Download China`s gross domestic product estimation

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Production for use wikipedia , lookup

Long Depression wikipedia , lookup

Nominal rigidity wikipedia , lookup

Đổi Mới wikipedia , lookup

Economic calculation problem wikipedia , lookup

Gross fixed capital formation wikipedia , lookup

Consumer price index wikipedia , lookup

Transcript
China Economic Review 15 (2004) 302 – 322
China’s gross domestic product estimation
Xianchun XU *
Department of National Accounts, National Bureau of Statistics of China, 75 Yuetan Nanjie, Xichengqu,
Beijing 100826, China
Accepted 4 September 2003
Abstract
This paper describes the process of establishing and improving Chinese gross domestic product
(GDP), estimation data information and dissemination procedure, main data sources, and calculation
methods and presents the main supplementations and adjustments to the historical data. The paper also
analyzes recent reforms adopted by the National Bureau of Statistics (NBS), discusses some of the
remaining problems, and offers a personal view and ideas for further reform of Chinese GDP estimation.
D 2003 Elsevier Inc. All rights reserved.
Keywords: GDP; Calculation methods; Reform measures
1. Establishment and development of China’s gross domestic product (GDP)
estimation
1.1. Establishment of China’s GDP estimation
National income, the key indicator in the Material Product System (MPS) by the Soviet
Union and East European countries has been long adopted in China, from its founding to
the early stage of its reform and opening up to the outside world. This statistic played a key
role in reflecting the activities of material production industries, that is, agriculture,
forestry, husbandry and fishing; mining and quarrying, manufacturing, electricity, gas, and
water; construction; wholesale, retail trade, and catering; and transport, post and telecommunications for the planned Chinese economy. However, it was unable to reflect the
activities of nonmaterial services industries.
Since China’s reform and opening up, nonmaterial service industries such as banking
and insurance, education, science and research, and consulting have developed rapidly,
* Tel.: +86-10-68570027 (O), +86-10-62647804 (H); fax: +86-10-68512445.
E-mail address: [email protected] (X. Xu).
1043-951X/$ - see front matter D 2003 Elsevier Inc. All rights reserved.
doi:10.1016/j.chieco.2003.09.021
X. Xu / China Economic Review 15 (2004) 302–322
303
playing an increasing role in the national economy. The macroeconomic management
agencies demanded a well-designed description of the economic situation to promote the
development of service industries and to coordinate different industries. To meet those
demands, the National Bureau of Statistics of China (NBS) began studying the GDP in the
United Nations System of National Accounts (SNA) beginning in the early 1980s. On
March 19, 1985, the NBS submitted a ‘‘Report on Establishing Tertiary Industry
Statistics’’ to the State Council (Administration Office of State Council, 1988). This
document summarized the necessity of carrying out statistics on tertiary industry and GDP
estimation1 in China. The State Council later approved the report. On April 5, 1985, the
State Council released a notice requiring GDP estimation in addition to national income
estimation. In line with the report, the NBS established a system for GDP estimation at the
national and provincial level that same year (National Bureau of Statistics, 1985).
1.2. Developmental of China’s GDP estimation
1.2.1. From production-side estimation only to both production- and expenditure-side
estimation
At first, Chinese GDP was only estimated from the production side. The NBS adopted
the expenditure approach for GDP estimation as an experiment in 1989, and then formally
in 1993.
1.2.2. From indirect to direct estimation
At first, production-side GDP estimates were not calculated independently but derived
from national income estimates by making the relevant adjustments and supplements. This
required deducting the consumption of nonmaterial services from the net output of each
material production industry and then adding the depreciation of fixed assets to obtain
value added for each material production industry. Then the value added of each
nonmaterial service industry was estimated by processing the final accounting data of
revenue and expenses from fiscal authorities, data about financial statements gathered
from financial institutions, tax data from the taxation authorities, data on wages and
employment of nonmaterial production industries from the NBS, etc. GDP estimates were
obtained by summing the value added for all industries. In 1992, the NBS switched to
direct estimation from raw data sources (National Bureau of Statistics, 1990; 1993).
1.2.3. From an ancillary indicator to a key indicator
Although it did not cover nonmaterial service industries, national income was a key
indicator, whereas GDP served as a supplemental indicator. Concurrent with reforms and
increasing openness, the agencies of macroeconomic analysis and management paid more
attention to GDP estimation. The NBS accordingly reinforced GDP estimation, making it the
main indicator. In 1993, national income estimation was officially cancelled and the GDP
became the key indicator of Chinese economic activity (National Bureau of Statistics, 1992).
