Economists look foolish by not seeing what Central Bankers see… small output gap
Antonio Fatas writes that the central bankers are estimating smaller output gaps in Europe, the UK and the US. I have been saying for a year that potential GDP is far less than the CBO estimate. But other economists have been sticking to the claim that the output gap is still quite large. Economists do not understand effective demand yet. We see this reflected in what Mr. Fatas writes…
“Measuring potential output or the slack in the economy has always been challenging. One can rely on models that capture the factors that drive potential output (such as the capital stock or productivity or demographics) or one can look at more specific indicators of idle capacity, such as capacity utilization or unemployment rate.”
Do you see any demand factors for determining potential output in his quote? No. So if productive capacity stays the same, so should potential output, right? Well, wrong… if demand is undermined, for example by a 6% fall in labor share of national income/output, the economy will not be able to sell its potential output. There is a demand constraint.
Keynes wrote about effective demand, which when weak, can keep the economy from reaching potential output. We are seeing this now. Well, at least, I am seeing this happen. Other economists have not caught on… yet… But it is inevitable that they will.
Here is a model that plots aggregate supply and effective demand (US economy). It is similar to the model of aggregate supply and aggregate demand, but effective demand is used to show the effective demand limit upon output.
The blue dots show real GDP increasing to the right over the past 8 quarters.The up-sloping lines are aggregate supply for real GDP for the given inflation rate. The down-sloping lines are effective demand. The down-sloping lines of the effective demand limit have stayed within a fairly tight band. Effective demand has fallen into this band 6 of the past 8 quarters.
There are two lines that fell out of that band. One was the labor income surge at the end of 2012 before the payroll tax changes of 2013. The other was the capital income surge at the end of 2013. Apparently businesses are competing to show better profits or simply protecting themselves as the minimum wage movement grows in strength. The effective demand limit corrected back to the band after the labor income surge of 2012. I expect the effective demand limit to correct back to the band in 2014.
The effective demand band is pointing to a meeting place with the increasing real GDP, as shown by the red zone. Real GDP in the US will reach the red zone of the effective demand limit between $16.0 and $16.1 trillion. The red zone is the actual potential GDP as constrained by effective demand. This model of effective demand is showing what the central bankers are seeing… potential output is much less than the CBO estimate above $16.7 trillion.
The tightness of the band is an important character-defining aspect of effective demand. After recessions, effective demand falls into a band that eventually points to the end of a business cycle expansion (red zone). Because of this consistency of effective demand, we are able to foresee the end of an expansion a couple of years in advance. Let that sink in for a moment.
What the central bankers are now seeing could have been seen 2 years ago, if they had this model. As it is, economists have no model for effective demand. They will blame the stunted potential GDP on the recession itself, and not on weak demand. As Mr. Fatas writes…
“business cycles can leave permanent (or at least very persistent) scars on output through the effects they have on the capital stock or the labor force. But it is important to understand that the permanent effects are the consequence of the recession itself. If we could manage to reduce the length and depth of the recessions we would be minimizing those permanent effects. And in that sense, accepting these changes as structural and unavoidable is too pessimistic, leads to inaction and just makes matters worse. If you read the evidence properly, you want to do the opposite, you want to be even more aggressive to avoid what it looks at a much bigger cost of recessions.
There is no word about raising wages and labor share in his prescription to overcome a lower potential output. By not understanding properly how demand constrains output, economists are looking foolish in the eyes of businesses and central bankers. The bankers see the numbers from actual businesses. Economists look at past data for production without incorporating a measure of demand. Economists are disconnected from reality.
What does it mean that the economy is closer to potential than even the Federal reserve projects in their forward guidance? For investors, it means that profit potential going forward is much less than they think on average. For the general population, the error in judgement creates more economic instability and thus more risk of a harsher contraction. For economists, it means that they have an opportunity to learn a lesson about effective demand. Yes, Keynes wrote about it back in the 1930’s, but economists have come to equate it with aggregate demand. A mistake that can be corrected.
There’s been some destruction of potential output, e.g. early forced retirements, easier to go on disability, unemployment benefit extensions, and other social programs.
Yet, there are still over 20 million Americans unemployed or underemployed. Job growth has increased a little faster than population growth and replaced some of the lost jobs from the recession. However, we haven’t even dug out of the last recession yet. We’re in a 21st century deep depression. Dollar hamburgers and dollar stores are the new soup lines.
Your mistaking “job growth” from all eras. Job growth before the 60’s wasn’t overly impressive from that standpoint, even in good years of growth. The Carter and Reagan era, when the Boomers were at their peak in some respects, didn’t create enough jobs to drive unemployment below 5.0% despite the huge numbers.
