PensionReforms' summary and comments
The mounting international evidence seems to show that governments have relatively little influence on the overall saving behaviour of households - see here for example.

If "success" were to be measured just by the amount of money invested in compulsory Tier 2 arrangements, Australia's "Superannuation Guarantee" or SG looks as though it might be working.  This report from the Reserve Bank of Australia seems to confirm that.  After 15 years of full compulsion and a further six years of a lesser version, there is undoubtedly a lot of money invested in the Australian SG.  And 90% of the work force belongs to the SG.

If a country forces people to save for retirement, household wealth (and retirement incomes) should increase.  Despite the large number of countries that do this, there isn't much evidence of the effect compulsory saving has on behaviour.

So has Australia's SG increased wealth; affected voluntary saving or changed retirement ages?  According to this report, the answers are 'yes', 'not much' and 'no'.

"Australia's experience with compulsory pension accounts is a useful case study since the reforms have been operating for over 15 years. This paper finds that Australia's pension accounts increased household wealth, with an extra dollar in their compulsory pension accounts adding between 70 and 90 cents to household wealth. This result is consistent with some households facing financial constraints that prevent them from fully offsetting the compulsory contributions employers make on their behalf by reducing other assets or borrowing. Voluntary saving for retirement also appeared to increase slightly. This result may be due to the Superannuation Guarantee making households more aware of the need to save for retirement, or the added convenience of being able to make contributions directly into pension accounts set up by their employer. Finally, empirical estimates suggest that there is no significant effect on intentions regarding the timing of retirement. These results suggest that compulsory pension accounts can increase household saving and expected retirement incomes."

PensionReforms has a number of issues with this analysis.  First, it's not clear what the counterfactual might have been in the absence of compulsory Tier 2 provision in Australia.  Given Australia's strong economic performance over recent years, what might citizens have done privately?

Secondly, there is a complex interaction in Australia between private wealth and income after the State Pension Age and Australia's Tier 1 pension that is both income and asset-tested.  Eventually, the compulsory Tier 2 scheme will largely replace Tier 1.  So, when measuring the growth in private wealth through the growth of Tier 2, any Australian analysis of this kind should really take account of the reduction in retirees' public wealth that will eventually follow.

Thirdly, there is no mention in the report of the large and rising cost to the tax system of the compulsory Tier 2 arrangements (and other tax-advantaged savings).  The Australian Treasury reports this to be about 2% of GDP.  Taxes have to be higher to pay for those concessions and higher taxes are not costless, either directly or through the indirect, deadweight costs to the economy.  Even collecting and administering the compulsory Tier 2 contributions will have their own deadweight consequences (again, not referred to in the report).

Fourthly, Australia's aggregate saving rates seem not to have been markedly improved by the story presented in this report.  That accords with PensionReforms' feeling that, while governments have the capacity to change specific behaviour in parts of the retirement saving spectrum, the rest of the economic environment tends to make adjustments to compensate.

The reports from the Reserve Bank of Australia are becoming progressively more positive - see here (2000) and here (2004) for past examples.  PensionReforms wonders though whether the large direct and indirect costs of a compulsory Tier 2 scheme and the consequent disruptions have yet been justified in Australia.  (File size 210 KB)  161