Tuesday, June 30, 2009

OECD Misrepresentation Of Public Pensions

The recent OECD Report Pensions at a Glance 2009: Retirement-Income Systems in OECD Countries (Paris, June 2009) uses and misuses the concept of “public spending on pensions. In OECD countries “The public-pension budget has been increasingly rapidly, growing at a rate 17% faster than the national income between 1990 and 2005. Pensions account for 17% of total public spending, on average… Typically, only health and education are more expensive items in the government’s budget”. “OECD countries currently [in 2005] spend an average of 7.2% of national income on public pensions”, but half of the 30 open advanced economies that are OECD members are above this average, with Italy at the top with almost a double share at 14% (30% of the government budget, compared with an OECD average of 16%, and a rise 23% higher than income growth in 1990-2005). Austria, France, Greece, Poland, Germany and Portugal are all above 10%. The United States and the United Kingdom are at 5.7% and 5.4% respectively, Iceland at 2%, Korea at 1.6%, Mexico the lowest at 1.3%. The full picture is given in Figure 1 below.

Figure 1. Public pension expenditure 2005


The implication drawn by the OECD is that the higher the share of “public spending on pensions, the worse is a country’s performance, for three reasons: 1) affordability and fiscal sustainability; 2) “The risk … that public pension spending crowds out other desirable expenditure, both in social policy (on benefits for children and parents) and elsewhere (on education, for example)” [OECD Report, commenting on Italy]; 3) the desirable “shift towards funding and private sector management within statutory pension systems” that has taken place in recent years – a trend that “has been especially strong in Latin America and Central and Eastern Europe, but it also extends to OECD countries such as Australia, Finland, Iceland, Norway, Switzerland and Sweden”.

Public and private pensions


International differences in “public spending on pensions” are due to many reasons.


First, accounting conventions. For instance, Italian pension expenditure includes items that have nothing whatever to do with pensions, such as lump-sum, golden handshakes on employment termination or retirement (a “deferred wage”, corresponding to 1.3% of GDP), and the cost of persons accompanying invalids. Also, Italian pensions are subject to standard income tax, equivalent to 3.6% of GDP, while in other countries such as Germany they enjoy a particularly favourable treatment, with tax corresponding to only 1.5% of GDP (1.3% in the UK, until the recent tax rise on low incomes Gordon Brown perversely introduced). When differences in tax treatment are taken into consideration, Italy’s position with respect to Germany improves by 2.1% of GDP (See R.F. Pizzuti, Rapporto sullo Stato Sociale 2008, UTET Università, Roma 2008, pp. 70-71). Once these accounting peculiarities are taken into account Italy loses top ranking on public pensions among OECD countries, while retaining a low ranking for social provisions other than pensions (6.8% of GDP against the European average of 7.5%).


Second, international rankings depend on specific features of the public pension systems, such as the pension/earnings replacement ratio, and other circumstances also common to private pension systems, such as: population ageing. contributions relative to earnings, retiring age, the average period spent on retirement (life expectancy minus retirement age).


But there is a third, major explanation of international differences in the share of GDP (or of government budget) earmarked for pensions: namely, the relative weight of Pay As You Go (=public) and funded (=private) pension schemes. A country with 100% PAYG pension schemes will have all pensions classed as a charge solely on public expenditure, whereas a country with a 100% fully funded, private pension system will have a zero incidence of pensions on the public budget regardless of all the other factors mentioned above; all factors will be relevant in intermediate cases. Thus “public spending on pensions“ is not really comparable across countries. A fuller picture is given below, in a table available for 2006 from the OECD Factbook 2009 (Paris, pp. 22-23).



Figure 2. Public and Private Pension Expenditure, 2006

Affordability and Fiscal Sustainability


It is incomprehensible why a compulsory pension contribution compulsorily invested in pension funds should be regarded as private saving, in spite of its undeniable forced saving nature, while an identical pension contribution that gives an entitlement to a public pension should be regarded as a tax, and such a pension treated as a net claim on public expenditure competing with health and education as if it was not financed out of pension contributions that are not normally available for alternative uses. And if the public pension contributions replicate the performance of pension funds (as in the Funded Notional individual Accounts discussed by Nick Barr in a comment to the previous post) there is no reason whatever to treat the two systems any differently.


For the public purse, the true net claim of public pensions on public expenditure is not the total pension expenditure but the difference – if any – between current total pensions and total current pension contributions. If pension payments match current contributions, or exceed them by an amount that the government regards as affordable and sustainable, then the entire public pension system is affordable and sustainable. And – as Paul de Grauwe recently reminded us – the fact that a policy is not sustainable from now to the end of time is no reason for not adopting it for a while, especially at a time of crisis.


