PensionReforms' summary and comments
In PensionReforms' experience, there is a quite low understanding of the net present value of a life annuity, even amongst the financially literate.  That disconnect becomes greater if the value of automatic increases and a contingent dependant's annuity must be factored in to today's price.

This 2007 report confirms that information gap amongst those with a future entitlement to Tier 2 Social Security benefits in the US.  The report tested whether they might be prepared to give up part of that pension in exchange for a lump sum.  It used answers to questions from 1,000 participants in the 2004 Health and Retirement Study.

"Given the prior literature on annuities suggesting that there are important benefits from annuitization, the new evidence presented here is striking: most respondents indicate a preference for exchanging Social Security's lifelong annuity payment for a lump-sum combined with a reduced annuity benefit.  This finding is robust across virtually every demographic subgroup in the sample.  In other words, most individuals do not appear to value annuities as highly as standard economic models would suggest, a finding that also casts doubt on the ability of several leading hypotheses to explain this aversion to annuities."

Unsurprisingly, in PensionReforms' view, the report found that older people (aged in their early 60s) placed a higher value on annuities than did younger.  And those in poor health, sensibly, thought that a lump sum would be better value than a life annuity; as did those who thought they had a lower expectation of life.

"The latter finding underscores the potential for adverse selection in annuity markets: to the extent that individuals have private information about their health status or longevity expectations that lead shorter-lived individuals to avoid annuitization, this will tend to raise average annuity prices."

However, preferences seemed not to be driven by the longer-lived being more likely to buy an annuity.  Rather, they seemed framed by the negative - those who were in poorer health would prefer cash rather than a pension.

"We also find many factors that have limited / no influence over the annuity vs lump-sum choice, despite the prominent role played by some of these variables in theoretical models.  These include such as risk aversion, sex, marital status, income, wealth, having a pension plan, or having children.  We do find that, conditional on education, more financially sophisticated individuals prefer to annuitize.  We also find that individuals who place a higher probability on the outcome that future Social Security benefits will be cut are more likely to choose the lump sum option, suggesting that individuals do discount future benefits for political risk."

The options discussed in the paper are not of immediate moment as the lump sum alternative isn't available for a Social Security pension but could be significant, given that some reform options include having Defined Contribution accounts as part of a reformed structure.  Might future retirees prefer an annuity choice, rather than be forced into swapping the Defined Contribution balance for a pension?  Received wisdom has it that retirees should have regular annuity-based income because it will be better for them, over all.

"In contrast, our results imply that a majority of individuals not only value the Social Security annuity less than would be predicted by the life-cycle model, but, in fact, value the Social Security annuity at less than its actuarial value. As such, it raises the possibility that compulsory annuitization is not as welfare-enhancing as standard life cycle theory would suggest, or at least that it is not perceived as such."

PensionReforms thinks it is perhaps a step too far to suggest that citizens' preferences at the point of retirement should drive the design of State-administered retirement income support.  It is in the nature of all retirement income designs that some will be better off under either a Defined Benefit or a Defined Contribution arrangement if the only measure is the net present value of those benefits at either the date of retirement (effectively used in the report), or at death when full account can be taken of all the necessary contingencies.

But that kind of analysis misses the point in PensionReforms' view.  Public policy on this issue should not about welfare-enhancing at an individual level but rather at a societal level - balancing the needs of different groups against the community's resources.  And anyway, PensionReforms suspects that attitudes may have changed since the global economic crisis.

PensionReforms takes particular issue with this from the report:  If retirees tend to consistently underestimate the true value of an annuity, ".then it suggests that one potentially feasible path to reducing the long-term liabilities facing the U.S. Social Security system would be to allow individuals to voluntarily convert part of their Social Security annuity to a lump-sum at less-than-actuarially favorable rates.  Even in the presence of (mild) adverse selection, our results suggest that such a voluntary approach could reduce net liabilities of the Social Security system."

PensionReforms suggests that is probably the worst reason for introducing a Defined Contribution/annuitisation option.  The costs will almost certainly fall more heavily on those least able to cope - the poor and the very long-lived.  What should matter is how to ensure that state interventions limit or even prevent poverty amongst the old.

The report acknowledges the potential exposure of more individuals to the downside of longevity risk might lead to "higher rates of poverty at advanced ages."  PensionReforms observes that poverty levels amongst the old in the US are already high (34% according to an OECD report reviewed here).  Reducing that, rather than running the risk of increasing it should be a key objective of any change. (File size 230 KB; 29 pp) 380