"Bad times, hard times, this is what people keep saying; but let us live well, and times shall be good. We are the times: Such as we are, such are the times."

Wednesday, 4 March 2015

Greece's shrinking working-age population: response to a critic

In discussing my recent post on Greece's alleged humanitarian crisis, I was challenged by one reader for saying that Greece's poverty stats could not be much worse in 2014 than they had been in 2013. I based this on both GDP rising and unemployment falling in 2014, and my critic took issue with the latter point on a factual basis.

His argument was that Greece's 2014 fall in unemployment was due to the working age population shrinking, and that gains in employment were in fact cancelled out by a fall in the working age population - this, he explained, was due to people leaving the country in search of work.

The argument has been played out over Twitter; I don't intend to reproduce it here just to score points. The end result is this: As of Nov 2014, the Greek Labour Force Survey (source of almost all labour market statistics) told us that unemployment was down 107k yoy. Employment was up 53K yoy. Economic inactivity was down 3K yoy. The size of the working-age population implied by LFS (see my note on this below) was down 57K yoy. The ratio of employed to total working age population was up by about 1pc point.

Where are those extra 57k working-age folks? Have they really emigrated between 2013-4? Some of them probably have. My guess, however, is that most of them are simply older than 64 now (people age automatically, regardless of what the economy is doing), or they are dead, and not enough young people have joined the working-age population to replace them. It's mostly demographics.

But I'm getting ahead of myself - LFS is actually one of the worst means possible of measuring the size of the working age population; particularly if one is trying to test a hypothesis related to migration. The reason for this is simple: LFS uses the latest Greek Census data as its sampling frame, and uses age groups for the purposes of post-stratification. That is to say, the LFS starts by assuming that the size and age structure of the overall population is more or less the same as it was during the last Census. Until as late as mid 2013, ELSTAT was using the 2001 Census as its sampling frame, but estimates were rebased in early 2014 to reflect the 2011 Census. There isn't another Census planned until 2021, so we're stuck with this sampling frame for quite a while.

So I guess the answer I should have given my critic is that Greece's LFS can't measure emigration because it cannot measure the working age population. It wasn't built for this purpose. Other EU countries' labour force surveys could be used to piece some of this information together, but getting the right data beyond Europe can be tricky.

Since we can't have a Census each year, the next best thing to help us understand what has happened to the Greek working age-population is a good counterfactual, and, amazingly, we have one. The UN produces national population forecasts on a regular basis, based on projected patterns in fertility, longevity and migration. The detailed 2012 forecasts can be easily accessed here. It's no longer very easy to get hold of the 2010 vintage of the UN projections, but you can see the central forecasts here. Crucially, both sets of forecasts use projected 2010 figures as their baseline, and neither of them incorporate economic factors in their analysis. This is because the UN forecasts are focused on building credible scenarios of population growth for the very long-term, not the next economic cycle. While I still think this is a weakness generally, it is a blessing in this particular case, as the forecasts are unaffected by austerity and provide a good counterfactual.

The 2010 UN population forecasts suggested Greece's working-age population (15-64) would shrink by some 54k people (11K per year) between 2010 and 2015, despite a constant 10K of net immigration per year. The 2012 revision widened this working age loss estimate to 146k, or 29K per year, and revised the working age population of 2010 downwards by a whopping 209k people. Why the difference? You can check the methodology here, but mostly it is to do with better projections on mortality and fertility; the UN also has to rely on Census figures, so they're bound to get things a bit wrong in the year just prior to a Census. Note also that the UN estimates for Greece make no assumption of changes in the rate of net migration - they assume net immigration of 10K per year. The impact of macro-economics is mostly assumed to be included in the various scenario forecasts - which, in Greece's case, are actually identical to the thousands in both the best- and worst-case scenarios for 2015.

The 2011 Census figures for our working age population, however, came in much lower than even the latest UN estimates; at 7.138m, the UN was a whopping 211K off. Where did those people go? Were they Memorandum refugees?

