Has the Fed “assumed broad responsibility for nominal output growth”?
Deutsche Bank says yes:
The Fed’s dual mandate of maximum employment and price stability implies substantial oversight of the private economy. We develop a proxy for nominal private GDP growth, perhaps the broadest measure of private sector activity, and decompose this proxy into two segments, one that is explicitly within the scope of the dual mandate and one that is not. Weak nominal private growth in the present cycle has been almost entirely due to the latter, in particular the abysmal trend in productivity. We conclude that while a narrow reading of the Fed’s dual mandate might suggest that it is far behind the curve in terms of rate hikes, the Fed has favored a very cautious approach because it has implicitly assumed broad responsibility for nominal output growth. As the productivity slowdown is unlikely to get resolved in the near term, we expect the Fed to remain on hold through much of this year, and possibly into next year.
I highly recommend Gregory’s Clark’s review of Robert Gordon’s book on the growth slowdown. He does a nice job of explaining why productivity growth is going to remain low for the foreseeable future:
The core of Gordon’s pessimism about future technological advance is that the modern US economy is now heavily based around services, accounting for 80 percent of output. Manufacturing, traditionally a sector with higher efficiency advance, has shrunk to 12 percent of the economy. . . .
A surprising share of modern jobs are the timeless ones of the pre-industrial era— cooking, serving food, cleaning, gardening, selling, monitoring, guarding, imprisoning, personal service, guiding vehicles, carrying packages. Food production and serving, for example, now employs significantly more people (9.1 percent) than do production jobs (6.6 percent). One in ten workers is employed in sales. The information technology revolution to date has left these jobs largely untransformed. Workers in these types of jobs in Europe in 1300, if transplanted to modern America, would need little retraining.
Even outside services, we can find jobs with no gain in productivity since the Industrial Revolution. Builders’ price books in eighteenth century London show the rate at which bricklayers laid bricks in house construction. In 1787 this was 75 bricks laid per hour. For modern England the rates are lower, 225 years later, at around 50–70 per hour.
Combine slow productivity growth with at most 2% inflation and a working age population that’s growing very slowly, and you are left with 3% NGDP growth and very low interest rates for as far as the eye can see. Get used to it.
HT: Federico
Tags:
19. May 2016 at 07:10
Scott, do you believe in the long-term neutrality of money?
19. May 2016 at 07:27
“Combine slow productivity growth with at most 2% inflation and a working age population that’s growing very slowly, and you are left with 3% NGDP growth and very low interest rates for as far as the eye can see. Get used to it.”
nice to see you’ve succumbed to the secular stagnation theory. welcome aboard
19. May 2016 at 07:40
on a more serious note I do believe that Gordon is right about how innovation and it’s ability to boost growth isn’t what it use to be. But I’m also aware that Gordon may be underestimating A.I and the potential of bio tech.
19. May 2016 at 07:43
Interesting article in the Economist: http://www.economist.com/news/leaders/21699121-americas-next-president-should-modernise-federal-reserve-system-right-kind-reform.
I was especially struck by this:
“This matters because banks tend to profit from higher interest rates. Regional-Fed presidents tend to be the most hawkish members of the FOMC, as their dissenting opinions suggest.”
If true, this does seem particularly worrisome. But it also suggests that regional fed presidents are too daft to see that their higher interest rates would be better achieved through higher NGDP growth rather than tightening.
So what’s worse, the conflict of interest, or the sheer shortsighted stupidity?
19. May 2016 at 08:03
Oderus, Yes, I think that’s a useful approximation of reality, although not literally true in extreme cases (i.e. severe deflation, hyperinflation, etc.)
James, You said:
“welcome aboard”
Shouldn’t it be the other way around? I’ve been saying this for many years.
Joe, You said:
“If true, this does seem particularly worrisome. But it also suggests that regional fed presidents are too daft to see that their higher interest rates would be better achieved through higher NGDP growth rather than tightening.
So what’s worse, the conflict of interest, or the sheer shortsighted stupidity?”
Shortsighted stupidity is far worse. As you correctly point out, it’s tight money that causes low interest rates. The hawks are their own worst enemies. Look at how the Germans delivered low i-rates to Europe.
19. May 2016 at 08:16
I hate this fatalism. Today, productivity comes from dense networks of human capital. That is the story of our time. And we have managed to create Detroit-level domestic outmigration from the places where this productivity happens. It’s as if 19th century Britain outlawed coal mining and then all its economists sat around going, “Well, guys, it looks like we’re shit out of luck. Probably just something to do with entropy or something. There’s just no hope for progress.”
19. May 2016 at 08:29
Obviously, believing that monetary policy can alleviate the “problem” of sticky prices requires that you believe in the reality of long-term neutrality, not just as an approximation.
