The German jobs miracle (and German data bleg)

This is a fairly important post, but it will take me a while to get to the punch line.  I’ve been trying to collect German data, but have not had a lot of success.  As far as I can tell, between the 1st quarter of 2006 and the 4th quarter of 2012, the German unemployment rate fell from about 12% to 5.4%.  But there seem to be different indices.  In any case, that’s the jobs “miracle.”  So then I began to look for other data.

RGDP rose by 8.8%, while NGDP rose by 16.94%.  At annual rates that’s about 1.3% RGDP growth and about 2.4% nominal.  So why such a big drop in unemployment?  The German population does seem to be gradually falling, but as far as I can tell that’s not the main factor.  German employment seems to have risen by around 7% to 10%, depending on the data source.  Can someone help me?  So that suggests either fewer hours worked per worker, or flat productivity.  But as far as I can tell productivity (RGDP/hour) rose around 4% total (with incomplete data), and hours per year only fell by around 1%.  But the hours number may be wrong, as I can’t tell if it includes part-timers.

So how did unemployment fall so much with slow NGDP growth?  Perhaps a bit of everything.  Low inflation helps; that means more RGDP for each euro of NGDP.  But 8.8% RGDP growth is not very much over 6.75 years.  So you then add in slow growth in worker productivity, or less hours worked per worker, or some combination.

BTW, employment in the US is down almost 1% over the same period where German employment has risen by 7% to 10%.  Our RGDP is up by 6%, but that’s not good for a country with robust population growth.

Now here’s where things get interesting.  NGDP grew very fast from 2006:1 to 2008:1, by 9.8% total, or about 4.8% per year.  Then NGDP started falling, and despite a rebound by 2012:4 was only up another 6.2%, barely 1.3% per year.  Market monetarism says that Germany should have done much worse after 2008:1 than before.  Initially they did do poorly, but over the entire 2008:1 to 2012:4 period they did quite well, with the unemployment rate falling from 8.2% to 5.4%.  So what’s wrong with market monetarism?

The short answer is that I took some shortcuts, instead of staying true to the “musical chairs model.”  In the past I’ve often argued that a fall in NGDP causes unemployment because there is less income to pay workers, and yet hourly wages are sticky.  Some workers end up sitting on the floor.  The logic of that model suggests that the real problem is not unstable NGDP, but rather instability in a component of NGDP, namely total wages and salaries.

Fortunately I was able to find German data for aggregate worker compensation (wrong link, try this.):

2006:1   286,190 mil. euros   (50.5% of NGDP)

2008:1  303,710 mil. euros  (48.8% of NGDP)

2012:4  347,820 mil. euros  (52.5% of NGDP)

To say these number blew my mind would be an understatement.  I immediately saw the annual rates were similar, and my calculator showed 3.0% growth during the huge German boom of 2006:1 – 2008:1, and then an almost identical 2.9% during the global recession and recovery, when their NGDP growth rate plunged from 4.8% to 1.3%.  No wonder the Germans love the ECB’s tight money policy!!

The logic of the musical chairs model is that you should stabilize the path of aggregate nominal wages and salaries, because hourly wages are sticky.  When I started blogging I assumed wage targeting would be politically impossible, and knew that NGDP targeting was already a well-regarded concept.  So I latched on to NGDP targeting.  But in retrospect I wish I’d latched on to aggregate employee income.  Call it “income targeting.”

Recessions are not caused by less spending; they are caused by less income going to workers.  Usually the two go hand-in-hand, but the German miracle tells us that when they diverge, it is employer employee income that matters most.

PS.  Now that market monetarism riding high, I figured it was time for a vicious internecine struggle for the soul of market monetarism.  Consider this the first shot.  🙂

PPS.  Just to anticipate one objection from my fellow MMs, I do realize that German NGDP was above trend in 2008:1, and that one could argue that it is currently near the trend line.  But in that case the unemployment rate should have fallen during 2006-08, and then risen.  But it fell fairly steadily.

PPPS.  Another advantage of total compensation over NGDP is that it is easier to measure, not susceptible to the sorts of definitional problems discussed by James Hamilton here.  (I.e., is R&D part of GDP?)


Tags:

 
 
 

40 Responses to “The German jobs miracle (and German data bleg)”

  1. Gravatar of Arnold Kling Arnold Kling
    6. May 2013 at 05:57

    What happened to the ratio of German nationals to others in the work force, and how did this affect the various measures?

