Sunday, July 11, 2010

Setting monetary stimulus straight

In light of my recent link to Gennady Stolyarov's post about the gloomy future of the economy (especially for young people), I thought it would be a good idea to put it against the backdrop of mainstream economics and the "experts'" solutions.

A characteristic post is this one by the relatively libertarian Scott Sumner. Like pretty much every day, his idea is for the Federal Reserve to do a "monetary stimulus" by injecting money into the economy to prop up nominal GDP. (Yes, nominal GDP -- you know, the one that doesn't mean anything until you adjust it to real GDP and even then commits you to a easily-abused framework.) This, it would do by various mechanisms, all of which aim to "get banks lending". Stop paying interest on reserves, buy more of banks' (junky) securities, rapidly debase the currency ("quantitative easing") so they have to loan or else hold worthless cash, etc.

In frustration at such a stupid policy, I made this sarcastic comment on that post:

Yes, the economy will definitely collapse if the Fed doesn’t print up more money to make shoddy loans for purchases people don’t want, and it’s a shame that folks at the Fed are stopping Bernanke from such a wise action.

And to my utter surprise, Sumner replied:

Silas. I agree. :-)

Note: the smiley was in recognition of my sarcasm, not to indicate he's changed his mind.

So, Sumner realizes exactly what he's asking for, and still thinks it's a good idea. But since it apparently isn't obvious to everyone what's wrong with such a policy, I thought I'd spell it out clearly for once:

Banks aren't lending (in sufficient numbers). Mainstream economists want to prod them into lending. But why won't they lend in the first place? Because they don't expect the future loan payments to justify the loan. Now, when you grip them so tightly that they have to, for some reason or another, make these loans, have you changed the factors causing banks to believe loans won't be paid? No, you haven't. So, the loans will just throw money after wasteful projects, destroying output and making everyone poorer.

Note: even if you -- quite reasonably -- care about unemployed workers, and you dismiss this concern about wastefulness on the grounds that, "hey, at least it will lift off the joblessness albatross for so many families", that still wouldn't make such policies a good idea. The wastefulness means that reality will eventually rear its head and force these projects to be abandoned. Then, all the new skills workers could have developed while working on sustainable projects that satisfy actual demand, instead don't get developed, and whatever they did do has just retooled them for a useless activity, leaving them even worse off. Doesn't sound too compassionate to me ...

But let's say I'm wrong about that. Let's put aside, for the moment, our skepticism about economists' claims that the same policy that forces banks to lend, also causes these loans to work out and get repaid, making them not such stupid loans to begin with. Even then, you're still causing inefficient activities to happen that cause workers and investors to dig themselves deeper on unsustainable activities.

Looking back, one has to wonder how economists ever came to the consensus that making ultra-underpriced loans to clumsy, inflexible banks could ever possibly be a good idea. My suspicion is that it is a kind of Goodhart phenomenon: at the time these economic models were created, the metrics economists cared about did serve as good proxies for general economic health. But as they were targeted by policy, they lost their value as indicators.

Furthermore, economists failed to continually ground their concept of a "good economy" in what is meant by the term in common parlance. They don't keep checking back to see whether their policies would mean that people get the best combination of work, leisure, and consumption (all broadly defined). No: if an improvement doesn't show up as a cash exchange, it doesn't matter. If people aren't spending enough, then obviously that's hurting the economy and they should spend more.

You would almost think the economy is some god that demands sacrifices, given the way economists talk, rather than a characterization of our collective ability to satisfy wants.

So please, understand my anger when I read about how young people have all of their options cut off by the earlier generation, how they can't save or invest because of how much will be taken to make up for the failures of poorly run enterprises, how genuinely productive ventures are quashed by an outdated mentality of how the world should work ... and then Scott Sumner swings in to tell us that the best way to improve "the economy" is with ridiculously underpriced loans from newly-printed money to aging, inefficient companies that just wasted trillions of dollars destroying our productive capacity.

Advice for economists: Ask whether, not why.

-Don't ask, "What can we do to increase aggregate demand?"
Ask, "Why should we increase aggregate demand?"

-Don't ask, "What can we do to keep people from saving so much?"
Ask, "Why does 'the economy' so crucially depend on people not saving, and why do I care about the health of the 'economy' in that sense?"

-Don't ask, "What can we do to get (traditionally measured) output back up?"
Ask, "Why is it necessary for that measure of output to go up? Would it be so terrible for people to produce less, if that's what they really want, based on honest assessments of the future?"

Get the picture?

10 comments:

Robert Fellner said...

I not only understand your anger I'm quite familiar with it as well!

Great post.

Anonymous said...

The only thing worse than the neo-keynesians demand for boosting aggregate demand through public borrowing/spending - regardless of the long term consequences - is the monetarists demands for an increase in money supply via forced private credit creation.

At least the neo-keynesians realize the private sector is incapable of expanding their balance sheets. Whether they know why is another issue entirely.

jsalvati said...

I often see your comments on Less Wrong and OB, so I know you're a smart guy. You're being too hasty here. Your arguing without really groking the nature of money which is fundamentally macroeconomics is about. You argue about whether more spending in the economy is a good thing without understanding why adding more money increases spending (or the reverse) in the short run. You treat monetary policy as a Mysterious Lever.

