You are here: Home / Publications / Income Growth, Price Variation and Health Care Demand: A Mixed Logit Model Applied to Tow-Period Comparison in Rural China

Income Growth, Price Variation and Health Care Demand: A Mixed Logit Model Applied to Tow-Period Comparison in Rural China

Martine, Audibert; Jacky, Mathonnat; & Yong, H. E. (2010). Income Growth, Price Variation and Health Care Demand: A Mixed Logit Model Applied to Tow-Period Comparison in Rural China. CERDI, Working Papers: 201035. CERDI.

Martine, Audibert; Jacky, Mathonnat; & Yong, H. E. (2010). Income Growth, Price Variation and Health Care Demand: A Mixed Logit Model Applied to Tow-Period Comparison in Rural China. CERDI, Working Papers: 201035. CERDI.

Octet Stream icon 665.ris — Octet Stream, 2 kB (2234 bytes)

1989-2006 is a period of the start and the end of deregulation of Chinese health care sector and of disintegration of rural cooperative insurance system. During this period, the government health policy has turned healthcare providers all alike into profit seeking entities. Face to perverse effects, by 2003, Chinese government begun to restore rural cooperative insurance system. From CHNS data source, we constitute two samples: 89-93 and 04-06 with respectively 2117 and 2594 rural patients surveyed roughly in the same villages in 9 Chinese provinces to compare their health choice behaviors with the evolution of price, income, distance, insurance, age, and regional inequality. Using Mixed Multinomial Logit (MMNL) estimations, we have obtained three main results. First, even in both periods there is clear price effect, in 04-06 it tends to be weaker, and heterogeneity in price preference has increased. This corresponds well the fact that between the two periods price level has significantly increased and price variation reduced. Second, there is a stronger negative distance effect and heterogeneity in 2004-06, while in 89-93 this negative impact was lower and absent for providers farther than 10km. One interpretation is the existence of a substitution effect: when patients have less possibility to discriminate providers by price, they increase their preference in choice by distance. Third, while, wealth effect exists in some choices in 89-93, it becomes absent in 04-06. Explanations may be that one the one hand both supply side and demand side conditions on health care have been improved even, to less extent though, for the poor, and on the other hand, health care is necessary goods and is price inelastic. But meanwhile, we observed catastrophic effect for the poor: the poorer patients have their share of consumption in income more decreased after health care.




RPRT

CERDI, Working Papers: 201035


Martine, Audibert
Jacky, Mathonnat
Yong, H. E.



2010









CERDI






665