Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

For $a,b \in \mathbb R$, $p\geq2$ I try to show $$\left|\frac{a+b}{2}\right|^p+\left|\frac{a-b}{2}\right|^p\leq\frac{1}{2}|a|^p+\frac{1}{2}|b|^p.$$

Is this a popular inequality (At least I could not find it in the list of popular inequalities from wikipedia)? It seems to be related to convexity but I did not succeed to show it. A related inequality seems to be for $p \geq 1,a,b\geq0$

$$\left(\frac{a+b}{2}\right)^p\leq \frac{1}{2}a^p+\frac{1}{2}b^p,$$ which directly follows from the convexity of $x^p$ for positive numbers.

share|cite|improve this question
It's a special case of Clarkson's inequalities (when your measure space is a single point with counting measure). – Jose27 Dec 31 '11 at 5:08
up vote 21 down vote accepted

We have for all $x_1,x_2\geq 0$: $x_1^p+x_2^p\leqslant (x_1^2+x_2^2)^{p/2}$. Indeed, it suffice to show it when $x_2=1$, otherwise apply it to $\frac{x_1}{x_2}$. $f(t):=(t^2+1)^{p/2}-t^p-1$ is non-negative, since its derivative is $p(t^2+1)^{p/2-1}t-pt^{p-1}\geqslant 0$ and $f(0)=0$. Applying it to $x_1=\frac{a+b}2$ and $x_2=\frac{a-b}2$, we get \begin{align*} \left|\frac{a+b}2\right|^p+\left|\frac{a-b}2\right|^p&\leqslant \left(\left|\frac{a+b}2\right|^2+\left|\frac{a-b}2\right|^2\right)^{p/2}\\\ &=\left(\frac {2a^2+2b^2}4\right)^{p/2}\\\ &=\left(\frac {a^2}2+\frac{b^2}2\right)^{p/2}\\\ &\leqslant \frac 12|a|^p+\frac 12|b|^p, \end{align*} since the map $t\mapsto |t|^{p/2}$ is convex ($p\geqslant 2$).

share|cite|improve this answer
Why is it sufficient to show it for $x_2=1$? – Listing Dec 31 '11 at 0:10
It's clear if $x_2=0$. Otherwise, you apply it to $\frac{x_1}{x_2}$. – Davide Giraudo Dec 31 '11 at 0:12
Thank you, I got it now. – Listing Dec 31 '11 at 0:26
Does the last step even follow from the convexity of $t\mapsto |t|^{p/2}$? I don't see it :/ – Listing Dec 31 '11 at 14:42
Thank you, I did not see that one could remove the abs inside because of the squares. – Listing Dec 31 '11 at 15:07

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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