• REMEMBER THAT NUMBERS ARE PEOPLE LOOK FOR COCK-UP BEFORE CONSPIRACY • ALWAYS CITE PRIMARY SOURCES


Sunday, 26 February 2012

TRADE DEFICIT DATA WIN, PART I: UPDATE WITH XTRA PWNAGE

Dear readers,

It never rains but it pours. I've got a second post for you in two days, and this one is a data post so brace yourselves.

Some time ago, I looked at Greece's falling importance as a trading partner to Germany and found that, contrary to the rhetoric of the drachma brigade, Germany was increasingly less reliant (not that it ever was) on Greece as a trading partner in the post-Euro era. One commentator called me out on this saying that I should instead examine Greece's dependence on Germany as an import market, which is rather the point.

I couldn't do this for some time because getting such data is not easy. You have to go here and torture your eyeballs for some time. I thoroughly recommend it to statpornographers; it's got data on imports and exports by trading partner AND by product. So you can find out how much Greeks spend on German cars in 2001 and how much Germans spent on Greek olive oil in 2010. It's got limitations but hey, what data source doesn't.

The results are surprising even to a grizzled Euro-cynic such as myself. People generally assume that the introduction of the Euro led to a gradual increase in Greece's dependence on German imports. Greek defaultniks also occasionally claim that it led to a gradual increase in Germany's dependence on exports to Greece. Some even go so far as to claim that the purpose of the austerity-backed bailout is to keep up Greece's purchases of German exports. This despite the obvious fact, that imports in Greece, indeed in most of the world, are income-elastic: the fall in Greek consumption of German imports will almost always be steeper than the fall in our total consumption.

Anyway, the actual data show that none of the above claims are true. Don't get me wrong, Greece did become less competitive and our trade deficit grew (see further below). But we lost ground to pretty much the whole world at a greater rate than Germany, and, most importantly, the Germans neither rely on our lack of competitiveness specifically nor have any particular incentive to keep us in the Euro.

Let's start with the Greek side of things:

Amazingly, following Greece's Euro accession, Germany's share of Greek imports fell slowly but steadily, and, apart from a brief jump in 2003, so did Germany's share of the Greek trade deficit.

And now for Germany's side of the story:

Greece accounts for less than 0.7% of Germany's exports, and even in the good days it rarely accounted for more than 0.8%. That's zero-point-eight per cent guys. And that number remained more or less stable, until of course 2010, when it fell of a cliff. If we only look at the German trade surplus, it's clear Euro accession initially had the opposite effect of what people assume: Greece's share of the German trade surplus fell off an even steeper cliff than the one in 2010.

Puzzled? Then you must be very young, or have a short memory. Back in 2002, people used to write endless editorials about how Germany had made a mistake in joining the Euro and killed its own export industry because it had been suckered into accepting an uncompetitive fixed D-Mark to Euro rate. How the times change. Read, for instance, this article from back in 2002 and tell me whether it reminds you of editorials about any country you know (see more articles along the same lines in the comments section). Isn't that funny? Or Lapa-witzig, as the Germans might joke if they followed the Greek defaultnik literature?

The Germans of course sucked it up and pursued a strategy of - you guessed it - internal devaluation. Perhaps the fact that it worked for them, despite the pain, might be the reason they think it just might work for us. The point is that, even if the dynamics of the Euro are a zero-sum game, who sits on either side of the zero-sum equation is not set in stone; it is endogenous.

For those of you who are hungry for more detail, check the original source of the data here. Or you can download my selection of the data from here.

By the way there are inconsistencies in the data. Greece's imports from Germany don't seem to add up to Germany's exports to Greece, and vice versa. This could be due to reporting asymmetries (Germany may not report the same in imports as we report in exports due to accounting conventions), or taxation, or maybe I've done something wrong. Check it out for yourselves, though, the trends are much the same either way:



Now, for the hungrier ones:

The following tables give you a summary of Greece's top import and export markets as well as the sources of our biggest deficits and surpluses over time. I've chosen 2000 and 2008 as milestones to enable comparisons between pre-and post-Euro Greece as well as pre- and post- crisis Greece. Feel free to download.




