I’m doing this as a separate post because I want to keep the arguments distinct… Let’s go back to this example from Scott Sumner:
NGDP is $10, and each of ten workers [gets] a dollar in wages. Now they negotiate a 10% wage increase, in anticipation of 10% more NGDP. So wages rise to $1.10. Now there is only enough NGDP to emply nine workers, because NGDP unexpectedly stayed flat at $10. And that is true no matter how high productivity rises.
What is so intriguing to me about Scott’s worldview is that he looks at things that, in my mind, are effects, and attributes causal power to them. And yet, it’s hard for me to attack his position, since our differences are almost metaphysical.
So for one thing–and I don’t mean this as a cheap shot–when is the last time you negotiated a wage increase “in anticipation of 10% more NGDP”? It’s not even true that employers do that. They might concede to wage hikes because they anticipate higher P, and for sure if they anticipate higher PxQ of their output, but I don’t think too many people are walking around, thinking about what NGDP will be next year. I say this, because I have to be careful to define NGDP whenever I write about Scott in polite company.
Let me reiterate my reasons for bringing up productivity in the original gauntlet I threw down to the quasi-monetarists. Remember, they are saying that the main problem with the economy is that total spending fell in late 2008. They concede that with perfectly flexible wages and prices, that wouldn’t be a problem; it would just change nominal values, but nothing “real.” But shucks, in the real world, some prices and especially wages don’t fall easily, and so a sudden drop in total spending (NGDP) will lead to high unemployment.
OK, so then I pointed out that NGDP has recovered, and in fact is currently the highest in US history. So why hasn’t full employment resumed?
Now if I were going to explain this, using sticky prices and wages, I would give a simple numerical example like this: Before the recession, each of ten workers got $1/hr in wages, each produced 1 widget per hour, and the price of widgets was $1.
But now, productivity has risen. Each worker can produce 1.1 widgets per hour. Yet because wages and prices are stuck at $1, the employer can’t afford to keep all ten workers employed. Since he is stuck paying them $10 total per hour, consumers only have $10 to spend on each batch of widgets. That was fine before, when widgets cost $1 each and there were ten produced per hour.
But now, if the employer holds all 10 employees, every hour an extra widget piles up, because the workers produce 11 per hour but can still only collectively afford to buy 10 per hour.
Now if things were flexible, it would be simple: The employer could either raise wages to $1.10 per hour, or he could cut the price of widgets to about 91 cents each. Either way, the real wage would increase, and workers would earn enough income to be able to buy enough product to keep them all employed.
Now note, I hate thinking in these terms, but if I were forced at gunpoint to do it, this is how I would go about it. I don’t even know what it means to talk about, “What if total spending is $10, but everyone expected it to be $11?” We know what the determinants of “total spending” are, so I would rather discuss theories of what moves those things.
Let me make sure you get what I’m saying: In my example above, I think you could tell an NGDP story to “explain” why there was unemployment, and then how it got solved. When W and P were stuck at $1 each–while worker productivity went up 10%–NGDP didn’t grow; it stayed flat at $10.
Now consider the solution scenario where W goes up to $1.10 per hour. In this case, the workers are earning $11 total per hour, and they produce 11 widgets total per hour, so they spend the $11 buying it. So NGDP (per hour) rose from $10 to $11. “Aha!” Sumner exclaims. “Once we boosted NGDP by 10%, the unemployment problem was solved!”
Now consider the other solution scenario, where P drops to 91 cents per hour. In this case, the workers are still getting paid $1 per hour. They can each afford to buy 1.1 widgets per hour, and so the 10 workers collectively buy 11 widgets for $10 total. So NGDP stays constant at $10. Sumner says, “Aha! Just as my theory predicts. Constant NGDP is consistent with rising real GDP and full employment, so long as prices can adjust downward.”
I haven’t quite put my finger on it… I just think there is something fishy in reasoning from “NGDP.”
Let me take one more stab at this and I’ll drop it for a while… If I’m looking at the world in terms of NGDP, where “total spending” is a concept that has explanatory power, I might reason like this: “During a recession, consumers are only spending, say, $13 trillion on output, but at the current level of output prices, that only corresponds to 97% of last year’s output. So that means employers have no choice but to lay off, say, 10% of the workers, so that the remaining (and most productive) 90% of the workers can produce 97% of last year’s output. So the unemployment rate goes up to 10%, and real output falls by 3%; we’re in a bad recession. The way to fix this, is to reduce the unit price of output. That way, with $13 trillion in spending, consumers can now afford to buy the same amount of stuff as last year. So employers can go ahead and hire back those laid-off workers, since there is now the demand to soak up the excess capacity.”
I think that has a certain plausibility, if you reason in terms of NGDP causing recessions. But if you think of it in terms of microeconomics, it makes no sense: It says the way to fix a glut in the labor market, is to make real wages go up. When macro reasoning seems to contradict micro reasoning, I feel on much safer ground by going with the micro.
Now I think there must be some resolution to this apparent paradox, in that things happen with “the velocity of money” (another term that doesn’t make sense at the individual level) and so on, so that NGDP changes as it needs to, to correspond with what we know from micro analysis.
Either that, or I’m going down a blind alley. It’s been a long week, you guys can hash it out in the comments.