If it’s an identity does that mean I’m right?

My musical chairs model of the economy assumes nominal hourly wages are sticky. In that case fluctuations in NGDP may be highly correlated with changes in the unemployment rate. Arnold Kling doesn’t like the empirical evidence I found in support of the model:

Scott is fond of saying, “Never reason from a price change.” I say, “Never draw a behavioral inference from an identity.”

I think Arnold knows it’s not an identity. If you graphed MV and PY, you’d observe only one line, as the two lines would perfectly overlap. In the graph I presented the two lines were correlated but far from perfectly correlated. So it’s no identity. I’d guess the gaps would be far larger in many other countries, such as Zimbabwe.

A better argument would be that the correlation doesn’t prove that causation goes from NGDP to unemployment. After all, changes in NGDP could cause changes in hourly nominal wage rates, leaving unemployment roughly unchanged. In that case the correlation I found would be spurious. If I left the impression that the correlation proved nominal wages were sticky that would have been a “behavioral inference,” and hence a mistake. What I tried to do was assume nominal wages are sticky (as they obviously are), and then show the effect of NGDP shocks in a world where nominal wages are sticky.

My musical chair model doesn’t have the “microfoundations” that have been fashionable since the 1980s. But which model with microfoundations can outperform the musical chairs model? Indeed let’s make the claim more policy-oriented. I claim that fluctuations in predicted future NGDP relative to the trend line strongly correlate with changes in future unemployment. And of course NGDP futures prices are 100% controllable by the monetary authority.

There is no NGDP futures market? You can’t use that against my claim as I’ve been advocating such a market since 1986. It’s an embarrassment to the economics profession that this market doesn’t exist (yet.)

So yes, it sort of seems like a tautology, but perhaps that’s because I’m right.

PS. Imagine one of the great controversies in physics (say string theory) could be settled with a particle accelerator that cost $500,000 to build. But the physics profession was too lazy to ask the NSF for the money. Replace string theory with the debate over demand-side vs. RBC models and you have described economics circa 2014.

Tags:

This entry was posted on February 09th, 2014 and is filed under Misc., Monetary Theory. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response or Trackback from your own site.

63 Responses to “If it’s an identity does that mean I’m right?”

TD
9. February 2014 at 19:05

What is your estimate of the costs of setting up an adequate NGDP futures market?
Benjamin Cole
9. February 2014 at 19:38

Wages are not sticky?

I ask any economist to truthfully answer this question:

1. Inflation is running at seven percent, and your university, citing budget constraints, manages to come up with a four percent raise. They say they tried and managed to come up with the funds.

2. Inflation is flat, but citing budget constraints, your university cuts your wage by two percent. The said they tried, but could not come up with money for your salary.

It is simply impossible not to regard the second scenario as either one in which your employer lacks appreciation for your services and all your hard work, or the employer is outright being punitive, and maybe tenure is not coming….

And that’s if you have a PhD in economics…

Employers know they cannot cut nominal wages. It crushes morale, except if they are extensive extenuating circumstances, and management and ownership takes cuts along with everyone else….
Major_Freedom
9. February 2014 at 20:29

A better argument would be that the correlation doesn’t prove that causation goes from NGDP to unemployment. After all, changes in NGDP could cause changes in hourly nominal wage rates, leaving unemployment roughly unchanged.

Don’t forget the other argument consistent with the correlation: That changes in wages paid causes changes in NGDP, ceteris paribus, and, as a corollary, if the Fed inflated to raise NGDP back up, that wages would not necessarily rise.

On a side note, the sticky wage doctrine should not be taken for granted. One should analyze the cause(s) for wages not immediately adjusting to changes in nominal demand for labor such that the labor market clears. It might be good to start with imagining how sticky wages would be if we lived in a society where production of goods outpaced the production of money, such that there was “healthy” price deflation. Would employers and wage earners be more or less willing to take pay cuts? Would pay cuts be less or more psychologically acceptable?

——————

My musical chair model doesn’t have the “microfoundations” that have been fashionable since the 1980s. But which model with microfoundations can outperform the musical chairs model?

ABCT. It explains why so many investors would, at the same time, make forecasting errors in revenues (and expected cash holding preference).

In using the musical chairs model on the other hand, one is forced to believe that errors in revenue and cash holding forecasting are not errors in revenue and cash holding forecasting after all, but rather an error of the central bank in issuing insufficient currency such that enough individual firm revenue forecasts are incorrect, or at least overestimated ex post.

The musical chairs model also does not explain why it is that in every recession, the more capital intensive industries suffer relatively more than the less capital intensive industries. Or, another way of stating the same question, the musical chairs model does not explain why it is that during the boom, too many resources and labor were allocated to the more capital intensive industries (and thus too few resources and labor allocated to the less capital intensive industries).

