IF YOU reach the age of 65 in the OECD, you can expect to live for another 19 years  or so (more if you are a woman, less if you are a man). If you stop work earlier than 65, and live a bit longer than average, you could easily be retired for 25-30 years, almost as long as you were in work. But people find it very hard to get interested in pensions (even Mrs Buttonwood), even though their financial future depends on them; retirement is too distant a prospect and the issue seems too complicated.

This blog has written a lot on the subject so it is time to summon some farewell thoughts. The executive summary: pensions are more expensive to fund, employers are less willing to do so, so you will need to save more (a lot more) and/or retire later. 

All pensions are paid for by the next generation. This may seem counter-intuitive; aren't we contributing money every month? State pensions are paid for by current taxpayers (yes, there is a US Social Security fund, but it is invested in Treasury bonds, which are a claim on taxpayers). Most state pensions run on a pay-as-you-go system; this year's benefits are paid for out of this year's tax revenues. When it comes to private sector pensions, many schemes have a fund, which is invested in bonds, equities, property etc. But what gives those equities, bonds and properties value in 2030, or 2050, when the pensions get paid? It will be the workers and taxpayers of 2030 and 2050 who will generate the income needed to pay the dividends on the equities, the interest on the bonds and the rent on the properties. As a consequence of this...

Pensions are a heavier burden when the elderly population expands, relative to the working cohort. In other words, fewer young people are supporting more retirees. In 1950, there were 13.9 people in the OECD aged over 65 for every 100 of working age (20-64). As of 2015, that figure had doubled, to 27.9; by 2050, it will have nearly doubled again, to 53.2. In some countries (Greece, Italy, Japan, Portugal, South Korea and Spain), the ratio will be more than 70 by 2050. In part, this is down to lower fertility rates (below the 2.1 needed to keep the population growing) and in part because of improved longevity. Since 1970, the life expectancy of the average OECD retiree aged 65 has risen 4-5 years. At its heart, this is why state retirement ages have been trending upwards. On top of this...

Even ignoring longevity, paying pensions has become more expensive. A pension is an income and the yield on income-producing assets has fallen. In other words, the amount of income that can be generated by a given savings pot has declined. The way that savers can guarantee an income in retirement is to buy an annuity; in Britain, the income from a £100,000 annuity has dropped from £15,000 in 1990 to £5,000 today, a decline of two-thirds. That explains why annuities are unpopular, and why the British government stopped the requirement for people to buy them. But the alternative is to put your money in cash, where rates are even lower, or in riskier assets like equities. The latter may generate a higher income but there is a risk of capital loss (imagine if you had retired in 1929 in America, or 1989 in Japan). 

Companies that provide defined-benefits (DB) pensions face exactly the same risks as individuals. There is a market for offloading a corporate pension liability; it is called the buyout market and it involves the insurance sector. The cost of a buyout has risen sharply as yields have dropped. Accountants and regulators can see this and have required companies to account for the risk; that implies higher contributions. So private sector employers have retreated from DB pensions and offer defined-contribution (DC) pensions instead. These don't guarantee to pay anything; the worker gets a pot on retirement. The employee now carries the investment and longevity risk. 

What you put in is not always what you get out. That is pretty clear for public pensions. People may have to contribute for a set number of years. By and large those who earn more, pay more into the system whereas the pension is either a flat rate, or a benefit that pays a higher proportion of final salary to the lower-paid. In the case of private sector DB pensions, it is the employer who usually makes the biggest contributions and is on the hook for any shortfall.* Were workers to get their own contributions back (plus investment return) they would get a disappointing pension.

But in a DC scheme, what you put in matters hugely. Employers may match contributions but even in a generous scheme, you would do well to get 15% of payroll. The cost of a DB equivalent is 25-30% or more. The running costs of a DC scheme are usually higher. So a DC employee cannot expect to get the equivalent of DB benefits (two-thirds final salary after 40 years). Worse still, the vagaries of the market mean two people who contribute exactly the same cash amount over their lifetime could get wildly different pensions. So you need to save more (at least another 10% of salary), and/or prepare to retire later. This is not really complicated; it is just maths. Alas, very few people get it.

* There are some who would argue this point, on the grounds that pension contributions are deferred wages. While that is technically true, people don't act as if it were so. If it were, people who work for companies that offer DC pensions (where employers contribute much less than in DB) would get much higher pay than those in DB schemes. They don't. Note also that if the company goes bust, guarantee schemes like the PBGC and PPF cap the amount of benefit, regardless of how much you contributed.