The wisdom of Scott Sumner

by on June 29, 2011 at 6:18 am in Economics, Uncategorized | Permalink

On structural unemployment:

…there is no hard and fast distinction between cyclical and structural unemployment. For instance, if structural unemployment in American has risen closer to European levels, it may be partly due to the decision to extend unemployment insurance from 26 weeks to 99 weeks, and to increase the minimum wage by over 40% right before the recession. Does that mean that demand stimulus cannot lower unemployment? No, because the maximum length of unemployment insurance is itself an endogenous variable. If stimulus were to sharply boost aggregate demand it is quite likely that Congress would return the UI limit to 26 weeks, as it has during previous recoveries. For similar reasons, the real minimum wage would decline with more rapid growth in demand. Aggregate supply and demand are hopelessly entangled, a problem that many economists haven’t fully recognised.

Read the whole thing.  I would add a few points.  First, structural and cyclical hypotheses interact in another way, namely that the degree of nominal stickiness for unemployed workers will depend on structural factors.  Second, the partially structural nature of unemployment is becoming increasingly clear with time; wages are not sticky forever and we are not at risk of a downward deflationary spiral from a round of wage-cutting among the unemployed.  Third, the unemployment itself is becoming increasingly structural, even if you think it was mostly cyclical in the first place.  Not working is bad for people.  Fourth, structural unemployment does change the appropriate mix of monetary and fiscal policies, although the net effect is indeterminate in theory.  Monetary policy should be expansionary, but structural forces behind unemployment can make traditional fiscal policy either more or less effective (it is more important to target disaffected workers, but also harder to do so) and you can think of that as the frontier policy question of the day.

Here is Scott’s bleg.  Could Scott be blogging again?  The consumer surplus from the internet just went up.

Lance June 29, 2011 at 8:57 am

I believe he said he was returning to blogging after the 4th.

TallDave June 29, 2011 at 8:59 am

Aggregate supply and demand are hopelessly entangled, a problem that many economists haven’t fully recognised.

I’m increasingly confused as to why we try to “fix” AD — if we know the problem of 2008 was that we weren’t as wealthy as we thought we were, then trying to address the recession by boosting AD seems a bit like losing your job and deciding to spend just as much anyway — a recipe for ruinous debt. The more rational approach would seem to be looking at today’s AD and saying “this is how much AD our actual productivity and debt levels can support.” Real growth in PPP GDP per capita should then be encouraged to come from productvity improvements due to innovation, etc.

It bothers me that so many people seem to think another sugar high is the answer to the last crash.

J Thomas June 29, 2011 at 10:21 am

The more rational approach would seem to be looking at today’s AD and saying “this is how much AD our actual productivity and debt levels can support.”

The problem is, we don’t know how much AD our actual productivity can support.

In 2008 we thought we were richer than we actually were, as shown by later events. Do we now think we are poorer than we actually are? How would we know? Wait for later events to tell us?

I say, if we don’t know how much AD we should have, we might do better to have the government try to regulate something else instead.

Like, try to remove subsidies and also taxes on production, while taxing consumption. We could make a sales tax that would be progressive and also completely fair, given modern technology.

For example, give everybody a government debit card. They can put money into their account to buy things, or even have their wages deposited directly into that account. Each week the government puts a little money into everybody’s account, and this provides a negative tax that can help poor people survive. Everybody else gets it too — so it’s fair. They get to spend X dollars each week without paying any sales tax. Then they get to spend an additional Y dollars with a small sales tax, and Z dollars with a larger tax, and Z’ dollars with a whopping tax. Everybody pays the same tax, but people who spend a lot pay more. The tax is deducted from your government debit account so businessmen don’t have to waste their resources collecting it. People can of course barter to avoid the taxes, but that’s inherently limited. And needless to say, this approach makes it difficult to be an illegal alien. Many places you will be able to buy things for alien currency, at a markup which may often be pretty large.

It might seem perverse to tax consumption when we want to increase demand, but consider — the less tax on our production, the cheaper things get and the more competitive our products get as exports, too.

When we tax or subsidise production, we distort production. Subsidies might result in our selling our products to other nations at a loss. The producers get their profits after subsidies, but the nation would be richer without those sales. We might even find ourselves importing raw materials at a high cost, and then subsidising businesses to make things which get exported for less than we paid for the imports!

If consumers pay the undistorted prices for what they buy, they might very well be able to afford more stuff. If we want them to buy more stuff than they can actually afford then the government can give them the money.