ABCT argues that it is not NGDP coupled with sticky wages that is causing unemployment, but rather it is the cause of NGDP suddenly falling so much despite the central bank not destroying dollars by decree, that causes unemployment. The musical chairs model leads one to believe that the cause for the sudden plunge in NGDP can be eliminated if inflation is increased so that NGDP is prevented from falling. This of course is still not an explanation for why more inflation was needed to prevent NGDP from falling. It only addresses the symptom.

In general, MM theory is a theory of treating symptoms, not the cause for the symptoms.
Conor
9. February 2014 at 21:01

Are you saying that nominal wages are sticky across the board, or just for workers paid by the hour?
Major_Freedom
9. February 2014 at 21:02

Benjamin Cole:

I ask any economist to truthfully answer this question:

1. Inflation is running at seven percent, and your university, citing budget constraints, manages to come up with a four percent raise. They say they tried and managed to come up with the funds.

2. Inflation is flat, but citing budget constraints, your university cuts your wage by two percent. The said they tried, but could not come up with money for your salary.

Where is the question?

—————–

If inflation is running at 7%, that doesn’t mean a given university’s revenues are increasing by 7%. It could be growing revenues at a slower, or, more likely, at a much faster pace (since inflation from the central bank does not affect all firm/institution revenues equally).

If a firm/institution is expected to grow revenues at 7% annually, then its owners and potential capital providers would of course have a greater incentive to invest more funds, including most likely wage payments. But let that not lead to the notion that real investment would be higher.

Absent the anti-capitalist causes for wages not immediately adjusting downward in response to an expected decrease in future revenues, there is no reason to believe wage rates will remain above the clearing rate permanently, or for years on end as what occurs in existing conditions.

My view is that the more often incorrect cash holding forecasts are rescued through inflation, the more likely will incorrect forecasts be made, because people are not able to learn from their mistakes, since their mistakes are not being revealed.

———————–

I am pretty sure most people on this blog accept that it would be absurd to believe that increased canned food hoarding should be matched by an equivalent increase in canned food production, always and without question, such that canned food trading does not fall.

Yet this same absurdity applies no less to money, even if people were paid in canned food and canned food wages were sticky. This is because resources are scarce, and should be allocated according to individual subjective preferences constrained to profit and loss via private property production and trading, not artificial, politically forced canned food targets. If canned food trading were forced to grow at 5% annually, then it is possible for too many resources, or too few resources, to be devoted to everything required to produce canned food, vis a vis market preferences.

One of the most common mistakes I see being made when it comes to inflation, is underestimating, and I mean stupendously underestimating, the full resources needed to maintain central banking and inflation. Most people believe the costs of central banks inflating the money supply are the direct, “variable” costs, for example the pennies it takes to print a new dollar bill, or increasing the digits on a computer. In reality, the costs of inflation are the costs of maintaining whole governments, for without governments, central banks would be impossible.

So what is the cost of maintaining “our” government that makes central banking possible? We have to include the full costs of government, because if all the government did was tax and regulate people in order to impose a central bank only, then that government would likely be considered nothing but a criminal enterprise. So a government has to offer “goodies”. Those “goodies” must be included in a consideration of the costs of governments and central banks.

So, central banking effectively costs what the government spends per year, approximately $4 trillion, plus the non-spending opportunity costs incurred from killing, imprisoning and regulating people who would otherwise be producing wealth. Don’t have a number to attach to this latter factor, but another $6 trillion is not unreasonable, considering just how productive people could be in economic and political freedom.

With the above analysis, it is not unreasonable to argue that the dollar system is one of the biggest examples, if not the biggest example, of partial relative over-investment the world has ever seen. It is in such a bubble in fact, that lethal force is necessary to maintain it. In contrast, the housing and tech stock bubbles required artificially low interest rates and credit expansion. Nobody from the housing or tech stock markets threatened investors with lethal force if they didn’t go along. But with dollars, it is necessary, because of the incredible dichotomy between what investors and consumers can value in a free market, and what they are using/valuing in a central banking world.
ssumner
9. February 2014 at 21:33

TD, $500,000

Conor, Across the board.
Don Geddis
9. February 2014 at 21:36

MF: Such a delight to see your off-topic walls of text again, after your lengthy absence. I also note that you and your doppelganger Geoff, seem rarely able to rant simultaneously. Curious.