There is a cherry on the Italian public pensions cake, a gold coin baked inside it, an Easter Egg surprise that OECD officials should consider and ponder: in 2006, the last year for which final budgetary data are available, Italy’s public pensions exceeded pension contributions by almost €17billion, the smallest deficit in the last 15 years except for 2001. But considering that almost €29billion were returned to the budget as taxes on pensions, the actual budgetary cost was lower than pension contributions by over €11billion. Thus in 2006 the public pension system actually made a positive, significant contribution to Italy’s government budget, by 0.8% (Rapporto sullo Stato Sociale 2008, cited, p.21). For the future, in spite of the serious ageing of the Italian population, doubling by 2050 the ratio between the over-65s and the population aged 16-65, on current trends it is expected that “public spending on pensions” should fall slightly until 2015, rising towards 15.3% around 2038, then falling back to the current values of around 14% in 2050 (ibidem).

Public versus Private Pensions


From the viewpoint of the pensioner the relevant difference between the two isystems (PAYG and fully funded) s the yield obtained on the funded system relatively to the notional yield implicit in the PAYG system. The maximum yield obtainable on a balanced PAYG system corresponds to the growth rate of total wages (for a constant rate of contributions); the actual PAYG yield may will be lower if the system yields a surplus as in Italy, or higher if the government budget devotes net resources to its subsidisation.


At the end of World War II the Italian pension system, that was then fully funded, left pensioners with a hatful of rain. In the current global crisis the assets of pension funds lost a very great deal of their value, according to the OECD between 8% (Mexico) and 38% (Ireland) in 2008, most probably more in 2009. Iceland in 2006 was at the bottom of the OECD “public spending on pensions” league, with under 0.5% of GDP (see Figure 2; reportedly penultimate at 2% in 2005, see Figure 1) and its not terribly generous pension system (with total public and private pensions corresponding only to 4% of GDP, Figure 2) relied massively on private funded pensions. Given the country’s state of near bankruptcy since mid-2008, it is going to be very interesting, though probably distressing, to see how Icelandic pensioners will have fared relatively to their Italian contemporaries in the next couple of years – even allowing for differential contributions. The OECD parrots the World Bank in claiming, in exactly the same words, that “No pension system is immune to the crisis”, in total denial about differential performance (this may be unfair, for it is not clear who parrots whom; maybe there was collusion, which would be even worse).

Ageing, Retiring Age, Life Expectancy


The above reflections are not meant to suggest that all is well in the world of public – or for that matter private – pensions. Population ageing is a secular trend that has been getting all the publicity (see Figure 3; the latest issue of The Economist also has a report on Ageing). But there are two other factors that combine with population ageing to create serious problems for pension systems in general and public pensions in particular. One has been the reduction of retiring age since the 1950s for men and since 1970 for women, amplifying the impact of ageing on the viability of public pensions until 1999, when the trend was reversed (see Figure 4). The other factor is the rise in life expectancy which, in spite of the rise in retirement age postulated to continue until 2050, is expected to be faster than the rise in retirement age and therefore raises the expected retirement duration. It turns out that after the once-and-for-all gradual drop of such duration in the current decade, associated with the reversal in 1999 of the earlier strong reduction in retiring age, the expected retirement duration resumed its inexorable growth (see Figure 5).


These are real problems. The questions arise of how much of public and private pension expenditure deserves a government subsidy. Or whether in anticipation of adverse trends some reserve should be built up in the public pension system. And which of the pension parameters – retiring age, minimum years of service, pension levels, pension contributions – are best adjusted to balance the books or at least to contain their imbalance. And whether the switch from PAYG to a funded system, and the associated surfacing of a the “implicit” pension debt of the PAYG system, is worth the transition to a system less vulnerable to demographic factors but more vulnerable to a systemic crisis like the current one. These issues warrant a serious, profound and extended debate. But the pretence – today spectacularly falsified – of an alleged superiority of funded private pensions over public ones, and the use and misuse of the notion of “public spending on pensions”, are a most unsound foundation for such a debate.


Figure 3. A century of population ageing


Figure 4. Pension Ages in the OECD, 1960-20050


Figure 5. Expected time in retirement 1960-2050

Note: expected retirement duration calculated at normal pension age for each country based on national life expectancy data and projections. Source: OECD, Pensions at a Glance, cit.


1 comment:

D. Mario Nuti said...

In an e-mail message a reader confirms that "to focus on formal presentation is misleading:

(a) Figures are likely to be dubious even taken at face value (e.g. do the OECD figures include in public pension spending the tax concessions for private pensions - which they should?)

(b) Much of the public/private distinction is in any case more theological than anything else, e.g. sickness funds in Germany are (I believe) technically private, yet they are de facto social insurance."