If, on the other hand, you're not happy with this but would rather play with and plot the data youself, then download the file below instead. It's an Excel file, with data for each year on imports/exports from/to every country imaginable. UPDATE: As a treat for returning readers, I've added two extra tabs to the excel file, essentially performing a regression analysis of Greek imports from Germany. Although the sample is tiny and the method is crude, for now the suggestion is that it was Maastricht, not Euro accession that changed the game when it came to German imports, although some might say it's all the same thing.



Enjoy!

THIS! IS! NOT! ARGENTINA! (UPDATE)


The other day, I got an email from Dan Beeton, the International Comms Co-ordinator at the CEPR. As veteran readers will recall, Dan and I have occasionally worked together on debunking the Weisbrot hoax. Dan is a colleague of Mark Weisbrot, so he has an interest in this.

Dan shared with me the latest CEPR paper which draws some compelling comparisons between Greece and Argentina and uses these to make the case against further austerity and in favour of Greece defaulting and leaving the Euro. I’ve promised to help disseminate this paper both on Twitter and on this blog. Not because I agree with it, but because I believe that since a Greek default is desired by a great deal of the population and ultimately inevitable, it is important for its advocates of default to become more rigorous in their argument.  

For the record, as you will know, I’m all for defaulting once we’re running a primary surplus, though I would caution that this will not in itself make Greece’s finances sustainable.

That said, as much as I respect Weisbrot and his colleagues at the CEPR, I’m sick to the back teeth of comparisons between Greece and Argentina. They are just too simple. Argentina was on a foreign currency peg; Greece is in a monetary union, which is like a foreign currency peg. Argentina had an IMF intervention, Greece had an IMF intervention. Argentina defaulted; Greece is expected to default. Hey presto, Greece is Argentina. What worked for one, must surely work for the other. It’s so easy to predict the future when you already know what you want it to look like. So I thought I’d talk you through some of the main ways in which Argentina in 2002 was not like Greece in 2011. I know it won’t convince the defaultniks but at least I’ll get this stuff off my chest. 

First, it’s important to appreciate how big the Argentine budget deficit was when that country defaulted. As I’ve said on this blog ad nauseam, a country with a primary surplus can default anytime – it can still pay its way regardless of what creditors do. A country with a primary deficit has a problem. Argentina’s budget deficit in 2001 was a paltry 3%. At the time, Argentina was paying 3.58% of GDP in interest (same source), so in fact they were running a primary surplus of 0.8%, courtesy of the IMF’s bitter medicine, compared to our projected 2011 primary deficit of 2.3%, which may yet turn out to be much higher, but is only at this non-eye-watering level courtesy of the IMF too. Search the latest CEPR paper for references to Argentina’s primary surplus. Not one. Yeah, I thought so. Even rigorous-minded Keynesians don’t believe deficits matter. You know why? Because cool guys don’t look at explosions.

Similarly, the CEPR paper makes the point that Greece’s exports are much higher as a share of GDP than Argentina’s back when it defaulted, and therefore better placed to drive growth. First, I would point out that Greek exports have grown, by nearly 10% in 2011. Second I need to explain that the share of exports on GDP is not the only thing that determines the sustainability of falling off a currency peg.

In early 2002, Argentina’s current account was nearly balanced, with a deficit of 1.4% of GDP in January. Nearly all gains in competitiveness went into growth. Greece is a different story: we’ve got a current account deficit of over 10% of GDP. Hence, while Argentina was able to grow at the very substantial rate it did using a devaluation of about 65% between 2001 and 2005, it would take a much greater devaluation to replicate this effect in Greece. Remember, even with massive inflation, you can’t devalue by more than 100%, although Keynesians will no doubt find a way (perhaps a tax on holding the national currency?). But even with their ‘modest adjustment’, the Argentines paid dearly for this strategy. You can check here what happened to inflation in the aftermath. It topped 25% in the first year, and was almost 15% the year after that, and has since never recovered to pre-default levels. In fact, Argentina is still blatantly manipulating inflation figures to keep up appearances – so much so that they have to threaten and fine anyone who dares publish more accurate figures (unless they are a union, and then only if they keep their estimates of real inflation for the negotiating table only).

