One of the problems of being a ‘dissenting voice’ is that it is hard for others to get a grip on a yardstick for evaluating what you are saying. Normally I am considered ‘gloomy’. But if what I am arguing against is a concoction of all the ‘best case’ scenarios rolled meticulously into one, it might be fair for me to ask, aren’t those who point the finger really guilty of presenting an excessively rosy panorama.
Latest case in point are the consensus projections for life expectancy, as highlighted by the forthcoming UK pensions Commission interim report, details of which are ‘leaked’ in today’s FT:
The government actuary and pension providers may still be underestimating the rise in life expectancy and the costs it imposes on the state and private pension schemes, Adair Turner, the chairman of the Pensions Commission, will warn on Tuesday.
Mr Turner will say that the failure by actuaries to get these projections right is a cause of the pensions-saving shortfall.
Mr Turner will say that the consistent underestimation of the rise in life expectancy has “huge implications for both public and private pension policy”.
In 1950, a man aged 65 could expect on average to live for another 12 years. In 1981, the government actuary projected a rise to 14.8 years by 2004. In fact, life expectancy for a 65-year-old man today is 19 years, he said.
That meant the state earnings-related pension scheme had been designed on unrealistic assumptions about its cost. “It also means that anybody running a private sector defined-benefit company scheme in 1980 who was thinking about the underlying cost was getting it wrong by 30 per cent,” Mr Turner said at a recent meeting organised by the Association of British Insurers.
The worry, he said, is that “we may still be underestimating life expectancy”. Official projections are that life expectancy at 65 will rise from 19 years today to 21 by 2030. “But that would involve a major slowing down of what is happening,” given that life expectancy has increased by one year for every five over the past 25.
Source: Financial Times
This is not a new or surprising finding. Demographers at the now defunct Mountain View Research, in a specially commissioned report for the G7 a couple of few back forcefully made the point about the potential implications of understating this issue. According to their report:
The most dramatic miscalculations are being made by Japan, where official forecasts appear to significantly underestimate life expectancy at birth almost immediately. Their estimates appear to be at least 2.7 years low for 2010 and may be even further off than that. By 2030 they appear to be underestimating life expectancy at birth by five to six years, and by 2050 by nine years or more. Japan is mildly pessimistic regarding the recovery of birth rates. Nevertheless, its old-age dependency ratio in 2050 is likely to be underestimated by 24%. There is a one-sixth probability that Japan’s official projection of its dependency ratio in 2050 is underestimated by at least 33%.
Italy’s official estimates are the least realistic among European G-7 countries. Official forecasts probably underestimate life expectancy by at least 0.7 years in 2010; 1.7 years in 2030, and 4.0 years by 2050. Italy is also optimistic regarding the recovery of fertility. Official estimates of the working population are likely to be high by at least 2% in 2030, and 11% in 2050. As a result, the old age dependency ratio in 2050 is probably underestimated by 26%. There is a one-sixth probability that the dependency ratio in 2050 is underestimated by at least 39%.
More details can be found here. Clearly it is impossible to make definitive forward projections on life expectancy, and clearly the idea that we may all live longer than expected should come as good news, but we do need to make some forward projections, and it is important that these projections be as realistic as possible.
The issue however, as can be seen from the extract, is much bigger than simple life expectancy projections. There is a general tendency to underplay life expectancy, hold upbeat expectations on fertility, assume rosy economic growth projections, hold unrealistic assumptions about the way our ageing societies can respond to the needs of rapidly changing technological environments, understate the potential of developing countries like China, India, Brazil (and Turkey?) to emerge on the global scene, downplay the mid-term consequences of the spread of IT leveraged high-end jobs across the third world, and understate the needs of our societies for systematic inward migration. This is what I would call the ‘rosy consensus’.
Obviously in comparison to this my outlook is ‘gloomy’, but everything depends on how realistic your baseline assumptions are.