Daily Benefit Amounts and Triggers
At least two and up to six levels of benefits are required for the CLASS plan32. We propose two levels of benefits:
- $50 initial daily benefit at 2-3 ADL deficiencies
- $60 initial daily benefit at 4+ ADL deficiencies or cognitive impairment.
The amounts are chosen to discourage beneficiaries from claiming a higher level of disability than is warranted.
After the first year of an individual’s enrollment, the daily benefit will increase according to the increase in the Consumer Price Index for Urban Consumers (CPI-U, or simply CPI). This is more generous than the required benefit increase in the statute. One potential interpretation of the statutory language is that the increase in the benefit amount begins only when a beneficiary receiving benefits (that is, everyone who claims starts with a $50 daily benefit). Another potential interpretation is that the increase in the benefit amount starts from the 6th year of the CLASS Program. Note that increases in long-term care service costs have been historically higher than the CPI33. While this proposed benefit indexing method may not cover future inflation increases fully, it does help to keep the premiums affordable.
In addition to the ADL and cognitive impairment triggers, the Secretary is authorized to define another similar benefit eligibility trigger34. A possible third trigger would be one for mental illness. We believe the cognitive trigger already accommodates beneficiaries with mental retardation and some people with intellectual disability. As well, there is no established test similar to the standardized test for cognitive impairment. A standardized test with sufficient experience data is a highly desirable for estimating claim incidence rates in an insurance setting. Adding this third trigger at this time will make the CLASS Benefit Plan more expensive and increase the likelihood of inadequate premiums for the Program. If the Program shows favorable results, we should revisit this issue.
Reduced Benefit Amounts
According to the statute, benefits are “not subject to any lifetime or aggregate limit.”35 As private insurance has painfully learned, because beneficiaries have no incentive to preserve their benefits when there is no lifetime limit, unlimited benefit plans are expensive and risky. Most policies with lifetime benefits had significantly unfavorable experience. Premiums for lifetime benefits can be more than 50% higher than the corresponding premiums for a limited benefit period plan. The vast majority of group long-term care insurance policies today have a limited benefit period.
Our recommendation is to satisfy that statute by paying 100% of the daily benefit for the first five years and then only 20% thereafter, without limiting the aggregate amount of duration. This feature is intended to meet the requirements in the statute, keep CLASS premiums reasonably close to that for private insurance, and minimize one potential source for claim variability.
With regard to the form of benefits for the eligibility beneficiaries, the statute states the following (emphasis added): “Cash benefits paid into a Life Independence Account of an eligible beneficiary shall be used to purchase nonmedical services and supports that the beneficiary needs to maintain his or her independence at home or in another residential setting of their choice in the community…”36. Accordingly, we question whether payments of straight cash are required. Instead, cash equivalents in individual accounts could be used to purchase services and supports for the beneficiaries. The provisions of the statute appear to grant a great degree of autonomy to the beneficiaries on the choice and specifics of the services and supports.
We are not opposed to paying cash when it is appropriate. We believe it is difficult to determine whether it is appropriate or not. Sound insurance systems require a demonstration of a real loss. With cash payments as benefits without limitations, the enrollees are incented to claim whether there is a real loss or not. Unfortunately, we believe the statute gives conflicting messages regarding benefits. On one hand, it appears to limit the usage of the benefits. One the other hand, it appears to suggest cash payments without accounting for its use.
We utilized some of the learning from Cash and Counseling demonstration projects in the following proposed benefit structure. A counselor will be assigned to provide advice and assistance in planning of services and support, in particular, assistance in developing a plan of care. The cost of counseling will be part of claim administrative expenses and will not reduce the benefit amounts. Funds will be deposited into an individual account (the Life Independence Account) of the beneficiary on a weekly basis. Under the plan of care and the advice of the counselor, the beneficiary can direct available funds in the account to obtain the services and supports needed as long as they are not on a list of exclusions. This list would include items such as food, rent, liquor and luxury items. Within guidelines set by the CLASS Office, the counselor can make exceptions to the list. Associated with the account, the beneficiary will receive a debit card with which permissible services and supports can be purchased. Cash could not be withdrawn from the account via the debit card.
