A follow-up on Germany

Most people seem to have misinterpreted my previous post, probably because I “buried the lede” at the end.  Just to be clear, I have no problem explaining why German unemployment fell between 2006 and 2012.  Although NGDP growth was fairly slow, Germany did lots of labor market reforms, which I’ve discussed in previous posts, and this opened up lots of low wage jobs.  That could also explain the relatively weak productivity figures, although as I indicated earlier there are lots of issues with German data, so I’m not sure exactly what happened to productivity.

Mostly I was asking if anyone knew of any accurate data for German employment, hourly productivity, and hours per year between 2006:1 and 2012:4.  No one seemed to know, so I assume the data is very hard to find, if available at all.  And even if I had the data, it wouldn’t really have any causal implication, it would simply be an accounting-like partitioning of growth into its components (labor employed, hours per year, and productivity per hour.

What I found amazing was the fact that total nominal compensation kept growing at roughly 3% per year, even as NGDP growth slowed from 4.8% to 1.3%.  In case anyone doesn’t think that is amazing, consider that the invaluable Mark Sadowski did research on this and found that nothing like that happened in any of the other eurozone members:

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.

I would encourage readers to not jump to conclusions when thinking in terms of the “musical chairs model.”  There’s a temptation to say; “the real problem was X,” when X was simply a manifestation of the musical chairs problem, i.e. the tendency for the unemployment rate to rise when wages are sticky and nominal income growth slows sharply.  Let me be more specific.  Suppose someone said the real explanation was the labor market reforms of around 2003-05.  Here’s how I would respond:

1.  Yes, that’s why Germany could create more jobs with a given amount of NGDP growth than the other eurozone members.  Call that the “miracle of 2006-12”

2.  But no, that doesn’t explain why German unemployment continued falling after 2008, when NGDP slowed sharply.  But the fact that aggregate nominal employee income continue rising at roughly 3% does explain why German unemployment kept falling after 2008.  Call that the “miracle of 2008-12”.

Those are two very different “miracles,” which require very different explanations.

[Update:  Marcus Nunes has an excellent post that presents a different interpretation, one more favorable to NGDP targeting.  I’m still thinking about the issue.]

PS.  This was one of the most successful labor market reforms in history.  The data suggests it increased the share of German GDP going to the working class.  However by creating low wage jobs it may have also made German wages more unequal.  So naturally the party responsible for one of the most successful labor market reforms in history now wants to scale back the reforms, and even the Christian Democrats are calling for a minimum wage.  Here’s The Economist:

WHEN Gerhard Schröder took to the podium in the Bundestag on March 14th, 2003, Germany was called the “sick man of Europe”. More than 4m Germans (11.6% of the workforce) were on the dole. A widespread assumption was that unemployment could never be defeated, merely “administered,” says Wolfgang Clement, who was the former chancellor’s labour and economics minister. Countering that spirit, Mr Schröder unveiled a package of reforms that he called Agenda 2010. The leader of the opposition, a little-known physicist from east Germany called Angela Merkel, derided it as unambitious. But it soon became clear that the agenda would transform Germany’s labour market.

Ten years later, what is the verdict? Financial crises be damned, Germany stands as an economic beacon, with record employment and the lowest youth unemployment in Europe (see chart). Some countries are studying Agenda 2010 as if it were a manual. Mr Schröder is being feted at conferences all over the world.

Only the Germans themselves, and notably Mr Schröder’s own Social Democrats (SPD) and the Greens, who were his coalition partners, seem unsure whether Agenda 2010 was a blessing or a curse. Now in opposition, but hoping this year to defeat Mrs Merkel, who replaced Mr Schröder as chancellor, the SPD cannot exactly disavow its own reform. But it is signalling to its blue-collar base and its own left wing that it might undo parts of it.

So the one European economy with a healthy economy wants to undo the reforms that made it successful.  Go figure.



37 Responses to “A follow-up on Germany”

  1. Gravatar of marcus nunes marcus nunes
    6. May 2013 at 15:21

    Mark is not quite right. Germany´s compensation had a very small drop between 2008.IV and 2009.II. If he´s correct relative to the others, it was Austria who never had any drop (big or small)in labor compensation before, during or after the crisis.

  2. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    6. May 2013 at 15:28

    In politics, no good deed goes unrepudiated.