  2. Gravatar of anon anon
    6. May 2013 at 05:57

    Just a heads-up, the “aggregate worker compensation” link is wrong. It actually has the text of a comment reply that you most likely intended to post as a comment to an earlier post.

  3. Gravatar of John Hall John Hall
    6. May 2013 at 06:01

    My understanding was that German labor market reforms in the mid-2000s were responsible for the strong employment growth. It seems like if half of the employment growth came from workers classified as unemployed, you can easily get a substantial decline in the unemployment rate.

  4. Gravatar of Petar Petar
    6. May 2013 at 06:11

    Thnx for an interesting post mr. Sumner. Do you think of Agenda2010 reforms that made labor markets more flexible were a factor here?

    The link to aggregate worker compensation is broken – btw… I dont really understand how can 300 billion be around 50% of german NGDP which I think is something like 2650 billion ear for 2012. On Destatis (https://www.destatis.de/DE/ZahlenFakten/GesamtwirtschaftUmwelt/VGR/Inlandsprodukt/Tabellen/Einkommensverteilung.html) you have “arbeitnehmer gehalter” which is 1 375,52 and that is something above 50%

  5. Gravatar of Petar Petar
    6. May 2013 at 06:11

    sorry, heres in english https://www.destatis.de/EN/FactsFigures/NationalEconomyEnvironment/NationalAccounts/DomesticProduct/Tables/DisposableIncomeSavingNetLending_NetBorrowingTotalEconomy.html

  6. Gravatar of Ashok Rao Ashok Rao
    6. May 2013 at 06:14

    “Low inflation helps [wrt unemployment]; that means more RGDP for each euro of NGDP. But 8.8% RGDP growth is not very much over 6.75 years.”

    In the long-run, yes, but in the short-run other things equal greater employment would increase inflationary pressure. More RGDP per euro/low inflation encourages capital investment, which is what we see in the 2% fall of labor share you cited.

    “But in retrospect I wish I’d latched on to aggregate employee income. Call it “income targeting.” ”

    Would this work in the long run? Let’s say America committed to this in 1950. Look at this graph: http://research.stlouisfed.org/fredgraph.png?g=i9h

    The secular decline of labor income would make long-run wage targeting sufficiently inflationary. As the marginal rate of technical substitution of capital relative to labor increases, labor share will fall requiring inflation. We would succeed at the cost of capital formation.

    This is a big “shot”, I’m interested to see where it leads.

  7. Gravatar of Tyler Cowen Tyler Cowen
    6. May 2013 at 06:32

    I like this post, but somehow the phrase “Many German wages fell” should enter more centrally into the discussion.

  8. Gravatar of Luis Pedro Coelho Luis Pedro Coelho
    6. May 2013 at 06:38

    Labour reform is, at least, a major confounder:

    http://en.wikipedia.org/wiki/Hartz_concept#Hartz_IV

    If you follow-up and argue that it was labour reform that caused the share of GDP going to workers to increase, then you’ll make some left-wingers heads explode 🙂

  9. Gravatar of MG MG
    6. May 2013 at 06:49

    I echo John Hall’s comment, and while I am far from an expert in Germany, I recall that these reforms even included a successful program to encourage job sharing. So, part of the headline miracle stat that made you dig into the rest of the data may not be as miraculous. That being said, you may have a point (and a big one): monetary policy could aim to target a suitable growth in nominal compensation assuming a baseline level of structural reforms in the labor markets is in place and (where it matters) trends in the country’s productivity/competitiveness.

  10. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    6. May 2013 at 07:03

    Oscar The Commie says, Maybe Milton was right;

    http://www.seattlepi.com/business/article/German-who-helped-start-euro-now-calls-for-its-end-4491112.php

    ————quote————-
    The leftwing opposition politician who was Germany’s finance minister when the euro was adopted says it’s time to think about abandoning the common currency in the wake of the enormous economic problems in southern Europe.

    Oskar Lafontaine says on his website that struggling eurozone countries have not been able to devalue their currencies to help them cope with the economic crisis that has seen several slip into recession.

    Lafontaine proposes a restructuring of Europe’s banking system, stricter controls on lenders, and a return to the European monetary system that preceded the euro.

    Lafontaine is now out of federal politics but is still an influential member of the Left Party, the successor to former East German communists.
    ————-endquote————-

  11. Gravatar of ssumner ssumner
    6. May 2013 at 07:10

    Arnold, My hunch is that it isn’t a decisive factor, but I can’t be sure.

    anon, Thanks, I fixed it.

