Look at the way in which the continental European welfare model was designed for a world where there were five or more workers for every dependent and unemployment was below 4 per cent. That model cannot continue in a world when, in another 20 years' time, there will be fewer than three workers for every dependent, and where the labour taxation to pay for the system helps create double-digit unemployment.
So what might be the structure of the labour market in another couple of generations' time, when the people now entering the workforce will be drawing their pensions? Well, 50 years really is a very long way off, but we can catch some ideas about the way the labour market might develop in the early years of the next century by looking at present trends, pondering how secure they are and how far they might run.
The trouble is that the trends are all over the place. Start with something as basic as employment, the proportion of people of working age that are employed. In the US and Japan that has been steadily rising over the past 25 years but in the UK it has been stable and in continental Europe it has been falling. The general trend of unemployment has been down in the US over the past 15 years whereas in continental Europe it has been up. True, there has been a universal growth of the proportion of women in the workforce. There has also been widespread growth of part-timers as opposed to full-time workers, but even here you have to be careful because the trend seems to have stopped in the US, as the graph on the left shows. Finally, while there has been a general rise in self-employment, the rise has been particularly strong in the UK and has not been nearly so strong on the Continent.
When the figures don't help one has to rely on instinct and hope the figures will catch up. So here are some necessarily tentative propositions about the workforce of developed countries in the future.
Proposition one is that the shift of jobs away from men and towards women will tail off. The shift was a one-off event associated with the decline in manufacturing employment and the rise of service industries. Now that manufacturing is down to about 18 per cent of total employment in the UK, much of that shift is over. It may fall further but the dramatic collapse is past; by contrast, while service sector employment will continue to grow, the new sources of employment will draw in men as well as women.
Proposition two is that employment opportunities in general will rise, as they have been doing in the US. The States has gone from having 64 per cent of its working age population in jobs 25 years ago to 72 per cent now; we have gone from 72 per cent down to 68 per cent and the levels in Europe have fallen even faster. The argument behind this proposition is that Europe will gradually shift to a US-style labour market, as the UK has done, on the grounds that it cannot afford (either financially or morally) to keep so many people out of jobs.
Proposition three is that the growth of part-time working will flatten off, or rather, that the somewhat artificial division between part-time working and full-time working will blur to such an extent that it becomes meaningless. That is because the nature of the work contract will itself become less "one size fits all" and more closely tailored to the individual.
This leads to proposition four, which is that the division between temporary and permanent workers will also blur. The general trend in much of the developed world has been for temporary workers to be hired to replace permanent ones: they are either hired directly with people on fixed contracts replacing people on staff, or the entire function is outsourced. In the latter case, though the workers may technically remain full-time workers, their employment will depend on the contractor continuing to win the contract for the work which has been outsourced. (In the US, while there has been considerable growth of temporary employment, when times are good companies prefer to take on permanent staff, as the right hand graph suggests.)
Under these circumstances, many people may prefer to employ themselves, which is proposition five, the continued switch towards self-employment. Indeed it is even possible that self-employment will become the normal form of employment, though perhaps more likely is the possibility that the distinction between employment and self-employment will itself blur.
These propositions may prove right, or they may prove wrong. But since there must be a reasonable chance that at least some of them will be right, any redesign of the welfare state ought to plan to take them into account. It would follow from one that any reform has to be gender neutral: that there has, for example, to be a common retirement age for men and women, insofar as there should be any specific retirement age at all. It would follow from two that any redesign has to be skewed to encourage a rise in participation in the workforce: if the ratio of people of working age to dependents becomes more adverse the only way to offset it is to increase the proportion of people of working age in jobs.
But it is propositions three to five which would have the most striking effect. If the distinctions between part-time and full-time, temporary and permanent, employees and self-employed all blur, then the welfare system has to be designed to cope with infinitely flexible patterns of work.
That is surely the challenge. We still hear the rhetoric that jobs are distributed by employers, despite the fact that nearly one in seven of the workforce is self-employed. If jobs are no longer "one size fits all", the social support system cannot be "one size fits all" either.