19. May 2016 at 08:32
I’d prefer to see the Fed try for 4% NGDPLT, or even 4.5% or 5%. If that results in 3%-4% inflation, so be it. The Fed can’t make us more productive per se, but it can facilitate the conditions for the private economy to do its best. The short-sighted stupidity is beyond frustrating. IOR is a powerful enabler. It’s so easy for the hawks to focus on “it’s just a measly 25 bps more” instead of seeing that they’d have to sell over a trillion from their balance sheet to move the Fed funds rate. Normalize IOR now (ie, let’s get IOR back to zero).
19. May 2016 at 08:45
Scott:
Isn’t your answer–“Yes, I think that’s a useful approximation of reality, although not literally true in extreme cases (i.e. severe deflation, hyperinflation, etc.)”– a bit like saying bacteria won’t hurt you except when there’s enough of it to overwhelm your immune system and kill you.
I have trouble believing that changes in the money supply have a binary effect: either they are neutral or they reach some threshold at which too much or too little of this neutral thing causes severe and permanent damage to prices and the economy. I would think the damage is cumulative.
19. May 2016 at 09:08
PS – right now, I’d take the “under” in a 3% NGDP bet over the next couple years. [bittersweet smiley face]
19. May 2016 at 09:52
Shouldn’t Investments and technology make restaurant workers more productive?
Aren’t we getting better and better at building manufactured homes and other buildings? Aren’t we now seeing people leave traditional brokerage and advisors to move towards automated advisors?
It seems to me that more investments of this kind during the 2008-2016 period would have made us much better off (and more productive). They were not made because of demand constraints. Even now, I think a lot more investments could be made, but are not because of weak future aggregate demand expectations.
I would advocate a 5% NGDP (or wages) growth target. If Sumner/Gordon are right (1.2% rGDP growth is the max), than we get 3.8% inflation for a few years. This would hardly be a catastrophe. If they are wrong, we get higher growth (say 3%) and smallish inflation (say 2%). In my opinion, these are very asymmetric risks.
19. May 2016 at 10:11
I’m with LK Beland’s suggestion. I’d bet we get a result in the middle. 2% growth and 3% inflation.
19. May 2016 at 11:23
Déjà Vu: In the late 1930s, Hansen argued that “secular stagnation” had set in so the American economy would never grow rapidly again because all the growth ingredients had played out, including technological innovation and population growth. The only solution, he argued, was constant, large-scale deficit spending by the federal government.
19. May 2016 at 15:57
Can’t say I am a fan of defeatist sentiments, which crop up in every depression and recession.
Why the productivity boom of the 1990s? That long and forgotten time.
I am surprised we do not see more Automat-style restaurants, or buffet-cafeterias. Maybe because labor is relatively cheap in the USA?
19. May 2016 at 17:56
“I am surprised we do not see more Automat-style restaurants, or buffet-cafeterias.”
Last year, a restaurant without wait staff or cashiers opened in San Francisco. You order and pay for your food via an app and then pick up your order from a cubby hole when it’s ready.
http://www.businessinsider.com/eatsa-fully-automated-restaurant-chain-2016-2
19. May 2016 at 18:20
BTW, in regards to the excerpt from the Gordon book about no improvement in productivity over hundreds of years in brick laying, there is now a brick laying robot that can construct a brick home in 6 weeks less time than it currently takes:
http://www.gizmag.com/hadrian-brick-laying-robot-fastbrick/38239/
And a few years ago, a machine was created to automatically lay down brick paved roads and paths:
http://www.gizmag.com/tiger-stone-lays-paving-bricks/16951/
19. May 2016 at 18:53
@scott
I guess the 60% increase in the capital gains tax rate and the drop in fixed non-residential investment doesn’t have anything to do with the drop in growth.
20. May 2016 at 02:17
@scott
And I totally disagree with Gordon about the pace of innovation. He sounds like an ill-informed curmudgeon who is totally out of touch with the changes that are taking place in the modern world.
20. May 2016 at 08:59
Oderus, You said:
“Obviously, believing that monetary policy can alleviate the “problem” of sticky prices requires that you believe in the reality of long-term neutrality, not just as an approximation.”
No, it doesn’t. This is a strange argument for me, because 99% of the time I’m defending money neutrality. But if tight money leads to WWII (and it did) then real effects can persist for a very long time. East Germany is still poorer than West Germany because of tight money in 1929-32.
Carl, Maybe I’ll do a post on that.