1.2.4. From annual estimation only to both annual and quarterly estimations
Since 1992, the NBS has conducted quarterly GDP estimation. Quarterly GDP is
estimated for eight industries: agriculture, forestry, husbandry and fishing; mining and
304
X. Xu / China Economic Review 15 (2004) 302–322
quarrying, manufacturing, electricity, gas, and water; construction; transport, post and
telecommunications; wholesale, retail trade, and catering; banking and insurance; real
estate; others (National Bureau of Statistics, 1997b).
1.2.5. Adjustment and further industry breakdown
In 1995, according to the new standard of national economic industrial classification,
the NBS adjusted the industry classifications for annual GDP estimation. The original 12
industries (agriculture, forestry, husbandry and fishing; mining and quarrying, manufacturing, electricity, gas, and water; construction; transport, post and telecommunications;
wholesale, retail trade and catering; services for household and consultation; public
utilities; banking and insurance; real estate; science, education, culture, health care, sport
and social welfare; government agencies, parties and social organizations; others) were
further disaggregated to 16 industries: agriculture, forestry, husbandry and fishing;
mining and quarrying, manufacturing, electricity, gas, and water; construction; services
for agriculture, forestry, husbandry and fishing; geological prospecting and water
conservancy; transport and storage; post and telecommunications; wholesale and retail
trade and catering; banking and insurance; real estate; social services; health care and
social welfare; education, culture and arts, radio, film and television; scientific research
and technological services; government agencies, parties and social organizations; others
(Xu, 2001).
1.2.6. Standardization of data sources and calculation methods
In 1995 the NBS complied and published The Estimation Methods of China’s
Annual Gross Domestic Product, The Estimation Methods of China’s Quarterly Gross
Domestic Product, and the Manual of China’s Gross Domestic Product Estimation.
Each has standardized the data sources and calculation methods for step-by-step GDP
estimation (Xu, 1999a).
2. Data formation and dissemination procedure of annual Chinese GDP estimation
Data formation proceeds in three stages: preliminary estimation, primary revision, and
final revision (National Bureau of Statistics, 1997a; Xu, 1998a).
2.1. Preliminary estimation and data dissemination
Preliminary estimation occurs early in the following year. Because the data sources
for annual GDP estimation are limited at that time, the preliminary estimation is carried
out mainly using preliminary statistics in specific fields (rapid reports) from different
NBS departments. This results in preliminary estimation, which meets the need for
preliminary analysis and judgement of the macroeconomic situation, awaiting revision
by more reliable and more encompassing data sources. The preliminary estimation
figures of annual GDP are released in China’s Statistical Communique in February of
the following year and in China’s Statistical Abstract published in the first half of the
following year.
X. Xu / China Economic Review 15 (2004) 302–322
305
2.2. Primary revision and data dissemination
Primary revision occurs in the second quarter of the following year, when annual
statistical data collected by the NBS, annual statistical data by related ministries of
the State Council, and part of the administrative records become available. However,
most administrative records and data on final accounts are still not available at that
time. The unavailable records include final fiscal accounts, financial accounts of
institutions in banking, insurance, railway, civil aviation, post and telecommunications, etc. Therefore, the primary revised GDP figures are more accurate than the
preliminary estimates, but still need further revision due to the unavailability of a
number of important data sources. The revised figures at this time stage are
generally released in the China Statistical Yearbook in the second half of the
following year.
2.3. Final revision and data dissemination
Final revision is made in the fourth quarter of the following year, when all of the
statistical, accounting, and administrative data for annual GDP estimation are available.
The final revised figures based on the comprehensive and detailed data sources are
more reliable than the previous ones. The figures are released in the China Statistical
Abstract and the China Statistical Yearbook in the second year following the
accounting year.
2.4. Adjustments to the historical data and their dissemination
In addition to preliminary estimation, primary and final revision, fixing GDP in
historical terms requires adjusting the historical data. Historical GDP data are adjusted
when one or more of the following conditions arise:
New data sources are found or generated, where the data differ from the original.
Changes are made to GDP-related classifications, e.g., changes in the classification of
industries or the classification of final expending.
Basic concepts, accounting principles, or calculation methods change significantly.
Changes in these areas often lead to changes in the total value, structure, or growth rate
of GDP. The data would lose comparability if the adjustments were not made to the
