Please use this identifier to cite or link to this item: http://hdl.handle.net/10419/61051
Authors: 
Cremer, Helmuth
Roeder, Kerstin
Year of Publication: 
2012
Series/Report no.: 
CESifo Working Paper: Public Finance 3843
Abstract: 
This paper examines whether myopia (misperception of the long-term care (LTC) risk) and private insurance market loading costs can justify social LTC insurance and/or the subsidization of private insurance. We use a two-period model wherein individuals differ in three unobservable characteristics: level of productivity, survival probability and degree of ignorance concerning the risk of LTC (the former two being perfectly positively correlated). The decentralization of a first-best allocation requires that LTC insurance premiums of the myopic agents are subsidized (at a Pigouvian rate) and/or that there is public provision of the appropriate level of LTC. The support for the considered LTC policy instruments is less strong in a second-best setting. When social LTC provision is restricted to zero, a myopic agent's tax on private LTC insurance premiums involves a tradeoff between paternalistic and redistributive (incentive) considerations and we may have a tax as well as a subsidy on private LTC insurance. Interestingly, savings (which goes untaxed in the first-best but plays the role of self-insurance in the second-best) is also subject to (positive or negative) taxation. Social LTC provision is never second-best optimal when private insurance markets are fair (irrespective of the degree of the proportion of myopic individuals and their degree of misperception). At the other extreme, when the loading factor in the private sector is sufficiently high, private coverage is completely crowded out by public provision. For intermediate levels of the loading factors, the solution relies on both types of insurance.
Subjects: 
long-term care
myopia
optimal taxation
JEL: 
D91
H21
I13
Document Type: 
Working Paper

Files in This Item:
File
Size
308.94 kB





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.