P.S. “…start by imagining…deflation” Let me guess. You can answer your own questions, without ever leaving your imagination. Yes? You have no need for empirical data. It seems to me that your epistemic process does not distinguish between “hypothesis” and “conclusion”.
Conor
9. February 2014 at 22:02

Scott, thanks for answering so quickly and at so late an hour. Could you (or anyone else) point me to evidence that nominal hourly wages are sticky for salaried workers. That strikes me as odd. I’m much more inclined to think that workers paid by the hour would be more unwilling to accept an hourly wage cut than that salaried workers would be unwilling to work more hours.
TD
9. February 2014 at 22:07

Scott, would you like me to help you set up a Kickstarter fundraiser for an NGDP futures market?
Don Geddis
9. February 2014 at 22:20

Conor: How do you imagine getting a salaried worker to work more hours? Isn’t the whole point of a non-hourly salary, that the position has a specific job to do, and the employer doesn’t keep precise track of exactly how many hours it takes to do the job? In a situation where corporate revenues fall, what strategy do you propose for getting more productive hours out of current salaried workers (while leaving their salaries unchanged)?
Conor
9. February 2014 at 22:57

Don Geddis, first off, I admit that my intuition is entirely based on anecdotal information. Everyone I know would be more than willing to put in extra saturdays or stay until 6 instead of 5, or take fewer breaks and sick days to avoid upsetting the boss during a recession (or lackluster recovery). I doubt if they would be so timid in the face of a nominal wage cut, even if the real effect is the same. I am probably completely wrong, but the idea that hourly wages are as sticky as wages paid to one who collects the same paycheck every week regardless of hours worked is hard for me to accept.
Conor
9. February 2014 at 23:01

Over the internet concessions always seem Colombo-esque to me: pretend to be unsure, so that you can demonstrate brilliance later. However, my ignorance is completely honest. I really do believe that I am wrong and confused.
c8to
9. February 2014 at 23:41

Conor – the whole thing is sticky. Its hard to instantly fire salaried (or any type of) worked. So as the demand for the firms output drops the salaried workers probably initially get a boost in effective per hour pay. A big project gets cancelled and they go home on time rather than working back.

In fact even in hourly based shift work I saw this happen. When the gfc hit the airlines were effected pretty badly. They didn’t fire anyone at my work or really change peoples base schedule. (Ok overtime could go down instantly so there is a wage drop at the margin) what they do is wait someone out who is leaving anyway and then don’t replace the headcount. This takes months.

So I see the whole firm, even the whole economy as sticky. NGDP shocks can move fast eg the financial contracts that blow up revealing I guess misallocation of resources or just plain triggering animal instincts (not sure what the main part is here) – anyhoo the ngdp can drop fast but the economy takes months to years to adapt.

Some of it may never adapt – I think maybe the shock can be the first tremor in a structural reconfiguration. Eg the reservation wages of the unemployed, discouraged workers, changes to benefits etc may all occur during the painful ngdp shock that screw up the real economy (or adjust it to some new equilibrium) for decades ahead.

Re the ngdp market – its probably not the setup cost rather subsidising it might be required by paying market makers.

The fed could do this easily by paying them directly or paying interest somehow (woolsey is probably the guy to ask about this)

Kickstarter could probably launch a private funded one (pay for setup) but I’d say liquidity would be a real problem without subsidy (at least initially I have to think more about long term)
Jim Glass
10. February 2014 at 00:11

Don Geddis, first off, I admit that my intuition is entirely based on anecdotal information. Everyone I know would be more than willing to put in extra saturdays or stay until 6 instead of 5, or take fewer breaks and sick days to avoid upsetting the boss…

I can speak to this from personal experience on both sides, as both employer and employee, supplementing the mountain of empirical evidence…

The first thing that happens when a firm reduces pay for work and makes working conditions worse (“from now on you have to come in on weekends”) is that it not only loses workers (by supply-and-demand: less pay, fewer workers will accept it) but it also loses its best workers.

Its best workers, who are in demand in the market and know it, decide correctly “I don’t have to take this” and scoot off directly to the competition — a very unhappy circumstance for that firm, as not only does it lose by suffering a reduction in the quality of its workforce but its competitors win directly at its expense.

The business school textbook example of this IBM, which suffered mass losses of it best employees, with only the poorer-quality ones staying on loyally, when it tightened pay company-wide while fighting for survival in the face of historic losses circa 1990.

