Long-term Care insurance: What Do People Want? Results from a Discrete Conjoint Experiment

Tuesday, June 14, 2016: 9:10 AM
B26 (Stiteler Hall)

Author(s): Ben Allaire; Derek S Brown; Joshua Wiener

Discussant: R. Tamara Konetzka

Almost half of all people turning 65 between 2015 and 2019 are projected to need long-term support and services (LTSS). With limited ability to pay for those resources, many elderly rely on Medicaid, family members, or go without needed care. Private LTC insurance (LTCI) could help pay for LTSS. However, demand for LTCI is low (about 10% or less nationwide), leading to a current equilibria with limited choices, or restricted supply. Policymakers widely agree that care for the elderly could be improved through better functioning of the LTCI market.

In this study, we use a stated preference (SP) survey to elicit preferences of US adults for LTCI plans. We estimate demand for LTCI, willingness to pay, and the value of various policy alternatives. A conjoint analysis, or discrete choice experiment (DCE), survey was developed and fielded to a nationally representative online sample of n=15,298 US adults ages 40-70. The survey included a variety of controls including personal characteristics, attitudes, and experiences with LTCI.

Respondents completed 8 paired comparisons, selecting which of 2 LTCI plans they would purchase or a neither LTCI plan alternative. A full profile conjoint design was used, consisting of 7 attributes: daily benefit, benefit duration, deductible period, health requirements, private/public insurer, premium cost, and mandatory/voluntary enrollment. Assuming a random utility model specification, we estimated preferences using mixed effects logit (random parameters logit) and latent class logit. After the main preference model, we estimated willingness to pay, predicted demand, and consumer surplus

Female respondents were substantially less likely to choose an LTCI plan (OR=0.72) as were those in fair/poor health (OR=0.88). Predicted demand increased with the level of assets owned by the respondent. Assets over $100,000 represented the strongest predictor of LTCI uptake (OR=1.47). A majority of respondents felt it was their own responsibility to pay for their LTC: 57.5% agreed or strongly agreed with this statement and those that did agree were more likely to opt into a plan (OR=1.42).

Preferences for LTCI plan features were logically ordered and consistent. The most important attributes were premium cost, benefit duration, and mandatory/voluntary enrollment; the least important was deductible period. Based on these preferences, we estimated demand for a simulated plan with a moderate level of LTC coverage ($100 daily benefit amount, 3 year duration). Demand for this LTC plan (vs. no plan) ranged from 54% at a premium cost of $25/month to 27% at a cost of $250/month. Under voluntary enrollment with a private insurer, the estimated consumer surplus is $185/month.

We find that a moderately generous public LTC plan could be welfare-enhancing in a simple benefit-cost analysis if the public finance cost of this plan was less than $185/month. This study demonstrates a potential role for alternative, welfare-enhancing policies to promote flexible LTC plans.