Looks like Buttons is up for taking on any of his fellow bloggers behind the bike shed after class.

Nothing like a good bit of blogger biffo.

The EPI memo which the offending post is based on claims to take into account pensions in its analysis of compensation. The summary does not say how this is done, and I am suspicious of its conclusions in any case, but it does mention the concerns you are pointing out.

But in the general discussion I do strongly agree that people do not take adequate account of the huge incentive which a guaranteed pension represents.

So, what I'm finding amusing in this "debate" is that on the one hand, it is clearly, demonstrably (yet not quite undeniably) true that public pensions are a huge looming train wreck. It is not clearly true and so far not demonstrated that the existence of public employee unions or collective bargaining make that situation. If anything, it's the protests that seem to affect the legislators as much as the contracts or even the corruption.

So we seem to have one side saying the pensions are the problem and the other side saying "the union is not the problem," and my inclination is to agree with both.

It seems to me the real villians are the past legislators who approved the pension terms which present legislators are now finding burdensome. The hackneyed phrase, "Kick the can down the road" resonates. The employees took the jobs knowing the terms and performed the work expecting the contracts to be honored. The legislators took the votes knowing someone else would have to pay up. It is a shame time doesn't seem to run backwards.

Hmmm, I don't know that comparing the pensions for the Bank of England and the teachers in Wisconsin is exactly apples-to-apples. I'll bet they compare more closely to the pensions for American members of congress than for front-line public sector workers.

But I do like introducing pensions as percentage of payroll as another number into the "debate" (as Doug correctly fingerquotes).

The problem with it, however, is that it becomes hard to fit comparisons exactly.

How much extra DefinedBenefit does a public sector worker get in relation to payroll? I would assume it's a graduated scale based on seniority, and people who split early lose out.

So while there's likely a strong incentive to stay in a public job for 20 years, there's likely a negative incentive to stay only five. Though the numbers there are, of course, nonsense -- the inflection point is going to differ for every single job.

Consider the opposite effect of the DC system: at best, the only extra financial strings an employer gets to put into their pensions is to make their 401k matching funds vest over a period of years.

Seems to me like navigating the real incentives on the table would be to work your first 25 years in the private sector cooking up higher wages and some small DC pension, and then your last 25 years in the public one at a lower wage but for a long enough time to score yourself real security for when you live to 110.

Given that I'm turning 40 in the Entertainment Software Industry, and playing the startup-lottery still hasn't made me a multimillionaire, such ideas keep looking better and better to me.

55% of payroll versus 10%. Quite a difference. In the case of the bank, they simply print more money to give themselves bigger pensions (confiscate more resources from the private sector). In the case of the private sector, they actually have to make more useful products to be able to give themselves bigger pensions. This pretty much explains the difference. Its much easier to take resources away from helpless fixed income private sector retirees (print) than it is to make additional useful products.

The option is particularly hard to value because governments can still default on their promises. Constitutions are not law of physics and can be changed. If a government made promises that turn out to be unrealistic in practice, it's almost certain they will at least partially default. Ultimately it's a bargain between the pensioners and the working age population at the time the pensions are paid, where the pensioners are not in a particularly strong position: the elderly generally make poor rioters, and the working population can simply leave any location where pensioners have excessive entitlements, which puts a hard limit on the bargaining power of pensioners.

Actually, Migs, I used my toes.

Thought experiment: A retired teacher passed away today. That teacher was hired in 1960 and retired in 1990. During the retiree's working lifetime, 10% of pay contributions were made to the pension fund based upon an 8% assumed rate of investment return. In the event, the fund actually averaged 6% investment returns over the 51-year period.

What should have been the contribution rate had the 6% investment returns been known in advance? About 15% of pay. And in such a theoretical case, the plan sponsor would have eventually put in more than 10% of pay while the person was working in an attempt to make good (somewhat, but not fully) on the short-of-expected investment returns.

Had the sponsor put in money on a risk-free rate of return (let's say 4%) basis, the initial contribution rate would have been 23% of pay. As time went on, the 23% of pay contributions would have proven to be too much, and presumably the employer would have reduced the contribution rate at some point.

The idea of discounting at a risk free rate more or less makes sense if a plan invests in risk free assets (although to the extent that no risk-free assets are available beyond a certain duration, there is nothing risk free about the required reinvestment). It also makes sense if you want to estimate the liability were it to be funded with risk free assets. But after the fact, if a fund supports a benefit with risky assets, and it gets better or worse returns than the risk free rate, it is ultimately the actual returns which determine the needed level of contributions.

