PensionReforms' summary and comments
The report summarises the different ways that governments treat the calculation of taxes on employment income, on capital income and the way in which pensions payable by the state are calculated.  It was partly inspired by work the author was doing on pensions (a report is reviewed here).

This report warns against the adoption by economists of findings from "individual studies literally as a basis for policy thinking, rather than seeking inferences from an individual study to be combined with inferences from other studies that consider other aspects of a policy question, as well as with intuitions about aspects of policy that are not in the models."

Having set the scene, the report observes that the amount retirees receive from pensions usually depends heavily on the history of the individual - career incomes, years of contributions or the contributions themselves, age, years of residence etc.

There are rarely similar kinds of historical influences in the calculation of the tax that individuals pay each year and economic models that look at the 'optimal' designs of tax systems do not usually incorporate end-of-life pensions into their calculations.  And models of pension systems typically take no account of the "annual taxation of labor and capital incomes".

The report suggests that these are connected issues and it seems impossible to look at the implications of one without also considering the implications of the other.  In that context, the report raises for consideration the issue of tax-favoured retirement savings, "a topic where the two subjects come together."

PensionReforms notes that, in this discussion, we need to distinguish between publicly and privately provided pensions.  The latter are the product of private arrangements that usually are part of the employer-employee relationship.  While tax incentives to encourage such arrangements are certainly a matter of public interest, the basis on which the pensions themselves are calculated are not.  So, when the report talks of "...pension benefit determination depend[ing] on individual history far more than taxes do" we take the public policy component in that to be of direct relevance to the calculation of public pensions.  This raises a key question. - should they be so dependent?  PensionReforms thinks generally not, except with respect to the consideration of age (which is really a proxy for dependency).  It seems inappropriate for the state to provide "lifetime earnings insurance" that is at the heart of the models discussed.  Poverty alleviation; even providing sufficient income for a degree of 'participation and belonging' seem to PensionReforms to be different considerations.

As the report notes, tax is not similarly related to "individual history".  Generally, the amount paid by a taxpayer is dependent on the amount of income received in the year of calculation.  

PensionReforms thinks some might suggest that Social Security contributions to secure 'entitlements' could be included in the list of things that affect tax calculations in an elaborate and indirect way.  These are certainly affected by earnings and, despite their apparent linkage with eventual benefits could be seen as just another tax that is required to support current PAYG benefits.  Others see Social Security contributions as a forced saving, building up claims on tomorrow's taxpayers and so fundamentally different to 'ordinary' taxes.

The report notes that it is generally recognised with respect to changes to pension arrangements that long periods of notice are required - what the report calls policy "inertia" is perhaps justified..  Similar considerations seem not to apply to tax where policies can change from year to year.

The report suggests there might be a case for a tax regime that applies just to the accumulation of pension savings:
"This is suggestive of a possible role for capital income taxation that varies with the age of the saver and/or with the time lapse between savings and later consumption.  And it points to a potential welfare gain from tax-favoring retirement savings, since retirement saving tends to be for longer times.  Also, the role of capital income taxation when future earnings are uncertain suggests that capital income tax rules might well be different for those at ages when people are mostly retired, a common feature of tax-favored retirement accounts."

It wonders whether the fact that withdrawing retirement savings usually incurs a "penalty" perhaps justifies a penalty that varies with the length of time the savings have been accumulating.

The report also looks at the literature and finds gaps:
"I conclude by repeating my call to avoid over-reliance on any single model and with the usual researchers' plea for more research.  In particular, I think we have done too little study of the issues around tax-favored retirement savings accounts, studies that need to recognize uncertainty in future earnings, uncertainty in future spending needs, diversity in savings behavior and earnings opportunities, and uncertainty about future tax rates."

PensionReforms is unconvinced.  We agree that models can go only so far and tax models that ignore pensions (and vice versa) are probably unhelpful in this context.  However, the report's thesis for a possible justification of favouring retirement savings depends on 'voluntary' retirement savings as having a fundamentally different character from other savings.  While 'other' savings may turn into 'retirement' savings (because they are consumed after retirement), there seems little justification to treat them differently from the outset on that ground alone.

And then there is the other collateral damage that such favoured savings create.  First, they are very expensive; next they are regressive; they are distortionary because they favour one type of saving over others; they are complex to administer (and need thickets of regulation to ensure they are applied to their intended purpose); they tend to be inequitable and probably raise financial risks of concentration on the favoured institutions.  They also insulate savers from the full impact of negative outcomes and allow uncompetitive financial firms to survive.  But those are not their greatest difficulty - they also seem not to raise saving levels. It will be interesting to see how the economic models allow for these difficulties.  (File size 139 KB; 28 pp) 334