historical data. The time span for adjusting historical data depends on the interval over
which GDP estimation has been affected.
3. Important supplementation and adjustment of historical data
Accounts have made two important supplementations and one important adjustment to
the historical data.
306
X. Xu / China Economic Review 15 (2004) 302–322
3.1. Two important supplementations
Since the initiation of GDP estimation in 1985, China has made two important
supplementations to its historical data to maintain continuity and comparability of data
for microeconomic analysis and management. The first was made in 1986 –1988 to the
data ranging from 1978 to 1984, the second in 1988– 1997 to the data from 1952 to 1977.
Both made the same supplementations, including GDP estimation from both production
and expenditure sides.
The same methodology has been adopted for the two supplementations. Productionside GDP estimates are measured by the following steps: First, adjustments are made to
net output in the MPS in five material production industries—agriculture, forestry,
husbandry and fishing; mining and quarrying, manufacturing, electricity gas, and water;
construction; transport, post and telecommunications; wholesale, retail trade, and
catering. Value added is obtained for each industry separately by deducting the payment
on nonmaterial services (e.g., expenditures on banking and insurance services, advertisement, consultation, etc.) and adding the depreciation of fixed assets. Next, the value
added is calculated for each nonmaterial service industry. Finally, GDP is obtained by
adding the value added of all material and nonmaterial production industries. GDP from
the expenditure side is measured in the following way: First, calculate household
consumption expenditure and government consumption expenditure by deducting the
value of materials as intermediate consumption in nonmaterial production industries
from household consumption and government consumption from national income and
adding the household and government total express on services provided by these
nonmaterial production industries, respectively. Then calculate gross fixed capital
formation in GDP by adding depreciation of fixed assets to net fixed capital formation
in national income. Third, obtain net exports in goods and services by adding imports
and exports of nonmaterial services to imports and exports for both goods and services
in national income, respectively.
Figures from the first supplementation in brief have been published in the China
Statistical Yearbook 1998. The Gross Domestic Product of China, 1952 –1995 combines
the detailed data for both supplementations.
3.2. One important adjustment
The first important adjustment to GDP was completed in 1994– 1995, shortly after the
first census of tertiary industries in China.
China had long been focusing on material production and the MPS regardless of
statistics on the production of nonmaterial services. From the first years of GDP
estimation, data sources concerning nonmaterial services production were feeble. The
years after reform and opening up have seen rapid development of non-state-owned
businesses in wholesale and retail trade, catering, and transport, yet official statistics do
not cover all these industries. To solve the problem, China conducted the first
nationwide census of tertiary industries between 1993 and 1995 (years 1991 and 1992
as census years). The census results were used to adjust the historical GDP data from
both the production and expenditure sides for the years from 1978 to 1993. Production-
X. Xu / China Economic Review 15 (2004) 302–322
307
Table 1
Retrospective upward percentage adjustment to GDP from production side
Year
GDP
Tertiary
industry
Transport,
post and
telecommunication
Wholesale
and retail trade
and catering
Nonmaterial
services
1978
1980
1985
1986
1987
1988
1989
1990
1991
1992
1993
1.0
1.1
5.1
5.3
5.8
6.1
5.7
4.8
7.1
9.3
10.0
4.4
5.2
20.6
21.2
23.0
23.4
20.3
17.2
24.7
33.1
32.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
2.7
10.4
9.5
11.7
0.0
0.0
52.2
58.1
62.3
65.1
66.7
67.6
67.6
88.7
73.4
9.3
9.6
11.9
12.4
13.2
10.7
8.8
8.5
13.9
21.7
24.8
side adjustments entailed a revision of value added in tertiary industries. The expenditure
side saw revisions in final consumption expenditures. Table 1 provides the adjustment
ratio for transport, post and telecommunications, wholesale and retail trade and catering,
and nonmaterial services.
A summary of the adjusted GDP data from both the production and expenditure sides
was first published in the China Statistical Yearbook 1995. The detailed results are found
in The Gross Domestic Product of China, 1952– 1995.
The two supplementations to the historical GDP data met the need for macroeconomic
analysis and management. One important adjustment enabled the historical GDP data to
reflect the development of tertiary industries more accurately and provided a superior basis
for the country to formulate rational industrial policies (Xu, 1998b).
4. Main data sources
There are three main data sources used in Chinese GDP estimation. Statistics collected
by the NBS on agriculture, forestry, husbandry and fishing; mining and quarrying,
manufacturing, electricity, gas, and water; construction; wholesale, retail trade, and
catering, fixed-asset investment; labor; wages; price; household income and expenditure,