But one doesn’t need a business school education to grasp this. Any small business owner knows that if he cuts the pay of his best key people they are going to bolt and he is going to be left operating with the dregs while faced with the big problem of replacing his best key people. When you are the owner in that situation you don’t even have to think about it, you feel it in your gut – and you are right.
Tom Brown
10. February 2014 at 00:58

Re: stick wages: David Glasner has and interesting post up on the subject:

http://uneasymoney.com/2014/02/06/why-are-wages-sticky/

This paragraph in particular caught my eye:

“All of these theories were powerfully challenged about eleven or twelve years ago by Truman Bewley in a book Why Wages Don’t Fall During a Recession. (See also Peter Howitt’s excellent review of Bewely’s book in the Journal of Economic Literature.) Bewley, though an accomplished theorist, simply went out and interviewed lots of business people, asking them to explain why they didn’t cut wages to their employees in recessions rather than lay off workers. Overwhelmingly, the responses Bewley received did not correspond to any of the standard theories of wage-stickiness. Instead, business people explained wage stickiness as necessary to avoid a collapse of morale among their employees. Layoffs also hurt morale, but the workers that are retained get over it, and those let go are no longer around to hurt the morale of those that stay.”
Tom Brown
10. February 2014 at 01:38

Mark Sadowski OT: I know you’re familiar with John Williams and his alternative ideas on inflation and RGDP:

http://www.shadowstats.com/alternate_data/gross-domestic-product-charts

You pointed out once that he hasn’t changed his subscription price in eight years.

I’m guessing you’re not a fan of his work. I’m not asking for a full debunking here, but what would be your #1 comment regarding his alternative real GDP figures? What would be your top couple of complaints? Do you suppose he starts with the same NGDP the gov does and then uses a different deflator? That would make sense (I think) since he doesn’t agree w/ gov inflation figures.

Rationalwiki pokes a bit of fun:

http://rationalwiki.org/wiki/Shadow_Government_Statistics

Would you agree?
Ben J
10. February 2014 at 02:44

Tom Brown,

http://blog.jparsons.net/2011/03/shadow-stats-debunked-part-i.html

That link is the most damning example to me of how shadow stats is clearly false.

If you believe his housing price and CPI numbers concurrently, it implies the real price of housing has never been higher than in the late 1980s. Pretty hilarious.
W. Peden
10. February 2014 at 03:53

Tom Brown,

I think someone worked out that he just increases the rate of change of the GDP deflator, with no specific technical rationale. That’s also how he calculates his CPI figures.
W. Peden
10. February 2014 at 03:59

Ben J,

That’s a good link.
Benjamin Cole
10. February 2014 at 05:03

Major Freedom:

Well, I think you missed the thrust of my comment, and that was wages are sticky as workers regard nominal wage cuts as punitive.

Why did we have a tech stock boom in 1999 when corporate AAAs were at 7.5 percent?
ssumner
10. February 2014 at 05:26

Conor. I didn’t say salaried wages are as sticky as hourly wages, just that they are sticky.
Mark A. Sadowski
10. February 2014 at 06:33

Tom Brown,
“I’m guessing you’re not a fan of [John Williams’] work.”

That would be a good guess.

“I’m not asking for a full debunking here, but what would be your #1 comment regarding his alternative real GDP figures? What would be your top couple of complaints?”

To do a complete debunking of John Williams would require getting ahold of his data in order to back out his methods, which I suspect are rather amateurish. So far the only people willing to pay the subscription price are people who don’t have all their oars in the water. I imagine eventually some people will pool together for a subscription and do a complete dissection of Shadowstats which ought to be pretty amusing.

Here’s his RGDP page:

http://www.shadowstats.com/alternate_data/gross-domestic-product-charts

Note that apart from a blip in 2004, year on year RGDP growth has been negative since before 2000, so according to John Williams the US economy has been contracting for about 15 years now. Do I have to explain why that is so ridiculous?

Also, just eyeballing the difference in growth between his estimate and the official estimate back to 1982 it implies that US RGDP is only about a third as large as is claimed. If true, this would imply that US RGDP (PPP) is less than half of China’s RGDP and barely more than that of Japan’s or India’s RGDP. It also implies that US RGDP per capita is about half that of the Euro Area, Japan or the UK and lower than every EU member with the exception of Bulgaria and Romania.

The alternative is of course that there is a giant international conspiracy with some 200 nations and international organizations all altering their RGDP figures by about the same proportion as the US. Now keep in mind I’m not claiming estimating RGDP is a perfect science. On the contrary, what I’m pointing out is that such a conspiracy would require a degree of coordination and competence among all the organizations charged with estimating national accounts across the globe that simply defies the imagination.

“Do you suppose he starts with the same NGDP the gov does and then uses a different deflator? That would make sense (I think) since he doesn’t agree w/ gov inflation figures.”

Yes, in fact I suspect it’s done in such a mickey mouse fashion it would be easily detectable were someone to stoop to paying for the data.

“Rationalwiki pokes a bit of fun:…Would you agree?”

Yes, it’s a reasonable summary.
flow5
10. February 2014 at 07:16

“There is no NGDP futures market?”

You don’t have to wait for the crops to be harvested to trade, for example, wheat. But to anticipate the BEA’s advanced, preliminary, & final figures (& their revisions & reconstructions), is more akin to gambling.
Major_Freedom
10. February 2014 at 09:38

Don Geddis:

“MF: Such a delight to see your off-topic walls of text again, after your lengthy absence. I also note that you and your doppelganger Geoff, seem rarely able to rant simultaneously. Curious.”

My post was not off topic.

“P.S. “…start by imagining…deflation” Let me guess. You can answer your own questions, without ever leaving your imagination. Yes? You have no need for empirical data. It seems to me that your epistemic process does not distinguish between “hypothesis” and “conclusion”.”

Don’t blame me if you don’t feel good about the answer that makes the most sense to you.

No need for empirical data? Depends on the context. If we’re talking about historical events, then of course empiricism is necessary.

If you have been paying attention, which I doubt, then you will know that the historical data that you believe proves or confirms or verifies your personal economic theories, actually does not do so. All of history is consistent with multiple, mutually inconsistent theories. History cannot possibly be the final arbiter of competing theories.

What you are totally ignorant about, because you never bothered to question what you have been told on school, is that economics is not a science like physics or chemistry. In economics, history is categorically separate from theory.

Now, do you have any substantive arguments to make in response to what I wrote? Or are you going to continue believing that hatred, resentment, hostility, and frustration are going to one day apodictically refute my arguments?

If you want to demolish what I write, you are going to have to educate yourself in my theories. I have done that with your theories, which is why I don’t need to engage in the same petty and churlish behavior that you do as the only way to express my disagreement or skepticism.

You’re never going to win if you continue doing what you’re presently doing, or more accurately, what you’re not doing.
Tom Brown
10. February 2014 at 11:03

Mark, thanks again. If you want to take up a collection to get a subscription, I’d be willing to chip in $20. I’ve still got that in my Paypal from Vincent. 😀
Tom Brown
10. February 2014 at 11:15

Ben J & W. Peden: just now saw your responses. Perhaps a subscription for Mark isn’t necessary then. Thanks.
Doug M
10. February 2014 at 11:27

It doesn’t take $500,000 from anybody to start a new futures market. It just takes enough people who want to play both sides to create two-way flow.

But, if people don’t want to trade, you can’t make them.
Tom Brown
10. February 2014 at 11:43

Mark, regarding this:

http://www.themoneyillusion.com/?p=26133#comment-317961

you are officially not banned at pragcap, just to let you know.
Daniel J
10. February 2014 at 12:01

Sorry this is random but it gave me a chuckle. Rick Santelli interviewing Eugene Fama.

http://www.youtube.com/watch?v=YfdKBYdXDD4
Bob
10. February 2014 at 14:49

There’s also the non-psychological effect of being forced to make the next move, and how that affects firms.

Salary increases and drops do not happen in a continuum, with people getting paid more every day. So the way in which the change happens matters.

Imagine firms A and B change the salaries they offer yearly, but at different times of the year. In practice, this means that there’s a bit of time of the year where salaries in A are updated while those in B are not.

If both increase salaries, A has an advantage getting employees, because there is a window of time where working for A pays more. So B might end up wanting to move salary increases up the calendar, to match A.

But now imagine what happens when we are making salary cuts. A cuts salaries 10%. At that point, we don’t know if B will cut salaries or not. They might instead fire the employees they find less productive. So, at that point in time, B has a window of opportunity for poaching employees from A, safely. What happens if B drops salaries a few months later? A never gets an advantage, so A can’t poach people back from B.

Therefore, as far as employee quality is concerned, it pays to be the first mover while offering higher rates, while it does not pay to be the first one to drop salaries. So you really might be better off delaying rate cuts as much as possible, or just do layoffs instead. So we either get layoffs or firms playing chicken.
Morgan Warstler
10. February 2014 at 14:59

Roger Farmer quoting Glasner takes a bat to Sticky Wage theory:

http://rogerfarmerblog.blogspot.com/2014/02/keynes-and-sticky-pricestime-to-think.html

Mark Sadowski,

I think you should pull his data down from the link and check it, and also SEND HIM YOUR RESUME.
Morgan Warstler
10. February 2014 at 15:01

It really looks like Real Wages / Prices are sticky, and Nominal ones are not, no?
CA
10. February 2014 at 16:40

Wow, that’s an interesting post from Farmer. Scott, what say you?
Jim Glass
10. February 2014 at 19:47

Roger Farmer quoting Glasner takes a bat to Sticky Wage theory:…

“Of course the fact that the CPI and the wage moved in lock step means that the real wage did not fall and that, I would guess, is the fact that Paul and Simon would point to.”

“I guess”. Real wages didn’t fall a bit with unemployment at 25%, and that says wages *aren’t* sticky?

A wiffle bat!
Major_Freedom
10. February 2014 at 20:37

Jim Glass:

“Real wages didn’t fall a bit with unemployment at 25%, and that says wages *aren’t* sticky?”

That’s quite unlikely, even absurd. With a 25% reduction in the workforce, productivity cannot possibly remain the same, let alone grow, unless there was a 25% increase in the average productivity of labor. 25%? Highly unlikely.

Real wages fell as unemployment rose to 25%.
ssumner
10. February 2014 at 21:31

Doug, You said;

“But, if people don’t want to trade, you can’t make them.”

Of course you can. I don’t want to trade, but if you give me $500,000 I will.

Morgan, Farmer’s data is not accurate—don’t know where he got it.
Jason
10. February 2014 at 23:01

It appears that the labor market transfers signals from AD (NGDP) to the labor supply while the wage market does not transfer signals from AD to nominal wages.

This is not to say why this happens, but from an information theory perspective, there is a flow of information captured by the quantity of people employed, not their price (wage).

http://informationtransfereconomics.blogspot.com/2014/01/two-kinds-of-stickiness.html

http://informationtransfereconomics.blogspot.com/2013/10/sticky-wages.html

This actually allows you to derive Okun’s law (change in the number of people employed is equal to change in RGDP) …

http://informationtransfereconomics.blogspot.com/2013/08/scott-sumners-model-part-2_30.html
Kevin Erdmann
10. February 2014 at 23:37

Daniel J. – funny clip.
Saturos
11. February 2014 at 00:24

Mercatus says that the whole drop in unemployment last year was from participation changes: https://www.facebook.com/photo.php?fbid=10152449069369239&set=a.146868129238.110831.19474609238&type=1
Morgan Warstler
11. February 2014 at 05:55

Scott, are you sure you didn’t mean Real wages?

Here is the spreadsheet data:

http://rogerfarmer.com/NewWeb/Spreadsheets/Data%20for%20JEDC%20Paper.xlsx
Morgan Warstler
11. February 2014 at 06:03

Wouldn’t North Carolina be a great test of sticky wages?

http://talkingpointsmemo.com/news/north-carolina-becomes-first-state-to-eliminate-unemployment-benefits

We should see no drop in avg. wages as the unemployed “enter” workforce.

Scott, looking at off-the-books Craigslist gig economy, you don’t think wages are sticky, right?

So IF the economy is more and more gig economy, and it is….
TravisV
11. February 2014 at 06:16

Justin Wolfers via Twitter:

@JustinWolfers
2/11/14, 7:50 AM
A thought: As the unemployment rate proves a less reliable guide, Fed will focus more on inflation data to infer whether slack remains.

@JustinWolfers
2/11/14, 7:51 AM
Implication: Don’t be surprised to hear a lot more discussion about price and wage inflation numbers over the coming year.
Morgan Warstler
11. February 2014 at 06:57

Look how TX passes IL in wages…. how does that support sticky wage theory?

http://illinoispolicy.org/texas-surpasses-illinois-in-median-household-income/
Kevin Erdmann
11. February 2014 at 09:58

Saturos, I realize I’m obsessive on this topic, but the Mercatus post nudged me to post a review of LFP, if you’re interested:
http://idiosyncraticwhisk.blogspot.com/2014/02/review-of-tricky-issues-in-labor-force.html
ssumner
11. February 2014 at 10:04

Jason, That sounds right.

Saturos, I find that somewhat misleading. The labor market is clearly healing. We will be back to the natural rate sooner than people expect.

Morgan,
I mean nominal wages, and the Craigslist effect is still not the main story. Wages are still sticky.

Texas Roolz.

Travis, I agree.
Mark A. Sadowski
11. February 2014 at 10:32

Morgan Warstler,
“I think you should pull [Roger Farmer’s] data down from the link and check it,…”

Scott,
“Morgan, Farmer’s data is not accurate””don’t know where he got it.”

I’ve looked at Roger Farmer’s data. Farmer takes annual data from the BEA on aggregate nominal compensation to employees, and the number of full-time equivalent employees, and converts them to quarterly frequency through interpolation. His wage index is derived by taking the quotient of aggregate nominal compensation by the number of full-time equivalent employees. In other words it is essentially the average compensation of a full-time employee.

This is of course very different from Scott’s prefered measure which is simply nominal hourly earnings. By that measure real wages rose during the Great Contraction because nominal hourly earnings did not fall as fast as prices:

http://research.stlouisfed.org/fred2/graph/?graph_id=141391&category_id=0

And then, in the second hald of 1933, nominal wages rose much faster than prices causing real wages to soar. This obviously presents a very different picture than the one Farmer portrays.

Incidentally, David Glasner has written a response to both Krugman and Farmer, and my initial reaction is that I find myself agreeing 100% with David:

http://uneasymoney.com/2014/02/10/paul-krugman-and-roger-farmer-on-sticky-wages/
Morgan Warstler
11. February 2014 at 12:06

Mark / Scott,

Doesn’t that show that Nominal wages fell 1929-1933?
Mark A. Sadowski
11. February 2014 at 13:16

Morgan,
Yes, nominal hourly earnings fell in 1929-33, but not as much as prices. Thus real hourly earnings actually rose during the contraction, implying downward nominal rigidity.

Farmer is claiming wages were flexible during the Great Depression. He argues this on the basis of the fact that average annual nominal earnings of a full-time equivalent employee appear to fall by the same amount as prices, and thus wages did not change in real terms.
Morgan Warstler
11. February 2014 at 14:05

Mark,

I don’t think so, but I assume this is a communication error.

Scott said “Nominal hourly wages are sticky” I figured he meant “Real”

I read Farmer as saying, no Nominal wages are flexible, but then he admits, well Real wages are sticky bc prices fall / wages fall.

I read you as saying Nominal wages are flexible, but real prices / wages can fall or rise relative to each other.

What i can’t understand is how Scott says Nominal wages are sticky, even Scott data shows they fall.
Mark A. Sadowski
11. February 2014 at 15:15

Morgan,
To say nominal wages are “sticky” is to say they are downwardly rigid, in the sense that it takes time for wages to fall far enough to adjust to a new labor market equilibrium induced by a decline in NGDP. It’s not that nominal wages are inflexible, it’s that they’re insufficiently flexible to restore labor market equilibrium instantaneously.

Farmer seems to think that showing that nominal annual earnings per worker fell by the same proportion as prices during the Great Depression shows that wages were sufficiently flexible to maintain labor market equilibrium. Obviously that was not the case.

Scott argues that nominal hourly earnings would have to fall by roughly the same proportion as NGDP per capita does in order to maintain labor market equilibrium. This is the “musical chairs model” of the economy:

http://research.stlouisfed.org/fred2/graph/?graph_id=134646&category_id=0

Recall that prices didn’t fall by the same proportion as NGDP did in 1929-33. Thus even if nominal annual earnings fell by the same proportion as prices, that means that the aggregate number of people employed fell.

And even if nominal annual earnings had fallen by the same proportion as NGDP per capita, hourly earnings didn’t fall by the same proportion, and consequently the same number of people would have been employed for fewer hours.
Mark A. Sadowski
11. February 2014 at 16:18

Here’s how the musical chairs model fares during the Great Depression:

http://research.stlouisfed.org/fred2/graph/?graph_id=160103&category_id=0

It’s probably not the best wage measure (manufacturing only), and it’s not my favorite unemployment rate series for the period (FERWs are counted as unemployed), and I can’t find a population series in quarterly frequency for the period (so it’s not on a per capita basis), but the coefficient of determination (R-squared) is equal to 0.8375.
Morgan Warstler
11. February 2014 at 17:02

I get what you are saying, but what I don’t see is Scott’s assertion that Nominal wages did not fall.

And it seems important.

MP isn’t premised on dealing with the stickiness of real wages and real prices.

It’s premised on you having a Money Illusion, that keeps you from being willing to go from $.59 per hour in 1929 to $.46 per hour in 1933.

And it looks like nominally, no matter what happened in real prices / wages – people took the pay cut.

And it is interesting to see the other stuff….

BUT IF Farmer is right that people did no suffer Money Illusion, Scott needs to tell me why!
Scott Sumner
11. February 2014 at 19:19

Morgan, Yes, wages fell in 1929-33. Sticky wages doesn’t mean wages that don’t fall, but rather wages that fall by less than the equilibrium level of wages.
ssumner
11. February 2014 at 19:23

Mark, Thanks, that graph is probably worth a post.
Morgan Warstler
12. February 2014 at 04:08

You mean wages are sticky if they fall from 59 cents to 46 cents, but there is still unemployment?

Then FOR SURE, you need another word that is not nominal sticky wages.

(Watching you guys and Glasner and Farmer go at this subject is easily most interesting thing so far this year)

Mark / Scott please read this piece from Jason Smith yesterday (its about this discussion):

“Sticky wages, information transfer and piece work”

http://informationtransfereconomics.blogspot.com/2014/02/sticky-wages-information-transfer-and.html
Daniel
12. February 2014 at 05:24

Morgan & co

So let’s see – NGDP in 1933 was 54% of what it was in 1929.
Yet nominal wages only fell to 75% of what they were in 1929.

And somehow that proves wages aren’t sticky ?

Here’s a question for you – when you wilfully misinterpret data to further you agenda, what does it say about intellectual honesty ?
Morgan Warstler
12. February 2014 at 05:58

Gah!

I have only one agenda GI/CYB

The Money Illusions means PEOPLE STARE AT THE NUMBER ON THEIR MONEY AND LIVE BY THE ILLUSION.

Yes Daniel, we all agree that when Real Prices fall and Real Wages fall we can say Real Wages are sticky.

But Scott says NOMINAL WAGES CAN FALL, BUT IF THERE IS STILL UNEMPLOYMENT, THEN NOMINAL WAGES ARE STILL STICKY

And I think that directly contradicts money illusion, people are supposed to resist taking a nominal pay cut.

And saying there is unemployment BECAUSE people wouldn’t take ENOUGH of a nominal pay cut, seems strained.

People obviously could take a NOMINAL PAY CUT.

but they couldn’t take any more of a real pay cut.

And Dan, that because “NGDP in 1933 was 54% of what it was in 1929.
Yet nominal wages only fell to 75% of what they were in 1929.”

People were NOT nominally adverse, there was no Money Illusion, they just couldn’t possibly live on less in real terms.
Daniel
12. February 2014 at 06:25

So we have the following facts –

1. NOMINAL wages pretty much went along with the CPI – meaning they stayed roughly the same, in REAL terms
2. NGDP was halved
3. wages fell by a quarter

And somehow that means they just couldn’t possibly live on less in real terms ?

The only logical conclusion is that there is unemployment BECAUSE people wouldn’t take ENOUGH of a nominal pay cut.

If that seems “strained” to you, it’s because of your own cognitive blocks.
Morgan Warstler
12. February 2014 at 08:40

“NOMINAL wages pretty much went along with the CPI – meaning they stayed roughly the same, in REAL terms”

WTF? Seriously WTF?

Real terms = Real wages

We do not disagree!

NOMINAL wages pretty much went along with the CPI = Nominal wages FELL.

I’m not being pedantic here, you are using Nominal and Real interchangeably.

Money Illusion is the idea that NOMINAL WAGES ARE STICKY.

Because of money illusion, we can’t give you a pay cut, so it is easier to inflate away your buying power.

Right? Yes.

But wages and prices are NOMINALLY flexible, and it sure look like from 1929-1933 they were…

Then WHAT IS MONEY ILLUSION????

If someone says “Sure cut my wages from 59 cents to 46 cents bc prices are falling, not as much as my wages, but still they are falling, and I’ll keep coming to work, keep applying for jobs”

Well that doesn’t sound like Money Illusion is real, does it?

We’re supposed to be growing money supply bc wages / prices are sticky in NOMINAL TERMS. You are supposed to REFUSE to take that pay cut EVEN IF the price of things is falling.

I don’t have an ideology here, words mean things. Real and Nominal are different.

I don’t have a problem with Scott saying, “nominal wages are not sticky, they can fall BUT they can’t fall enough to clear the labor market”

And I’m pretty sure thats what Scott is saying.

But I wouldn’t use his shorthand of “Nominal wages are sticky” and expect people to add in the other stuff.

I assumed, like others would, that The Money Illusion is why Nominal Wages are sticky.

If someone can accept 46 instead of 59 cents.

But they can’t accept 45 cents.

That’s not because of Money Illusion. ThatS REAL.
ssumner
13. February 2014 at 07:36

Morgan, Wages that don’t change are rigid, wages that respond with a lag are sticky.

Daniel, Good point. I’d add that I don’t believe the CPI existed in 1930. Wages fell far more slowly than the WPI in the early 1930s. Also far more slowly than NGDP.
Morgan Warstler
13. February 2014 at 21:02

Scott I don’t see a lag in 1929-1933, I see nice flexible wages until they hit rigid.
ssumner
14. February 2014 at 14:40

Morgan, You probably are not looking at the right series, wages fell much more slowly than prices. The graph someone linked to is wrong.
Daniel
15. February 2014 at 05:07

I think what Morgan is trying to say is that the very fact that wages fell at all somehow falsifies the sticky wages model.

I think that’s nonsense, but it makes sense it would come from a guy who praises “conservative efforts to un-stick wages”.