    John and Petar, I agree, but the main issue is the smoothness of the decline, despite unstable NGDP.

    Petar, Those are quarterly numbers, not annualized.

    Ashok, No, I don’t think it would be particularly inflationary–wage growth under income targeting would have been lower than the actual wage growth that we observed. In any case, inflation doesn’t matter, NGDP growth matters.

    Tyler, Yes, but there is the separate issue of explaining the remarkable stability before and after 2008.

    MG, I’ll do another post to explain my argument better.

  12. Gravatar of Suvy Suvy
    6. May 2013 at 07:12

    “So that suggests either fewer hours worked per worker, or flat productivity. But as far as I can tell productivity (RGDP/hour) rose around 4% total (with incomplete data), and hours per year only fell by around 1%.”

    You know that a bad economic theory exists when people say that as a population becomes less productive, employment is supposed to go up. This absurd idea would imply that Ethiopia should have a lower unemployment rate than the US.

    This also shows that NGDP isn’t the most important factor and low levels of NGDP isn’t why our unemployment rate isn’t determined by only NGDP. The German economy’s private sector isn’t nearly as indebted as the US private sector.

  13. Gravatar of Suvy Suvy
    6. May 2013 at 07:19

    NOT EVERY COUNTRY SHOULD HAVE 3% RGDP GROWTH. In Germany, they can’t sustain the same RGDP growth as the US because they don’t have a rising population. Infinite growth forever in a world of finite resources is impossible, especially as natural resources start to become scarce.

    How can you talk about real GDP growth without talking about the size of a country’s population and without talking about population growth?!

    I remember when I got into an argument with someone about how Japan could have 3% RGDP growth in a sustainable manner. This shows the absurdity of economic theory. Japan is a country that has had a falling workforce for around 20 years and is aging faster than any country I can think of in world history. How Japan can even have 1% RGDP growth sustainably is beyond me, never mind 3% RGDP growth. Ditto for the countries in Europe.

  14. Gravatar of Suvy Suvy
    6. May 2013 at 07:20

    “At annual rates that’s about 1.3% RGDP growth and about 2.4% nominal.”

    We have a rising population, Germany does not. It’s that simple.

  15. Gravatar of Ashok Rao Ashok Rao
    6. May 2013 at 07:28

    “Ashok, No, I don’t think it would be particularly inflationary-wage growth under income targeting would have been lower than the actual wage growth that we observed. In any case, inflation doesn’t matter, NGDP growth matters.”

    You don’t think it has any distortionary effects on capital formation, though? If NGDP growth matters and moves fully hand-in-hand with wages, then targeting one is the same as targeting the other. If targeting wages are more optimal, it allows flexibility in expected NGDP growth.

    This means is NGDP growth < Wage growth, capital share of income is falling, unnaturally.

    By the way, would a more optimal regime be targeting of NGDP and wages (both have a buffer room) weighted by the relative stickiness of wages to prices?

    Because wages are stickier in the short-run of business cycle, this would be a move towards wage targeting, but not all the way.

  16. Gravatar of Geoff Geoff
    6. May 2013 at 07:31

    This post is a relief, because it is a movement towards sound economics and away from the “simplified” models that contain blatant logical contradictions.

    “The logic of that model suggests that the real problem is not unstable NGDP, but rather instability in a component of NGDP, namely total wages and salaries.”

    Wages and salaries are not direct components of NGDP. NGDP includes spending on consumer and capital goods (and services), i.e. spending that generates profits (or losses) on capital invested. Wages and salaries are indirect components of NGDP, that is, it is one form of spending that precedes (logically and temporally) NGDP spending. I pay a worker $100, but that doesn’t show up in NGDP. Later on, once that worker spends that money on consumer or capital goods (or services), then $100 will be added to NGDP.

    “I started blogging I assumed wage targeting would be politically impossible, and knew that NGDP targeting was already a well-regarded concept. So I latched on to NGDP targeting. But in retrospect I wish I’d latched on to aggregate employee income. Call it “income targeting.””

    “Recessions are not caused by less spending; they are caused by less income going to workers. Usually the two go hand-in-hand, but the German miracle tells us that when they diverge, it is employer income that matters most.”

    Wow. This is a step closer to the credit circulation theory of depressions. You’re almost there.

    The argument that recessions (increases in unemployment) are caused by a drop in income going to workers, is another way of arguing that recessions are caused by a drop in saving and investment.

