1
$\begingroup$

Find the returns to scale for he following production function- $(x_1 + 1)^.5 (x_2)^.5$

My attempt, Let $f(x_1,x_2)=(x_1 + 1)^.5 (x_2)^.5$ And $g(x_1,x_2)=(x_1)^.5 (x_2)^.5$ Now, $(x_1 + 1)^.5 (x_2)^.5$ $>$ $(x1)^.5 (x_2)^.5$ Therefore, $f(x_1,x_2)>g(x_1,x_2)$ Or, $f(tx_1,tx_2)>g(tx_1,tx_2)$ for some t>0 $>(tx_1)^.5 (tx_2)^.5$ $=t(x_1)^.5 (x_2)^.5$ $=tg(x_1,x_2)$ The function g exhibits constant returns to scale. The function f being bigger than it should exhibit increasing returns to scale. The answer given, however, is decreasing returns to scale. Where am I wrong? Thank you in advance.

$\endgroup$
3
$\begingroup$

What you have proven is that $$f(tx_1, tx_2) > tg(x_1,x_2)$$

which translates "scaled $f$ is higher than some other, homogeneous function $g$ scaled by the same factor". This does not prove anything about the returns to scale related to $f$, although I can see why it may appear otherwise.

You can go in reverse, starting from

$$tf(x_1,x_2) = t^{0.5}t^{0.5}f(x_1,x_2)=...$$

and compare to $f(tx_1, tx_2)$.

Also beware about the following: what are the permissible values for $t$ when returns are not constant? Are you sure we should obtain the result for $t>0$, or maybe we examine the property for a smaller interval?

$\endgroup$
  • $\begingroup$ For IRS and DRS the constant t should be greater than 1, got it, thank you! However, since I've been asked to find out the returns to scale and thus when I still don't know whether it is CRS or otherwise, in that case, what should be my assumption about the constant t be? $\endgroup$ – Chd Jul 5 '18 at 13:18
  • $\begingroup$ @Clarissa Is it possible for this function to exhibit CRS? $\endgroup$ – Alecos Papadopoulos Jul 5 '18 at 13:25
  • $\begingroup$ Err no, I got your point, thanks for replying @Alecos Paradopoulos $\endgroup$ – Chd Jul 5 '18 at 13:49

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.