Haskell have the Maybe monad, which is defined as Just a | Nothing. One may argue then that languages with nullable integers implicitly define this type as Maybe Int.
Right! that definition was the type, thanks for the correction :) still, the Maybe monad is used for computations that may fail, which is often the purpose of the null pointer.
I meant that in some way one could think of, for instance, C's int as a sum type of just integer and null :) in practice the obvious difference is that a language like Haskell enforce considering both options with its type checker.