Wednesday, May 12, 2010

Economics 12/05/2010: How not to do austerity...

How not to do austerity? Well, Ireland is a good example.

For all the tough talk about reforms and changes to spending habits of the public sector, the new employment in civil service document released two weeks ago, drawn up by the Department of Finance envisions that staffing levels will fall from 37,376 estimated for the end of 2010 to 36,594 at the end of 2012. That’s a whooping (or in terms of SIPTU/ICTU savage) drop of 782 workers, or less than 2.1%. The resultant savings, assuming jobs cut will be at the media level of pay for the civil service, will total a massive €39.41 million per annum. Translated into our public sector’s spending habits, that’s about 16 hours and 20 minutes of our deficit financing for the first 4 months of this year. Not counting the banks costs.

The Government has told the nation before that the new public service pay and reform deal negotiated with unions at Croke Park last month will "substantially" reduce the number of State employees over the coming years. Hmm... guess 2.1% is philosophically ‘substantial’, even if not economically substantive.

But wait, these are gross savings, pathetic as they might be. To get to the net figure, we must factor in early retirement incentives doled out to civil servants by Brian Cowen in Supplementary Budget 2009 and golden handshakes for voluntarily leaving staff.

So take a rule of thumb - the cost of laying off civil service workers ranges around 15-20% of their total annual salary per year of service – once the value of pensions and redundancy payments are factored in. This is very, very much conservative, given the one-off payments and other perks accruing to retiring public sector workers and given that their tax liabilities collapse upon the retirement, especially over the first year. Take 15% on the lower end and assume that average tenure of the workers leaving the service is around 15 years (lower-end assumption as those taking early retirement would more likely to be more senior than that).

What do you have? The cost – and not all of this obviously will hit the taxpayers at one single shot, but most will – will be around €133,400 per worker reduced. And that’s at the lower end.

Savings of €50,294 per annum, at a cost of €133,400 means that given our Government’s innate inability to manage its own workforce, the first time we, the taxpayers, will see positive net savings on the deal (assuming opportunity cost of funds at 5% and automatic stabilizers on the salary payments to public sector workers at 30% - income tax, levies, etc - none of which are going to apply under voluntary retirement) September 2015!

I am not kidding you – September 2015! By which time, of course, the Unions would have forced the Government to get a new Benchmarking going…

Folks, we are now truly turning the corner!

11 comments:

Anonymous said...

Your comments display a basic lack of understanding on your part. The "cost" figures you have used are totally nonsensicleI won't go into the details but you can read comments on it here:

http://boards.ie/vbulletin/showthread.php?t=2055909983&page=5

TrueEconomics said...

Dear OMD.

I am saddened by the fact that you had no time to post here your more detailed arguments. And I followed some of your arguments on the post linked above. Needless to say, I disagree with them.

Apparently, you skipped the first two paragraphs of the note which does estimates of savings on gross basis. Even these are hardly impressive by all possible means. Let me remind you - the country is running a deficit that will add up to roughly speaking 34-36 billion over 2011-2012, the range over which these figures apply. Gross savings are under 40 million. Any idea just how disproportional your entire diatribe on boards.ie is to task at hand?..

Now to your critique of my assumptions:

1) Your sssertion that prior to the figures estimated by the DofF note there have been reductions in PS numbers is simply unrelated to anything that I wrote about in the note as we are talking about 2011-2012 figures per DofF numbers.

2) Assertion that my numbers overestimate the costs due to excluding natural attrition rate is a red herring. I do not impute a number of future costs (increments or subsequent reforms of public pensions that can result in increased provisions that will be foregone). Furthermore, per April 2009 note from DofF: "vacancies arising from the [early retirement] scheme may not be filled without its specific sanction and that secretaries general will be expected to reorganise or restructure work or business units in order to protect service levels as far as possible." In other words, natural attrition and prior early retirement uptakes are either already factored into the latest DofF numbers, or the entire exercise of posting the latest numbers and making noises about reforms in the Croke Park was a pure PR exercise. Take your pick. Figures across the departments contained in DofF note actually show that in some cases (the case of CSO, if I recall correctly) there will be an increase in employment levels followed by a decrease - over the span of 2011-2012. This, surely, shows that at least some of the changes in numbers are not envisioned to come from natural attrition alone.

