Willingness-to-Pay for Parallel Private Health Insurance: Evidence from Laboratory Experiment

Debate over the effects of public versus private health care financing has been, and continues to be, active in both academic outlets and policy circles. Theoretical literature on parallel health care financing is often built on untested behavioural assumptions and the empirical evidence generally depends upon the institutional details of the specific health care systems under analysis. This paper contributes to the literature on parallel health care finance by developing and executing a revealed preference laboratory experiment based on the theoretical model of parallel health care finance in Cuff et al. (2008). The theoretical model involves individuals in the labour force with varying severities of illness who demand health care from a limited supply of health care resources. Health care resources are purchased by the public sector and rationed free of charge to individuals, or purchased by individuals through a private insurance market. The general theoretical model is converted into a discrete experimental representation of a large-scale economy where individuals are price takers, the probability of receiving public health care is exogenous and the willingness-to-pay (WTP) for private health insurance is elicited from subjects. The experimental design includes two within-subject factors based on the theoretical model: the public sector rationing rule (rationing based on need or severity versus rationing based on a random allocation) and the probability of being publicly treated (high versus low). The experimental design also includes two between-subjects treatments based on the frame of the experiment (neutral frame versus health frame) and on the distribution of private health insurance prices (high prices versus low prices). The results show the public system's allocation rule and the probability of receiving health care from the public system both significantly affect an individual's WTP for private health insurance in the predicted direction, although the WTP values tend to be above the actual theoretical predictions. When the public system allocates health care based on need, the average WTP is lower than under random allocation. A higher probability of receiving health care from the public system elicits a lower WTP regardless of how the public system allocates health care. It is also found that when the public system allocates health care based on need, the WTPs are significantly higher under a neutral frame than a health frame.