The fact is, the economy only needs to grow 2.5% to reach trend. That means if the economy grows 3.7% in 2013(considering the long term revisions, the current number is irrelevant), it had a pretty boastful year. Overall average job growth before final revisions is 194 just on NFP alone in 2013. That would be like 250,000 when Carter was Pres or 230,000 when Clinton was Pres.
The output gap just isn’t smaller, the jobs gap is as well to produce a clearing of labor markets and start producing inflation pressures in some markets. This is what I have been hammering for 2 years now. We are a aging society that will get worse over the next 10 years before is starts getting better.
John C,
You make a good point about the jobs gap…
Also like you say, the economy only needs 2.5% growth to reach trend, but some central bankers are seeing only 1% to 1.5% spare capacity left to reach potential. So growth would be stunted before it reaches trend.
The central bankers are concerned because monetary policy is still very aggressive going into a lower potential.
It seems most economists recommend pushing on the gas pedal even harder, but potential cannot be raised a la Say’s Law like that. That is what Keynes tried to say in Chapter 3.
John, when the country is at or near full employment, or after a mild recession, job growth doesn’t have to be strong to reach full employment (also, full employment is 5% unemployment rate, because of frictional unemployment).
Given high household debt, millions of Americans want to work harder and longer, which adds to future growth. However, either Americans are being given compensation by the government for not working or cannot find decent jobs (i.e. have higher reservation wages, and when you lose a good job, in a severe recession, you may only be able to find a much lower paying job).
In the Reagan recovery, which was a strong recovery after a severe recession, some months created over one million jobs, and that was when the labor force was much smaller than today.
The following chart shows the huge output gap over three generations, since 1960:
http://www.advisorperspectives.com/dshort/charts/indicators/GDP-per-capita-overview.html?Real-GDP-per-capita-since-1960-log.gif
The huge downshift in real growth happened suddenly rather than slowly.
And, from from the UCLA Anderson forecast last year:
“U.S. real GDP is now 15.4 percent below the normal 3 percent trend. To get back to that 3 percent trend, we would need 4 percent growth for 15 years, 5 percent growth for eight years, or 6 percent growth for five years, not the disappointing twos and threes we have been racking up recently, which are moving us farther from trend, not closer to it. It’s not a recovery. It’s not even normal growth. It’s bad.”
Peak Trader,
When you say full employment is 5%, you are looking at past data. This is the problem with economists… They try to make the present situation fit past data. The central bankers are seeing something different. The situation has changed.
So economists look foolish by forcing past dynamics on current dynamics if full employment turns out to be over 6%.
When you say 5% is full employment, you have to show that the current dynamics are pointing to that. I show that they aren’t.
Edward, you’d have to be a poor economist just to look at past data and ignore the trends in the data.
When more jobs are created, the labor force participation rate will rise and the unemployment rate will rise, because many discouraged Americans out of the labor force will come back in.
And, there will always be frictional unemployment. So, full employment will always be around the 5% unemployment rate.
What’s sad is Americans were willing to work harder and longer, to pay-down high levels of household debt and postpone retirements, if they had the opportunity, which would add to future economic growth.
What has changed over the past few years is some destruction of potential output. The output gap can be closed either by raising actual output or destroying potential output.
Demographics, or the aging of the Baby-Boomers, destroys some potential output. However, the destruction was sharp and sudden, over just a few years.
I suspect, those Baby-Boomers, who live longer than their parents, want to work longer to pay for the overconsumption (low prices and interest rates induced demand), and strengthen their (household) balance sheets heading into retirement.
The Baby-Boomers who kept their (good) jobs in the recession are working longer. However, many Baby-Boomers were forced out of the labor force, one way or another, while many young workers are waiting for older workers to retire for their jobs.
If we were using something like the “finance-neutral” output gap calculation , as recommended by Borio at the IMF , we wouldn’t be suffering these delusions of excess capacity. We’re so used to having worker incomes supplemented by compounding levels of debt that we don’t know how to act now that we’ve drained that particular well.
We either need a fresh supply of NINJA borrowers , or we have to pay our workers more. Otherwise , it’s “secstag” as far as the eye can see.
I’m not a big fan of Robert Gordan , but I’m afraid he’s going to be closer to the mark here than all the polyannas will be , absent big changes in the way we’ve structured our economy :
http://www.nber.org/papers/w19895#fromrss
“The United States achieved a 2.0 percent average annual growth rate of real GDP per capita between 1891 and 2007. This paper predicts that growth in the 25 to 40 years after 2007 will be much slower, particularly for the great majority of the population. Future growth will be 1.3 percent per annum for labor productivity in the total economy, 0.9 percent for output per capita, 0.4 percent for real income per capita of the bottom 99 percent of the income distribution, and 0.2 percent for the real disposable income of that group…..”
