Thursday, 12 August 2010

US Lessons on the Failure of Pension Tax Arrangements

Sinéad Pentony: Last week, Professor Teresa Ghilarducci's spoke at a pensions seminar co-hosted by TASC, TCD Pension Policy Research Group and the INTO. See her presentation here. Professor Ghilarducci is the Bernard and Irene Schwartz Chair of Economic Policy Analysis and the Director of the Schwartz Center for Economic Policy Analysis at the New School for Social Research, New York.

Her presentation focused on the 401k pension system and showed that pension tax reliefs, as presently structured, result in a significant increase in inequality in the US and that a change in the balance of pension provision in favour of public rather than private pensions is necessary to provide a guaranteed income in retirement. As Ireland starts the process of implementing the National Pensions Framework, we need to take a step back and examine the experience in other jurisdictions. The 401k system has clearly not worked in the US, yet we are proposing to go down a very similar road here.

Ghilarducci’s research on the 401k (defined contribution) system of pension provision clearly shows that this system has failed to provide an adequate replacement income in retirement for millions of Americans. This system of pension provision is similar to Ireland’s PRSAs and they tell a very similar story:
• They are voluntary.
• They have failed to increase pension coverage - half of all workers in Ireland do not have a private pension and 64 million Americans at retirement age are without adequate pension provision.
• They are costly - because of the fees and charges of private providers.
• Tax reliefs have failed to increase pension coverage and they disproportionately benefit high earners -in Ireland 80 per cent of pension tax reliefs accrue to the top 20 per cent of earner; in the US, Ghilarducci contends that while pension tax reliefs in the USA are regressive, they are less regressive than in Ireland.
• They fail to provide an adequate income in retirement - as increasing the value of pension funds is largely dependent on the performance of the stock market and to a large degree, the gains in value are eroded by fees and charges, which is the case in both Ireland and the US.

Having researched 30 years of defined contribution pensions in the USA, Ghilarducci has developed a proposal for a Guaranteed Retirement Account (GRA), which resonates very stongly with the TASC/TCD model of pension provision. Her proposal highlights the importance of the social security pension, as this is what most people rely on for an income in retirement.

Her proposals are:
• A supplementary, mandatory and state-led system that requires all workers and employers to contribute through the social insurance system.
• The state also contributes through tax credits.
• Investments are managed by the State.
• People are provided with a guaranteed income (adjusted for inflation) in retirement.

Ghilarducci has estimated that state contributions to GRAs would be cost neutral if tax reliefs were redistributed (through tax credits) in favour of low and middle income earners.

The aim of the National Pension Framework is to deliver choice – but real choice can only be delivered if people are given the option of saving in a state-led and state-guaranteed supplementary system of pension provision. The evidence clearly supports the need for such a system. The proposals set out in the NPF are centred on a pension system based on tax reliefs which are costly, inefficient and inequitable, even at 33%; and managed by the private pension industry – an industry that has demonstratably failed to deliver security for many Irish workers in retirement.


Anonymous said...

How come no-one on this site ever complains about the obvious inequities associated with public service pensions?

(Hugely disproportionate benefits to better paid public servants, especially those with the potential for late-career promotion).

Oh yeah, I think I just figured it out. The clue is in the affiliations of the contributors.

Sinéad Pentony said...

Thanks anonymous,

Teresa Ghilarducci was asked a question on inequities associated with public service pensions. She called it 'pension envy' and said that the goal should be to bring everyone up to the same level of pension provision rather than bring everyone down.

I would agree with her on this, but also think that pensions should be subject to progressive taxation, so if you're on a very high pension (public or private)you pay income tax.

Anonymous said...

Thanks for that perspective Sinéad

Strange though that you don't see any contradiction here. To object to the state's largesse in the form tax relief for private pensions is like totally righteous, whereas to even question the extreme generosity from the very same source in the form of public sector pensions is simply the green eyed monster at work.

I wonder also do your reform proposals in the interest of "equity" include any or all of the following:

* limit tax relief on public sector superannuation contributions
* limit tax relief on pension levy payments
* apply benefit-in-kind tax to the difference between the actuarial cost of public sector pension entitlements minus superannuation + pension levy
* apply income tax to the 150% of final salary lump sum

And if not, why not?

Brendan Quinn said...

• A supplementary, mandatory and state-led system that requires all workers and employers to contribute through the social insurance system. - Bad idea
• The state also contributes through tax credits. - very bad idea
• Investments are managed by the State. - really terrible idea.
• People are provided with a guaranteed income (adjusted for inflation) in retirement. - Yes, but through basic income process.

Mack said...

Sinéad -

If you are guaranteeing me the same pension, provided by the state, as public servants - then sign me up.

