case where $$A_{t}^K K_{t} = A_{t}^L L_{t}$$ should we then be indifferent to the selection of values of $\pi$? $\endgroup$ – David Apr 26 at 10:41
  • 1
    $\begingroup$ @David Your previous question was harder. $\endgroup$ – Giskard Apr 26 at 13:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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