Edward ,
I’ll post this with the hope that you might find it interesting , since you’re into economic puzzles , graphing and such :
Like many other countries , we’ve gotten less gdp bang for our debt buck over time. Before 1980 , we got almost $.80 per dollar of debt ( using domestic nonfinancial , or TCMDODNS at FRED ) , while now we only get about $.40 per dollar. I wondered how much extra gdp we’d have if we had used the same amount of debt , but got the old marginal return on it. Secondly , I wondered if some of the excess has flowed into asset values , as speculation / financialization has risen over the same period.
Look at the blue line in this first graph. It shows gdp/tcmdodns , or dollars of gdp per dollar of debt , over time ( left-hand scale ):
http://research.stlouisfed.org/fred2/graph/?graph_id=161216&category_id=0#
The red line , on the right scale, shows dollars of net worth per dollar of debt over time. The figures on the right and left scales are effectively the “multipliers” for gdp or net worth per additional dollar of debt.
Now , what I did was netted-out the gdp and net-worth pre-1975-1980 or so , by subtracting the appropriate amount from the multipliers( set so as not to generate a negative value , but as close to zero as possible ) , then applied those new multipliers to the debt that came on board over time , to see how much gdp and net worth resulted , which is shown in this graph ( blue is the “missing” gdp and red is the added net worth ) :
http://research.stlouisfed.org/fred2/graph/?graph_id=161217&category_id=9489#
I was amazed to see both calculations come up so close to each other. In fact , if not for the bubble nature of the net worth plot ( dot-com , RE bubble , and now ) , they might match almost exactly. If you go to the transformation plot for both lines and set them to “natural log” and replot , you can see that the match is pretty impressive.
I’ve done this with TCMDO , which includes financial debt , and added in business net worth , and get similar results.
Keen’s formulation is now stated as GDP at time t equals GDP at time t-1 plus M2 velocity times delta debt. I think he’s screwing up , because the revealed velocity for debt “money” is as shown in the first graph , ranging from .4-.8 over the last 60 years or so. M2 velocity has ranged from 1-2 or thereabouts , so I don’t see how he makes that work
Keen seems to have dropped the catch-all he once had at the end of his formula , standing for “asset turnover” or some such. Based on what I’m seeing , if he brings the extra term back as “added net worth” , and uses the correct debt multiplier , he might find out that things add up to the nickle ! Ha!
Here’s a link to the Keen paper , if you haven’t seen it yet :
I love the graph. it reminds me of the last visit to the Met, where several similar pieces were on display.
http://www.drewnussbaum.com/wp-content/uploads/2013/06/linear-abstract-on-teal-wall.jpg
http://www.artmajeur.com/en/artist/experimentalstudio/collection/abstract-and-contemporary-fine-art/1455910/artwork/linear-composition-number-1/6698059
Jack,
A picture is worth a thousand interpretations… : )
I like the linear abstract on teal wall.
I messed up the first graph , above. It should be this :
https://research.stlouisfed.org/fred2/graph/?graph_id=161299&category_id=9489#
The second one was OK , but probably does’t make sense with the messed up graph and messed up explanation.
My thinking , which still may be off , was that pre-1980 we got a max of .85c of gdp per dollar of debt. If that had been the return throughout , we would have got .85 x ~$41trillion ( of domestic nonfinancial debt ) , or ~$35 trillion of gdp. Our realized GDP was ~17 $trillion , so “missing” gdp was ~$18 trillion.
On the net worth side , I specified the “normal” level of net worrth to gdp at 3.17 x , and subtracted that ratio from the higher ratios that followed , to arrive at the multiplier to apply to gdp to generate the “excess” net worth. This worked out to be about in the same range , if you split the bubble valuation , maybe $18-20 trillion. Adding in the 3.17 multiple of current ngdp , ~$54 trillion , you get a total of ~$72-74 trillion , not far off the current HH net worth of ~$77 trillion.
If this is right , it’s intriguing. It would mean you could write Keen’s formula as GDP(t) = GDP(t-1) +velocity x delta debt (t) , and then also Net Worth(t) = net worth (t-1) + [delta debt(t) – (velocity x delta debt(t)) ]. In other words , all the debt would be accounted for as either GDP or net worth.
If this is right , I’m going to have to start believing in the EMH , and I don’t want to do that , so I hope someone will pick this apart. 😉
Ed,
The problem for me is that your chart and the two linear abstractions look little different to the lay (not statistically sophisticated, that is) public. Economic data needs to be understood in order to be appreciated. Just my opinion, of course, but when the discussion becomes a maze of formula and abstract lines I think tha the average reader of sites like AB begins to lose contact with the meaning of the message. I’m not expecting excessive simplicity, but a statistical barrage can be, and is often, used to obscure the meaning of the data. We can then only rely on our opinion of the discussant to be able to trust their presentation of the meaning of the facts when those facts are a jumble of esoteric numbers and algorithms.