  3. Gravatar of ssumner ssumner
    6. May 2013 at 15:41

    Marcus, I interpreted Mark as indicating that the 2008-12 period was similar to 2006-08, which would allow for a slight drop in 2008-09. BTW, I added an update with a link to your new post. I encourage everyone to take a look at it.

    Patrick, Yup.

  4. Gravatar of Doug M Doug M
    6. May 2013 at 16:31

    When you say German GDP slowed sharply… form where to where and when?

    I was under the impression that Germany NGDP had been Growing at about 2% since the mid 90s, and never really deteriorated from “trend growth.”

  5. Gravatar of Mark A. Sadowski Mark A. Sadowski
    6. May 2013 at 16:39

    Your interpretation is absolutely correct.

    For example Austria had the smallest drop in the average rate of change in aggregate nominal compensation of employees after Germany. From 2006Q1 to 2008Q1 the rate of change averaged 5.0%. From 2008Q1 to 2012Q4 it averaged 3.2%.

    At the opposite extreme you have Estonia which went from 22.5% to 0.1%. Of course Greece, Ireland, Portugal, Spain and Slovenia actually had a decrease in aggregate nominal compensation of employees since 2008Q1, but the change in the average rate of change was smaller than Estonia’s.

    P.S. I haven’t actually looked for the employment/productivity/hours data for Germany. I know the OECD and the Conference Board has such data on an annual basis, but not quarterly. I shall see what I can scrape up.

  6. Gravatar of xtophr xtophr
    6. May 2013 at 17:52


    If we are considering “income targeting” why not consider Gini targeting? The current recession hasn’t hurt wealthy Americans in the least. We pretend to care about NGDP/RGDP only as an academic exercise, not because we care in the slightest about how the poor are faring. Why not base MM on a measure that optimizes wealth creation for say, the upper 5% of income earners?

  7. Gravatar of Benjamin Cole Benjamin Cole
    6. May 2013 at 19:37

    It might be worth noting that the minimum wage in the United States is about one-third lower today (in real terms) than in the 1960s.

    This may have played a positive role in the low inflation-low unemployment 1990s (in real terms the minimum wage was lower in the 1990s than in the 1960s, also. The MW peaked in the 1960s).

    I think the lower real minimum wage, and the de-unionization of the US private labor force, and other structural reforms—deregulation in rails, phones, trucks, airlines, banks, and a reduction in the top federal marginal tax rate from 90 percent to under 40 percent, and a huge increase in imports—mean the US economy is far, far, far less inflation-prone today than in the 1970s.

    All the more reason the Fed should pour it on. 1.1 percent inflation PCE y-o-y is not a threat, and two sickos armed with cooking utensils is not threat to America.

    People need to get their bearings again.

  8. Gravatar of Fed Up Fed Up
    6. May 2013 at 19:42

    How about this?


    trade surplus and work sharing?

  9. Gravatar of Mark A. Sadowski Mark A. Sadowski
    6. May 2013 at 20:05

    Eurostat has quarterly data on productivity and employment. I can’t find annual hours worked but if you have RGDP it’s of course a residual.

    Furthermore, I found out that the Conference Board gets its annual data from Eurostat and the OECD, who collect their employment data using a joint questionaire, so it’s all effectively consistent identical.

    The data series names are as follows:

    RGDP (2005 euro) – namq_gdp_k
    Productivity (RGDP per hour in 2005 euro) – namq_aux_lp
    Employment (domestic) – namq_aux_pem
    Unemployment rate (quarterly) – une_rt_q

    The data is as follows:

    RGDP (annual rate)
    2006Q1 – 2,263,994.4 million euro
    2008Q1 – 2,429,934.8 million euro
    2012Q4 – 2,463,022.8 million euro

    The average rate of growth fell from 3.6% to 0.3%

    2006Q1 – 40.7 euro
    2008Q1 – 42.6 euro
    2012Q4 – 42.6 euro

    The average rate of growth fell from 2.3% to 0.0%

    Employment (1000)
    2006Q1 – 38,940.0
    2008Q1 – 40,250.0
    2012Q4 – 41,719.0

    The average rate of growth fell from 1.7% to 0.8%.

    Hours worked at an annual rate (calculated)
    2006Q1 – 1,428.5
    2008Q1 – 1,417.2
    2012Q4 – 1,385.9

    The average rate of increase fell from (-0.4%) to (-0.5%).

