Collaborative Proposal: Integrating Expenditure and Production Estimates in International Comparisons

我们研究的一个领域有望对使用单位价值进行各国价格和购买力平价（PPP）比较的概念和方法问题做出重大贡献。国际贸易和生产数据通常用于通过将生产、出口或进口的类别的价值除以各自的数量来估算价格。数量本身是相当详细的，但仍然是异质的，通常是将不同容量和/或不同质量的泵组合成一个数量。概念上的问题是，我们不知道某些国家的价格是否会更低，因为商品的数量和/或质量较低。这项研究通过使用详细的海关数据，同时观察各国对高质量商品的需求和供应，解决了贸易数据中的这一问题。加州大学戴维斯分校（University of California/Davis）国际数据中心（Center for International Data）开发的初步结果表明，这是解决单位价值贸易数据使用基础问题的一种令人兴奋的方法。此外，基础数据集将对该领域的研究人员有价值。这项研究的另一个非常重要的方向是在一致的基础上使用购买力平价综合生产和支出方面的估计。到目前为止，支出方面的比较比较容易进行，作为佩恩世界表（PWT）的一部分，从1950年到2004年的部分或全部时间，涵盖了180多个国家。然而，这些估计只允许在GDP水平上对生产率进行粗略的比较，而大多数兴趣集中在零售贸易、IT等部门。在国内生产总值水平上，根据上述单位价值研究对贸易条件进行调整，可以大大改善对各国总产出的估计。格罗宁根增长与发展中心（GGDC）对许多国家的生产部门进行了比较。我们的目标是将全球生产总值估算值与尽可能多的PWT国家的估算值联系起来，这将在方法上和各国生产率比较方面做出重大贡献，从而使该项目具有重大的全球研究影响。随着上述研究的进展，我们希望在国际比较中形成贸易、支出和生产流动之间的进一步闭合循环。这里的一个主要障碍是，进入一个国家国际贸易的商品价格往往与该国生产的商品价格大不相同。因此，即使一个人通过比较出口和进口的价格解决了各国之间的质量问题，这也不一定会产生两国之间的比较，就像比较国内生产一样。像中国这样的国家，出口各种高质量的产品，也生产国内版本的低质量和不同的组合。这一重要问题将是作为本项目一部分计划的讲习班的一个主要主题。世界银行正在协调2005年最新的详细基准国际比较项目（ICP）比较。将这些估计值纳入PWT将是对这一广泛使用的数据集进行重大升级的一部分。这将是介绍作为正在进行的研究的一部分出现的一些方法变化的机会，这些变化有些是由我们合作项目的成员进行的，有些是由独立研究人员进行的，有些是在编写国际比较方案手册时进行的。

One area of our research promises to make a major contribution to conceptual and methodological problems in the use of unit values to make price and purchasing power parity (PPP) comparisons across countries. International trade and production data are often used to approximate prices by dividing values of categories produced, exported or imported by their respective quantities. The quantities themselves are at a fairly detailed level but still are heterogeneous, often combing say, pumps of different capacities and/or different qualities into one quantity. The conceptual issue is we do not know if the prices may be lower in some countries because of the mix of items in the quantity and/or the lower quality of the goods. The research tackles this problem for trade data by simultaneously looking at the demand and supply of higher quality items by country using detailed customs data. Preliminary results developed by the Center for International Data at the University of California/Davis suggest this is an exciting approach to a problem that is fundamental to the use of unit value trade data. In addition, the underlying data sets will be of value to researchers in the area.Another very important direction of this research is related to the integration of production and expenditure side estimates on a consistent basis using PPPs. Up until now expenditure side comparisons, which are simpler to carry out, cover over 180 countries for some or all the years 1950 to 2004 as part of the Penn World Table (PWT). However, these estimates only allow approximate comparisons of productivity at the GDP level, whereas most interest attaches to sectors like retail trade, IT, and the like. At the GDP level, the adjustments of the terms of trade that flow from the unit value research described above permit very much improved estimates of aggregate production across countries. Comparisons of productive sectors have been carried out by Groningen Growth and Development Center (GGDC) for a significant number of countries. It is our goal to link the GGDC estimates to those of as many countries as possible in PWT, which would be a major contribution both methodologically as well as for comparisons of productivity across countries, thereby giving this project a major global research impact.As the above research progresses we hope to formulate a further closing of the circle between trade, expenditure and production flows in international comparisons. A major hurdle here is that the prices of goods entering international trade of a country are often quite different from the prices of goods produced in a country. So even if one has solved the quality program across countries in terms of comparing prices of exports and imports, this will not necessarily produce the same comparison between two countries as would a comparison of domestic production. A country like China, which exports a variety of high quality products, also produces domestic versions of those items of lower quality and different mix. This important issue will be a major subject of the workshops planned as part of this project.The latest detailed benchmark International Comparison Project (ICP) comparisons for 2005 are being coordinated by the World Bank. Integration of these estimates into PWT will be part of a major upgrade of this widely-used data set. It will be an opportunity to introduce a number of methodological changes that have emerged as part of ongoing research, some by members of our collaborative project, some by independent researchers, and some developed in conjunction with the preparation of an ICP handbook.