2017 Milliman LTCI Survey

Claude Thau & Allen Schmitz & Chris Giese
July 2017 Issue

The 2017 Milliman Long Term Care Insurance Survey is the 19th consecutive annual review of long term care insurance (LTCI) published by Broker World magazine. It analyzes the marketplace, reports sales distributions, and details available products. 

The data includes certificates or individual policies sold to multi-life groups (primarily small groups) with discounts and/or underwriting concessions, but excludes group policies aimed only at the large group market. 

Analysis of worksite sales will appear in the August issue of Broker World magazine.

Unless otherwise indicated, references are solely to U.S. stand-alone LTCI sales, excluding exercised future purchase options or other changes to existing coverage. “Stand-alone” refers to LTCI policies that do not include death benefits (other than returning premiums upon death or waiving a surviving spouse’s premiums), annuity, or disability income benefits. 


Highlights from This Year’s Survey

Thirteen carriers participated broadly in this survey. Four others provided sales information so we could report accurate aggregate industry individual and multi-life sales.

Although not displaying products, Northwestern LTC provided background statistical information. Auto-Owners, John Hancock, MedAmerica, and United Security contributed to the sales total but did not provide other statistical information.

Sales Summary

Participants’ individual claims rose 6.9 percent and group claims rose 4.7 percent. Overall, the stand-alone LTCI industry incurred $9.7 billion in claims in 2015 based on companies’ statutory annual filings, raising total incurred claims from 1991 through 2015 to $107.8 billion. (Note: 2015 was the most recent year available when this article was written.) Most of these claims were incurred by insurers that no longer sell LTCI. This compares with $8.7 billion of incurred claims in 2014, roughly an 11 percent increase.

The average time from receipt of the application until a policy is issued dropped from 44 days to 38 days, the fastest time since 2012.


About the Survey

This article is arranged in the following sections:

Market Perspective (more detail in subsequent parts of the article)

We asked additional questions this year relative to the worksite market. Insurers that do not sell in the worksite market often did not answer these questions. Distributors might answer some of these questions differently from insurers.

We asked some questions about “both-buy” discounts (percentage discounts when, generally, a husband and wife or two life partners both purchase coverage). We stated the premise that “With unisex pricing, single person pricing generally reflected an expectation that most buyers would be female. Couples’ both-buy discounts were on the order of 30 percent to 40 percent, partly because couples’ business was very close to 50/50 in gender distribution. That is, gender distribution contributed to the large both-buy discounts.”

The potential impact of regulation continues to be unclear. Last year, we wrote that, “The detrimental impact of the Affordable Care Act on worksite LTCI sales should wane, as the ACA is likely to demand less attention from employers and employee benefit brokers going forward. The new Department of Labor fiduciary rules do not appear to impact LTCI directly, but increased need for detailed documentation may leave financial advisors less time for ancillary services such as LTCI sales. Furthermore, some advisors may become more reluctant to discuss LTCI with their clients, concerned that they lack the expertise to meet fiduciary standards.” As this 2017 article is being drafted, uncertainty relative to the Patient Protection and Affordable Care Act (ACA) and fiduciary rules continues to impact the LTCI market, with no clear resolution in sight.

Only five participants offer coverage in all U.S. jurisdictions and only one worksite insurer does so. When a jurisdiction is slow to approve a new product, restricts rate increases, or has unfavorable legislation or regulations, it contributes to insurers’ reluctance to sell in that jurisdiction.


LTCI claims paid by insurers no longer selling LTCI might differ significantly from data reported below because their claimants might be more likely to have facility-only coverage, be older, have smaller policies, etc. 

Table 1 shows claim distribution based on dollars of payments, whereas Table 2 shows distribution based on number of claims. In the distribution based on number of claims, if someone received care in more than one venue, that person is counted more than once. Claims will shift away from nursing homes because of preference for home care and ALFs and because newer sales are overwhelmingly “comprehensive” policies (covering home care and adult day care, as well as facilities), whereas many older policies covered only nursing homes. Claims that could not be categorized as to venue were ignored in determining the distribution by venue type. 