Amazingly, we have a counterfactual on emigration that can shed light on this. In December 2009, there was a Eurobarometer study of EU citizens' intentions to work abroad. You can see the full results here, but pay special attention to pp. 178-188. As of late 2009, 8% of working-age Greeks had hoped to move abroad for work purposes. Not all of these will have moved of course, but we have two ways of pinpointing the ones who did. First, 27% of that 8% said they were planning to move in the next 2 years or less - taking us to the 2011 Census. Another way of looking at this is that 44% of those who said they would consider moving had taken some concrete steps to facilitate the move. Without access to the raw data, my estimate would be that between 2% and 3.5% of Greece's working age population were planning to move pretty seriously as of late 2009 - pre-austerity.

Correcting the latest UN estimate of our 2010 working-age population for pre-crisis emigration intentions actually explains most of the disparity with the 2011 Census. At the top range of the estimates, it explains all but 20K of the loss of work-ing age people by 2011, and at the lower end, it explains all but 50K. So my estimate of the incremental emigration in 2011 as a result of austerity is between 20K and 50K.

And what about the post-Census period? Well, for this we have nothing to go on except LFS, which has 'found' (remember, it can't really find anything in this regard) an average loss of 59.5K working age people per year between Q3 2010 and Q3 2014, after a minor tweak by me to bring it in line with Census figures (60K without the tweak). A simple linear extrapolation of the counterfactual UN figures, also adjusted to bring estimates in line with Census, would suggest an average loss of 38K (56K without my tweak). This suggests net emigration of about 21K per year in the post-Census period.

One dataset that can help us piece more details together, and which I plan to explore more in future, are Eurostat's estimates from administrative data, which even include detailed estimates at the regional level. These also don't map across to the Census correctly - in fact they're more in line with the old UN forecasts, since people can leave Greece without notifying the authorities. These data still show that the working age population in Greece shrank by 330k between 2009 and 2013, one-fifth of which, however, was due to workers who were not Greek nationals leaving the country. On the same basis, Eurostat has put together some slightly iffy emigration estimates for 2011 and 2012. You can see them here. And some estimates of where the Greeks ended up.

IMHO this means the following:

1. The Greek working-age population has been shrinking for the last four years or so, though LFS figures alone are a very poor means of testing this.
2. Absent a strong immigration trend, the Greek working-age population was going to shrink anyway  post-2010, regardless of policy, as a result of demographics.
3. The bulk of post-crisis emigration by Greeks probably occured pre-2011 as a result of emigration plans that pre-dated austerity. Some 190K people left the country in this way between 2009 and 2011, and they were arguably the ones with the most marketable skills.
4. On top of these, Greece also lost between 20K and 50K working-age individuals per year as a result of austerity. This 'extra' loss was greatest pre-2011, and has been tapering off since, as Greece slowly runs out of exportable people.
5. Immigration to Greece for work purposes has trickled down to a halt post-crisis; this is unlikely to change.

Please stay tuned as I polish my estimates. I will publish all figures once they are ready.

Thursday, 26 February 2015


Greece, in brief.

If you're reading this blog then you have no doubt followed the dramatic series of events involving Greece and its creditors that followed the parliamentary election that took place on 25 January 2015. But in case you've not...

With nearly 64% of voters casting a ballot, and 36.3% of those voting for Syriza (the Coalition of the Radical Left), Greece got itself a new government. Surprisingly, Syriza held their noses and entered a coalition with the far(cical) right party, the Independent Greeks, who had barely scraped into parliament after losing more than half of their share of the vote since the 2012 elections. Their victory was hailed by anti-austerity activists and commentators everywhere as an unquestionable mandate to not merely end austerity in Greece, but also reshape Europe while we're at it.

Unfortunately for all of us, when Syriza took their mandate to the bank, announcing that the troika was unwelcome and we demanded debt relief and a lower surplus target, our creditors predictably refused to cash it - an eventuality once derided by our new Prime Minister as having "not one [chance] in a million." What followed was not entirely predictable, but not far from either. First, the ECB made a point of not playing fair and Samaras' trap for Syriza snapped shut, thus combining with Syriza's own uncompromising ways to bring us briefly to the edge of at least a nasty bank run if not Grexit. In this toxic environment, Syriza's bargaining cards, never top trumps to begin with, started to eat into their own hands, like something out of Saw IX.

As the Eurogroup (whose legitimacy we had openly questioned) went into meeting after meeting to resolve the impasse and avert Grexit, we learned that other austerity-ravaged nations most certainly did not have our back, that the Eurogroup itself was possibly not such an august institution after all, with contradictory draft communiques leaking in every direction, and that parts of the German press were, basically, assholes (so where parts of ours). In the end, a 4-month extension to our programme was agreed, to be supervised by the Institutions Formerly Known as the Troika. Despite some commentators decrying the triumph of 'nastiness' over 'democracy', everyone has now gone back to sell the compromise to their respective electorates - our government describing it as Greece taking ownership and control of reforms, and others as Greece bowing to the inevitable and denouncing la revolucion. As I write these lines, our reform proposals have just been submitted to our partners and got a general thumbs up.

In the days leading up to the compromise, Greeks were visibly uplifted by, even proud of, the new government's confrontational style and the personality cult surrounding our new finmin, economist, game theorist and notorious Modest-Proposal-touting econ-blogger Yanis Varoufakis. A sense prevailed that we were finally in the driver's seat,  and were no longer going to be pushed around. Even the regular clashes between Varoufakis and his counterparts as well as the first rounds of Eurogroup madness were treated with a combination of nail-biting euphoria and crass humour that brought to mind the Euro cup knockout rounds of 2004. The inevitable concessions made to the Eurogroup have caused some notable reactions, but I doubt the new Government is in any danger.

This offers me some hope that either Syriza will be able to sell much-needed reforms that other, discredited parties, would have struggled to, or it will educate its voters in realism, compromise and prioritisation in a way that other parties could not. Either way, Greece wins.

Yani say what?

For his share of the negotiations, Varoufakis employed a range of devices of varying validity and tact - from pretty solid ones, such as arguments on the sustainability of our own debt or the under-estimated impact of austerity on Greek GDP and our tax base, to less helpful ones such as declaring other countries insolvent; bringing up German post-war debt relief and the rise of the Neonazi Golden Dawn in Greece, to downright cringeworthy flattery.

But the defining claim of Varoufakis', echoed by many commentators and examined by virtually none, is that Greece needs fiscal slack and/or debt relief in order to deal with a man-made humanitarian crisis caused by years of austerity. Predictably, this phrasing didn't go down well. It was echoed in the now-infamous Moscovici draft but then reportedly struck down by Germany, eventually making no appearance in the 'final' 18/02 Eurogroup communique.

Of course, everyone brings their own biases into this discussion and so do I. But there are definitions of the term out there (see here) and claims of a 'humanitarian crisis' are testable on more or less objective grounds. I promise you that, as with many such mythbusters, the truth is neither on Greece's side nor against us. It's on the side of people everywhere who want to understand and solve problems.

Do-it-yourself Statporn

Where do you start checking claims of a "humanitarian crisis"? Most people start with macro indicators such as unemployment and GDP - and no doubt Greece has suffered enormously on both counts; the Economist's round-up is particularly useful as a starting-point. But even an economic depression does not a humanitarian crisis make (no one calls the Great Depression such a thing, after all). Besides, as of Nov 2014, unemployment had been falling for a full year (though half of this seems to be due to people exiting the working age population altogether, as one critic pointed out to me), and GDP had grown for three quarters before blipping downwards in Q4 ahead of the elections. Whatever it was we'd suffered, the macro figures suggested we were on the mend, despite ongoing austerity. To justify talk of a 'humanitarian crisis', and the policies to address it, you better have facts on the actual hardship experienced by humans on the ground.

Enter EU-SILC, Eurostat's survey of incomes and living conditions, and the source of all poverty-related statistics in the EU member states. I've covered its figures in 2011 and 2012, and a new batch of figures is available for 2013 as we speak (even 2014 in some countries, but not Greece unfortunately). Using EU-SILC, you can explore the most important dimensions of poverty for yourself using the links below:

Food poverty - are you unable to afford meat, fish or pulses every other day?
Financial precarity - are you unable to deal with unexpected expenses?
Fuel poverty - are you unable afford to keep your house warm?
Arrears - are you behind on your bills?
Physical safety - are you confronted by violence, crime or vandalism in your area?
Access to healthcare - are you unable to access the right kind of medical care due to cost?
Even better, access to healthcare by level of activity limitation - this distinguishes between people with serious, ongoing medical problems and those without. Unfortunately the data series ends at 2012.

