Speed Read
  • The quality of economic data across the Asian & Pacific region varies considerably across countries, mainly as a result of lack of statistical capacity, making comparisons of GDP of dubious value.
  • Many countries have seriously out of date base years and it is estimated that bringing these up to date could add 1.2% on average to existing levels of GDP, equivalent to nearly US$0.54 trillion.
  • Most Asian & Pacific countries use a national income accounting standard which does not record the informal economy which could account for an estimated additional average of 28.0% of GDP across the region.
  • China’s economic data faces a large problem of credibility because of the lack of independence of its statistical office and a number of anomalies between published data and externally estimated data.

Asian & Pacific Economic Data: The problems

In this paper the quality and the reliability of official national income accounting data is investigated for 45 countries in the Asia & Pacific region.[1]

1. Resources Available to Measure National Income

The quality of economic statistics in the Asian & Pacific region varies considerably across countries. There are several reasons for this, but one of the most important is the differences in the resources available to national statistical offices for the collection, processing and analysis of economic statistics on output and prices. There is a global standard set by the United Nations for measuring national income,[2] but poorer countries generally report poorer quality statistics because of their lower capacity in terms of resources and trained staff in national statistical offices to implement best practice and the paucity of comprehensive household and business surveys available.[3] Four of the countries analysed are higher income OECD members – Japan, Australia, New Zealand and Korea who are part of an organisation committed to best practice in the provision of economic statistics, but the others differ significantly in their level of economic development. Given the differences in per capita income across the region outside of this group ranging in purchasing power parity (PPP) from US$82.8 thousand in Singapore in 2014 to US$2.4 thousand in Nepal some differences in the quality of economic statistics would be expected. All the GDP PPP data are from the World Bank’s WDI database[4] apart from the GDP PPP estimates for Myanmar/Burma, New Caledonia, and Taiwan, which are taken from the CIA World Factbook.[5]

Morton Jerven has undertaken extensive research into the negative relationship between the level of GDP per capita in African countries and the quality of national statistics. He cites a response by the IMF to one of his enquiries which encapsulates the basic problem:

“…there is a need to strengthen source data for national accounts. However, these countries are poor with inadequate revenue to fund regular ongoing data collections for NAS[6] purposes. In general, there is a need to strengthen data collection for agriculture, fishing, informal sector and services activities. There is also scope to improve prices collections (agriculture, producer prices). Compilation staff and resources are also constrained, due to budget, limiting the range and quality of the statistics produced. Existing staff need further training and development. The constraints in compilation are reflected in the limited dissemination.”[7]

This problem of statistical capacity has a serious impact on the quality of economic statistics across a significant portion of the Asian & Pacific region where out of the 45 countries analysed, 25 of them had a GDP per capita below US$10,000 in 2014. This set of nations also includes some of the most populous countries in the region: India, Pakistan and Bangladesh, which collectively in 2014 accounted for 40.7% of the region’s total population and 20.1% of total GDP. China, with a GDP per capita of US$13,206 lies just outside of this arbitrary threshold, but also faces the problem of scarce resources devoted to statistical capacity relative to the size of the economy. Limited statistical resources present the most difficult measurement issues for geographically large, populous, developing nations with serious pockets of poverty and much informal unrecorded economic activity.

India, the second largest economy in Asia, for example, has a number of well-documented economic data deficiencies. Official GDP statistics are revised frequently, just as in developed economies, but often as result of faulty reporting rather than the accretion of more information. To give a recent example, in June 2013, the Central Statistics Office, revised Industrial Production data twice in a week.[8] Apart from frequent clerical errors Indian statisticians are only able to sample a very small unrepresentative proportion of the firms and agricultural enterprises operating in the country and estimates are likely to be biased. There has also been problems for many years with the measurement of inflation. Since 1947 the Reserve Bank of India has relied on a Wholesale Price Index (WPI) to target inflation despite its defects including the omission of services prices, the collection of price data at different stages of the production process and the high volatility of the resultant index values. These issues have been reported by Sturgess (2011), but although India launched a new Consumer Price Index this has not replaced the WPI because of continuing reliability issues.

