Although I'm sure that this isn't what he meant to do Danny Blanchflower has just come out with some proof that the world, or at least the UK part of it, is a lot less Keynesian than it used to be. The proof is in this paper:
There have been unprecedented
since the start of the recession
falls in real wages in the UK
triggered by the financial crisis
of 2008. This did not happen in previous
economic downturns: median real wage
growth slowed down or stalled, but it
did not fall. Indeed, in past recessions,
almost all workers in both the lowest and
highest deciles of the wage distribution
experienced growing real wages. It was
the unemployed who experienced almost
all the pain: they lost their jobs and much
of their incomes, and many were
unemployed for a long time.
But in the Great Recession and its
aftermath, the economic hurt has been
spread more evenly, with wages taking the
strain this time. The real wages of the
typical (median) worker have fallen by
around 8-10% – or around 2% a year
behind inflation – since 2008. Such falls
have occurred across the wage
distribution, generating falls in living
standards for most people, with the
exception of those at the very top.
Recall what it is that Keynes said about wages: that they are sticky moving downwards. Without this property we'd be in something much more like a New Classical or Real Business Cycle world (or, if you prefer, something akin to the old classical models). If there was some exogenous shock to the economy that required wages to change in order to keep employment levels constant then that adjustment would happen. Keynes, and New Keynesianism uses this as a central part of the construct, pointed out that people really, really, don't like declining nominal wages. They'll put up with declining real wages if nominal wages stay flat because of the money illusion. And it is this that makes adjustment to exogenous shocks so difficult, so protracted, that we must therefore send half of all we produce to those wise people in government so they can spend it for us have fiscal stimulus to get us out of such recessions.
Of course, this is all being a little extreme. Taken to the logical conclusion the NC or RBC view says that recessions and their attendant unemployment just aren't possible. And the Keynesian view is that without that interference we'd never recover at all. Both of those are very much extreme descriptions. A rather better way of thinking of it is that each view contains some of the truth and none of them contain all of it: that's certainly my view of pretty much all macroeconomic theories anyway.
Now look at what Blanchflower is saying. That we would have expected very much larger unemployment as a result of the 2007 crash. It didn't happen because real wages fell rather than those wages staying static and unemployment soaring. Which means, ineluctably, that something has happened in the UK to move us from a less Keynesian world and into a more New Classical or Real Business Cycle one. A reasonable guess there would be the labour market reforms of the 80s and 90s. Even, perhaps, those of the 00s for I can see an argument that larger in work benefits will temper the reluctance of people to accept falling real or nominal wages.
The amusement of this of course is that it was Blanchflower who was among those shouting loudest that without significant Keynesian fiscal stimulus we would have exactly the unemployment rise that didn't happen: because wage rigidity.
There is one thing that grates in this paper though and it's this chart:
The idea of comparing UK median wages with US median wages has merit. But it is necessary to compare like with like. And UK wages are very close to total compensation. There's very little that employers offer over and above what is recorded as wages in the statistical system. US wages are much more unlike total compensation for the system of providing health care insurance is usually through one's employer. And the cost of that has, as we all know very well, changed considerably over recent decades: it must have done, the health care sector has doubled as a portion of GDP over this same time period. US workers have therefore had at least some rise in total compensation, a rise that isn't picked up in the total wages that is being used here.
In more detail, as you can see from this paper, the Current Population Survey does not include employer 401 (k) contributions nor employer costs for other benefits, while the National Compensation Survey does. And given that both of those have been rising in recent decades it's not really true to start asserting that median incomes have been static in the US. You can only get there by leaving out some of what is, quite obviously, income derived from working, that is, health care insurance and retirement contributions.
