Contoyannis Paul, Hurley Jeremiah, Grootendorst Paul, Jeon Sung-Hee, Tamblyn Robyn
Centre for Health Economics and Policy Analysis, McMaster University, Canada.
Health Econ. 2005 Sep;14(9):909-23. doi: 10.1002/hec.1041.
The price elasticity of demand for prescription drugs is a crucial parameter of interest in designing pharmaceutical benefit plans. Estimating the elasticity using micro-data, however, is challenging because insurance coverage that includes deductibles, co-insurance provisions and maximum expenditure limits create a non-linear price schedule, making price endogenous (a function of drug consumption). In this paper we exploit an exogenous change in cost-sharing within the Quebec (Canada) public Pharmacare program to estimate the price elasticity of expenditure for drugs using IV methods. This approach corrects for the endogeneity of price and incorporates the concept of a 'rational' consumer who factors into consumption decisions the price they expect to face at the margin given their expected needs. The IV method is adapted from an approach developed in the public finance literature used to estimate income responses to changes in tax schedules. The instrument is based on the price an individual would face under the new cost-sharing policy if their consumption remained at the pre-policy level. Our preferred specification leads to expenditure elasticities that are in the low range of previous estimates (between -0.12 and -0.16). Naïve OLS estimates are between 1 and 4 times these magnitudes.
处方药需求的价格弹性是设计药品福利计划时一个至关重要的参数。然而,使用微观数据估计弹性具有挑战性,因为包含免赔额、共同保险条款和最高支出限额的保险覆盖范围会产生非线性价格表,使得价格具有内生性(是药品消费的函数)。在本文中,我们利用加拿大魁北克省公共药物保险计划中成本分担的外生变化,采用工具变量法来估计药品支出的价格弹性。这种方法纠正了价格的内生性,并纳入了“理性”消费者的概念,即消费者在做出消费决策时会考虑到根据预期需求在边际上面临的预期价格。工具变量法改编自公共财政文献中用于估计税收政策变化对收入反应的一种方法。该工具变量基于如果个人消费保持在政策实施前的水平,他们在新的成本分担政策下将面临的价格。我们首选的设定得出的支出弹性处于先前估计的较低范围(在-0.12至-0.16之间)。简单的普通最小二乘法估计值是这些幅度的1至4倍。