Sunday, November 25, 2012

Government Pensions in Greece

Government spending on state pensioners more than doubled from €2.9 billion to over €6.6 billion in 2000—2009 (+€3.7 billion). Why?

The number of pensioners rose by a mere 13—19% percent depending on whether one uses the data from the ministry of finance or ELSTAT (see sources and notes below). This graph plots the data from ELSTAT. Clearly, the number of pensioners cannot, alone, explain the increase in spending.

To understand what drove more spending, I have decomposed the rise in three parts: inflation, more pensioners and higher pensions. “Inflation” assumes that the number of pensioners is constant and just grows the 2000 average pension at the rate of inflation. “More pensioners” takes that constant purchasing power euro and multiplies it by the number of pensioners—the delta between that and inflation can be attributed to the fact that there are more people receiving a pension. “Higher pensions” is what remains unexplained by either inflation or the number of pensioners.

Of the €3.7 billion increase, inflation accounted for 26%, the higher number of pensioners made up 23% and higher pensions contributed the remaining 51%. Thus, half of the increase in government spending for state pensioners came from giving out more generous pensions. Military pensions were 50% higher than inflation alone would justify and civil pensions were 35% higher.

One final comment on the number of pensioners. For this, I turn to the ministry of finance, which has more complete and more updated information. The chart shows the number of pensioners from 2000 to 2012.

From 2000 to 2005, the number of pensioners was growing at 0.6% a year. Then, in 2006—2007, it started to grow at 1.5% a year. Finally, in 2008—2012, it grew by 3.1% a year. The pace of retirements accelerated by 2.5 times in less than a decade! I don’t have enough information to explain this. Perhaps there was a demographic bulge. But I suspect that early retirement incentives played a role as well.

Notes and Sources

The Greek state classifies pensioners in two ways.

In the Hellenic Statistical Authority’s (ELSTAT) Public Finance statistics, they are divided into civil, military, war and other; ELSTAT, Public Finance 2000-2006, 23, here; ELSTAT, Public Finance 2002-2008, 23, here.

In the Ministry of Finance’s budgets, they are classified into civilians, military, war, rail, local government, national resistance and clergy; Ministry of Finance, State Budgets, various editions, here.

Both data series have their advantages and they generally match with two exceptions: the ministry of finance data is for October of each year (ELSTAT does not specify whether this is year-end), and the ministry also includes a count for national resistance fighters, which do not show up in ELSTAT. But the numbers are pretty close otherwise.

For all charts, except the last, I have relied on ELSTAT. There is one more exception: ELSTAT reports spending for 2009 in its Statistical Yearbook 2009—2010, but it does not report the number of pensioners in the same way. I have used the data from the Ministry of Finance to estimate the total number of pensioners in 2009. The results are the same: the split between inflation, more pensioners and higher pensions is similar for the data through 2008 (48% explained by higher pensions).

Saturday, November 24, 2012

Transportation in Focus: Hellenic Railways Organization (OSE)

Part three of a series on transportation in Greece. Hellenic Railways Organization (OSE) runs the country’s railway system (passenger and freight) and operates bus routes overseas. OSE was re-structured in 2006—2007, so the data stop at different points (data from ELSTAT, Statistical Yearbook of Greece, various editions, here). OSE is one of Greece’s longest standing and most loss-making companies, so in contrast to my other posts, this one goes back in time.

Passenger transportation has fluctuated, and in 2008, OSE carried 25% fewer passenger-kilometers than in 1975. Passenger transportation can be classified in three periods: (a) a U-shaped period, from 1975—1989, when passenger transport peaked; (b) a decline (-43%) from 1989 to 1994, which marked a 30+ year low in overall transportation (one has to go back to 1967 to find a lower figure); and (c) relative stability from 1995 to 2008, which a slight upward trajectory.

Cargo transportation shows similar variability: in 2008, OSE transported 16% fewer freight ton-kilometers than in 1975. Cargo transportation can be similarly divided in three periods: (a) from 1975 to 1994, a steady decline with occasional (1982—84 and 1986—89) spikes; (b) from 1994 to 2002, a period of relative stability; and (c) a period of relative boom. From the composition of traffic, it is clear that the decline was driven as much by domestic as by foreign transport, but the recent spike has been driven by foreign deliveries (OSE credits Bulgaria’s entry into the EU as adding to cargo transport).

