STUMP » Articles » Public Pensions Followup: The Effect of Assumptions and Materiality » 30 September 2014, 14:39

Where Stu & MP spout off about everything.

Public Pensions Followup: The Effect of Assumptions and Materiality  


30 September 2014, 14:39

You may have noticed that I reposted a prior post, with a few fixes in there…but there’s more to the story.

I talked with Jason Franken of Foster & Foster last week, with regards to that post, and he gave me information about the particular comment that referred to the disability assumption.

Here’s the deal: the disability assumption that Tia Goss Sawhney complained about? It was in place of a no disability assumption.

You can go to the Comprehensive Annual Financial Report in question: and look on page 89, there is a table of the historical reasons for changes in the unfunded actuarial liability of the pension in question.

Here is the table I’m referring to:

I circled the particular component: in 2012, the unfunded liability increased by $7M due to the addition of this disability assumption. It did not decrease the liability value, which is of concern.

But let us dig a little more, using numbers from the CAFR: this is a $7M increase in a liability value when the total liability value was $2.2 billion at the end of 2013. Even if we suppose the liability were only $2B at the end of 2012, the effect of the change of disability assumption increased the liability amount by less than half of a percent.

That’s not material.

Yes, I know that $7 million sounds like a lot of money to a regular person, but when the actual amount in question dwarfs it by several orders of magnitude, you can look at it as amounting to a rounding error.

In any case, in discussing the situation with Jason, he mentioned that in doing an experience study for the pensions, the actual disability rate was lower than the assumption set he used, as old as it was. I looked up the table, and I believe the experience used to build it is about 100 years old… I think it’s unsurprising that they had higher disability rates back then.

But as Jason noted to me, unlike mortality, which has a very strong trend of improvement this past century and more, one cannot assume the same for disability, as the nature of the disabilities can change due to various reasons. Some disabilities recognized now for benefits would not have been allowed 100 years ago, and some disabilities from back then probably wouldn’t occur now (or would be able to be fixed). The point is you do need to check your assumptions against experience periodically to make sure you’re not getting off track (much more on that later)

But let’s ignore that experience, and let’s ignore the whole size of the unfunded liability — let’s compare that $7M against all causes for the increase in the unfunded liability.

For the time period in question, 2004 – 2013, the unfunded portion of the liability increased by $635M. The $7M change due to an added disability assumption amounts to 1.1% of the total increase.

On the other hand, that the governmental entity in question did not make the recommended contributions made up 45% of the cause of change. And that the assets underperformed the assumed discount rate was 56% of the problem.

Yes, I know that this adds up to more than 100%, but that’s because changes in the salary scale actually reduced the liability…. again, at a much higher percentage (about 12.5%) than the effect of changing the disability assumption.

My point is that one needs to have perspective, whether it’s a change in assumptions or a change in benefits — some of the changes being made are extremely small, and we need to criticize accordingly.

I will be looking at mortality assumptions later, but even in this particular case one can see that the effect of the mortality assumption has been much less than that of actual underfunding and a too-high asset return assumption.

And in both of those cases, the actuary can only recommend — generally for public plans, the actuary does not get to set the discount rate nor the amount actually contributed.

Look to the pension plan trustees, and in some cases state law, for those shortcomings.

Related Posts
Friday Trumpery: No Exemptions for State-Run "Private" Pensions [UPDATED]
Kentucky Meltdown: Teachers in Revolt, Bill in the Sewers, and Nothing Actually Solved
Friday Trumpery: DOL pick, take two