The trick is not to induct on $m$, just do the induction on $n$. You don't **have** to induct for both variables. Let $m\ge 0$ be arbitrary instead, then the base case $n=0$ is verified since $1+0=(1+m)^0$. Next assume we have proven this for some $n\ge 0$. Then
$$1+mn\le (1+m)^n$$
Multiply both sides by $(1+m)$ to get
$$1+mn+m+m^2n\le (1+m)^{n+1}$$
now we see that $m(n+1)=mn +m=m(n+1)\le m(n+1)+m^2n$ so that we can subtract off $m^2n$ and get
$$1+m(n+1)\le (1+mn)(1+m)\le (1+m)^{n+1}.$$