Suppose $L=\{w \in {a,b}^* \colon \#b(w) = \#a(w) \}$, the language of all strings with an equal number of occurrences of $a$ and $b$ in all possible arrangements. Furthermore, this language can be generated by the grammar $G$ with rules:
$$S → aSbS\text{ }|\text{ }bSaS\text{ }|\text{ }ε$$
I realize that the standard approach is to prove (i) $L(G)\subseteq L$ and (ii) $L \subseteq L(G)$ by induction. How would one prove $L \subseteq L(G)$ by induction given that the string $w$ can contain an equal number of $a$ and $b$ characters in any arrangement?
Hint:
I hope this helps $\ddot\smile$