Greece: mapping recent job gains

An anemic recovery is currently taking place in Greece. This has led to some job creation in a labour market still reeling from the depression that the country went through. Since little attention has been paid to detail thus far it would be more than interesting to try and map the characteristics of any job gains accrued. Certain aspects of this anemic bout of job creation are somewhat surprising and definitely very different from those of the last such wave that took place during the 00s.

For the aforementioned goals to come into fruition, data from Labour Force Survey's quarterly series were used. This dataset has some pros and cons, the main pro being its detail and the main con being that it's not seasonally adjusted, something not particularly helpful in the case of Greece's tourism-heavy economy where seasonality is ever-present. To overcome this last issue and smooth seasonal variations out, 4-quarter moving averages of all the series will be used, something that while not ideal does not distort the underlying picture.

The first dimension that has to be looked into is the sectors that created the jobs.

source: ELSTAT, own calculations

In the current recovery the leader in job creation is the tourism sector, with retail trade in second place and manufacturing in third (i.e. two of the three top job-creating sectors belong to the tradable sector).

The exact opposite was true for the 2001 - 2007 period. 

source: ELSTAT, own calculations

In this case, the 6 top job-creating sectors were of the non-tradable variety with just Tourism (which ranked 7th) belonging to the tradable family. So, a distinct feature of this recovery is that tradable sectors account for most (or at least for a big part) of those "new jobs".

Another lens through which to examine the jobs created is their spatial distribution. During the 00s, more than half of the new jobs created were in the Attica region. This time around though Attica accounted for just 17% of new jobs.

source: ELSTAT, own calculations

In fact, as far as where the number of employed persons stands regarding to employment's pre-depression peak, Attica ranks dead last among all Greek provinces.  It is worth mentioning that Central Macedonia, whose capital is Thessaloniki, was a larger contributor during the current recovery.

source: ELSTAT, own calculations

If we move on and take a look at the age distribution of the persons who filled those new jobs we come across something unexpected that seems to belie quite a lot of anecdotal evidence making the rounds in Greece. Or does it? Namely, the vast majority of new jobs where filled by people belonging in the 45-64 age class. What's even more remarkable is that at the same time unemployment for people in that age group is indeed pretty sticky (anecdotal evidence were not wrong after all). So how did this age bracket account for the lion's share of job creation? The answer is because its inactives came back into the labour force. Someone more cynical than me would ask, "were those people really inactive or just working informally who for some reason or other decided to make their employment formal?" and put this whole post's raison d'être into question in the process. Well, your guess is as good as mine...

source: ELSTAT, own calculations

If we move on to the educational background of the people filling those new jobs one can find out that the majority of positions were filled by technical education/vocational degree holders while the second largest group was people with a secondary education background.

source: ELSTAT, own calculations

A few conclusions can be reached if the figures in the chart above are juxtaposed with the percent of total population that falls under each educational background.

source: ELSTAT, own calculations

First, there seems to be a premium attached to somebody having a post-graduate degree since people who do so account for a much larger chunk of jobs created than their respective population share. Second, people with a technical/vocational education background also account for a significantly larger share of new jobs than what their share of total population would imply. Third, people with a secondary education background are also over-represented, implying that the technological content of a big chunk of new positions is not particularly high. Finally, people with lower educational backgrounds are not represented at all in job creation, a fact indicating that these are almost solely people beyond retirement age who have (or are in the process of) dropped out of the labour force.

To sum this up, the effort to map the distribution of recent job gains according to a number of different variables produced some quite interesting and, in some cases, surprising results. In an effort to sum findings up, we could say that contrary to the last recovery, this time around the top job creator seem to be the country's tradable sector (mostly tourism) while, also unlike in the last expansion, the lion's share of new jobs were created outside the province of Attica (where Athens is located). In perhaps the most surprising result produced by the data-crunching undertaken to write this post, it turns out that more than 70% of new jobs were filled by people aged between 45 - 64 years old. On another note, postgraduate qualifications as well as technical and vocational degree holders seem to be in high demand while people with a basic educational attainment also don't seem to go to waste implying that a lot of the jobs created are in sectors with a middling to low technological content. That is all for now, thank you for taking the time to read the post.

Assessing the breadth of the current Greek recovery

We keep hearing about the "Greek recovery". International media outlets have got on the wagon and the Greek government is certainly doing its best in making it sound as if Greece has finally turned the page after the monumental crisis we went through the past (not so) few years. But is that the case?

For some time now I wanted to do a post assessing the breadth of the Greek recovery. The novelty of the current post is twofold. First, I used nominal (and not real) variables. The reason is that, on top of everything else, I wanted to gauge how the recovery is being felt on the ground at the moment and since we're living in a nominal world (and since very small enterprises account for the lion share of the Greek corporate universe - and these are not paradigms of sophistication in this respect) I thought that this is the way to go. The other novel aspect of this post is that I wanted to look at the whole thing from the production side. So I took a look at sectoral turnover indices that ELSTAT publishes to see how big a chunk of those sub-sectors are in expansion. A sub-sector was deemed to be in expansion if it experienced year-over-year (YoY) turnover growth.