Contributions + investment income = benefits paid + plan expenses

The issue with discount rate selection for liabilities backed with risky assets is how much risk premium to recognize up front. Recognizing no risk premium seems unduly dire for a decades-long venture. (If you are budgeting for your own retirement through a DC plan, are you really going to set that budget based on gilt yields?) There is always a temptation to recognize more risk premium today and let the next generation hold the bag if the expected returns do not materialize, but that temptation is not sufficient cause to recognize no risk premium whatsoever.

Stubborn reader gets last word... for now?

This is a silly debate. School teachers in my area work a total of 183 days last year. No matter how much I hear teachers work overtime, they do not work 73 days of it. Anyways it is no match for everyone in my office that is college educated and works 250 days per year and gets two weeks of vacation per year. Our 401k retirement plan gets destroyed every ten years by the market and no one will pay me when I retire. The teachers at our local Catholic School have no such benefits and give a fine education for less than $7,000 per year while we pay at least $9,000 per year per student in California public schools and that does not include the cost of the pension and health benefits for the retired teachers.

I love Buttonwood and agree with this column. The Republicans in America could vote on some things, however, that would be seen as heroic and solidify their position. The first that comes to mind is a "Fair or Flat" Tax. Why has that slipped out of sight?

In the American case the pensions may not be inflation indexed. As many are pointing out the next gutting of the middle class may well be through inflation.

Whether or not public sector employees get more compensation is not really the issue in Wisconsin. The issue is whether they should be allowed to bargain with the state to negotiate terms of wages and benefits that both can accept and will honor in the future. The most important point in the DiA post is this: "as Ezra Klein points out, the public-sector employees got rooked: they accepted lower pay in exchange for retirement benefits, and now the retirement benefits look unlikely to come through." If states are allowed to declare bankruptcy as many Republican legislators would like, then the employees will be totally screwed. Basically the employees agreed to let the state invest what would have been their additional wages, and now the state is deciding to keep the money for themselves.

"as Ezra Klein points out, the public-sector employees got rooked: they accepted lower pay in exchange for retirement benefits, and now the retirement benefits look unlikely to come through."

This is a load of BS. There is an interesting myth starting to grow about public sector pay -- that somehow it is automatically lower than private sector counterparts. This is a highly misleading statement.

Public sector pay (at least in the federal government), is intended to be comparable to private sector pay for similar jobs, and for the most part it is pretty close. In some respects, it is even more generous, especially when you factor in not only retirement benefits, but also the extremely generous health benefits, disability benefits, transportation subsidies, childcare subsidies, student-tuition assistance, and other federal fringe benefits that are made available to a great number of federal employees. When you couple this with the extraordinary job stability (your employer never goes belly up), your nearly ironclad job security (it requires moving mountains to fire a federal employee), and statutorily guarranteed pay raises every year, there are a host of reasons why a federal job is extremely lucrative in ways that don't immediately show up on a W-2.

I am a federal government attorney, and therefore know exactly of what I am speaking about. In fact, comparing lawyers in private sector jobs can be an illuminating exercise.

When I left the private sector to join the government after 9/11, I took a nearly 50% pay cut. However, I had been in a very lucrative, big-firm environment. Many of my classmates had left law school to hang up their own shingle, and were making a fraction of what I was making at the time. People like to think of all private sector lawyers as filthy rich, but in reality the income inequality within the private sector is quite huge -- if you are with a big firm or a big company, you can make a fortune, but if you are a solo practicioner or in a small firm that handles minor crime, slip-n-falls, or other lower level types of litigation, very often you are not much wealthier than your clients. There are a great many private sector lawyers scrabbling along at 40-50 K per year and paying for their own benefits who would kill for a public sector job that potentially pays double that (or more than double, when you consider benefits).

It is true that no one joins the government to get rich. However, it is also true that I have never met a government employee that was unable to save for retirement, put kids through college, live in a decent neighborhood, have a nice vacation, pay for doctor's visits, and enjoy at least a few frivolous pleasures from time to time. A federal job is a good gig, no matter how you slice it.

Interesting how the discussion manages to ignore the elephant in the room. "benefit is based on your three highest years of earnings (your "final average earnings")"

Now wherever I have been, my pension contribution depended on my then-current earnings. And my pension benefit (whether DB or DC) depended on how much had been contributed -- over the whole term of my employment. No special top up if my final position got paid a lot more than my earlier ones.

