I would say there are two primary benefits:
- Encapsulation: code in libraries that shouldn't be called from outside the library can be hidden, preventing misuse, and easing changes to the internal structure of the library while preserving the external interface. In a good OO design, changes are introduced more easily once the code is complete.
- Abstraction: instead of dealing with arrays of arrays you're dealing with employees, departments, and so on. This means you can focus on the business logic, and write fewer lines of code. Fewer lines means fewer bugs.
Reuse I wouldn't strictly qualify as an OO benefit, because in a purely procedural model smart library organization lets you reuse code as well.
On the other hand, using lots of objects in PHP tends to decrease performance compared to a procedural model, because there is too much object construction overhead for every request. Finding a good balance between procedural-style and oo-style code is imperative.