The CLASS Office will establish a process to train, monitor and evaluate the performance of the counselors. If beneficiaries need direct-paid care services (for example, unlicensed home health aide), they must use a fiscal manager to handle various employment and reporting requirements. The expenses associated with the fiscal manager will be deducted from the account balance. Specific activities and time spent by the paid direct care worker (including paid family members) must be fully documented.
Our goal is to strike a balance between beneficiary-directed benefits and the potential for induced demand from the perception of a free-flow of cash. According to the statute, “nothing in this title shall be construed as prohibiting benefits paid under the CLASS Independence Benefit Plan from being used to compensate a family caregiver…”37 We remain concerned regarding unwarranted payments to family members, in particular, the spouse. It is difficult to differentiate spouse’s free time from the time spent that needs to be compensated. Moreover, payments to a spouse are effectively cash to the beneficiaries. We are considering a reduced payment amount to a spouse (for example, $10 daily).
Regardless of our concerns, the proposed benefit structure is a significant departure from typical long-term care insurance benefits and should be viewed as an attractive feature of the CLASS plan.
Long-term care insurance claim patterns are typically characterized by very few claims in the early years of a program and a significantly higher number of claims in the later years. Level premiums develop a relatively high fund balance in the early years when premiums exceed claims and expenses. This relationship reverses in later years.
For the CLASS Program, an increasing premium schedule provides several advantages:
Increasing premiums make the initial premiums lower than the corresponding level premiums. This may result in higher enrollments. This, in turn, increases the spread of risk and improves the chance for program sustainability.
An increasing premium plan provides less inflow to the trust fund in the early years than a level premium plan. The Program is subjected to less investment risk in matching the cash flow from assets with the cash flow from claim and expense obligations. Lower cash inflow also places less reliance on the expected relatively moderate fund returns available to the CLASS Program which is limited to an investment portfolio of Treasury Securities.
As increasing premium schedules are still uncommon in private insurance, they could give the CLASS Program a temporary competitive advantage.
From the enrollees’ perspective, an increasing premium schedule starts out significantly lower than the corresponding level premium but eventually exceeds it in later years. Lower initial premiums mean more efficient use of enrollees’ discretionary dollars. Premiums go up relatively in line with increases in plan benefits and general wages. Also, if enrollees decide to lapse, their cumulative outlays would be less than their cumulative outlays for level premiums for the same period of coverage. However, a potential disadvantage of an increasing premium schedule is that enrollees may not be able to afford the premiums in their later years when the enrollees’ income is relatively fixed. This can be addressed by exempting enrollees from scheduled premium increases after a certain age or period of enrollment.
One method of implementing an increasing premium schedule is to index the premium by the CPI in the same manner as the daily benefit amount increases by the index. This has the undesirable feature of ever-increasing premiums beyond the retirement years. Also, there will be uncertainty each year regarding the amounts of future increases.
We propose an increasing premium schedule with a 4% compounded annual increase that stops at age 70 (or after 5 years of enrollment, whichever is later). Premiums are level thereafter38. The 4% is chosen to provide an attractive entry price point to most enrollees. Other increasing premium schedules, such as indexing up to a specific age, are also under consideration.
After enrollment, if an enrollee finds the increasing premium schedule unaffordable, the enrollee will have the option to freeze future premium increases with a corresponding freeze in benefits. However, if premiums need to increase for other reasons (not related to the set annual increase), enrollees who have frozen their premiums will still be subject to those increases.
Other Plan Features
Waiver of premiums during the time a beneficiary is receiving benefits is not a statutory requirement but is common in private long-term care insurance. To be competitive, the CLASS plan should also include this feature, except when the Limited Initial Benefit provision applies.
Premium discount for spouse coverage is also common in private insurance in recognition that couples tend to have lower claim costs than singles. However, we believe that premiums can only vary by the age at enrollment in CLASS. A response to the premium discount would be to allow a couple to share their benefits. If a spouse is a beneficiary for more than 5 years, the benefit amounts would not be reduced (that is, to 20%) to the extent that the other spouse’s 5 year period has not been used up.