When we do want to distort the economy, we can do that. We can for example put high taxes on fossil fuels, and put the revenues into everybody’s debit cards. Anything that requires fossil fuels becomes more expensive, and yet people can still afford those things — if they prefer them to services and products which require less fossil fuel.

More than just trying to adjust demand up or down, we should look for ways that government can reduce its tendency to encourage uneconomic choices. Remove both subsidies and taxes on production (except when we truly want to distort production). Avoid making specific products artificially cheap or artificially expensive, except when we really want to.

Craig June 29, 2011 at 10:51 am

“In 2008 we thought we were richer than we actually were, as shown by later events. Do we now think we are poorer than we actually are? How would we know? Wait for later events to tell us?”

One of the more popular ideas in answering this puzzle has been to look at the long-term trend lines in the economy. Of course we have no guarantee that the future will look like the past, but at least it’s something.

There is no reason to believe that the famines are more “real” than the bumper crops. And in particular, if you run into someone who is suffering from hypoglycemia, a lovely glass of orange juice is just the thing for them. To assume that lapsing into a diabetic coma is just the natural state of things is to carry recency bias a bit too far.

J Thomas June 29, 2011 at 10:22 am

It bothers me that so many people seem to think another sugar high is the answer to the last crash.

Well said!

dirk June 29, 2011 at 12:45 pm

It bothers me that so many people are buying into that metaphor. Goes to show how powerful a metaphor is, however inapt.

J Thomas June 29, 2011 at 2:20 pm

How do you decide whether it’s inapt or not?

dirk June 29, 2011 at 4:30 pm

My point is that metaphors powerfully capture the public imagination whether they are apt or not. An inapt metaphor will beat a well-reasoned argument any day.

In this case, Sumner’s argument is not that monetary policy should be too loose but that it shouldn’t be too tight. I don’t see what that argument has to do with sugar highs, yet getting drunk or intoxicated in some way on monetary policy seems to be the metaphor of choice these days. I suspect our puritan culture lends moral power to that image.

No, sugar highs aren’t the answer but neither is hypoglycemia.

TallDave June 29, 2011 at 5:18 pm

Well, I wasn’t disagreeing with Sumner’s point on monetary policy or characterizing it as a sugar high, so much as agreeing with and amplifying his concern about AD being tied to supply.

For example, fiscal stimulus is a sugar high. Most policies attempting to tweak AD are a sugar high, a short term nonsustainable improvement in GDP — real economic gains come from the ability to produce more stuff more efficiently.

J Thomas June 30, 2011 at 9:13 am

No, sugar highs aren’t the answer but neither is hypoglycemia.

That makes sense.

I would see that metaphor as not exactly an answer to Scott Sumner’s point, but a general point of view.

It’s like the dot.com boom was a big sugar rush, but then when it didn’t immediately pay off in sustainable production, we got a shock. And then the real estate financing boom was like another sugar rush, and when that failed we got another shock. And keeping the primary attention on manipulating money isn’t likely to lead to real growth unless money manipulation is the limiting factor.

In that context, I’d figure that focusing our attention on how to artificially stimulate the economy is not like getting a sugar rush. It’s like taking a hit of methedrine.

Can we fit Keynesian theory into this metaphor? How about this — somebody’s anorexic and depressed. They’re starving and as a result they are not hungry, and don’t want to do anything much. Maybe if you take them to an amusement park and put them on some rollercoasters and feed them funnel-cakes and get them to apply for exciting jobs, they might perk up.

But by this point the metaphor is stretched too far. The human body has a whole lot of feedback mechanisms in place to aid its survival. It has evolved that way over 200 million years, improving on many previous models. Our economy lacks those mechanisms. Vertebrate bodies have a whole lot of homeostasis because it promotes survival — billions of failures have led to robust systems. National economies have hundreds of years of testing, with usually less than 200 national economies competing. Of course they aren’t stable. It’s surprising they limp along as well as they do.

TGGP June 29, 2011 at 8:56 pm

I like Karl Smith’s metaphors better.

dirk June 30, 2011 at 10:58 am

I agree with Karl Smith on this.

Silas Barta June 29, 2011 at 5:39 pm

Agree 100% Glad to finally see someone pointing this out.

Matt Waters June 30, 2011 at 1:38 am

You are misunderstanding AD. Holding your analogy constant, “fixing” AD just means bringing our income back to pre-recession levels.

A drop in AD is a drop in the nation’s NGDP, the dollar value of our output as a county. Ideally the RGDP, the amount of stuff we produce (NOT consume, btw), would stay constant. But let’s say that you are managing a company which sees its revenues grow 5% a year up to 100 million, and then suddenly revenues drop to 90 million. Your revenue drops, but your debt level stays at the same level. You also can’t drop wages and prices easily and instead you cut back on workers to not go out of business.

