The Keynesian critique of British fiscal policy
Back in 2011 I did a post entitled “The job-filled non-recovery.” Now in 2013 The Economist has an article entitled:
The Job-Rich Depression
BRITAIN’S economy has had an odd five years. In output terms, things have been terrible. The slump that started in 2008 is far worse than the 1930s depression; only the years after the first world war were harsher. Consumption has been dragged down by weak real wage growth, investment has been held back by tight credit and exporters have struggled with weak demand in the euro zone. The initial estimate of GDP growth in the fourth quarter of 2012, due shortly after The Economist went to press, was expected to contain more bad news.
. . .
Yet the job market is humming. Data released on January 23rd show that employment has topped previous peaks (see first chart). The combination of economic slowdown and plentiful jobs means output per worker has fallen 12% further than at the same stage in previous recessions. That is equivalent to the loss of the entire manufacturing sector. Britain is now startlingly unproductive compared with other rich countries. What is going on?
. . .
Some point to Britain’s growing army of part-timers, who account for a third of the 1.3m net new private-sector jobs created since 2010. Interns and other unpaid workers are classified as employed but may produce little output while learning their trades.
Still, neither answer solves the puzzle. Average hours worked have increased even as part-time jobs have become more common. And the 275,000 or so unpaid workers are a tiny fraction of Britain’s 30m- strong workforce. Britons really are producing less per hour worked. It is not the data that are odd. It is the British economy.
What can we make of this? Keynesians feel the problem is too little aggregate demand, and point to the very low RGDP growth rates. But RGDP is not the right way to measure AD, so this creates some problems for the Keynesian model. Old Keynesians tended to focus on the unemployment rate, but the British labor market is no worse than most other developed countries (US employment is far from the peak.) So where is the evidence that austerity is a problem?
New Keynesians use inflation as their AD indicator, but that’s run way over target for the past 5 years. So by that measure there is no AD problem at all.
I think the Keynesians are partly right, the UK does have an AD problem. I say ‘partly right’ because I suspect they also have big AS problems, which the Keynesians tend to underestimate. But I’d like to focus on where I agree—they have an AD problem.
Now I’d like to suggest an AD indicator that is far superior to either jobs or RGDP—you guessed it, NGDP. The growth rate of British NGDP has recently been quite low, and that’s keeping the unemployment rate above the natural rate.
So I’d encourage Keynesians to focus on NGDP as a policy indicator. It’s the variable that will best describe the AD problems they worry about. If you focus on RGDP, that raises the question of why so many extra workers haven’t created any extra output. New Classical skeptics will say there must be a productivity problem, and the Keynesians won’t have a good answer. After all, in the Keynesian model more AD leads to more NGDP—how that gets partitioned between RGDP and P depends on the slope of the SRAS curve, which demand-side policymakers cannot control.
In contrast, the NGDP “musical chairs” approach says that a sharp slowdown in NGDP growth will almost always cause excess unemployment, regardless of the supply-side of the economy.
PS. I won’t have time to answer comments over the next few days.
Tags:
31. January 2013 at 07:14
Yes, exactly.
I wonder if some of the jobs/output divergence is explicable by a failure to measure inflation properly under such tight NGDP growth; i.e the deflator is overstated and RGDP should be a little higher. There are lots of examples of this: widespread discounting, substitution to lower priced vendors; budget airlines, hotels, retailers are all booming. I don’t trust the CPI measurement to capture that at all. But it’s all anecdote with no data, and five years is a long time for excuses.
31. January 2013 at 07:19
“In contrast, the NGDP “musical chairs” approach says that a sharp slowdown in NGDP growth will almost always cause excess unemployment, regardless of the supply-side of the economy.”
So true…
http://thefaintofheart.wordpress.com/2011/12/03/the-hole-theory-of-employment/
31. January 2013 at 08:02
Real wages in the UK are significantly below 2007 levels. Leaving aside NGDP/AD considerations I think that lower real wages will encourage hiring into lower productivity jobs and could explain some part of the employment/productivity “puzzle” in the UK. In addition I think that Britmouse’s analysis above is also worthy of further investigation.
Real wages, index 1989-2012
Annual average
2007 119.6
2008 115.2
2009 106.8
2010 106.1
2011 108.9
2012 111.7
31. January 2013 at 08:05
Please ignore my last comment – just realized that my data source was invalid.
31. January 2013 at 08:41
I think that much of the unemployment in the UK and even in the US is structural. I think around 50% of businesses are still having difficulty finding jobs and filling vacancies. The problem is that there is a shortage of skilled labor. I also think that there is an AD problem as well.