This whole thing took a bit of data crunching to come into fruition as some of the turnover indices that ELSTAT publishes are in monthly format while others are in quarterly format so some adjustments had to be made. Unfortunately not all sectoral indices could be used after all due to spotty data. Luckily, the ones dropped account only for a small part of total gross value added and employment. On the other hand, it has to be said that the sectors for which ELSTAT publishes such indices account for 47% of total gross value added and 53% of total employment. But still one has to make do with what he/she has. 

The sectors included in the following calculations are Industry, Retail Trade, Wholesale Trade, Tourism and Professional, scientific and technical activities; administrative and support service activities. One's got to admit that the aforementioned list seems much more satisfying and full than what the chunk of total value added and employment its components account for would imply at first sight.

It's finally time to take a peek at what the breadth of the recovery for the dataset components with the most sub-sectors looks like. Here's the chart for industry.

source: ELSTAT, own calculations

One can see that while the 4-quarter MA of the number of Industry's sub-sectors in expansion has not exactly matched the heights reached during the 00s expansion, it is not lagging far behind.  One worrying aspect is that the number of sectors in expansion has not surpassed 2016's highs.

source: ELSTAT, own calculations

Moving on to Retail Trade, we can see that the 4-quarter MA of the number of sub-sectors in expansion significantly lags the levels seen during the 00s' expansion, something that is consistent with the fact that households' final consumption expenditure has decreased for the past 3 quarters under the weight of over-taxation (among others). The lofty primary surpluses that the country is supposed to achieve for the next many years as well as the current mix of measures picked, certainly do not fill someone with confidence as far as private consumption's chances to be the driver of the recovery for the foreseeable future.  

source: ELSTAT, own calculations

The picture is more or less the same as far as Services are concerned, the only difference being that the number of sub-sectors in expansion is not only lagging behind the heights reached in the 00s but also cannot seem to be able to completely recover from the tumble that the disastrous first half of 2015 brought on. What one has to factor in here is that these particular service activities are rather heavily taxed at the moment. 

If we put all those sub-sectors together and we calculate what percentage of them is in YoY expansion we get the following chart.

source: ELSTAT, own calculations

The 4-quarter MA of the percentage of sectors in expansion is currently ~24% lower than the highs of the 00s' expansion and ~15% lower than the 2001 - 2007 period average. The good thing is that after consolidating during 2015 - 2017 the MA of the number of sectors in expansion seems to be growing again. Let's hope that this keeps. 

To wrap this up, the current Greek recovery certainly feels more lackluster at the moment than what a "normal" recovery would. This could be due to the fact that it is still in its early stages but at least part of the blame has to be put on the overly restrictive current fiscal policy stance (while I'm anything but against fiscal discipline, I think that this is way over the top) as well as the severely sub-optimal mix of fiscal measures chosen. Finally, while the supposition that the current recovery is still at its early stages leaves some room for optimism, the fact that monetary policy worldwide is gradually becoming less expansionary as well as the potentially growth-harming policies being enacted worldwide, certainly curtail part of that optimism.

Looking back: the catalysts of Greece's trade balance adjustment

It certainly is no secret that Greece has experienced a massive trade balance (or goods & services balance if you prefer) reversal since 2008-2009. Now that the said adjustment appears to be going through a phase of consolidation it is interesting to see what the catalysts of the move were thus far.

source: Bank of Greece, own calculations

The basis of our calculations will be the absolute change in Goods & Services Balance components  between February 2018 (latest data available) and September 2008 (the month that the highest deficit was posted). We will use the 12-months moving sum of those components in order to smooth out monthly fluctuations and discern trends more clearly.

During the aforementioned period Greece's good & services balance deficit contracted by 28,55 billion EUR. Goods imports decreased by 21,49 bln EUR, so we got the biggest contributor right here, while at the same time, services exports fell by 7,28 bln. Moving on to the other side of the ledger, goods exports grew by 5,94 bln while services exports decreased by 6,17 bln EUR.

source: Bank of Greece, own calculations

If we add up the contributions of total exports and imports we can see that exports made a -223 mln EUR contribution while imports accounted for the whole of the adjustment. It is interesting to drill down a bit more and see how different categories of exports did, which sectors were salvaged and which ones are mostly to blame for this dismal headline figure performance.

The thing with goods exports is the dispersion of readings that different data sources provide. If one used ELSTAT's figures for merchandise exports he would arrive at the conclusion that they increased by 8,12 bln EUR in the period examined. If Eurostat's SITC figures were used, the same figure would stand at 8,036 bln while, as mentioned above, Bank of Greece's data pegs that same number at 5,94 bln. And exports potentially increasing by ~2 bln more in the period we are looking at makes a big difference.

source: Bank of Greece, ELSTAT, Eurostat, own calculations

Anyway, we'll use SITC data to see how each different sectors' merchandise exports performed because it's the only one out of the three sources that provides sectoral data.

source: Eurostat, own calculations

Oil Products make up for 57% of goods exports' increase, while Food Products are in second place with 16,2% of total and Chemicals in 3rd with 7,8%. The top spots for biggest contributors to Greece's merchandise exports growth are taken up by sectors with a low technological content (with the exception of Chemicals). No surprise there, sadly.