If you are offering the rest of us a different product while they get to keep that - I'd be suscpicous that what you are offering is as good as you say. If private sector workers are to give up tax reliefs & choice - what is the quid pro quo from the public sector, if we are not talking about the same system being put in place for all?

Anonymous said...


The key phrase in the TASC/TCD pension document is:

"contributors would be guaranteed 50 per cent of their final wage or salary up to a specified maximum"

(emphasis mine)

The actual amount of this limit, or whether it will apply equally across the board, is conveniently left unsaid.

But I can guess where this is going ... set the limit fairly low in the interests of "equity", while excluding current public servants from the restriction (legitimate expectations, yada yada, not really a pension at all more like deferred income, yada yada, we didn't cause this mess).

The net effect is to consolidate the existing advantages of the well-pensioned, while locking many more contributors into the public system. The huge number of extra contributors (mostly on lesser benefits) guarantee long-term viability of the public pension system, otherwise the unfunded public sector pension liabilities would surely trigger drastic reform or even outright collapse within a couple of decades as demographic changes lead to the PS pension outgoings totally outweighing current superannuation contributions.

An Saoi said...

Anonymous, There is very little between the average Public Service pension & a person receiving a SW OAP with adult dependent.

You are correct that late career promotion or more likely imputed added service for certain professional grades and pensionable allowances are very costly. None of those involved with TASC has ever defended excessive pensions within the Public Sector. However the numbers receiving very large pensions are very small. A maximum pension level should be set

However let us take a retirmennt grade of Asst. Principal. His max salary is now €80,679, giving him a pension of €40,340 less single OAP of €11,976 or an occupational pension of €28,364. His occupational pension is 35.16% of his final salary. The actuarial cost of his occupational pension is approx. €620,000.

The problem with tax based pension schemes are numerous.

1) Most of the tax relief has accrued to a very small number of people, perhaps 2/3 of the relief going to just 10,000 people.

2) Many of those SSAP Schemes were not used to provide pensions, rather as investment & CAT avoidance vehicles.

The current Public Service schemes have some flaws, but provides the State with very good value for money. The accrual rate should be perhaps revised so as to take into account improved life expectancy and health outcomes, however it should be noted that Public Sector workers are now paying up to 16.5% of their salaries for their pensions. The State contribution is now minor.

Anonymous said...

An Saoi,

A few points about your example APO ...

If she retires before the end of the year, her pension will actually be calculated on the pre-paycut level of €84,746 as opposed to the €80,679 as you stated.

However if she was to retire within that timeframe, its most likely she was a pre-1995er on a slightly lower payscale, but with little or no PRSI contributions through her career and hence no right to the state OAP.

The OAP adult dependent argument is weak, as increasingly into the future the spouse will have their own independent pension. It'll be approximately 2035 before the post-1995ers start retiring and at that point the proportion with fully dependent spouses will be small (given the historical trends around female participation in the workforce). Up until then, the state OAP is irrelevant as the pre-1995 retirees would never have paid PRSI.

Presumably you're counting the pension levy and superannuation payment as basis for the 16.5% contribution quoted? You realize of course that it will be nearly 2050 before anyone retires having made a full career worth of pension levy contributions? Even allowing for that, the actuarial cost of the PS pension has been calculated at approx 25% of salary for civil servants, 50% for Garda and in excess of 70% for judges. So there's still a massive contribution coming from the state.

I'd love to hear the basis for your calculation of an actuarial of €620,000, but given that €120k would be accounted for straight off the bat by the tax free lump sum, €500k seems a bit light for a pension that rises faster than inflation (in line with current post-holders).

Brendan Quinn said...

I have a solution to the pensions problem. Stop investing in stocks and shares or limit to 20% of investment. Get government pension trusts to buy land abroad. Not just property, land in different parts of world. Then live off the rent created by the economic hard work of others. This would provide a continuous income for Ireland and her citizens that would be used for pensions etc

Sinéad said...

Thanks for your comments.

The TASC/TCD pension proposals aim to provide adequate income in retirement for low and middle income earners, through a universal state pension and supplementary earnings-related social insurance pension. The proposals are based on research undertaken by TCD and the ESRI on the impact of tax reliefs for private pensions. In other words, the aim is to remove tax breaks that disproportionately benefit high earners, and use some of those resources to benefit low and middle income earners and the rest to support a more sustainable pension’s model. TASC’s proposals do not involve contributors to the proposed system subsidising or funding public sector pensions.

There isn’t as much data available on public pension tax reliefs so it’s not yet possible to get into the specifics of the tax treatments of public pension tax reliefs. However, in principle, TASC would support changes to the public sector tax reliefs that make the system more progressive in favour of low and middle income earners. TASC does not support advantaging public sector workers at the expense of private sector workers. In this context we’d also support capping and/or taxing public sector pensions. It should also be noted that a new “career average earnings-related” public pension system is due to be introduced this year.

