In the following post, please consider reading films in the same catagory as procedures. Film reading is not a procedure. However, it's concrete like a procedure and it's a big revenue generator like procedures typically are.
If generating enough revenue for the department is an important goal for procedural specialties, can't a department just hire more clinicians to staff the clinics and then have a couple more MD/PhD on the side to do majority research? It'd be a win-win situation for all.
It doesn't work that way. It's very hard to recruit clinical-only or mostly clinical physicians to academia in procedural departments. Thus, there is a chronic staffing shortage among procedural academic departments. In Radiology, the shortage is so severe that many academic hospitals have to outsource their Radiology to private practice groups. The section chief positions within departments go unfilled for many years. Many residency programs are "weak" in a subspecialty of Radiology because there's just one or none left in the department and so the films are being sent to a private practice group. The ABR has gone so far as to start forcing all residents to do fellowship to get boarded, so that there's still someone left in academics. Similarly, they allow foreign trained Radiologists to become board certified after 3 years training in the states. Again, this all helps to fill very sparse academic departments for the couple of years while they are "training" and also helps flood the market so the private practice positions might fill up and make private practice more competitive/less lucrative.
There are some incentives to academia. Such as that there is typically less responsibility because you can make residents and fellows do your scut and call. Thus mostly clinical academics can be more lifestyle and intellectually friendly even if the pay is less. The caseload tends to be less and the studies are often more interesting than a stack of normals. It's one place that tends to tolerate you almost entirely reading whatever films you were fellowship trained in, while private practice is much more general. That "lifestyle" goes out the window if you're competing for grants with serious research, however, since you're still expected to carry clinical load and run a full-time lab and obtain serious funding. These are all parts of the reasons why only about 50% of academic MD/PhDs are doing 50% or more research according to the survey.
But private practice still earns on the order of double or more. You also have the satisfaction and independence of owning your own business, or being a partner in a practice. Of course this business side doesn't appeal to everyone, but the heirarchy, competitiveness, and often antagonistic behavior of academic hospitals doesn't fly in private practice. My point is there are positives and negatives to both sides. You have no idea about this as a pre-med student, and I still think it's ridiculous to expect a 22 year old fresh out of undergrad to have any sense of what the real world is like. Even if they did, saying you will have a research career 15 years from now is like saying you'll be a millionaire 15 years from now. It's a nice goal, but you may not get there for a variety of reasons.
I think it's silly to make it sound like private practice is automatically more work than academics in procedure heavy fields. Often it is, but it's because the attendings are raking in lots of money and directly see the fruit of their increased labor. In academics, it is often not this way thanks to a fixed salary. Though many departments are rewarding based on numbers of procedures done. This further removes the incentives for anyone to do research within academics and is one of many ways the pay gap is widened between researchers and clinicians.
Even so, I know a few private practice guys who left academics. They upgraded their vacation from 4 weeks to 4 months a year and tripled their salaries in the process. They would say they work about the same amount when they're on service as they were in academics. So, I guess Radiology is just a "bad specialty for research"? No. The research seems to be as good as anywhere else, maybe even better if you have the right technical skills and also since the field is so empty. It's just a great specialty for clinical work. If you want to follow the people and the talent, you just have to do what I said earlier, and follow the $$$$.