I'm trying to understand intuitively what a homothetic preference is, and I am still not quite there. I understand the definition, that a homothetic preference implies that the slope of the indifference curve remains constant along any ray from origin.

But intuitively, like in real life, how would we know that a preference is not homothetic? If its not strictly monotonically increasing (i.e. you don't prefer more of a product to less of it always), does it mean that the preference is not homothetic? Can someone give an example of what exactly is not a homothetic preference so I understand what is?


1 Answer 1


Are homothetic preferences strictly monotonically increasing?

Homotheticity requires that

$$ \alpha^\gamma U(x,y) = U(\alpha x, \alpha y) $$

This is not defined over the "increasing" part of strictly monotonously increasing. Indeed, you can have decreasing preferences $U(x,y) = -x -y$ for whom homotheticity holds.

Are nomothetic preferences weakly monotonic?

However, the "strict" part (in both jointly) is required, i.e. $U(\alpha x, \alpha y) \neq U(x,y)$ for $\alpha \neq 0$.. You cannot have homothetic and weakly increasing or weakly decreasing preferences.

Proof outline To prove that, we assume otherwise. There is some homothetic $U$ with a strictly increasing part and a flat part. You will see that the $\gamma$ corresponding to the increasing part can potentially be any number (it depends on the functional form). However, the $\gamma$ corresponding to the flat part needs to be 0.[1] Hence, there cannot be a $\gamma$ that holds for the utility function as a whole, for any general increasing part.

[1]: Unless of course, $U(x,y) = 0$ at the flat part. However, as utility functions are immune to monotone transformations, we can ignore such cases.

  • $\begingroup$ By weakly monotonic do you mean monotonic but not strictly monotonic? If yes, what about the counterexample in the answer below? $\endgroup$
    – Giskard
    Commented May 24, 2015 at 12:34
  • 1
    $\begingroup$ @denesp: Weakly increasing in both jointly. With Leontief preferences you have it that your utility always increases when you increase both $x$ and $y$. $\endgroup$
    – FooBar
    Commented May 24, 2015 at 12:41

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.