The TASC/TCD pension proposals don’t recommend the abolition of private pensions – but instead recommend the creation of a supplementary state-led system of guaranteed pension provision aimed at low and middle-income earners. High earners can continue to exercise the choice to take out a private pension, but this shouldn’t be subsidised by the state.

Notwithstanding the issues concerning tax reliefs, there is a bigger problem with the private pension system – problems that have been highlighted by the financial and economic crises. Many people have lost all or part of their pensions. Half of all workers don’t have a private pension and many of those who do, won’t have enough in their fund to provide them with an adequate income in retirement.

We had the worst performing pension funds in the OECD in 2008. While they have recovered some of their value, it will be many years before they return to 2008 levels. The Pension Board, in its recent annual report for 2009, highlighted the problem of risky investment by private pension providers. An OECD Survey of Investment Regulation of Pension Funds highlights the lack of regulation of Irish pension funds with regard to risk management. We are one of the few OECD countries that do not have quantitative restrictions in place; i.e. setting minimum levels of investment in secure Government bonds and maximum limits in high risk investments. The countries that have these restrictions in place experienced fewer losses in their pension funds.

The TASC/TCD proposals aim to address at least some of these problems by putting a supplementary state-led system in place through social insurance.

Mack said...

Sinéad -

It's a big improvement that the state would manage a fund out of which future pensions would be paid (as opposed to the pay as you go / social insurance model). I think a two tier public pension system is probably unworkable. On the one hand it appears that above average earners in the private sector are being asked to fund this scheme, while equivalent workers in the public sector remain exempt. I imagine this might make gaining wide spread support more difficult (i.e. you might well face opposition, not only from the pension industry but also from a lot of motivated voters in the private sector unless the measures are deemed even handed).

I'm in the middle of reading Progressive Economy contributor, Stephen Kinsella's book Ireland in 2050. He seems to think that the public pension system is going to start breaking down around 2020, as the dependency ratio starts to worsen.

I see the age 65 mentioned frequently in the report, but moves are already afoot to increase the retirement age to 68. If life-expectancies continue to rise, (and with them medical costs) - will these pensions - with a retirement age of 65 - still be affordable? Or once we hand complete control over to the state are we likely to see the date of our retirements pushed ever further out?

Mack said...

Also, given the current raiding of the National Pension Reserve Fund (and the history of Robert Maxwell type raids elsewhere) - do you propose any taxpayer / future pensioner protections for your fund?

kirneh98 said...

@ Anonymous
The implementation of the Tasc proposals would mean that tax relief on public service superannuation and pension levy deductions would be limited in the same way as private sector pension contributions as Revenue treats both public and private sector pension contributions identically. Regarding taxing the lump sum, are you suggesting this for all workers who can take a lumpsum at retirement or just public servants? Likewise, do you expect private sector workers to pay benefit-in-kind on the employer contribution to their pensions?

Regarding actuarial calculations of the value of PS pensions, the most up-to-date figures are from the C&AG Special Report 68 (2008). The figures area lot lower than the ones you are quoting. The figure for Gardai is 27.9% gross before contributions and pensions levy are deducted, not 50% as you posted. And if the Government used the discount rate assumption of 7% real which is used in selling the National Pensions Framework, the alleged value of PS pensions would be much lower still.
Have you any ideas for improving the lot of 1,000,000 Private Sector workers who have no occupational pension at all and who would be the big winners if the Tasc proposal was implemented?

Anonymous said...


Are you familiar with Moloney and Whelan's work on valuing pensions, and their contention that risk-free pensions are under-valued?

But leaving aside such technical arguments, even the roughest back-of-an-envelope calculation would show the estimate of 27.9% of salary as the value of the Garda pension to be utterly non-sensical.

A Garda may retire at 50 after 30 years service on a full pension. Their life-expectancy at 50 would be about a further 30 years (higher for retiring Ban Gardai). Hence a Grada retiring at 50 is likely to draw their pension for longer than they actually worked! They also get the usual lump sum of 150% of final salary. Their pension rises with the salary of the current post-holder (historically much faster than inflation). So we can very roughly value the pension at (30+3) times (50% of final salary) over (30 years worked). The factor of 3 is the lump sum. Given that the final salary is likely to be at least 20% higher than the inflation-adjusted career-average salary, a valuation of 50% of actual salary seems on the low side and 27.9% is just silly.

Regarding taxing the lump sum, are you suggesting this for all workers who can take a lumpsum at retirement or just public servants?

Yes, certainly for any worker regardless of sector who gets a lump as of right that does not reduce the value of their recurring pension payments.

For workers who reduce the value of their annuity by taking a lump sum, the tax treatment could be a little less harsh.