There are also two measures of  extreme poverty that I'm very keen on:
Multiple deprivation - do you face multiple poverty-related problems at once?
Deprivation, risk of poverty and low work intensity - are you experiencing poverty and have little prospect of working your way out of it?

What the numbers tell us 

This is a lot of data to take in so you need to look at it in an organised way. My preferred method of analysis is to take each type of household in Greece in 2013 (latest SILC data for Greece) and compare it to the equivalent abroad for each category of poverty or hardship. I've particularly isolated two 'live-alikes' for each category and each group - the countries with the percentage closest to Greece's. This helps put our performance into context, and, as you will see if you check for yourselves, the differences are almost always statistically negligible.

What the figures will tell you is that Greece is generally in a bad shape. On most counts, and for most segments of the population, we do worse than at least 20 of the 31 countries for which Eurostat provides comparisons (the EU 27 plus Iceland, Serbia, Switzerland and Norway). By EU standards we have unacceptably high numbers of households in arrears (i.e. behind on their bills, loan payments, rent etc), numbers of people feeling unsafe in their neighbourhoods, and people unable to work their way out of poverty (a small but particularly hard-hit minority). We do better when it comes to financial precarity and food poverty - ranking closer to the EU-average.

Greece is a worse place to be for some. If you're a single woman, or a single parent, Greece is a pretty tough place to be - mostly it feels like Bulgaria or Romania. If you're a young man who can afford to live on your own, it can be pretty OK (most people who were once in your shoes have been forced back into the family home, so you're probably earning well). Incomes and prices aside, life in Greece can feel like living in a recent accession country in many cases.

But I can't help but wonder how these figures justify the suggestion of a 'humanitarian crisis'. Should Bulgaria or Romania get debt relief until they catch up with the rest of Europe? Should the Czech Republic?

Is it getting worse?

An assumption that underlies all talk of a humanitarian crisis in Greece is that, unless addressed quickly, the current trend (which is assumed to be negative) will either get out of hand, or take forever to reverse on its own. But the evidence for this is assumed to be there, and is assumed to be common knowledge. It actually is neither.

My crisis test is two-fold:
  • first, conditions need to have been worse in 2013 than pre-crisis. I use the average of the 2003, 2004 and 2005 readings as the 'pre-crisis' figure, as 2006-7 was the peak of the cycle and it's unfair to use it as a basis for comparison. The figures you see on the table are the % changes between pre- and post-crsisis. To avoid having my results biased in favour of the austerians by a statistical blip, whenever the 2013 reading is better than the 2012 reading I use the average of the two. 
  • second, conditions need to have got steadily worse since 2007. For this test, I fit a linear trend between the seven readings from 2007 onwards and the number of years elapsed since 2007, normalise the slope based on the average pre-crisis reading, and multiply by four. It's not robust of course but on the other hand if things are by and large not getting worse post-crisis, we're probably not in humanitarian crisis territory, are we? Please note that, even though I have lazily called this a 'correlation' in the table, it isn't. 
The findings look as per the tables to the right. For the colour coding I had to use some judgment on what is a significant number or not, going back and forth between graphs and formulas - the biggest tweak happened in order to make my assessment of arrears more realistic, since arrears have followed a U-shaped trajectory from 2003 onwards. I did not want my formula to imply that arrears hadn't risen post-crisis when they clearly had.

As you can see, things have got significantly worse since the height of the cycle in many respects, but compared to the pre-peak period only the sense of safety and access to healthcare  (both crucially important, make no mistake) appear to have deteriorated strongly across all demographics. Financial precarity and arrears have risen significantly, but only to roughly pre-crisis levels. This is quite important, as no one has ever suggested early noughties Greece was going through a humanitarian crisis.

To see what this looks like in practice, look at the graphs to your right. For the worst-hit, households with dependent children, the crisis was far from over in 2013. Food poverty and multiple deprivation plateaued, which I guess is good news, but financial stress continued to rise steeply. For single parents, on the other hand, 2013 was an unexpected reprieve across all fronts (see caveat below).