Certain group and most individual long-term care insurance plans have preferred risk discounts. We believe the CLASS Program has no apparent answer to this feature.
The statute contains the following provision regarding administrative expenses (emphasis added): “In determining the monthly premiums for the CLASS program the Secretary may factor in costs for administering the program, not to exceed … an amount equal to 3 percent of all premiums paid during the year.”39 Expenses for protection and advocacy services and advice and assistance counseling are counted as administrative expenses40.
It is our opinion that there is no explicit restriction on the actual administrative expenses. Other than the initial roll-out, there is no guarantee of additional federal funding in support of these services. It would be unrealistic and undesirable to limit actual administrative expenses to only 3% of premiums (see our estimate below). Such a limitation would allow for little or no advocacy services or advocacy and assistance counseling. Claim payments might not be made on time. Enrollment might have to be curtailed. Experience monitoring might not be done to ensure timely premium adjustments in order to maintain solvency. Fraud and abuse prevention and monitoring might be limited or non-existent. These events would adversely affect the integrity and ultimate existence of the Program.
There is an important distinction between expense assumption in the premium development and actual future expenses which are not guaranteed to be realized. The process of determining premiums involves making assumptions regarding future claims, investment returns, persistency, as well as expenses. None of these assumptions can be guaranteed to hold during the lifetime of the enrollees in the Program. In order for the plan to be actuarially sound and the Program to be sustainable, these assumptions must be realistic. In selecting these assumptions, we anticipate that favorable experience for one of these factors may offset unfavorable experience for another such that the plan is overall actuarially sound. For example, favorable claim experience can offset unfavorable investment experience, leaving the Program actuarially sound overall. The statutory provision may necessitate choosing assumptions that are expected to be reasonable in the aggregate, with some conservative margin incorporated into the other assumptions to compensate for the aggressive 3% expense assumption.
Our current estimates of Program expenses expressed as percentages of premiums are as follow:
These expenses are derived from a survey of the pricing expense assumptions of group long-term care insurance and experience from the Federal Long Term Care Insurance Program. We expect that these expenses will rise according to the CPI. Because certain components of these costs are fixed or not a function of premiums, these percentages reflect the relatively low estimated average premium (under $1,000 annual premium) compared to other forms of insurance (for example, health insurance). Thus metrics from other insurance are not always directly transferrable to the CLASS Program.
The statute is silent on the use of investment returns from the trust fund. Because premiums collected are expected to exceed benefits and expenses in the early years, we expect considerable build-up of the balance in the trust fund. Investment returns from the trust fund are a source of revenue in addition to premiums. This revenue can be used to cover expenses.
Rather than modeling expenses solely as a percentage of premiums, a more appropriate set of expense assumptions would delineate the following:
- fixed expenses,
- expenses that vary by the number of enrollees,
- expenses that vary by the number of beneficiaries and the length of the claims, and
- expenses by vary by premiums.
Compared to this simplistic assumption based on premiums alone, this approach will minimize the likelihood of actual expenses exceeding expected expenses due to incorrect estimation of the average premium per enrollee.
Minimum Earnings Requirement and Nominal Premium
Under the statute, eligible beneficiaries must meet a minimum earnings requirement for 3 out of the first 5 years of enrollment equal to the amount of wages to be credited with one quarter of Social Security coverage41. This amount is $1,120 for 2011. However, the Secretary is authorized to make exceptions to this rule for certain population42. A low minimum earnings requirement allows enrollment of low-income workers who, as a group, are generally in poorer health43.
Furthermore, the statute requires a $5 nominal monthly premium for enrollees whose income is below the poverty line or who are full-time students under age 2244. This provision poses a potential threat to the financial viability of the CLASS Program. Without any mitigation for adverse selection, this is a tremendous incentive for a poor worker, who is near or already met the benefit triggers, to enroll. The ‘returns’ in the form of benefits (approximately $18,000 a year for the first year and indexed higher thereafter) far exceed the ‘investments’ of 5 years of $5 monthly premiums ($300). Because other enrollees will be subsidizing the poor, the overall premiums will need to be higher than without this subsidy. This adds to the competitive issue with private insurance. In addition, estimation of the appropriate premiums will not be reliable due to the unpredictable mix of nominal and regular premiums.