Likewise, do you expect private sector workers to pay benefit-in-kind on the employer contribution to their pensions?


That's the way PRSAs work currently.

kirneh98 said...

@ Anonymous

I am familiar with Moloney and Whelan's work on valuing risk and also with Whelan's work on the lower cost of PAYG pensions as compared with funded pensions (ESRI Quarterly Economic Commentary summer 2007).

Re the garda pension valuation, I didn't invent the figure (I know you didn't suggest that I did). It was actuarially computed by the same private actuarial firm who came up with earlier, much higher figures. You can't dismiss the actuarial figures you don't like in favour of ones you do especially since you propose using actuarial evaluations to compute benefit-in-kind.

Also, while PRSA employer contributions may attract benefit-in-kind, is the same true for non-PSRA DB or DC pensions in the Private Sector? I think not, although I would be interested in a reference if I'm wrong on that. Would you have a reference for your PRSA point?

Still wondering if you have any suggestions for improving the pensions environment for woprkers who have no occupational pensions?

Anonymous said...


Re the garda pension valuation ... It was actuarially computed by the same private actuarial firm who came up with earlier, much higher figures.

Hmmm ... the devil as always is the details. Perhaps they assumed in their calculations that most Gardai will voluntarily work on past age 50 for no extra pension years, but with the chance of additional benefit in the form of a promotion while in their 50s. Or just to keep themselves busy and get out of the house ;)

Also, while PRSA employer contributions may attract benefit-in-kind, is the same true for non-PSRA DB or DC pensions in the Private Sector? I think not, although I would be interested in a reference if I'm wrong on that. Would you have a reference for your PRSA point?

You're correct, employers contributions to normal DC schemes are not subject to BIK.

In the case of PRSAs however, there is a BIK liability, see:

Still wondering if you have any suggestions for improving the pensions environment for workers who have no occupational pensions?

The first step would to get our hands around the problem in more detail. It would be useful for example to estimate the average gap between the state OAP and a 50% replacement income rate for these workers without occupational coverage. In many cases, this gap will be small or non-existent, as they might have been working part-time or be low-paid (in which case the gap could be made up with relatively small increases in the OAP, funded by the sort of taxation measures that I suggested). Also it would be useful to estimate the proportion of those 1 million pension-less workers who will retire with a substantial asset in lieu of a pension (e.g. a farm, business, or buy-to-let property) or have a spouse with substantial pension coverage.

Not trying the minimize the problem or anything, which I realize is serious. However it would be useful to drill down into the detail or what proportion of that million workers are in need of substantial help.

kirneh98 said...

@ Anonymous
Thanks for the PRSA reference. This seems to me like double taxation. The employer can set aside the contribution against tax but the employee then gets hit for BIK and pays tax on the eventual pension. I would think the fairer option would be to exempt employees from BIK if DB and DC scheme members in both private and public sectors are not hit for it.

Re the garda figure, the set of assumptions used are standard and the report contains a detailed explanation of how the figures are arrived at and why they differ from previous figures in the Commission on Public Service Pensions and the Second Report of the Public Service Benchmarking Body.

My main contention regarding these figures is that they use a false paradigm that confuses value with cost. If a bottle of water costs 70c in Tesco but e1.50 in a corner shop, what is its value? It depends where you buy it. Private sector workers are condemned to buy their pensions at the corner shop with high management fees eating up their money. The Green Paper on Pensions shows that fees of just 1.5% reduce the final pension by 30%, a staggering figure. Put that together with the poor performance of Irish pension companies and it will take a miracle for many private sector workers to retire with dignity a miracle that seems highly unlikely.

Those costs do not exist for Public pensions includeing the state pension and are the reason that Whelan shows the cost benefits of PAYG versus funded. Yet they are always included when coming up with comparative figures for public-private pension comparisons.

I agree entirely with your point about needing more detail. But I also think we need to look beyond these shores to how pensions are provided in other countries where they don't speak English. Whelan has also pointed out that if no changes whatsoever were made to our current pension regime and you use the standard population projections, by 2050 our pension spending as a proportion of GDP would be less than the European average today.

Mack said...

kirneh98 -

It's possible to purchase etf's with Total Expense Ratio's of less than 0.2% (for example Vanguard's Total Stock Market ETF has a TER of just 0.07% ). I don't doubt that some people are paying way over the odds, and you can't yet purchase ETFs in your pension in Ireland, but there is no reason why they should have to suffer those fees.

The PAYG model works fine as long as the number of workers to retirees is increasing, when it goes into reverse (and it will in some style, thanks to longer life expectancy and lower birth rates) it will force tax increases on the young. Laurence Kotlikoff has some scary maths on the size of the unfunded liabilities stateside.