PensionReforms' summary and comments
It's not often that PensionReforms comes across a report that is as poorly founded as this one.  The Australian superannuation industry commissioned it and has got the report it paid for.  There is so much wrong with the report's logic and economic underpinnings that PensionReforms has, after some internal debate, decided to run it as a 'negative exemplar' - what not to do.  The report's conclusions have been widely cited by the press and commentators and that is another reason for running this abstract.

Australia's compulsory Tier 2, "Superannuation Guarantee" (SG) started in 1986 amongst union-based employees with full coverage for all Australian employees in force from 1992.

"This report outlines the evidence that superannuation continues to benefit Australians despite recent falls in the share market, and changes to legislation."

According to the report, the SG helps employees save for retirement; "...benefits the Australian economy because it drives economic growth through its impact on national savings and investment" and "...reduces the amount the Government needs to spend on the [Tier 1] Age Pension..."  PensionReforms suggests that there is either little or no conclusive evidence cited to support any of this.

The report notes that the Age Pension is income- and asset-tested so a higher SG benefit can mean a lower Tier 1 pension.  PensionReforms suggests this then leads to the second potential reason - if everyone has significant private provision, the 'pressure' to increase the Tier 1 pension might presumably be reduced in the future.  But there is another problem associated with what's known locally as 'double-dipping' (not mentioned in the report).  Australians who retire early can spend their SG benefit and still fall back on the Age Pension.  The government intends eventually to align the so-called 'preservation age' and the State Pension Age but that is some way off.  Inexplicably, the report thinks that financing early retirement is a positive contribution to Australia's retirement income regime.

The report reaches the unsurprising conclusion that retirees with superannuation savings (past or present) "...have a significantly higher gross weekly income than those without superannuation and so generally enjoy a higher standard of living."  That also applies to some 'case study' families though, "holding all else constant" the report puts the differences down "...to the more favourable tax treatments and the government co-contribution."  PensionReforms expected that result.

There is apparently some discussion in Australia as to whether the undoubtedly larger amounts of money in the SG schemes might be at the expense of other savings or higher debt.  The decline in the macro 'household saving' data seems to have been unaffected by the introduction of compulsion.  The report offers the following in support:
"Empirical evidence indicates that the superannuation guarantee may have increased the household saving rate by up to 1.5-2 per cent of gross domestic product (GDP).  That is, government policies encouraging superannuation have added to both household saving and wealth, albeit that they appear to have been 'swimming against the tide' of other strong factors reducing saving, and disposing people to incur debt."

Taking that as the starting point and "[u]sing a macroeconomic growth model that relates GDP to the amount of capital and labour in the economy, we estimate that without superannuation, investment would have been [AU]$14 billion, or 4.5 per cent lower than it actually was in 2008 ([AU]$312 billion) and capital stock would have been almost [AU]$144 billion less.

"Extrapolating from these capital and investment growth figures we estimate that without superannuation, GDP would have been an estimated 1.8 per cent lower in June 2008 than it actually was.... This is a difference of almost $20 billion.  Without superannuation, GDP may have been only [AU]$1.06 trillion in 2008, as opposed to the actual figure of [AU]$1.08 trillion."

PensionReforms suggests that the report's justification for this is unfounded - more on this below.  Regardless, the report ploughs on:
"In terms of per capita differences, we estimate that in 2008 without superannuation, individuals would have been worse off by [AU]$928, or almost [AU]$2,400 per household.....  As GDP per capita in 2008 was [AU]$50,586, this equates to a 1.8 per cent difference in per capita incomes. For 2009 we estimate a per capita difference of around [AU]$996, or around 1.9 per cent."

The report suggests that, over time, the difference will increase:
"GDP with superannuation in 2020 is estimated to reach [AU]$1.7 trillion, based on current growth levels. Without superannuation it is likely GDP would fall short by 3.2 per cent, reaching only [AU]$1.6 trillion."

There are other apparent benefits of the Australian arrangements:

. The financial services industry is obviously better off.  Superannuation now apparently accounts for 45% of the "finance and industry sector in Australia".

. The proportion of Australian listed shares held by superannuation schemes has grown "from 8.5 per cent in 1998 to 16.5 per cent in 2007".  A year later, that has seemingly grown to 23%.

. Superannuation has a growing proportion of 'venture capital' - currently "...55 per cent of the total of funds committed toward venture capital and later stage private equity... as at 30 June 2008."

. Retirees are paid benefits of about $40 billion a year and that apparently stimulates the local economy.  PensionReforms doesn't understand why the retirees are picked out for this special mention.