If you are advocating foreign-currency default but don’t know why inflation is bad for a country, I’m tempted to suggest that you deserve to have its massive double-digit slapped across your face. But let me explain nonetheless: inflation means your money is worth less in relative terms and is harder to store, i.e. convert into wealth such as savings or a house. Inflation is a tax on everyone who earns a wage or benefits but has little negotiating power with the government, as well as everyone who saves money in a bank or owns a cash-poor business. It is, on the other hand, a subsidy to anyone whose income comes from investment, anyone who owns gold, anyone who owes money in the national currency, anyone who has the political connections to negotiate higher salaries, and any business able to borrow cheaply.

Think really hard of which of the two sides you’re on and you’ll know whether it’s good for you. Rich people and banks are typically winners from massive inflation. The working poor are typically losers. That’s partly the reason why, despite Argentina’s success in battling income inequality, wealth inequality (as measured by the wealth Gini coefficient) has not only failed to come down, but actually increased since the default, from 74% in 2000 to about 75% in 2010, and is in fact way higher than Greece’s.

That’s one point. On to the second. Whatever the fiscal and current account deficit figures, we need to get one thing clear: the Greek state is much, much bigger than Argentina’s was when it defaulted. More than double in size in fact. Even without paying any interest, the Greek state would still be leeching much more money out of its economy than Argentina’s ever did, either pre- or post- default. Hence the potential for growth post-default would be much smaller without, you guessed it, more austerity! Don’t forget, unlike government investment, government consumption does slow down growth.

I can sense your disbelief at this point. The assumption among most commentators is that anyone who flicks the finger at the IMF must be a socialist who believes in an all-encompassing state. But in fact, by 2001, the last full year before it defaulted, Argentina was spending just 18% of GDP in primary government expenditures. Contrast this to Greece’s projected primary spending of 42.7% of GDP in 2011 (which, again, could turn out to be higher). Remember, I’m using the primary spending figure because clearly in a default(nik) scenario we wouldn’t be paying interest at all.

This comparison, incidentally, tells you a little bit about why Argentina grew so fast post-default.

If Greece were to default and revert to spending, minus interest, what Argentina did as a share of GDP in 2001, just before it defaulted, we’d save 24.7% of GDP and, in my preferred scenario, return it to the taxpayer through tax cuts. And according to the IMF’s calculations, a tax cut of that monumental size would boost the Greek economy’s output by 1.3x24.7%=31.1% within 2 years (or anyway that’s how much an equivalent increase in taxes would shave off our output). Libertarians would of course love this, and provided we could find that extra 2.3% of GDP through some other revenue (such as privatisations) it would actually be possible, but defaultniks would typically hate both small states and privatisations, so I don’t expect them to come out in support of us becoming like Argentina in this regard.

Not that we could do that, either. Too much of our primary spending is locked in. The second major difference, you see, between Greece and Argentina is demographics. Greece has for many years been a much older society than Argentina. In 2002, Argentina’s old-age dependency ratio was a mere 16%, while Greece’s ratio for 2011 is just over 28%. You can check this for yourselves here (select ‘detailed indicators’ on the right).
Now by looking at the correlation of health, survivor, sickness and disability spending with old age spending, it is possible to approximate the total impact of ageing: it adds up to about 17% of GDP per annum (if you don’t like my estimate, try your own using the COFOG government spending data provided by Eurostat). Remember I’m only adding to age-related spending the additional spending on other things attributable to ageing.

