Two of the world's foremost experts on the economics of pensions and social security team up in this work to produce an invaluable examination of the principles and practice of pension reform. At once guide and handbook, it outlines the challenge of an ageing society as a preface to a systemic discussion of the problems of supplying better pensions. Successive chapters and their numbered sub-sections eschew the hyperbole and alarmism found in journalistic treatments: instead they focus soberly on the difficult choices that have to be made. Many readers of this journal will be familiar with Barr's major study, The Welfare State as Piggy Bank (2001), with its carefully framed response to critiques of the welfare state. Barr now teaches at the London School of Economics and was adviser to the World Bank at a time when that organisation was distancing itself from its earlier crusade for wholesale pension privatisation. Peter Diamond teaches at the Massachusetts Institute of Technology and has written widely on the US Social Security system. Both authors have been in demand as advisers to governments contemplating pension reform or, in some cases, governments anxious to correct past mistakes committed in their names. This book has special chapters on Chile and China which draw on the reports the authors drew up for the governments of these countries.

Any government that wishes to grapple with pension problems would be well advised to read this book at an early stage and to make sure that its essential arguments have been registered by both public opinion and technocrats. One way or another, as population ageing advances nearly everywhere, many governments need to overhaul pension provision. Only about one-fifth of the world's population have serious pension coverage and even this is unsatisfactory in various ways, and between 1995 and 2005 about 30 governments were persuaded to replace pay-as-you-go (PAYG) public pensions with commercially supplied alternatives.

Carr and Diamond patiently itemise and elucidate the problems. They stress the large sums required to keep growing elderly populations from poverty. They warn that pension funds can be eroded by charges or can capsize if exposed to economic squalls. An annual fund charge of 1 per cent is likely to result in a reduction in yield of 19.6 per cent (p. 162). In the United Kingdom, fund managers successfully lobbied to raise the charge threshold for ‘stakeholder’ pensions from 1 to 1.5 per cent. Costs are pushed up in such systems by the need for marketing and customisation. By contrast PAYG systems involve blanket enrolment and can usually be financed by a simple payroll tax. Beyond the question of economic cost, the authors also stress limits to administrative capacity which may lead to the adoption of approaches which are easy to implement. Another pitfall to watch out for is the transitional costs of moving from PAYG, which easily swamp any notional advantage.

The authors counsel the need for a holistic assessment of pension systems. Contributions and deductions, on the one hand, and pension income on the other, can easily amount to a tenth or more of Gross Domestic Product (GDP), so both will have significant macro-economic impact. A chapter on pension finance and labour markets gives some illustration of this point. The mobility-limiting effects of classic corporate ‘defined benefit’ schemes are noted and the authors also warn that such schemes have significant risk because the sponsor may not be in a position to meet their obligation. Indeed we know that millions of US employees in steel, airlines and automobiles have lost both jobs and entitlements in consequence.

The authors devote little attention to Europe or to the argument that high payroll taxes raise labour costs and reduce aggregate demand, hence potentially boosting unemployment. This was one of the World Bank's key arguments for dismantling pension regimes in which payroll taxes and public pensions were the dominant pillar. The authors make a similar point, however, when they observe that so-called ‘employers' contributions’ are really an addition to labour costs in a market economy. For nearly three decades unemployment levels in continental Europe ran at about double the US level, a gap that has narrowed as a consequence of the credit crunch but may not have disappeared. There is a widespread tendency to say that ‘pension reform’ under such conditions means downsizing public pensions and reducing both contributions and payouts. Given the rapid ageing of European populations – which the authors document – European pension delivery will need to run hard simply to remain in the same place, that is to prevent pension payments per capita from declining as a proportion of GDP. While increasing pension payments would be good for demand, and hence boost employment, there remains the problem of how they should be financed. Keeping payroll taxes at the level of 40 per cent does not favour employment. Such an effect can be mitigated by allowing new workers or young workers exemption from such taxes but really needed are different imposts, namely taxes on higher incomes, financial transactions, shareholdings and rises in property values.

Although Reforming Pensions gives little attention to the largest European countries, it does return on several occasions to Sweden and its ‘notional defined-contribution’ (NDC) approach. This aims to pay pensions that reflect contributions while keeping them within the limits of overall growth and hence are ‘more affordable’. The essence of the scheme is that, unlike other defined-contribution schemes, no fund is built up but the co-ordinates of the scheme mimic such a facility by using specified interest and annuity conversion rates. The authors point out that Germany and France use a points system to gear pension entitlements to the history of contributions, but their pension regimes are not thought of as NDC systems. This is probably because the pension benefit is easy to specify in terms of a replacement rate. Barr and Diamond cite Annika Sundén to the effect that in Sweden ‘replacement rates in the new system are likely to be lower compared to the old’ (2006: 144). EU projections suggest that Swedish pensions will significantly lag Danish, let alone Norwegian, pensions by 2050.

Pension reform often serves as a euphemism for reducing pension entitlements and exposing them to market risk. That is not the approach adopted by Barr and Diamond though their critique is quite restrained. They are not hostile to public systems but are keen to see them adopt efficient and equitable procedures, and to be clear about objectives. They show how greater gender equity can be incorporated and they are interested in new funding mechanisms. Occasionally they resort to conventional formulations that could get in the way of such advances as when they claim: ‘Consumption by retirees is at the expense of consumption by workers, or spending on investment, or both’ (p. 152). Left out of account here is consumption based on property incomes and capital gains. Anyone who wanders around the centre of London or New York will soon discover that the (super-) rich are still with us. Of course their consumption expenditures out of income from capital are still a small slice of GDP and quite insufficient to cover the needs of retirees. So some retiree income will indeed be at the expense of workers' consumption (or investment), but this still leaves the possibility that income from capitalists' consumption might help to meet some retiree costs in an ageing society. The authors themselves cite this very possibility when discussing China's pension reforms.

They point out that one way in which governments can both honour past agreements and extend coverage to new workers is by placing shares in privatised companies into a publicly-held pension trust fund (p. 275). They explain, ‘the proposal is not that the trust fund would sell shares to pay benefits, rather it would use the flow of dividend payments as a source of ongoing [pension] finance’ (p. 276). In China the state or public banks still own large holdings in supposedly privatised corporations. Barr and Diamond suggest that they be transferred to a National Social Security Fund (NSFF). This could have the further advantage, they write, of ‘improved corporate governance’ if the NSSF had qualified and public-spirited officials.

Some of this sounds a little like the scheme proposed by Rudolf Meidner, the Swedish trades union economist, for a share levy to set up a network of social funds. Meidner's ideas were never fully implemented because the 20 families who own so much of Swedish industry were able to mobilise opposition. In the Chinese context, there is still enough of a public stake in corporations and banks to render very plausible the funding device suggested by Barr and Diamond. At the time the book was drafted, no-one could have foreseen that the US and UK governments would have taken such large ownership stakes in entities such as Citibank, AIG, GM, RBS and Lloyds. In the post-credit-crunch environment, the intriguing possibility is posed of keeping the public stake in these enterprises but vesting them a Barr-and-Diamond-style NSFF, dedicated to helping to meet pension needs and to promote a needed improvement in corporate governance. Reforming Pensions is a well-informed, timely and authoritative examination of the best way to meet the challenge of pension provision in the ageing society. It examines the main approaches with great fairness and comes up with its own original synthesis.