Stressors in the long term care industry, like premium increases, have resulted in lawsuits, market consolidation and bad press. This turmoil has encouraged conservatism in pricing and guardedness in benefit designs.

Criticism has recently targeted innovative benefit designs such as so-called “cash” products. Cash products pay eligible insureds cash based on standard benefit triggers, versus the standard model of reimbursing paid services. Some believe cash products cannot be actuarially sound. An executive at the 2007 Intercompany Long-Term Care Insurance Conference in Dallas, said, “Cash at 100% is stupid and wrong.”

Why are people worried about cash products? Fears presumably include adverse selection and moral hazard, resulting in changing behavior of insureds and the eventual discovery that such products have been underpriced.

In our opinion, however, cash products can not only be properly priced, but will be more stable over the long run. Disability rates are stable and declining, so this is not a pressure on pricing. Changes in care delivery or legislative pressures place greater price adjustment risk on typical reimbursement products, while the inherent flexibility of cash products compensates for this.

Pricing of cash products

Cash product pricing is entirely consistent with the pricing of reimbursement products, except for the morbidity risk. Both depend on disability rates among insureds. For reimbursement products, the probability that someone will make a claim is based on (a) disability under the policy’s definition (generally, disability in two activities of daily living or cognitive impairment), and (b) receipt of paid services for that disability. The percentage of disabled people assumed to seek paid services is known as the “paid help” assumption. A cash product must account only for the first risk (a).

Disability rates

Population disability rates have exhibited two important trends. First, disability rates are decreasing (note the increasing rate of “nondisabled” in the table from the National Long-Term Care Survey). However, this risk could be made unstable by not allowing for changes in care delivery, as with more restrictive reimbursement policies.

Paid help

The second important trend is that care delivery is changing; the proportion of disabled people likely to make home healthcare or alternative care (non-institutional) claims is growing (note the decreasing rate of “institution” in the chart relative to other groups). The paid help percentage assumption, combined with this non-institutional proportion, is the key pricing issue to differentiate between a cash product and a reimbursement product.

According to the National Medical Expenditure Survey, Supplement on Aging (1984), 36% of impaired older males and 52% of females received paid help. These rates are heavily influenced by access to financial resources such as insurance. In a federally-sponsored demonstration in 1980, for example, non-institutionalized disabled elders who were given access to community-based care exhibited paid help rates of roughly 90%.

Thus different care delivery systems produce different paid help rates. One source suggests that if disability rates continue to decline, the number of older adults using paid home healthcare services could increase by three quarters between 2000 and 2040.

For reimbursement products, underestimated paid help assumptions can lead to potentially large premium adjustments. Suppose that the implicit paid help percentage assumption for non-institutional care is 33%. This implies that the percentage of non-institutionalized insureds making claims could be 3 times too low if changes in care delivery entice all eligible insureds to use services. If the actual paid help rate was 90%, premiums would need to increase by roughly 50%.

Cash products using a 100% paid help assumption, however, have compensated for these risks, leading them to be less risky with respect to future care delivery changes.

Risks of cash products

This does not mean that cash products are without risk; they are more likely to induce adverse selection and moral hazard. It is important to monitor and manage these risks.

Adverse selection

Presumably, part of the industry’s reservation is that cash products may attract riskier and costlier customers than anticipated. Firms must consider 3 issues:

1) Customers must understand that benefits are not income replacement for temporary disabilities. Benefit triggers require specific, chronic conditions.

2) Sales messages must be clear and accurate. The term “disability” should be used carefully.

3) All long-term care insurance underwriting is becoming more difficult as customers get younger (and have fewer medical records). The average issue age for an individual policy fell from 66 in 2001 to 62 in 2006, according to a Society of Actuaries experience study.

Moral hazard

Policies may create incentives to seek benefits where one otherwise would not. This risk is much greater with cash products for non-institutional care than nursing home care, because making claims involves seeking payment with no services required. Several considerations arise:

1. Actuaries should assume some “ADL creep,” in which people raise their disability status under the policy.

2. Eligibility determination must be more active (and thus more costly) than for reimbursement products, particularly during the initial period of contestability when disability rates should be low.

3. Conversely, claims administration for the non-institutional benefits of cash products is simpler and less costly once eligibility is established, because administrators need not track individual services.


Older, less flexible policies have come under fire lately. An Iowa Insurance Division report notes that many older policies do not cover many, now common, nursing home alternatives. The report proposes that these carriers offer a one-time opportunity to purchase coverage for these alternatives. This guarantee would be problematic for insurers, inviting adverse selection, according to a LTC insurance study done by the Iowa Insurance Division. As other types of care delivery emerge, this pattern will likely be repeated.

Cash products with a 100% paid help assumption could have more stable premiums and be more useful for beneficiaries in the long run, while standard reimbursement products are not designed or priced to account for changes in care delivery.

As baby boomers age, does anyone really think the delivery system will remain unchanged?

Jon Shreve, FSA, MAAA, and Jill Van Den Bos, MA, are healthcare benefits consultants in the Denver office of Milliman. You can contact them via e-mail, respectively, at or .