But public employees get to count only their last 3 (or 4 or 2, depending on the system) years' income. Not just salary, but also overtime, unused accumulated vacation, unused sick leave(!), etc. It's an easy system to game, and people do. As a result, we see people whose retiring pensions are twice as big as the salary of their last position, or even more. No wonder that public patience, especially from those making a fraction of those salaries, let alone those pensions, is wearing thin.

Nice post. Couple of quick points:

1. An oddity is that public employees rarely get social security - and if they do, from some job worked before, during or after public employment, some 15 states reduce benefits. Private pensions need only cover the overage above social security because you do pay into that system as a private worker. Take Illinois, which is the poster child for underfunding. The average teacher pension is $15k over social security. That then looks like a $43k pension - roughly - but that's really equivalent to a $15k or so private pension.

Question: did you consider this in your numbers? I didn't see any evidence of that. Rerun the numbers taking out the equivalent of social security and see where that comes out. It will reduce the public incentive.

2. Public employees almost always contribute to their pensions. They usually contribute more than private workers pay into social security. Examples: Illinois state worker average is 8.5% and teachers contribute 9.4%; MA teachers now contribute 11%.

3. A changeover to defined contribution isn't simple. The system now pays out a guaranteed benefit from funds in the trust, while a dc plan pays out account by account. You have to take all the money contributed into the fund by each employee and convert that into an account - which raises the question of earnings for each account - but the state still has to pay the benefits accrued under contracts (that the courts enforce). There is thus an immediate funding issue because you now have a bunch of accounts with money in them and a benefit pool that needs to be paid (the old debt). In private cases that results in a big contribution now that then with earnings assumptions can pay the old debt. This is what has happened in big bankruptcy cases and takeovers. This could be a huge number.

If public sector workers were overpaid,which is not the case, that would be a reason to strenghten private sector unions, not to destroy public sector ones.

Know whos really overpaaid: Economist writers*

*Ok, not really, even journalists at places like the economist are relativly low paid, at least compared to their city finance pimps.

In the US, yes, many public employees don't participate in Social Security. What proportion, I'm not sure. And it is not difficult to qualify for a reduced Social Security benefit outside of public sector service. There are a lot of factors to consider in comparing benefits and contribution levels, but the important point is that Social Security participation needs to be considered to get a full picture.

Public employee pension contributions are also important. However, these contributions are often "picked up" (i.e. paid for) by the employer. Moreover, in many cases, these contributions enjoy a guaranteed rate of return well above gilt yields.

Here is a link to the 12/31/09 actuarial valuation for Wisconsin Retirement System:
http://etf.wi.gov/boards/agenda_items_2010/etf20100624_items/ji-item4a.pdf

Page I-1 shows a total contribution rate of 11.7% for general participants. Police and firefighters have higher rates. It appears these rates (which includes "participant" contributions which are apparently mostly paid by employers) do not yet reflect a portion of 2008 investment losses.

Page I-4, third paragraph beginning "The statutory allocation..." is an interesting read.

Page I-6 shows key benefit provisions.

Re: LexHumana "This is a load of BS... "

No it's quite accurate. 'Less' pay of course meaning lessor than if they had chosen to forgo pensions for more pay in negotiations. All deals made for benefits are tradeoffs. They got these pensions in negotiations not by legal decree. If the public sector workers took DB pensions, then its fair to say that they were not asking for something else or not given something else in return when these agreements were reached.

Perhaps one could argue that governments discounted the cost of DB pensions when they gave them out. But that just suggests that EK is right, and that they were playing fast and loose with their financial obligations and were always willing to declare bankruptcy over raising taxes or making the right contributions.

Incidentally, lets not pretend that DB pensions don't exist in the private sector, they're hardly only a feature of the public sphere. (Though a lot of people seem to think this, bolstering their belief that these benefits are some extra legal special status as opposed to negotiated compensation....) Heck I got one, as I know there's little value in a pension that offers no actual security.

But then taking the modern examples into account; maybe I just got scammed too.

@ Kevin Sutton,
"They got these pensions in negotiations not by legal decree."

No, you are confusing private sector unions with public sector unions. In the private sector union scenario, all wages and benefits are subject to negotiation, and the workers get what they bargain for and no more.

In contrast, public sector wages and benefits are not set by negotiation -- they are set by statute. Every unionized federal employee has their wages or salary set according to law, and all federal benefits are also set according to law. If you are curious, you can look up all the wage/salary tables and benefits descriptions at the Office of Personnel Management (OPM). www.opm.gov