In various places in ProVal’s output, in both the OPEB and pension modes, you may encounter the term “head count”. Two such places are the “Projected Headcount and Benefits” report, available when you view the output from the Valuation command under the Execute menu, and the Projected Benefits tab of Valuation Output (available on the Output pane).

The term “head count”, as used by ProVal, is shorthand for one of the following definitions, depending on where it is used:

1. The number of participant families or households that are either receiving benefits or have the possibility of receiving benefits in the future. This definition is used for:

• active records in the pension and OPEB modes and

• pension mode inactive records (in a Core Projection, only initial inactive records).

2. The number of benefits payable, or potentially payable in the future, to anyone in the participant’s family or household (generally, the member or spouse). This definition is used in the pension modes for emerging inactives.

3. In OPEB mode, the number of participants (member or spouse) who are receiving now, or who may potentially receive in the future, one or more benefits from the plan. This definition is used for:

• emerging inactives of a Core Projection (that is, the portion of members who are active initially but become inactive during the period of the forecast) and

• inactive participants (in a Core Projection, initial inactive records).

In the pension modes, an active member who decrements switches from Definition 1 to Definition 2; thus a member who contributes only one family to the head count while active may contribute several benefits to the head count when projected to be an emerging inactive. Head counts for multiple benefits are combined by adding them together. Therefore, ProVal’s head count will not be accurate, for example, if both a retiree pension benefit and postretirement life insurance coverage commence at the retirement date. The best way to obtain a more accurate count of members and spouses who have a liability associated with them is to run a separate Valuation wherein the “duplicate” benefits are eliminated, leaving one benefit per participant. (For example, in U.S. qualified mode, eliminate the “REA”, post-decrement death, benefit, leaving just the deferred life annuity benefit.)

In OPEB mode, an active member who decrements switches from Definition 1 to Definition 3.

Participants, families or benefits stop contributing to the head count when there is no longer a liability associated with them.

 

Head Count Details Available for View

Head counts can be viewed separately for active families (i.e., the count is associated with members who are still active, that is, still working) and inactive families or benefits (i.e., the count is associated with members who are former employees). In pension mode Valuations, ProVal allows you to see the head counts at each future decrement date for actives, initial inactives (those members who are inactive on the valuation date) and emerging inactives (the portion of members who are active on the valuation date but assumed to become inactive before each decrement date). In OPEB mode Valuations, ProVal also allows you to see a further split of head counts by benefit and whether the benefit is payable pre-Medicare or post-Medicare.

 

Post-decrement probabilities

For emerging inactives, head counts for multiple benefits are combined by adding them together after adjusting by any post-decrement probabilities. Thus, when combining benefits, ProVal implicitly assumes that benefits with post-decrement probabilities are mutually exclusive. For example, if there are two benefits, a life annuity and a lump sum, each with a 50% post-decrement probability, ProVal assumes that half of the participants get one benefit and the other half get the other benefit and thus simply adds the head counts together, resulting, in this example, in an accurate total of 100% (i.e., half the members elect the annuity and half the lump sum). If, however, the benefits are not mutually exclusive, and each benefit is elected (or declined) independently, the head counts displayed will not be accurate. For example, if a participant has a 90% probability of electing the medical benefit and, separately, a 50% chance of electing the dental benefit, the post-decrement probabilities will be coded in the same manner but the total head count will be 1.4, not 1.0, thus “double-counting” participants who elect both medical and dental but not reflecting the fact that 5% of participants (50% of the 10% not electing medical), on average, will elect neither benefit and should therefore not contribute to the head counts.

 

Impact on Deterministic and Stochastic Forecast Results

In the U.S. Qualified mode, the PBGC flat rate premium and, under PPA law, the loading of the at-risk liability and at-risk normal cost depend on the inactive head count. Please note that these calculations may be overstated.