In order to mitigate this threat, raising the minimum earnings requirement for certain enrollees so that it is always above the poverty level could be explored. This should be somewhat effective in controlling adverse selection by those who expect to receive a high level of benefits for very little premium.
As stated above, the Secretary is authorized to promulgate regulations on exceptions to the minimum earnings requirement for certain populations. Under the Phased Enrollment plan design, those enrolled through individual enrollment are certainly a population that needs special attention since most of the adverse selection is expected to come from this group. Under the Temporary Exclusion design, the potential problem is less severe but not entirely eliminated. In either case, raising the minimum earnings requirement on this group would help. In addition, since it is not explicitly prohibited, defining “actively employed” to require a minimum number of work hours at time of enrollment can also help mitigate adverse selection. Finally, there should be a specific regulation to prevent companies from forming for the sole purpose of enrolling employees into the CLASS Program.
Reenrollment after Lapse
For reenrollment after a lapse from the CLASS Program, the statute prescribes separate treatments based on the time period between lapse and reenrollment:
- less than 90 days,
- 90 days to 5 years, and
- over 5 years45.
Potential for gaming exists for reenrollment after 90 days. For lapse period between 90 days and 5 years, the statute merely requires 2 years of continuous payments (of premiums for a new enrollee at the same attained age at reenrollment) and the initial 5 year vesting period before claiming. This encourages a ‘skip-and-go’ scheme where an enrollee would lapse, reenroll within 5 years from lapse, pay a premium, lapse again, and reenroll permanently only when his or her health deteriorates. If the lapse period is over 5 years, a similar but slightly less devious gaming opportunity also exists if no control is in place.
We propose the following rules in order to treat the reenrolled individuals and the in-force enrollees equitably. If the lapse period is less than 90 days, payment of due premiums is required to maintain enrollment. If the lapse period is between 90 days and 5 years, the individual must:
- Pay all back premiums except for the first lapse,
- Pay future premiums based on the attained age at reenrollment, and
- Meet the minimum earnings requirement again, or restart the 15 year exclusion period if the plan included Temporary Exclusion.
If the lapse period is over 5 years:
- Meet the minimum earnings requirement again,
- Pay future premiums based on the attained age, and
- Pay an actuarially sound premium that anticipates the enrollee will be very likely to claim or restart the 15 year exclusion period if the plan included Temporary Exclusion.
Minimum $50 Average Daily Benefit
The statute requires that the daily benefits paid be at least $50 on average as determined based on the reasonably expected distribution of beneficiaries receiving benefits at various benefit levels46. There are numerous ways to interpret this requirement. The $50 minimum could apply to each year’s claims and the averages could be determined on a present value basis. We have identified the following 6 methods of calculations that can fit this requirement:
At the beginning of each year, the present value of all expected future benefits from all enrollees divided by the total expected future days of claims from all enrollees.
At the beginning of each year, the present value of all expected future benefits of new beneficiaries during the past year divided by the corresponding total number of claim days.
At the beginning of each year, the present value of all expected future benefits of all beneficiaries during the year divided by the corresponding total number of claim days.
The total expected daily benefits available each year for all beneficiaries divided by the total expected number of claim days.
For each new beneficiary, the expected total future payments divided by the expected total number of days in claim.
The expected total future payments to each beneficiary for each 12 month period divided by the expected future number of days in claim during such period.
Another related issue is whether the minimum $50 is indexed by CPI or not. While it is clear that benefits are indexed after claim, it is not clear that benefits are required to be indexed while the enrollee is active past the first 5 years of the program, but not in claim status47. Note that we have proposed the benefit amounts starts indexing on the second year of an individual’s enrollment, which is the most liberal interpretation.
We take the position that the $50 minimum average is not indexed and elects the first calculation method described above. We plan to monitor actual results and adjust benefits if necessary in order to comply with the requirement.