And that’s the issue with AD dropping: you need to “fix” AD because a drop indirectly causes a drop in RGDP due to sticky wages and the increasing real value of previous debt. Fixing AD also does NOT mean fixing consumption. I can’t emphasize that strongly enough. It means fixing NGDP, either through growing real production, which includes exports, or increasing inflation. You obviously want the former, but the latter, somewhat higher inflation to keep NGDP from dropping, is still preferable to unemployment going from 5% to 10%.

J Thomas June 30, 2011 at 8:46 am

…. Your revenue drops, but your debt level stays at the same level. You also can’t drop wages and prices easily and instead you cut back on workers to not go out of business.

So, you cut back production, but not prices, and your debt stays the same. How does this help?

If I remember it right, traditionally you were supposed to produce as much as you could sell above your variable cost, even if you had to sell below total cost. Because you get more money that way than you do if you sell less. But that’s out of date?

How about this. Say that 10 years ago you got a deal from a foreign government. They’d give you low-interest loans, and no environmental regulation or other expensive paperwork, and cheap labor, if you’d move your production there. And you did it. But you kept some production going in the USA too, because after all a lot of your fixed costs were sunk costs, and you could sell above variable cost. And there was always the chance the foreign government would change their mind and impose heavy taxes, or nationalize your stuff, or suffer a revolution or something.

But now, 10 years later, demand is down. Which plants do you shut down, the profitable foreign ones or the less-profitable US ones?

OK, we want the government to manage things so that money instability doesn’t mess up business. And it would be good if when business is generally slow, wages went down and interest rates on existing loans went down.

We already have a system which sometimes helps wages and other things during periods of high inflation. The government publishes inflation rates, and some wages get cost of living adjustments to make up for it. Of course, this only works well for employees who have a strong bargaining position. If there’s a reasonable chance they’re going to fire you, they won’t agree to a COLA clause. But maybe we could do something like that with loans. Have adjustable-rate loans, that depend on some government economic statistic. When business generally slows down, interest rates on existing business loans automatically go down too.

That way, debt payments wouldn’t stay level in hard times. And lenders would have an incentive to keep the economy moving.

Matt Waters June 30, 2011 at 5:58 pm

You’re correct about variable cost, but here’s the issue with variable cost. The biggest component of variable costs, wages, are very hard to reduce across the board. The inflexibility in wages feeds into an inflexibility in prices once prices are reduced to a certain level. Once prices get down to VC, the firm can’t reduce prices further and “make it up on volume.”

This price and wage floor basically acts like a minimum wage. The equilibrium prices are below the actual prices. Prices higher than equilibrium bring in less demand but more supply. Just like the minimum wage, you get more unemployment and idled capacity because of this market failure.

As far as debt, I think we’re in agreement that fixed-rate debt becomes more expensive in real terms due to deflation. Firms and customers must cut back on everything other than paying back debt, which leads to yet more deflation due to lower monetary velocity. Adjustable-rate debt helps somewhat, but the LIBOR and short-term treasuries have a zero-floor. The adjustable-rate debt can go down to a LIBOR of 0%, but not further. Past that point, deflation just makes debt more and more expensive.

“That way, debt payments wouldn’t stay level in hard times. And lenders would have an incentive to keep the economy moving.”

I think you’ve just summed up how the Fed typically responds to recessions. They make money cheaper when something like the .com crash happens, and the lower rates offset the decline in investment. From 2001 to 2003, the Fed Funds rate went down from 6% to 1%. The lower interest rate offset the crash of the tech boom and the general uncertainty surrounding stocks after Enron.

The Fed also lowered interest rates after Lehman failed, but this time they hit 0%. At that point, unorthodox policy is needed to stimulate the economy. In late-2008, the Fed did expand the discount window and engaged in unorthodox policy. However, the policy just lent to banks at 0%, who would then keep the cash in their vaults. It wasn’t until the Fed bought bonds other than short-term risk-free bonds in QE1 and QE2 that the economy turned around.

J Thomas June 30, 2011 at 7:35 pm

The biggest component of variable costs, wages, are very hard to reduce across the board.

Well, let’s change that. We sometimes have an automatic COLA adjustment for wages. Arrange for a second adjustment, perhaps based on average corporate profits, or consumption GDP, or something like that. When the economy heats up then wages automatically increase. When demand goes down, wages automatically decrease. Possibly make it mandatory. (If we assume that whoever has the upper hand in negotiation wouldn’t want it that way, then it does nothing unless it’s mandatory.) Then wages are not so sticky. In good times, wages automatically go up which increases demand but not as much as it increases prices. A drag on booming economies, which needs to be somehow tuned. In bad times wages go down, decreasing demand but also reducing wage stickiness.