Gordon, That’s exactly the point. You can invent a bricklaying machine, but they are so hard to use that they don’t materially impact the building industry. Growth in productivity in the building industry has been abysmal, which is why the relative price of houses, compared to manufactured goods, keeps rising.
dtoh, Even if the rise in the cap gains tax did explain the slowdown (it doesn’t), you’d just be proving my point that monetary policy can’t fix the problem. I’d be the first to agree that policy reforms would boost growth for a certain time period
And isn’t it possible that the people who are “out of touch” are those who don’t look at the aggregate productivity data, and instead rely on anecdotes about smart phones.
20. May 2016 at 12:19
@scott,
You said, “Even if the rise in the cap gains tax did explain the slowdown (it doesn’t), you’d just be proving my point that monetary policy can’t fix the problem. I’d be the first to agree that policy reforms would boost growth for a certain time period”
1. So the rise in cap gains tax and slow down in growth are just coincidental?… and tax rates on capital don’t have any impact on investment or growth? Did I just fall down the rabbit hole?
2. “if the rise in the cap gains tax did explain the slowdown you’d just be proving my point that monetary policy can’t fix the problem….”
The existence of one solution precludes the possibility of a second solution? I must of dozed off when they covered that in my undergrad logic course.
3. I’d be the first to agree that policy reforms would boost growth for a certain time period.
For a certain period of time? Do you mean temporarily? If so why?
4. Good monetary policy > less GDP volatility > less risk on investment > higher risk adjusted return > more investment > higher growth. So I think monetary policy can raise long term growth rates or is my logic wrong?
20. May 2016 at 13:52
“You can invent a bricklaying machine, but they are so hard to use that they don’t materially impact the building industry.”
Scott, would you elaborate on this? There was nothing in the article that indicated this robot was hard to use and was facing resistance. I could see this kind of automation failing to take hold if human capital is less expensive than the robot. But that’s a different issue.
21. May 2016 at 05:28
dtoh, You said:
“So the rise in cap gains tax and slow down in growth are just coincidental?…”
Not, because they did not coincide. The growth slowdown started in 2004, and the cap gains tax increase occurred in 2013. It might have slowed growth by 0.1% or so.
3. See my earlier posts where I discuss the difference between a level shift and a permanent growth rate shift. Almost all policy changes produce a level shift, not a permant grwoth rate shift.
4. More investment raises the level of GDP, not the permanent growht rate. Diminishing returns to capital, etc.
Gordon, You said:
“There was nothing in the article that indicated this robot was hard to use and was facing resistance.”
Nothing in the article that you linked to, but there was in the article that I linked to.
21. May 2016 at 06:31
@scott
“Not, because they did not coincide. The growth slowdown started in 2004, and the cap gains tax increase occurred in 2013. It might have slowed growth by 0.1% or so.”
No…part of the increase (estate taxes) increased in 2011. And the others were expected (in fact legislated) well before (2003). Or maybe expectations only apply to monetary policy not to tax policy. As for the 0.1%….. any basis for that? Or is it just a SWAG?
3. “Almost all policy changes produce a level shift, not a permanent growth rate shift.”
Maybe but only if the expectation is that they are temporary. If you get a permanent increase in expected after tax returns on investment, then you will have a permanent shift from consumption to investment and a permanent increase in the growth rate.
4. “More investment raises the level of GDP, not the permanent growth rate. Diminishing returns to capital, etc.”
As I said, if it’s one time…sure. If returns increase permanently you get a permanent increase in investment (and growth) rates.
Diminishing returns to capital? When non residential investment is less than 12% of GDP in a good year? ROFL.
21. May 2016 at 17:16
“Nothing in the article that you linked to, but there was in the article that I linked to.”
I’m not able to access that article. But in searching around the web I see that Robert Gordon points to fact that automated checkout machines have failed to displace cashiers and that ATMs have failed to displace bank branches as examples of technology failing to improve productivity. And I agree that technology will fail to improve productivity if the service provider implementing a new technology imposes a learning cost on the customer without providing any incentive. But Gordon overlooks one case in which the customer was given an incentive – self-service gas pumps. A surcharge for using a full service pump provided enough incentive for customers to make the switch. What would happen if grocery stores charged customers a fee for using a cashier or provided a discount for using the automated checkout? I also saw that Robert Gordon points out that robots are not restocking grocery store shelves. I suspect this is just technical ignorance on his part. Robots are not sophisticated enough yet to do such tasks. But they’re getting close despite this compilation of robots falling down during a DARPA challenge:
https://www.youtube.com/watch?v=g0TaYhjpOfo
21. May 2016 at 18:26
@gordon
You said,
>Robots are not sophisticated enough yet to do such tasks.
Wrong. Robots are not CHEAP enough yet to do such tasks.
22. May 2016 at 10:37
Yeah, a big part of the story is just Baumol’s Disease. Some tasks are just difficult to automate. I also think falling wages are a factor (pro-wage distortionary policies might be a decent automation-tech subsidy, and net beneficial in the long-run due to path-dependence).
Finally, I don’t think we can ignore increasing share of economic rents. Wouldn’t that explain a lot of the economic mysteries of the last few decades? Productivity slow-down, productivity-wage gap, intangible capital in equity valuations, low return to capital, etc.
22. May 2016 at 19:41
dtoh, You asked:
“As for the 0.1%….. any basis for that?”
Yup, there has never been any reputable study that suggested increasing the top cap gains rate by 8.8% would have a major impact on growth.
If you were right that expectations matter regarding taxes, then the Bush tax cut should not have mattered, as it was temporary. You can’t have it both ways.
I’m actually to the right of 90% of economists, who would say much less than 0.1%.
You said:
“then you will have a permanent shift from consumption to investment and a permanent increase in the growth rate.”
No, a permanent increase in investment will not lead to a permanent increase in growth, due to diminishing returns.
22. May 2016 at 20:58
I haven’t read Gordon’s book, and Clark’s review is paywalled for non-academics ($9.50 is cheaper than Elsevier’s $19.95 I suppose…), but he’s too pessimistic about food service automation. Chili’s has automated their cashier and drink reordering functions with at-table touchscreen credit card readers. McDonalds and Wendy’s are field testing order kiosks. I’m looking forward to Wendy’s automation because they *always* get my order wrong, no matter which store I visit.
Burger flipping is easy to automate, but stacking lettuce, tomatoes and bacon on top of a bun is harder. Famed robotics researcher Rodney Brooks has observed that picking clean, uniform parts like nuts and bolts out of a bin is one of the great robotics challenges where very little progress has been made. Picking an irregular slice of greasy bacon out of a pan full of other slices takes talent that mammals have been evolving for millions of years.
The other Gordon’s comment above about incentives to use automation is spot on. I started using ATMs for banking in 1987 when BayBank Boston began adding a $1 surcharge for using a real teller. The incentives can also be created by understaffing the human-serviced function. My Bank of America branch recently had someone standing at the door redirecting customers to their ATM because they had a shortage of regular tellers and those lines were so long.
Supermarkets and other big-box stores are beginning to do the same thing. I’ve seen a nearby Home Depot with only one human-staffed checkout aisle, while their 6 self-check stations are always ready. That’s a 600% improvement in productivity for that checkout worker.
The loss of employment opportunities for waitstaff, fast food order takers, and checkout workers is not their concern.
22. May 2016 at 22:01
@scott
1. There has never been any reputable study that suggested increasing the top cap gains rate by 8.8% would have a major impact on growth.
– Top cap gains rate increase by 60% not 8.8%
– There has never been a reputable study suggesting an increase in top cap gains rate would NOT have a major impact on growth.
– It doesn’t take an empirical study to prove this. It just take logic unless you believe demand curves slope up and supply slopes down. This can only be non-obvious to an academic who has never invested their own money in new business.
Diminishing returns???? With non-residential investment at or below 12%, the capital stock is barely replenished after deprecation. Honestly you can’t be serious with this argument?
23. May 2016 at 09:02
Anyone who travels knows that demand curves don’t always slope down and supply curves don’t always slope up. It’s known as the Law of Traffic Congestion (sometimes “Down’s Law”, sometimes “The Fundamental Law”). When you attempt to reduce the cost of travel on congested highways by building more lanes, demand goes up faster than capacity so that congestion increases. A first order theory can’t account for this kind of induced demand.
Or as Yogi Berra is supposed to have said, “that place is so popular nobody goes there any more”.
23. May 2016 at 14:56
dtoh, You said:
“Top cap gains rate increase by 60% not 8.8%”
No, with taxes what matters is the percentage point change, not the percentage change. Do you think if cap gains taxes rose from 0.01% to 0.1% (a 1000% increase using your math) it would matter very much?
My diminishing returns comment referred to the claims of a long run effect.
GmcK, You said:
“When you attempt to reduce the cost of travel on congested highways by building more lanes, demand goes up faster than capacity so that congestion increases.”
I don’t agree.
23. May 2016 at 15:32
@scott
I’ve explained this before, but when you look at the cap gains tax rate, what is really important is the effective expected after tax rate of return on investment. When you have asymmetric returns, e.g. 1 winner and 9 losers (as you do with new business formation,) then small changes in the tax rate will have hugely amplified effects on the expected after tax return.
The only way this doesn’t have a huge effect on investment and growth is if you are rational expectations denier. Do the numbers. This is just simple arithmetic.
As for long term diminishing returns on capital, what are we talking? A millennium? If so, you may be right.