Before 2007-2008, we had a massive debt fueled investment boom. If you tread AD as not just NGDP or I, but add on dD/dt and NAT and now you have an easy demand side explanation. Whether or not AD=I+dD/dt=NGDP+NAT is a different issue; however, using that view for AD can give you a very different explanation. When you have debt increasing at 10% GDP and have that drop to -7%, that would definitely explain the AD problem.
31. January 2013 at 10:38
“Now I’d like to suggest an AD indicator that is far superior to either jobs or RGDP””you guessed it, NGDP.”
But yesterday you said that GDI was better indicator than GDP, so which is it.
NGDP as a stand in for AD — what about involuntary inventory accumulation — it is NGDP that is in excess of actual demand.
31. January 2013 at 10:42
Regarding Great Britton, my hunch says it is that the “definancialization” of the British economy has something to do with it. At one time there was a small group of very highly paid poeple who created new financial products. While some people would say that these people didn’t produce a thing, the national income accounting metrics said that these were the most productive people on earth. If the “city boys” have been pushed out of finance that would show as a drop in productivity.
31. January 2013 at 12:32
Doug M.,
NGDP can be superior to employment data or RGDP yet still inferior to NGDI.
31. January 2013 at 13:22
PS. I won’t have time to answer comments over the next few days.
That’s a pity, because I wanted your views on this.
31. January 2013 at 14:41
Hi Scott,
I’m curious to see how Keynesian policies stack up throughout history when using NGDP growth as our gauge for success?
Thanks.
1. February 2013 at 00:32
I assume this will interest readers here:
http://www.economist.com/news/leaders/21571138-how-bank-englands-new-governor-and-chancellor-should-stimulate-british
1. February 2013 at 05:36
If only inventories of final goods are considered, there is an advantage to ignoring them and only using final sales. There is also a disadvantage.
The advantage is that unplanned inventory investment reflects a mistake. Planned expenditures are too low or too high.
However, the disadvantage of leaving off inventories of final goods is that planned inventory investment is ignored. And that includes corrected for past unplanned inventory investment.
More importantly, however, is that inventory investment also includes inventories of intermediate products. It is only inventory investment that includes changes in the demand for materials and partially finished goods.
Leaving off inventories would be a serious mistake in my view. I think leaving off changes in the demand for commodities (changes in price and quantity) would be a serious mistake.
1. February 2013 at 05:58
Since the inflation measures are wrong the RDGP measures are wrong. That explains everything. Especially the PPP puzzle, where the UK is way above its previous peaks GDP in price purchasing power terms. I suspect the RGDP growth rate is actually in the upper end of 1-2%, in which case its consistent with jobs, there is no massive fall in productivity.
Also, it dovetails neatly with my previous thoughts that when the trimmed mean is much less than the mean you should prefer the trimmed mean as a measure of inflation, which takes about two percent off the CPI.
1. February 2013 at 07:53
Phil, I just read your blog. Fascinating data on the CPI, great stuff.
1. February 2013 at 09:10
Suvy,
I seem to remember seeing this before. I’ve noticed other Keen channeling in monetarist blogs recently. Could it be that people are beginning to understand that his effective demand equation isn’t actually in conflict with the usual identity for Y? After all, new money has to enter the economy somewhere.
He has posted two videos of interest about this:
http://www.youtube.com/watch?v=pOxZixjorho
http://www.youtube.com/watch?v=UzxQcTOs4JA&list=PLqs7-zw9kiAJu9rsKB35f3cZWg6aq24V-
I wish people would get with this stuff, because the real mindbending problem is to reconcile it with stuff like:
http://ftalphaville.ft.com/2013/01/02/1319583/on-the-new-purpose-of-government-debt/
There’s no reason they can’t both be right, but I’m not sure what the consequences would be. It would seem to sharpen the horns of the Triffin dilemma considerably.
1. February 2013 at 09:44
Doug M
31. January 2013 at 10:42
“Regarding Great Britton, my hunch says it is that the “definancialization” of the British economy has something to do with it. At one time there was a small group of very highly paid poeple who created new financial products. While some people would say that these people didn’t produce a thing, the national income accounting metrics said that these were the most productive people on earth. If the “city boys” have been pushed out of finance that would show as a drop in productivity. ”
That is often put forward as an explanation. However, the financial sectors contribution to productivity growth is way smaller than people generally believe.
http://www.edmundconway.com/wp-content/uploads/2013/01/financeproductivity.jpg
http://www.edmundconway.com/2013/01/stop-press-were-not-doomed-yet/
The apparent violation of Okun’s Law is not new. The puzzle has just become a big talking point over the last six months, but JP Morgan analysts were pointing it out as long ago as 2009, they reckoned GDP was being inaccurately measured.