The fundamental idea is that the demand for money is volatile and it is inefficient to force prices to adjust to changes in the demand for money, as the process of such adjustment is difficult. It is much more efficient to allow the supply of money

I encourage you to to read the posts at the very beginning of Sumner's blog. Nick Rowe's blog Worthwhile Canadian Initiative and Woolsey's blog Monetary Freedom are also informative on this topic. The Cato Unbound issue on this issue is also informative (http://www.cato-unbound.org/archives/september-2009-monetary-lessons-from-the-not-so-great-depression/). I wish I could offer more precise reading advice, but I don't think I can. I promise you that there really is something there.

Silas Barta said...

Thank you for your input, jsalvati. My comments:

You argue about whether more spending in the economy is a good thing without understanding why adding more money increases spending (or the reverse) in the short run.

Yes, because if more spending *isn't* necessarily a good thing, that would obviate the entire question of *how* to increase spending!

The fundamental idea is that the demand for money is volatile and it is inefficient to force prices to adjust to changes in the demand for money, as the process of such adjustment is difficult. It is much more efficient to allow the supply of money [to expand to meet demand?]

Alright, that's making a bit more sense of the policies that Sumner-types advocate, and I hadn't heard it phrased in quite this way before. But I don't see how it's responsive to my point that any method of increasing the money supply involves, in effect, bankrolling projects that couldn't get private funding (or public support), and therefore will, on average, just compound the pain of any existing inefficiency.

I encourage you to to read the posts at the very beginning of Sumner's blog. Nick Rowe's blog Worthwhile Canadian Initiative and Woolsey's blog Monetary Freedom are also informative on this topic. The Cato Unbound issue on this issue is also informative... I wish I could offer more precise reading advice, but I don't think I can. I promise you that there really is something there.

I actually already read the September 2009 Cato Unbound exchange recently, and I still didn't find it to be responsive to the reservations I have about questionable monetary policies (like deliberate inflation, making underpriced loans to clumsy banks, etc). For example, what if less spending -- such as by doing more production (e.g. cooking) for oneself -- really is the efficient re-allocation of resources, given the newly discovered non-productivity of the economy?

That, plus your own inability to paraphrase the logic behind such policies causes me to be skeptical of what more the other sources have to offer. I'll still give them a try, but I humbly suggest that you should also probe your own assumptions. Like I pondered in the link above, can you ground your reasoning all the way down to the layman level understanding of what a "good economy" means?

Even the professional economist I asked had not done so ...

jsalvati said...

I have a somewhat response here: http://goodmorningeconomics.wordpress.com/?p=476

Silas Barta said...

That returns a dead link. Did you mean this?

The best way I can think of to address your position is: the premise behind Sumner's argument (and that of the macroeconomic consensus he references) is that total spending should be kept at some arbitrary level and growth rate.

But why? Why is more spending necessarily good? When people spend money, they are getting what they need via specialization and through a cash economy from (mostly) strangers.

Certainly, that has *usually* been a good metric of economic health: only a real undeveloped economy has people bartering or producing everything for themselves in their households. But this still runs into the Goodhard problem I complained about: you cannot take that general historical correlation and infer that, on the margin, decisions to spend less are "hurting the economy" -- not unless you're going to redefine the economy as "passing around pieces of paper", which is about on the level of defining morality as "jumping off cliffs" and accusing people of being immoral because they're cliff-jump-refrainers.

If I stop hiring my maid to clean my place, and do it myself, does that really indicate a loss of efficiency? There are any number of reasons I could be doing so. When you contort my incentives to the point that I reverse that decision, aren't you just throwing the inefficiency right back in?

Spending is not, in and of itself, good. Rather, good spending is good. (Less, tautologically, the spending that people would do without manipulation, and because of an honest assessment of their situation, is good.)
Spending that exists only because incentives were jimmied until it looks like people's best option ... is not good.

jsalvati said...

Yes... I mean that link; I am not sure how that could happen.

I think this is a good conversation to have and for other people to hear, so I will respond on my blog.

Doc Merlin said...

While I agree with some of your post, I disagree with this bit:
"Yes, nominal GDP -- you know, the one that doesn't mean anything until you adjust it to real GDP and even then commits you to a easily-abused framework."

This isn't entirely true. As debts are nominally denominated, NGDP can be thought of as aggregate income which which to pay off debts. When NGDP drops, it means that we don't have enough income to pay off debts so we must restructure.

For an example, I'll use a representative agent model:
So a person buys a house in 1970 with a loan. Then there is huge inflation that makes their real earnings unchanged, but their loan payment is much less as a % of their income. This is a very real change in income for them.

This representative member of society is very much like the entire economy (It too has debt ratios, etc). So, in short, Unexpected changes in nominal GDP have real effects in the short term.

Doc Merlin said...

I should clarify, in my example, there was inflation, but their income rose at the inflation rate.

in page analytics said...

That's really good stuff. My buddies at work will definitely be awestruck ! Thanks for sharing.