Table 3 shows average size individual and group claims since inception. Because claimants submit claims from more than one type of venue, the average total claim should generally be larger than the average claim paid relative to a particular venue. Nonetheless, ALFs consistently show high average size individual claims, probably because:

The following factors distort our average claim sizes:

  1. Roughly 15 percent of the inception-to-date individual claims are still open. Our data does not include reserve estimates for future payments on open claims.
  2. People who recover, then claim again, are counted as though they are multiple insureds. We are not able to add their various claims together.

In-force claim data understates the value of current sales because:

  1. The many small claims drive down the average claim. The purpose of insurance is to protect against a non-average result, so the amount of protection, as well as average claim, is important.
  2. Older policies had lower average maximum benefits and were sold to older issue ages compared with current sales, resulting in smaller claims for shorter periods of time than might result from today’s sales. 

The average group claim is smaller than the average individual claim, probably because of shorter benefit periods, lower maximum daily benefits, fewer benefit increase features, and more common reduced benefits for home care. 

Statistical Analysis
Twelve insurers contributed significant background data, but some were unable to contribute data in some areas. Four other insurers (Auto-Owners, John Hancock, MedAmerica, and United Security) contributed their number of policies sold and new annualized premium.

Sales characteristics vary significantly among insurers. Hence, year-to-year variations may reflect a change in participants or changes in market share, as well as industry trends.

Market Share
Table 4 lists the top 10 carriers in 2016 new premium. Northwestern and Mutual of Omaha repeated as the top two carriers, with Mutual of Omaha cutting Northwestern’s lead by 40 percent compared with 2015. The two top carriers produced 45 percent of annualized first year premium in 2016. They are followed by four insurers with 7.0 percent to 8.5 percent market share each, and then three insurers with 5.0 percent to 5.5 percent market share each. Four of the top 10 had higher sales in 2016 than in 2015. 

Characteristics of Policies Sold
Average Premium

The average premium per new sale was $2,480, almost unchanged from $2,497 in 2015. Two insurers had average premiums between $1,505 and $1,565, while three insurers were between $3,070 and $3,129. The average premium per new purchasing unit (i.e., one person or a couple) was also similar to last year, dropping from $3,525 in 2015 to $3,496 in 2016. The average in-force premium rose 0.7 percent to $2,116.

Issue Age
Table 5 summarizes the distribution of sales by issue age band based on insured count. The average issue age was stable (55.8 in 2016 vs. 55.9 in 2015), but still a record low. Table 5 shows that a record high 17.9 percent of buyers were between 30 and 49 and a record low 3.6 percent were 70 or older. All but two insurers had their most sales in the 55-59 block; one had their most sales in the 60-64 block and one had their most sales in the 65-69 block. Three participants have a minimum issue age of 40, two won’t issue below 30, and two won’t issue below 25. 

Benefit Period
Table 6 summarizes the distribution of sales by benefit period. The combined percentage of sales for benefit periods of four years or less in 2016 remains the same as in 2015. While most carriers continued to gravitate toward shorter benefit periods, one insurer sold many eight-year benefit period policies, hence the average notional benefit period increased slightly from 4.01 years to 4.07 years. Because of shared care benefits, total coverage was higher than the 4.07 average suggests. Endless (lifetime) benefit period sales may register on this distribution in 2017 because one insurer is now offering such policies.

Maximum Monthly Benefit
A record 81.0 percent of 2016 policies were sold with a monthly or weekly maximum, which is superior to a daily maximum from a consumer’s perspective given the additional flexibility to use benefits. It was included automatically in 69.6 percent of the policies. Where it was optional, only 37.5 percent of purchasers opted for monthly or weekly determinations.

Table 7 summarizes the distribution of sales by maximum monthly benefit at issue. The average maximum benefit decreased about 0.5 percent, staying at about $4,800 per month.

Benefit Increase Features
Table 8 summarizes the distribution of sales by benefit increase feature. Future purchase options (insureds buy more coverage in the future at attained age prices, 36.0 percent), deferred options (purchasers can add level premium compound benefit increases within five years of issue if they have not been on claim, 3.8 percent), and benefits scheduled to be flat (15.2 percent) now comprise 55 percent of LTCI sales, and “other compound” has surged to 10.3 percent of sales as future purchasing power is sacrificed to produce lower premiums.