3) I do not explicitly account for the lower provisions to cover pensions due to a shorter length of pensions contributions that result from earlier retirement. These would increase further the costs of early retirement.

4) Present value of future contributions (and benefits) are more than offset in my calculations, since benefits arrive at a slower rate, initially, than costs. But I factor in only 5% as the opportunity cost of direct deficit financing today. If I take account of automatic stabilizers on the net debt incurred to fund these early retirements, this cost actually would rise for my calculations to ca 6.0-6.5%. Should I bother to add the cost of issuance and administration of this debt, the true cost of deficit financing of early retirement in public sector will increase to 6.1-6.6%. And if I were to use economic opportunity cost, this figure would be in excess of 8.5%. Get the idea?

As it stands, one cannot really precisely estimate the effects of these (or any) reforms or the actual exact timing of the savings to be delivered. My assumptions are transparent and the data I used comes straight from public sources, such as CSO.

Part 2 to follow...

TrueEconomics said...

Part 2 of reply...

So now, let us entertain ourselves and take your own numbers from boards.ie:

"Say you are earning 50K a year with 39 years service due to retire next year on full pension. So you take early retirement this year. You get €73k lump sum and a pension of €24,375 (total cost to government €97500 this year but a saving of 50K (wages not paid) so overall cost €47,500. However you were going to retire next year anyway getting: Lump sum: €75,000 and a pension of €25,000. Over the 2 years. If you retire early the overall cost is (73,000 + 24,375 + 24,375) €121,750. If you keep working the extra year and then retire the overall cost is (€50,000 + 75,000 + 25000) €150,000. So the government saves 28,250 by retiring one year early."

Not really, OMD.

Suppose the cost of early retirement for this employee is indeed €121,750. You claw nothing on 73K in terms of taxes and you forego taxes on 50K annual salary. So cash flow for the Government over years 1 and 2 is:
Gain +€100,000 (2 years worth of salary payments)
Loss -€121,750 (paid out in pension benefits and lump sum)
Loss -€27,800 (approximate tax liability at €13,932 per annum over 2 years)
Loss -€30,000 (approximate value of underfunded pension at 30% annual income – note, it is actually estimated to be higher by the Life Strategies research)

Net loss: €79,550

Now, suppose, for simplicity, that this net loss accrues in year 2 (to avoid PDV adjustments). Since this loss is recoverable out of savings generated from not paying this individual her salary in years 3 and on, again, abstracting away from PDV adjustments, in year 3 and forward we have
Exchequer saves €50,000
Exchequer loses €13,932 in tax returns that would have accrued, plus €24,375 in pension cost.
Net savings in year 3 are therefore just €11,693… And the same in year 4… and so on. Until we pay down the Net Loss of €79,550… And forget the trivialities of NPDV. Your numbers just do not add up, OMD.

Now, if you care to apply this to 50% of the ‘expected’ reductions, so as to reflect an assumption that 50% of the reductions might come through attrition and the rest through early retirement (to give our Government a benefit of the doubt here), you get pretty much the same date as I did – around end of 2015 – for a break even on average.

See the light yet, OMD? It’s not about ideology, it’s about numbers… Let me elucidate: you see, if one lives an insolvent household, not buying a new Mercedes Benz today is not really a saving of money – it’s avoidance of spending that one cannot afford at any rate. This is the problem with the Croke Park logic – it is about ‘saving money by reducing what we might have paid tomorrow’. Alas, we are broke. We need to save today – i.e we need to cut current spending. Not future expected spending. Current spending.