Other significant data problems found across the Asian & Pacific region, which affect inter-country comparisons, are:

  1.   the failure to update base years;
  2.  the use of outdated standards of national income accounting; and
  3.  the degree to which the shadow economy (including informal activities, illicit activities and the income derived from outright crime) are under-recorded.


These problems can all be rectified by increasing the resources available for national income accounting exercises, but there is also a specific problem within the region: the extent to which China’s economic data is subject to political manipulation remains a live and continuing issue. These problems distort the measurement of the relative size of national economies and the validity of inter-country comparisons of living standards and growth rates across the Asian & Pacific region, are discussed in more detailed in section 5 below.


2. Base Year Problem

Real growth in the activity of an economy is estimated by comparing GDP at current and at constant prices. Constant price estimates use the price relatives of a particular year, known as a base year or benchmark year, to weight the volume components of production. But since the structure of production and relative prices over time are dynamic, the pattern of relative prices and the industries surveyed in the base year become less relevant over time. Even when the economic data produced by national statistical offices are published regularly serious distortions arise from the use of outdated base years. This biases estimates of the relative size of economies and the speed at which they are growing in real terms. This means that the estimates of GDP used in country rankings are likely to be seriously inaccurate with the degree of bias for each country being dependent on the length of time since the last base year was updated by the national statistics office.


The time elapsed between current estimates of GDP and the base year is of crucial importance to the accuracy of national income data. According to Jerven (2013) failures to regularly update a base year in which the structure and relative prices of an economy are monitored can lead to serious bias in the estimation of GDP. He refers to the case of Ghana when estimated GDP was uplifted by 60% overnight as a result of an update in the base year from 1993 to 2006. In 2014, the Nigerian National Bureau of Statistics released new estimates of GDP showing a rise of 59.5% after the economy’s base year was updated from 1990 to 2010. One of the reasons for these large uplifts is that better estimates of some economic activities such as telecommunications, tourism and financial services were made by statisticians while some previously unrecorded activities from the informal sector were included.

Table 1 shows the results of a sample of recent rebasing exercises for 5 Asian & Pacific countries. A scatter graph of this data shows that there is a strong positive relationship between the numbers of years since the last rebasing exercise and the size of GDP uplift (See Figure 1).

Uplifts in the measurement of GDP have been found among some of those countries in the Asian & Pacific region that have recently updated their base years: India, Pakistan, Sri Lanka, Thailand, and Vietnam. GDP generally increases under a new base year by the addition of the measurement of new or under-recorded elements in the economy.

However, a positive uplift effect is not always guaranteed. In 2013, a rebasing exercise carried out by the Pakistan Bureau of Statistics (PBS)[9] resulted in an additional 7.8% increase in estimated value added for the year 2012-13 equivalent to US$5.7 billion when the base year was updated from 2000 to 2006. But while agriculture and the service sectors’ shares of GDP both rose as a result of the exercise from 20.3% to 23.0% and 52.8% to 56.0% respectively, the estimate of industrial output actually shrank by US$3.13 billion in value. This was due to contractions in the scale of measured output of large and small scale manufacturing which caused the industrial sector to fall by 20.9% of GDP.[10]Neither are the uplifts always large as a result of rebasing. In Sri Lanka, which updated its base year from 1998 to 2002, GDP was revised upwards by only 4.2%, while in India, GDP actually shrank by 2.0% after rebasing from 2004 to 2011 [See Table1].

The best current international practice is to update a country’s base year every five years to minimise sudden changes in a GDP series, but OECD countries now adopt the practice of chaining where price relatives are updated every year. This was one of the recommendations of SNA 1993. Chaining also allows continual updates to be made to the structure of production and consumption by adjusting for the actual goods produced, but it requires considerable expenditure on resources by statistical offices. Only six countries of the richer countries use chaining in the region: Australia, Japan, Hong Kong, Korea, and New Zealand.  