From 2001 to 2003, the Fed Funds rate went down from 6% to 1%. The lower interest rate offset the crash of the tech boom and the general uncertainty surrounding stocks after Enron.

I’m no expert in such things, so I want to go over the boring details. If the Fed rate is 6% and you have a loan at 7%, then when the rate goes down to 1% you can get a new loan at 2% and use the money to pay off the 7% loan early? So you aren’t stuck with the high rate? But you have to go through a lot of paperwork and some expense to do that, and maybe you have an early-payment penalty? And maybe now the bank thinks you aren’t as good a credit risk at 2% as you were before when your business was doing well, so they might charge points or demand a higher rate.

I say, what if the government keeps a statistic that measures how well the economy is doing, something kind of like GDP. And set interest rates to reflect that statistic. So when you buy a 40-year bond, you don’t really know what rate you’ll get. It will vary over the years with the economy.

If fixed-interest loans were rare, then fixed-interest bonds would not be a safe-haven when the economy goes bad. Everybody would have a stake in the economy growing in a stable way.

The Fed also lowered interest rates after Lehman failed, but this time they hit 0%.

Obviously nobody will lend money at less than 0%, because they would have more money if they didn’t make the loan. The government could subsidise loans — give borrowers some of the money so they effectively get a below-zero loan even while lenders get some interest.

But if loans were adjustable-rate, adjusted by the economy, then when the economy goes real bad then existing loans could turn negative-interest. The principle starts evaporating just because there’s a depression! So lenders would have a strong incentive to avoid depressions, like everybody else.

Matt Waters June 30, 2011 at 8:40 pm

I would agree that your idealistic version of a new economy could offset the effects of deflation. If wages could be reduced more easily, and adjustable-rate mortgages could go negative, it might offset the detrimental effects of deflation.

However, that isn’t actually how the economy works. I have never seen COLA’s outside of union employment and indeed they are extremely rare outside of public-sector workers. The vast majority of workers work for some wage set in advance and the wage is very difficult to reduce across the board. I have also never heard of COLA’s which work the opposite way. A COLA did that for Social Security checks, and the government immediately stepped in to stop that. In reality, people are extraordinarily tied to their nominal compensation.

Adjustable-rate mortgages also can’t really set negative rates in practice. Most are set according to the LIBOR, which is the rate big banks lend to each other. No bank would ever lend to another bank for less than zero. They would keep cash instead. The only way to set a negative LIBOR is for the Fed to be the only lender to banks in the overnight market, which would be political suicide. You can imagine what zerohedge would do with that.

On lower interest rates, you are halfway correct about the effects of lower interest rates. After rates are reduced, homeowners and businesses do refinance at lower rates which saves them money. They then spend or invest that money. However, lower interest rates also make new borrowing more attractive. More people buy houses at a prime rate of 5% vs. 8%. More companies also issue bonds at 3% vs. 5%. More loan demand offsets the decline in AD from some adverse event which spooks consumption and investment.

The issue in 2008 was that Lehman’s bankruptcy REALLY spooked both consumers and investors. The Fed reduced interest rates, but hit the zero bound and could do no more to offset Lehman’s bankruptcy. The answer is to go past the zero bound by penalizing banks for holding excess reserves, like Sweden has done. An excess reserve penalty in effect is a negative interest rate on cash holdings. The Fed can also go down the yield curve, buying bonds with longer maturities, which is what QE2 did. It could also go into unorthodox markets, such as asset-backed securities, which is what QE1 did.

Unfortunately, the Fed actually put in place a small positive interest rate on reserves. Bernanke did this as political cover since it’s a future tool to fight inflation. However, it was stupid because it encouraged banks to not lend even more.

J Thomas July 1, 2011 at 4:27 am

I would agree that your idealistic version of a new economy could offset the effects of deflation.

It would probably be necessary to impose it legally. I don’t know what it would take to make that politically feasible. But if I assume the politics can’t allow it, then I will never find out what it takes.

I have never seen COLA’s outside of union employment and indeed they are extremely rare outside of public-sector workers.

If COLAs were legally required, they would be inflationary during times of inflation, too. But I’m thinking about basing it on something else. Perhaps on per capital GDP divided by COLA. Make it a measure of productivity. Apart from raises and bonuses, require wages to go up when production increases, and down when production decreases. I expect some generally good results from that.

