[matt_kantor]

Fans of static type checking use that as a rebuttal against "but I'm so much more productive in $dynamically_typed_language". I've also seen it in discussions about Scala, Haskell, TypeScript, etc.

Intuitively it makes a lot of sense, and it jives with my subjective feelings (especially for longer-lived projects), but I'm not sure if there have been any studies to back up that claim.

[root_axis]

From an organizational perspective, the benefits of static type checking are quite obvious. I'd go so far as to say it's a software engineering best practice to prefer languages with static types when building new software. The reason for this is because many, many, many engineering hours are saved at an organizational level by having a compiler that is able to enforce correct data structures throughout the application. Eschewing types for speed is just a form of technical debt that has to be repaid in perpetuity whenever it comes time to maintain or refactor the software, not to mention the fact that every new developer working on the application is now burdened with this debt.

[steveklabnik]

I am not aware of any studies that definitively show an advantage for static or dynamic typing; there have been ones that didn't provide anything conclusive, and you could argue that the experiment design wasn't representative.

https://danluu.com/empirical-pl/ is a good place to get started on this topic.