Table 1: Results of Rebasing in Asia Pacific


Figure 1: Asia-Pacific % GDP Uplift and Years since last Rebasing

Reassessing GDP in Asia & Pacific

In this section we report on an exercise to estimate what the size of Asian & Pacific region’s GDP might be if most countries updated their base years to 2013. When this exercise was carried out for Africa and the Americas it was estimated that GDP might be about 21% and 5% higher respectively. Given the inadequate resources allocated to national income accounting in some of the poorer countries across the world, a rebasing exercise, often carried out with external technical and financial support, usually does far more than the updating relative price weights used to estimate real GDP. Some sectors of the economy are measured more accurately using better survey methods, while some economic activities are monitored that were not previously monitored such as mobile telephony and media services. This explains the 60% uplift experienced by Ghana.

The methodology we use to estimate what would happen if all Asian & Pacific countries rebased up to 2013 is crude, but simple and involves applying an estimated constant cumulative annual rate of growth to the years between the last reported base year and 2013.The longer the period between 2013 and the last base year and the higher the assumed cumulative growth rate, the higher the uplift that would be expected from rebasing. The resultant country uplift factors are then applied to the 2013 GDP PPP country data for the Asian & Pacific region (See Table 3). Analysing the existing economic data shows that aggregate GDP in this region may be underestimated and could be around 1.1% higher.

Two assumptions were necessary to carry this exercise out


(1)     First the choice of a base year data series to apply on a country by country basis, a task made problematic given the inconsistencies between the WDI, the IMF, and the UN.


(2)     The second assumption is to estimate the cumulative annual rate which would act as a proxy by replicating over time the effect of the uplift created in one year by rebasing GDP (See Figure 1).


Inconsistent Base Year Data:
Many countries in the Asian & Pacific region have not even been able to update base years to update base years every five years, but calculating how out-of-date national economic data for each country is, is hampered by the inconsistencies in the publicly available information on base years from three main international sources. These include, the World Bank, IMF, and UN. Alternative sources are the base year recorded by each country’s national statistical offices (NSOs). The reported base years from each of these sources are shown by country in Table 2. Given the differences in reported base years, the source(s) generally used to calculate how many years the latest base year was out-of-date was highlighted in bold.

The number of years each country’s base year is out of date is also shown in Table 2 for the 45 countries in the Asian & Pacific region. These are calculated as the difference between the last reported base year and 2014. The average time elapsed between the base year and 2014 was just above six years. However, fourteen countries had a base year that was 10 and more years out of date: Afghanistan (2002), Bhutan  (2000), Cambodia (2000), Laos (2002), Maldives (2003), Marshall Island (2003), Micronesia (2004), Nepal (2000), New Caledonia (1990), Papua New Guinea (1998), Philippines (2000), Solomon Islands, and Sri Lanka (2002).

 Table 2: Reported Base Years by Source


Cumulative annual rate: The assumed annual rate that would act as a proxy by replicating over time the effect of the uplift created in one year by rebasing GDP was 0.37% for the Asian & Pacific countries. This was based on an econometric model using data from the sample of 5 GDP national rebasing exercises, shown in Table 1 and Figure 1.


Results: Analysing the existing economic data shows that the aggregate Asian & Pacific region’s GDP is underestimated by the World Bank and it could be around US$0.54 trillion higher. In 2014, the sample of 45 Asian & Pacific countries had a combined GDP of US$43.6 trillion measured in PPP international dollars. Applying the uplift estimate discussed in assumption (2) above produces a revised figure for aggregate GDP of US$44.1 trillion, an average uplift of 1.2%. However, this is misleading since it includes Australia, Hong Kong, Japan, Kazakhstan, Korea, Kyrgyzstan, New Zealand, Tajikistan, Thailand and Uzbekistan (0% uplift) as these countries have adopted chaining which is in effect rebasing every year. The uplift estimate also produce differences in the size rankings of Asian & Pacific countries and the results are shown in Table 3 which displays each country ranked by current GDP and rebased GDP.