    Unemployment Rate
    2006Q1 – 10.6%
    2008Q1 – 8.0%
    2012Q4 – 5.4%

  10. Gravatar of Mark A. Sadowski Mark A. Sadowski
    6. May 2013 at 20:37

    All of these data series are available seasonally adjusted or non-seasonally adjusted. The figures reported above are all seasonally adjusted.

  11. Gravatar of mbk mbk
    6. May 2013 at 21:15


    I was about to mention Austria. The thing is this – Germany is big and supposedly efficient and went through these big labour reforms in the mid 2000s. So all of these may be “real” factors as to why Germany worked out as a model vs other Eurozone members during the crisis. And since Germany is also influential you may suspect that ECB policy somehow targets German welfare. So this may be a “nominal” factor for German success.

    But how come then Austria? It’s not influential. It’s not big. It used to depend a lot on the German economy as a cheap labour country for suppliers but now I hear (anecdotally) it’s the Germans coming to Austria for work and Austria does well exporting to Asia, maybe better than Germany.

    So are we to believe that the ECB targets Austria too? (nominal explanation). That would be an odd coincidence, no? I mean, does anyone seriously believe that the ECB factually targets the welfare of the German economy above all others, when in fact the political risk in the EU right now is very obviously coming from crisis hit Southern European countries?

    Or are we to believe that Austria, like Germany, simply has economic policies that work well with current ECB monetary policy, even though it didn’t have spectacular labour reforms lately, isn’t known for being particularly efficient etc.? If so, what prevents others from employing the same economic policies and be just fine? This is of course just what the average German would suggest. Note, the average Austrian shuts up while laughing all the way to the bank.

    Finally, if the nominal mattered so much, why don’t non Eurozone countries (UK…) do much better than Eurozone countries? With their independent monetary policies, shouldn’t they at least do as well as Germany?

  12. Gravatar of libertaer libertaer
    7. May 2013 at 00:44

    Maybe this is interesting:

    The first line is GDP, below is productivity per hour worked, the third line is working population, the fourth line is hours worked and the bottom line is hours worked per capita.

    Since the labour reforms “hours worked” are up but still lower then 1991! Meanwhile, each year more people are working. So hours worked per capita is continually going down.

    For me this looks like we Germans solved the musical chairs problem, by putting 2 unemployed on one chair. 🙂 Chair sharing!

  13. Gravatar of Saturos Saturos
    7. May 2013 at 01:45

    The Noah Smith – Brad DeLong bet: http://delong.typepad.com/sdj/2012/07/is-the-us-at-risk-of-becoming-argentina-the-bet-with-noah-smith.html

  14. Gravatar of Saturos Saturos
    7. May 2013 at 01:51

    What happens to arguments like those of Lars Christensen and David Glasner on targeting aggregate nominal income as “the least interventionist monetary policy” when you decide to target aggregate money income to workers specifically instead?

  15. Gravatar of Saturos Saturos
    7. May 2013 at 01:58

    Do poorer people have less demand for holding money? (the MM interpretation): http://marginalrevolution.com/marginalrevolution/2013/05/are-food-stamps-the-best-macro-stabilizer.html

    Theirs is just a model, but we would expect it to be more true at the ZLB, where the marginal propensity to save becomes the marginal propensity to hoard.

  16. Gravatar of Saturos Saturos
    7. May 2013 at 02:11

    Bob Murphy sums it up: http://consultingbyrpm.com/blog/2013/05/applying-krugmanian-lessons-90s.html

    And Tyler has a new disclaimer: http://marginalrevolution.com/marginalrevolution/2013/05/why-is-no-one-talking-about-jamaican-fiscal-policy.html

  17. Gravatar of Saturos Saturos
    7. May 2013 at 02:24

    Finally, Nick Rowe shows EconomistHulk (amongst others) how to “smash” bad economics: http://worthwhile.typepad.com/worthwhile_canadian_initi/2013/05/monetary-stimulus-vs-financial-stability-is-a-false-trade-off.html#more

  18. Gravatar of Petar Petar
    7. May 2013 at 02:46

    Dont know if its relevant but during the recession german government set up a “kurzarbeit” scheme where companies reduced working hours but the difference in income was covered by the government. Some unions also have these types of deals where they agreed to reduce working hours and pay per worker during recessions. When times are “good” a lot of german companies dont raise wages but have yearly performance based payouts, especially the carmakers


  19. Gravatar of Martin Martin
    7. May 2013 at 02:53

    “Mostly I was asking if anyone knew of any accurate data for German employment, hourly productivity, and hours per year between 2006:1 and 2012:4. No one seemed to know, so I assume the data is very hard to find, if available at all.”

