Not applicable for Long Term Disability assumptions

Election probabilities are conditional probabilities utilized in situations where an additional condition, besides eligibility and decrement caused by the contingency initiating the benefit, must be met at the time of decrement to receive a particular benefit. That is, receipt of the benefit is not certain, even after a participant becomes eligible and decrements from the appropriate cause.  

Lapse probabilities apply after decrement to relect that the benefit may be forfeited each year after decrement.

The Election & Lapse Probabilities topic lists all Benefit Definitions that exist in the current Project (or were previously defined for a Valuation referencing either this Valuation Assumption set or a set from which it was copied by “saving as new”) and with which election or lapse probabilities are to be associated, i.e., those Benefit Definitions for which the “Election probabilities apply” or "Lapse probabilities apply" box has been checked (in Input | Plan Benefit Definitions | Benefit Definitions). The probability, if any, that has already been associated with the benefit, whether a constant value, a table or “none” (discussed below), is displayed next to the name of the Benefit Definition. To specify Election or Lapse Probabilities, select the desired benefits from the pertinent section, click the Edit button, and complete the parameters in the resulting dialog box. 

If you wish, you may select a Plan Filter, from the dropdown list, to reduce the list of possible choices to those Benefit Definitions associated with a particular Plan Definition. Please see Filters for details.

There are three options for defining the associated Probabilities:

1. Select the Constant option to apply the same probability of benefit receipt for all active records meeting the eligibility requirements and decrementing from the appropriate cause. Instead of a single constant value, you can specify that election probabilities Also vary by the contents of a coded database field or by calendar year of decrement (the relevant parameters are discussed below). Enter each probability as a number between zero and 1 (not as a percentage).

2. Select the Table option to specify probabilities defined by tables that vary by age, service and/or sex. Select from the list of Election Probability Reference Tables or Lapse Probability Reference Tables that have been saved and unhidden in the current Project or click the  button (to the right of the Probabilities column) to create a new table. Instead of a single table, you can specify that election probabilities Also vary by the contents of a coded database field or by calendar year of decrement (the relevant parameters are discussed below).

3. Select None (i.e., probability of receipt is 1) to assign an election probability of 1 or None (i.e., probability of receipt is 0) to assign a lapse probability of 0, that is, all persons elect the benefit or no participants lapse the benefit. This option is useful when a benefit has been defined to accommodate election and/or lapse probabilities, but no election and/or lapse probabilities are to be reflected in this particular assumption set. Click “OK” to return to the Election & Lapse Probabilities library dialog box, where “None” now appears next to the name of the Benefit Definition.

To indicate that the election probabilities Also vary by the contents of a coded field on the database, that is, different constant values or tables should be used for different groups of records, identified by values of a coded database field (e.g., Division), check the Coded database field box and then select the field from the list of all coded fields unhidden in the current Project. A Code column now appears in the grid above. For each code, enter the numeric value (if the Constant option is selected above for Probabilities) or select the desired election probability table library entry (if the Table option is selected above for Probabilities).

To indicate that the election probabilities Also vary by calendar year of decrement (i.e., the calendar year of the anniversary of the valuation date in the year of decrement), check the Calendar years box and enter (in the text box below) for all constant or tabular election probabilities that you will specify, the first calendar year to be associated with a particular constant value or table. For example, if the election probabilities will be 0.25 through 2008, 0.2 for 2009, 0.15 for 2010, 0.1 for 2011, 0.05 from 2012 through 2014 inclusive and 0.25 in 2015 and later years, then enter “2009 - 2012, 2015” in the text box. Two columns, From and To, now appear in the grid above, and rows appear according to the “breakdown by year” indicated by the text box entries. In our example, rows appear for years through 2008, for 2009, for 2010, for 2011, for years 2012 through 2014 and for years 2015 and later. For each set of years, enter the numeric value (if the Constant option is selected above for Probabilities) or select the desired election probability table library entry (if the Table option is selected above for Probabilities).

To indicate that the election probabilities Also vary by both a coded field on the database and the calendar year of decrement, combine the instructions in the two paragraphs above; the grid above will populate with both the Database Code column and the From/To columns. Specify the desired constant or table probability for each combination of code value and decrement year.

Because a Benefit Definition in OPEB mode can represent both benefits payable to the member and benefits payable to the spouse, further discussion about using election probabilities to code participation assumptions is in order. For member benefits, input the election probability that the member will elect coverage given that he/she decrements and is eligible for coverage at the time of decrement. For spouse benefits, input the probability that the spouse will have the right to coverage at the time the member decrements and that coverage will be so elected for the spouse. For example, consider a retiree medical plan that makes medical coverage available to the spouse only if the retiree elects coverage for himself (or herself). If the probability that the member elects coverage is 80% and, of the married members so electing, 60% of their spouses also elect coverage, then 0.8 should be entered as the election probability for the member Benefit Definition and 0.48 (0.8 multiplied by 0.6) as the election probability for the spouse Benefit Definition. Note that the probability that the member is married is not a factor in calculation of the election probability value; because, typically, the percentage of members that are married is entered for the Fraction of population that is married parameter under the Other Valuation Parameters topic.

When a dual Benefit Definition (containing both a Member Payment Form and a Spouse Payment Form) is used, it is presumed that coverage has been elected for the spouse whenever coverage has been elected for the member. Therefore, if not all spouses of members with medical coverage will also have medical coverage, as in our example in the preceding paragraph (where only 60%, not 100%, of the spouses with a right to elect coverage actually did so), separate Benefit Definitions for member and spouse benefits must be defined. Similarly, when a dual Benefit Definition (containing both a Member Payment Form and a Spouse Payment Form) is used, it is presumed that the same lapse assumptions apply to both the member benefit and the spouse benefit, (for a table, looked up using the age and sex of the member or spouse respectively). If the assumptions differ, separate Benefit Definitions for member and spouse benefits must be defined.

For election probabilities, the Table service based on parameter is used to define service for look up if the table has a service dimension. The preset option, “<rounded attained age – rounded hire age>” indicates that service will be computed as the difference between rounded attained age (as of the decrement date) and rounded hire age. (See also the discussion of the date of hire (or hire age) parameter under the Active Data topic of the Census Specifications command.) Alternatively, Table service may be based on a database Field; select from the list of database fields in the current Project. You may refer to a date field from which to measure service or to a numeric field containing service as of the valuation date. Note that selecting, by name, the date of hire field of the Active Data topic of the Census Specifications might not produce the same result as the preset option. This is a consequence of the fact that if a database field is selected, truncated service based on the field is used. That is, if you select the date of hire field by name, ProVal subtracts the hire date from the decrement date and then truncates. The result may be lookup service of one year more or less than lookup service computed as the difference between rounded ages at the valuation date and at hire. For more information about table lookups, see the discussion of table interface. To use fractional service attribution (e.g., hours-related service) or rounding (e.g., completed years), select from the library of Service Definitions. The  button accesses the library to create and modify Service Definitions. (See also the discussion of Service Definitions and attribution in the Technical Reference article entitled "PUC and UC attribution".)

For lapse probabilities that are defined by a table, the Table lookup for spouse benefit based on parameter indicates whether the table lookup for a spouse benefit should be based on the member age and sex or the spouse age and sex.

Click the Clear button to remove the selected benefits' election or lapse probabilities from the Valuation Assumption set you are editing. These probabilities will no longer appear when you View this set of Valuation Assumptions (although you might still have associated election probabilities with these benefits in another set of Valuation Assumptions).