Now, if our old-age dependency ratio were the same today as Argentina’s back in 2000, it’s only a matter of simple math (and an assumption of linearity which I admit may be wrong) to deduce that we would be spending 7.1% of GDP less on our aged and see a primary spending figure of 38.1%. Still more than twice what Argentina was spending when it defaulted.  Although that would, in theory, put paid to Greece’s primary deficit, cutting that much off pension and health spending would have disastrous effects on Greek society.  In fact, let’s go balls-deep in the defaultnik scenario and assume that Greece cut all of its toxic public procurement budget as well – no new roads, no new arms spending, no nothing – that would save us 13% of GDP at most. We’d still only get down to 25% of GDP, which is still way higher than what Argentina used to spend when it defaulted.  

Note by the way that Argentina didn’t have to worry about destroying its pension system when it defaulted, because by the time it defaulted it had privatised its pensions system, courtesy of the same hideous reforms that defaultniks hate so much. In an earlier paper, the CEPR credits this with losing Argentina about 1% of GDP in government income annually (presumably, of course that was meant to be the people’s money, not the Government’s, but hey let’s humour them), but they forget to mention that it is precisely this that made it possible for Argentina to default without destroying or confiscating the pensions income of its own citizens.

But in any case, this little exercise suggests that, even if our population was currently as young as Argentina’s when it defaulted, we’d still be in a far worse position to default than they were. But at least we would be able to default.

Remember, ageing is not reversible. Pre-austerity, it used to add a significant 0.46% of GDP to our primary spending per year. At this rate, if we were to balance the 2011 books in one fell swoop, by 2020 we would be running a 4.6% primary deficit once again. Unless we cut pensions and benefits savagely, of course. Which we have, to the towering rage of defaultniks everywhere.

Why is pensions income so important? Well because they are a huge part of the Greek household income. Now there’s no actual calculation available of this, so I’ve had to come up with my own. I can only offer a mix of 2011 and 2010 figures, but that’s a start. Greece’s pensions funds paid out EUR25.6bn on pensions in 2011 (pg. 95 here), against total household consumption of EUR165.8bn in 2010. Household consumption has probably fallen since, so that resulting ratio of 15.4% is definitely an underestimate. So basically, destroying pension funds via default would eat into private consumption in a massive way (it already has). Another big handbrake on growth, I think.


Defaultniks often point out how investment would supposedly soar in a country freed from the burden of debt. but age once again reaches for the handbrake. Without the ability to tap global capital markets, only domestic savings will be left to finance investment and that will not be nearly enough in countries like Greece (discussion here). Savings rates don't typically go up as a country ages; past a certain point they go down: people reach their maximum savings rate at a certain age and then start eating into their savings to pay for the kids' school fees, for healthcare, for whatever have you. Pensioners are, in fact, de facto negative savers. Have a look at the graph below if you don't believe me (source):




If you're feeling really gloomy right now, you've got my gist. The last decade was actually Greece's golden age of age-related saving propensity, if there is such a term. It was the time in our near history when we were most disposed towards saving, and our governments made sure we didn't. Game over.

The difference in household savings rates between 2011 Greece and 2002 Argentina implied by the above graph is about 8% of GDP. That's 8% of GDP less that we Greeks will be able to put into investment, post-default, than Argentina was able to. That's twice our 2010 gross national savings, bottled up for the foreseeable future by the irresistible force of human destiny. It's also more, by the way, than what we're currently paying in interest.

Here’s one final comparator that people often forget, and it’s related to all of the above. 2002 Argentina was way more entrepreneurial than 2010 Greece:  the average Argentine adult is almost three times as likely as the average Greek to be in the process of starting a business as the average Greek adult, even though entrepreneurial activity fell by half post-default: as growth returned, the appetite for enterprise fell quickly. Entrepreneurs are important to growth because they help convert all of that free post-default cash into sustainable wealth. One reason why we’ve got so many fewer of them is, once again, age: it’s easier to take a huge gamble with bankruptcy when you’re 28 (the median age in 2002 Argentina) than when you’re 41 going on 42 (the median age in 2011 Greece), and who would blame you?



