If I take a fragile drinking glass and forcefully throw it on the surface of a hard floor, then it wouldn't surprise me to learn that others would think I'm the cause of the subsequent broken glass. Had the glass not been fragile, and had the floor not been hard, then it's still the case I caused the glass to brake, as I threw it with sufficient force to break it given the conditions. I take full responsibility for being the cause behind why the glass broke. I admit, I caused it to break. We'll call that "cause of the first kind."
If a government entity or private organization changes a policy, we might find that things no longer happen as they might have before. There seems to be a tendency by many to attribute the change in policy as the cause for the real world change that subsequently occurs. For instance, (and I'm probably going to regret my impromptu attempt to give an example), an increase in the allowable speed limit in a given area might find future records of accidents conveying to us an increase in accidents and fatalities along that area. One might be inclined to blame the policy as being the cause. Okay, but if that's a cause, then it's certainly a cause of a different kind (and not a cause of the first kind).
The glass that broke wouldn't have broke had it been strong enough to withstand the force of my throw, but I'm still to blame regardless, for I threw the glass and the glass had no alternative but to react as it only could. The policy gave permission to drive faster along the roadway (and please, don't let this example rule over the point trying to be made), but the people could have continued to drive as they always have. This is very much unlike the glass. Yes, the people might have been led to think it was safe whereas it might not have been safe, but saying the change in policy caused the increase in traffic incidents cannot be the same kind of cause. It's a cause of a different kind, and I'll dub it a cause of the second kind.
I'm hesitant to regard it as an indirect cause, mainly because of the ambiguity potential. For instance, it could be a direct cause in the sense of it having a direct correlation yet an indirect cause as the change in policy itself has no immediate physical effect like the glass example. Other people have to respond (and through the choice not to respond, unlike the glass) before there are later physical accidents.
Any thoughts?