The number of employees at OSE can also be divided in three periods: (a) in 1977—1982, the company’s staff declined by 11%, driven by a 36% reduction in temporary staff; (b) in 1982—1985, OSE’s personnel increased by 27%, and almost two-thirds (60%) of the increase came from permanent staff; in 1985—2006 (latest data for OSE as a single entity), staff declined by 50%, while the company almost ceased to employ temporary staff.

OSE’s revenues from passengers and from cargo have covered a progressively smaller share of its total expenses. In 1975, the company’s revenues from operations covered around two-thirds of expenditures. Over time, this share declined steadily until it reached ~11% in 1997. It has since declined by a smaller share. By 2010, the company was losing about €800 million a year, and it depended on state financing or state-guaranteed loans to keep operating. Its cumulative liabilities were roughly €8 billion (from OSE’s financial accounts, here).

To sum: OSE has transported fewer people and fewer goods over time, and, despite a recent rise in cargo transportation, the company carried fewer people and goods than it did in 1975. Its personnel registered a sharp uptick in 1982—1985 (+27%), and it has since been declining at a slow but steady pace. The decline in staff, however, has not prevented the company’s deteriorating financial position: in the late 2000s, revenues from passengers and cargo covered only a tenth of total expenditures, down from two-thirds in 1975. As a result, the company has needed more cash from the treasury, its losses amounted to €800 million in 2010, and its debt was around €8 billion.

Friday, November 23, 2012

Transportation in Focus: Attiko Metro S.A.

Part two of a series on transportation in Greece. Attiko Metro runs the Athens metro system. The table below gives the basic operational and financial numbers (data from ELSTAT, Statistical Yearbook of Greece 2009-2010, 405—406, here).

From 2005 to 2009, the company carried 5.4% more passengers against a 7.2% increase in kilometers run, leading to a slight decline in utilization. Both figures, however, fell in 2009; in 2005—2008, the number of passengers grew by 11% and kilometers run by 15.7%.

Revenues increased by 17%, driven by a 15.4% increase in sales of tickets and passes—of which, ticket validations decreased but monthly passes rose by 42%. Metro thus made its revenues more permanent and relied less on ticket-by-ticket sales. Revenues per passenger traveled rose by 9.5%.

Expenditures have two parts: system rent and other. System rent depends on profits—if the company makes profits, it makes payment towards the system (in the English translation system rent includes the words "interest payments"). Expenditures also exclude investment in new lines and stations.

To understand profitability, look at expenditures excluding system rent: from 2005 to 2009, these rose by 64% (+€44 million). Of that increase, 70% came from higher spending on personnel and wages (+€30.4 million). Another 8% came from more money for cleaning and security.

The increase in wages came both from more personnel and from higher wages. The number of people working at Attiko Metro rose by 30.6% from 2005 to 2009, with the largest increase in 2009. Office clerks (including ticket sellers) and maintenance staff rose accounted for 75% of the increase. Since overall spending on wages rose even faster, the cost per employee increased by 32% in 2005—2009, or 2.6 times the rate of inflation.

In sum: From 2005, when Attiko Metro paid almost €11 million in system rent, in 2009 it lost €20 million without any system rent—a net change of almost €30 million. The drop in profitability was not the result of lower revenues (revenues rose) but of higher costs: the company had 30% more people in 2009 than in 2005, and costs per employee rose by 32%, 2.6 times the rate of inflation.

Tuesday, November 20, 2012

Transport in Focus: Company of Thermal Buses S.A. (ΕΘΕΛ)

Part one of a series on transport in Greece. The Company of Thermal Buses (ΕΘΕΛ) has since acquired the trolley company of Athens and Piraeus to form a new venture—it no longer exists in this form. Even so, some statistics on its activities are instructive. The table below gives the basic numbers (data from ELSTAT, Statistical Yearbook of Greece 2009-2010, 409, here).

From 2006 to 2009, the company’s output (kilometers run) was flat but it transported 11.5% more people with the same volume of kilometers. In other words, its buses were fuller.