The result is that even if Greece could somehow pump all of that money we spend on interest into the economy, it’s doubtful whether it would ever turn into growth of the sort the CEPR and our local defaultniks are hoping for without an exogenous boost to entrepreneurship. Don’t be quick to blame the Euro; Argentina had a currency peg too pre-2002, remember? That’s the whole point. In fact, I think the difference here is largely down to demographics and the size of the state, but it’s harder to prove this, so let’s park it for now.
This is not the end of the long list of differences between 2002 Argentina and 2011 Greece. It’s just all the stuff I could come up with at relatively short notice. But I hope it helps clarify how tenuous and wishful the Greece-Argentina analogies are.


UPDATE: 


Demetri Kofinas, aka @CoveringDelta, has paid me a tremendous compliment by inserting a reference to this analysis on the very successful show he produces on Russia Today - check it out here. Thank you Demetri!

Monday, 13 February 2012

ABOUT THAT PRIMARY SURPLUS: A NOTE TO @BBCSTEPHANIE

Dear readers,

I know there are serious things afoot in Greece and I'm not contributing very much, but the truth is that politics is not my thing; I don't have anything particular to add to the newsflow and am privy to no information beyond that. This is the time for insiders, and I'm not one of them, so why add to the torrent of pointless speculation?

That said, I have been hounded for some days now by Tweeps who have read this article and don't know what to make of it. In particular, friends have got very excited by the suggestion that Greece is currently running a whopping EUR1.8bn primary surplus.

While it's usually best not to lock horns with @BBCStephanie, the BBC Economics editor, this is an important issue, so I'll try to explain why, despite the noteworthy progress so far, the 'surplus' is an illusion.

First, it is important not to extrapolate annual figures from semi-annual ones. Second semester deficits have, for the past three years, been consistently smaller than first semester ones, and not because of fiscal adjustment alone. The three graphs below, based on the latest Budget Execution bulletins, show the monthly and semiannual primary deficits we've recorded over the last three years and should help @BBCStephanie reconsider how the primary deficit is evolving:


Second, it is important not to take the Greek government's primary deficit figures at face value because these are subject to VERY substantial revisions, only ever pointing upwards, and also because they omit a number of items that still make it onto our borrowing requirements. The reason for the first limitation is simple: it takes months for all of the government agencies and ministries to turn in their data and for the Finance ministry to quality-check them.

Compare, for instance, the original December 2010 budget execution bulletin, citing a EUR19.4bn deficit, to the December 2010 figures included in the latest budget execution bulletin for December 2011, citing a deficit of EUR21.5bn. The difference is a whopping EUR2bn more spending (about 1.5bn of this in December's monthly revision alone) and in fact, if you consider the total borrowing requirement, it grows to an amazing EUR4.9bn. We won't know the true size of the 2011 deficit, or indeed that of the second semester, until about a year later, and something tells me it will turn out to be higher as well.

Third, as Philip notes in the comments below and as I've also noted in the past, a lot of the tax measures involved in cutting the primary deficit are unsustainable one-offs. When these wear out the result could be a return to the bad old days, but the damage to business and consumer sentiment will be lasting.

For now, the second half of 2011 has indeed been a very good time for the primary deficit and long may this trend continue. It's fair and appropriate to report on this but another thing entirely to deduce policy implications for it, or a change to Greece's negotiating strength. Greece's policy-minded moderates have been looking for a sign of a primary surplus for a long time and to give them one prematurely could force the Greek government's hand in unpredictable ways. Don't forget, the second semester of 2011 was relatively benign in that expectations of the extent of GDP contraction were more or less borne out, unlike the surprisingly bad first half. What's the chance of any further negative surprises in 2012? Yeah, I thought as much.

I too want to see a primary deficit that will restore some level of sovereignty to Greece, and I do believe we're slowly getting there. But if we don't care how far we still have to go then we might as well have defaulted in the spring of 2010.