But keep on commenting, and reading...

AL said...

Great post Dr CG.

Its no wonder the civil servants didnt have the neck to show their face on the streets... we need a proper axe taken to the civil service, massive pay cuts and privitisation of as many services as possible (starting with Passport service).

Anonymous said...

Public servants receive 10% of their tax free lump sum now and the rest at their normal retirement date. (your figures incorrectly say they get 100% now)

The actual tax free lump sum will be lower as they will have less years contributions. So in your example the person on 50K retiring after 39 years service will receive €7,300 not the €73,000 you are saying. So the government pays €7,300 this year and €66,700 next year.


If the person stayed at work for the extra year the government would pay him €75,000 next year. Not only that but the government is saving €625 a year for the rest of that employees life as well as the €26,625 they save this year for not having that person employed.

So again person retires early on a salary of 50K a year with 39 years service. Cost to government: Lump sum €7,300 and pension of €24,375 which totals €31,675. The government saves €50,000 by not having that person employed. Savings to government in year 1 is €18,325.
In year 2 the saving is €1,875 (from reduced tax free lump sum) and €625 from reduced pension.
Saving in year 3 and every year thereafter is €625


You also ignore the fact in your last post that the lump sum and pension are going to be paid anyway they are just being paid a year early.

To AL: I am not a public servant but I think a discussion should be based on facts.

Anonymous said...

Most of your comments are some sort of lecture on cutting costs in Public service. I fully agree with this. Early retirement with a deferred and reduced tax free lump sum will save money. Obviously there are issues like taxing lump sum that could be done to save money but as you made no reference to this I did not address it either. Unless you are talking about fundamental pension reform (which your post did not do) then you have to accept that the pensions will have to be paid at some stage.

TrueEconomics said...

OMD, I took your own figures in my reply from the boards.ie.

I totally agree, the debate should be factual and it should be free of ad hominem arguments - something that the commentators on the linked boards.ie thread do not seem to care about much. And I agree with you on the need for pensions reforms for public sector and, indeed, for private pensions provision.

Since you mentioned it - yes, pensions and lump sums will be paid a year earlier. Which relates to the issue of what type of 'savings' we are looking at. The point is that, assuming the state is actually going to continue honoring these 'contracts' on public pensions, the savings to be generated from PS numbers reductions via early retirement are going to be absolutely negligible. And that's the whole point of my post.

Unknown said...

You are misintreptting what I said. You said:
"Since you mentioned it - yes, pensions and lump sums will be paid a year earlier."

This is not what I said. Indeed it is almost the oposite of what I said, and is the part you ignore in your comments: The tax free lump sum is not paid a year early. Only 10% of it is paid early. The remaining 90% is paid at normal retirement age. This is the vital issue that you ignore. Add to this the fact that the overall lump sum (and pension) is lower. Add again the fact that the pension is not underfunded by any more by retiring early (indeed it is less underfunded)and you can see that the figures you use are in fact fantasy.

Early retirement will save money from year 1 (not 2015)which is totally different to the main thrust of your original article.

TrueEconomics said...

Des, I actually used your figures... These forced me to deal with a one particular example - an employee who is one year away from retirement.

Also, by saying that civil servant working one less year does not underfund his/her pension by any more than a civil servant who works an extra year, are you saying that these workers length of tenure has nothing to do with funding of their pensions? Really? I might just agree with you on this one, but that simply means that their pensions bear no connection to their work, whatsoever...

Accepting your logic on numbers and using for illustrative purposes,once again your figures:

If you retire early the overall cost is (73,000 + 24,375x2 + Foregone Tax on earnings of €13,932 + Foregone pension funding for 1 year (6%?) - €625 saved on future liabilities annually)~ €121,750+13,932= €138,682.

If you keep working the extra year and then retire the overall cost is (€50,000 + 75,000 + 25000 - Tax on year 1 earnings of €13,932 - Pension levy (6%?)) ~ €133,068

No break even, still...

