PensionReforms' summary and comments
This report qualifies as a PensionReforms' 'must read'.  It starts with a first principles approach to economic issues associated with pension reform and identifies some "lamentably common errors".  It then addresses what all this might mean for policymakers.

The report is based on work done by the authors for the Chinese government (that PensionReforms has reviewed here.  It is also prompted by a 2005 World Bank report that the authors found "incomplete" - as did PensionReforms (see here).

"Simple theory assumes that individuals make optimal choices and that labour markets, savings institutions, and insurance markets exist and function ideally.  Formulating policy within that first-best framework is analytically simple, but a bad guide to pension design in a world with limited policy tools and major market imperfections ."

The report suggests that the reasons for this are that people don't know what they want, there are too many choices, regulators don't look at the system as a whole and also seemingly don't seem to know what they want to achieve.

The report itself summarises what it describes as the "core arguments":
". The roots of the problems of pension finance are long-run trends, not a short-run 'crisis'.

". Pension systems have multiple purposes, diverse institutions, diverse histories, diverse politics and diverse constraints.

". Thus there is no single best system.  Pension systems differ widely.  This is as it should be.

". Options widen as capacity constraints relax.  But. even in the most advanced economies there are benefits from keeping choices simple.

". Advice about pension reform needs understanding of the underlying principles.  There is room for disagreement about choice between sensible strategies; there is wide agreement on some aspects of bad policies.

". Examples of bad policies include staying too long with an unsustainable system; prematurely adopting a system that exceeds implementation capacity (.the introduction of mandatory funded individual accounts in China in 1998 is an example); and incorporating excessive implicit taxes - many countries have systems where benefits rise insufficiently in response to later retirement.

"The main solutions to problems of paying for pensions. In responding to long-run trends, any improvement to the finances of a pension system must involve one or more of
. higher contribution rates,
. lower benefits,
. later retirement at the same benefit,
. policies, such as increased saving, designed to increase national output.

"That statement remains true whatever the degree of funding."

"..housing aside, a pensioner's living standard in old age will depend on his or her ability to consume goods and services produced by younger workers.  PAYG and funding are both ways of organizing claims on that output.  It is therefore mistaken to focus excessively on how pensions are financed while ignoring future national output and its division between workers and pensioners."

As the report notes, pre-funding future pension promises need not improve the security of those promises at all.  It may just shift the battleground from one part of the economic system to another.

"It is clear that a developed country has a full range of choices. Thus it is not surprising that the richer countries have very different systems one from another. But the fact that the range of options is not greatly constrained by issues of feasibility should not be misinterpreted: that a country is capable of implementing an administratively demanding plan does not mean that such a plan is a good idea or that it is necessarily superior to a less administratively demanding system. New Zealand has a simple pension system through choice, not constraint."

PensionReforms thinks that 'good' pension design need not be complicated as long as the objective of that piece of public policy is clearly identified.  

PensionReforms suggests there are some gaps in the issues covered by the report.  For example, much of the debate on pension reforms in developed countries starts from an assumption that citizens are not saving enough for their own retirement.  However, as PensionReforms has noted on several occasions, the evidence for this is scarce.  In fact, what little we know about this issue is surprisingly comforting - citizens as a whole do not seem to be irrational (see here, here and here for just three examples from, respectively, the US, the UK and New Zealand).  So, are citizens really not saving enough?  How do we know that?  Wealth at and in retirement (including pension wealth) is the only thing that really matters.  Measuring that would be a start.

Even if citizens were behaving irrationally, does that mean governments must intervene at Tier 2 and/or Tier 3 to correct this communal 'mistake'?  

Should a compulsory Tier 2 scheme really be on the agenda of even the most administratively capable country, as implied by the report?  PensionReforms thinks we need better evidence than we have that compulsion actually works - the jury is still out on this as PensionReforms has noted on a number of occasions.  For example, it will be particularly interesting, in the face of the 2008 global economic crisis, to see whether more countries adopt Argentina's recent approach to effectively abolish Tier 2 or Slovakia's option that lets citizens to transfer from Tier 2 to Tier 1.

PensionReforms was slightly surprised to see no real discussion on the role of tax incentives in the public policy debate.  The report suggests that governments should probably not encourage the development of Tier 3 Defined Benefit saving plans through tax incentives and that those incentives should be designed (in developed, but not "low income" countries) to limit their regressive tendencies.   PensionReforms agrees that tax incentives are regressive (favour the rich over the poor) but that is only one of many problems.  They are also inequitable, very complex, very expensive but, worst of all, they seem not to work (increase saving).  PensionReforms thinks that is a significant public policy issue in all countries.  

As the report notes, there is no real evidence for the supposed links between savings, investment and growth - there is, in fact, much controversy about both of those links.  More savings may (but may not) lead to more investment; more investment may (but may not) lead to higher growth.  So any policy that assumes more savings will produce the growth needed to resolve the issue of paying for tomorrow's pensions makes unproven assumptions that are likely to lead to disappointment.

There is an important underpinning theme to the report - the apparent need, in any discussion of reform, to 'start from here'.  This implies building on whatever is there now.  PensionReforms agrees with the statement that: "Policy makers will attach different relative weights to the objectives, including views about the importance of poverty relief and about how risks should be shared within and across generations."

But PensionReforms thinks this begs an important question - should they do that?  What risks might policymakers run from a 'first principles' look at the things that really matter - in other words, the things the report talks about?  PensionReforms thinks a 'clean sheet' is always a good place to start.  The existing framework must affect any transition process but need not bear on the destination.

PensionReforms also takes issue with the report's summary of the apparently limited options available to what the report identifies as countries that are "very poor" or "at a low level of development".  The cases covered by PensionReforms under the topic "Poorer country strategies" (go to the Search & options tab here and select that topic) suggest that governments in those countries can, in a relatively simple way, make a big impact.  In fact, PensionReforms thinks that, if rich countries took a close look at what seems to work in countries "at a low level of development", they might learn some important lessons for their own situations.

Finally, PensionReforms notes that New Zealand used to have the "simple pension system" noted in the report but it has been much complicated by recent change - the partial pre-funding of the Tier 1 New Zealand Superannuation; the new (and already much changed) auto-enrolment, tax-subsidised 'KiwiSaver' and new complexities in the tax treatment of pooled saving vehicles.  What New Zealand now has is still simpler than that of most countries, but it ignored many of the key issues raised in this report when it made what were apparently unneeded changes to a "simple pension system" that seemed, on the best evidence available, to already be working.  (File size 225 KB; 28 pp)  266