Moving on to services, whose dismal performance must have come as a surprise to most people as Greece abroad is mostly associated with Shipping and Tourism. Well, it is exactly due to Transportation Services' exports plummeting by 10,34 bln during the said period that services' exports contribution was negative since Travel Services and Other Services were up by 2,94 bln and 1,22 bln respectively.

source: Bank of Greece, own calculations

It is worth noting that transportation services' dreadful performance can be traced to two different reasons. The first is the fact that the sector never recovered from 2009's Great Financial Crisis and the second one is the imposition of Capital Controls in Greece in the end of the first half of 2015 which resulted in a 46,3% drop (if one compares the month before the CCs imposition with the post-CCs trough).

source: Bank of Greece,own calculations

To wrap this up, the monstrous goods & services balance adjustment that Greece experienced these past 10 years was effected through a contraction of imports. Contrary to what most people would expect, goods exports did better than those of services (due to the dismal performance of shipping). The Greek economy's structure did not assist in making the adjustment more balanced and  easing the pain. Hopefully, the country's current plight has driven home the message that the country needs to move beyond relying solely on consumption to stoke growth and that attention must be paid to the supply side too. Of course, for the economy's fundamentals to change, copious amounts of fixed investment are needed and well, this is an area that things are currently "a bit" slow over here...

The forgotten ones: Greece's ultra long-term unemployed

Unemployment in Greece has been decreasing these past 4 years but I'm afraid that this only means it went down from ridiculously high levels to what can be charitably characterized as pretty damn high levels. Still, it is a start (hopefully) and it is definitely better than nothing. 

source: ELSTAT, own calculations

It is interesting to drill down a bit deeper than headline figures and take a look at long-term unemployment and contrast its behaviour with that of short(er)-term one.

First of all, it should be noted that in Greece, the long-term unemployed (defined as those people that are unemployed for more than 12 months) account for a much higher share of total unemployment, than they do in the Euro Area as a whole. What's more, the spread between the two figures has grown during the current depression, while at the same time long-term unemployment's share even seems to be increasing in Greece.

source: Eurostat, own calculations

Eurostat publishes a detailed breakdown of unemployment figures according to the length of the unemployment spell. If we re-base those separate series to each one's respective peak we can assess a couple of different things. First, if shorter-term unemployment peaked earlier than longer-term quintiles and second, if unemployment duration is correlated to how fast each series decreased after it reached its peak.

Here's the relevant chart. I apologize for it being a bit bungled up but I guess that in order to make visual comparisons "possible", it couldn't be any different.

source: Eurostat, own calculations

The answer to both questions posed above is more or less positive. Short-term unemployment (less that 12 months) did indeed peak first and decreased faster after that. At the same time, ultra long-term unemployment (more than 48 months) peaked last with 2017 being the first year during which it has posted a hesitant decrease. Interim long-term unemployment cohorts broadly follow along these lines (12-17 months unemployment peaked and started decreasing faster than 18-23 and 24-47 but on the other hand, 24-47 unemployment decreased faster than 18-23).

If one wants an alternative representation of unemployment in Greece based on how different duration cohorts evolved here it is.

source: Eurostat

The chart makes more than obvious how ultra long-term unemployment (> 48 months) is not really budging.

It would be interesting to try and compile the profile of the people that find themselves locked into that seriously detrimental state that is ultra long-term unemployment.

While international and local media make a lot of noise about youth unemployment, the elephant in the room, as far as long-term unemployment is concerned, is old-age unemployment. People over 45 years old account for about 40% of the long-term unemployed with their ranks having swelled considerably after the depression broke out. To be precise though, the trend towards older-age unemployment accounting for a bigger chunk of long-term unemployment has been there for quite some time and is more structural in nature.

source: ELSTAT

The sectors from which the long-term unemployed come from are all over the spectrum. This makes sense since, especially before the crisis, virtually all sectors of Greece's economy were mostly inward-oriented meaning that the sudden-stop and ensuing collapse in domestic demand hit all of them.

source: ELSTAT

On top of the rank one finds "persons that cannot be classified" which probably stands for people that either refused to answer the relevant question or for younger people that have yet to hold their first job.

To wrap this up, while everyone is going on about youth unemployment a thought should be spared for the older-age unemployed too. These people face considerable challenges, namely skill-sets that have become obsolete, social exclusion, a lower probability of having family to support them through this etc. Also, the most-efficient re-allocation mechanism (i.e. the market) has chucked them out and they mostly have to depend on active labour market policies (ALMPs) and the Greek state's limited financial means to fund these. Fingers crossed for all the people that find themselves in that psychologically (and potentially physically) scarring situation and let's hope that an investment boom (now how improbable does that sound?!) will lift all boats and get them back into the fold.