PI: Prices and Inequality

虽然财富不平等和收入不平等已经得到了广泛的研究，但人们对价格对不平等的影响知之甚少。PI将把价格和消费作为不平等研究的核心，分三步进行：首先，PI将在消费者偏好随收入变化时制定新的价格指数，这将构成该项目其余部分的理论基础，并可供统计机构今后采用。风险很高：新的非相似价格指数将改变对不平等以及长期收入增长和生活水平跨国比较的衡量。第二，PI将利用“大数据”为政策制定者提供通货膨胀不平等的全面情况，方法是：（i）收集国际信用卡数据数据库，以便计算20个国家的通货膨胀不平等，（ii）收集涵盖美国所有重要部门的微观价格和支出数据集，（iii）使用机器学习技术对服务进行新的享乐回归，以及（iv）估计新的住房通胀指数，以解决现有的偏差。这些新的数据集和技术将通过一个项目网站与科学界分享，以最大限度地扩大影响。第三，PI旨在通过放松当前范式的假设，通过将经济的生产方和内生价格重新纳入最优税收政策设计，带来公共经济学理论的革命。这种方法强调了生产和消费之间的相互作用以及不同通货膨胀率的重要性。PI第二部分的综合微观数据集将用于价格关联模型的学科。因此，PI远远超出了测量，并将产生改进的价格关联模型和最优政策。

While wealth inequality and income inequality have been studied extensively, little is known about the impact of prices on inequality. PI will put prices and consumption at the core of the study of inequality, proceeding in three steps.First, PI will develop new price indices when consumer preferences change with income, which will constitute the theoretical backbone for the rest of the project, and which could be adopted by statistical agencies going forward. The stakes are high: the new non-homothetic price index will change the measurement of inequality as well as of long-run income growth and cross-country comparisons of living standards. It will also have practical implications such as the indexation of the poverty line and better targeting of social transfers.Second, PI will leverage "big data" to provide policymakers with a comprehensive picture of inflation inequality by (i) gathering an international database of credit card data allowing to compute inflation inequality in twenty countries, (ii) collecting micro price and expenditure datasets covering all important sectors in the United States, (iii) using machine learning techniques to run new hedonic regressions for services, and (iv) estimating a new housing inflation index to address existing biases. These new datasets and techniques will be shared with the scientific community via a project website to maximize impact.Third, PI aims to bring about a revolution in public economics theory by relaxing the assumptions of the current paradigm, by incorporating the production side of the economy and endogenous prices back into optimal tax policy design. This approach highlights the interaction between production and consumption and the importance of heterogeneous inflation rates. The comprehensive micro datasets from the second part of PI will be used to discipline models of price incidence. PI thus goes far beyond measurement and will yield improved models of price incidence and optimal policy.