WHY ARE THE GREEKS SO GLOOMY?

Because I'm not very smart, I ventured onto Reddit the other day. A user (whom I'm not naming to avoid harrassment) was questioning why every Greek they speak to is so negative about the state of the economy when things look like they're getting better. 

The user linked to this article, written by Michalis Argyrou, the Chair of Greece's Council of Economic Experts - a body directly appointed by the Greek Minister of Finance. Here, Argyrou focuses on a recent (Oct 2024) Eurostat publication on subjective poverty, and asks a very interesting question: 

How is it that Greeks rate themselves subjectively as 'poor' much more frequently than one would expect from more objective indicators of material poverty? Why do their subjective perceptions fail to reflect the significant post-crisis improvements in employment and other economic indicators? 

You could put this down to some level of partisanship; it certainly reads as though Argyrou is rage-baiting certain parts of the Greek internet, stating as he does that

'In Greece, of every 100 people who claim to be poor, only 28 are truly poor.' 

Except, there are receipts. Apparently 67% of Greeks are classed by Eurostat as reporting 'subjective poverty'. Fewer than 19% are at risk of poverty as this measure is defined by Eurostat.  

More importantly, Greece really stands out from the rest of Europe in this regard: Argyrou points to a shocking graph, not purpose-built by some ND partisan but included in the original Eurostat publication, which shows how much of a deviation there is between self-assessed poverty and 'objective' poverty across different European countries. Greece stands out so much it practically needs its own secondary axis. 

Worse, the two figures are of good quality and properly comparable - they come from the same study, SILC, which veteran readers will recognise as one of this blog's favourite surveys. SILC is where all European poverty statistics come from and it's been surprisingly robust to the rise of non-response across Europe.

You don't have to share Argyrou's institutional perspective or have particular sympathies for the current government to raise an eyebrow at this. I did too - and the fact that Balkan countries seem to dominate the ranks of poverty over-estimators seemed to point to a kind of cultural bias - a 600-yr-old legacy of treating all state institutions as predatory and only grudgingly acknowledging them if they can be weaponised against one's enemies.   

So is Argyrou right to speak of an over-reporting bias? More to the point, is he right to claim that 'In Greece, of every 100 people who claim to be poor, only 28 are truly poor.' ?

To get to the facts, we first need to understand what Eurostat means by 'subjective povery' and the 'risk of poverty', how objective the latter is, and how Eurostat figures are used to approximate them. 

First -subjective poverty. 

Subjective poverty is not a term invented by Eurostat or Argyrou - it has a long history (see eg here h/t Cubiclogic). Eurostat appears to have adopted it as shorthand for a specific measure derived from SILC.  

You can read the questionnaire from which this measure is derived here.  The precise wordings in English and Greek are as follows. If you say you can only make ends meet 'with difficulty' or 'with great difficulty', then that makes you 'subjectively poor'. If you're not convinced that's how Eurostat calculates this, see Eurostat's explanation here and the actual data for C5 here and add the two responses up for yourself. 

To my eyes at least, we have a bit of a methodological issue here. The Greek and English versions of this item are not semantically identical. That's always a risk with international surveys but here I think the deviation is significant. The Greek version literally asks 'how do you meet the usual needs of your household?'. There is no negatively charged word in it. The English version asks people whether they can make ends meet, which to the best of my understanding of English, does have a slightly negative energy. One can ask even a wealthy person how they 'meet the needs of the household' and get a complex answer. The bar for saying 'I make ends meet with difficulty' in English is higher than the bar for saying 'I meet the needs of the household with difficulty' in Greek.  

If this does not convince you, you needn't worry. My conclusions do not rely on this observation. 

Now on to 'objective' or 'true' poverty

In fairness nobody truly uses such terms. Argyrou does contrast the self-described 'poor' with the 'truly' poor, but as you can see here, the Eurostat release that he cites only reports on the 'risk of poverty'.  

Specifically, a person is 'at risk of poverty' if their income, after tax and other deductions such as national insurance contributions, and after social benefits received, and adjusting for household size, is lower than 60% of the country's median income. This is a standard definition, widely used internationally, which feeds into a lot of EU policy and Eurostat figures. It's also baked into SILC statistics, so we can cross-tab almost every poverty statistic in Europe against this. 

How much is this 'poverty threshold' in today's Euros? See for yourselves.

• ca EUR12,600 per year for a couple with two dependent children
• ca EUR 9,900 per year for a couple with no children
• ca EUR6,000 per year for a single person with no children

I hope I'm not alone in thinking this -

This is not a lot of money! In fact when you compare the Greek poverty threshold to those of other countries in purchasing power terms, it's pretty clear that a Greek person at the poverty threshold is, after allowing for the cost of living in each jurisdiction, much poorer than their peers in most of Europe. A person on the statistical threshold of poverty in Germany has about twice the purchasing power than their threshold-straddling peer in Greece.

Whatever your politics, it seems really perverse to insist that a person earning say 10pc or 20pc above Greece's scandalously low statistical poverty line can't be experiencing hardship of some sort. 

As many friends have pointed out, this measure of poverty is only 'objective' in a trivial sense - the same formula can be used to calculate it in all cases. The same income however can mean wildly different things depending on one's cost of living, mode of tenure and family circumstances. So we'll refer to persons 'at risk of poverty' from this point on. For brevity, I will also speak of 'above-the-line' and 'below-the-line' households and persons. The 'line' being the threshold for 'risk of poverty'.  

We can now restate Argyrou's challenge thus:

• First, 60% of Greeks above the 'at risk of poverty' threshold are reported to be subjectively poor. Can that really be the case?
  
  
• Second, it seems as though Greeks above the threshold are about as likely as those below the threshold to be reported as subjectively poor (though the balance of 'with difficulty v 'with great difficulty' shifts). Again, can that really be the case? 

To avoid subjective measures, we can try to correlate risk of poverty with other objective indications of hardship. You won't see above-poverty-line households struggling to eat, and even the percentage of such above-the-line households that are unable to keep their homes warm is modest. So, what gives? 

The answer presents itself clearly once you consider housing, debt and arrears. 

• Above-the-line Greeks are much more likely than any of their fellow European above-the-liners to have unpaid bills, unpaid rent payments, etc. About 40% of them have such unpaid liabilities. Watch out for the massive outlier, does it remind you of some other graph you've seen recently?
  
  
  
  
• If we try to break this down into constituent debts, Greece's above-the-liners stand out a bit in terms of rent and mortgage arrears (2.8x the Eurozone average) but more so when it comes to arrears on utility bills (4.9x the Eurozone average) and even more so for arrears on loans and hire purchase agreements (6.9x the Eurozone average). 
• Households with children are very significnalty more likely to report utility arrears (nearly 2x as childless households), and similarly more likely to report unpaid consumer loans; while arreats on mortgage or rent are less affected by the number of dependents.
• If you assume that someone who keeps getting final demands from the bank or utility companies and cannot pay these immediately is likely to say they make ends meet 'with difficulty,' well the mystery is solved.  
• Overall, utilities and loans are a bigger part of the "subjective poverty" story than unaffordable housing. Above-the-line Greeks do pay a bigger share of their income into housing costs (29%) than any of their fellow above-the-liners, but admittedly this is not a huge outlier and the share of Greek above-the-liners that Eurostat calculates as facing a 'heavy financial burden' due to housing costs alone is hardly the highest in Europe.

But wait! There's more!

Over on Bluesky, longtime friend of the blog @mperedim noted that Greece has always, even in the best of times, had a very high percentage of above-the-line-ers in arrears of all kinds, completely outside the norm in other EU countries. This in unequivocally true. 

The reason for this is clearly not the 2010- financial crisis, but a fundamental distortion in the market for credit in Greece. Due to the skewed development of lending practices, a good share of Greeks have traditionally used utilities and landlords as sources of credit instead of more formal finance. The poor do it because they have to and the well off do it because they can. This is also reflected (and @mperedim made that point as well) in the late payment of commercial debts, which is a long-standing issue in Greece. 

 

In fact the trend since 2014 is one of every other country growing out of this dependence and Greece not doing so. Why is this? It is a very long discussion that's fascinating to have but I can't conclude it in this article. What I will say is that when it comes to rent and mortgage arrears, Greece's over-the-line-ers did not stand out among their EU peers until around 2009, so I don't think the treatment of landlords as creditors is culturally ingrained - though it may have become so post-crisis. 

Utility bills on the other hand have always been a source of credit, and of bad debt, even in the best of times. We've also had a very consistent record of over-the-line-ers being unable to repay commercial credit providers, including before the crisis. 

So here is what I think. 

First, Greeks responding to the SILC survey are not describing themselves as poor. This isn't an option in the survey. 

Eurostat ascribes to them a 'subjective poverty' status based on them reporting some difficulty in managing household expenses. The word 'poverty' and its derivatives are never presented to the respondent.I think the slight difference in wording does matter, but I don't believe it accounts for much of the supposed 'over-reporting' called out by Argyrou.

The supposed 'over-reporting' emerges from the fact that 'low income' is only one way in which a person might find themselves in a bad place financially. If expenditures are high, even people with reasonably good incomes could do so. 

In the case of Greece, it seems clear that 'above-the-liners' (ie people with more than 60% of the country's median income, which to be clear is not much) are significantly more likely than their peers in Europe to be struggling with debts of different kinds, primarily related to utilities. To be clear, Greek households don't have more debt than fellow European households do on aggregate, but they do have a long history of using overdue utility bills as a source of credit, and have more recently taken to doing the same to overdue rent and even overdue mortgage payments.  

Of the ca 60% of 'above-the-line' households who report difficulties making ends meet, a good 30% seems to be down to this kind of debt overhang. A small percentage is probably also attributable to the phrasing of the question. Another 20% is likely due to the fact that people just over the 60% 'at risk of poverty' line are probably still not doing so well - Greece's statistical poverty line is much lower in purchasing power terms than other European countries. Taking these together I am reasonably sure that 50% of the 60% of above-the-liners who report difficuties (which is my transformed measure of the 'bias' posited by Argyrou) can be explained by matters irrelevant to Greek culture or politics.      

Come on, this blog is called LOLGreece, surely the Sainted People are also to blame?

Those commentators who point to a cultural angle do need to consider what would be good evidence of this. I tried one possible way of testing this but it's not very effective. I figured that, since the objective measures of hardship that I use to justify why above-the-liners might well struggle to make ends meet all correlate with having children, one would also expect the 'making ends meet' metric of subjective poverty to correlate with having children. It does not, at first glance at least. If it did I would consider this case closed. 

It doesn't; to me, there are three ways of reading this fact:

• There is some element of the poverty over-reporting bias that is irreconcilable to more objective circumstances and its origin is cultural or political - people like to complain, especially if they are politically opposed ot the ruling party. It's hard to see how more than a fifth of the observed over-reporting can be attributed to this, however, and if it was about politics you'd expect the sentiment to shift over time as parties alternate in power.
  
  
• There is some element of the poverty over-reporting bias that is irreconcilable to more objective individual circumstances but results from people internalising the economic conditions of their families and social circles. One thing that people often overlook about Greece and some of our neighbouring countries is that personal budgets are substantially pooled, sometimes even across households - think unemployed son or working mom subsidised by dad's pension. I've covered this here and it could mean that people could struggle to make ends meet even if their individual income is reasonably good; more importantly this also provides a channel for perceptions of economic well-being to flow, with less-well-off family members disproportionately affecting the narrative. 
   
• Being unable to make ends meet is partly about poverty, partly about debt and partly about skills and support networks. The reason why perceived poverty doesn't correlate with child-rearing in the same way that objective economic challenges do might be simple survivorship bias; only those with the ability to make ends meet despite challenges (because of family support, of being debt-free and of their own skills) are more likely to have children in the first-place - their social and human capital, and maybe a good deal of luck, allow them to prevent objective challenges translating to subjective well-being.      

In conclusion, what looks, to Argyrou and most casual readers, like a bias in favour of thinking of oneself as poor is, in fact, the product of debt overhang and a malfunctioning credit market, possibly coupled with the influence of Greece's pooled household budgeting (ie our close-knit family structures). 

It's not a hysterical trope but a piece of solid (or not solid, in another sense) economic fundamentals that Greece's Council of Economic Experts should be treating as a high priority risk - not dismissing as mere snowflakery. It creates a reluctance to form households, or have children, or save, invest, start businesses, indeed even to consume. 

Perhaps many of the Greeks who claim to be having a hard time deserve some hardship. Maybe they were not prudent enough in their spending or borrowing back in the day. Perhaps they deal with their creditors in bad faith. But in macro terms, the rights and wrongs of indebtedness do not matter. The resulting economic anemia will hurt even those folks whom one considers virtuous and thrifty.