A (positive) German shock?

Eurozone Watch has two articles about Germany and Italy that offer support for an optimistic view of the European economy. For a start, Sebastian Dullein argues that a comparison of Germany today and the US after the early 90s recession shows that Germany might be on the brink of a productivity surge. Dullein argues that labour productivity growth at the moment is being depressed by the re-absorption of the long-term unemployed, which also happened in the US in the early 90s. He quotes a figure of 7.6 per cent for productivity change (per employee, rather than per hour worked) in the metalworking industries (in Germany, a term that covers most of the industrial sector), which is positively stellar – after all, the US didn’t pass 2 per cent per-hour until 1998, well into the boom.

He also criticises Wolfgang Munchau for arguing (in essence) that there had been no structural reforms that accounted for productivity growth, and therefore that there was no growth. At this, I think I heard J.K. Galbraith’s ghost chuckle into his martini – it is indeed a fine example of all that is wrong with economics as a discipline that one can argue that we must all reform because there is a crisis, the evidence of that crisis being that one’s reforms have not been adopted.

An alternative argument would be that there was not all that much wrong with German firms in the first place. It is suggested that R&D spending is too low, but Dullein argues that it’s picking up. And anyway, their products can’t be that bad, as the rest of the world wants to buy German exports more than anything else. He also notes that there has been a wave of capital investment since 2002.

This possible German shock is already reverberating interestingly. Italy, for example, is experiencing better economic times, with growth picking up and strong industrial order books – especially on orders from France and Germany for capital goods. The growth is despite an increase in the tax take, with the result that the government is likely to have a chunk of change on hand. The OECD and the EU Commission would rather like to see that used to cut the monster public debt, still running at over 100 per cent of GDP. But the political situation might make that unlikely.

That might be the good news, though. When wasn’t the Italian government up to its eyes in debt? And it’s almost traditional that political turmoil in Italy is accompanied by good economic news. The difficult bit, though, is that Italian inflation is running somewhat slower than German – this implies, of course, an improvement in the terms-of-trade. Probably, Italy has done some internal disinflation, being unable to devalue – but this implies that wages have suffered relatively. The question is how to redistribute the benefit of the German shock without killing the golden goose.

21 thoughts on “A (positive) German shock?

  1. Hi Alex,

    I hate to be tiresome, but I don’t really go with this whole reading at all. And the principal reason I don’t agree with it should be obvious: the analysts you refer to have one simple thing in common, they don’t consider the demographic component in all of this to be at all relevant, in fact normally they don’t even mention it.

    Now it is impossible for me to go here into any depth on all of this, but the EU economies have a dirty little secret, serious imbalances exist, and the situation is getting worse rather than better. And *part* of the explanation for these imbalances (part, please note, not all) is differential demography.

    Some EU15 economies – Germany, Austria, Sweden, Finland – are structurally now forced to run current account services if they want to seriously grow (and keep their welfare systems afloat), since domestic consumption demand remains comparatively weak.

    Others – the UK, Ireland, Spain, Greece, France – have strong internal consumer demand, housing booms, and tend towards current account deficits.

    And, what a coincidence, the former group have a median population age of over 41, while the latter one are all under 40.

    Funny that no one seems very bothered about this evident correlation.

    On the German structural dependence I have written two recent lengthy posts (and here).

    Basically I have a number of posts in the pipeline for over the summer on the general topic of EU imbalances, but I just thought I’d alert you now to this alternative reading. Obviously, since economists themselves generally can’t agree about all of this, I appreciate it is very hard and frustrating for others to decide.

    “Dullein argues that labour productivity growth at the moment is being depressed by the re-absorption of the long-term unemployed, which also happened in the US in the early 90s.”

    This may well be the case, but you need to remember that there is a huge difference in labour force dynamics here, since the US labour force expanded at a record rate throughout the 90s (partly due to the arrival of migrants in unprecedented numbers, as is happening now in Ireland, Greece, and Spain), while the German population of working age is now declining (as I explain here, and can be seen from the chart here). Germany is like Latvia on the back-burner, at some stage they are going to run out of sufficient suitable labour to fuel the export growth.

    “Italy, for example, is experiencing better economic times, with growth picking up and strong industrial order books”

    I’m afraid you’ve got this one wrong, since Italy is now visibly slowing, and todays manufacturing numbers only tend to confirm this impression.

    Actually Italy is a rather good example of why this *isn’t* all about demographics, since according to the age structure of the population Italy should be export-lead and running a hetfy CA surplus. For a variety of (non-demographic) reasons, like the inability to reform fast enough, it isn’t able to achieve this, hence the horribly low level of long-run trend growth, which means, of course, that it is going to have real problems putting that enormous public debt on a sustainable path.

    And talking of government deficits, demography and Wolfgang Munchau, all of this now seems likely to come sharply into focus during the Portuguese presidency.

    In terms of deficit dynamics, it is plain that France has far more short term leeway, than say Germany does. So Sarkozy wants, yet one more time, to break ranks on the Stability Pact. Wolfgang Munchau draws attention to the possible political implications of what is happening, although since he seems to be a fully paid up member of the “demography doesn’t matter” club, the underlying significance of what is happening tends to escape him (and it is indeed an irony that it is poor Portugal, who was turned into the sacrificial lamb of the old version of the Pact -and still hasn’t recovered from the consequences – while Germany and France effectively got off stock free should be holding the Presidency at this point).

    It is important to realise here that France has effectively a 20 year demographic time advantage over Germany at this point – that is France is now where Germany was say in 1985 in ageing terms. So France does have a lot more room for manoeuvre, and while they obviously need to address long term structural issues in their pension and medical funding, they aren’t under the same short term pressure to run a balanced budget (very similar to the US in many ways all this). I don’t doubt these are the sort of points Sarkozy will be making at the forthcoming meeting.

  2. Edward, you dwell on “the demographic component…” and then mention “housing booms” as if they were something positive under the circumstance of a declining population.

    How do you explain this?

  3. Writing from a US perspective, I have to say that I think that Dullein’s piece has completely misinterpreted the US productivity story in the 1990’s. While welfare reforms did in fact reduce the number of individuals on welfare, I am positive that the number of persons who moved from long-term unemployment to employment was not “in the millions”. This link to a study by the US Bureau of Labor Statistics on long-term unemployment directly contradicts Dullein’s assertion that “The U.S. recovery after the 1990/1991 recession started out with very strong job growth”. The study states that

    “The most obvious reason for
    the slow improvement in longterm
    unemployment following the
    two most recent contractions was
    the relatively slower pace of job
    growth. Following each of the recessions
    of the mid-1970s and
    early 1980s, employment rose by
    1.5 percent within a year. In contrast,
    employment was virtually
    unchanged in the year following
    the 1990–91 and 2001 recessions.
    As shown in the accompanying
    table, even by the time long-term
    unemployment had started to decline,
    employment had risen by 1.0
    percent or less. Also, in contrast
    to the recessions of the mid-1970s
    and early 1980s, the employment-to-
    population ratio continued to
    decline far longer following the recessions
    of 1990–91 and 2001.”

    Further, rapid growth in US productivity later in the 1990’s was largely attributable to heavy investment in information technology which was driven by rapidly increasing capabilities of information technology and the massive increase in Internet-related business and consumer activity. I don’t foresee any kind of major technological shift occurring in the next few years that would be centered in Germany, particularly considering that German firms have been scrimping on research and development, as Dullein noted.

    With respect to housing booms in the countries where the median age is under 40, the booms were created by reduced interest rates and easier credit conditions that allowed younger households to purchase homes. Now, it is clear that housing markets have oversupply in these countries, but domestic growth was aided by the booms. In the aged countries, there is little chance that GDP growth will be aided by housing construction, as the number of households in these countries will remain flat or begin to shrink soon.

  4. Hi Suvi,

    “Edward, you dwell on “the demographic component…” and then mention “housing booms” as if they were something positive under the circumstance of a declining population.”

    Well look, this is a fairly complex problem, and I can only really start to touch the surface in a relatively short comment.

    The fist point to get hold of is that there is a big difference between housing booms, and housing bubbles. Housing booms may be quite normal and positive (due to their impact on eg consumption) from an economic point of view, while bubbles most certainly are not desirable, since when they burst the consequences are normally not pleasant.

    The difficult thing here is to identify bubbles before they burst. I say this, since for all the talk about this topic in recent times we have not yet had one case of a supposed bubble which has burst. What we have seen are property prices in some countries rising very rapidly over a short period of time, and then undergoing some form of correction. The UK, the US and the Netherlands (in the late 90s) would be examples of this kind of phenomenon. The correction may be a little bit more or a little bit less painful, but in no case can they be compared with the impact of a bursting bubble.

    Now inside the eurozone we have three economies who have had rates of growth in construction activity and house prices which some have argued may constitute bubbles – in Spain, Ireland and Greece. Indeed part of the justification for the recent bout of ECB interest rate tightening would be that they fear the development of housing bubbles in these countries.

    I personally (I live in Spain) used to think that a bubble was developing here. I have now changed my mind. Basically to decide whether there is a bubble you need to have it burst, and whether or not this (or a very substantial correction in house prices) takes place depends on two things:

    a) the future path of long term interest rates inside the eurozone
    b) the rates of migrant flows into these countries going forward.

    Now, oddly enough, (b) seems to depend on (a). In fact Claus Vistesen and I are doing some research on precisely this topic at this moment. The reason (b) depends on (a) is that the level of housing activity depends on affordability to some extent, and affordability depends on the principal cost for the first time buyer, which is constituted by the level of interest payments, and the level of migrant inflow depends – in those economies structurally dependent on internal consumption growth – on the level of housing activity, and indeed this latter seems to correlate very well with inward migrant flows.

    Indeed this correlation seems to be so good that the US Economist Dawn Mclaren at the University of Arizona has proposed that cross border crossings into the US could be considered now to be a “leading indicator” for the future level of US economic activity. I have a post on Demography Matters on precisely this point here. Initial findings from Claus and I tend to confirm this in the case of Spain, Greece and Ireland.

    Now, at this point, another thing needs to be said. Eurozone economies seem to be developing different structural characteristics. This is my whole beef with the way Sebastian and many others are treating what is happening, since they seem to be ignoring this issue. Now the most “normal” trend for societies moving from an industrial to a services profile is a growing role for the housing sector, and in construction activity and professional and financial services associated with the housing sector. This growing role for housing is associated with increasing levels of indebtedness and with growing consumption shares in the economy via the well known wealth effect. As I say, in the Eurozone the three economies which have moved furthest in this direction have been Greece, Ireland and Spain, but France has increasingly been moving in this direction (outside the zone the UK and Denmark would also seem to be examples). Now the curious thing is that what all the above mentioned economies have in common is that they are comparatively young (median ages <41 ). What is more the feedback process from pulling in large numbers of migrants (who normally have a younger age profile) has meant that median ages in these societies have been either stationary or declining slightly.

    To come back to whether or not Spain, Greece and Ireland will have a burst bubble on their hands at some point (rather than a price correction), as I say this depends in large part on the future course of interest rates. If we look at what is happening on the global liquidity front, there would seem to be no good reason to anticipate this outcome in the short run, since there is a relatively plentiful supply of savings (and part of the explanation for this may be demographic). Also, and from where I am sitting, there may be reasons for thinking that the ECB will not be able to sustain interest rates at their present "high" for that long, and in particular the turn of events in both Germany and Italy may well put pressure on the ECB to reduce at some point in the not too distant future (this is, of course, a personal view). So all in all I have the impression that the show may well be set to go on.

    And now (finally) to come to your question:

    "as if they were something positive under the circumstance of a declining population."

    This needs, I'm afraid some rephrasing. In the case of the countries I have been talking about population is not yet declining, indeed in some it is increasing rapidly. This has, as I say, a direct impact on the evolution of median age and this is what I would consider to be positive.

    Obviously this ageing process cannot be put off indefinitely, but it can be slowed down, and the more time we have to address the issues like pension and health system reforms the better. We also need to steadily get unhooked from Paygo pension systems, and here again time is useful.

    Finally, I would point out that all of this is very contested from some quarters, and especially by economists in those societies who are most subject to rapid ageing (Germany, Austria, Italy, Finland). These societies as far as I can see are now structurally (demographically) unable to generate the kind of housing and consumption booms on the level I have been talking about (this would be part of the reason they need to export to grow dynamically, and they do need to grow substantially to be able to support the increasing elderly dependence share. Now since there seems to be a consensus in the most ageing affected countries that all of this is relatively benign, all I can say (from the comfortable vantage point of Spain) is, after you gentlemen, you go first.

  5. Alex,

    Sorry if I have hijacked this a bit. I think the main point you are making is perfectly valid: German firms (especially in the export sector) have had a quite remarkable productivity performance in recent years, and this is obviously a huge plus.

    And Italy was (in 2006) benefiting significantly from the kind of positive feedback mechanisms you draw attention to. Really I am trying simply to contextualize this in a rather longer term perspective.

    I wouldn’t be so optimistic about the course of the Italian deficit even this year, although it is obvious that global growth is still incredibly strong.

    Taking a hard look at German exports the impact of the growth that is taking place in Eastern Europe is most striking. So Germany will be solid as long as East Europe is solid, and who knows for how long that will be.

    As and when the current “long boom” draws to a close, Germany and Italy will inevitably need to fall back on internal consumption for support (since the normal fiscal mechanisms in both cases will not be so easily available) and this is what I am trying to draw attention to. Lets just hope 2007 continues to be “another exceptional year”.

  6. I’m still not seeing how a stagnant or decreasing work age population *automatically* translates into decreased per capita income.

    To give two very different examples: both Russia and Singapore have seen their working-age populations decrease in recent years. But both have enjoyed several years of steady growth.

    Arguably Germany is indeed forced to run a current account surplus. Since Germany has no problem doing so, this doesn’t seem like a major issue. The world loves to buy German.

    “Germany is like Latvia on the back-burner, at some stage they are going to run out of sufficient suitable labour to fuel the export growth.” — Why?

    A shrinking labor pool does pose certain challenges. To name just one, you have to keep relentlessly re-investing in order to keep productivity up. But this is exactly what Germany is doing.

    Obviously exports can’t grow forever. But they can stabilize at a very high level for long periods of time. Labor is just one part of the equation; productivity and structural incentives count for as much, or more.

    I can see a number of failure modes for Germany. But so far, they seem to be holding up pretty well.

    Doug M.

  7. “Some EU15 economies – Germany, Austria, Sweden, Finland – are structurally now forced to run current account services [sic] if they want to seriously grow (and keep their welfare systems afloat), since domestic consumption demand remains comparatively weak.

    “Others – the UK, Ireland, Spain, Greece, France – have strong internal consumer demand, housing booms, and tend towards current account deficits.

    “And, what a coincidence, the former group have a median population age of over 41, while the latter one are all under 40.”

    Hmm. Greece doesn’t belong in that second group; it has a median age of 40.5. And Spain and France are both pretty close to 40 — 38.8 and 39.5, respectively.

    I don’t see how this makes them dramatically different from Sweden (40.6) or Austria (40.4). And I note that Spain and France are both older than much-maligned Latvia (39.1).

    Also, the correlation between age and current account seems pretty loose. The 20 oldest countries in Europe include a dozen countries that are running large current account deficits. Eyeballing the numbers, I’d say there’s either no correlation, or it works t’other way.

    Doug M.

  8. “Now the curious thing is that what all the above mentioned economies have in common is that they are comparatively young (median ages <41 )."

    There are only three countries IN THE ENTIRE WORLD that have median ages >41.

    Doug M.

  9. “Taking a hard look at German exports the impact of the growth that is taking place in Eastern Europe is most striking. So Germany will be solid as long as East Europe is solid, and who knows for how long that will be.”

    Um. Eastern Europe — here defined as the EU-8 — absorbs less than 10% of German exports.

    Throw in Russia, the Western Balkans, and the CIS, and it’s still just barely 10%.

    German exports to the EU-8 are hugely important to the EU-8, because their economies, combined, are less than a third as big as Germany’s. But If everything between Moscow and the Oder vanished tomorrow, Germany would still be running a trade surplus.

    Doug M.

  10. Hi Doug,

    Quite a string of questions. I’ll try and answer them one by one.

    Let’s take median ages first.

    The first problem seems to be the lack of really reliable data on this topic at present. In fact due to this problem Claus is hard at it preparing some excel sheets right now. We should have the data ready some time during the summer. In the meantime, the best source of data I have found which is systematic is the CIA 2007 edition of the world factbook which people can find here. (If anyone finds a better source please let us know, since the CIA data is often highly unreliable, which is why we are going to do the calculations for ourselves. In particular there doesn’t seem to be any systematic dataset which tracks how median ages have changed over the years. I guess up to now no-one has been interested).

    “Hmm. Greece doesn’t belong in that second group; it has a median age of 40.5.”

    Aha! Well spotted. It was precisely thinking about the Greek case that lead me to put the rough and ready working number at 41. Having said that I want to check what exactly has been happening to the Greek median age over recent years, since Greece has had substantial immigration and this – as I am arguing – brings the median age down, and the age of resident non-citizens may or may not be included in the CIA data set. I also want to have a look at what happened to US median age in the 90s, but with so many illegals this may be difficult to track. My back of the envelope guess is that it actually dropped.

    But in general terms, look, this isn’t an exact science yet. It is only a rough and ready working hypothesis, which needs a lot of refining. In particular, demographics are clearly not the only factor at work. So obviously we have a large grey area. This is why I stuck my neck out at 41. I am prepared to advance a falsifiable hypothesis here, and that is that a society with a median age of 41 and over will NEVER have a substantial housing boom of the order of those that we have recently seen in Autralia, the UK, Spain, the US, Ireland, New Zealand, Iceland, Greece. This is, I think testable.

    By September Claus and I hope to have a much tighter definition of “substantial” here, but as you will appreciate all of this is currently very much work in progress. We will also be able to examine all of this much more closely when we have some sort of correlates on size of housing sector, real interest rates, migrant flows, and impacts on median age. There seems to be a very large and systematic machine at work somewhere in the middle of all this, and it needs identifying.

    “And Spain and France are both pretty close to 40 — 38.8 and 39.5, respectively.”

    Yep, well this is just what I am saying, they are under 41, this is what seems to be more or less critical. This is definitely a non-linear process. There seem to be significant underlying feedback mechanisms, and this is why the traditional models haven’t been able to capture any of this. There is a “partgage des eaux” out there somewhere, and if you are on one side you slide one way, and if you are on the other, you go down the other side (given one and the same interest rate from the central bank, this is the beauty of the eurozone as an experiment).

    Identifying exactly where the “partage” line is (or whether in fact it is a region rather than a point, which is much more plausible) is going to be the tricky part.

    Look, this isn’t voodoo science or something, there is a solid background of economic theory (Modigliani’s life cycle model of saving and consumption) to back up the empirical finding. The trick is to be able to see that you can apply this theory to populations as median ages rise, and then to calibrate how it all works.

    “There are only three countries IN THE ENTIRE WORLD that have median ages >41.”

    Yes, this is the point, this is why I feel I am on solid ground:

    Japan 43.5
    Germany 43.0
    Italy 42.5

    But I don’t think you read my syntax right. I said <41, which means that the other group would be > or = 41.

    Then we could also add:

    Austria 41.3 (according to the CIA 2007)
    Belgium 41.1
    Sweden 41.1
    Finland 41.6

    But the big point you seem to be missing is that all these median ages are set to rise systematically, so this group is going to grow steadily. This is a whole demographic transition which is set, in principle, to run for at least the next 50 years. The only thing which can really work in the other direction is immigration, and since the housing sector seems to be so important in the modern migration process, this helps explain the importance I am attaching to this.

    Having said that Germany, Japan and Italy do seem to be cut adrift from everyone else in ageing terms, so it is precisely in these countries that all the various differing opinions are about to be tested.

  11. Ok, now lets look at the EU 8:

    “And I note that Spain and France are both older than much-maligned Latvia (39.1).”

    Look, first off, I’m not sure where you get the idea I am maligning Latvia from. I am CONCERNED about Latvia, not trying to give them a kick. I think the people from Latvia itself who have contacted me in recent days appreciate this. The think is to devise a viable exit strategy for them, and my proposal is to loosen the labour supply constraint, by encouraging inward migration (like, for example the Czech Republic and to some extent Estonia are currently doing).

    But on the broader point, I think there is something which I may not have made clear which needs to be made clear. When I talk about Germany, Japan etc and housing booms, I am talking about mature economies. The EU8 are still classified in economic terms as emerging economies and understandably so. That means that they have fundamentally different structural characteristics at this point to the societies I am trying to analyse here.

    But they will be hit by all this at some point, since many of these societies are already quite old (and may even be older than the data suggest, since they have had out- rather than in-migration).

    Bulgaria 40.9
    Croatia 40.6
    Slovenia 41.0

    At this point I have no “stylised facts” to offer about the future path of these economies, except to say that the whole situation needs monitoring, to be so old and to have such low per capita GDP is obviously not good news.

  12. OK, now let’s look at Germany.

    “I’m still not seeing how a stagnant or decreasing work age population *automatically* translates into decreased per capita income.”

    Maybe I’m being a pedant here Doug, but I sometimes feel I am having to argue on shifting sand. I mean where exactly do I talk about per capita income in this comments thread, and much more to the point, where do I use that little phrase *automatically*? I don’t recall saying this, nor would I agree with anyone who said it.

    What I have said is that Germany has become structurally dependent on exports, and this fact seems to be related to Germany’s median age. Per capita living standards can rise or fall, depending.

    What we can say is that it is a lot easier for a society to raise per capita incomes when the workforce constitutes a rising share of the total population (the demographic dividend period) than when it constitutes a declining share (the demographic penalty period).

    The only think which can offset the downward path of per capita incomes, is, as everyone recognizes, productivity. I would say the jury is still out on whether productivity can be maintained as populations age.

    But the important point here is the structural dependence on exports, and more to the point on GROWTH in exports to get growth in total GDP. This means that the whole economy becomes unduly dependent on one sector, and that it becomes increasingly difficult to maintain the dynamic.

    And the economy needs to grow to maintain the living standards of the rising dependent population.

    “Arguably Germany is indeed forced to run a current account surplus. Since Germany has no problem doing so, this doesn’t seem like a major issue. The world loves to buy German”

    Yes, but the problem isn’t quite adequately specified putting things this way. The issue is that for the German economy to grow exports need to GROW, and keep growing, at a certain given rate ‘x’% pa which still needs to be specified, if Germany is to find a sustainable path.

    In the present global expansion this isn’t proving TOO difficult (although what Germany and Japan are doing here are really quite remarkable achievements, it needs to be said. Cycling uphill is never easy).

    Let’s have a look at what this means. Exhibit 1 is this chart which shows the rates of German export growth 1995 – 2006. Now what can be seen at first glance is that Germany has recently been sustaining very high rates of growth in exports, 16.6% in 2006 for example, and 9.1% in 2005. What is also interesting to note is that back in 2000 German exports grew at a rate of 21%, while in 2001 exports only rose by 0.8%, and guess what, Germany then dropped off into recession.
    Now Claus kindy prepared this graph for me, (lets call it exhibit 2), and it tracks German GDP growth and German export growth, and I think it is clear that the two things are pretty closely connected, which leaves us with the big question, what happens to Germany (and Japan) when the present wave of global growth subsides?

    “Obviously exports can’t grow forever. But they can stabilize at a very high level for long periods of time.”

    Well this is the point, so what happens at the point when they can’t. No-one seems to be talking about this.

    “Um. Eastern Europe — here defined as the EU-8 — absorbs less than 10% of German exports.”

    I think we are back here to the point that this is all about GROWTH, and it just happens that the EU 8 are accounting for a very significant proportion of the additional export growth which Germany has been getting.

    I recomend reading this (pdf) paper by some IMF economists (and which seems to be the most authoritative piece of work to date on the German export phenomenon), since I got the data from them.

    Now if we look at this chart (exhibit 3), we will see that over the period from 1995 – 2005 exports to the EU8 grew at an annual rate of 12.5%, and that these countries, as you suggest, account for 8.6% (on average) of total exports. Now this is a decade long average, but it is clear that the share of export growth is weighted especially towards recent years.

    Another way of looking at this is to compare the EU8 share, with the China share, which was only 2.7% of total exports over the period.

    If you still doubt the importance of this phenomenon you could try looking at this chart (exhibit 4) which shows the top twenty sources of German imports and destinations for exports in 2006, and contains I think some surprises. Like, for example, the fact that exports to the Czech Republic alone were not that far short of exports to China despite the huge difference in the relative sizes of the countries. Exports to Poland were in fact greater than exports to China.

    Some indication of the scale of importance of Eastern Europe can be found from the latest edition of the BIS quarterly review (summarised here by Bloomberg), which informs us that:

    “Investment and lending have boomed in eastern Europe, pushing up wages and spurring consumer spending, as eight nations joined the European Union in 2004 and a further two followed this year. More than 60 percent of new credit to emerging markets went to European countries in the last three months of 2006, the BIS said today in a quarterly report.”

    This credit share is enormous, and gives us an idea of the magnitude of the phenomenon we are dealing with.

    OK, finally this question:

    “Germany is like Latvia on the back-burner, at some stage they are going to run out of sufficient suitable labour to fuel the export growth.” — Why?

    The biggest part of the explanation is that the version of the German unemployment story you are reading in the press is quite simply spin. Of course, unemployment in Germany is falling because new jobs are being created, but it is also falling because older people are leaving the labour force, and this latter component seems to be gaining in strength.

    If we look at the latest employment statement from the German Statistics Office we find “The adjusted number of people out of work fell 37,000 to 3.82 million.” (June), But if we look at the latest employment creation data (May) we find that there were a total of 14,000 new jobs created, so that the number of people leaving seems to be the bigger share.

    Here is the chart with the latest employment data (exhibit 5). Basically there are now 3.69 million people in Germany looking for work. This number is down by 1.5 million from the peak of unemployment in 2005. Obviously you need to think about trend and cycle here, but (without doing the calculations at this point) you need to ask with people leaving the labour force at the present rate just how long will it take to get below 2 million, and then 1 million, and then where do you go. No-one seems to be thinking about this at all, and yet we find a similar situation developing in Japan, and in Italy.

  13. Germany has a structural advantage, though – its economy is based on export-oriented manufacturing. Which these days is capital-intensive. It’s whizzy-joy-joy US/UK call centre bollocks that’s labour-intensive.

    That would also be the sector that defined economic history since the 18th C; the UK, then Germany, then the US, and now…?

    “There are more things, sir, in heaven and earth than are dreamt of in your philosophy..”

  14. Hi Alex,

    First off, sorry again if I have sort of hijacked discussion here (aided and abetted by Doug) and turned this into a sort of master class on contemporary German macro issues.

    In the end though this isn’t necessarily a bad thing, since many of the issues here are going to be central to the debate which is likely to emerge as we get into the autumn and winter. For example, the ECB is busy raising rates. Domestic consumption in Germany is congenitally weak, so how advisable is this? What is the impact we might see on German consumption moving forward of the combined 3% VAT rise, and the higher costs of borrowing? Both of these make it less likely that the domestic consumption component can pick up slack coming from any slowdown which might occur in export momentum. Thankfully US demand is holding up reasonably well despite the drop in housing activity, so at the moment this is one less worry for them, but this is also why I think we need to follow carefully events in the EU8, since any “correction” in activity there would be noted.

    But really what I am trying to do in the comments here is draw attention to a number of contemporary macro “urban legends” about the path of the German economy. Some of these are:

    1/. The Goldilocks, self-sustaining recovery idea. This is not a NORMAL recovery in Germany, but is extremely skewed towards export growth, this development is significant, yet in the popular press (including places like the Economist) this is barely recognised or discussed.

    2/. The rapidly falling unemployment story. Unemployment IS falling as a result of job creation, but as I am arguing this process is a much more complex one than people seem to recognize. There is good news AND bad news here, but you only get to read the good news in most of the press articles.

    3/. The wage inflation scare. What is most striking about all of this labour tightening that has been happening in Germany is that it is not producing the long forecast wage inflation. In fact y-o-y real german wages were only up 0.1% in Q1 2007 over Q1 2006, and this despite all the productivity gains we are talking about. We have the same story in Japan (where real wages are falling) and in Italy, as you note, wages are now starting to rise more slowly than prices. So why should this be happening as labour markets tighten? This is curious isn’t it?

    4/. The idea that Germany is outsourcing low value added work, and keeping high value added at home. The reality behind German outsourcing is much more complex than it seems. As this chart here – which was produced by the IMF economists – shows, the domestic value added component has been a steadily declining share of German industrial output in recent years.

    So something strange is happening here, especially since, as you note, German industry is becoming more capital intensive. Perhaps the answer is to be found, as Munich economist Dalia Marin argues, in what is happening to none-production workers. Based on the research of her PhD student Alexander Raubold – which was a study of the impact of the human capital shortage both Germany and Austria were experiencing given their declining young cohorts and weaknesses in the education system – Marin has been arguing that – at the level of non-production workers – German MNCs have been keeping the skill premium down by outsourcing a comparatively high share of the more skilled work from Germany (where the relevant labour is in increasingly short supply) to East Europe which has a relatively high quantity of the needed skills, and – as we all know – these are available relatively cheaply. Marin calls this process Maquiladoras in reverse.

    Germany seems to have had the singular bad luck to have been experiencing these declining young cohorts at precisely the moment when demand for more skilled labour as a share of the total (the shift towards the knowledge intensive economy) was growing rapidly.

    This graph shows reasonably well what has been happening here, and this outsourcing may well help us understand how such a dynamic expansion has been possible in the export sector at the same time as the relative value of German wages (in comparison with other EU states) has been falling.

    One example of the implications of all of this is to be found in the news today that Spain is now to pay a 2,500 euro premium for every new child (whether born to the parent or adopted internationally). The impact of such payments on long term fertility is, of course, unclear, but the central point is that – thanks to all the immigration in recent years – Spanish finances are in a position to permit such expenditure increases.

    Summing up: Germany is indeed undergoing something of an economic miracle, but sometimes miracles alone just aren’t enough.

  15. “a society with a median age of 41 and over will NEVER have a substantial housing boom of the order of those that we have recently seen in Autralia, the UK, Spain, the US, Ireland, New Zealand, Iceland, Greece. This is, I think testable.”

    It’s testable; but since only a handful of nations have median ages over 41, it may take some time before we can say it’s been tested.

    I will say, on one hand, that there’s a superficial plausibility here: you’d expect older societies to be less prone to a housing boom. I wouldn’t be surprised if this turned out to be true.

    On the other hand, you used the word “societies” instead of “nations”. I think that was vagueness on your part, but if it’s not, you’re getting into some murky waters. Because the housing boom in the US stretches over some pretty elderly regions.

    There’s no US state (yet) with a median age over 41.0. Right now the three oldest are Maine (40.7), Vermont (40.4) and West Virginia (40.3). All three of those states have shared in the national housing boom. (Casual googling suggestst that West Virginia has had a relatively smaller boom, while Vermont housing prices have grown even faster than the national average.)

    Going down to the 39-40 range, we find Montana, Florida, Pennsylvania and New Hampshire. Florida, the fifth oldest state in the US (39.3) is ground zero for the housing boom, with both prices and number of housing starts growing even faster than the national average.

    A significant element in the Florida superboom is the construction of… retirement communities. Hmm.

    Would you view a housing boom in a >41 American state as falsifying your theory? Because by the next census — 2010/11 — there will probably be a couple of them.

    Doug M.

  16. “Then we could also add:

    Austria 41.3 (according to the CIA 2007)
    Belgium 41.1
    Sweden 41.1
    Finland 41.6”

    Hum. Belgium saw very sharp increases in both housing prices and housing starts in this decade. Whether it was a “boom” or not is a question of definition, but we’re talking price increases of ~~10% yoy, sustained over 5 years or more. To me that looks at least boom… ish.

    “Bulgaria 40.9
    Croatia 40.6
    Slovenia 41.0

    At this point I have no “stylised facts” to offer about the future path of these economies, except to say that the whole situation needs monitoring, to be so old and to have such low per capita GDP is obviously not good news.”

    Here we agree. For the EU-8 plus Romania and Bulgaria, I view the next 15 years as a race between growth and demographics.

    Note that these countries still have a little time left. Broadly speaking, birthrates remained high across Eastern Europe through the 1960s, fell slowly through the 1970s and 1980s, and then crashed hard after 1990.

    There are exceptions to this pattern. Bulgaria and Slovenia had their birthrates fall faster and earlier, which is why they’re older now; Albania and Kosovo are about 10 and 20 years behind the curve, respectively, so they’re younger now. But they are exceptions, and the general pattern is clear.

    What this means is that for most of the EU-8, dependency ratios are still low. There are a lot of adults between 30 and 60, but not a lot of kids, teenagers, or old folks.

    However, these countries are about to be hit by a double whammy: the postwar baby boomers will start retiring, and the “empty cohorts” born in the 1990s will start (not) entering the job market. So dependency ratios are going to start rising rapidly after 2010.

    Additional wrinkle: right now most of those countries have low TFRs. But after 2010, they’re going to have low TFRs /and/ low numbers of women in peak childbearing years. This means they have another five or ten years to raise TFRs and dig themselves out of the demographic hole that they’re in. If they don’t, it’s going to get much much harder.

    Anyway. I’m not too optimistic for most of the EU-8+2. Romania, for instance, has seen rapid growth since 2001, but it’s still running a massive current account deficit. That’s just not going to be sustainable.

    I suspect one ugly solution will be to just allow the elderly to be immiserated.

    But Germany, Finland et al are something else entirely. Not only are those mature economies, but they’re highly efficient, very productive, and export powerhouses.

    Long-term net exporter -> lots of foreign investments -> macroeconomic air bag against the demographic car crash.

    In the case of Germany, you keep saying that German exports must grow. Must they? They’re already running a trade surplus of ~$100 billion/year. By the middle of the next decade, even without growth, they’ll own a couple of trillion dollars of foreign assets. That’s going to pay a lot of pensions.

    IOW, I agree that a demographic sea change is coming, but it’s going to have very, /very/ different effects on different economies.

    Doug M.
    Doug M.

  17. Hi again doug,

    “It’s testable; but since only a handful of nations have median ages over 41, it may take some time before we can say it’s been tested.”

    Yes, I agree. We now need to be rather patient.

    “Hum. Belgium saw very sharp increases in both housing prices and housing starts in this decade. Whether it was a “boom” or not is a question of definition,”

    Well spotted! I would say that apart from the Greek case we have already mentioned, the two countries which are causing me to scratch my head most at the moment are Belgium and Portugal.

    Of course, since I am not trying to be excessively rigid here, but simply to indicate a general area for future research, there is still a lot of room for refining the argument. I think sometimes the way to move forward is simply to ask yourself interesting questions (knowing of course that you can make mistakes).

    On the US, really I’m going to take the scoundrel’s path and take recourse to “US exceptionalism”. Which is another way of saying that I don’t know enough about how the US internal market works to feel I have anything very useful to say on this.

    It is part of the received wisdom in the housing literature that housing markets are essentially local (what I am trying to do with median ages is find a convenient proxy for something, if I find one I could then start to think in more detail about what exactly it is a proxy for). Japan house prices have been either stationary or trending downwards for over a decade now, but downtown Tokyo prices have been booming over the last 18 months I think.

    One trend is the move towards certain given urban areas. Another is the creation of a kind of global city network in some key city centres across the globe. In these cases property prices may accommodate more to those of the global peers than they do to local economic dynamics.

    The same with the elderly population, and the migration towards the sun.

    The property boom dynamics in Barcelona, Madrid and Valencia have been very different from those in Almeria, where the boom has been fuelled by a retiring population from the north of Europe moving south. They estimate that there are now 750,000 Brits alone who spend some of the year in a Spanish home they own.

    So I suspect, and I only say suspect since I don’t know, that some of this may be going on in the US, especially in Florida.

    On the other hand if you look at the median age for the US as a whole, there is still going to be plenty of life left for housing expansion once the current correction passes through the system.

  18. On Eastern Europe, it seems we agree in substance:

    “For the EU-8 plus Romania and Bulgaria, I view the next 15 years as a race between growth and demographics.”

    Yep. Agreed. The thing which worries me most is what happens when the financial markets wake up to all of this.

    Housing is already slowing in the Baltics, while unemployment is dropping incredibly rapidly in Poland (almost “too rapidly”).

    “Note that these countries still have a little time left.”

    Yep, but if any of them manage to dig themselves into a temporary hole, it may cost them a lot to climb out again. And hardly a day passes at the moment when I don’t find myself surprised by the rapidity with which everything is happening there.

    “Albania and Kosovo are about 10 and 20”

    Yep, incredibly Albania could find itself as a country which is very much in vogue in the decade 2010 – 2020.

    “However, these countries are about to be hit by a double whammy: the postwar baby boomers will start retiring, and the “empty cohorts” born in the 1990s will start (not) entering the job market. So dependency ratios are going to start rising rapidly after 2010.”

    “In the case of Germany, you keep saying that German exports must grow. Must they? They’re already running a trade surplus of ~$100 billion/year. By the middle of the next decade, even without growth, they’ll own a couple of trillion dollars of foreign assets. That’s going to pay a lot of pensions.”

    Good point. Really I haven’t gone into the implications of this part yet, in the same way as I am avoiding thinking about what happens if and when Germany and Japan ever where to reach a negative savings stage. All of that is so far out in the future that the topic becomes very speculative. I am trying to focus on what is happening now.

    And here you need to think about who own what, and who the Federal government can tax. Basically they can’t keep piling 3% on the VAT rate every other year. So they need to grow so that revenue can grow sufficiently quickly to meet obligations to the rising dependent population. At the moment we can see what happens when they have a good year, the government has funds, but you also have to think about what happens when they have a bad year.

    So what I am focused on here is the sustainability of government finances, over a fairly short time period – say between here and 2020 – and how the export dependence can impact on this.

    “IOW, I agree that a demographic sea change is coming, but it’s going to have very, /very/ different effects on different economies.”

    I agree. And I hope, indeed I imagine we all hope, that Germany will be one of the winners here, but we would not be rational animals if we did not ask ourselves the hard questions about whether our hopes are justified.

  19. Hi again,

    “Whether it was a “boom” or not is a question of definition”

    Yep, as I say I am actively working on this at the moment. The best definition of a “boom” I have found to date is the one I reproduce below from OECD economists David Rae and Paul van den Noord (who are really pretty on the ball about all this):

    “Between 1960 and 2004, 49 residential construction booms have occurred in 23 countries for which data is available. A boom is defined (rather generously) as a rise in the level of real per capita residential investment of at least 15% over a five year period. In order to avoid identifying false peaks and data blips, a peak is defined as the highest point in a window of the preceding four years and the subsequent three years. By construction, the latest peak that can be identified is 2002; the analysis therefore omits the housing booms that are currently underway. In the cycles that have been identified, the average increase in real per capita residential investment from trough to peak is around 40%. The largest occurred in Korea from 1973 to 1978 (where investment rose by 160%). The trough to peak increase has exceeded 50% in 16 cases.”

    By this definition Belgium has not had a housing boom in the last few years (and neither, though you hadn’t spotted this one, has Italy at the turn of the century). Of course, as they say, this is a “rather generous” definition of a boom.

    Indeed, some of these “booms” may be better classified as bubbles, as the following on hard landings makes clear:

    How common are soft landings? If a soft landing is defined as a relatively small reduction in the investment rate, they are not especially common. There have been only four cases where the decline in per capita residential investment has been smaller than one third of the increase that occurred during the boom years (these are the Netherlands after 1978, Belgium after 1990, the United Kingdom after 1998 and Finland after 2000). Soft landings are more common if they are defined as gradual declines, i.e. where it takes at least three years to hit the trough. There have been around 20 examples of these. But all of these were comparatively deep declines. If a soft landing is defined as something that is both mild and gradual, there has not been a single case out of the 49 boom bust cycles.”

    So, on the generous definition of soft landings we have 20 booms, and – by subtraction – 26 bubbles or semi-bubbles.

    More points of note, Belgium seems to have had a “boom” at the end of the 80s. Will this be the last one. The Netherlands had one at the end of the 90s. Again, the same question. And what has happened with Finnish housing. I think looking at all these kind of “borderline cases” will teach us a lot.

    In that sense, detailed study of German or Japanese residential construction activity would seem to offer relatively little in the way of sensitive information, since these markets seem to be well and truly “done”.

    Of course a more interesting question – and indeed the one I am presently thinking about – is why there are these booms in the first place, and what the whole pattern is all about. Decoding the information we have, this is the issue.

  20. The housing booms created by US retirees moving to Florida, or UK retirees moving to the Mediterranean, are in part fueled by relative youth in the places the retirees are moving from. Without people willing to purchase the retirees former residences, they would be less able to afford to relocate to a warmer clime.

  21. “are in part fueled by relative youth in the places the retirees are moving from.”

    This may very well be the case. In the case of the UK, of course, this may also be facilitated by people relocating from Poland.

    Maybe you think I joke here, but last year in Spain 1 mortgage in 4 was from a (recent, since there are hardly any others) migrant. Allowing multiple salaries to be attached to a single mortgage is one of the recent “financial efficiencies in mortgage markets”.

    Of course in the US, after the sub-prime bust, this may now be more difficult.

Comments are closed.