The number of employees declined by 3% from 2006 to 2009 (-202). Looking at the breakdown (not shown) there was a drop in the number of drivers (-319) and ticket inspectors (-48), coupled with an increase in the number of office clerks (+33) and maintenance technicians (+130). The biggest percentage change came from the 15% addition of maintenance technicians.

Revenues increased by 3.9%. “Subsidies or grants from the Public Sector or the European Union” make up 40% of total revenues, so it is good to look also at revenues from tickets and other activities (including “advertising, ticket fines, insurance indemnities, etc.”), which increased by 2.2%.

So this company carried 11.5% more people but its revenues rose by 2.2%. Receipts per person carried fell by 8.3%.

Expenses increased by 18% in 2006—2009. Gross salaries and social insurance contributions rose by 23.2% and 17.4%, respectively (+22% combined). Other expenses showed a modest increase, while spending on repairs and maintenance declined.

Excluding subsidies, losses increased by 25% in three years, an additional €60 million, to reach €295 million (2010 losses were €360 million according to the Ministry of Finance). In 2009, revenues from tickets and other activities covered 45% of the spending on salaries and wages, and they covered 28% of all expenditures.

To sum: This company transported 11.5% more people from 2006 to 2009, but revenues increased by just 2.2%, so revenue per passenger shrank by 8%. Staff declined by 3%, but costs for salaries and social contributions rose by 22%, salaries per person increased by 27%, and losses increased by 25%.

Sunday, November 11, 2012


Readers have challenged my recent posts on austerity—and my argument that Greece has yet to practice it—in various ways. Either, people say, the data is incorrect or it is irrelevant. Let me try one more time.

This is the starting point. On one hand, we say “the state is being starved and cannot perform basic functions.” On the other, Greece still have a *primary* budget deficit, which means that we still cannot raise enough money to pay the bills even if we just ignored the debt. So we have a contradiction: we cannot both being starved and be eating too much. It does not add up.

There are two hypotheses: either revenues are too low or spending is too high. There is no other way for us to have a budget deficit.

By the way, since publishing my posts, I have found that ELSTAT publishes a pretty similar table to the one I constructed with government revenues and expenditures back to 2000 (p. 22—23). So let us use these numbers—which have a 2 November 2012 date on them.

Revenues are at a 10-year high as a share of GDP. Yes, there is tax evasion and it is important. But tax evasion did not appear overnight—we did not have more evasion in 2009 than in 2005 or 2003, at least not systemically more so. Tax evasion did not show up in the last two years, and so it cannot be used to explain why things got so much worse.

Revenues as a share of GDP
2000 43.0%
2001 40.9
2007 40.8
2008 40.7
2009 38.3
2010 40.6
2011 42.3

If not revenues, it is expenditures. The ELSAT table comes to the conclusion I did: wages (compensation of employees) and social benefits are much higher than earlier in the decade. Goods and services (which I researched to attribute to weapons procurement) and capital transfers (which is public investment) have been slashed from 2009—2011 and they account for most of the reduction in primary spending as a share of GDP from 48.8% in 2009 to 44.6% in 2011.

Spending on “social benefits” as a share of GDP
2000 14.8%
2001 15.4
2007 17.9
2008 19.6
2009 21.2
2010 21.3
2011 22.6

Confronted with that information, I can ignore it. I can say, no, this is not relevant, look at “real people on the ground.” I have no illusions about the quality of Greek statistics (after all, I have spent my fair share looking at and trying to make sense of the numbers—I understand the weaknesses). Yet this is the budget and government finances. How can you ignore it? It is like ignoring your bank account when making a plan. This is the ultimate data--it is the ultimate test.

It is also, I have pointed us, the data that the troika looks at and tries to negotiate over. They also see compensation being higher than in 2008 (or earlier) and they also see social benefits at record levels, and they wonder: how could that possibly be true? How can we reconcile all these cuts with all these measures being at record levels? Either the data is fake or the cuts are.

Is the data great? Probably not. Will the numbers be revised again? Probably. But look at the ELSTAT table—social benefits went up from 15% to almost 20% of GDP by 2008. Five percent points of GDP is a massive movement. Yes, maybe it is not 5 but 4.5 or 5.5. But the trend is too overwhelming to ignore.

