Something that I think is worth noting about this is that, of course, Thatcher was operating in a system that put very few procedural constraints on the Tory majority in parliament. On the one hand, that allowed her to implement dramatic changes in U.K. public policy. But on the other hand, it meant that there was no tactical advantage to be gained by adopting public negotiating positions at odds with her real policy agenda. A British politician who believes that reducing high-end income taxes and replacing the lost revenue with regression consumption taxes is a good idea has no good options other than stating that this is the case and then doing it. An American politician with identical beliefs might nonetheless believe that the best strategy is to profess opposition to all forms of revenue and profess willingness to destroy the global economy in fanatical pursuit of that goal and then only very reluctantly accept the Thatcherite “compromise” once rival politicians are willing to put it on the table.His example, however, I think undermines his point. He notes that last time the GOP had control of both Congress and the presidency, instead of eliminating Medicare they actually added to it, and substantially. Fair enough! But it's awful hard to look at the first six Bush years as an example of a party acting responsibly in office because everyone knew they would be accountable. The Bush/DeLay Republicans cut high-end tax rates, but did they replace the lost revenue with regressive consumption taxes? Hardly. Did they pay for the expansion of Medicare? Certainly not.
Backing up a bit...the problem with Yglesias's thesis is that it depends on an electorate that has strong opinions on policy choices and would punish politicians that would deviate from them. If such an electorate existed, then he'd be correct. But it doesn't. Ideological voters are going to stick with their party in general elections regardless, while swing voters are going to react (on the whole) to economic and other big-picture results, not single policies.
Now, at the extremes, sure, Yglesias has a point. If President Bachmann, Speaker Paul, and Majority Leader Paul in 2013 immediately eliminate Medicare and Social Security, yes, that would almost certainly drive their approval ratings down to single digits. But note that even in the current situation, with no clear responsibility, none of them are calling for anything that dramatic. And at the level they're actually operating at, Republicans can afford to call for the things they're calling for because they just aren't voting issues for very many people.
It is true that House Republicans can risk crashing the economy through a debt limit crisis, or by fighting for an economy-crippling austerity program, secure in the knowledge that Obama would probably pay the price is the economy tanks. But I think the evidence is strong that what's driving Republicans on these policies is that they either truly believe in them (and don't forget, the Conservative Party in Britain is pursuing austerity), or that they are frightened of primary voters and organized groups within the party who really believe in them. In other words, I strongly suspect that President Bachmann, Speaker Ryan and Majority Leader Paul might well be implementing the same policies they're advocating today. And Speaker Boehner believes that he has to advocate the policies that Bachmann and the others advocate, because he believes (and rightly so) that if not, he'll be bounced.
If the GOP really has gone off the rails -- and I think it has -- the place to look for explanation is the rewards and incentives within the party, not the US Constitutional system.