Just to clarify, when I say I’m an aggregate demand skeptic, I don’t mean to deny that aggregate demand exists, at least in principle (though we might not measure it well). I also don’t mean to deny that it does indeed sometimes fall sharply. Further, I concede that these falls are caused by human activity, and not sunspots. However, what I am skeptical about is that these anthropogenic falls in aggregate demand are damaging to human welfare, and what I outright deny is that Ben Bernanke can make things better if we gave him a printing press and just let the guy do his job, for crying out loud.
I am going to write up a careful response to the quasi-monetarists (people like David Beckworth and Scott Sumner), but first I want to make sure I understand their worldview. So some questions:
(1) For Scott, you say that the recession that officially began in December 2007 was due in part to a fall in housing construction. But can that be, since you have demonstrated that housing construction started collapsing back in 2006? Why didn’t the maintenance of NGDP growth all through that period ensure that the decline in housing construction was offset by increases elsewhere in the economy? (Note: I know you distinguish between a mild “real” recession starting in late 2007, and the big boy “nominal” recession starting in June [?] 2008 when NGDP growth collapsed. So put the later recession aside. Just looking at the first one, how does your timeline work? How can a collapse in housing construction that started in early 2006, explain a recession that didn’t start until December 2007? And of course, whatever reason you give, then explain why I can’t use the same reason to get out of the cage you thought you had put over Arnold Kling and me, regarding unemployment. E.g. is a one-year lag OK, but a two-year lag is pushing it?)
(2) As I understand Beckworth and Sumner, their position is that nominal income collapsed in 2008, and so there wasn’t enough total spending to keep everyone fully employed. It’s true, if all prices and wages were perfectly flexible, there’d be no problem. But since wages in particular are sticky downward, the drop in total spending is accommodated by a fall in real output. Yet in that case, the following chart is interesting:
Nominal GDP–i.e. total spending–is now the highest it has been in US history. So why is unemployment still so high? I can imagine one possible answer: Because productivity has grown in the meantime, so that the mid-2008 level of total spending is now only sufficient to employ 91% of the workforce. Yet if that is the answer, then doesn’t it solve the “sticky wage” problem? In other words, if the problem as originally conceived was that falling prices combined with sticky wages led to above-equilibrium real wage rates, the shouldn’t an increase in productivity mitigate that problem? Employers just keep nominal wage rates at their “stuck” level, and the ever-improving workers raise the equilibrium wage rate up and up, shrinking the gap.
So to repeat, I think the explanation–at least if Beckworth/Sumner like the “productivity has grown” answer for why record-breaking NGDP hasn’t pulled us out of the recession–is internally inconsistent. If stagnant NGDP is only a problem because of sticky wages, then it doesn’t make sense (at least to me) to say, “Oh shoot, if only productivity hadn’t grown the last 2 years, we’d now be at full employment.”
This is a crucial point, so let me put it differently. Suppose there had never been any problem with AD. Instead, back in mid-2008, all of a sudden there was a huge leap in worker productivity–namely, workers back then suddenly became as productive as they are (in reality) right now. Now with sticky prices and wages, this could lead to a big spike in unemployment: The fully-employed factories are cranking stuff out, and inventories are piling up, because employees still have the same take-home pay, and the owners refuse to cut prices. So the workers don’t have enough total income to buy the higher output, at the original prices. So in this case, wouldn’t the goal would be to reduce the price level? But if so, then why in our environment, does Scott (not sure about David) think the Fed’s policies are starting to work when price inflation ticks up? (I have seen him say that on his blog; I’m not conflating “I think the Fed needs to do more to boost AD” with “I want higher price inflation.”)
(3) For my third question, I want David and/or Scott to give me their take on this graph:
Isn’t it a bit weird, if the alleged problem is one of wages and prices that are rigid downward, that both CPI and average hourly wage rates are at all-time highs? If we have a glut in the labor market, because at least one of those things (wages or prices) is stuck against a binding constraint, then why are they both moving upward? Furthermore, even the gap between the two has grown back to its level of the period during the peak housing bubble years, when there was no unemployment problem.
I think Scott may say, “When we speak of wages being ‘sticky downward,’ that’s sloppy. Actually wage contracts have built-in increases, and there is an expectation among workers that they will get raises. So it’s not that wages are flat and can’t sink to restore equilibrium, it’s worse than that: Wages are ‘stuck’ on an upward trend.”
If that is indeed his answer, then fine, but then I repeat my question from (2): If the problem is that nominal wages are rising too quickly, then wouldn’t the solution be to boost productivity? In that way, the workers could deserve their nominal wage increases. And yet, I thought we had to cite “rising productivity” as the reason for why record-high NGDP hasn’t gotten us out of the rut yet.
NOTE: I am not claiming that I’ve found actual contradictions in the Beckworth/Sumner worldview; I know there are internally consistent formal models with all the i’s dotted and t’s crossed. But I’m saying from the verbal discussion of those models, I can’t keep things straight. So please enlighten us, you monetarist Keynesians you.