As a rule, households with dependent children and multi-adult households have suffered the most, while pensioners living alone have suffered the least. These findings, however, are misleading, as they reflect the movement of hard-hit individuals back into larger households. In reality, the findings should be read in reverse: only those who have not been hard hit can afford to continue to live alone. The narrative of the young bailing out the old is probably true in Greece - though many of the old are returning the favour by subsidising the young. It's deeply unhealthy, but it's not a humanitarian crisis.

You can download all the data I've used for this analysis from here.

Greece's Shadow Welfare State

This analysis brings me to a vital discussion: how to get more out of the welfare state with (much) less. Greece’s welfare state was never set up to properly address poverty in the first place. According to the latest data, and the older data too, Greece’s welfare state has for years been the worst in the OECD at tackling poverty – in terms of how much it manages to reduce the risk of poverty per Euro spent. This was all pre-crisis studies - the times when we could afford our welfare state.

Ultimately, families were, and continue to be, Greece's de facto welfare state. Eurostat's groundbreaking work on household resource pooling back in 2013 (working on 2010 data) confirms this - with 82% of Greek households having fully pooled resources, Greece was third only to Malta and Romania in terms of household pooling in the EU. Barely half (54%) of all Greeks have full discretion in what they spend, with only Romanians and Bulgarians reporting less discretion. Income pooling was widespread but also flexible ; in 2010 Greece, fewer than 40% of individuals put all of their income into a shared household pool, while another 14.5% didn't chip in because they had no income.

The analysis by @AristosD in his book, Το αόρατο ρήγμα. goes further - Greek families are risk-diversifying, not merely resource-pooling institutions. At the heart of this function is an Anchor Income that was substantial relative to family outgoings and predictable - a solid pension, for instance, or a good professional job. Around this, the different family members diversify the family's income through a collection of low-wage, flexible work; variable entrepreneurial incomes with a significant upside (including incomes from agriculture); and marginal guaranteed income such as a small pension.

The graph to the right groups EU countries into families based on how similar the household income pooling situation was in the 2010 EU SILC figures. Sadly, France is missing due to a methodological issue which meant that answers were not comparable with other member states.

As a rule, the further apart two countries are, the more individuals' access to pooled household funds will differ. This graph will surprise some, because it doesn't fully bear out the core v. periphery, east v west, protestant v. catholic/orthodox narratives. Sure, Greece, Cyprus and Italy are very similar as might be expected. Spain is related to Portugal, though not to the other three peripherals. Ireland is close to the UK, but also to Poland and the Czech Republic, not to mention Bulgaria. Germany is closest to the Netherlands and Denmark, but also to Lithuania and Hungary. This is a deeper, more fundamental order in Europe that we are not exposed to very much these days; Europe's shadow welfare state.

In summary

Life in Greece is worse than macro aggregates such as GDP per capita might suggest, and has by and large got worse post-crisis. For most Greeks, quality of life is the same as a citizen of a recent accession country would expect. As a rule, we have live relatively better when it comes to bare essentials such as food, and the impact of the crisis has been less severe in that respect, but other important inputs such as access to healthcare and the feeling of safety have taken a huge hit. Some elements of poverty, especially fuel poverty, were inexcusably high regardless of austerity, and have got worse for some segments of the population.  But in many respects we're also just back to the pre-crisis normal, and claims of a humanitarian crisis are driven more by the ambient narrative than by hard fact. Having children is a key factor in how people have experienced the depression. Families with children may have not seen the threat of hunger rising in 2013, but their levels of financial stress were still rising.

You'll note I have not discussed incomes in this post. If you want to find out more about how incomes have developed, you can find a very similar but much more rigorous round-up by M. Matsanganis and C. Leventi here.

Greece's 'humanitarian' crisis, I think, is first and foremost the loss of the Greek family's ability to balance out its members' loss of income. It still aims to maintain this function of course; our culture is deeply embedded. But it cannot. It is clear to me that the last few years have seen younger family members (and even their dependents) return to large households, to the point where living on one's own is a sign of financial security. The tax and welfare system can be tweaked to restore some of the risk- and income-sharing function of families and protect single parents. By focusing more resources on the hardest-hit, and those without a family safety net, it really is possible to do a lot more with less.