In 2007 Greece spent 9.9% of GDP on age pensions. This was the fourth highest level of spending on pensions in the OECD (after Austria, Spain and Italy). Australia spent 3.2% of GDP, the fifth lowest level of spending in the OECD (ahead of Iceland, Ireland, Korea and Mexico). Greece also spends 2% of GDP on benefits for survivors, mainly widows, a level about 10 times as high as in Australia. The combined rate of employer and employee social security contributions is around 38% of the total wage bill.
But as we should all know by now the distribution of spending is important – as is the level of spending. And the distribution of spending is very different in Greece and Australia.
In Greece, pensions are provided through an earnings-related public scheme with two components plus a series of minimum pensions/social safety nets. (The safety nets are very low – in 2008 if total net income from all sources was less than EUR 7 750 you could get a benefit of around 230 EUR a month).
Effectively if your lifetime average earnings are less than 50% of the average wage, you get almost nothing from the public pension system. If your incomes were above this level your gross replacement rate is about 97% of your earnings – irrespective of your earnings. Taking account of taxes makes the system marginally more progressive – net (after-tax) replacement rates range from 114% at half average earnings to around 104% at twice average earnings.
Yes, in Greece when you retire you can be better-off than when you are in work. But of course most people don’t have full contribution histories. Moreover, the effective age of retirement in Greece exceeds that of Germany by about 27 months. Having said this, Greeks retire earlier than Germans after correcting for occupational distributions.
How is spending distributed?
The figure below shows spending on cash benefits to households with a head aged 65 years or over around 2005, comparing Australia and Greece. Spending for each decile is calculated as a percentage of overall average household income in each country, so this is a measure of relative generosity.
While Greece spends about three to four times as much on pensions as Australia relative to GDP, the bottom 30% of the Greek population receive lower relative pensions than the corresponding group in Australia. In Australia the middle income older people get the highest public pensions – this reflects the fact that some of the people in the lower income groups are probably not as poor as they appear to be – because they have higher levels of assets which exclude them from receiving pensions. In addition, even though the figures adjust for household size, the households in the middle are more likely to be couples rather than single people. However pensions paid to higher income groups are much more generous in Greece.
The second figure shows the same sort of comparisons for the overall transfer system.
Here the comparison is more stark, because Australia spends a higher share of its total transfers on benefits to working age households, and these are more progressively distributed. So in Australia a household in the second decile gets about 1.8 times the transfers paid in Greece. In contrast in Greece households in the richest decile receive about 20 times as much as in Australia. (Although this is relative to average household incomes, which are a lot higher in Australia than Greece.)
Despite spending a lot on age pensions, poverty in Greece amongst the aged is high – and this was before the crisis.
In July 2010 the Greek Parliament passed sweeping changes to the pension system. The main changes included setting a basic state pension of 360 Euros per month for all regardless of contributions; reducing the replacement ratio to 64% for contributions-related primary pensions based on earnings over the entire working life (rather than the best five of the last 10 years), topped up as before by supplementary pensions; consolidating all types of pension funds into three: for employees, the self-employed and farmers; raising the retirement age to 65 for both men and women, with gradual transition over three to seven years for women and those in special job categories; raising the minimum early retirement age to 60, including for workers with 40 years of contributions and those in heavy and arduous professions, starting in 2011; reducing the maximum allowable pension; and severely reducing the list of jobs deemed heavy or unhealthy, and which thereby qualify for early retirement (starting in 2011).
On the surface, the introduction of a higher basic state pension and the reduction in the maximum allowable pension are likely to have a progressive impact. The further pension cuts just announced mean that a person getting a monthly pension of 1,500 euros will face a reduction of 12 percent on the amount above 1,300 euros.
The effects of these changes may be more complex to interpret, however. Greece has a much higher share of multi-generation households than many other countries. For example, nearly 46% of men aged 30 to 34 years in Greece live with their parents – compared to 6% in the Netherlands. (Anyone who can find the comparable figures for Australia will get brownie points.) So pensioners living with adult children may be counted as living in relatively well-off households.-, but in some cases pensions will be cut for those who are also supporting newly unemployed younger people.
To end on a sombre note, the pension challenges facing Greece are by no means the most severe in Southern Europe, as can be shown by considering the how the distribution of all transfers compares across Greece, Spain, Italy and Portugal.
fascinating stuff Peter. I thought I knew how the insiders in Greece and Italy had twisted the system somewhat in their favour, but these details about the extreme lop-lopsidedness of the pension systems certainly make you shake your head even more in amazement at how much they have gotten away with.
Defending the Pension PIGS
For the record, I think it is worth stating the case for the apparently surprising patterns found in these graphs. We can think of three distributional objectives for pension systems: “Within-lifetime” (individuals moving their resources from working to retirement age), “redistributional” (moving from rich to poor), and “intergenerational” (moving between generations). The results shown in the figures are quite reasonable for a system which focuses primarily on the first of these.
The first thing to note about these graphs is that the vertical axis is not income shares, instead, it is average pension income for the decile group (normalised to be relative to the national overall average). (Peter does say this above, but I missed it the first time and maybe others did too).
Now, consider a pure within-lifetime pension system which is entirely effective in smoothing income across the life course (eg a more comprehensive version of superannuation). In this case people in the top decile when working will be in the top decile when retired, except now all their income will be from pensions. Similarly, the population share of the elderly will be the same in each decile (equal to the overall population share). If all income in retirement is from pensions, then the average pension incomes shown in these graphs should unambiguously increase across deciles – in exactly the same way as overall average incomes do.
I think most people will agree that both within-lifetime and redistributional objectives are important for a pension system, but there might be disaggreement over the relative importance of each. These figures simply show that the Southern European pension model places much greater emphasis on the former of these objectives than does the Australian (public) pension system.
Nonetheless, I agree that pensions are part of the problem, for two reasons.
– My understanding is that the funding side of this pension model relies too much on inter-generational redistribution (rather than within-lifetime distribution). The young are being asked to pay for these benefits.
– Especially in a crisis, the redistributional objective should trump the within-lifetime objective. These graphs indicate that there is ‘fat’ to trim from the latter which would not affect the former.
Bruce,
you are being way too kind. You should add to the above the knowledge of inflated wages in a vastly inflated civil service and a pretty heftily ageing society. Then calculate again who pays for what. There is no way this kind of generosity is manageable for the retiring baby-boomers on the backs of a dwindling supply of actually working youngsters. Can you blame the good young ones from running away from that kind of system?
In aggregate the pension reforms that have been introduced since 2009-10 in Greece mean that it is projected that by 2060 (if you believe projections that go out that far) public pension expenditure will reach 16% of GDP. In the absence of these policy changes it was projected that public pension expenditure would have reached 26% of GDP!