The report acknowledges that the tax breaks are regressive (favour higher earners), but:
"These arguments, however, ignore the significant contribution to taxation generated by the superannuation industry.  This occurs in three key ways.  Firstly, superannuation contributes directly to the tax base.  Secondly, by generating economic activity superannuation increases tax revenue.  Thirdly, superannuation eases the burden on the Government in terms of paying for the Age Pension.  These three impacts - both direct and indirect - sum to just over $15 billion (2009-10)..."  This includes the savings to taxpayers of the reduced Tier 1 Age Pension as a result of the means tests.

PensionReforms thinks this analysis also has little to support it.  While the cited numbers may be a best guess, the analysis ignores the counter-factual.  What would Australians have done had the employer's compulsory 9% contribution been extra pay?  Putting aside the direct cost of the tax breaks (covered next), it's just possible that the return to taxpayers through income and expenditure taxes may actually have been higher than the modelled SG arrangements.  And then there is the direct cost of the concessions that the report acknowledges:
"It would however, be misleading to claim that superannuation contributes significantly to tax revenue without acknowledging the value of superannuation tax concessions, which in 2009-10 were worth nearly $25 billion .... Superannuation tax concessions obviously cost the Government a significant proportion of tax revenue."

According to the report, this high cost is justified because "... without the tax concession, few Australians would contribute to superannuation beyond the mandatory component..." 

Yes but, PensionReforms wonders, even if it were possible to justify tax concessions for contributions in excess of the compulsory SG's 9% of pay (and that's quite a large ask), how can it possibly justify having tax concessions on the compulsory contributions that Australians can't avoid?  If there were any justification for incentives, that might be to influence a change in behaviour.  It might have been nice if the report had looked at what PensionReforms thinks would be an obvious improvement to the SG arrangements - getting rid of tax breaks.

Anyway, even if there were no reduction of other saving because of the compulsory scheme there is, as the report itself effectively acknowledges, no increase in net saving because the cost of the tax concessions (AU$25 billion as above) outweighs the estimated increase in household saving (around AU$20 billion).  That sounds to PensionReforms like a money-go-round to not much effect.

The high cost of tax concessions is not their only problem - Australia amply illustrates their highly regressive nature, their complexity, inequity and also their apparent inability to increase saving; all not covered in the report.  PensionReforms agrees that there may be higher levels of household saving in the presence of compulsion than in its absence.  But the specific impact of tax incentives (as opposed to compulsion) on this was not separately analysed and, in PensionReforms' view, should have been.

Then there is what PensionReforms regards as the report's most serious shortcoming.  It accepts as a given the link between saving, investment and growth; a link that is at least problematic and even possibly nonexistent.  Higher savings may mean more investment and higher investment might mean more growth but not necessarily in either case.  The case for these 'inevitable' linkages was demolished by Nicolas Barr in a 2000 report reviewed here.  Barr also disposes of other policy underpinnings for Australia's compulsory regime.

The Allen Consulting report's analysis essentially assumes that, given a particular relationship between capital and labour (in Australia's case, the historical value, in relation to GDP is 40% capital and 60% labour), adding to capital will necessarily improve the total GDP.  This is simplistic.  Even if that were the past case, just adding capital might reduce the returns to all capital and/or reduce the returns to labour.  Assuming otherwise seems brave and is unsubstantiated.  It also ignores what happens in the real world.  There are examples of countries with low savings accompanied by both high and low growth; also of countries with high levels of savings with both low and high growth.  As Herbert Simon suggested 20 years ago (Herbert A. Simon "On parsimonious explanations of production relations", The Scandinavian Journal of Economics 81:4 (1979), pp. 459-474), "An examination of the evidence suggests instead that the observed good fit of these functions to data . are very likely all statistical artifacts."  PensionReforms thinks that physical capital may, or may not, have anything to do with growth.  PensionReforms concedes there is disagreement and a large literature on this but it would have been nice to see evidence of either in the report.

The report's key message is at best equivocal: "Superannuation can stimulate investment, which enables Australian companies to expand and drives economic growth." (page 35 - emphasis added).  PensionReforms suggests that this is scarcely a resounding conclusion and is scant justification for the very large amounts of money that Australians send to financial service providers (and the high costs of keeping it there).

PensionReforms does, however, understand why the financial services industry is such an enthusiastic supporter of the Australia's Tier 2 scheme and why it paid for this report.  (File size 1.22 MB; 59 pp) 357