At that point, you say, okay but look at the cuts that hear about. The issue here is relativity—what do you compare against? If your starting point is that 2009 was “fair,” then maybe the state has in fact been butchered (even though social spending is up as share of national income, which means that the spending cuts have not been proportional to our ability to finance them).

But if you look at these numbers versus what other European countries do, it helps you put the 2009-2011 adjustment in perspective. Look at this table, for example, on spending for old age in OECD countries:

Public spending on old age as a share of GDP, 2007
Italy 11.7%
France 11.1
Austria 10.7
Greece 10.0
Portugal 9.2
Sweden 9.0
Japan 8.8
OECD 6.4

This is up to 2007 only so it ignores the crisis. Only three OECD countries spent more on old age support than Greece did. Greece, in this measure, spent 10% versus an OECD average 6.4%. Perhaps we say that Greece has more older people. Fine. Let’s adjust for that by looking at how many old people countries have relative to their working age population (data from here). Here are the top countries (sorted by the ratio of old age dependents per 100 working age population from here):

Social spending and old-age dependency
Japan 8.8% / 31.9
Italy: 11.7 / 30.2
Germany: 8.7 / 29.7
Greece 10.0 / 27.5
Sweden 9.0 / 26.7
Belgium 7.1 / 26.2
Portugal 9.2 / 25.9
France 11.1 / 25.3
OECD 6.4 / 21.0

Greece spends more than Germany even though it has fewer old people per younger population. This is not a new issue. For the past few months, I have been read every single report on Greece that the OECD has written since 1962. For decades, it has been noting that Greece has some of the most generous (but underfunded) pensions around. This is not a new thing and the OECD data above confirms it. Our debt of 170% over GDP did not come from Greece being stingy. So yes, there are cuts, but these come only because: our past spending was unsustainable relative to what we earned, and it was also high relative to what other people spend or what Greece spent just a few years ago.

So unless the budget is irrelevant, what the state raises and spends is irrelevant, how much Greece spends relative to other countries is irrelevant, then we have to scratch our heads and come up with a convincing answer to explain the wide disparity between what we see in the daily news and what the ministry of finance has to say in its numbers. The only answer that convinces me is:
(a) we don’t fully understand how much we overspent in 2006-2009; 
(b) we focus too much on anecdotes and on income groups that do indeed suffer without appreciating how the totals look and how high total spending really is; 
(c) we don’t fully appreciate how much we spend relative to others in Europe; and
(d) we don’t see that our adjustment has not come from wages and pensions but from spending on goods and services (including weapons) and from investment.

Friday, November 02, 2012

The First Real Memorandum?

If parliament approves the 2013 Budget and its associated measures, Greece will finally have a real plan. The first memorandum was hastily drawn and ill understood, least of all by the parliament that voted for it. It was never fully endorsed and it was hardly implemented. The second memorandum was negotiated by an interim government without a proper mandate, and its main elements were to be agreed upon later (now). If this new deal passes (and the risk it will not is not trivial), it would be the first memorandum signed by parties elected for that purpose. In itself, this is a milestone.

Yet the fragility of the ruling coalition is evident in the hesitation of the two leftist parties to accept changes in employment laws. Once again, ambivalence threatens reform. PASOK’s equivocation effectively doomed the first memorandum, paralyzing reforms and leaving the country adrift. That same ambivalence is now spread between two parties, fewer deputies, and with more radical dissent from the far left. The calculus of reform has not changed, but the politics of change look harder.

The politics of change, however, also look simpler. The reason is that the left is largely a parliamentary rather than an executive partner in the current government. Of course, if the coalition disintegrates, this distinction will hardly matter – no parliamentary majority, no government. But if it survives, even weakened, this is important. Reforms during the Papandreou government failed both on a macro level (inability to come up with a grand plan on big questions) and a micro level (tendency of even modest reforms to get stuck on minister's desk).

In theory, this deal provides clarity on the big picture. An agreed big picture is important and, evidently, difficult. But the true test will be implementation: not just voting in favor of measures but carrying them out one by one. In the PASOK years, much of the stasis came from non-implementation. A New Democracy-led government committed to change could overcome these hurdles. The haggling is almost over. The time of action is coming. And while it is exciting to have a sense of consensus over the big picture, I am mostly eager to see little things being done – many, little things that have been stuck due to irresolution.