"Should we pension off the state pension?" asks Robert Colville. The answer is: emphatically no. Quite the opposite. There's a case for increasing it.

Robert is right that the cost of the state pension will rise over time. The OBR expects it to rise from 5.3% of GDP now to 7.3% in 2060 if the triple lock continues.

But this is irrelevant. Unless we euthanize people in their 60s, we'll have to pay pensioners an income. The question is whether this comes from taxes in the case of a state pension or from dividends in the case of a private pension financed by stock market investment. Whichever it is, there's a burden upon future working people.

And I reckon there's a good case for pension provision to be done by the state more than via individual private pensions.

I don't say this merely because some people are too short-sighted or poor to save, important though this is. It's because even the most prudent saver will struggle to provide a private pension for herself.

One reason for this is that the costs of private pensions are eye-wateringly high. It's common for fund managers to charge £1500 a year on a £100,000 pension pot - which is a fortune for what should be a simple administrative job.

But there's a bigger reason. It's impossible to know how much we should save. This isn't just because we don't know how long we'll live or because stock market returns are volatile. It's also because we don't even know what long-run equity returns should be.

Of course, we know they've been high in the past: 5.7 per cent per year in real terms since 1900. But we can't be at all sure this is relevant. For much of the 20^th century, investors feared there was a chance of something really nasty happening: military defeat, nuclear annihilation, revolution or severe depression. For the most part, these risks didn't materialize and so returns were good, at least in the US and UK. But we can't be sure that our luck will continue, and it's not obvious that stock markets now are so under-priced as to offer decent long-run returns.

In fact, if we add the equity premium (pdf) estimated by Mehra and Prescott to current real bond yields, we get negative prospective real returns on equities.

I suspect this is too pessimistic. But that's not my point. The point is that we cannot know how much we can save for a pension. And remember - the costs of misjudging savings cut both ways. Under-saving gives us an impoverished retirement. But over-saving means we deprive ourselves of nice experiences in our youth, and so have a depleted stock of happy memories in our old age.

A good state pension is the solution here. The state is better able to bear risk than private citizens.

This isn't just because tax revenue is more predictable than equity returns. It's also because the state can better pool some risks. For example, whilst individuals face longevity risk (the danger of outliving our assets) this is not so troubling for the state; aggregate life expectancy isn't so uncertain as individual life expectancy.

And there are some types of equity risk the state can bear. Retired savers face distribution risk: the danger that profits and dividends will fall as wages rise. This isn't a problem for governments which can tax labour as well as capital. Equity investors also face creative destruction risk - the danger that future growth will come from as-yet unlisted companies rather than from listed ones (as happened in the 70s (pdf)): insuring against this risk by holding private equity funds is difficult, as these expose us to considerable fund manager risk. But again, this needn't trouble the state as it can tax profitable firms whether they are listed or not.

You might question how much fiscal space governments have. But they have a damned sight more than individuals do.

For me, all this argues for the state undertaking the job of pension provision far more than it does. This isn't to say there's no place for private provision: insofar as doing so involves investing overseas, we can get future pension incomes from foreign workers not just UK ones, which is sensible risk-pooling.

I suspect there are simple reasons why this case is rarely made. One is that people tend to fetishize the public finances, and forget that some jobs need to be done and the question is only how to pay for them. Another, of course, is that low state pensions suit the parasitical fund management industry very well - and it has considerable influence over politicians and journalists.

There is, though, another point here. Given that future pensions must be paid by future workers, it is imperative that they produce a pie big enough to feed future oldsters as well as themselves. This means that policies to increase productivity are essential not least as a way of better providing future pensions.