Some of you keep mentioning a relationship of the axle height to the drawbar. Not important.
You've got force #1, the contact patch of the rear tires on the ground trying to push forward. And you've got force #2, the attachment at the drawbar trying to pull backwards. Since those two forces are not colinear, they result in a torque. Since force #2 is above force #1, that torque acts to rotate the tractor up and over backwards.
The only things that will prevent a backflip are the lack of either enough power or enough traction to complete the job.
Force 2 acts at the axle. At the ground point, you have the tire lever arm pushing backwards.
No force #2 does not act at the axle. It acts where it happens. In this case it acts where the chain is attached to the draw bar.
Let's get rid of the whole tractor and just imagine there is a rigid steel triangle running from the drawbar attach point down to the contact patch of the two tires. Three points; drawbar hole, right contact patch, left contact patch. Rigid steel triangle made of angle iron or something. That's what's being acted upon by the two forces. The top point is leaning against a flag pole and loosely chained to it. You grab the lower two points and move them further away from the pole. What happens? As the lower two points are pulled further from the pole, the triangle must incline further back against the pole and the upper chained point must slide a bit down the pole.
Now it so happens that as that triangle gets inclined further back, so does everything else that happens to be rigidly attached to it. Which basically means the whole rest of the tractor, and you too if you're sitting in the seat.
So sure, it has an effect at the axle, just as it has an effect everywhere else on the tractor. That effect can be calculated/predicted at the axle just like it can be for any other point. And the result of any such calculation, for any point on the tractor, is that the forces will tend to make any chosen point rotate backwards. But none of that means the force is acting AT the axle. I mean there ARE forces going on at the axle of course, but not either of the two simple vectors needed to deal with the question asked by this thread.
xtn