 Table 3:  Rebased GDP data and % Uplift

3. Outdated National Accounting Standards

The harmonized System of National Accounts (SNA) was created by the international community to facilitate the comparability of economic statistics and standards. Since 1953 there have been three revisions to recommended SNA standards, in 1968, 1993, and in 2008, all approved by the United Nations Statistical Commission. Most of the Asian & Pacific countries analysed have now adopted SNA 1993 while Australia, Hong Kong, India, Indonesia, Malaysia, New Zealand, Pakistan, Philippines, Singapore, South Korea, Taiwan and Timor-Leste have adopted the latest SNA or 2008 standard (See Table 4).

These standards are not mandatory and the adoption by a country of a standard such as SNA 1993 does not imply that all of the recommendations are implemented. The near general use of SNA 1993 means that, apart from the individual discrepancies across countries mentioned above, economic statistics across the Asian & Pacific region are broadly comparable in terms of the definitions used and accounting methodologies applied. However, the United Nations Statistical Commission formally adopted the SNA 2008 standard in 2009.[11] Significant differences between SNA 1993 and SNA 2008 include the treatment of government accounts, capital expenditure, intellectual property and improved measurement of the informal sector and illegal activities, areas which are particularly important in emerging markets. The longer it takes the countries of the Asian & Pacific region to adopt this standard the less reliable will be the economic comparisons between themselves and the developed world.

Table 4: Accounting Standards by Country



4.  The Shadow Economy Problem

All economies have many unrecorded economic transactions which bias downwards official estimates of GDP, employment and income per head. The SNA 1993 national income accounting standard recognised the importance of the informal economy particularly in developing economies defining it as belonging to the household sector, but there was no explicit methodological recommendation as to how to measure its size.

The SNA 2008 standard goes a lot further in defining the informal sector and in placing the onus of measurement on national statistical offices adopting the standard to include estimates within the framework of national income accounting. The SNA 2008 standard recommends a number of methods of estimating the informal economy, although at present without a set of full regular surveys researchers rely on the work of Freidrich Schneider a leading expert on the size and importance of shadow or grey economic activity across the world. In his latest estimates for 2007 which covers 31 of the countries investigated in this analysis, Table 5 shows the relative size shadow economies across the region ranged from 10.3% of GDP in Japan to 48.2% in Thailand with an average level of 28.0%.[12]


Table 5:  The Shadow Economy % of National Income


5. The Chinese Credibility Problem

The quality of national income statistics depends on relevant theoretical constructs, the application of consistent national accounting standards and the accuracy of the sampling methods employed. It is important to recognise, however, that all GDP data is ‘man-made’, open to manipulation and often subject to revision from preliminary estimates as more information is received and analysed. According to Crabbe (2014):

“It is that China (a country of 1.3 billion people) releases its quarterly GDP data two weeks after each quarter end, and never revises that data, Hong Kong (population 7 million) takes six weeks, while the US (population 314) takes eight weeks to publish quarterly data, and provides constant revisions.” [13]

China has attracted much controversy about the reliability of its economic data. From the 1990’s onwards, as China’s significance in the global economy grew, concern about the accuracy of official data produced a debate about the relative size of the Chinese economy and the velocity and accuracy of its reported real growth rate following the market-led reforms of the late 1980s.[14] There have also been disagreements about the accuracy of the purchasing power parity (PPP) price indexes used as converters instead of market exchange rates to compare the relative size and growth rates of China in relation to the US and the rest of the world.[15]

China’s official economic statistics have suffered for some time from a problem of reliability. A recent US government report published in 2013 concluded:

“China’s official statistics are not as reliable as those produced in the United States and Europe. Several findings support this conclusion. The first is that there are serious deficiencies in the way the Chinese government gathers, measures, and presents its data. Although China’s National Bureau of Statistics (NBS) now uses sample surveys to measure the economy, survey coverage remains incomplete, particularly in services and the private sector. Economic censuses, in turn, prompt inordinately large revisions of statistical data, and are themselves not on par with international standards. At the same time, many industrial enterprises still report their output directly to the government, keeping in place a Soviet-style reporting system based on state-owned enterprises.”[16]

Holz (2013) blames the problem on the institutional arrangements of data compilation in China since the NBS has no degree of formal independence and there is no transparency in how the NBS compiles data. Furthermore, the limitations of publicly available data do not allow independent checks on the accuracy of the official data. He notes that the NBS has scope for intentional manipulations in order to achieve a politically desirable GDP nominal or real GDP growth rate. It can “falsify” nominal data by expanding its data compilation method and can falsify real growth rates through its choice of deflation methodology, as well as through its choice of retrospective revisions to nominal values and implicit deflators in benchmark revisions.

However after applying a statistical test of whether or not Chinese data issued by the NBS conforms to normal data regularities, Holz (2013) concluded:

Reviewing past and ongoing suspicions of the quality of Chinese GDP data, one is hard pressed to find evidence of data falsification.”[17]

Nevertheless, there are a number of commentators who still question the accuracy of Chinese official statistics. Suspicion of the quality of Chinese economic statistics has attracted domestic criticism. In 2007 the current Prime Minister Li Keqiang is alleged to have said that he did not trust “man-made” figures when examining the Chinese economy. At that time he was head of the Communist Party in Liaoning Province and considered electricity consumption, rail freight and bank loans to be more accurate measures of economic activity than GDP. The Economist[18] created a “Keqiang index “based on a weighted average of these three indicators which showed that economic activity appeared to be far more volatile than was evidenced in the official statistics produced by the National Bureau of Statistics (NBS).[19]

One officially recognised domestic problem is the existence of large gaps between central and provincial estimates of economic activity which has been called the “phantom province” problem. As Crabbe (2014) explains the “province” results from the continued divergence between the central government’s GDP estimate and the sum of the estimates of China’s 31 provincial and centrally run municipal governments. The sum of provincial GDP estimates regularly exceeds the national figures issued by the NBS. For example, in 2013 the NBS estimated national GDP at 56.9 trillion yuan (US$9.3 trillion), while the aggregate of all but three of the independently released provincial estimates of economic activity was nearly 2.06 trillion yuan higher. This discrepancy has been growing over time. According to data provided by Crabbe (2014) the divergence rose from 2.5% in 2004 to 10.0% by 2012. In that year the discrepancy reached Renmimbi (RMB) 5.76 trillion equivalent to the size of Guangdong province.

The deliberate falsification of local economic data reported centrally by overzealous or fearful officials has historical roots, but the current explanation is that the discrepancies are based less on dissimulation than on errors resulting from disjointed and un-coordinated accounting. According to Zhang Liqun at the Development Research Centre of the State Council the problem is largely due to the double counting of the turnover and activities of large corporations whose subsidiaries operate across Chinese provinces.[20]

A more serious problem noted by Balding (2014) in World Economics is the manipulation of inflation statistics, particularly house prices and the cost of renting through reducing their weights in the calculation of the Consumer Price Index (CPI) and by under-recording price rises. The impact of this, it is alleged, systematically biases the estimated inflation rate downwards which has the effect of raising real GDP. Balding adjusts China’s official inflation data to take account of his methodological criticisms and finds that over the period 2000 to 2011 real GDP should be reduced by up to 8-12%.



Most of the Asian & Pacific region uses a consistent, but outdated national accounting methodology, which means that comparing their economies with OECD countries will cause problems as the latest SNA 2008 is progressively adopted across the developed world. The information already available on estimates of the size of the shadow economy economies (See Table 5) suggests that significant additional upward revisions might be expected when Asian & Pacific countries adopt the SNA 2008 themselves since the new standard requires greater efforts to be devoted to measuring informal and illicit activities.