Marko, it should be noted, since 1980, U.S. living standards improved much faster than U.S. GDP growth. For example, U.S. trade deficits became increasingly larger from 1980 to 2006, reaching $800 billion a year (or 6% of U.S. GDP). I stated before:
In the Keynesian consumption function (or identity), Y = C + I + G + NX; trade deficits, or negative net exports (NX), subtract from GDP growth, because consumption (C) is overstated.
For example, in a $10 trillion economy, a 5% one year increase in GDP expands the economy by $500 billion. Also, if imports are greater than exports, e.g. $1 trillion of imports and $600 billion of exports, then net imports are $400 billion in one year, which adds to our trading partners GDPs and subtracts from U.S. GDP. However, the total added to the U.S. economy is $900 billion ($500 billion from GDP growth in one year and $400 billion from net imports in one year), because the U.S. is consuming more than producing in the global economy.
It should be noted, with trade deficits, the U.S. exchanges dollars for foreign goods and foreigners exchange those dollars for U.S. Treasury bonds. However, the U.S. is able to maintain trade deficits in the long-run, mostly, because foreigners lose through changes or differences in inflation, interest rates, and currency exchange rates. For example, over the past few decades, the Japanese yen appreciated from 360 to 100 per dollar, which means Japan received fewer and fewer yen per dollar.
Also, to a lesser extent, many of those foreigners moved to the U.S., and foreign firms began producing in the U.S., because it’s better to use those dollars than exchange them for their currencies. Moreover, those foreigners can raise their standards of living by selling their assets, exchanging their currencies for dollars, and moving to the U.S., or attending a U.S. college, for example. Furthermore, tourism (by foreigners in the U.S.) adds to U.S. exports.
PT ,
And Edward – disregard all that crap above. I think I simply devised a creative way to get FRED to tell me that 1+1= 2.
My brain is fried. I’m going to bed.
Also, domestic growth cannot expand beyond potential output for a long period of time. However, the U.S. had one of the greatest eras of prosperity at the height of the Information Revolution, from 1982-07, because the U.S. economy moved much closer towards optimization, producing and consuming more output, in the global economy, with fewer inputs. The only way to move from one economic revolution into the next is through efficiencies.
Since the U.S. became more of an open economy around 1980, it became a Black Hole in the global economy, attracting imports and capital, and even attracting the owners of that capital themselves. Unfortunately, the U.S. federal government became a Black Hole in the U.S. economy, spending and squandering too much and building-up massive debt rather than refunding enough dollars for U.S. consumers in the form of tax cuts to continue the virtuous U.S. cycle of consumption-investment.
The virtuous U.S. cycle of consumption-investment would’ve eventually slowed anyway, because of diminished marginal utility (or the opposite of pent-up demand), along with the aging of the Baby-Boomers. However, tax cuts would’ve lessened “deleveraging” and strengthened household balance sheets, which would’ve resulted in a milder recession or stronger recovery.
From an email to a friend last May.
——————————————————– Start ——————————————————–
Total Consumer Debt at the end of 2012 was shown as $11.34Trillion.
Note: http://data.newyorkfed.org/research/national_economy/householdcredit/DistrictReport_Q42012.pdf
The peak Total Consumer Debt was in the 3rd quarter of 2008 at $12.68Trillion. So Total Consumer Debt is currently down by $1.34Trillion from the peak.
That left me curious. What was it earlier? So I found their historical data at the link “Pre2003 Data X Excel” on this page: http://data.newyorkfed.org/householdcredit/historical-reports.html
Total Consumer Debt at the end of 1999 was shown as $4.76Trillion. Just for reference, from 2000 to 2010 the population increased by 9.7% according to the US Census Bureau. So there were not that many more people to run up more debt.
Great Caesar’s Ghost!!! At this rate it would take about 23 years from the peak until Total Consumer Debt gets back to 1999 levels. Surely not, they must be depending on inflation to do some of the heavy lifting. The truth is that consumers are in hock up to their eyeballs and that will continue, since their wages are still not keeping up with inflation.
——————————————————– End ——————————————————–
Total Consumer Debt at the end of the 2013 was shown as $11.52Trillion.
I can not even imagine a tax cut that would change this dynamic, let alone getting it enacted into law.
Labor share will have to be dramatically increased, or huge amounts of consumer debt will have to be written off. Or we can limp along with low Effective Demand for a decade or two.