university ofgroningen

groningen growth anddevelopment centre

GGDC RESEARCH MEMORANDUM 151

Cross-country income levels over time: did the developing world suddenly

become much richer?

Robert Inklaar and D.S. Prasada Rao

December 2014

Cross-country income levels over time: did the developing world suddenly

become much richer?

Robert Inklaar

Groningen Growth and Development Centre University of Groningen, The Netherlands

D.S. Prasada Rao

School of Economics The University of Queensland, Australia

December 12, 2014

A previous version of this paper was circulated under the title “Cross-country income levels of time: can ICP 2005 and ICP 2011 be reconciled?” The authors would like to thank Angus Deaton, Arvind Subramanian, Marcel Timmer and seminar participants at the University of Graz and the University of Groningen for helpful comments and the World Bank for providing detailed researcher data underlying ICP 2005 and ICP 2011. This work was undertaken when Rao was a visiting researcher at the Groningen Growth and Development Centre in August, 2014 and at UNU-MERIT during October, 2014. Correspondence to: Robert Inklaar, University of Groningen, Faculty of Economics and Business, PO Box 800, 9700 AV Groningen, The Netherlands (R.C.Inklaar@rug.nl).

2

Abstract

The latest global survey on relative prices and income levels showed revisions to income

levels that were larger in lower-income countries, thereby shifting the world income

distribution. The aim of this paper is to establish whether changes in measurement

methodology and price sampling methods between the latest price survey for 2011 and the

previous survey for 2005 can explain these large differences. We construct a counterfactual

set of relative prices for 2005 that harmonizes measurement and we find that the systematic

differences in income levels are substantially reduced or even eliminated, implying that

international income inequality had been overstated.

Key words: purchasing power parities; International Comparison Program (ICP); extrapolation; price levels; price sampling bias; real incomes; inequality

JEL Codes: C43, E01, E31, C83, O57

1

1. Introduction

Our understanding of differences in living standards across countries, international income

inequality and global poverty relies heavily on the periodical global price comparisons of the

International Comparison Program (ICP).1 Without ICP relative prices, comparisons of

income levels across countries would have to rely on exchange rates, which tend to

systematically underestimate the purchasing power of consumers in low-income countries.2

But despite the conceptual appeal of ICP relative prices, they are often also a source of

controversy. For instance, Deaton (2010) showed how results from consecutive ICP rounds

led to upward revisions in income inequality across countries. Similarly, Ciccone and

Jarociński (2010) and Johnson et al. (2013) show how various results from the cross-country

growth literature change depending on the vintage of relative income data that is used.3 More

broadly, changing the methods and model used to measure relative prices can have a notable

effect on the resulting relative income estimates, as shown by Neary (2004), Almås (2012)

and Feenstra et al. (2013).

The most recent source of controversy is the publication of ICP 2011 (World Bank, 2014a,b).

Compared with earlier estimates based on (extrapolations from) ICP 2005 (World Bank,

2008), the average country saw its income and consumption levels increase by about 25

percent. Revisions for developing countries were particularly large, resulting in major

changes to the economic geography of the world and a downward adjustment to global

inequality. Given the size and systematic pattern of these revisions, they can have far-

reaching consequences, ranging from a different number and geographical distribution of the

world’s poor to a different view on income convergence or appropriate policy.4 It is thus no

surprise that the revisions following ICP 2011 have already been the topic of substantial

commentary and debate.5

1 See e.g. Anand and Segal (2008), Deaton (2010), Deaton and Heston (2010), Chen and Ravallion (2010a) and Milanovic (2012). See also Feenstra, Inklaar and Timmer (2013) on how the ICP results are a crucial input in the Penn World Table and World Bank (2013) for details on ICP. 2 The Balassa (1964)-Samuelson (1964) effect, whereby prices of non-traded products tend to be lower in lower-income countries. 3 Differences between data vintages in these studies reflect revisions to relative prices – the focus of our paper – and revisions to national GDP statistics – a source of differences we exclude. 4 See Chen and Ravallion (2010a) on the impact of ICP 2005 on poverty, (e.g.) Pritchett and Summers (2014) on growth and convergence and Aghion, Akcigit and Howitt (2014) on productivity levels and appropriate policy. 5 See Dykstra, Kenny and Sandefur (2014), Chandy and Kharas (2014), Deaton and Aten (2014) and Ravallion (2014).

2

Yet these early commentaries may be misleading because of the major changes to how

relative prices were measured in ICP 2011 compared with ICP 2005.6 A particular concern,

as identified by Deaton and Aten (2014), is the so-called linking of regions. In both ICP 2005

and 2011, prices are first collected and compared across the countries within a region and in

the second stage, the regions are linked to each other to allow for a global price comparison.

Some regions consist mostly of low- and middle-income countries (Africa, Asia-Pacific) and

some mostly of high-income countries (Eurostat/OECD), which means that changes to the

linking approach can shift the prices of lower-income countries relative to higher-income

countries. Since the linking approach was considerably improved and refined in 2011,7

Deaton and Aten (2014) conclude that the linking data and methods are a good starting point

for trying to understand why the results from ICP 2011 were so different from the ICP 2005

results.

The goal of this paper is to identify the impact of the 2011 methodological and price

sampling changes and provide better estimates of the revisions to relative prices and income

levels. To this end, we construct a counterfactual price comparison for 2005 using ICP 2011

methods and adjusting for biases in the sampling of the 2005 price data, in particular

regarding the linking of regions. We draw upon the early diagnostic work of Deaton and Aten

(2014), but provide a more comprehensive and quantitative assessment of the impact of the

measurement changes. Given the theoretical problems in measuring relative prices (Van

Veelen, 2002) and the practical challenges (Deaton and Heston, 2010), even the

counterfactual we construct will show some differences.8 But if those differences remain

systematically larger in lower-income countries, this calls into question the common practice

of using a single set of cross-country relative prices to estimate relative income levels across

many years,9 and the research results that depend on such data.

To construct our counterfactual price comparison for 2005, we draw upon the original data

and methods used in the compilation of ICP 2005 and ICP 2011, including detailed price and

expenditure data, and we go through the full compilation of PPPs as detailed in World Bank

(2008) and World Bank (2014b). In presenting our results, we focus on the implications for

6 See World Bank (2014b) for a list of major improvements in ICP 2011 methodology compared to ICP 2005. 7 See in particular World Bank (2013). 8 See also McCarthy (2013b) for a more practical discussion of why subsequent price comparisons may not be consistent with observed relative inflation patterns. 9 This is standard practice in the World Bank’s World Development Indicators, the Maddison Project Database (Bolt and van Zanden, 2013) and for parts of the Penn World Table (Feenstra et al., 2013).

3

GDP per capita levels, which are most relevant for research into cross-country economic

performance, and consumption per capita, which is used most often in the analysis of poverty

and inequality. We make two sets of adjustments to the original ICP 2005 comparison; in the

first set, we take all 2005 price data as given, but apply the ICP 2011 methods. In this

‘harmonized methodology’ counterfactual, the average upward revision to relative income

and consumption levels goes down from 23-24 percent to 17-18 percent, but especially for

consumption, revisions are still systematically larger in lower-income countries. In the

second set of adjustments, we correct for a number of biases in the price data collected in

2005. Most importantly, we find clear evidence that the price data used in linking the regions

in 2005 was based on a product list that was only representative in higher-income countries; a

possibility that, notably, Deaton (2010) called attention to. In the ‘sampling bias adjusted’

counterfactual, the average upward revision goes down to 9-12 percent; making both sets of

adjustments simultaneously reduces the average revision further, to 4-10 percent. For both

GDP and consumption, lower-income countries no longer show systematically larger upward

revisions. Furthermore, the root mean squared revision decreases from 34-35 percent to 17-

22 percent, which is similar to estimates of the standard error of relative prices.10

The results from this analysis have important implications. Most importantly, the difference

between the original ICP 2005 and ICP 2011 suggested that datasets based on extrapolating

GDP per capita levels from a benchmark comparison to earlier years were inherently biased.

This, in turn, could have had serious consequences for research based on such datasets, which

include the World Development Indicators of the World Bank, the Maddison database (Bolt

and van Zanden, 2014) and parts of the Penn World Table (Feenstra et al., 2013). Our success

in eliminating the systematic differences with our counterfactual price comparison suggests

that such a fundamental reconsideration is not necessary. Furthermore, one concrete outcome

of our analysis is two sets of price comparisons based on the same methods and comparable

data. This means they can be fruitfully used in combination, such as in the part of the Penn

World Table that presents relative price and income levels based on all available ICP

benchmarks. Finally, from a statistical point of view, we have demonstrated the usefulness of

backward revisions to relative prices, which is similar in spirit to the backwards revision of

National Accounts time series when introducing new methods.

10 See e.g. Deaton (2012) and Rao and Hajargasht (2014).

4

2. Basic framework

The basic aim of the ICP is to measure the overall price level P in a country j relative to any

other country k. Implicit in this aim is the idea that the law of one price does not hold,

because if it did hold we would only need to observe market exchange rates to put GDP or

consumption on a comparable basis. The presence of non-traded goods in an economy can be

one reason for the law of one price to fail, but also see Rogoff (1996) for other reasons. As a

result, a procedure is needed to estimate the economy-wide price level using data on prices p

and expenditures v for individual products i. One could take a welfare-theoretic approach and

aim to compare the cost-of-living, as in e.g. Neary (2004). However, as argued by Deaton

and Heston (2010), such an approach leads to questions that are extremely hard to provide

sensible answers on, such as whether tastes or identical or, if not, how best to envisage

comparing countries that do not only differ in price and income levels but also along

dimensions such as climate or religion. What is done in ICP can thus best be thought of as

comparing an index of the price level, not the cost of living.

If we choose a so-called superlative index for this purpose, we know from Diewert (1976)

that such an index will provide a second-order local approximation to arbitrary preferences.

Let us, for expositional purposes, assume the superlative Törnqvist index. The relative price

level of country j relative to k at time t can then be written as:

log Pjt P

kt 1

2 sijt s

ikt i log p

ijt p

ikt , (1)

where sij are the expenditure shares of product i, v

ijv

iji . An important feature of equation

(1) is that the relative price level between j and k depends on expenditure shares in both

countries.11 How will this relative price index evolve over time? The standard approach, used

in e.g. the World Bank’s World Development Indicators, is to assume the change in the

relative price level is equal to relative inflation. Let country inflation also be measured by a

Törnqvist index:

log Pjt log P

jt P

jt1 1

2 sijt s

ijt1 i log p

ijt p

ijkt1 (2)

11 In the more general, so-called multilateral indexes used in ICP, the price level between j and k also depends (to a lesser degree) on expenditure shares in all countries, see Diewert (2013).

