http://www.barefootrunning.fas.harvard.edu/4BiomechanicsofFo...
Dredged that out of my memory from a couple years ago.
Heel strike - Effective mass is approximately the foot plus the lower leg, which equals 6.8% of total body mass.
Forefoot strike - Effective mass is the forefoot and some portion of the rearfoot and leg, which equals 1.7% of total body mass.
"We have found that even on hard surfaces (a steel force plate) runners who forefoot strike have impact forces that are 7 times lower than shod runners who heel strike. Rates of loading are equal to or less than rates of loading for shod runners."
On the other hand, if you're a heel striker, the reduction in impact might be worth it. Stomping on concrete with your heels sends a lot of force through your ankles and knees.
* Running (with shoes) is not a predictor of osteoarthritis (see http://www.jaoa.org/content/106/6/342.full). Elite athletes (running vastly more than typical recreational runners) do exhibit more signs visible in X-Rays, but do not report symptoms at a higher rate than any other group.
* Reliable predictors of acute injury are "too much, too soon", history of past injury, and, oddly, old shoes. (See http://www.time-to-run.com/footwear/mechanisms.htm). Matching the type of shoe to type of gait (as running shoes stores often attempt to do) has never been shown to reduce injury rate.
* It is possible to run with shoes and without injury. (I did it for 10 years; my old coach for 30+).
* It is possible to run barefoot on concrete and asphalt without injury. (I've done it for 5 years; others have for longer).
Those last two anecdotes don't speak to long-term injury, but I'm not terribly worried for two reasons: The first is the lack of correlation between running in shoes and osteoarthritis. The second is that the foot-achilles-knee is a spring-like appendage with vastly more play than 1/2 inch of EVA foam. Anecdotally again, I feel no pounding when running barefoot on concrete.