[Spyro7]

> "means that they're the only govt institution that isn't a timebomb of future obligations"

This was the spirit in which the requirement was enacted. I completely agree with the idea. My problem with this line of reasoning is this:

Why fund the estimated lifetime costs of health care that will not be provided for maybe 40 or so years into the future? Why not fund a moving window of ten or twenty years of obligations instead?

There is a huge opportunity cost associated with stashing such a large amount of money away for this purpose rather than using it for current operating expenses.

> "Yes, current projections say that they're overpaid, but these projections have a way of going horribly wrong."

Indeed, projecting costs for the next half-century is a dangerous thing to do. This is why I question the rationality of handicapping the USPS with the responsibility for funding far-future liabilities based on those projections.

Even if you ignore the opportunity costs of this fund, the inflexibility of the rule as well as the fact that it puts so much faith in projections makes it a highly suspect decision.

[anamax]

> This is why I question the rationality of handicapping the USPS with the responsibility for funding far-future liabilities based on those projections.

You're ignoring the fact that those liabilities are being incurred today. If you can't pay them with current revenues, how are you going to pay them with future revenues that also have to cover the liabilities that you're incurring when you're receiving said future revenues.

I'm sympathetic to the idea that there's no way to properly estimate the NPV of this sort of open-ended liability, but that's an argument for not incurring such a liability. It's not an argument for paying less than your best guess of said NPV.

> Even if you ignore the opportunity costs of this fund,

There is no "opportunity cost" in not paying NPV now.

Suppose that your landlord offered to let you pay each month's rent starting 1 year from now and over the 10 succeeding years. How much should you set aside? The only answer that keeps you solvent is to set aside the NPV of the payment stream for each month's rent payments each and every month.

Do the arithmetic. After each month, you've got a new liability that you're going to owe after you stop receiving any corresponding benefit. At the end of year N+1, you're making payments on N years worth of rent (capped at 10) and those payments will continue for 10 years after you stop renting. (Yes, they'll decrease over time, but since you won't be getting the benefit of whatever you were renting, it's unclear why you're happy paying for it with "new money" that you probably need for replacement digs.)

[anamax]

> Why fund the estimated lifetime costs of health care that will not be provided for maybe 40 or so years into the future? Why not fund a moving window of ten or twenty years of obligations instead?

Because it has incurred those obligations and nothing that happens in the next 10-20 years will make those obligations go away. However, 10-20 years from now, it won't have today's revenue to cover those obligations. In addition, ten years from now it will be incurring additional obligations.

Those obligations have a net present value. If you don't fund them now, you have to fund the remainder later out of future revenues while you're also trying to fund at least part of the obligations that you're incurring then.

Let's assume steady state. If you always fund the net present value of 10-20 years of the obligatations that you incurred in the past and present, you'll eventually end up funding the equivalent of the obligations that you're incurring, but you're behind by the amount that you didn't fund during the ramp. You have to make that up in addition to the equivalent of just funding NPV of what you're incurring.

Do the arithmetic over time - funding the total net present value of future obligations is only sustainable way to handle said obligations.

A pension promise is not like a mortgage - it's just a debt with no collateral; there's no equity to sell if you can't make the payments.

[michaeldhopkins]

The General's office's report said that the projections would have been accurate but for the USPS earning a higher than expected amount of interest on their assets. To recalibrate based on this windfall seems foolish to me.

If you take the view as I do that an institution ought to make choices that will keep it around for hundreds of years, spending money conservatively enough to afford promises made to one's employees becomes a wise choice instead of a suspect decision.

That said, I do agree with the General that the USPS is unnecessarily entangled with the US government, and would be able to adequately fund its defined-benefit obligations without legislation (as private companies do), and also regulate overpayments and underpayments better without relying on the GAO or Congress.