Similarly, instead of having fixed rate interest, and instead of expecting adjustable rate based on LIBOR to help, require loans to be adjusted according to national production.

So when national productivity goes up, existing wages and existing loans become more expensive. Marginal businesses find times getting harder — because the economy is doing well. Lots of people find they don’t want a runaway boom.

When productivity goes down, wages go down and the interest on existing loans goes down — perhaps to less than zero. “I’m sorry you’re losing money, but there happens to be less stuff being made, so your dollars are not in fact worth as much as they used to be. If you can do something to help the economy improve we’ll all benefit.” There could be side effects from that. The ones I foresee can all be worked around; they aren’t fundamental problems but oddities in the way the system currently works. But I often fail to foresee everything. Maybe there’s a fundamental problem that would make this a bad idea.

If it’s good, I should start shopping it to politicians and see if any of them like it. I guess making wages less sticky would be a lot more popular when the economy is doing well and voters would get automatic raises, than now.

Andrew' June 29, 2011 at 9:55 am

I decided this morning that I disagree with Tyler’s TGS benediction. It’s never coming back. Technology is fungible. They can have whatever they want to tinker with AD. It’s not going to matter. That’s not to say you can’t be optimistic, but noone has articulated a real reason yet.

J Thomas June 29, 2011 at 10:27 am

Technology is fungible.

You were so brief that I did not at all follow your reasoning.

Are you saying that if we depend on new technology to give us something more like a traditional middle-class lifestyle, that foreign nations can use the new technology to outcompete us? So we’ll still be outcompeted, until our consumption fits third-world standards and our workforce accepts third-world compensation?

Andrew' June 29, 2011 at 12:07 pm

Pretty much. The US can’t count on non-unique factors to restore us to the status established by unique factors.

Matt Waters June 30, 2011 at 1:42 am

By that logic, Silicon Valley workers won’t enjoy middle-class lifestyles because technology is fungible and Google could set up in Mississippi instead and pay lower wages.

J Thomas June 30, 2011 at 8:21 am

Matt, to the extent there’s something truly unique about Silicon Valley, then they’re safe.

But if the advantages can be reproduced elsewhere by government subsidy etc, then China can make an equivalent, and India, and Taiwan, and maybe Israel, etc.

And if the bigger, newer, more-heavily-subsidised versions turn out more profitable, it might indeed turn out that SV can’t survive.

Bollywood etc have not cut into Hollywood’s market, right? Because Hollywood has something special they can’t duplicate. If they could do as good a job as Hollywood does, they could win on price.

Scott Sumner June 29, 2011 at 1:12 pm

Thanks for the link. Some workers go on SSDI when they become discouraged by long term unemployment.

If monetary policy targets NGDP, then fiscal policy can no longer affect aggregate demand. In that case fiscal policymakers need to work on reducing labor market inefficiencies. Replace UI with self-insurance (as much as possible) with unemployment savings accounts. Eliminate the minimum wage, and replace it with some sort of EITC. Welfare reform. End occupational licensing laws. Etc.

One commenter asked whether the government should even be trying to control AD. As long as the government produces money, they determine AD, the only question is how much instability they will introduce into the economy.

J Thomas June 29, 2011 at 10:38 pm

One commenter asked whether the government should even be trying to control AD. As long as the government produces money, they determine AD, the only question is how much instability they will introduce into the economy.

Let me make an analogy. People used to hang carrots in front of cart horses, with the idea the horse could try to get the carrot but the carrot would always be just out of reach, and maybe every now and then the horse would get a little bite of it.

The government produces money that businesses try to get by selling stuff. If they get the idea that it’s too hard or risky to make a profit, then they won’t try. That’s like hanging the carrot too far away.

If they get too much of the money, they won’t work hard to get the last little bit. That’s like letting the horse eat the carrot. After he has it, why keep pulling?

If the money loses value too fast, that’s like putting an old moldy carrot in front of the horse that he doesn’t want anyway.

So, we have “bankers” who already have so much money that they see no reason to risk it hoping for more. If we allow too much inflation they will strenuously object, because that takes wealth away from them. Why should we let them keep the wealth? No reason at all except that they earned it fair and square by hard work and careful planning. Their financial acumen let them win the economy. Now that they own it, why should they let anybody motivate them to keep working hard trying to get even more? They have all the carrots they want.

anon June 29, 2011 at 2:30 pm

How often is there a Sumner piece that doesn’t say “NGDP” in it? I like Tyler better because the TGS disease isn’t terminal yet.

Contemplationist June 30, 2011 at 2:59 pm

Sumner is a case study in how effective repetition can be.

Comments on this entry are closed.

Previous post:

Next post: