New results from the ICP including purchasing power parities (PPP), price level indexes and PPP-based expenditures for reference year 2017 are now available at icp.worldbank.org. This blog series, edited by Edie Purdie, covers all aspects of the ICP and explores the use made of these data by researchers, policymakers, economists, data scientists and others. We encourage users to share their data applications and findings in this blog series via email@example.com.
The International Comparison Program (ICP) is one of the world’s largest and most complex statistical initiatives. It is implemented as a global partnership of national and regional agencies and managed by the ICP Global Office at the World Bank, under the auspices of the United Nations Statistical Commission (UNSC). The main objective of the ICP is to estimate purchasing power parities (PPPs) and price level indexes for the world’s economies. PPPs are similar to market exchange rates in that they convert national currencies into a common currency, most often to the US dollar. However, PPPs differ from market exchange rates by controlling for differences in price levels between economies, thereby equalizing the purchasing power of the resulting currency conversions. In doing so, PPPs enable cross-country comparisons of gross domestic product (GDP), and its expenditure components, which reflect only differences in the volume of economic outputs, unlike market exchange rate-based comparisons, which reflect both volume and price differences. Price levels vary considerably between higher- and lower-income economies and thus market exchange rate-based comparisons overstate the size of economic output where price levels are high and understate output where they are low.
The recent release of ICP results for the reference year 2017 stimulated a familiar set of questions regarding the ICP’s outputs, such as those posed in a recent article by the Economist. These queries typically ask how the ICP ensures comparability—and at the same the representativity—of the price data it collects, how it accounts for different purchasing patterns and behaviors across the globe, and how it achieves an appropriate weighting of goods and services priced by the participating economies. In this blog we explain how a robust and well-established ICP methodology addresses these challenges, and how the
Overall, the ICP methodology has three major elements. The first is the ICP Classification of Expenditures which relies on System of National Accounts (SNA) definition of final expenditures on the GDP. The second is the basket of goods and services from which items are selected for pricing. The third element is the estimation of PPPs for each participating economy. We discuss these elements in the paragraphs below.
The ICP Classification of Expenditures
The ICP Classification of Expenditures ensures that the economies compile their national accounts expenditures using the same definition of GDP and its expenditure components, and adhere to the internationally agreed concepts, definitions, classifications, and accounting rules of the 2008 SNA. This conceptual framework warrants that all economies are equally exhaustive in their measurement of economic activity.
The ICP Classification is based on a number of international classifications, such as the Classification of Individual Consumption by Purpose (COICOP), the Classification of the Purposes of Non-Profit Institutions Serving Households (COPNI), the Classification of the Functions of Government (COFOG) and the Statistical Classification of Products by Activity (CPA).
GDP is broken down into six main aggregates in the ICP Classification: individual consumption expenditure by households; individual consumption expenditure by nonprofit institutions serving households (NPISHs); individual consumption expenditure by government; collective consumption expenditure by government; gross capital formation; and the balance of exports and imports. These main aggregates are further broken down into 28 expenditure categories, 63 expenditure groups, 126 expenditure classes, and 155 basic headings. The basic heading level, the lowest level in the ICP classification, is important as it is at this level national accounts expenditures are defined and estimated, goods and services are selected for pricing, and PPPs are first calculated.
The chart below visualizes the ICP classification according to both the number of basic headings within each aggregation and the average global expenditure weight as a percentage of the GDP applied to each. Filter the view and click the pie pieces to explore the data further.
Note: For the purpose of the visualization, the total number of basic headings in this chart is 154 as the two basic headings, (a) Exports of goods and services and (b) Imports of goods and services are combined as one basic heading called “balance of exports and imports.” The average expenditure weights of basic headings with negative expenditure values are not shown in this chart.
The ICP basket of goods and services for the GDP
The ICP’s basket of goods and services is based on the ICP classification. The basket is developed collaboratively by the national, regional and global implementing agencies, and it builds upon baskets from previous ICP comparisons in order to ensure temporal consistency across the periodic comparisons.
The 2017 ICP basket of goods and services contained 940 items at the global level, representing four of the main aggregates, 16 categories, 39 groups, 79 classes and 98 basic headings in the ICP Classification. For the remaining headings PPPs were estimated through “borrowing” PPPs from similar headings or using other proxies. Taken together, the ICP basket covers the full range of economic activities under the GDP.
The chart below visualizes the basket of global items within each aggregation level.
Basket of goods and services for Household Consumption
The aggregate Individual Consumption Expenditure by Households reflects the goods and services consumed by households and it encompasses expenditures on food and clothing through to transport, health and education. This aggregate typically carries a significant expenditure weight within the GDP.
In order to ensure the representativeness of household consumption baskets across the regions and participating economies, each region within the ICP comparison selects a set of global and regional items to be priced. Global items ensure sufficient overlap across the regions, while regional items—identified as those items typically consumed by households of economies within a region—ensure that the baskets are locally representative.
The ICP basket for household consumption also strives to include items from the participating economies’ national consumer price index (CPI) baskets, to improve the representativity of the ICP basket and harness synergies and realize efficiencies between ICP and CPI data collections.
Definitions of goods and services
In order to ensure the comparability of prices, and hence the reliability of the resulting PPPs, it is crucial that all participating economies price comparable goods and services. To achieve this, each item in the ICP basket is defined by a structured product definition (SPD), which describes every main price-determining characteristic of a particular good or service in a clear and systematic manner.
Ensuring that participating economies are able to collect prices for nationally representative goods and services is also important. To accomplish this, the ICP basket contains three types of SPDs: i) SPDs for goods where the selection of brand or label is not pre-defined, or a brand is not considered to be price-determining; ii) SPDs for goods where a certain brand, and possibly model, to be priced is pre-defined; and iii) SPDs for services. The second category typically covers pharmaceutical products, motor vehicles, electronics and industrial machinery, for which brand and model details are considered to be key price-determining characteristics. Specifying the brands and models for these items ensures that the prices collected for them across economies and regions are comparable.
For the global ICP 2017 household consumption basket, as an example, SPDs without pre-defined brands or labels made up 67% of the total basket, those with pre-defined brands 13%, and SPDs for services 20% of the total basket.
Estimation of PPPs
PPPs are first estimated at an elementary level for a similar group of items within the basket, that is, at the basic heading level. These basic heading PPPs are subsequently aggregated using national accounts expenditures as weights, in order to arrive at appropriately weighted PPPs for each participating economy and for all levels within the ICP classification.
The resulting PPPs are used to deflate the same national accounts expenditures to provide comparable cross-country volume estimates of output.
This blog provides an overview of the ICP methodology which has evolved over several decades through the collaboration of national and regional implementing agencies with the ICP Global Office at the World Bank and drawing on the guidance and counsel of world-renowned technical experts. The forthcoming 2021 ICP cycle will apply the same high-quality processes used in the 2011 and 2017 cycles, ensuring temporal comparability alongside the extensive spatial comparability of results. The Program’s outputs of comparable economic estimates largely meet the needs of a wide range of policymakers, economists, statisticians, analysts, academia and other users.
More information on the ICP approach and methodology is available through the ICP 2017 report, the ICP Operational Guidelines, Measuring the Real Size of the World Economy: The Framework, Methodology, and Results of the International Comparison Program (ICP), as well as through the ICP website. ICP results and data are available through the World Bank’s DataBank.