The level-premium three percent compound increase provision, once called the “new five percent,” has dropped from 30.1 percent of sales to 23 percent of sales in two years; however, 3.2 percent of that 7.1 percent drop has shifted to 3.5 percent compound and step-rated. Five percent compounded for life, which represented more than 47.5 percent of sales each year from 2006 to 2008, now accounts for only 2.3 percent of sales. 

The “Age-Adjusted” benefit increase features typically increase benefits by five percent through age 60, by three percent compound or five percent simple from 61 to 75, and by zero percent after age 75.

“Indexed Level Premium” policies are priced to have a level premium, but the benefit increase is tied to an index such as the consumer price index (CPI).

A small error in the 2015 distribution has been corrected for one carrier, primarily shifting data from the “Other” category to various compound benefit options.

We project the age 80 maximum daily benefit by increasing the average daily benefit purchased from the average issue age to age 80. We project benefits according to the distribution of benefit increase features, using current future purchase option (FPO) election rates and assuming a long-term three percent CPI. The maximum benefit at age 80 (in 2040) for our 2016 average 56-year-old purchaser projects to $281/day. Had our average buyer bought an average 2015 policy at age 55, his age 80 benefit would be $300/day. The age 80 coverage for 2016’s average buyer is six percent less than if that person had bought in 2015 and 24 percent less than a purchase in 2014. Combining the reduction in sales with the reduction in coverage at age 80 for the average sale, the stand-alone LTCI industry sold about 26 percent as much coverage in 2016 as it did in 2012. The drop in coverage is really greater primarily because the average claim payment age (as opposed to the claim start age) is greater than 80. However, some of the difference has been covered by combination policies with LTCI benefits and policies with accelerated death benefits.

Six insurers provided both the number of available FPOs (at attained age rates) in 2016 and the number exercised. Table 9 shows 32.8 percent of insureds exercised FPOs that were available in 2016 based on their data. By insurer, election rates varied from 17 percent to 74 percent. The high percentage reflects an insurer using a “negative election” approach; i.e., the increase applies unless specifically rejected. Most carriers use “positive election” (the increase occurs only if specifically requested). 

Elimination Period (EP)
Table 10 summarizes the distribution of sales by facility elimination period. More than 96 percent of issued policies have facility elimination period selections of 84 days or longer.

The percentage of policies with zero-day home care elimination period (but a longer facility elimination period) has dropped from 38.9 percent in 2013 to 21.3 percent in 2016, which is largely due to change in sales distribution among carriers. In 2016, 35.6 percent of the policies had a calendar-day elimination period definition, compared with only 31.6 percent in 2015. When a calendar-day EP was available, 45.5 percent of policies had the feature; in some cases, it was automatic. 

Sales to Couples and Gender Distribution
Table 11 summarizes the distribution of sales by gender and couples status. It shows that 44.4 percent of couples insure only one spouse/partner. Sometimes one spouse already has coverage (perhaps left over from a previous marriage). Sometimes one spouse is declined and the other buys. The percentage of single people was low, but the percentage of females among single insureds was high. It appears the one-of-a-couple sales include a good percentage of males.

When one spouse is declined, the other spouse completes the purchase 71.4 percent of the time.

Shared Care and Other Couples’ Features
Table 12 summarizes the distribution of sales by shared care and other couples’ features. It shows that a lower percentage of both-buying couples bought couples’ features than in 2015:

Changes in distribution by carrier and product designs impact year-to-year differences in results in Table 12, particularly because sometimes survivorship or joint waiver is embedded automatically. 

Because some insurers don’t offer these features, Table 12 also shows the (higher) percentage that results from dividing the number of buyers by sales of insurers that offer the feature. On this basis, the percentage of people buying shared care did not drop as much, partly because we refined the calculation this year to reflect a carrier that makes shared care available only for some benefit periods.

Table 13 provides additional breakdown on the characteristics of shared care sales. As shown on the right-hand side of Table 13, eight-year (37.7 percent), three-year (33.2 percent), and four-year (33.1 percent) benefit period policies are most likely to add shared care. Because three-year benefit periods comprise 42.2 percent of sales, most policies with shared care are three-year benefit period policies (as shown on the left side of Table 13).

Above, we stated that shared care is selected by 36.0 percent of couples who both buy limited benefit period policies. However, Table 13 shows shared care comprised no more than 37.7 percent of any benefit period. Table 13 has lower percentages because Table 12 denominators are limited to people who buy with their spouse/partner whereas Table 13 denominators include all buyers.

Existence and Type of Home Care Coverage
One participant reported home-care-only policies, which accounted for 1.0 percent of industry sales. Four participants reported sales of facility-only policies, which also accounted for 1.0 percent of total sales. Ninety-seven percent (96.6 percent) of the comprehensive policies included home care benefits at least equal to the facility benefit. These percentages were all within 0.2 percent of 2015 results.

Partial cash alternative features (which allow claimants, in lieu of any other benefit that month, to use between 30 percent and 40 percent of their benefits for whatever purpose they wish) were included in 30.3 percent of sales. This is more than double the 14.2 percent of 2014, because the two insurers that dominate such sales include these features automatically and each sold more business in 2016 than in 2014 in an industry in which total sales declined. Another insurer covers some family care that is part of the plan of care. Overall, therefore, the percentage of 2016-issued policies that cover some family-provided care was 43.6 percent.

Other Characteristics
Return of premium (ROP) features were included in 26.7 percent of all policies (up from 24.8 percent in 2015 and 10.5 percent in 2014). ROP returns some or all premiums (usually reduced by paid LTCI benefits) when a policyholder dies. Approximately 93 percent of policies with ROP arise from ROP features embedded automatically in the product. Embedded features are designed to add minimal costs to premium, typically decreasing the benefit to $0 by age 75. 

Thirteen percent (13.2 percent) of policies with limited benefit periods included a restoration of benefits (ROB) provision, up from 9.8 percent in 2014. ROB provisions typically restore used benefits when the insured does not need services for at least six months. Approximately 79 percent of policies with ROB arise from ROB features automatically embedded, compared with 70 percent in 2015.

Shortened benefit period (SBP) nonforfeiture option was included in 1.2 percent of policies, the same as 2015. SBP makes limited future LTCI benefits available to people who stop paying premiums after three or more years.

Only one insurer issued non-tax-qualified (NTQ) policies, which accounted for 0.1 percent of industry sales.

“Captive” (dedicated to one insurer) agents produced 48.9 percent of the policies, a percentage that has risen by nearly 2.5 percent annually for the past few years. At one time, “captive” agents who sold LTCI tended to specialize in LTCI. Now many are agents of mutual companies.

Sales distribution by jurisdiction is posted on the Broker World website.

Limited Pay
Only two insurers sold 10-year-pay, 20-year-pay, paid-to-65, or paid-to-75 policies in 2016, collectively accounting for 0.7 percent of sales (down from 0.8 percent). Limited-pay features can be attractive for consumers because they can help avoid a potential future rate increase that would apply for lifetime-pay policies. With today’s policies priced based on substantially more data, and thus with a lower likelihood of premium rate increases, such a motivation recedes. Furthermore, with today’s higher prices, the additional cost of 10-year-pay for pass-through entities (employers that report earnings to their owners to be included on the owner’s tax return, as opposed to paying income tax directly to the federal government) is almost all paid in after-tax dollars. In the past, some of the additional 10-year-pay premium was tax-effective for pass-through entities.

Nonetheless, limited-pay and single-pay policies are attractive to assure reduced future expenditures, for §1035 exchanges and other situations. In 2017, more limited pay policies potentially will be purchased, as a carrier has entered the market with single pay and 10-year pay.

Partnership Programs
When someone applies to Medicaid for long-term care services, states with Partnership programs disregard assets up to the amount of benefits received from a qualified LTCI policy. Partnership sales were reported in 43 jurisdictions in 2016, all but Alaska, District of Columbia, Hawaii, Massachusetts, Mississippi, New Mexico, Utah, and Vermont. 

Four participants offered Partnerships in 39 jurisdictions (none exceeded 39). One participant has never certified Partnership conformance and insurers sometimes delay certifying policy forms as Partnership because of other priorities (e.g., needing time to comply with state-specific requirements to notify existing policyholders or offer an exchange). Such delay is not harmful, as certification is retroactive to policies already issued on that policy form. Few currently sell Partnerships in the original Partnership states: California (one insurer), Connecticut (four insurers), Indiana (five insurers), and New York (three insurers).

We estimate that only 31.9 percent of policies sold nationwide (plunging from 38.9 percent in 2015 and 48.0 percent in 2014) would qualify under the Deficit Reduction Act (DRA), but we estimate that 47.1 percent of the policies in DRA states satisfy the Partnership requirements. Clearly, DRA Partnerships encourage stronger benefit increase features to be purchased, even though the program is faltering because of less robust sales. Unfortunately, in the original Partnership states, only 7.6 percent of the policies qualified for Partnership, the highest percentage being 14.9 percent in Connecticut. California has passed legislation to allow compounding less than five percent to qualify; however, we are not aware of any insurer having certified a design other than five percent compounding yet.

Minnesota led all states with 77.4 percent of participant policies being Partnership-qualified, followed by Maine at 70.5 percent. In 2014, six states topped 80 percent and seven others topped 70 percent.

Partnership programs could be more successful if:

  1. Advisors offered small maximum monthly benefits more frequently to the middle class. A $1,500 initial maximum monthly benefit covers about four hours of home care every two days and, with compound benefit increases, may maintain buying power. Many middle-class citizens would like LTCI to help them stay at home while not “burning out” family caregivers and would be motivated further by Partnership asset disregard. (Note: this approach does not work in the original Partnership states as their Partnership minimum sizes preclude small policies.)
  2. Middle-class clients were better educated about the importance of benefit increases to maintain LTCI purchasing power and to qualify for Partnership asset disregard.
  3. Benefit increase rules were loosened. A state helping the industry by establishing a Partnership program is entitled to whatever requirements it chooses. Encouraging benefit increases makes sense, to avoid major reduction in purchasing power by the time long term care is needed. However, any purchase helps defray Medicaid costs and, with a dollar-for-dollar offset, a state can’t lose on any insured and will definitely gain overall. If a state wanted to expand Partnership but still encourage compounded benefits, reduced credit could be given to benefits from policies that do not include compounding.
  4. More jurisdictions adopt Partnership programs.
  5. Programs to privately finance educational direct mail from public agencies were adopted more broadly.
  6. Financial advisors were to press reluctant insurers to certify their products.
  7. More financial advisors were certified. Some people argue that certification requirements should be loosened. At a minimum, the renewal certification process could be improved.
  8. Insurers modified worksite programs to encourage Partnership-qualified core benefits.

Underwriting Data
Case Disposition

Table 14 shows the distribution of case resolutions among LTCI applicants. Eleven insurers contributed case disposition data to Table 14. In 2016, 60.9 percent of applications were placed, including those that were modified, which is a little higher than 2015’s record low of 60.7 percent.

Low placement rates can be problematic for insurers, as cost per placed policy increases. Financial advisors have a more serious problem. In addition to wasting time and effort encouraging clients to apply for LTCI, they create disappointment for clients who are rejected. Years ago, when financial planners complained about decline rates, it was customary to explain that, if they proactively discussed LTCI with clients instead of simply responding to clients who specifically asked for LTCI, they would have better results. Such a response does not address today’s problems; low placement rates make it much harder for wholesalers to encourage advisors to discuss LTCI with their clients.

Decline rates have exceeded 20 percent since 2010. In 2016, they set a new record, 23.3 percent of applications and 25.8 percent of insurer’s decisions. 

Nearly 40 percent of applications (39.1 percent) were declined, suspended, or withdrawn, despite the larger concentration in ages with the lowest decline rates. 

Our placed percentages reflect the insurers’ perspective. It is important to recognize that a higher percentage of applicants secure coverage because applicants denied by one carrier may be issued either stand-alone or combination coverage by another carrier or may receive coverage with the same insurer after a waiting period.

