From one perspective, this is odd, because annuities have two great virtues. They give us a secure income free from the risks of investing in equities or property. And they give us an income for life, thus ensuring that we won't outlive our wealth.
So powerful are these virtues that back in 1965 the Israeli economist Menachem Yaari advised retirees to annuitise all of their wealth. Although subsequent thinking has tempered this view, many economists still believe there's a strong case for at least partially annuitising.
There is, therefore, a big gap between economists' conventional advice and what people are actually doing. This gap isn't new. There has been for years in the US a so-called 'annuity puzzle': demand for annuities has been smaller than economic theory predicts.
There are good reasons for this gap, and bad ones.
One bad reason is the belief that annuities are poor value. Of course, annuity rates are low. But this is mainly because gilt yields are low. And gilt yields are low because investors believe that the outlook for the economy is still weak and uncertain. This is hardly an environment in which to prefer risky shares and property to safe annuities.
Indeed, Jonquil Lowe at the Open University Business School has estimated that annuities are fair value in the sense of offering value for money given life expectancy and gilt yields. This is more than can be said for many high-charging financial products. If it's value for money you want, you wouldn't prefer actively managed unit trusts over annuities.
Another bad reason to shun annuities is what Elise Payzan-LeNestour calls the picking up pennies bias; some people prefer slightly higher returning assets to safer ones even if the risks are big.
Nevertheless, there are some good reasons not to annuitise.
One is simply that you might want to leave a bequest. But this doesn't explain why some retirees are using their pensions freedom to buy holidays.
Another is that you might not live long enough to get value for money from an annuity. However, this is only true for a minority of retirees - albeit perhaps a significant one.
There are, though, four other good reasons to use our pensions freedom to avoid annuities.
One is simply that we all do, in fact, partially annuitise merely by getting a state pension. At just over £6,000 a year, this is equivalent to annuitising £180,000. In this sense, all of us come closer to economists' advice to partially annuitise than you might think.
Secondly, some of us have a sufficient appetite for risk to prefer risky equities to safe annuities. Researchers at the LSE have shown that lower levels of risk aversion allied to moderate bequest motives can explain low demand for annuities.
A third reason lies in the fact that the decision to buy an annuity is, in effect, irreversible: you will soon be able to cash in an annuity, but doing so is unlikely to be value for money. This implies that not annuitising is like holding a call option - it gives us the opportunity to annuitis e later. Given uncertainty about annuity rates and our personal circumstances, this option is worth holding onto.
There's something else. It might be quite rational for younger retirees to spend a lot early in their retirement even if it means living in straitened circumstances later.
For one thing, some spending is a form of saving. Just as saving gives us a stock of wealth later, so spending on holidays and nights out gives us a stock of happy memories later. And spending on books, music or art gives us a stock of consumption capital: it increases our enjoyment of such things even years later.
Also, the presumption that we should smooth out our spending over time might be mistaken. Holidays are less enjoyable if we can't walk far; cars are useless when your eyesight is fading, and so on. In economists' jargon, we should spend when the marginal utility of doing so is high - which is likely to be in our 50s rather than our 80s. In this sense, annuities which give us a constant income whether in our 50s or 80s are sub-optimal.
You might think all this is a story about retirement planning. But there's more to it than this. What's at issue here is an old question in political philosophy which is echoed in the debate between conventional and behavioural economists, namely: if people have freedom, will they use it well or badly?
On this question, I am in the former camp. Although I am generally no fan of Mr Osborne, I applaud his decision to give us more pension freedom.