    Eurostat as Mark indicates is a possible source for data. However you can, I believe, find the data you need at the Deutsche Bundesbank website.


  20. Gravatar of Germany’s labor market reform since 2003 has been one of the most successful in history. But increasing the share of GDP going to the working class by slashing unemployment increased wage inequality. So now the political pressure is on to repeal it. Germany’s labor market reform since 2003 has been one of the most successful in history. But increasing the share of GDP going to the working class by slashing unemployment increased wage inequality. So now the political pressure is on to repeal it.
    7. May 2013 at 03:01

    […] Source […]

  21. Gravatar of ssumner ssumner
    7. May 2013 at 05:41

    Doug, My previous post has the data.

    Thanks Mark–very helpful.

    xtophr, You can do Gini targeting, but obviously not with monetary policy.

    BTW, I don’t think the Gini for income is a good way of measuring economic inequality.

    mbk, I think the data strongly suggest that both nominal and real policies matter. The interaction is quite complex, as I’ve explained in other posts.

    liberater, I like the “chair sharing” metaphor.

    Saturos, You asked:

    “What happens to arguments like those of Lars Christensen and David Glasner on targeting aggregate nominal income as “the least interventionist monetary policy” when you decide to target aggregate money income to workers specifically instead?”

    It seems more interventionist, but it isn’t. That’s because these are nominal targets, which have no long run impact on the real wages of workers, or their share of total income. You are merely trying to smooth out the cycle, prevent distortions caused by wage stickiness.

    Thanks for the links, I’ll take a look.

    Thanks Martin.

    Petar, Yes, I do recall reading about those policies, and they undoubtedly helped.

  22. Gravatar of Geoff Geoff
    7. May 2013 at 06:09

    “Call that the “miracle of 2006-12″³

    Or you could call it that which I have argued is possible, many times on this blog, due to my understanding of…150 year old economic principles:

    “We now pass to a fourth fundamental theorem respecting Capital, which is, perhaps, oftener overlooked or misconceived than even any of the foregoing. What supports and employs productive labour, is the capital expended in setting it to work, and not the demand of purchasers for the produce of the labour when completed. Demand for commodities is not demand for labour. The demand for commodities determines in what particular branch of production the labour and capital shall be employed; it determines the direction of the labour; but not the more or less of the labour itself, or of the maintenance or payment of the labour. These depend on the amount of the
    capital, or other funds directly devoted to the sustenance and remuneration of labour.” – John Stuart Mill, The Principles of Political Economy, Book I, Ch V.

  23. Gravatar of ssumner ssumner
    7. May 2013 at 06:11

    Fed Up, Definitely not the trade surplus. Job sharing may play a role, but Mark’s data suggests that hours worked per employee fell about the same rate before and after 2008.

  24. Gravatar of Jon Jon
    7. May 2013 at 06:14

    Scott, you don’t need to look to labor compensation. You need to look to expectations. When realized NGDP oscillates but the expected growth path remains the same, labor markets will be stable.

    This is particularly true in Germany where custom is to retain workers through brief downturns. The result is that you can smooth NGDP because everyone accepts the constructin that the future will return to trend soon.

    US tends to follow more of a JIT model, though expectations still moderate employment swings in skilled industries.

  25. Gravatar of ssumner ssumner
    7. May 2013 at 06:23

    Saturos. Brad will almost certainly win, but of course Noah got very favorable odds.

    Regarding food stamps, I don’t think fiscal policy has much of a role in stabilization, regardless of which type is used. The one exception might be supply-side fiscal policy.

  26. Gravatar of ssumner ssumner
    7. May 2013 at 06:24

    Jon, Perhaps, but how does that fit earlier German business cycles?

  27. Gravatar of Arthur Arthur
    7. May 2013 at 07:21

    Scott, I think you should make the case that a central bank can target agregate wages.

    “W” being agregate wages and “C” being profits and rents and stuff.


    It’s clear that unless C is infinitely elastic to M, a central bank can target aggregate wages. Although it’s quite clear that this cannot be true on the long run, I think it’s not that clear that it cannot be true on the short run.

  28. Gravatar of mpowell mpowell
    7. May 2013 at 09:23

    I don’t follow German politics that closely, but aren’t a lot of those new jobs $2-3/hour type jobs with state assistance bringing the worker above the poverty line? And isn’t searching for such jobs non-optional for receiving state assistance now? You call it the biggest historical labor market success story, but I think you are overstating the accomplishment. Sure, there were headwinds, but a 4% employment gain in super low end jobs, which probably aren’t leading to a higher skill level workforce by the way, doesn’t represent much advance in human happiness (probably net negative on this score). I’d like to see some more productivity growth before I really start lauding labor market reform. I imagine there are a lot of middle class Germans happy to have cheap housecleaning, but it’s not shocking that isn’t that popular in Germany.

  29. 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 10:34

    […] fruit doesn’t fall far from the tree, of course; in his second post he attributes the “miracle” largely to “structural”/supply-side […]

  30. 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 10:34

    […] fruit doesn’t fall far from the tree, of course; in his second post he attributes the “miracle” largely to “structural”/supply-side […]

  31. Gravatar of Geoff Geoff
    7. May 2013 at 10:53

    The full quote regarding demand for commodities and demand for labor, from J.S. Mill, is worth reading (plus it is likely that nobody here would read the original source material if only a link were provided anyway):

    “Suppose, for instance, that there is a demand for velvet; a fund ready to be laid out in buying velvet, but no capital to establish the manufacture. It is of no consequence how great the demand may be; unless capital is attracted into the occupation, there will be no velvet made, and consequently none bought; unless, indeed, the desire of the intending purchaser for it is so strong, that he employs part of the price he would have paid for it, in making advances to work-people, that they may employ themselves in making velvet; that is, unless he converts part of his income into capital, and invests that capital in the manufacture. Let us now reverse the hypothesis, and suppose that there is plenty of capital ready for making velvet, but no demand. Velvet will not be made; but there is no particular preference on the part of capital for making velvet. Manufacturers and their labourers do not produce for the pleasure of their customers, but for the supply of their own wants, and having still the capital and the labour which are the essentials of production, they can either produce something else which is in demand, or if there be no other demand, they themselves have one, and can produce the things which they want for their own consumption. So that the cemployment afforded to labour does not depend on the purchasers, but on the capital, c I am, of course, not taking into consideration the effects of a sudden change. If the demand ceases unexpectedly, after the commodity to supply it is already produced, this introduces a different element into the question: the capital has actually been consumed in producing something which nobody wants or uses, and it has therefore perished, and the employment which it gave to labour is at an end, not because there is no longer a demand, but because there is no longer a capital. This case therefore does not test the principle. The proper test is, to suppose that the change is gradual and foreseen, and is attended with no waste of capital, the manufacture being discontinued by merely not replacing the machinery as it wears out, and not reinvesting the money as it comes in from the sale of the produce. The capital is thus ready for a new employment, in which it will maintain as much labour as before. The manufacturer and his work-people lose the benefit of the skill and knowledge which they had acquired in the particular business, and which can only be partially of use to them in any other; and that is the amount of loss to the community by the change. But the labourers can still work, and the capital which previously employed them will, either in the same hands, or by being lent to others, employ either those labourers or an equivalent number in some other occupation.

    “This theorem, that to purchase produce is not to employ labour; that the demand for labour is constituted by the wages which precede the production, and not by the demand which may exist for the commodities resulting from the production; is a proposition which greatly needs all the illustration it can receive. It is, to common apprehension, a paradox; and even among political economists of reputation, I can hardly point to any, except Mr. Ricardo and M. Say, who have kept it constantly and steadily in view. Almost all others occasionally express themselves as if a person who buys commodities, the produce of labour, was an employer of labour, and created, a demand for it as really, and in the same sense, as if he bought the labour itself directly, by the payment of wages. It is no wonder that political economy advances slowly, when such a question as this still remains open at its very threshold. I apprehend, that if by demand for labour be meant the demand by which wages are raised, or the number of labourers in employment increased, demand for commodities does not constitute demand for labour. I conceive that a person who buys commodities and consumes them himself, does no good to the labouring classes; and that it is only by what he abstains from consuming, and expends in direct payments to labourers in exchange for labour, that he benefits the labouring classes, or adds anything to the amount of their employment.

    “For the better illustration of the principle, let us put the following case. A consumer may expend his income either in buying services, or commodities. He may employ part of it in hiring journeymen bricklayers to build a house, or excavators to dig artificial lakes, or labourers to make plantations and lay out pleasure grounds; or, instead of this, he may expend the same value in buying velvet and lace. The question is, whether the difference between these two modes of expending his income affects the interest of the labouring classes. It is plain that in the first of the two cases he employs labourers, who will be out of employment, or at least out of that employment, in the opposite case. But those from whom I differ say that thig is of no consequence, because in buying velvet and lace he equally employs labourers, namely, those who make the velvet and lace. I contend, however, that in this last case he does not employ labourers; but merely decides in what kind of work some other person shall employ them. The consumer does not with his own funds pay to the weavers and lacemakers their day’s wages. He buys the finished commodity, which has been produced by labour and capital, the labour not being paid nor the capital furnished by him, but Coy the manufacturer. Suppose that he had been in the habit of expending this portion of his income in hiring journeymen bricklayers, who laid out the amount of their wages in food and clothing, which were also produced by labour and capital. He, however, determines to prefer velvet, for which he thus creates an extra demand. This demand cannot be satisfied without an extra supply, nor can the supply be produced without an extra capital: where, then, is the capital to come from? There is nothing in the consumer’s change of purpose which makes the capital of the country greater than it otherwise was. It appears, then, that the increased demand for velvet could not for the present be supplied, were it not that the very circumstance which gave rise to it has set at liberty a capital of the exact amount required. The very sum which the consumer now employs in buying velvet, formerly passed into the hands of journeymen bricklayers, who expended it in food and necessaries, which they now either go without, or squeeze by their competition, from the shares of other labourers. The labour and capital, therefore, which formerly produced necessaries for the use of these bricklayers, are deprived of their market, and must look out for other employment; and they find it in making velvet for the new demand. I do not mean that the very same labour and capital which produced the necessaries turn themselves to producing the velvet; but, in some one or other of a hundred modes, they take the place of that which does. There was capital in existence to do one of two things-to make the velvet, or to produce necessaries for the journeymen bricklayers; but not to do both. It was at the option of the consumer which of the two should happen; and if he chooses the velvet, they go without the necessaries.

    “For further illustration, let us suppose the same case reversed. The consumer has been accustomed to buy velvet, but resolves to discontinue that expense, and to employ the same annual sum in hiring bricklayers. If the common opinion be correct, this change in the mode of his expenditure gives no additional employment to labour, but only transfers employment from velvet-makers to bricklayers. On closer inspection, however, it will be seen that there is an increase of the total sum applied to the remuneration of labour. The velvet manufacturer, supposing him aware of the diminished demand for his commodity, diminishes the production, and sets at liberty a corresponding portion of the capital employed in the manufacture. This capital, thus withdrawn from the maintenance of velvet-makers, is not the same fund with that which the customer employs in maintaining bricklayers; it is a second fund. There are, therefore, two funds to be employed in the maintenance and remuneration of labour, where before there was only one. There is not a transfer of employment from velvet-makers to bricklayers, there is a new employment created for bricklayers, and a transfer of employment from velvet-makers to some other labourers, most probably those who produce the food and other things which the brick layers consume.

    “In answer to this it is said, that though money laid out in buying velvet is not capital, it replaces a capital; that though it does not create a new demand for labour, it is the necessary means of enabling the existing demand to be kept up. The funds (it may be said) of the manufacturer, while locked up in velvet, cannot be directly applied to the maintenance of labour; they do not begin to constitute a demand for labour until the velvet is sold, and the capital which made it replaced from the outlay of the purchaser; and thus, it may be said, the velvet-maker and the velvet-buyer have not two capitals, but only one capital between them, which by the act of purchase the buyer transfers to the manufacturer, and if instead of buying velvet he buys labour, he simply transfers this capital elsewhere, extinguishing as much demand for labour in one quarter as he creates in another.

    “The premises of this argument are not denied. To set free a capital which would otherwise be locked up in a form useless for the support of labour, is, no doubt, the same thing to the interests of labourers as the creation of a new capital. It is perfectly true that if I expend 1000L in buying velvet, I enable the manufacturer to employ 1000L in the maintenance of labour, which could not have been so employed while the velvet remained unsold: and if it would have remained unsold for ever unless I bought it, then by changing my purpose, and hiring bricklayers instead, I undoubtedly create no new demand for labour: for while I employ 1000L in hiring labour on the one hand, I annihilate for ever 1000L of the velvet-maker’s capital on the other. But this is confounding the effects arising from the mere suddenness of a change with the effects of the change itself. If when the buyer ceased to purchase, the capital employed in making velvet for his use necessarily perished, then his expending the same mount in hiring bricklayers would be no creation, but merely a transfer, of employment. The increased employment which I contend is given to labour, would not be given unless the capital of the velvet-maker could be liberated, and would not be given until it was liberated. But every one knows that the capital invested in an employment can be withdrawn from it, if sufficient time be allowed. If the velvet-maker had previous notice, by not receiving the usual order, he will have produced 1000L less velvet, and an equivalent portion of his capital will have been already set free. If he had no previous notice, and the article consequently remains on his hands, the increase of his stock will induce him next year to suspend or diminish his production until the surplus is carried off. When this process is complete, the manufacturer will find himself as rich as before, with undiminished power of employing labour in general, though a portion of his capital will now be employed in maintaining some other kind of it. Until this adjustment has taken place, the demand for labour will be merely changed, not increased: but as soon as it has taken place, the demand for labour is increased. Where there was formerly only one capital employed in maintaining weavers to make lO00L worth of velvet, there is now that same capital employed in making something else, and 1000L distributed among bricklayers besides. There are now two capitals employed in remunerating two sets of labourers; while before, one of those capitals, that of the customer, only served as a wheel in the machinery by which the other capital, that of the manufacturer, carded on its employment of labour from year to year.

    “The proposition for which I am contending is in reality equivalent to the following, which to some minds will appear a truism, though to others it is a paradox: that a person does good to labourers, not by what he consumes on himself, but solely by what he does not so consume. If instead of laying out 100L in wine or silk, I expend it in wages, the demand for commodities is precisely equal in both cases: in the one, it is a demand for 100L worth of wine or silk, in the other, for the same value of bread, beer, labourers’ clothing, fuel, and indulgences: but the labourers of the community have in the latter case the value of 100L more of the produce of the community distributed among them. I have consumed that much less, and made over my consuming power to them. If it were not so, my having consumed less would not leave more to be consumed by others; which is a manifest contradiction. When less is not produced, what one person forbears to consume is necessarily added to the share of those to whom he transfers his power of purchase. In the case supposed I do not necessarily consume less ultimately, since the labourers whom I pay may build a house for me, or make something else for my future consumption. But I have at all events postponed my consumption, and have turned over part of my share of the present produce of the community to the labourers. If after an interval I am indemnified, it is not from the existing produce, but from a subsequent addition made to it. I have therefore left more of the existing produce to be consumed by others; and have put into the possession of labourers the power to consume it.

    “There cannot be a better reductio ad absurdum of the opposite doctrine than that afforded by the Poor Law. If it be equally for the benefit of the labouring classes whether I consume my means in the form of things purchased for my own use, or set aside a portion in the shape of wages or alms for their direct consumption, on what ground can the policy be justified of taking my money from me to support paupers? since my unproductive expenditure would have equally benefited them, while I should have enjoyed it too. If society can both eat its cake and have it, why should it not be allowed the double indulgence? But common sense tells every one in his own case (though he does not see it on the larger scale), that the poor rate which he pays is really subtracted from his own consumption, and that no shifting of payment backwards and forwards will enable two persons to eat the same food. If he had not been required to pay the rate, and had consequently laid out the amount on himself, the poor would have had as much less for their share of the total produce of the country, as he himself would have consumed more.

    “It appears, then, that a demand delayed until the work is completed, and furnishing no advances, but only reimbursing advances made by others, contributes nothing to the demand for labour; and that what is so expended, is, in all its effects, so far as regards the employment of the labouring class, a mere nullity; it does not and cannot create any employment except at the expense of other employment which existed before.

    “But though a demand for velvet does nothing more in regard to the employment for labour and capital, than to determine so much of the employment which already existed, into that particular channel instead of any other; still, to the producers already engaged in the velvet manufacture, and not intending to quit it, this is of the utmost importance. To them, a falling off in the demand is a real loss, and one which, even if none of their goods finally perish unsold, may mount to any height, up to that which would make them choose, as the smaller evil, to retire from the business. On the contrary, an increased demand enables them to extend their transactions-to make a profit on a larger capital, if they have it, or can borrow it; and, turning over their capital more rapidly, they will employ their labourers more constantly, or employ a greater number than before. So that an increased demand for a commodity does really, in the particular department, often cause a greater employment to be given to labour by the same capital. The mistake lies in not perceiving that in the cases supposed, this advantage is given to labour and capital in one department, only by being withdrawn from another; and that when the change has produced its natural effect of attracting into the employment additional capital proportional to the increased demand, the advantage itself ceases.

    “The grounds of a proposition, when well understood, usually give a tolerable indication of the limitations of it. The general principle, now stated, is that demand for commodities determines merely the direction of labour, and the kind of wealth produced, but not the quantity or efficiency of the labour, or the aggregate of wealth. But to this there are two exceptions. First, when labour is supported, but not fully occupied, a new demand for something which it can produce, may stimulate the labour thus supported to increased exertions, of which the result may be an increase of wealth, to the advantage of the labourers themselves and of others. Work which can be done in the spare hours of persons subsisted from some other source, can (as before remarked) be undertaken without withdrawing capital from other occupations, beyond the amount (often very small) required to cover the expense of tools and materials, and even this will often be provided by savings made expressly for the purpose. The reason of our theorem thus failing, the theorem itself fails, and employment of this kind may, by the springing up of a demand for the commodity, be called into existence without depriving labour of an equivalent amount of employment in any other quarter. The demand does not, even in this case, operate on labour any otherwise than through the medium of an existing capital, but it affords an inducement which causes that capital to set in motion a greater amount of labour than it did before.”

    Understanding this passage is so crucial to understanding economics that it should be a part of every economics program in the country, indeed the world. Moreover, every economist ought grasp the full implications of it as well.

  32. Gravatar of Fed Up Fed Up
    7. May 2013 at 11:15

    “Fed Up, Definitely not the trade surplus.”

    So what happens if Germany’s trade surplus fell by reducing exports?

  33. Gravatar of Geoff Geoff
    7. May 2013 at 19:41

    Marcus Nunes has an excellent post that presents a different interpretation, one more favorable to NGDP targeting.

    “Scene one: Shows that NGDP and labor compensation ‘hugged’ each other along a rising trend from 1992 to late 2001. At that point NGDP ‘reduced speed’ and compensation stopped completely, even going a bit backward towards the end (late 2005).”

    Talk about seeing what you want to see in the data via the supremely scientific method of “eyeballing”!

    Try really, really hard, and you’ll see the “reduced speed.” I mean really really really try hard.

  34. Gravatar of ssumner ssumner
    8. May 2013 at 07:48

    Arthur, The Fed can target any nominal variable, because they can target the price level.

    mpowell, Surely the alternative (unemployment) is far worse. Not having a job is associated with deep declines in reported well-being.

  35. Gravatar of Ashok Rao Ashok Rao
    8. May 2013 at 09:57

    Geoff, I tried really really hard and it didn’t work. But when I tried really really really hard with two cherries on top it did.

    “mpowell, Surely the alternative (unemployment) is far worse. Not having a job is associated with deep declines in reported well-being.”

    What are your thoughts on direct employment/work contingent welfare benefits. If you don’t believe in “crude” Keynesianism, but accept long-run benefits to immediate employment would direct state/local employment financed by federal bonds be okay?

  36. Gravatar of Geoff Geoff
    9. May 2013 at 07:24

    “Geoff, I tried really really hard and it didn’t work. But when I tried really really really hard with two cherries on top it did.”

    Try staring at the clouds at look for a face shape. I mean try really really really hard. At some point, after thinking about a face for long enough, you’ll see one suddenly pop out in plain view!

    This is called MM intuition. Only the most intelligent can do it.

  37. Gravatar of Greece & Germany: Things tend to get worse before they get more worse | Left Outside Greece & Germany: Things tend to get worse before they get more worse | Left Outside
    14. May 2013 at 11:44

    […] Scott, the German labour market reforms of the mid-2000s took place against the more benign global and […]

Leave a Reply