Underwriting Tools
Table 15 summarizes by application how often various tools are used in the underwriting process. Ten insurers contributed data to Table 15. Medical records, phone interviews, and prescription profiles continued to increase in 2016. MIB*, face-to-face exams, parameds, and medical exams decreased. Change in distribution of sales among insurers from year to year significantly impacts results. Reduced maximum ages result in fewer face-to-face exams. Note: an insurer might underreport the use of an underwriting tool because it may lack a good source for that statistic. *MIB Underwriting Services alert underwriters to errors, omissions, misrepresentations and fraud on applications for life, health, disability income, long term care and critical illness coverage.  MIB, Inc. provides its Underwriting Services exclusively to authorized individuals in MIB Group, Inc. member companies.

Underwriting Time
Table 16 summarizes the time from receipt of application to mailing the policy. It shows a reduction for the third straight year. The average policy took 38 days from receipt of application to policy issue, compared with 44 (2015), 46 (2014), 51 (2013), and 34 days (2012). The worst year for underwriting speed remains 2013.

Rating Classification
Table 17 summarizes the distribution of sales by underwriting class. The percentage of cases placed in the two most favorable risk classifications was higher in 2016 than in any year since 2012. 

Table 18 summarizes by issue age for fully underwritten policies the percentage of policies issued in the insurer’s most favorable rating class. Table 18 shows that, for issue ages under 60, a lower percentage of fully underwritten policies was issued in the most favorable rating in 2016 than in 2015, but for ages 60+, the reverse was true.

Premium Rate Details
Our Premium Exhibit (available on the Broker World web-site, shows level annual lifetime premiums for the insurer’s most common underwriting class, for issue ages 40, 50, 60, and 70 for single females, single males, and heterosexual couples (assuming both buy at the same age), reflecting the following features:

For worksite products, we’ve shown the gross premium, reflecting couples’ discounts but prior to any worksite discount.

Each insurer’s distribution by underwriting class and b) their premium adjustments (from our published prices) by underwriting class are available here.

Click here for the 2017 Milliman LTCI Survey Product Exhibit.

We thank insurance company staff for submitting the data and responding to questions promptly. We also thank Nicole Gaspar, Derek Montgomery, and Sam Pianetto of Milliman for managing the data expertly.

We reviewed data for reasonableness and insurers reviewed their product exhibit displays. Nonetheless, we cannot assure that all data is accurate.

If you have suggestions for improving this survey (including new entrants in the market), please contact one of the authors. 

Author's Bio
Claude Thau
founded Thau, Inc. to help build a sound long term care insurance industry. He does that by: wholesaling LTCI through financial advisors nationwide in conjunction with Target Insurance Services; consulting for insurers, other consulting firms, employers, regulators, etc., (for example, he was a consultant to the Federal LTCI program); and by doing pro bono work related to LTCI and long term care. He believes LTCI is a secondary industry, as there is no purpose to LTCI if people can’t get good quality care. Thau’s LTCI experience is unusually broad and deep. After a career as an actuary, he accepted responsibility for a major company's LTCI division, which then grew five times as fast as the rest of the LTCI industry for each of three consecutive years. When the entire company was sold, he set up his own company in 2000. Since 2005, he has been the lead author of the annual LTCI surveys printed in Broker World. In 2007, he was named one of the 10 “Power People” in the LTCI industry by Senior Market Advisor. A former inner-city public school teacher, Thau enjoys mentoring financial advisors to help them decide how they can grow their business and educate their clients. He can be reached by telephone at 913-403-5824. Fax: 913-384-3781. Email:

Allen Schmitz, FSA, MAAA
Schmitz, FSA, MAAA, is a principal and consulting actuary with the Milwaukee office of Milliman. Schmitz can be reached at 15800 Bluemound Rd., Brookfield, WI 53005. Telephone: 262-796-3477. Email:

Chris Giese
FSA, MAAA, is a principal and consulting actuary in the Milwaukee office of Milliman, Inc. He can be reached at 15800 Bluemound Road, Suite 100, Brookfield, WI 53005. Telephone: 262-796-3407. Email: