Study of Long Term Care Applications Indicates Marriage Discounts Strong Proxy for Gender Ratings
Long term care insurance carriers have recently announced that they intend to implement gender based premium rates. Examination of current applications against national population data indicates that wealth and age qualified single women are already underrepresented in the application pool. It also shows that marital discounts are an excellent proxy for gender based rates and that implementing them will likely have minimal impact on carrier results.
Application data was collected from private brokerage for 2012 and 2013. Applications from single women were underrepresented compared to the qualified national population, especially at ages greater than 65. This was true of males as well, that represented only 4% of applications and 10% of the wealth and age qualified population. Based upon these results, it appears that the marital discounts represent an excellent proxy for gender based rating and that, while further examination of the impact of policy placement and longevity might provide additional insights, instituting the practice is likely to provide minimal advantages to long term care carriers
The data contained 10,089 records of persons that applied for long term care insurance with a representative set of carriers through private brokerage from the beginning or 2012 until the last day of February 2013. See ACSIA.com for a typical agency. The fields included an anonymous 6 digit identification key, Date of Birth, Date of Submission, Marital Status, and Gender. Each record was grouped into 5 year age ranges in 2 ways:
- 1. Nearest age modulo 5
- 2. Specified 5 year groupings from age 50 to age 74 and those younger and older than 50 and 74
There were 212 incomplete records, missing date of birth, gender or marital status. These were removed. That data was not pulled randomly, but is the complete population for the dates pulled and these agencies and should be fairly representative.
Table 1 shows the distribution of submissions by marital status, gender and age at issue to the nearest age modulo 5. Ages under 33 and over 82 were removed.
Table 1: Gender and Marital Status by Age at Submission - Submission Dates 1/1/2012 through 2/28/2013
Table 2 shows the distribution of submissions by marital status and gender broken up into age ranges of 5 years.
Table 2: Gender and Marital Status for Age ranges at Submission - Submission Dates 1/1/2012 through 2/28/2013
Results and Analysis
The average age of all applicants is 62. For married females it is 61 and 2 months. For married males it is 62 and 10 months. Singles are roughly equal at 62 and 4 months for females and 62 and 6 months for males. Over 85% of the applicants are married. Under 10% are single females. About 4% are single males. This is compared to 20% of the female population ages 55 to 74 that is single and 13% of the male population between 55 to 74 that is single. (See http://www.census.gov/population/www/socdemo/hh-fam/cps2011.html)
The industry's discount program for couples has had a strong impact on the distribution of sales. Over 85% of submitted applicants are married, while the overall population has 36% unmarried between 65-74 and 33% unmarried applicants between 55-64, based upon 2011 census data. The results are not materially different for higher incomes overall, but show a much greater percentage of older males married, 84% and 81% in the 55-64 and 65-74 age groups earning over $100,000 compared to 59% and 38% married for similar females. It is remarkable that with 51% of females ages 65 to 74 with incomes over $40,000 single, only 20% of the same aged females applying for Long Term Care are single. (See http://www.census.gov/population/www/socdemo/hh-fam/cps2011.html )
The domestic partner category includes a wide variety of living arrangements and would be recorded only when the cohabitants are applying for a discount, so this number would likely understate the national population in these arrangements. The total is less than 1% and not material, however.
Gender Based Rates
Because of the gender distribution of the married population at older ages, the marital discounts provided by long term care insurance carriers have been excellent proxies for gender based rates. In the first place, the total population of age and wealth qualified single males is a small section of the qualified population. In the second place, single female applicants are already below 10% of applicants. Paying a premium may both increase premium and/or reduce single female participation, but because the participation is already small, the impact will be minor overall.
The percentage of the population of single males between 55 and 74 is 13% of the total population in that age range. It is 10% if the population is further restricted to single males with greater than $40,000 in annual income. Reduced rates may bring in more single males, but the target group is small. Marital discounts have had a much greater impact on male participation than gender based rates.
There are other possible impacts beyond the application pool that might provide a benefit to carriers with gender based rates. In cases where one spouse is declined, the premium discount may cause more males to retain the policy and more females to reject the policy and decrease the overall carrier costs. However, these populations are again less than 10% of the total application pool and are likely to have minor total impact. Another area with some improvement for carriers, is that when a spouse dies, the remaining spouse, if female, will carry a higher premium. It is unclear how much impact this will have, without some modeling, but the impact is again likely to be modest.
Recent sales of long term care insurance are highly biased toward married couples and singles are greatly underrepresented compared to the overall population. This is especially true of single women in the 65-74 age group. The carriers' current marital discounts have been highly effective as a proxy for gender based pricing. An additional layer of gender based pricing is likely to provide only modest impact on carrier results, though additional work is needed to model the impacts due to examining the placed policies compared to the applicants and the further impacts due to the increased longevity of females in premium payments.