1. February 2013 at 09:48
Indermit Gill and Naotaka Sugawara have some good observations on European productivity and employment;
http://www.voxeu.org/article/central-european-reforms-beacon
‘What might not be known widely is that since a few years ago, “Doing Business” rankings no longer include measures of the ease of hiring workers. Of course, this is an important part of the investment climate.
‘Using measures of the ease of hiring and firing workers makes the investment climate look better for countries that already do well in the Doing Business rankings – such as Denmark, Finland, Estonia, Latvia and Lithuania – and makes the investment climate even worse for Italy, Portugal and Spain. Figure 7 shows which parts of Europe have made it easy to hire workers, and where it is cumbersome. The Swiss and other EFTA members do the best, the Spaniards and the south the worst. In the EU, the best performers are the three Baltic economies. It is perhaps not a surprise that their labour productivity growth measures have been the best.’
1. February 2013 at 09:52
Marcus,
Read your “Hole Theory” post with interest. How you define the labor force is important. I saw a study once comparing Japanese unemployment with U.S. unemployment. The official Japanese unemployment survey asks if you have looked for work in the last week. The U.S. asks if you have looked in the last 6 weeks. According to the study, if you use the U.S. methodology in Japan, Japanese male unemployment increases by 60% and female unemployment doubles.
1. February 2013 at 13:56
Scott
The Economist has endorsed NGDPT!
http://thefaintofheart.wordpress.com/2013/02/01/the-economist-in-favor-of-ngdp-targeting/
1. February 2013 at 21:09
Alas, Marcus, The Economist justifies NGDPT by mistakenly believing that it will necessarily keep interest rates low (“…By promising to keep monetary conditions loose until nominal GDP has risen by 10%, the Bank would provide certainty that interest rates will stay low…”).
2. February 2013 at 08:18
Britmouse, That’s possible, although NGDP growth is also rather low.
Suvy, Yes, both problems exist.
Doug, You are mixing up two separate issues; which is the best variable, and which is the best way of measuring NGDP. NGDP is the best variable, and NGDI income is the best way of measuring NGDP. In theory, both NGDP and NGDI are identical, hence in pointing to NGDI I am in no way moving away from NGDP, just suggesting a more accurate way of measuring it.
Bababooey, I’m afraid I don’t have anything enlightening to say on that issue.
PrasanthK, You’d have to be more specific. Which Keynesian policies–old or new?
Vivian, I have a post.
Bill, If we are going to leave anything out, I’d consider non-labor income. The labor market is the problem. But I see big political problems in trying to target labor income and letting profits soar–it would be misunderstood.
Patrick, Thanks for that info, it’s an issue people should pay more attention to.
dtoh, I find that very plausible.
Marcus, I don’t think the Economist actually endorsed NGDP targeting. See my new post.
gofx, Good point.
2. February 2013 at 12:04
‘…it’s an issue people should pay more attention to.’
Yes, people like Barack Obama, whose health care legislation made it more difficult to hire people.
4. February 2013 at 00:16
I wonder if NGDP growth brought about by central banks can cause future NGDP slowdowns brought about by the market.
4. February 2013 at 06:33
Geoff, Not unless the central bank allows it.
4. February 2013 at 18:41
Dr. Sumner:
“Geoff, Not unless the central bank allows it.”
This doesn’t really address my concern. I was wondering if what the central bank “won’t allow” is itself caused by prior central bank activity.
I’ll use an analogy to elaborate on what I mean:
“I wonder if taking medication X to help with symptom Y can cause future symptom Y which might lead a doctor to believe that the patient needs further doses of medication X.”
response:
“Not if the doctor won’t allow the patient to experience those side effects.”
5. February 2013 at 07:40
Geoff, I don’t follow that analogy. If monetary stimulus causes more NGDP today, that won’t cause less NGDP in the future.
5. February 2013 at 11:43
Dr. Sumner:
“Geoff, I don’t follow that analogy. If monetary stimulus causes more NGDP today, that won’t cause less NGDP in the future.”
The analogy is that monetary stimulus is the medicine, and the market is the body.
I was thinking about the effects of higher or lower current NGDP on the market, and then the market having a decreasing or increasing effect on future NGDP later on.
The motivation for my concerns here is that I don’t think that the only significant effect of higher or lower NGDP is on nominal statistics.
In other words, I am basically asking if higher NGDP today might have a particular effect on the market in some way, apart from prices and other nominal statistics, after which the market then has a decreasing effect on NGDP in the future (a heretofore non-considered side effect) without the Fed’s intentions.