David Leonhardt on the state pension shortfall

This is exactly right:

But 8 percent [equity returns] still seems like an aggressive assumption for state and local governments to be making.

If state and local governments instead assumed a future return of 7 percent, their funding gap would nearly double, to $1.3 trillion, according to Alicia Munnell and her colleagues at the Boston College retirement center. If they assumed a 6 percent return, the funding gap grows to $1.8 trillion.

Even if you believe – as I do – that government workers are not grossly overpaid, you can see that states and local governments have not set aside nearly enough money for their employees’ retirement.

Here are related comments from Ezra Klein, though I see less agreement than he does.  Here is Baker, writing on social security privatization (see point seven), claiming we can expect rates of return on the stock market of four to five percent, not the seven to eight percent in the municipal context.

Comments

Forward-looking returns have implications for everyone saving for retirement, not just public employees with pensions. If stock returns over the next 30 years are only 4-5%, then we as a society will have bigger problems with retirement than under-funded public pension schemes.

Most pension funds have an allocation to fixed-income, typically around 40%. Luckily, bond rates are high right now...oops.

Note that QE policies which artificially inflate asset prices axiomatically lower the forward return on those assets, (certeris parebus on cash flows - grant me that the main effect of QE is to lower the discount rate). The nominal returns on fixed income assets are presently at all time lows (across the risk spectrum - the high yield market is exhibiting normal spreads, but all-time lows in absolute yields).

One consequence of QE is thus to move pension funds and others into higher risk assets. Is this an unintended consequence? Or wasn't that the point?

Leonhardt: "the valuing of government pensions go a long way toward determining whether government workers appear to be overpaid."

Of course government workers, with a union, should be overpaid to those in the private sector, who don't have a union by and large. That's what unions are for. Are unions supposed to fight so that their workers are underpaid?

If you don't want government workers to appear overpaid compared to the private sector, bring back unions in the private sector. Problem solved.

The fundamental problem is a simple one. You can't promise a future performance.

I think that capitalizing future risk premiums at some mismatched discount rate amounts to a gross error. The future returns and cash flows from stocks have to be discounted at the risk-adjusted discount rate. The result is (identically) the value of assets today. So the task of assessing the funding level is simple: Discount the expected future cash outflows to participants at the appropriate discount rate, which in this case is the risk free rate, and compare the result to asset values today. Anything else is something between a misunderstanding and fraud.

The US has the pension accounting completely wrong. The UK is better.

More facts at http://papers.nber.org/papers/w16453

Why not let the market take care of the valuation? Prohibit the state from assuming any liability for future pensions. The state can buy retirement annuities for the workers from a private firm. The workers can negotiate with the state over the size of the annuity and its safety by requiring that the issuing firm commit to a specific investment strategy. The market can evaluate the cost of the pension, and the state can recognize the cost immediately. There would be a dramatic increase in transparency. This may have adverse tax effects for the workers - perhaps they would need to recognize the income immediately as well. However, it's not clear that this is a bad thing, and even if it is we can amend the tax laws.

Please watch the details before you try to play "gotcha." Baker's comments relate to the real return on stocks, not the nominal return. The paragraph containing his estimate describes a methodology that is exactly the same as in the new piece you referenced yesterday.

As for pension finance, the most important issue is whether future costs are sustainable -- the measurement of a single "liability" is only one part of this broader issue. In fact, the type of liability typically used by pension actuaries in calculating funding rates includes a much larger portion of projected future benefit cash flows than the type of liability used to measure currently accrued obligations for a plan wind-up type of calculation. For one large state fund, I recently calculated that their traditional ("funding target liability" using a 7.75% discount rate) was only about 5% lower than their liability for currently accrued benefits using a discount rate of 6% (which is a rate that would typically be used right now for a corporate pension plan calculation). I am a pension actuary, and I confess that our profession has not made it particularly easy to understand all this -- but the level of mis-understanding is getting to be of mind-blowing proportions.

Rahul,

The problem of saving for retirement is not unlike the old story problem of a driver who needs to be 30 miles awat in 30 minutes (you can see he needs to do 60 mph at the beginning of the trip). However, the next question is, how fast does he need to go if he begins the trip by driving the first 10 miles at 30 mph? the first 15 miles at 30 mph?

The longer you spend undersaving for a large goal (and a retirement annuity is a very large goal) the more quickly you'll find that your catch up payments become unbearably large. State pensions are finding that they were saving 1/2-2/3s as much as was really necessary for 2/3-3/4 of the lumpy pensioners (boomers) careers.

Not to mention that all the pensions/401(k) plans etc are going to have a tremendously difficult time liquidating their investments to meet projected cash flow needs (hence all the interest in purchasing infastructure).

Rowland, nowhere does he indicate he's talking about real returns. He says "total returns" are equal to dividends plus GDP growth, which he does not define as real. He switches to a real number in the final sentence to exaggerate his point. But I was actually responding to commenters who think the actuarial rate of return assumption for the portfolio should be 4-5%. I could take lots of issues with his argument in point 7, though.

You are implying that Baker is inconsistent, but you are comparing his nominal return assumption now with his real return assumption in the linked article. If he assumes 3% inflation rate in the near term and 3.8% ultimately, the numbers are fully consistent. Those inflation rate assumptions are high - 1% to 1.5% above most economists' projections - but the two writings are also 13 years apart. I don't think it's fair to fault someone for shifting his views on return rates by 1% over 13 years.

That said, I don't think liabilities should be valued using the expected rate of return on assets, but that is an entirely different issue.

Has anyone talked about a lower assumption and issuing refunds as a bonus for high performance in a given year? (i.e. assume 5%, when 6% happens, you give out money to the people paying in.)

People arguing against that would have to expose their preference for unfunded contributions.

I understand that paying out in good years ruins the averages in the bad, but thats implementation details.

The various studies allegedly demonstrating that public workers are not overpaid are poorly constructed. They are not adjusted for lifetime hours worked or the present value of pensions on the retirement date. These studies use the annual employer contribution to value the pension benefit. This contribution is far below what the pensioner actually gets at retirement; even accounting for market returns of those contributions.

As an example, in my hometown in California our current assistant city manager is about to retire with a $140,000 per year pension with annual COLAs. The present value of his pension is over 2 Million dollars. This far exceeds the contributions (including market appreciation of those contributions) used to fund that pension. Further, his lifetime hours worked is far below an equivalent private sector manager.

My example is fairly typical of what is going on, at least in California. Go to the State Controllers website and check out the salaries of various city employees. You will be astounded. Then remember these guys are getting 70-90% of that at retirement. And many of these same guys have lifetime retiree medical benefits (which are usually completely unfunded) as well as separate "side pocket" deferred comp plans.

In California the average public employee, accounting for the above, is paid AT LEAST double equivalent private sector workers with much greater job security. The reporting on this has been extremely weak. I am not sure why.

Please watch the details before you try to play "gotcha." Baker's comments relate to the real return on stocks, not the nominal return. The paragraph containing his estimate describes a methodology that is exactly the same as in the new piece you referenced yesterday.

As for pension finance, the most important issue is whether future costs are sustainable -- the measurement of a single "liability" is only one part of this broader issue. In fact, the type of liability typically used by pension actuaries in calculating funding rates includes a much larger portion of projected future benefit cash flows than the type of liability used to measure currently accrued obligations for a plan wind-up type of calculation. For one large state fund, I recently calculated that their traditional ("funding target liability" using a 7.75% discount rate) was only about 5% lower than their liability for currently accrued benefits using a discount rate of 6% (which is a rate that would typically be used right now for a corporate pension plan calculation). I am a pension actuary, and I confess that our profession has not made it particularly easy to understand all this -- but the level of mis-understanding is getting to be of mind-blowing proportions.

When you say "the governments and municipalities have not set aside enough money"...you mean they haven't taken enough from the private citizens..correct?

@WAyne: Well, since the private citizens elected the government to act on their behalf, you could say that the citizens didn't put enough money aside to pay for the obligations their agents made on their behalf or didn't exercise enough diligence to ensure that their representatives acted in their interests. Democracy give you the government you deserve, not the one you want.

Comments for this post are closed