Lastly, in all of the above, we are both assuming that the civil servant retiring early and not being replaced was either completely unproductive (in which case surplus value of this employee over and above his/her wage to the employer was zero) or his/her work can be assumed by other employees (in which case on the net the entire system was made more productive by the early retirement. Either way, we clearly have an implicit assumption built in that prior to the early retirement, the system we had was wasteful.

Ok, at this stage, I am moving on - good arguments, though... keep them coming!

Anonymous said...

The underfunding of pension is a given. Public servants do not fund their pension adequately. We all know that. If you work 1 year less yes you pay in less but you also get a lower lump sum and a lower pension. So by retiring earlier your pension is not as underfunded as it would have been.

On this point:
If you retire early the overall cost is (73,000 + 24,375x2 + Foregone Tax on earnings of €13,932 + Foregone pension funding for 1 year (6%?) - €625 saved on future liabilities annually)~ €121,750+13,932= €138,682.

If you keep working the extra year and then retire the overall cost is (€50,000 + 75,000 + 25000 - Tax on year 1 earnings of €13,932 - Pension levy (6%?)) ~ €133,068




You have included tax on income but seem to assume the pension is tax free.

Also you count the tax of 13932 twice which is incorrect.It can only be counted once. Also the figure is incorrect anyway.
The actual figures are:
The underfunding of pension is a given. Public servants do not fund their pension adequately. We all know that. If you work 1 year less yes you pay in less but you also get a lower lump sum and a lower pension. So by retiring earlier your pension is not as underfunded as it would have been.

On this point:
If you retire early the overall cost is (73,000 + 24,375x2 + Foregone Tax on earnings of €13,932 + Foregone pension funding for 1 year (6%?) - €625 saved on future liabilities annually)~ €121,750+13,932= €138,682.

If you keep working the extra year and then retire the overall cost is (€50,000 + 75,000 + 25000 - Tax on year 1 earnings of €13,932 - Pension levy (6%?)) ~ €133,068

There are a number of problems with your figures. You ignore tax on the pension, you miscalculate tax due on income (a person earning 50k will pay 12800 a year in pension contributions, levies and taxes)
But most basic of all is the mathematical error of counting the tax twice. ie adding it to the figures in the first example and subtracting it in the second. You can only count it once. Once you see the error you can see a saving is made in year one.

For example:

If you retire early the actual overall cost is (73,000 + 24,375 X 2) = €121,750.
This is the actual cost to the government. ie the amount the government has to pay. It is mathematically incorrect to add lost tax here. The government does not pay out this lost tax.

In the second example:

If you keep working the extra year and then retire the overall cost is (€50,000 + 75,000 + 25000 - Tax on year 1 earnings of €12800 =€137200.
This is the amount the government actually pays out. 50K salary less tax and pension contribution of €12,800 + tax free lump sum plus pension

So you can see the government has saved almost €16,000 by year 2 even before we discuss tax on the pension received. Far more than break even.

Anonymous said...

Sorry in my previous post I was doing a lot of cutting and pasting. This section should be deleted

"On this point:
If you retire early the overall cost is (73,000 + 24,375x2 + Foregone Tax on earnings of €13,932 + Foregone pension funding for 1 year (6%?) - €625 saved on future liabilities annually)~ €121,750+13,932= €138,682.

If you keep working the extra year and then retire the overall cost is (€50,000 + 75,000 + 25000 - Tax on year 1 earnings of €13,932 - Pension levy (6%?)) ~ €133,068




You have included tax on income but seem to assume the pension is tax free.

Also you count the tax of 13932 twice which is incorrect.It can only be counted once. Also the figure is incorrect anyway.
The actual figures are:
The underfunding of pension is a given. Public servants do not fund their pension adequately. We all know that. If you work 1 year less yes you pay in less but you also get a lower lump sum and a lower pension. So by retiring earlier your pension is not as underfunded as it would have been."