5

Since the price level in equation (1) depends on both country’s expenditure shares, while

national inflation only depends on national expenditure shares, the change in the price level is

not exactly equal to relative inflation, as shown by Deaton and Aten (2014). If we assume

that expenditure shares do not change over time, we can write the change in the price level by

combining equations (1) and (2) as:

log Pjt P

kt log P

jt log P

kt 1

2 sij s

ik i log pijt log p

ikt . (3)

This equation shows that the expected change in PPPs is given by relative inflation – the first

term in brackets – and a term reflecting differences in expenditure patterns and (average)

relative price changes across products. If there are no differences in expenditure patterns, no

changes in relative prices or if differences in expenditure patterns are uncorrelated with

average changes in relative prices, this term will be zero. In the absence of detailed data

about product level inflation, log pijt , it is not feasible to implement equation (3) in full.12

To extrapolate PPPs from 2005 to 2011, we will therefore rely only on data about relative

inflation, which according to equation (3) should be a fair approximation on conceptual

grounds. On practical grounds, discrepancies may well occur if, for example, country-specific

products are used to estimate national inflation but (for lack of cross-country comparisons)

are not used in ICP. McCarthy (2013b) details a comprehensive set of practical

considerations, but for none of these would we expect a larger effect in lower-income or

higher-income countries.

Given this framework, we now turn to a comparison of ICP 2011 with extrapolations from

ICP 2005. For ICP 2011, we use the results as published in the Summary Report (World

Bank, 2014a) and specifically the purchasing power parities (PPPs) and expenditure levels

for GDP and final consumption expenditure (‘consumption’). For ICP 2005, we use the PPPs

as originally published in World Bank (2008).13 The combined data set is constrained mostly

by ICP 2005, which covered 146 countries. Argentina, Syria and Lebanon participated in ICP

2005 but not in ICP 2011 and Zimbabwe participated in ICP 2005 but the coinciding period

of hyperinflation led to unreliable results. This leaves us with a data set of 142 countries,

12 Experiments with inflation data for (at most) four products in the consumption bundle shows small relative price changes over the 2005-2011 period, which suggests that the second term in equation (3) will be small. However, more detailed data could in principle reveal larger changes, so definite statements are not possible. 13 Or, to be more precise, the ratio of the PPP over the exchange rate. This avoids having to explicitly control for changes in currency units, such as the adoption of the euro in a number of countries.

6

which is used in the remainder of the paper. To extrapolate GDP PPPs, we use GDP

deflators; to extrapolate consumption PPPs, we use the consumer price index (CPI).14

Let YjtP be the PPP for GDP (Y) in country j at time t based on a benchmark year and let

Yjtp be the GDP deflator in year t in country j. Whenever t the PPP is extrapolated from

the benchmark year as:

Yjt YjYjt Yj

Ybt Yb

p pP P

p p

(4)

where b refers to a base or reference country – usually the United States. We can then

compare GDP per capita (y) or household consumption per capita (c) based on the PPP

extrapolated from ICP 2005 with the corresponding numbers based on the new ICP 2011

benchmark. We use the following relative difference measure:

2011 20052011 2011 2011

2011 2015 20112011 2011 2011

1 1j Yj YjYj

j Yj Yj

y P Pd

y P P (5)

As all the subsequent comparisons are for the year 2011, we drop the subscript 2011. Figure

1 shows a plot of dY against the log of GDP per capita converted using exchange rate E,

ln ( )j jy E , and dC against the log of consumption per capita, ln ( )j jc E . The GDP and

consumption per capita numbers are those reported in World Bank (2014a).

The figure illustrates the very large upward revisions to income and consumption levels

resulting from the downward revisions to PPPs when moving from extrapolations from ICP

2005 to ICP 2011.15 As the range of the graphs shows, relative income and consumption

levels have increased by more than half in a large number of countries. Differences in GDP

per capita extrapolations are between 30 to 40 percent also for large countries like Brazil,

China and India and differences are quite large in oil-rich economies including Indonesia.

However, the relative differences exhibit a slightly different pattern for per capita

consumption. Differences for India are around 50 percent whereas the difference for

Indonesia is more moderate. More generally, there is a clear negative correlation with the

14 An alternative would be the implicit price deflator of household consumption expenditure from the National Accounts but the results are very similar. 15 The figure is qualitatively similar to Figure 1 in Deaton and Aten (2014, p. 31), though for OECD/Eurostat countries, their consumption levels are extrapolated from a more recent benchmark comparison than ICP 2005.

7

highest-income countries typically showing only small revisions. The lower-income countries

typically show larger revisions and this effect is stronger for consumption.16

Figure 1, Differences between extrapolating ICP 2005 and ICP 2011

To summarize the patterns in Figure 1 and compare these to alternative PPP series in the

remainder of this paper, we use three statistics based on the difference measure in (5).

Mean difference across the set of J countries: di

1

J

Pij2005

Pij2011

1

, for i Y ,C

j1

J

Root mean squared difference (RMSD): di2

1

J

Pij2005

Pij2011

1

2

j1

J

, for i Y ,C

Slope coefficient of the regression: dij a

i b

iln e

ijE

j uij, for j 1, , J and i Y ,C ,

where eij is either GDP or consumption per capita. These summary measures are best

understood by viewing the problem as a forecasting exercise. Until the release of the

benchmark ICP 2011 PPPs, the typical approach was to forecast PPPs for 2011 by

16 The difference between the GDP and consumption pattern suggests that revisions to the price levels of investment and government consumption are not related with income levels.

8

extrapolating 2005 PPPs using equation (4).17 Statistic is then a measure of the forecasting

bias, RMSD is a measure of forecasting uncertainty18 and an indication whether the bias

varies systematically with GDP or consumption per capita.19

Given the challenge of comparing prices of more than a thousand individual products across

almost 200 countries, it is not surprising that measurement error is present, which would lead

to RMSD being non-zero.20 Indeed, this can be a rationale for conducting regular benchmark

price comparisons: since the PPP of a given country in a given year will be measured with

error, it is good to have measurements in multiple years to avoid relying too heavily on a

single (possibly error-prone) observation.21 That said, RMSD is still a useful measure since a

method of extrapolating PPPs that would lead to a lower value of RMSD would be

preferable. The average bias id and the systematic variation in the bias measured by ib are

arguably the most worrisome measures as they imply a shift in the world income distribution.

This is most obvious for ib , as this measures whether lower-income countries show

systematically larger (or smaller) differences. But a non-zero is also worrisome because

PPPs are given relative to the United States and this thus implies an average shift in prices –

and thus expenditure levels – relative to US.

Table 1 shows summary statistics corresponding to the two panels in Figure 1. The average

country has expenditure levels in ICP that are about 25 percent higher than those based on

extrapolations from ICP 2005. Put differently, the average country is 24 percent richer

relative to the US than estimated previously, implying a sizeable decline in cross-country

income inequality. The root mean squared difference is also comparable between the GDP

17 An exception was Ravallion (2013, 2014), who has argued for a so-called ‘dynamic Penn effect’. However, Inklaar (2013) showed that the forecasting performance of the dynamic Penn effect was inferior to the standard inflation-based extrapolation. 18 In the terminology of the forecasting literature, this would be the root mean squared prediction error, RMSPE, see West (2006). 19 Note that the average difference and root mean squared difference are not independent of the numeraire country. However, since typical comparisons are almost exclusively made with the US as the numeraire this is not a major drawback. However, see Diewert (2009) for alternative measures. 20 See Deaton and Heston (2010) for a discussion of the conceptual and practical challenges of cross-country price comparisons. Note also that measurement errors may not just be a problem for measuring PPPs but also for tracking inflation at the national level. The errors in inflation measurement should typically be smaller since it is more likely that a particular product can be priced from one period to the next than in to (potentially) disparate) countries, though information on spending patterns are not regularly updated, measured inflation could well start to substantially deviate from true inflation. 21 See also e.g., Deaton (2012) and Rao and Hajargasht (2014) on the estimation of PPP standard errors from the variation of individual product prices around the overall GDP or consumption PPP.

id

ib

id

9

and consumption, at 33–35 percent. The main difference between the GDP and consumption

extrapolation errors is that the coefficient on expenditure per capita is much larger for

consumption that for GDP, though both are significant. As was apparent from Figure 1, the

differences tend to be smaller for non-oil-producing countries and the mean difference is

even larger for developing economies than for the sample as a whole.

Table 1, Differences between ICP 2011 and extrapolation from ICP 2005 –summary

statistics

All countries

Non-oil countries

Developing economies

GDP Mean difference 0.237*** 0.199*** 0.278*** Root mean squared difference 0.336 0.284 0.370

Coefficient on log(expenditure/capita) -0.020* -0.033*** 0.025 (0.012) (0.010) (0.019)

Consumption Mean difference 0.254*** 0.218*** 0.308*** Root mean squared difference 0.349 0.300 0.380

Coefficient on log(expenditure/capita) -0.063*** -0.066*** -0.027 (0.011) (0.009) (0.019)

Notes: summary statistics based on differences in Figure 1 for 142 countries. Countries where fuel exports

contribute more than 10 percent of real GDP are labeled as ‘oil countries’ (source: Penn World Table, version

8.0). Developing economies are the set of low-income and middle-income countries as defined by the World

Bank, so with a GNI/capita level less than $12,746. Robust standard errors of the regression coefficients are

shown in parentheses below the coefficients. * denotes a variable significantly different from zero at a 10%-

level, ** at 5%-level, *** at a 1%-level.

In the remainder of this paper, the results in 1 will serve as a baseline for comparison with

statistics based on counterfactual versions of ICP 2005 that harmonize methodology and

resolving sampling biases. Ideally, those adjustments will lead to a closer alignment between

the extrapolated (counterfactual) ICP 2005 results and ICP 2011 thus providing a clue as to

whether the developing world has indeed suddenly become richer. In the following section,

we describe the main methodological improvements introduced in 2011 ICP, drawing on

World Bank (2014) and explain how the counterfactual versions of ICP 2005 are constructed.

3. Methodological innovations in ICP 2011 and the counterfactuals for ICP 2005

The International Comparison Program (ICP) is a major international statistical initiative in

which prices and national accounts data are collected and processed to compile PPPs and

10

internationally comparable real expenditures. The early phases of the ICP, until 1985, were

essentially world comparisons where data from all participating countries were treated as a

single set. A major drawback of such a world comparison was that it proved hard to collect

price data for the same products across such a wide range of countries and the requirements

for global comparability stood in the way of comparability between the more similar

countries in each region. A start was made to move away from a single world comparison

with ICP 1993, but ICP 2005 was the first comparison with a fully-implemented regionalized

approach.

The regionalized approach involves compilation of PPPs in two stages. The first stage is at

the regional level, where countries would collect and compile data according to regional

product specifications, leading to more reliable relative prices within each region. In the

second stage, the regional comparisons are linked to complete the global comparison. The

methodology for linking regional comparisons was pioneered while data collection for ICP

2005 was in progress. So it was perhaps not surprising that after the release of ICP 2005, an

assessment of the methods used in 2005 for linking regional comparisons indicated several

deficiencies. These deficiencies are discussed in Rao (2013), on the methods for linking

prices for product categories – referred to as ‘basic headings’ in ICP – and in Diewert

(2013b), for linking at the aggregate level. To remedy the signaled deficiencies, several

methodological innovations have subsequently been implemented in ICP 2011.

Major changes in methodology raise the question how the results from ICP 2005 and ICP

2011 can be compared. A direct comparison of ICP 2011 results with an extrapolation from

ICP 2005 is like comparing apples and oranges in view of significant improvements in the

methods used in the ICP 2011 round. In this paper, we offer a solution to this problem by

constructing a counterfactual for ICP 2005 based on ICP 2011 methodology. We briefly

describe the major methodological innovations in the ICP 2011 – with further details

available in World Bank (2014b). We then construct counterfactual 2005 PPPs based on data

used in the ICP 2005 computations. The basic data used are the PPPs and national

expenditure data at the basic heading level for all the 146 participating countries. In addition,

productivity adjustment data for 2005, information on dwellings;22 the 2005 ring product list

22 We thank Alan Heston and Nada Hamadeh for providing the data and the worksheets used in the computation of linking factors for dwellings.

11

and prices; the 2011 global core list products and their prices in all the participating countries

have been obtained from the ICP Global Office at the World Bank.23

3.1 Harmonization of methodology

Aggregation at the regional level

In ICP 2005, the African region employed the transitive and additively consistent aggregation

method, the Iklé-Dikhanov-Balk (IDB) method, to compute aggregate PPPs within Africa,

while all other regions used the Gini-Elteto-Koves-Szulc (GEKS) method.24 Bringing all the

regions to line, the GEKS method was implemented in all the regions in ICP 2011. The first

step in the construction of the counterfactual is thus the replacement of the 2005 PPPs within

the African region based on IDB method with PPPs computed using the GEKS method.

Productivity adjustment to relative wages

Government consumption, including spending on public administration, health and

education, have always been considered ‘comparison-resistant’ in ICP since market prices for

the output of most of these services cannot be observed. So instead of relative output prices,

ICP compares government consumption by comparing input prices, specifically wages and

salaries. A drawback of using input prices is that relative productivity differences are not

taken into account, which means that in lower-income countries, a simple measure of relative

wages would understate relative output prices and overstate real expenditure under this

category.

In ICP 2005, the Asia-Pacific and African regions recognized the need to make a productivity

adjustment to relative wage data, but no other regions implemented any adjustments. An

implicit assumption in this limited use of productivity adjustment is that productivity of

government employees in all the remaining economies is the same as that in the United States

which is unlikely to be true. Consequently, a fully-fledged productivity adjustment for all the

economies participating was implemented in ICP 2011, see World Bank (2014b) for details.

Following the approach used in ICP 2011, we implement productivity adjustments based on

the contribution in 2005 of differences in (physical) capital per worker to GDP.25 We find a

marginal increase in the share of Africa, Asia-Pacific and Eurostat/OECD countries when a

23 These data are routinely provided to all researchers upon request to the ICP Global Office. 24 See Diewert (2013a) for detailed descriptions of the IDB, GEKS and other aggregation methods. 25 See World Bank (2014b) for details on the methodology and the Penn World Table, version 8.0 for data on comparative capital stocks and the share of labor income in GDP.

12

comprehensive productivity adjustment is implemented, see Appendix Table A1 for regional

shares before and after comprehensive productivity adjustment.

Linking dwellings using quantity indicator data26

To compare relative rents of dwellings across regions, a procedure was followed for ICP

2005 and ICP 2011 that is different from other types of consumption, see Heston (2013) and

World Bank (2014b). Ideally, information would be available on the actual (or imputed) rent

paid on different types of dwellings, but data on the number of dwellings and some quality

characteristics are more widely available.

In ICP 2005, the available rental price data was used to link the relative rents for dwellings

across regions, see Heston (2013). For ICP 2011, a similar approach was tried but the rental

data supplied was judged to be of insufficient quality. Instead a method based on quantity

indicators was applied. First, the number of dwellings per capita27 is used to construct a

regional quantity index for dwellings with OECD-Eurostat set at 100. The quantity index is

then adjusted for quality differences using a quality index computed as the arithmetic average

share of dwellings with electricity, inside water and private toilet.

To harmonize this aspect of the methodology, we apply the ICP 2011 methodology on

dwellings quantity and quality indicators data for 2005 supplied by the Global Office. We

combine the per capita volume index with a per capita value index, computed as the ICP

2005 expenditure in the ‘Actual and imputed rentals for housing’ basic heading category,

converted to US dollars using the current exchange rate and divided by population. This gives

a price level index for each region. The steps involved in implementing the 2011 linking

methodology for dwellings on data from 2005 are shown in Appendix Table A2. In Africa,

Asia-Pacific and Latin America, the prices according to the ICP 2011 methodology are lower

than under the ICP 2005 methodology, suggesting an upward revision to income levels in

countries from these regions.

26 We are grateful to Alan Heston who provided details of the approach used in 2005; Nada Hamadeh for providing data used in 2005 as well as in 2011; and to Paul Konijn for providing us with a copy of the document “Linking the regions: the case of housing” outlining the methodology used in linking dwellings data in 2011 ICP. The Konijn document describes the actual procedure implemented by the Computational Task Force for ICP 2011. 27 Though data on usable surface in sq. meters was collected, it was not of sufficient quality for use in the linking process. Therefore, linking was essentially based on number of dwellings per capita data.

13

Aggregation above basic heading level

In the implementation of the ICP, the very last step involves linking of regions at higher

levels of aggregation, such as household consumption and GDP. The methodology used for

linking at the GDP and component level used in 2005 was replaced by the country

aggregation and redistribution (CAR) method in 2011.

The starting point for linking above the basic heading level is data on national expenditure

and PPPs for each of the 155 basic headings, usually with the US=1. The problem in

aggregating these data for global comparisons of GDP is the additional requirement of fixity.

The fixity requirement stipulates that the relative levels of consumption or GDP of countries

within a region must be identical to the relative levels established at the regional level,

thereby ruling out the standard aggregation methods. Diewert (2013b) summarizes a range of

suggested methods for global comparisons that maintain fixity. The method used in ICP 2005

treats each region like a super-country28 and requires regional quantity and aggregate price

vectors. To construct these, first the implicit quantity for each basic heading in each country

is derived by dividing national expenditure by the basic heading PPP. These basic heading

quantities are then added across all the countries in each region, resulting in one quantity

vector for each of the six regions for each basic heading. In the second step, the price of the

basic heading for the region can be computed as total regional expenditure in the units of

currency of regional reference country divided by the quantity measure. This yields a price

and quantity vector for each of the regions, where prices are expressed in the currency units

of the reference countries in each region. The GEKS method is then used to get relative

regional prices at the aggregate level, which are combined with the within-region relative

prices to compare prices across regions.

As detailed in Diewert (2013b), though, it was later found that the method implemented in

ICP 2005 was not invariant to the choice of the reference country in each region. For ICP

2011, the use of the Country Aggregation with Redistribution (CAR) method was

recommended. This method starts with a global comparison of prices using the GEKS

method and basic-heading-level data for all individual countries. The resulting relative prices

are used to compute real, PPP-converted GDP for each country and these GDP levels are

added to arrive at total regional real GDP. In the second step, the shares of each country in

within-region GDP – an outcome of the regional comparison – is applied to total regional real

28 The term is from Deaton and Aten (2014).

14

GDP. In constructing the counterfactual comparisons for 2005, we implement the CAR

approach with the basic heading PPPs and national expenditures obtained after using the

linking procedures detailed above.

Effect of harmonized methodology on ICP 2005

The first step in our approach to construct a counterfactual for ICP 2005 prices is to

harmonize the methodology that used in ICP 2011. In Table 2, we show how the original ICP

2005 and the different counterfactuals compare to ICP 2011. Comparing the first and last

column, it is clear that harmonizing the methodology results in lower mean differences and a

lower RMSD for both GDP and consumption. Though the mean difference for the fully

harmonized counterfactual is still sizeable at 17 to 18 percent, this is 26 percent smaller for

GDP and 31 percent smaller for consumption than the original. Similarly, the RMSD has

gone down by 15 (GDP) to 18 (consumption) percent. Going through the various stages of

the harmonization, the most substantive change results from harmonizing the within-Africa

PPPs, while implementing the CAR method, productivity adjustment and dwellings

adjustment matter relatively less.

Table 2, Differences between ICP 2011 and extrapolations from ICP 2005 –

counterfactuals based on harmonized methodology

Original CAR method

+ Within-Africa harmonization

+ Productivity adjustment

+ Dwellings adjustment

GDP

Mean difference 0.237*** 0.216*** 0.175*** 0.192*** 0.184*** Root mean squared difference 0.336 0.349 0.280 0.290 0.285

Coefficient on log(expenditure/capita)

-0.020* -0.043*** -0.018* -0.016 -0.015

(0.012) (0.013) (0.010) (0.010) (0.010)

Consumption

Mean difference 0.254*** 0.254*** 0.187*** 0.187*** 0.174*** Root mean squared difference 0.349 0.410 0.296 0.296 0.286

Coefficient on log(expenditure/capita)

-0.063*** -0.102*** -0.066*** -0.066*** -0.063***

(0.011) (0.017) (0.010) (0.010) (0.010) Note: column labeled ‘Original’ is from Table 1. ‘CAR method’ uses the regional linking methodology of ICP

2011, ‘+ Within-Africa harmonization’ uses the GEKS method rather than the IDB method to compare prices

within the African region; ‘+ Productivity adjustment’ also implements the productivity adjustment for public

wages for the Eurostat/OECD region and Latin America and ‘+ Dwellings adjustment’ also implements the ICP

2011 method for comparing the (rental) price of dwellings across regions. This column corresponds to a set of

PPPs where the methodology has been harmonized to the extent possible. Robust standard error of the

regression coefficients shown in parentheses below the coefficients. * denotes a variable significantly different

from zero at a 10%-level, ** at 5%-level, *** at a 1%-level.

15

In addition to the methodological innovations in 2011 discussed here, ICP 2011 also

introduced a new method for measuring relative prices in construction. ICP 2005 used the

‘basket of construction components’ (BOCC) method was – see McCarthy (2013a) for details

– but it was replaced in ICP 2011 by a simpler method based on the prices of basic

construction materials, different types of labor, and the hire of machinery and equipment.

However, the information required to implement the ICP 2011 construction methodology was

not available from the 2005 data sources.

3.2 Accounting for price sampling biases due to ring country linking methodology in

ICP 2005

This section outlines the adjustments made to 2005 ICP in the process of constructing a

counterfactual that incorporates the main improvements to methodology introduced in ICP

2011. World Bank (2014b) emphasizes changes to linking methodology as a major

innovation in ICP 2011. Linking regional comparisons to arrive at the overall global

comparison is a critical step in ICP. ICP 2011 introduced a global core list (GCL) of products

and pricing of GCL products by all 177 participating countries in ICP29 in the process of

linking PPPs. In contrast, the linking procedure in ICP 2005 was based on data for only 18

countries, the so-called ring countries,30 which were deemed to be representative for the full

set of countries in their respective regions. The ring countries collected prices for items on

the global ring product list and those prices were then used to link PPPs from different

regions; see also Deaton and Heston (2010) and Vogel (2013).31

After the completion of ICP 2005, several issues were identified with the linking procedure

based on ring countries and the ring products. The first is whether the particular selection of

18 ring countries would induce a bias compared to the use of data for all the participating

countries. We term this bias the ring country selection bias. The second issue relates to the

items included in the ring product list. A close examination of the products included in the

ring list and the product specifications used for pricing purposes revealed a strong rich

country bias in the products. The list included items like Uncle Toby oats; a bottle of

Heineken beer, Bordeaux wine; and a Peugeot 408, which is suggestive of a rich country bias

29 Though ICP 2011 covered 199 countries, only 177 countries participated fully and the remaining are Pacific Island countries whose participation was limited to comparisons of household consumption. 30 The ring countries are Comoros, Egypt, Kenya, Senegal, South Africa and Zambia in Africa; Hong Kong, Malaysia, Philippines and Sri Lanka in Asia-Pacific; Brazil and Chile in Latin America; Jordan and Oman in Western Asia; and Estonia, Japan, Slovenia and the United Kingdom in Eurostat/OECD. 31 See Rao (2013) on the linking methodology at the basic heading level and Diewert (2013b) on linking at the aggregate level.

16

in the ring list. Deaton (2010) presented further suggestive evidence that the use of the ring

product list led to relatively high price levels (and thus lower real incomes) in low-income

ring countries. We refer to this as the ring product selection bias. In addition, we also

consider the widely discussed urban bias in prices of household consumption goods in China

in ICP 2005. In this section we describe how we quantify these biases and use them in

constructing a counterfactual for ICP 2005.

Urban bias in consumption prices from China

It is well documented that consumption prices provided by China to the Asia-Pacific regional

comparison in 2005 were collected from only 11 capital cities and their surrounding areas

(see ADB, 2007 and Word Bank, 2008).32 The general consensus from Deaton and Heston

(2010), Chen and Ravallion (2010a,b) and Feenstra, Ma, Neary and Rao (2013) is that the

Chinese consumption prices used in ICP 2005 were biased upwards relative to national

average prices by 20 percent or more.33 In contrast, the Chinese price survey for ICP 2011

was much more comprehensive, covering urban and rural areas in each of the country’s

provinces (see ADB, 2014). In constructing the counterfactual for 2005, we make a 20

percent upward adjustment to consumption prices for China.

Ring country selection bias

We cannot use data for 2005 to establish whether there was any bias from the specific

selection of ring countries. However, given the price data for all 177 countries in 2011, we

can examine if selecting the 18 ring countries from 2005 would have introduced any

systematic bias in the (2011) linking factors at the basic heading level. The assumption here

is that the ring country bias we find in 2011 would have been present to the same degree in

2005 as well.

To measure this bias, we exploit the basic structure of the country-product-dummy (CPD)

model used in the ICP.34 We specify a simple regression model that explains log prices P of

product i in country j located in region r using dummies for each product, Di, each region, Dr,

32 The Global Office of the ICP at the World Bank and the regional coordinating agency, Asian Development Bank, devised a mechanism to derive national average prices for China based on price data from the 11 cities. However, the method employed failed to account for rural-urban and regional price differentials. 33 PWT 7.1 and 8.0 implemented a downward adjustment in the preparation of their PPP and real expenditure extrapolations. A similar adjustment was used in Chen and Ravallion (2010a,b) in estimating absolute poverty in China. 34 Details of the CPD model and its use in the ICP can be found in Rao (2013).

17

a ‘ring country’ dummy R (which takes value 1 if the country is one of the 18 ring countries)

and interactions between the regional dummies and the ring country dummy:

logP

ijr

ii Di

rr DrR

rr DrR ijr (6)

Following the linking procedure used at the basic heading level (see Rao, 2013 for details),

prices are first converted into regional numeraire currencies using within-region basic

heading PPPs. This model, without the ring country dummy and interaction terms, is used in

ICP to derive the linking factors for each basic heading, see Vogel (2013). The r parameters

are of main interest since a significant coefficient would indicate that pricing in the ring

countries in that region are significantly different from pricing in the other countries in the

region.

Table 3, Estimating the difference in 2011 pricing of global core list products

in ring countries versus other countries

Region coefficient Region x ring country coefficient

Africa 1.8023*** Africa x ring country −0.0023 (0.0288) (0.0117) Asia-Pacific 1.8221*** Asia-Pacific x ring country −0.0695*** (0.0309) (0.0202) Latin America 0.6997*** Latin America x ring country 0.0158 (0.0288) (0.0209) Western Asia −1.3541*** Western Asia x ring country -0.0107 (0.0294) (0.0152) Ring country dummy 0.0033 (0.0079) Number of observations 57,415 Note: table shows regression results where the log of consumption item prices for 2011, converted to the

regional base country’s currency, are explained by dummies for each region, whether a country was a ring

country in ICP 2005, interactions between this ring country dummy and the regional dummies, and dummies for

each of the 1192 items (not shown), see also equation (6). Items are weighted by their importance, see main text

for discussion. The Eurostat/OECD region is taken as the base region and hence the coefficients should be

interpreted as price differences relative to the Eurostat/OECD. Robust standard errors of the regression

coefficients shown in parentheses below the coefficients. * denotes a variable significantly different from zero at

a 10%-level, ** at 5%-level, *** at a 1%-level.

r

r

18

Table 3 reports the results of a pooled regression across all basic headings, based on 57,415

price observations35 and using the ‘importance weights’ that are also used in ICP 2011.36 The

regional coefficients in the left panel provide the average regional linking factorsr. For

example, the table shows that 6.15 Hong Kong dollars (exponential of 1.8023, the coefficient

for the Asia-Pacific) for one US dollar is the linking factor when all the countries are used in

linking the regions. We also observe that the coefficient of ring country dummy is

numerically very small and statistically insignificant. This means that the selection of ring

countries on average has no influence on the linking factors.

Turning to the r parameters, we find insignificant coefficients for ring countries in all the

regions with the exception of the Asia-Pacific region. The significant interaction term for the

Asia-Pacific ring countries indicates that prices in the ring countries (Hong Kong, Malaysia,

Philippines and Sri Lanka) from this region are, on average, 6.95 percent lower than in other

Asia-Pacific countries. In our counterfactual ICP 2005, we divide prices by the exponent of

this coefficient to counteract this ring country selection bias.

Ring product selection bias

In ICP 2005, the 18 ring countries priced items from the ring product list and ideally, this list

should contain products that are part of a representative consumption bundle in every country

of their respective region. In practice, this is much harder, as discussed in more detail by

Deaton and Heston (2010). For instance, a bottle of Heineken beer was on the ring product

list, but if this is a premium brand in poor countries, while a standard choice of beer in rich

countries, beer prices in the poor country will be relatively higher than would be the case for

beer in general. Deaton (2010) presented suggestive evidence that points to a ‘rich country’

bias in the ring product list.

Given this potential for bias, a global core list (GCL) of products was developed for ICP

2011 which is representative of goods and services that are consumed in all regions of the

world. The ICP 2011 GCL product approach seems to have been fairly successful as the GCL

had significant overlap with the regional product lists. Of the 692 consumption products on

the GCL, 610 matched with the regional list in Africa; 412 with Asia; 394 with OECD-

35 The regressions includes data from 148 countries. If these had priced all the 692 GCL products, we would have 102,416 observations, so not all products were priced in all countries. 36 A product that is deemed important in a particular country gets a weight of 3 and other products get a weight of 1.

19

Eurostat; 489 with Latin America and the Caribbean and 606 with the list used in Western

Asia.

To explicitly test whether the product selection in drawing up the ring product list was a

source of bias, we exploit the fact that there were multiple ICP 2005 ring countries in each

region and that these ring countries participated in two sets of price comparisons: one based

on the regional product list as a part of the regional comparisons and the other based on the

ring product list for the purpose of computing linking factors. As the regional lists were

drawn up to be representative of products in each region, it seems reasonable to assume that

the results based on the regional price comparison are a more accurate measure of relative

prices between the ring countries within the region than the relative price differences implied

by the ring product list. Indeed, in the compilation of ICP 2005 the ring product list prices are

only used for linking prices between regions, not for price comparisons within the region. So

if we find that the relative prices within each region based on the ring product list prices

differ systematically from the relative prices implied by the regional product prices, we take

this as evidence of a bias from the particular selection of the products on the ring product list.

We use the following variant of the CPD model to test for the presence of product selection

bias:

logPijr

iDi

rBr

ijri (7)

where P represents item price (converted using the basic heading PPP), i refers to item, j

refers to country and r represents region. The new dummy variable Br, the base country of

region r. Importantly, ijrP measures the price of item i in country j which is part of region r

relative to the basic heading PPP for the product group of which item i is a part. The basic

heading PPPs are expressed relative to the base country of each region. The base countries in

different regions in ICP 2005 tended to be the highest-income countries of the set of ring

countries in the region: South Africa in Africa, Hong Kong in Asia-Pacific, United Kingdom

in Eurostat/OECD, and Oman in Western Asia. The exception was Latin America, where

Brazil was the base country but Chile had a higher income level.

If there were no ring product selection bias, we would expect the estimate of r not to differ

significantly from zero. However, based on the arguments of Deaton (2010) and Deaton and

Heston (2010) that the ring product list was a ‘rich-country’ product list, we would expect

20

negative estimates of r in Africa, Asia-Pacific and Western Asia. Since the regional base

countries are the highest-income (ring) countries in the region, the ring products would be

(more) representative of consumption patterns in the base country, while in the other ring

countries in the region the ring products would be unrepresentative and thus relatively

expensive.

Table 4, Estimating the difference in ring product prices in 2005 between regional base

country and other ring countries

Region (base country) Base country dummy Country Country dummy Observations Products Africa (South Africa) −0.0940*** Comoros 0.1708*** 2981 711

(0.0247) (0.0299) Egypt 0.0958***

(0.0313) Kenya 0.2127***

(0.0295) Senegal 0.0788***

(0.0287) Zambia −0.1334***

(0.0314) Asia-Pacific (Hong-Kong) −0.1556*** Sri Lanka 0.2781*** 2171 681

(0.0159) (0.0223) Malaysia 0.1309***

(0.0176) Philippines 0.0834***

(0.0185) Eurostat-OECD (UK) 0.0035 Estonia −0.0082 2004 636

(0.0143) (0.0166) Japan −0.0022

(0.0234) Slovenia 0.0012

(0.0160) Latin America (Brazil) 0.0103 Chile −0.0103 594 297

(0.0251) (0.0251) Western Asia (Oman) −0.0661** Jordan 0.0661** 710 355 (0.0322) (0.0323)

Notes: Left part of the table shows regression results where the log of consumption ring item prices for 2005,

converted to the regional base country’s currency using basic heading PPPs, are explained by a base country

dummy and dummies for each of the items (not shown), see also equation (7). The right part of the table reports

regression results where the base country dummy is replaced by individual country dummies. Robust standard

errors of the regression coefficients shown in parentheses below the coefficients. * denotes a variable

significantly different from zero at a 10%-level, ** at 5%-level, *** at a 1%-level.

Equation 7 is estimated separately for each region. The results from estimating equation (4)

for each region are reported in the left panel of Table 4 and the results conform to our prior

expectations. The ring product list prices are systematically lower in the regional base

countries in Africa, by 9.4%, in Asia-Pacific by 15.6% and in Western Asia by 6.6%. This

means that ring product prices for the reference country South Africa in Africa are on average

21

9.4 percent below the prices of observed in other ring countries in Africa (Egypt, Kenya,

Senegal and Zambia). The ring-country effect is most pronounced in the Asia-Pacific region

where Hong Kong ring prices are found to be 15.56 percent lower, on average, than the ring

countries of the region, viz., Malaysia, Philippines and Sri Lanka. These results are contrary

to the expectation based on Balassa-Samuelson and Penn effect that price levels in lower

income countries are generally lower. The insignificant coefficient in the Eurostat/OECD

region provides further indication that the products on the ring product list were

representative of rich-country consumption patterns.

Table 5, Estimating the difference in global core list product prices in 2011 between

regional base country and 2005 ring countries

Region (base country) Base country dummy Country Country dummy Observations Products

Africa (South Africa) 0.0034 Comoros -0.0005 2770 567

(0.0270) (0.0306)

Egypt 0.0260

(0.0361)

Kenya -0.0083

(0.0305)

Senegal -0.0195

(0.0334)

Zambia -0.0239

(0.0329)

Asia-Pacific (Hong-Kong) 0.0702*** Sri Lanka 0.0812** 1478 440

(0.0242) (0.0352)

Malaysia -0.1618***

(0.0283)

Philippines -0.0859***

(0.0284)

Eurostat-OECD (UK) 0.0242 Estonia -0.0170 1330 380

(0.0158) (0.0197)

Japan -0.0592***

(0.0225)

Slovenia -0.0087

(0.0189)

Latin America (Brazil) 0.0020 Chile -0.0020 508 254

(0.0313) (0.0387)

Western Asia (Oman) -0.0254 Jordan 0.0254 920 460

(0.0220) (0.0238) Notes: Left part of the table shows regression results where the log of global core list consumption item prices

for 2011, converted to the regional base country’s currency using basic heading PPPs, are explained by a base

country dummy and dummies for each of the items (not shown), see also equation (7). The right part of the table

reports regression results where the base country dummy is replaced by individual country dummies. In both

22

sets of regressions, robust standard errors of the regression coefficients shown in parentheses below the

coefficients. * denotes a variable significantly different from zero at a 10%-level, ** at 5%-level, *** at a 1%-

level.

To provide further support for the findings in the left panel of Table 4, we reformulate

equation (7) with dummy variables for each of the ring countries within each region except

for the regional numeraire and the right panel in Table 4 shows the results. The coefficient of,

for example, Kenya implies that ring prices (relative to the regional PPPs) are 21 percent

higher than those in South Africa. The coefficient of Zambia is somewhat puzzling since it

shows that prices of ring products in Zambia are 13 percent lower than implied by the

regional comparison, making it the only country with a sign counter to expectations.37 In the

Asia-Pacific region, the signs and magnitudes are as expected, with e.g. Sri Lankan ring

prices 27.81 percent higher than in Hong Kong. None of the coefficients are significant in the

OECD-Eurostat and Latin American regions, which suggests that that ring prices are aligned

with regional prices.

Though we would expect smaller or no product selection bias based on the 2011 GCL, we

run similar regressions as reported in Table 4 based on those data. Results of the regressions

based on 2011 GCL prices are presented in Table 5. The results for the most part confirm

with our expectations as there is less evidence of product selection bias. The only exception

is the Asia-Pacific region, where prices in Hong-Kong were 7.02 percent higher than in the

other (2005) ring countries.

Product selection bias from Deaton and Aten (2014)

Deaton and Aten (2014) were the first to examine the sources of systematic and major

discrepancies between 2011 ICP and extrapolations from the 2005 benchmark. Their

approach to identifying a bias from the product selection in 2005 differs from ours. They

estimated country-level (consumption) PPPs for the 2005 ring countries based on the (2005)

ring product list and based on the (2011) GCL. They then compared the change in these PPPs

between 2005 and 2011 to the relative inflation rate as in equation (4), taking the UK from

the Eurostat-OECD region as the numeraire. For the high-income ring countries, they find

that relative inflation rates approximate the change in PPPs quite well but for the low-income

countries, they find smaller changes in PPPs than implied by relative inflation (see Deaton

and Aten, 2014, Figure 5). If one assumes that GCL prices are unbiased (which is supported

37 Using robust regression makes little difference, suggesting Zambia’s result is not driven by price outliers for individual products, but by a general pattern.

23

for most regions in Table 5), then 2005 relative prices for the low-income ring countries were

biased upwards. They conclude that the ring product prices may have overstated price levels

in the low-income ring countries in Western Asia, Africa and the Asia-Pacific by between 20

to 30 percent.

This is a larger adjustment than implied by our analysis in Tables 4 and 5. Table 4 showed

that prices in Africa were biased upwards relative to the regional base country by 9.4 percent

and a 6.6 percent bias in Western Asia. In Asia-Pacific, prices in 2005 were biased

downwards by 15.6 percent while prices in 2011 were biased upwards by 7.0 percent for a

total discrepancy of 22.6 percent.38 However, there is reason to believe that our estimates are

conservative. The analysis reported in Tables 4 and 5 can only identify a bias relative to the

regional base country, not a bias relative to the global base region, Eurostat-OECD.

Especially in Oman (Western Asia) and South Africa (Africa), consumption levels are

notably lower than in the UK, the Eurostat-OECD base country. This could mean that the

ring list products may not have been representative in those countries either, implying a

larger bias than we can identify in our regressions.

Accordingly, our preferred counterfactual is constructed on the basis of the adjustments we

derive for the methodological harmonization and ring country selection bias but we replace

the product selection bias estimates with those recommended by Deaton and Aten (2014). We

label this as ICP 2005C. The rest of the paper focuses on the preferred ICP 2005C

counterfactual, though for completeness we also present results for a conservative

counterfactual in the Appendix based on the regression coefficients in Tables 4 and 5.

Counterfactual for ICP 2005C based purely on sampling bias corrections

Based on the discussion and results presented in the previous section, we are in a position to

construct the counterfactual that is purely based on sampling bias corrections. The sampling

bias corrections include: (i) a 20 percent downward adjustment of Chinese consumption

prices; (ii) a downward adjustment of 6.4 percent to Asia-Pacific consumption prices to

correct for the bias due to the use of Hong Kong, Malaysia, Philippines and Sri Lanka as ring

countries instead of the full set of countries in the region; and finally (iii) a downward

adjustment for ring product selection bias of 25 percent in low-income ring countries.

38 These results suggest that the 2011 results could be adjusted rather than letting all adjustments fall on the 2005 relative prices. For the purpose of analyzing the differences between the extrapolation from 2005 and 2011, this does not matter.

24

Table 6, Differences between ICP 2011 and ICP 2005C

Counterfactual with corrections for price sampling bias

Original – Urban bias China

– Ring country selection bias

– Ring product selection bias

GDP

Mean difference 0.237*** 0.198*** 0.207*** 0.118*** Root mean squared difference 0.336 0.306 0.318 0.235

Coefficient on log(expenditure/capita)

-0.020* -0.021* -0.021* 0.010

(0.012) (0.011) (0.012) (0.010)

Consumption

Mean difference 0.254*** 0.223*** 0.238*** 0.088*** Root mean squared difference 0.349 0.327 0.346 0.196

Coefficient on log(expenditure/capita)

-0.063*** -0.064*** -0.067*** 0.000

(0.011) (0.011) (0.011) (0.009) Note: columns labeled ‘Original’ are from Table 1. Robust standard error of the regression coefficients shown in

parentheses below the coefficients. * denotes a variable significantly different from zero at a 10%-level, ** at

5%-level, *** at a 1%-level. Ring country selection bias correction is from this study and ring product selection

bias is from Deaton and Aten (2014).

The columns in Table 6 progressively correct for the differences price sampling biases. For

example, at the GDP level, adjusting for urban bias in China prices has reduced the mean

difference from 23.7 percent to 19.8 percent representing a 16 percent reduction in average

differences. The last column shows the results for the counterfactual which incorporates

corrections for all types of price sampling bias. At the GDP level the mean difference has

fallen from 0.237 to 0.118 representing, a 51 percent reduction. What is interesting is the

slope coefficient of the regression, which is close to zero and statistically insignificant. This

means that the country-specific differences do not exhibit any pattern around the mean

difference. The bottom panel of Table 6 for consumption shows that, the average difference

between the extrapolations and 2011 ICP is reduced by 66 percent from 0.254 to 0.087 and

the RMSD decreases from 0.349 to 0.195. In the next section we present the IRDA

counterfactual for 2005 which incorporates methodological harmonization and the adjustment

for price sampling biases.

4. The counterfactual for the ICP 2005C benchmark international comparisons –

implications for global inequality

This section presents what we consider is the most appropriate counterfactual for 2005 that

could be meaningfully compared to ICP 2011. The counterfactual, ICP 2005C, accounts for

all the methodological changes adopted in 2011 with the exception of the new method used

25

for construction and also accounts for price sampling biases associated with 2005 due to the

use of ring countries and prices for ring list products collected from these countries.

The first question we ask is to what extent the systematic differences between extrapolations

from ICP 2005 with ICP 2011 are reduced or eliminated when moving to ICP 2005C. We

then discuss salient features of the ICP 2005C, including the impact on the size of the world

economy, on major economies such as China and India and the effect on international income

inequality. In the Appendix, we present the counterfactual for 2005 based on the more

conservative adjustment for product selection bias based on the regression coefficients in

Tables 4 and 5.

Differences between ICP 2011 and extrapolations from ICP 2005C

In Table 7 we provide an assessment of the differences between ICP 2011 and ICP 2005C.

Figure 2 then shows a scatter plot of the differences for different countries and for GDP and

consumption, akin to Figure 1. Extrapolating from ICP 2005C leads to substantially lower

differences with ICP 2011, with the mean difference for GDP reduced from 0.237 to 0.100

(58 percent) and a large reduction from 0.254 to 0.040 (an 84 percent reduction) for

consumption. In the last column of the table we present population weighted differences. At

the GDP level, the counterfactual implies a reduction in the mean difference of 67 percent

while at the consumption level the differences are virtually eliminated. The counterfactual

has brought the ICP price comparisons a lot closer to the ICP 2011 with the remaining

differences accounted for by the usual factors that explain discrepancies between

extrapolations and benchmarks.39

The slope coefficient of the regression of differences on per capital nominal income, at the

GDP level, showed a change in sign but a reduction in absolute value with a change from -

0.020 to 0.016. However for consumption, the slope coefficient is practically eliminated from

a significant negative slope. This means that for consumption, the differences are randomly

scattered around the mean difference of 0.040. We also find that the remaining differences

are largely driven by the presence of “oil countries”. Column 3 shows that the average

difference reduced from 0.237 to 0.061 (a reduction of 74 percent) at the GDP level and

reduced by 94 percent from 0.254 to 0.014 for consumption. The root mean square error

differences are also substantially reduced for the counterfactual.

39 See McCarthy (2013b) and Inklaar and Timmer (2013) for a discussion on the discrepancies between benchmarks and extrapolations.

26

Table 7, Differences between ICP 2011 and extrapolations from ICP 2005C

counterfactual with methodological and sampling adjustments

Original Counterfactual

All countries

All countries

Non-oil countries

Developing economies

All countries, population-weighted

GDP

Mean difference 0.237*** 0.100*** 0.061*** 0.106*** 0.081***

Root mean squared difference 0.336 0.218 0.165 0.235 0.218

Coefficient on log(expenditure/capita)

-0.020* 0.016* 0.007 0.065*** -0.019

(0.012) (0.009) (0.007) (0.017) (0.015)

Consumption

Mean difference 0.254*** 0.041*** 0.015 0.045** 0.006

Root mean squared difference 0.349 0.168 0.132 0.177 0.181

Coefficient on log(expenditure/capita)

-0.063*** 0.000 -0.001 0.031** -0.019

(0.011) (0.008) (0.007) (0.014) (0.022)

Note: The first column (ICP 2005) is from 1. ICP 2005C refers to the counterfactual ICP 2005 prices that

harmonize methodology and correct for price sampling biases. The Deaton/Aten adjustment is a 25 percent

downward adjustment to African and (West) Asian consumption prices. Countries where fuel exports contribute

more than 10 percent of real GDP are labeled as ‘oil countries’ (source: Penn World Table, version 8.0).

Developing economies are low-income and middle-income countries as defined by the World Bank, so with a

GNI/capita level less than $12746. Robust standard errors of the regression coefficients shown in parentheses

below the coefficients. * denotes a variable significantly different from zero at a 10%-level, ** at 5%-level, ***

at a 1%-level.

It is also helpful to emphasize the differential roles played by methodological adjustments

(cf. Table 2) and price sampling adjustments (cf. Table 6). Comparing those results suggests

that the adjustment for the ring product selection bias has a larger effect than that resulting

from methodological adjustments. Our analysis also suggests the robustness of the use of

prices for global core list of products and the use of all the participating countries in the

linking process. However, results in Tables 3 and 4 on the estimation of ring country

selection bias, significant coefficients for the Asia-Pacific region suggests that there is need

for further fine-tuning and integration of the GCL product list in the Asia-Pacific region in

the next round of ICP.

At the GDP level the average differences are still significant at 10.0 percent and the RMSD

for GDP and consumption imply substantial country-specific differences. There a number of

possible reasons for this. First, we have not been able to implement any adjustment for the

new methodology for construction used in 2011. Second, ICP does not survey export and

import prices but uses exchange rates instead. This means that differences in terms of trade

27

are not taken into account, despite their importance in cross-country comparisons, see

Feenstra et al. (2013). McCarthy (2013b) provides a more extensive set of conceptual and

measurement reasons why two subsequent benchmarks may not align.40 Finally, ICP is a

survey of prices so measurement error, reflected in e.g. the RMSD, will always be present.

Estimates of the standard error of PPPs are in the order of 20 percent – Deaton (2012) and

Rao and Hajargasht (2014) – so the RMSD of 14–20 percent in the second column of Table 7

looks fairly good from that perspective. In that sense, these results confirm the need to

conduct future rounds of ICP since any given round can only measure relative prices across

countries with error. This is no different when estimating relative prices using information on

income per capita (i.e. the Balassa–Samuelson effect); the root mean squared error (RMSE)

of such regressions is also in the order of 20 percent, despite an R-squared of around 70

percent and highly significant coefficients.

Figure 2, Differences between extrapolating ICP 2005 counterfactual and ICP 2011

Size of the world economy and selected countries

At the time of the release of results from ICP 2005, commentaries from the economists and

researchers focused on the big reduction in the size of the world economy and of the

economies of China and India compared to the extrapolations from 1993 benchmark

40 See also e.g. Dalgaard and Sørensen (2002).

28

available at that time from the World Development Indicators and the Penn World Table 6.3.

Deaton and Heston (2010) and Deaton (2010) draw attention to the big downward adjustment

implicit in ICP. So it is useful to revisit this and place it in the context of ICP 2005 to see the

magnitudes of the differences implied by the counterfactual ICP 2005C.

Table 8, GDP per capita in 2005 in selected countries, ICP 2005 versus ICP 2005C

GDP per capita (in US$)

ICP 2005C/ICP 2005 – 1 ICP 2005 ICP 2005C

China 4124 5678 38% India 2234 2742 23% Kenya 1340 1636 22% Ethiopia 614 687 12% Brazil 8501 8175 -4%

World GDP ($trillion) 56.8 61.8 9%

At the time of release of ICP 2005 (World Bank, 2007), the general commentary focused on

the downward adjustment to the size of Chinese and Indian economies of 40 and 36 percent,

respectively, compared to what was expected at that time based on the World Development

Indicators (WDI, 2007). In Table 8 we find the ICP 2005C proposed in this paper pegs back

the downward adjustment. In the case of China, the adjustment is 38 percent bridging the 40

percent gap identified at that time whereas the counterfactual provides an upward revision of

23 percent of per capita income in India which is slightly lower than the 36 percent identified

in 2007. We find similar upward adjustments to Kenya, Ethiopia and most of the African and

Western Asian countries. In the case of Brazil, the counterfactual implies a downward

revision of nearly 4 percent. For most of the rich countries the revisions are marginal. An

immediate consequence of these revisions is a downward revision of the estimate of global

inequality in 2005 based on ICP 2005C. In Table 9 we confront our counterfactual based

extrapolations to 2011 with ICP 2011 actual and with extrapolations from ICP 2005.

The last column shows a significant reduction in the difference between the extrapolations

and actual 2011 when the counterfactual is used. However, some differences are still large for

countries like Zambia and Jordan. For India the difference reduces to 8 percent whereas the

difference goes down to –13 percent from +21 percent for China. This change is expected as

the price sampling adjustment includes a 20 percent downward adjustment to account for the

urban bias in Chinese prices in 2005.

29

Table 9, GDP per capita in 2011 for selected countries in PPP converted US dollars

Extrapolations from Actual ICP 2005 and Counterfactual ICP 2005

GDP per capita in 2011

ICP 2011 Extrapolations from ICP 2005

Difference ICP 2005 and 2011

Actual Actual Counterfactual Actual CounterfactualKenya 2136 1729 2111 0.24 0.01Zambia 3155 1745 2235 0.81 0.41India 4735 3566 4377 0.33 0.08China 10057 8335 11476 0.21 -0.12Jordan 11169 5844 6856 0.91 0.63Brazil 14639 11785 11333 0.24 0.29United Kingdom 35091 36153 37447 -0.03 -0.06Germany 40990 37729 39199 0.09 0.05United States 49782 49782 49782 0.00 0.00

Note: the column ‘ICP 2011’ shows 2011 GDP per capita as reported in World Bank (2014, Table 6.1), based

on ICP 2011 PPPs. Columns ‘ICP 2005’ shows 2011 GDP per capita based on the actual World Bank (2008)

ICP 2005 PPPs or the counterfactual 2005 PPPs developed in this paper, extrapolated to 2011 using the change

in GDP deflator in each country relative to the change in the United States. The columns ‘Difference’ are the

percentage change of the ICP 2011 column relative to the ICP 2005 columns.

International inequality

The ICP 2011 results showed a significant increase in per capita real income of low-income

countries compared to extrapolations from 2005 while there was no appreciable difference

for the high-income countries. This suggests that between-country income inequality is lower

according to ICP 2011 than based on ICP 2005. We analyze this more formally by examining

the trend in the population-weighted Gini coefficient. This provides an indication of what

Milanovic (2012) refers to as type-II international income inequality. This measure stops

short of a full global inequality measure since it does not take into account inequality within

countries, but (in contrast to type-I international income inequality) it does give greater

weight to more populous countries.41

Figure 3 shows three Gini coefficient series between 1990 and 2012: one based on ICP 2005,

one based on ICP 2011 and one based on ICP 2005C42. The most striking feature of this

figure is the consistent difference between inequality measures based on actual ICP 2005 and

2011 and the nearly indistinguishable profile of inequality from ICP 2005C and ICP 2011.

41 The Gini coefficient is related to the Lorenz curve, which plots cumulative income shares against cumulative population shares when ranking countries by their income/capita level. Specifically, the Gini coefficient is the area between the diagonal and the Lorenz curve divided by the area below the diagonal. 42 The Gini coefficients presented in this figure are presented in appendix table A3.

30

For instance, in 2011 the extrapolated Gini for GDP per capita from ICP 2005 was 0.525, the

ICP 2011 Gini was 0.484 and the ICP 2005C Gini was 0.489. Gini coefficients for

consumption per capita are similarly close when comparing ICP 2011 and ICP 2005C. This

means that from the perspective of international income and consumption inequality, our

suggested methodological and price sampling adjustments basically eliminate the differences

between the 2005 and 2011 price comparisons.

Figure 3, International inequality in income and consumption per capita,

population-weighted Gini coefficient, 1990-2012

Note: GDP per capita in PPP-converted dollars is computed in the benchmark year (2005 or 2011) and these

levels are extrapolated using growth in GDP per capita. The full set of 142 economies that participated in ICP

2005 and 2011 can be covered since 1990 and these 142 economies cover approximately 95 percent of the world

population.

5. Conclusions

The publication of the ICP 2011 results, in April 2014, immediately sparked much discussion

given the large upward revisions to GDP per capita in PPP-converted terms. The changes to

economic geography, global inequality and poverty implied by ICP 2011 are significant and

warranted an explanation. Deaton and Aten (2014) were the first to research this issue and

their findings point towards the data and methods used for linking the regions in the ICP

Actual'ICP'2005

Actual'ICP'2011

Counterfactual'ICP'2005

.5.55

.6.65

1990 1995 2000 2005 2010

GDP

.5.55

.6.65

1990 1995 2000 2005 2010

Consumption

31

2005 round. In contrast, Ravallion (2014) argues, using dynamic Penn effect models, that it is

the 2011 round that could be an issue.

The starting point for our research is the recognition that there have been major innovations

in the methodology used in the ICP 2011. These are now well documented in the recently

released Final Report for ICP 2011 (See World Bank, 2014b). Of particular note is the use of

the global core list of products and the use of all the 177 participating countries in the linking

process. There are other significant innovations including the use of the country aggregation

and redistribution (CAR) method for aggregation above the basic heading level; productivity

adjustments in dealing with government compensation and a new linking procedure for

dwellings. This raised the question as to what extent a counterfactual version of ICP 2005,

adjusted for these methodological changes, would be close to the ICP 2011 results.

The main finding is that making the price and methodology adjustments brings the

extrapolations from ICP 2005 counterfactual and the actual ICP 2011 rounds a lot closer on

average, though country level differences do remain. At the GDP level this gap, on average,

is reduced by 58 percent to 10.0 percent. The average difference in the consumption

aggregate has been reduced by 44 percent to 4 percent. While these results are broadly

reassuring, we find that the main drivers of the original differences are not the

methodological innovations but the price sampling biases. In this paper we have been able to

identify the main sources of bias induced by the use of 18 selected ring countries and the use

of the ring product list in ICP 2005. Our analysis suggests that the use of the global core list

of products and incorporating data from all participating countries in the linking process has

largely eliminated the biases associated with the linking approach used for ICP 2005.

We find that our counterfactual results imply a much lower degree of international income

and consumption inequality than the original ICP 2005 results. From a statistical point of

view our research suggests that constructing counterfactual PPPs is a useful exercise

following major methodological changes. Such counterfactuals can reduce the size of the

revisions following a new PPP release and highlight how much of the revision is due to the

methodological changes and how much due to new measurements in a different year. We

would also argue that our results are a case for more frequent price comparisons. Since a

price comparison in any year will provide results that are measured with a sizeable degree of

error, more frequent measurement will prevent outliers in any one year from distorting

international comparisons of living standards for a substantial period of time. Furthermore,

32

the introduction of new methodologies can more easily be phased in gradually if there is an

ongoing statistical infrastructure for surveying prices, again leading to fewer major revisions.

From a researcher point of view, our results can be seen as broadly comforting. Many

datasets used for cross-country analysis of economic performance rely on a single set of PPPs

to construct relative income levels, which are then extrapolated to other years using national

growth rates. Major differences across PPP sets that are systematically larger (or smaller) in

low-income countries would argue strongly against this practice and would call into question

research based on such datasets, which include the World Development Indicators, the

Maddison data and (parts of) the new Penn World Table. Our counterfactual for 2005

suggests that such revisionism is not necessary. At the same time, the large (idiosyncratic)

differences between the two benchmarks for individual countries suggests that relying on a

single benchmark for country figures is hazardous. Our counterfactual for 2005 provides a

helpful and consistent second point of reference to the new data for 2011.

33

References

Aghion, Philippe, Ufuk Akcigit and Peter Howitt (2014), “What Do We Learn From

Schumpeterian Growth Theory?” in Philippe Aghion and Steven N. Durlauf

(eds.) Handbook of Economic Growth, volume 2, Elsevier: Amsterdam, 515-563.

Almås, Ingvild (2012), “International income inequality: measuring PPP bias by estimating

Engel curves for food” American Economic Review 102(1): 1093-1117.

Anand, Sudhir and Paul Segal (2008), “What do we know about global income inequality?”

Journal of Economic Literature 46(1): 57-94.

ADB (2007), 2005 International Comparison Program in Asia and the Pacific – Purchasing

Power Parities and Real Expenditures, Asian Development Bank, Manilla, Philippines.

ADB (2014), 2011 International Comparison Program in Asia and the Pacific – Purchasing

Power Parities and Real Expenditures, Asian Development Bank, Manilla, Philippines.

Balassa, Bela (1964), “The Purchasing-Power Parity Doctrine: A Reappraisal” Journal of

Political Economy 72: 584-596.

Bolt, Jutta and Jan-Luiten van Zanden (2014), “TheMaddisonProject:collaborative

researchonhistoricalnationalaccounts”EconomicHistoryReview67(3):627‐651.

Chandy, Lawrence, and Homi Kharas (2014), “What do new price data mean for the goal of

ending extreme poverty?” Brookings Institution, May 5,

http://www.brookings.edu/blogs/up-front/posts/2014/05/05-data-extreme-poverty-

chandy-kharas.

Chen, Shaohua and Martin Ravallion (2008), “The developing world is poorer than we

thought, but no less successful in the fight against poverty” World Bank Policy

Research Working Paper no. 4703.

Chen, Shaohua and Martin Ravallion (2010a), “The developing world is poorer than we

thought, but no less successful in the fight against poverty” Quarterly Journal of

Economics 125(4): 1577-1625.

Chen, Shaohua and Martin Ravallion (2010b), “China is poorer than we thought, but no less

successful in the fight against poverty” World Bank Policy Research Working Paper

no. 4621.

Ciccone, Antonio and Marek Jarociński (2010), “Determinants of Economic Growth: Will

Data Tell?” American Economic Journal: Macroeconomics 2(4): 222-246.

34

Dalgaard, Esben and Henrik Sørensen (2002), “Consistency between PPP Benchmarks and

National Price and Volume Indices,” Paper presented at the 27th General Conference of

the International Association for Research on Income and Wealth, Sweden.

Deaton, Angus (2010), “Price indexes, inequality, and the measurement of world poverty,”

American Economic Review, 100(1): 5-34.

Deaton, Angus (2012), “Calibrating measurement uncertainty in purchasing power parity

exchange rates”, Research Program in Development Studies, mimeographed, Princeton.

Deaton, Angus, and Alan Heston (2010), “Understanding PPPs and PPP-based national

accounts,” American Economic Journal: Macroeconomics, 2:4, 1-35.

Deaton, Angus, and Olivier Dupriez (2011), “Purchasing power parity exchange rates for the

global poor” American Economic Journal: Applied Economics 3: 137-166.

Deaton, Angus and Betina Aten (2014), “Trying to Understand the PPPs in ICP2011: Why

are the Results so Different?” NBER Working Papers, no. 20244.

Diewert, W. Erwin (1976), “Exact and superlative index numbers” Journal of

Econometrics4:114‐145.

Diewert, W. Erwin (2009), “Similarity Indexes and Criteria for Spatial Linking” in D.S.

Prasada Rao (ed.) Purchasing Power Parities of Currencies: Recent Advances in

Methods and Applications, 183-216, Edward Elgar, Cheltenham, UK.

Diewert, W. Erwin (2013a), “Methods of aggregation above the basic heading level within

regions” in World Bank (ed.), Measuring the Real Size of the World Economy, Chapter

5, World Bank, Washington DC.

Diewert, W. Erwin (2013b), “Methods of Aggregation above the basic heading level: linking

the regions” in World Bank (ed.), Measuring the Real Size of the World Economy,

Chapter 6, World Bank, Washington DC.

Dykstra, Sarah, Charles Kenny, and Justin Sandefur (2014), “Global absolute poverty fell by

almost half on Tuesday,” Center for Global Development, May 2,

http://www.cgdev.org/blog/global-absolute-poverty-fell-almost-half-tuesday.

Dykstra, Sarah, Benjamin Dykstra and Justin Sandefur (2014), “We just ran twenty-three

million queries of the World Bank’s website” Center for Global Development Working

Papers, no. 362.

Feenstra, Robert C., Robert Inklaar and Marcel P. Timmer (2013), “The next generation of

the Penn World Table” NBER Working Paper no. 19255.

35

Feenstra, Robert C., Hong Ma, J. Peter Neary and D.S. Prasada Rao (2013), “Who Shrunk

China? Puzzles in the Measurement of Real GDP” Economic Journal 123(573): 1100-

1129.

Heston, Alan (2013), “Government services: productivity adjustments” in World Bank (ed.),

Measuring the Real Size of the World Economy, Chapter 16, World Bank, Washington

DC.

