Let $G$ be a $g$-dimensional Lie group and $H$ be an $h$-dimensional normal Lie subgroup of $G$. Is the dimension of the quotient group $G / H$ always $g-h$? This seems intuitively true, but I can't find such a statement anywhere.
For finite groups, we have that the order of $G/H$ is the quotient of the order of $G$ and the order of $H$, which suggests that there should be a simple relationship in the Lie case as well - but of course a nontrivial Lie group doesn't have finite order, so the result doesn't directly carry over.
The subgroup $H$ needs to be closed to ensure that $G/H$ is a Lie group. For example, let $G = S^1\times S^1$ and $H = \{(e^{2\pi i\theta}, e^{2\pi i\alpha\theta}) \in S^1\times S^1 \mid \theta \in \mathbb{R}\}$ where $\alpha$ is a fixed irrational number. While $H$ is a Lie subgroup of $G$, the quotient $G/H$ is not Hausdorff (the cosets of $H$ are dense in $G$ so the quotient topology on $G/H$ is the trivial topology) and therefore cannot be a Lie group.
It is worth noting that $G/H$ is a smooth manifold of dimension $g - h$ even if $H$ is not normal (as long as it is closed); of course, $G/H$ is not a Lie group in this case. The following is from Lee's Introduction to Smooth Manifolds (second edition).