I think this [1] is one of the better references I've seen on this question. It's an analysis of bug-counts across Github projects. There's a correlation between number of lines of code and bugs. And there's a correlation between languages that favor simplicity and bugs. But there's not a a correlation between static typing and bugs.
So, I think we can say that you don't choose static typing in order to write fewer bugs.
I've worked in lots of code-bases, both static and non. Indeed, the bug rates are fairly consistent across them. The worst code bases I've seen (from a maintainability perspective) were actually old C++ and Visual Basic ones.
Refactoring is easier in statically typed code-bases. But, in my experience, it's not significantly better. What makes a bigger difference is immutability, functional purity (within reason) and a preference for obviousness (as opposed to, say Ruby-like magic which makes me pull my hair out on a semi-daily basis).
So, I think we can say that you don't choose static typing in order to write fewer bugs.
I've worked in lots of code-bases, both static and non. Indeed, the bug rates are fairly consistent across them. The worst code bases I've seen (from a maintainability perspective) were actually old C++ and Visual Basic ones.
Refactoring is easier in statically typed code-bases. But, in my experience, it's not significantly better. What makes a bigger difference is immutability, functional purity (within reason) and a preference for obviousness (as opposed to, say Ruby-like magic which makes me pull my hair out on a semi-daily basis).
[1] https://dev.to/danlebrero/the-broken-promise-of-static-typin...