1A notable characteristic of the pension issue is the extreme simplicity of the underlying accounting equation, the one which describes the relationship between three parameters: retirement age, contribution rates and the relative level of pensions. The text published in 1946 by Paul Vincent is by no means obsolete in this respect; the terms he uses to describe this equation could well be found, practically unchanged, in most current publications. The arguments he deploys to rule out a general return to a funded pension system also remain very contemporary. He focuses on a particular form of this system: investments with fixed nominal returns directly exposed to the risk of inflation. Today, the debate would be centred rather on the performance of financial markets, their potential contribution to the financing of the economy, counterbalanced by other forms of risk to which they expose contributors. But what remains approximately true is that no system is independent of demographic constraints: pensions are necessarily drawn from the wealth created at the same time by the working population. This idea was used repeatedly at the height of the debate on funded and pay-as-you-go systems in the 1990s.

2Among the three terms of the pay-as-you-go “triangle”, Paul Vincent excludes an increase in contributions, for reasons that are easy to understand. The French economy in 1946 was an economy of shortage, with the priority being to rebuild and to minimize the burden upon working people. While ruling out an increase in contributions, Paul Vincent refuses the option of a downward drift in pensioners’ relative living standards: on the contrary, everything must be done to avoid such an eventuality. He thus argues for the third solution, for him the least unjust, namely an increase in the retirement age. But he suggests that it be combined with a population policy designed to limit its extent, or even make it unnecessary over the long term. This population policy is two-pronged: first, a higher level of fertility whose effects on the size of the working population would restore the pension balance over the long term; and second, over the short term, a targeted migration policy designed to repair the imbalances in the 1946 age structure, i.e. above all, the deficit of births during the First and Second World Wars.

3With 70 years of hindsight, how do these recommendations stand the test of time? The most outdated aspect is the author’s presentation of population ageing. The ageing that Paul Vincent diagnoses and fears is almost exclusively “from the bottom”, due to below-replacement fertility. This phenomenon still exists and is an important component of overall ageing in many countries; in these countries, family policies and greater recourse to migration are options that need to be considered. But in the French case, the main problem is not of this nature. France is a country where ageing occurs almost exclusively from the top, due to the lengthening of life. Family and migration policies can only provide temporary solutions to this form of ageing, unless we accept the unrealistic idea of compensating increasing longevity byperpetual population growth. Our contemporary understanding of the effects of the baby boom illustrates this change of approach. There is no doubt that the baby boom temporarily played the positive role that Paul Vincent expected from an upturn in fertility, but all this is now viewed as a long and non-reproducible parenthesis of history, whose closure must now be managed. After helping to finance pensions, baby boomers are now becoming a drain on resources: the large cohorts reaching retirement age account for a large share of the problems to be faced until 2035-2040.

4So the demographic situation has become very different. Ageing, associated with an increase in life expectancy, is seen much more positively than in Paul Vincent’s time: the pejorative French term vieillard he uses to qualify the over-60s was abandoned many decades ago and sounds shocking to modern ears. But at the same time, this ageing is much more pronounced and irreversible than that referred to in the article. Paul Vincent estimates that an increase in the retirement age of just two years would be enough by 1975 to counteract the effects on pensions of ageing. This increase was very moderate and postwar economic growth made it easy to take a totally opposite path by offering the means to finance longer and higher pensions while reducing the retirement age. Today, what is the result of the same calculation based on the latest INSEE projections (Blanpain and Chardon, 2010)? The ratio of over-60s to 20-59 year-olds was 40% in 2007. Based on the medium variant of these projections, the retirement age would have to be raised to 68 years in 2040, and to 69 years in 2060 to obtain this same ratio. Here, the effort required is on a very different scale, and it would probably be difficult to rely upon such a large increase as the sole instrument of adjustment.

5What additional levers could we use? We know that no further increases in contributions are necessary if we accept, over the long term, a decrease of around 20% in pensioners’ relative purchasing power, in addition to a pushing back of retirement age to around 65 years. This is what past reforms should bring about, assuming that economic growth follows the medium projections of the French Pensions Advisory Council (Conseil d’orientation des retraites, COR). The main mechanism of this decrease is the price-indexation of the main parameters of the system: prices are used to reevaluate past wages that are taken into account for calculating the first pension, and they are used again for indexing current pensions once people have retired. While this policy takes account of the need to protect pensions against inflation risks, it does not fully pass on the benefits of growth in real incomes of retirees.

6Would this drop in pensioners’ relative living standards that Paul Vincent wished to avoid be more readily accepted in a society whose standard of living is beyond all comparison with that of the postwar period, and where retirees currently enjoy a mean purchasing power practically equivalent to that of the working-age population? Or should we avoid it by further raising contributions, despite a much less propitious context than was the case during the postwar boom years? Even if the general context has changed enormously, the shape of the equation to be solved has not changed, and the same three options must be weighed in the balance.