Duncan Weldon asks whatever happened to deficit bias, and why was austerity so popular when it seems to hurt two crucial groups of electors. They are both excellent questions.
First, a quick recap of Duncan’s arguments. The standard view among economists before the financial crisis was that economies were prone to deficit bias: a tendency for government budget deficits and debt (as a share of GDP) to slowly increase.  Although there are a number of theories about why deficit bias occurs, several involve voters being at best unconcerned about it, and at worst conniving in it. Such theories are miles away from an electorate giving strong support to a government campaigning on reducing the deficit.
Second, austerity keeps interest rates low. So not only does austerity hit wages and profits, but it also hits those who rely on interest income to supplement their pension, a group who in case you need reminding have a high propensity to vote. Although asset values have gone up as well, many in this group will be reluctant or unable to turn this into income.
I try and answer both questions in this paper. As with many things, I think the answer lies in the financial crisis. Popular concern about government deficits will be much greater if these deficits are at 'record levels', which they inevitably were following the deepest global recession since WWII. A recession initiated by a financial crisis is also likely to see consumers reducing their own borrowing, and so (erroneous) analogies between governments and households resonate. A recession initiated by a financial crisis also makes the public receptive to the potential power of these markets, and therefore to claims by those ‘close to the markets’ that national disaster is just one more large deficit away. Arguments from economists that rational markets would not be concerned about government default when the central bank can create money are met with a widespread belief that the recent crisis shows markets are not necessarily rational. Markets become like a powerful god who can only be appeased by the sacrifices prescribed by its priests.
This is how, in the case of the UK, George Osborne was able to redefine the goal of macroeconomic policy from the normal desire to see higher living standards into the need to reduce the deficit. His motives for doing so may have involved a desire to reduce the size of the state, what I call deficit deceit, but uppermost in his mind was that his strategy was popular. So popular, both among the electorate and the media, that the Labour opposition eventually gave up on arguing that there was an alternative.
A key corollary of all this is that the popular appeal of austerity has a sell by date. Once people have stopped paying off their own excess debt, market panic becomes a more distant memory, and the need to control the government’s deficit seems less compelling. The underlying factors that created deficit bias can resurface. Quite when that sell by date will be depends on particular national circumstances. In the UK the game was up when the country voted for Brexit, but I suspect even without that austerity would not have won the Conservatives two elections. So the popularity of austerity is short term.
As to the political economy question, I think that be a genuine puzzle if austerity was a long term phenomenon. But as it is not we need to bring in (hopefully) short term failures of knowledge and information. I have not noticed campaigns to ‘save our savers’ also arguing for less austerity, and I suspect the reason is because they just do not see the connection. Who makes that connection for them? Here is Chris Dillow complaining that the media just does not do this kind of thing. Central banks should, but for their own reasons rarely do. People may not act in their own self-interest if they do not know what is in their own interest, and in the short term at least this knowledge may not be made available to them.
 To preempt the usual MMT comments, this bias relates to economies where monetary policy takes care of output and inflation stabilisation.