public policy variables are outside the market efficiency hypotheisis because. markets are therfore not efficient wrt to public policy in general. they only are efficient with repsect to a specificially dilineated (analytically tractavle) public policy context at a single points of time.

And specifically what externality do you think is at play here?

Market effeciency concepts have to do with incorporating information into a (well defined) market. Breaking the logic sequence of "well defined market", which can be done with trivial levels of "real world complexity", is problematic for EMH. This means you cant use EMH as a sort of backward logic to defend the status quo.

As to the concept of a market externality, I'm not sure I understand the question. If you just mean "uncaptured incentives", well surely there are myriads. But the bigger point is that this is not a simple solution. You cannot re-solve the maths as per a case of omitted variable bias.

The reason that this is so may be intuively grasped with an example. Let us posit that rent-seeking is a behavioral consequence of 'opportunistic' principals. And that rent seeking public policy architects are inevitable, because our politician is opportunistic (qua politician--behavioral assumption).

OK, but now I've just made mess for EMH, and this is why EMH has an 'H' in its name. Because general economic theory requires that we assume away several things we just introduced. And if you understand EMH you understand why having opportunistic, strongly authorized, unrestrained public policy makers is a problem. Namely, (1) they hold private information (outside emh); and (2) they are free to play zero-sum dynamic games with market participants; and (3) they can charge economic rents by implementing a 'protection racket' scheme whereby donor/lobbyists avoid (large) zero-sum losses by taking (smaller) zero-sum losses that benefit the policy holder (so called diversions).