Inklaar, Robert (2013), “Price levels and economic growth: making sense of the revisions to

data on real incomes: a comment” Review of Income and Wealth 59(4): 614-622.

Johnson, Simon, William Larson, Chris Papgeorgiou and Arvind Subramanian (2013), “Is

newer better? Penn World Table revisions and their impact on growth estimates”

Journal of Monetary Economics 60: 255-274.

McCarthy, Paul (2013a), “Construction” in World Bank (ed.), Measuring the Real Size of the

World Economy, Chapter 12, World Bank, Washington DC.

McCarthy, Paul (2013b), “Extrapolating PPPs and Comparing ICP Benchmark Results” in

World Bank (ed.), Measuring the Real Size of the World Economy, Chapter 18, World

Bank, Washington DC.

Milanovic, Branko (2012), “Global inequality recalculated and updated: the effect of new

PPP estimates on global inequality and 2005 estimates” Journal of Economic Inequality

10(1): 1-18.

Neary, J. Peter (2004), “Rationalizing the Penn World Table: True multilateral indices for

international comparisons of real income” American Economic Review 94(5): 1411-

1428.

Pritchett, Lant and Lawrence H. Summers (2014), “Asiaphoria meets regression to the mean”

NBER Working Paper no 20573.

Rao, D.S. Prasada (2013), “Computation of basic heading PPPs for comparisons within and

between regions” in World Bank (ed.), Measuring the Real Size of the World Economy,

Chapter 4, World Bank, Washington DC.

Rao, D.S. Prasada and Gholamreza Hajargasht (2014), “ Stochastic Approach to Computation

of Purchasing Power Parities in the International Comparison Program (ICP)”

mimeographed, University of Queensland.

Ravallion, Martin, Shaohua Chen and Prem Sangraula (2009), “Dollar a day revisited” World

Bank Economic Review 23(2): 169-184.

36

Ravallion, Martin (2013), “Price levels and economic growth: making sense of the revisions

to data on real incomes” Review of Income and Wealth 59(4): 593-613.

Ravallion, Martin (2014), “An exploration of the international comparison program’s new

global economic landscape” NBER Working Paper, no. 20338.

Rogoff, Kenneth (1996), “The Purchasing Power Parity Puzzle” Journal of Economic

Literature34(2):647‐668.

Samuelson, Paul A. (1964), “Theoretical Notes on Trade Problems” Review of Economics

and Statistics 46(2): 145‐154.

Van Veelen, Matthijs (2002), “An impossibility theorem concerning multilateral international

comparison of volumes” Econometrica 70(1): 369-375.

Vogel, Frederic A. (2013), “The ring comparison: linking the regions” in World Bank (ed.),

Measuring the Real Size of the World Economy, Chapter 9, World Bank, Washington

DC.

West, Kenneth D., “Forecast Evaluation,” in Graham Elliot, Clive W. J. Granger, and Allan

Timmermann (eds), Handbook of Economic Forecasting, Volume 1, 99-134, 2006.

World Bank (2008), Global Purchasing Power Parities and Real Expenditures. 2005

International Comparison Program, World Bank, Washington DC.

World Bank (2013), Measuring the Real Size of the World Economy, World Bank,

Washington, DC.

World Bank (2014a), Purchasing Power Parities and Real Expenditure of World Economies

– Summary of Results and Findings of the 2011 International Comparison Program,

World Bank, Washington DC.

World Bank (2014b), Purchasing Power Parities and the Real Size of World Economies – A

Comprehensive Report of the 2011 International Comparison Program, World Bank,

Washington DC.

37

Appendix Tables and Figures

Table A1, Share of regions in real GDP, with partial and comprehensive productivity

adjustment of public wages

Partial Comprehensive Africa 3.36 3.38 Asia-Pacific 24.30 24.34 Eurostat/OECD 64.82 65.05 Latin America 5.57 5.26 Western Asia 1.94 1.96

Note: under partial productivity adjustment, only the publicwages in Africa, Asia‐Pacific andWestern

AsiaareadjustedfordifferencesinthecontributiontoGDPofdifferencesinthecapitalperworkerratio.

Undercomprehensiveadjustment,publicwagesinEurostat/OECDandLatinAmericaarealsoadjusted.

Table A2, Harmonizing the treatment of dwellings for linking the regions.

Africa

Asia-Pacific

Eurostat/OECD

Latin America

Western Asia

Dwellings/capita (1) 0.53 0.66 1.00 0.64 0.44 Quality (2) 0.32 0.73 1.00 0.77 0.88 Volume/capita (3) = (1) x (2) 0.17 0.48 1.00 0.49 0.39 Expenditure/capita (4) 0.03 0.16 1.00 0.10 0.28 Price index (5) = (4)/(3) 0.19 0.33 1.00 0.21 0.72 ICP 2005 price (6) 0.21 0.43 1.00 0.34 0.68 Price adjustment (7)=(5)/(6) 0.91 0.77 1.00 0.63 1.06

Note: the elements in lines (1) and (2) are based on an arithmetic unweighted average of the number of

dwellings per capita (line (1)) and of the quality characteristics (the percentage of houses with electricity, water

and a toilet; line (2). The averages are divided by the Eurostat/OECD values to arrive at the figures in the table.

Expenditure per capita (line (4)) is from the basic heading ‘actual and imputed rents’, in exchange-rate

converted US dollars, per capita. The ICP 2005 price is from Heston (2013, p333).

38

Table A3, Population-weighted Gini coefficients for GDP and consumption per capita,

1990-2012

GDP Consumption ICP 2005 ICP 2005C ICP 2011 ICP 2005 ICP 2005C ICP 2011

1990 0.640 0.613 0.611 0.659 0.622 0.6251991 0.638 0.611 0.609 0.658 0.620 0.6241992 0.635 0.608 0.605 0.654 0.615 0.6201993 0.631 0.602 0.600 0.651 0.610 0.6161994 0.628 0.599 0.596 0.651 0.610 0.6161995 0.624 0.593 0.591 0.646 0.604 0.6091996 0.619 0.588 0.586 0.640 0.597 0.6031997 0.618 0.587 0.584 0.640 0.598 0.6021998 0.617 0.585 0.584 0.641 0.597 0.6021999 0.616 0.583 0.582 0.641 0.597 0.6012000 0.614 0.582 0.580 0.641 0.598 0.6012001 0.609 0.576 0.574 0.634 0.591 0.5942002 0.603 0.570 0.568 0.631 0.588 0.5902003 0.597 0.563 0.561 0.627 0.583 0.5852004 0.590 0.556 0.554 0.624 0.579 0.5822005 0.583 0.548 0.545 0.618 0.573 0.5762006 0.574 0.539 0.536 0.613 0.568 0.5702007 0.564 0.528 0.525 0.600 0.553 0.5572008 0.555 0.519 0.516 0.592 0.545 0.5482009 0.539 0.502 0.499 0.577 0.528 0.5322010 0.531 0.494 0.491 0.570 0.522 0.5252011 0.525 0.489 0.484 0.562 0.514 0.5162012 0.520 0.485 0.480 0.557 0.509 0.510

39

Appendix Figure A 1, Lorenz curves for ICP 2011, and extrapolations from ICP 2005

and ICP 2005C

40

APPENDIX: AN ALTERNATIVE COUNTERFACTUAL FOR ICP 2005

In the main text of the paper we presented a counterfactual for ICP 2005 which adjusts for the

methodological innovations in ICP 2011 and also address concerns on possible ring country

selection and ring product price biases arising out of the use of 18 selected ring countries for

the purpose of linking regional comparisons. This appendix describes an alternative which

may be described as a conservative counterfactual. As discussed in the main text, we use

detailed item-level prices to establish the presence of product selection bias in 2005 (for

Africa, Asia-Pacific and Western Asia) and in 2011 (for Asia). In our preferred

counterfactual, ICP 2005C, we used evidence from Deaton and Aten (2014) on changes in

relative prices for the 2005 ring countries compared with national inflation trends in making

adjustments for product selection bias in 2005. We believe this combination of cross-country

prices and national price trends allows for a more comprehensive estimate of the product

selection bias.

The alternative, which we explore here, would be to directly use the biases implied by the

regression coefficients reported in Tables 4 and 5. This approach leads to downward

adjustments of 23.1 percent in the Asia-Pacific; 9.4 percent in Africa and 6.6 percent in

Western Asia. In both the approach and results, these adjustments are more conservative than

the 25 percent adjustment in all three regions suggested in the Deaton-Aten analysis. As

before, these adjustments are implemented only for the low-income countries in the region.

We now consider the features of this alternative counterfactual (ICP 2005C2) compared to

our preferred counterfactual (ICP 2005C).

Table A4, Differences between ICP 2005 and the two counterfactuals and ICP 2011

ICP 2005 ICP 2005C ICP 2005C2 GDP Mean difference 0.237*** 0.100*** 0.149*** Root mean squared difference 0.336 0.218 0.256 Coefficient on log(expenditure/capita)

-0.020* 0.016* -0.003 (0.012) (0.009) (0.010)

Consumption Mean difference 0.254*** 0.040*** 0.116*** Root mean squared difference 0.349 0.168 0.231 Coefficient on log(expenditure/capita)

-0.063*** 0.000 -0.038*** (0.011) (0.008) (0.009)

Note: The first column (ICP 2005) is from Table 1, the second column (ICP 2005C) is from Table 9, and the third column (ICP 2005C2) is based on the more conservative counterfactual discussed in this appendix. Robust

41

standard errors of the regression coefficients shown in parentheses below the coefficients. * denotes a variable significantly different from zero at a 10%-level, ** at 5%-level, *** at a 1%-level. As shown in Table A4, the smaller adjustments for product selection bias in our more

conservative counterfactual translates to larger mean differences and root mean squared

differences compared to our preferred counterfactual. However, compared with the original

ICP 2005 differences, the reduction is still large. A similar result can be seen in Table A5,

where the population-weighted Gini coefficients for the year 2011 are compared. By this

measure, inequality according to ICP 2011 and our preferred counterfactual are very similar,

as discussed in the main text. Inequality according to the alternative counterfactual is clearly

higher than in these other two cases, but still notably closer to the ICP 2011 Gini coefficient

than to the original ICP 2005 Gini.

Table A5, Population-weighted Gini coefficients for 2011 based on alternative PPs

GDP ConsumptionICP 2005 0.525 0.562ICP 2005C 0.488 0.513ICP 2005C2 0.495 0.526ICP 2011 0.484 0.516