Stats NZ has a new website.

For new releases go to

www.stats.govt.nz

As we transition to our new site, you'll still find some Stats NZ information here on this archive site.

  • Share this page to Facebook
  • Share this page to Twitter
  • Share this page to Google+
New Zealand's involvement in the Purchasing Power Parity Programme

Abstract

New Zealand is one of 45 countries participating in the OECD-Eurostat Purchasing Power Parity (PPP) Programme, part of the World Bank’s wider International Comparison Programme (ICP).

Aggregate gross domestic product (GDP) and GDP per capita are frequently used to assess, respectively, a country’s economic size and the well-being of its residents. To make meaningful inter-country comparisons of GDP, prices must first be converted into a common currency, and so a robust method of comparison and conversion of prices must be used. This is where purchasing power parities (PPPs) come in.

This paper discusses the measurement and calculation of PPPs, their uses and limitations, the ICP, and New Zealand’s involvement in the PPP Programme.

top

1. Introduction

The measure of GDP per capita within a country is a significant measure of the economic well-being of its residents. It is usually expressed in terms of the prices and currency units of each country. When comparing across countries, GDP values should first be converted into a common measure of currency unit and price before any valid conclusions can be drawn.

Exchange rates are a simple way of converting different countries’ GDP values into a common currency unit. This method faces some criticisms related to how suitable exchange rates are for converting prices of goods and services between countries. One criticism is that exchange rates tend to explain prices for traded goods more than prices for services and non-traded goods. Another comment is that market exchange rates are affected by factors that may not directly affect the well-being of a country’s residents, such as the purchase and sale of contracts, which speculate on future currency movements. Others say that the volatility of market exchange rates is more unpredictable than the experiences of consumers of their level of wellbeing.

Purchasing power parities, or PPPs, provide one way of converting different countries’ GDP measures into a common currency unit and price. PPPs are a measure of the relative domestic prices of the components that form GDP in each country. Because prices are being directly observed, rather than inferred through exchange rates, more meaningful comparisons can be made between the volumes of goods and services purchased in different countries. The well-known ‘Big Mac Index’, published in The Economist magazine, is an example of a PPP comparison. The Big Mac is a suitable product for comparison because it is similar in quality between countries and it can be purchased in about 120 countries. The PPP programme follows similar principles but is on a much larger scale and yields more robust results.

This paper discusses:

  • the measurement of PPPs in more detail (section 2)
  • the calculation of PPPs (section 3)
  • the uses and limitations of PPPs (section 4)
  • the ICP Programme (section 5)
  • New Zealand's involvement in the PPP Programme (section 6).

top

2. Measurement

Measuring PPPs is a meticulous exercise that involves comparing (between countries) prices of about 3,000 goods and services that contribute towards expenditure on GDP. PPP numbers are price relatives – one country’s price divided by another's.

The price collection of goods and services are made over a three-year period. Results from each country are then compiled by the OECD. The PPP figures remove the effects of different price levels between countries to compare aggregate GDP and GDP per capita across countries for a particular year. As the comparison is spatial in nature, the effects of changing prices over the three-year period within a country are removed by utilising temporal price indexes supplied by the respective countries.

Each country participating in the programme is asked to provide sets of national annual average prices (or a close representation – usually prices from a country’s capital city at a specific time when prices are expected to be reasonably stable) so that prices are not higher or lower than what a country’s citizens usually experience. Countries also provide national expenditure weights for PPP calculations. The goods and services chosen for the survey have to meet certain criteria. PPPs are supposed to be comparisons between countries of the same goods and services, and so a rigid specification of items is necessary to ensure very similar goods and services are priced, although some flexibility is necessary too, as the same brands and styles may not be available in each OECD country.

In order to provide flexibility around pricing some goods and services, the OECD provides specifications that refer to high-end, middle-range or low-end brands in markets, referred to as ‘brand strata’. These inform what brands are acceptable for price collection.

Goods and services in PPP surveys are selected by Eurostat and the OECD, informed by the Classification of Individual Consumption According to Purpose (COICOP – a classification system of goods and services) and by input from countries involved in the PPP programme. The COICOP system aids in maintaining the comparability of goods and services across different countries, but the selection will be slightly restricted by a need to price goods and services that are also representative of individual countries’ economies. In addition, all prices should be at market transaction prices paid by purchasers, while goods and services should be priced at similar outlets. This is because, for example, specialty stores often have different price levels from larger, general stores.

For calculation purposes, countries are asked to specify which priced items are representative of the expenditure on GDP in their particular country. This is because even though a wide range of items are priced, their representativity within different countries may differ. Countries attempt to collect prices for all items, but one or more items are then designated as being representative of the basic heading by each particular country.

Altogether, there are six price surveys in the PPP programme that are staggered over a three-year collection cycle. These surveys cover:

  • food, drinks and tobacco
  • personal appearance
  • house and garden
  • transport, restaurants and hotels
  • services
  • furniture and health.

Some surveys’ names are self-explanatory, although the items in other surveys might require further explanation. Items surveyed for personal appearance include goods such as clothing, footwear and jewellery. The house and garden survey covers consumer durables such as whiteware, audio and video equipment, goods used in the maintenance of the house and garden, and other miscellaneous household expenses.

Transport, restaurants and hotels relate to goods and services such as the purchase, running and maintenance of motor vehicles, dining, and accommodation. Services covers repairs, transport services, education, and energy usage.

Other surveys not covered by store-bought goods and services are equipment goods (such as machinery), construction (such as houses and roads), government services (such as health and education), and accommodation rents1.

National expenditure weights are also provided for calculation purposes, discussed in more detail in the next section.

top

3. Calculation

The calculation of PPPs at the aggregate GDP level involves two broad stages of aggregation. The first stage is across products (that is, individual goods and services) within each basic heading to derive PPPs at the basic heading level. A country-by-product matrix of national annual average prices is used to calculate the PPPs as geometric means of price relatives at the basic heading level between pairs of countries.

The second stage of aggregation is across basic headings to derive PPPs at the aggregate GDP level2. A country-by-basic-heading matrix of PPPs (which was calculated at the first stage) and a country-by-basic-heading matrix of national expenditure weights are used to calculate PPPs at the aggregate GDP level as expenditure-weighted arithmetic means between pairs of countries of the PPPs from the first stage.

The calculation of PPPs incorporates the Eltetö-Köves-Szulc (EKS) method at both stages of aggregation. The EKS method is a calculation method using Laspeyres-, Paasche-, and Fisher-type indexes3. Put simply, the Laspeyres and Paasche indexes are calculated to compare the prices of items in two countries. First, the Laspeyres-type index can be calculated using the items that are representative in one country (termed the base country).

The Paasche-type index is similar to the Laspeyres, but the index is calculated using the representative items in the other country (termed the partner country). A Fisher index is simply a geometric mean of the Laspeyres and Paasche indexes – that is, the square root of the Laspeyres and Paasche indexes multiplied together.

The EKS method also ensures transitivity is maintained between countries (similar to exchange rates), where a direct comparison between New Zealand and Australia yields the same result as an indirect comparison.

In the EKS method, it is not a requirement that every good and service listed in the surveys is priced. It is required, however, that at least one item be priced and marked 'representative' under each basic heading for countries to be compared at that level.

top

4. Uses and limitation of PPPs

PPPs are used to convert GDP expenditures into common currency units and prices. The results are then used to compare GDP figures across countries. While exchange rates can be used to convert GDPs from different countries to a common currency unit, they do not expressly convert GDPs into a common price. This is where PPPs come in, as they can fulfil both requirements for inter-country comparisons of a common currency unit and a common price. These measures are used by universities, economists, journalists, and international organisations like the World Bank, OECD, Eurostat and the United Nations. The European Union also uses PPPs in determining as to how to allocate structural funds to regions with low GDP per capita4.

PPP statistics are commonly used for making GDP comparisons between countries in terms of aggregate GDP (used to compare the sizes of economies) and of GDP per capita. The GDP per capita informs about the differences in well-being between the citizens of different countries, assuming that more goods and services are preferred to fewer. Countries can then be ranked according to their per capita GDPs. The rankings are split into four ‘income’ groups (high, high-middle, low-middle and low income). It is important to note when discussing the rankings that small differences between countries’ figures should not be over-emphasised, because PPP estimates broadly represent price levels and contain some sampling and non-sampling errors. Instead countries should be compared based on the income groups they fall into.

Exchange rates can be compared with PPPs, to arrive at a ratio regarded as the purchasing power of a currency. In comparison with another country, if New Zealand's exchange rate was equal to its PPP – yielding a ratio of one – then it could be said that on average, one unit of New Zealand currency purchases the same amount of goods and services in New Zealand as it would in the other country (after being converted into foreign currency). If New Zealand's PPP was greater than its exchange rate, it could be said that one unit of New Zealand currency purchases fewer goods and services in New Zealand (due to a higher comparative price level in New Zealand) than it does in the other country.

If New Zealand's PPP with the other country was lower than its exchange rate, it could be said that one unit of New Zealand currency purchases more goods and services in New Zealand (due to a lower price level in New Zealand) than in the other country. It should be noted that this ratio is not a measure of the over- or under-valuation of a currency, but a measure of the purchasing power of the currency.

Another use of PPPs is to compare labour productivity between countries. Labour productivity is a measure of output (in terms of volume of goods and services) to input (in terms of labour). A measure of GDP (output) per person employed (input) can be constructed using employment data and then compared across countries using PPPs.

Due to the limitations of PPPs, such as their design to be aggregates and their resulting limited precision below this level, the OECD has recommended an approach for using PPPs based on their suitability for particular roles. A summary appears below.

Areas where PPPs are well placed:

  • volume comparisons of GDP:
    • GDP per capita
    • GDP per hour worked
    • size of economy
  • comparisons of relative price levels.

Areas where PPPs may be used, with limitations:

  • cost of living index across countries
  • time series analysis of relative GDP per capita
  • analysis of price convergence.

Areas where the use of PPPs is not recommended:

  • as an indicator for the over- or under-valuation of a currency
  • as a precision tool to establish rankings between countries
  • as a measure to generate output and productivity comparisons by industry (unless there are industry-specific PPPs)
  • as a way of constructing national growth rates
  • as a ‘volume’ measure of exports or imports across countries.

top

5. International Comparison Programme

The International Comparison Programme (ICP) was established in 1968 to enable PPP comparisons between countries. First used between 10 countries, the ICP today has about 150 participating nations that are split into six main regions (Africa, Asia and the Pacific, Western Asia, Latin America, the Commonwealth of Independent States, and the OECD-Eurostat countries). The programme is a major statistical undertaking overseen by the World Bank. The OECD’s role is to oversee the PPP programme in 45 OECD countries. PPP results are combined between the World Bank and OECD-Eurostat to derive comparisons between approximately 150 countries.

There are some differences between the OECD-Eurostat’s and the World Bank’s collection of PPP data. The OECD’s PPP programme collects survey data and reports results over three-year periods, while the ICP’s data is collected every three to five years, depending on the region, and reports are made over longer periods. Due to the larger number of countries involved in the ICP and the significant differences in their economies, the characteristics and amount of data collected varies more compared with data collected from the OECD’s PPP programme. The ICP introduced reduced information surveys in the 1993 round of data collection to minimise the burden on those statistical offices with limited resources. This survey divided GDP into a list of about 200 items for these offices (compared with over 1,000 items for most regions).

Because the ICP comparison involves a large number of countries with dissimilar economies, the problem of designing PPP surveys for goods and services (that are both comparable between countries and representative of expenditure on goods and services that comprise GDP in each country) becomes more pronounced. A method of calculating PPPs was developed for the 2005 ICP round, which uses ‘ring countries’ to link the different world regions surveyed in the ICP. This involves selecting multiple countries within each region to participate in a global PPP survey, which means some countries are involved in two sets of surveys. These countries' economies should be representative of the regions they belong to, and they should have a wide range of goods and services that are also available in countries from other regions5.

top

6. New Zealand’s ongoing involvement with the PPP programme

Data collection

The PPP survey is similar to New Zealand’s price collection for other price indexes. However, there are differences that should be noted. In the PPP survey, there are about 200 basic headings under the COICOP classification6. New Zealand’s Consumers Price Index (CPI), which supplements some of the PPP pricing, contains about 105 equivalent ‘classes’. Because the nature of the CPI is temporal and national, there is usually a range of goods and services priced, representative of what consumers purchase in New Zealand.

The Producers Price Index (PPI) is another temporal price index which may overlap with PPPs, especially in the equipment survey round. PPPs are comparisons between countries at a point in time, so it is necessary to ensure that the goods and services priced are of the same quality as in the specification. Because of these differences, there is some overlap between the collection of prices for temporal, national price indexes and PPPs, but not enough to fully provide the price collection for PPPs.

The main issues relating to pricing PPPs for New Zealand are:

  • that goods and services are primarily representative of economies in Europe
  • determining outliers among collected prices
  • determining representativity.

How goods and services reflect European economies is covered in more detail in the next section under 'Statistics NZ and the Australian Bureau of Statistics (ABS)'. The issue of determining outliers comes up because gauging whether the collected prices are good representations of an item’s ‘normal’ price is not always straightforward, and requires a reasonable knowledge of markets where these items were purchased.

Determining the representativity of goods and services priced in the PPP surveys can be difficult because data for weighting items in price indexes does not always give the level of detail desired for PPPs. People with good knowledge of specific goods and services give their input, while use is also made of other sources, such as AC Nielsen, which is a compilation of barcode data from the sale of supermarket items. The data can be used to determine which brands and products are most representative within basic headings for some items priced in the PPP surveys. There is also brand and product data available from GfK (a market research organisation) which provides information relevant to determining representativity for household items (such as televisions).

Prices are intended to represent national annual average prices, and so those for clearance stock are not desirable in this regard. Prices are usually collected in the capital city of participating countries, because this allows countries to collect within the timeframes set by the OECD, and prices here are regarded as reasonable approximations at the national level. Some countries collect prices in multiple cities, while others publish their own regional spatial price indexes that can be used for estimating national prices if collection is made only in the capital city.

National expenditure weights and price indexes are also supplied by the statistical office. The expenditure weights are used for weighting purposes (for calculating aggregate PPPs), while the price indexes are used to convert prices to a base year. For example, if 2008 was a base year, and the price of a new car was collected in 2007, a temporal price index for new cars could be applied to estimate a 2008 price.

Friends of the Chair

New Zealand is one of 13 countries currently participating in the Friends of the Chair (FOC) working group. This was established to help ongoing reviews of the ICP, more specifically to evaluate the data being collected and its methods of collection. To evaluate the ICP, it is important to asses the quality and credibility of the data collected and how extensively they are used. In 2001, the FOC group established a plan to focus on issues surrounding the governance structure of the ICP, as well as its resources and methodologies.

Statistics NZ and the Australian Bureau of Statistics (ABS)

In October 2006, representatives from Statistics New Zealand and the Australian Bureau of Statistics (ABS) met to discuss the Purchasing Power Parity Programme in the context of the two nations.

Among the topics discussed was that goods and services priced for the PPP programme should be identical. However, because often what is popular and readily available in one country is not necessarily so in another, this theory often does not work out in practice and therefore some compromise is needed. Commodities for the PPP programme (such as food and clothing) tend to be representative of what is purchased in Europe, but not elsewhere, which means that some goods and services specified for pricing are not available in New Zealand and Australia.

Some products are available but are not representative of what is purchased in New Zealand. The OECD has introduced more flexibility in product descriptions ensuring a more representative sample. Because the New Zealand and Australian economies have much in common, they therefore have common bases of experience from which to discuss PPP surveys in some depth.

Tokyo 2007 meeting

In March 2007, the annual meeting on PPPs for non-European member countries of the OECD was held in Tokyo, Japan. One of the senior members of the National Accounts team at Statistics New Zealand attended.

There were four main topics of discussion at the meeting:

  • the presentation of the preliminary results for the 2005 PPP round (results were published in November 2007)
  • the inclusion of these results into the world-wide ICP results, which will further raise the visibility of the new benchmark results
  • the treatment of non-market services in the Eurostat-OECD PPP Programme and the use of output measures to improve the quality of PPPs in this area
  • the 2008 round – organisation of the work, timetable and a review of the interim results of the first surveys.

top

Footnotes

  1. OECD (Jun 2007). "Eurostat – OECD Methodological Manual on purchasing power parities (PPPs).”
  2. Schreyer P, Koechlin F (Mar 2002). “Purchasing power parities – measurement and uses.” Statistics Brief No.3.
  3. OECD (Feb 2003). "Glossary of Statistical Terms", http://stats.oecd.org/glossary/detail.asp?ID=5525
  4. OECD (Sep 2007). "The Eurostat-OECD PPP Programme and the ICP – A Shared Commitment", http://www.oecd.org/document/15/0,3343,en_2649_33715_34487759_1_1_1_1,00.html
  5. The World Bank (Sep 2007). "Ring Comparison for Global PPPs", http://go.worldbank.org/I13AJC2HS0
  6. OECD Secretariat (Oct 2001). “On reducing the number of basic headings and products priced”.

For more information about the PPP Programme, please contact:

Cary Milkop
Wellington 04 931 4600
Email: info@stats.govt.nz

Back to Price Index News: January 2008

  • Share this page to Facebook
  • Share this page to Twitter
  • Share this page to Google+
Top
  • Share this page to Facebook
  • Share this page to Twitter
  • Share this page to Google+