Show that, for $a,b,c \in \mathbb{R}$ and $q \geq 2$: $$ |a+b|^{q} + |a-b|^{q} + |a+c|^q + |a-c|^q + |b+c|^q + |b-c|^q \\\leq \\|a+b+c|^q + |a+b-c|^q + |a-b+c|^q + |a-b-c|^q $$
When $c=0$, this follows from the result here.
Is some generalization of Clarkson's inequality in the real case for more than 2 variables known?
Edit: Numerical experiments indicate that $\frac{lhs}{rhs}$ is maximized when one of $a, b, c$ is non-zero and the other two are zero. Many inequalities involving convex functions (for example: for $p\geq 1$, $(x_1^2 + x_2^2 + ... +x_n^2)^p \geq ( (x_1^2)^{p} + (x_2^2)^{p} + ... +(x_n^2)^{p})$ satisfy this property (the inequality being tight when one variable is non-zero and rest all is zero). This gives the impression that the inequality in question follows from some convexity arguments.
Why does Fact 2 hold for all $q\geq2$? By taking derivative, I see that there exists a $q_0$ (depending on $a,b,c$) such that for all $q\geq q_0$, Fact 2 holds.
– Mathews Boban Jun 12 '22 at 12:27