Second, the long overdue rebasing of some Asian & Pacific countries means that GDP is being underestimated by around 1% across the region and that the rankings of some countries by GDP per capita within the region must be reassessed. However, although researchers face nothing like the scale of problems in interpreting African national accounting statistics much still needs to be done by some statistical offices that have not rebased for some time. Accurate and transparent statistics are essential indicators of economic potential and if the poorer economies wish to continue to attract rising investment interest, the issue of data quality needs to be urgently addressed.

The lack of reliable recent economic data is being addressed by World Economics which produces monthly surveys (Sales Managers Indexes) of business conditions across Asia[21]and for a number of individual countries: China, India, Mongolia and the Philippines. Indicators of the speed and direction of economic activity are available for these areas. In addition to a headline figure the Sales Managers Indexes (SMIs) also monitor Business Confidence, Market Growth, Sales Growth, Prices Charged and Staffing Levels in separate diffusion indexes. The SMI are published monthly providing data relating to the current month and are available on the World Economics ( website.


Benford, Frank. (1938) “The Law of Anomalous Numbers.” Proceedings of the American Philosophical Society 78, no. 4 (31 March): 551-72.


Dikhanov, Y., and Swanson, E.V. (2010), “Maddison and Wu: Measuring China’s Economic Performance”, World Economics, Vol. 11 (1), January-March: 199-203.


Holz, C. (2004), “Deconstructing China's GDP Statistics”, China Economic Review, Vol. 15, pp. 164-202. 


Holz, C. (2007), “China’s 2004 Economic Census and 2006 Benchmark Revision of GDP Statistics: More Questions than Answers.” Hong Kong University of Science and Technology, Discussion Paper,


Holz, C. (2013), “The quality of China’s GDP statistics”, Conference Proceedings.


Jerven, M. (2011), “Counting the Bottom Billion: Measuring the Wealth and Progress of African Economies”, World Economics:


Jerven, M. (2013), “Poor Numbers: How we are misled by African Development Statistics and what to do about it”, Cornell.


Maddison, A., and Wu, H.X. (2008), “Measuring China’s Economic Performance”, World Economics, Volume 9 (2), April-June: 13-43.


Schneider, F. (2013), “The Shadow Economy and Shadow Economy Labour Force. What do we (not) know?,” World Economics:


Sturgess, B. T. (2011), “Indian Wholesale and Consumer Price Indices”, World Economics, September:


Sturgess, B.T. (2013), “Emerging Market Data Cannot be Trusted”, World Economics, August:


USCC (2013), “The Reliability of China’s Economic Data: An Analysis of National Output, U.S.-China” Economic and Security Review Commission Staff Research Project:'sEconomicData.pdf

[1] A number of countries were excluded as there was no base year information available: America Samoa, Korea, Dem. Republic, and Northern Mariana Islands. French Polynesia was also excluded as no GDP PPP data was available for 2013.

[2] The United Nations System of National Accounts (SNA):

[3] See Sturgess (2013)

[6] National Accounts Statistics

[7] See Jerven (2011).

[11] In the U.S. some of the provisions of the SNA 2008 standard which led to an upward revision of GDP in September 2013 backdated to 1929 as a result of changes such as the inclusion of R&D investment expenditure and the change in the accounting for defined contribution pension plans from a cash base to an accrual basis. See:

[12] No Shadow Economy data was available for the following countries: Afghanistan, Guam, Kiribati, Marshall Island, Micronesia, New Caledonia, Palau, Samoa, Timor-Leste, Tonga, Turkmenistan, Tuvalu, Uzbekistan, and Vanuatu.

[13] See Holz (2004), (2007).

[14] See Maddison and Wu (2008) and Dikhanov and Swanson (2010).

[15] See USCC (2013).

[16] See Benford’s Law see Benford (1938).

[17] See Holz (2013) pp.27.