    After all, paying a wage is an act of saving, not an act of consumption (well, in most instances anyway…some wage payments are a form of consumption, such as someone hiring personal servants, on the part of those paying wages, but this consumption is small enough relative to total wage payments that it can be ignored).

    So if wage payments decline, then, ceteris paribus, there has a decline in saving and investment. They mean the same thing.

    So then the next question to ask is why did saving and investment significantly decline? First, let’s make clear when it happened. Let’s look at the trend of wages and salaries (semi-annual averaged):

    http://research.stlouisfed.org/fredgraph.png?g=i9n

    If you look at the period from 2002 to 2009, you will notice that wages and salaries grew from 2002 until 2006 when it peaked, and then wage payments started to decline, until 2009 when wage payments began to increase once again.

    For some reason, many business owners found it in their interests to start reducing saving and investing in wage payments in 2006. Why in 2006?

    The credit circulation theory states that non-market interest rates brought about by central bank activity redirect labor and capital into physically unsustainable investment projects. Once those errors are revealed, there naturally arises an incentive to stop financing those projects as they were intended to be financed.

    But if there were revealed errors, why did they get revealed in 2006, and not previously say in 2003 or 2004, nor later on say in 2008 and 2009 when NGDP collapsed?

    Let’s look at the same chart as above, this time including the key interest rate that the Fed “manages”:

    http://research.stlouisfed.org/fredgraph.png?g=i9o

    During the period of 2004-2005, the Fed raised the key rate from a low of 1% in 2004, to a high of 5.5% by 2006. It was during this time that the growth rate in wage payments declined, finally hitting zero growth rate in 2006, and then an absolute decline.

    The credit circulation theory applied to this information is the following:

    1. The Fed lowered the key interest rate from a high of 6.5% in 2000, to a low of 1% by 2003 and into 2004. This sent market signals to investors to allocate labor and resources in such a way *as if* society’s rate of real savings has increased to that extent. But real savings did not increase to that extent in actuality. Just because the Fed lowers the key rate, it does not mean that actual consumer spending and savings preferences have changed. The illusion that it did was monetary in origin.

    So the investments post-2000, in an environment a falling key interest rate (which is associated with falling interest rates in general), were physically unsustainable. There were not enough real savings to complete the projects started.

    Now, this physical unsustainability is, as mentioned above, revealed through individual investment errors. Why did these errors aggregate in 2006? It is because the Fed chose to prevent runaway inflation, rather than accelerate its activity to keep the investments propped up in the face of reality of society’s real saving and consumption. So the Fed raised and held the key rate steady above 5% by 2006. The rise in interest rates exposed all the previous investments that were dependent on lower interest rates, as errors.

    That is why saving and investment started to absolutely decline in 2006, after a year or so of the growth rate declining. It’s because the Fed reduced the fuel that made such investments seem worthwhile previously, and at some point, the incentive was for investors to cut their losses.

    Finally, does the data show that the problem wasn’t one of aggregate demand post-2006? If interest rates are the culprit, then surely those industries most sensitive to interest rates (capital intensive industries) would be revealed as more problematic than industries least sensitive to interest rates (service and retail industries). Let’s look at the various industry employment levels post-2006 (indexed):

    http://research.stlouisfed.org/fredgraph.png?g=i9p

    Well isn’t that interesting. From 2006 to 2008, there was an relative addition to employment in the retail and service sectors, and a relative removal of employment from the construction and durable goods industries. This is exactly what one would expect during a period of correction where investors realize that too much employment and resources were allocated to industries most stimulated by lower interest rates, and not enough to industries least stimulated by lower interest rates. The market was in the process of rebalancing the economy to be a reflection of higher interest rates (as if society’s real saving and investment did not increase after all).

    In the aftermath of post-2009, we see very clearly which industries suffered the most relative to other industries. We see that construction and durable goods suffered the most, which is the same thing as saying too much investment was devoted to those lines previously. We see that retail and service suffered the least, which is the same thing as saying too little was devoted there (given that the excess labor and resources devoted to construction and durable goods, were instead made available for retail and service).

    Overall, what happened can be explained with textbook credit circulation theory.

  17. Gravatar of Rob Rawlings Rob Rawlings
    6. May 2013 at 08:28

    “my calculator showed 3.0% growth during the huge German boom of 2006:1 – 2008:1, and then an almost identical 2.9% during the global recession and recovery, when their NGDP growth rate plunged from 4.8% to 1.3%.”

    Does this mean that had Germany been targeting NGDP they would have had much higher inflation during this period? Is so, why would this be ?

    Also, while this trend in wage growth is noteworthy it is pretty clear that it wasn’t being targeted by anyone – it came about by other means. Some of these other means seems to be fiscal policies aimed at raising employment such as job subsidies. (see: http://crookedtimber.org/2011/01/19/keynesianism-by-stealth-and-symbolic-austerity-german-fiscal-policy-and-the-post-2007-economic-crisis/)

    Does this partially explain why employment could recover despite NGDP being below trend ?

  18. Gravatar of Farid Elwailly Farid Elwailly
    6. May 2013 at 09:25

    “it is employer income that matters most” should, I think, be “it is employee income that matters most”.
    Good post. My question is did the ECB target German employee income or was this a nice accident?
    More interesting, is it even possible for the ECB to target aggregate European employee income since that would drive inflation rate at the core to a politically unacceptable rate?

  19. Gravatar of Mark A. Sadowski Mark A. Sadowski
    6. May 2013 at 09:39

    Eurostat’s “Compensation of all employees” data may be found here:

    http://appsso.eurostat.ec.europa.eu/nui/show.do?dataset=namq_nace06_c&lang=en

    The only eurozone member for which they don’t have equivalent data is Malta.

    I checked to see how Germany compares with the 15 other eurozone members for which there is data (using the most recent quarter for which they have data) and what I found was Germany is the *only* country *not* to have a big drop in the rate of nominal labor compensation growth since 2008Q1. This seems to support David Beckworth’s argument that ECB monetary policy is primarily geared towards Germany’s needs.

    Furthermore Germany had the largest arithmetic drop in the labor compensation to GDP ratio from 2006Q1 to 2008Q1 and the largest increase since among these 16 nations.

    As a side note, Germany also went from having the third highest labor compensation to GDP ratio in 2006Q1 to having the 9th highest in 2008Q1 to having the fourth highest most recently among these 16 nations. For all the talk of the Hartz Reforms the fact is Germany’s labor markets are still among the least liberal in the eurozone. Matthew Yglesias recently pointed to ratings by the Fraser Institute showing Germany went from the least liberal labor markets before the reforms to ranking just barely more liberal than Greece, Portugal and Spain, behind Italy and well below Ireland.

  20. Gravatar of Eliezer Yudkowsky Eliezer Yudkowsky
    6. May 2013 at 11:00

    If worker income as a share of NGDI is falling – I get the impression there’s a long-term secular trend here over time – then wouldn’t nominal wage targeting produce overly rapid NGDP growth and wouldn’t that have a destabilizing effect of the sort people used to attribute to inflation? I expect this is probably an overly basic question and the answer is something like “it’s a higher trend but that higherness is stable and the important thing is maximum stability in wage income” but would still like to hear it spelled out.

  21. Gravatar of J J
    6. May 2013 at 11:07

    Professor Sumner,

    FYI, I see that you are now on the list of people influenced by Milton Friedman on his Wikipedia page.

  22. Gravatar of Jacob Lyles Jacob Lyles
    6. May 2013 at 11:10

    Hey Scott,

    Since I’ve been reading your blog, I was reminded of a sci fi novel by Robert Heinlein influenced by monetarist thinking. The government produces a constant rate of money growth, but instead of using new money to buy financial assets it gives new money directly as wages to the population.

    What do you think of this policy as a macroeconomic tool? It seems like piping new money through worker’s paychecks might be more effective at generating inflation than putting money on bank balance sheets.

    The only downside I see vs. asset purchases is that it doesn’t have a slow-down mechanism.

  23. Gravatar of genauer genauer
    6. May 2013 at 11:25

    Scott, it is a long time since I commented here,
    But since Arnold, Tyler, and Mark joined it, maybe I make a few remarks.

    I think I understand the US system I significant detail, I once contacted some nice guy at the US statistics on how lower numbers of returns to the survey in the winter storm 2009 were accounted for : – )

    I tried once to explain to you, how the incorporation of US housing costs in the CPI is reasonable, well, at least to me, and why the FED is going by PCE ex energy and food, and the Euro HICP is different, but appropriate here.

    For you, and other Americans, to start to understand German numbers,
    you would first need to understand the following words, and complex programs behind them:

    1. Unemployment: official count by

    2. arbeitsagentur, our number 7.1% not comparable to ILO 5.4%, determination completely different to US procedure (Survey)

    3. three massive trends overloading

    4. Hartz IV, suddenly since 2005/2006 counting about 1 million, or 2.5% more

    5. 1-Euro Jobs, some weird program, with very unexpected results, intended as kind of harassment, but mostly scrapped for unintended effects

    6. Ich-AG, kind of zero-capital one man enterprise, having interesting, more positive results, not counting to unemployment, but nor sure about to employment numbers

    7. Existenzgründerförderung, higher level entrepreneurship subsidy, nice to abuse, LOL, cut back

    8. Beschäftigungsgesellschaft, typical done after insolvencies with at least some residual capital available, mostly only above a few hundred people fired, not counting as unemployment, but definitely is

    9. Arbeitszeitkonten , people work more during good times, but do not get paid, and then get paid for it during worse times, sometimes years ago.

    10. Ãœberstunden (your overtime) also not paid out in many cases immediately. Even in the US I carried some 2 months forward for years, in some hybrid contract

    11. Kurzarbeit (if a company is short on new contracts, people work e.g. 60%, but get paid 90%, some 20 % paid by the company, 10% by the unemployment office, specific highly dependent on the situation, the company, and maybe some other things

    12. Elternzeit (young fathers and mothers stay at home, for some time, regulations significantly changing ofver the period you are looking on, what counts as what, and who pays)

    13. Gründungszuschuss

    And some more. All of these programs changed substantially during the time, and accounted in various ways towards numbers like hours worked, worker compensation, GDP.
    Often smoothing out data, but sometimes, like 1Q2006 steepening changes.

    Mark wondering about “*not* to have a big drop ” comes from that. The balance of power in the labor market, especially in the low income sector has shifted, substantially.

    Your attempts to understand German data is like trying to repair a 7-series BMW with a sledgehammer.

    I tried a few times to explain your NGDP, which I certainly do not support, to people here, they simply couldn’t believe that somebody could be that crazy.

  24. Gravatar of Theo Clifford Theo Clifford
    6. May 2013 at 11:40

    I think the falling-unemployment story is a pretty obvious one. You allude to it in your post but you don’t tell it explicitly so I’m going to lay it out here.

    Germany did the Hartz labour market reforms. Unemployment benefits were cut and they introduced ‘minijobs.’ Low-paying part-time or second jobs face low marginal tax rates, minimal social insurance contributions, and have no minimum wage.

    As many as one in five Germans has one of these jobs.

    The effect of the labour market reforms was to encourage the marginal worker to get a job. But the marginal worker is less productive than the average worker, so productivity fell. And they worked fewer hours, because these ‘minijobs’ cannot pay more than €400/month. Falling productivity + fewer hours per worker = lackluster RGDP growth.

    Regarding the market monetarist schism you describe, I’m ambivalent. Clearly wages are more relevant to the musical chairs example. And the measurement point is *huge*. But wages aren’t the only part of the market monetarist story. If corporations see a fall in their expected future revenues, and the profit slice of the pie they get, they’re likely to retrench too.

  25. Gravatar of There´s “nothing wrong with Market Monetarism” | Historinhas There´s “nothing wrong with Market Monetarism” | Historinhas
    6. May 2013 at 12:05

    […] Scott Sumner has a (thought) provoking post on Germany´s job miracle. I highlight a few passages: […]

  26. Gravatar of Joe Eagar Joe Eagar
    6. May 2013 at 13:40

    Nominal income targeting has always had a certain appeal, but like Mr. Sumner says, the politics is nasty. People will naturally think the central bank is messing with their real wages, and of course they are correct (this criticism applies to inflation targeting as well, though in that case the central bank has the excuse that its actions are necessary to meet its target).

    I’m curious about the mechanics; would the current monetary policy transmission mechanisms suffice, or would new ones have to be made?

  27. Gravatar of ssumner ssumner
    6. May 2013 at 14:06

    Suvy, You misinterpreted the post on several levels. I wasn’t making the claims you said I was. I’ll do a follow up ost.

    Rob, Yes, inflation would have been higher because productivity growth was lower. And yes, the stable growth in nominal compensation was a happy accident for Germany.

    Farid, Thanks, I’ll correct that. And yes, a happy accident. They could target income growth, and inflation would not be a big problem if they did so.

    Mark, Great comment. My only quibble is the Fraser rating. Germany has no minimum wage, which allows for a large number of low wage jobs for the unskilled. Most other southern European countries do have minimum wages, and have few jobs for the unskilled. Perhaps Germany is less liberal in other areas, but the wage flexibility is a huge advantage for Germany.

    Eliezer, There are several trend and cyclical issues involved. If the trend in share of NGDP earned by workers is falling slowly, inflation would be slightly higher, but only if they didn’t compensate with a lower rate of growth in income. And the difference would be very small, less than 1%.

    There might be greater cyclicality of inflation, but even if there were the welfare costs of inflation might fall, as the costs that most attribute to inflation are actually more closely correlated with high and unstable nominal income growth.

    Thanks J.

    Jacob, In my “Cantillon effect” posts I argued it would make no difference, as the government doesn’t give away new money, it sells it at market prices.

    genauer, I don’t think that data has much of a bearing on my central claim–I’ll do a follow up oost.

    Theo, You said;

    “If corporations see a fall in their expected future revenues, and the profit slice of the pie they get, they’re likely to retrench too.”

    It’s a mistake to view this as an alternative to the sticky wage story, rather it’s simply another way of looking at the sticky wage story.

  28. Gravatar of genauer genauer
    6. May 2013 at 14:41

    And to give folks in the US some more ideas, how we organize our efficient job system:

    a) we have a 3-tier school system, selection is mainly done at the age of 10

    b) the majority does 9 or 10 years of school, not 12

    c) the majority goes on to apprenticeships, doing college is not the holy grail

    d) people doing apprenticeships are well respected members of society, some doing their Meister later on, starting their own business.

    e) But before this “social injustice” arguments comes: Our income distribution has a Gini coefficient of 27 vs the US 45.

    f) public schools and universities are well maintained. Private schools and universities are for rich misfits

    g) we played a little bit with (small, 500 per semester) tuition fees and abandoned that in this year completely. Nobody racks up student loan debt. That makes it also a lot easier, to abandon false dreams.

    h) Union rights are strictly enforced, by the police, as it happened with Walmart in 2007, if necessary. In large companies the Unions have half the board seats. They understand the business in detail, globally, and act accordingly, disciplined and responsible.

    i) wages are determined by employer and employees. There is no government administered minimum wage.

    j) there is some factual “minimum wage” driven by a guaranteed “social minimum”, some 27 k€ or effective 38k $ for a family of 4, numbers highly simplistic here, for the moment.

    k) we have universal health care and pensions, since 128 years, Every asylum seeker enjoys the same health care as me. Nobody goes hungry or cold to sleep.

    Just some food for thought : – )

  29. Gravatar of Fran Fran
    6. May 2013 at 15:47

    You may want to try the OECD source, especially if your goal is to make international comparisons:
    http://stats.oecd.org/Index.aspx?DataSetCode=STLABOUR

  30. Gravatar of Saturos Saturos
    7. May 2013 at 01:43

    “Recessions are not caused by less spending; they are caused by less income going to workers”

    It would be clearer to say, “less income to workers for given nominal wages”, otherwise you get the fallacy, “Nominal wage rigidity is a good thing because it sustains Aggregate Demand” http://econlog.econlib.org/archives/2013/05/straight_talk_a.html

    Also, any opinions on this paper?
    http://www.newyorkfed.org/research/staff_reports/sr615.html

  31. Gravatar of Britmouse Britmouse
    7. May 2013 at 02:14

    In my post on the “musical chairs” model I used “labour income” as total employee compensation plus mixed income. Is it appropriate to ignore the income of the “self-employed”?

    I have been meaning to do a post on all this. The UK data also shows a divergent trend between the growth of NGDP and employee compensation since 2010. NGDP growth in the three years to 2012 was 4.6%, 3.4%, 1.7%, whereas “labour income” was 1.5%, 2.3%, 2.9%.

    In that sense there is no “mystery” in the UK jobs data – in 2012 we had the fastest growth in total labour income since 2007, and with hourly wage rates only just positive.

    There is research for the UK similar along the lines described by Hamilton, which says that UK NGDP and RGDP have been understated by maybe 0.5% p.a in recent years. http://haskelecon.blogspot.co.uk/2013/02/can-intangible-investment-explain-uk.html

  32. Gravatar of Benny Lava Benny Lava
    7. May 2013 at 05:00

    Scott,

    You mention demographics and the falling population. But what about falling worker population. That is to say that the population of working age people is probably falling faster than the population at large. I suggest this because it was the case in Japan that worker population began to shrink in 2005 or earlier, before the general population began to shrink because of longevity issues. I’m busy these days but maybe one of your worker bees could dig up German worker population.

  33. Gravatar of Tom Tom
    7. May 2013 at 05:11

    Ohanian and Raffo created a quarterly hours series for OECD countries that took into account changing series’ definitions, reliability, etc. They may be aware of more recent vintages and/or efforts.

    http://www.nber.org/papers/w17420

  34. Gravatar of Tom Tom
    7. May 2013 at 05:22

    Also Stata-Bloggers (among others) pointed to a new data repository; Quandl.

    The search below has datasets from the UN, World Bank, and Bundesbank (which a commenter suggested in your next post).
    http://www.quandl.com/search/germany%20wage

    And Stata-Bloggers for completeness.
    http://stata-bloggers.com/

  35. Gravatar of ssumner ssumner
    7. May 2013 at 06:35

    Thanks Fran.

    Saturos, Yes I should have been clearer on the wage question.

    Regarding the paper, do their results control for NGDP changes?

    Britmouse, I’m not sure whether it makes more sense to include or exclude the self-employed. The UK data is quite interesting.

    Benny, Falling population can help explain the unemployment rate, but not the employment level.

    Tom, Thanks for that data.

  36. Gravatar of Angry Bear » Scott Sumner Goes Marxist, Proposes Targeting Labor’s Share of Income Angry Bear » Scott Sumner Goes Marxist, Proposes Targeting Labor’s Share of Income
    7. May 2013 at 10:28

    […] he says, he buries the lede in his first post. Here it is, from the end of the post (I’m reversing these two paras to make it flow even […]

  37. Gravatar of Asymptosis » Scott Sumner Goes Marxist, Proposes Targeting Labor’s Share of Income Asymptosis » Scott Sumner Goes Marxist, Proposes Targeting Labor’s Share of Income
    7. May 2013 at 10:30

    […] he says, he buries the lede in his first post. Here it is, from the end of the post (I’m reversing these two paras to make it flow even […]

  38. Gravatar of Geoff Geoff
    7. May 2013 at 19:43

    Dr. Sumner:

    “Another advantage of total compensation over NGDP is that it is easier to measure”

    Ah yes, the age old scientistic fallacy, that Hayek warned against, rears its ugly head once again. Observable data somehow hold a premium in terms of importance, whereas unobservable information takes on a secondary, or less important role.

  39. Gravatar of Edward Lambert Edward Lambert
    9. May 2013 at 15:20

    Scott, Labor share targeting for monetary policy is coming into vision. I developed a model just in the past week and then read today that you are writing about it.
    If my model can help, here is the link…
    http://effectivedemand.typepad.com/ed/2013/05/universal-model-for-monetary-policy.html

    It is based on the Keynes’ concept of effective demand. I have numerous equations built up. The data of the movement of the Fed rate fits nicely with the slopes of my model. But I still have a problem with inflation targeting from the Taylor rule. I wrote a post showing how to minimize that…
    http://effectivedemand.typepad.com/ed/2013/05/effective-demand-monetary-policy-adjusting-for-inflation.html

    Scott, I heard you think the same way about inflation targeting.

  40. Gravatar of Is There Something Wrong With Market Monetarism | Last Men and OverMen Is There Something Wrong With Market Monetarism | Last Men and OverMen
    21. February 2017 at 04:40

    […] 1.  I favor monetary stimulus combined with fiscal austerity. 2.  If the central bank continues to stubbornly target inflation, I’ve advocated cuts in employer-side payroll taxes and/or the VAT as a way of reducing inflation, and encouraging more monetary stimulus.  So if monetary policy refuses to play ball I’ve advocated effective forms of fiscal stimulus.  I don’t favor ineffectivestimulus, such as bloated military spending merely to give us more “empty GDP calories.”      http://lastmenandovermen.com/2013/05/05/sumners-nuanced-support-of-austerity/      As far as what’s wrong with MM, however, it’s not my question but MMer Marcus Nunes.       http://thefaintofheart.wordpress.com/2013/05/06/theres-nothing-wrong-with-market-monetarism/      Actually, Marcus is just responding to Sumner who first asked it.       “Now that market monetarism riding high, I figured it was time for a vicious internecine struggle for the soul of market monetarism.  Consider this the first shot.  ?        http://www.themoneyillusion.com/?p=21058#comments […]

Leave a Reply