etc. and by other government agencies under State Council on transport, customs, balance
of payments, etc. constitute the first source. Data also come from administrative records on
final fiscal accounting of revenues and expenditures. The rest of the data come from final
financial accounts by institutions such as banks, insurance companies, the Administration
of Civil Aviation, the Ministry of Railway, the National Bureau of Post and Telecommunications, etc.
The NBS data account for the highest proportion of all data. These data are
collected and aggregated in various ways. For example, the data sources needed for
calculating the value added of mining and quarrying, manufacturing, electricity, gas,
and water are collected and aggregated in two totally different ways. For the state-
308
X. Xu / China Economic Review 15 (2004) 302–322
owned enterprises and the non-state-owned enterprises with 5 million yuan of annual
sales or more, the report forms on production, sales, and financial status formulated by
the NBS are filled in by these enterprises, then the completed forms are returned to
local statistical agencies. While the agencies aggregate the forms, the NBS aggregates
these forms by itself and derives the enterprises’ value added based on these forms.
For non-state-owned enterprises with annual sales of less than 5 million yuan and
individual businesses, the data for value added calculation are collected in a sample
survey by the Enterprise Investigation Organization of the NBS. The process is
organized under the direct jurisdiction of the NBS. The sample survey is conducted by
separating non-state-owned enterprises with annual sales of less than 5 million yuan
from individual businesses. Then the provincial branches of the Enterprise Investigation Organization survey the first subpopulation according to a list of these enterprises
if they exist in the province or in clusters if the name list does not exist, and
investigate the second subpopulation in a clustering survey. Either the production units
sampled or special investigators fill in the questionnaires. The Enterprise Investigation
Organization derives the national total of the population based on samples. Another
example is fixed-asset investment statistics. Either local statistical agencies or the units
surveyed fill in report forms formulated by the NBS for fixed-asset investment data.
Then the completed forms are aggregated level by level up to the NBS (Xu, 1999b).
5. Basic calculation methods
GDP is estimated using both current and constant prices. Measured from the
production side, GDP is the sum of value added by industry; from the expenditure
side, it is the sum of final expenditures. Estimation at current prices uses prices in the
current accounting period. Estimation at constant prices adopts base year prices. At
present, 2000 is the base year for China. The basic methods used to calculate Chinese
GDP are briefly explained below (Xu, 2000b; National Bureau of Statistics, 2001).
5.1. Production-side estimation
5.1.1. Estimation at current prices
Production-side GDP at current prices can be calculated using either production or
income approaches.
The formula for using the production approach is as follows:
GDP using the production approach ¼
¼
X
X
Value added in the production approach
ðTotal output intermediate inputÞ
In this formula, S denotes aggregation across all industries. Total output refers to the
value of all goods and services produced in a given period by an industry. Intermediate
input refers to the value of all non-fixed-asset goods and services consumed and used in
the same period by an industry.
Table 2
Main data sources and basic calculation methods of value added at current prices by industries
Direct calculation part
Indirect
calculation part
Coverage
Basic
Main data sources
calculation
methods
Production NBS agricultural
approach production statistics
Coverage
Basic calculation
methods
Main data sources
Production Mining and quarrying,
approach manufacturing,
electricity, gas,
and water statistics
of NBS
Income
NBS construction
approach statistics
Enterprise
below cutoff
level and
individual
businesses
Enterprises
rated at
lower than
qualification
four
The rest
Value-added
ratio approach
Sampling survey on enterprises below
cutoff level and individual businesses
of mining and quarrying, manufacturing,
electricity, gas, and water issued by
the NBS
NBS fixed-asset investment statistics
Agriculture, forestry,
animal husbandry
and fishing
Mining and quarrying,
manufacturing,
electricity,
gas, and water
Whole
Construction
Enterprises rated
at four or higher
qualification
Enterprises
above cutoff
level
Transport, post and
Enterprises under Income
telecommunications the ministries
approach
Volume
extrapolation
method
Transportation and communication
statistics from the Ministry of Transport,
the Ministry of Railway, the National
Administration for Civil Aviation,
China Oil and Natural Gas Corporation,
the National Bureau of Post and
Telecommunications
309
Final accounts provided
by the Ministry of
Transport, the Ministry
of Railway, the National
Administration for Civil
Aviation, China Oil and
Natural Gas Corporation,
the National Bureau of Post
and Telecommunications
Value-added
ratio approach
X. Xu / China Economic Review 15 (2004) 302–322
Industries
Income
approach
Banking and
insurance
Main banks
and insurance
companies
Income
approach
Real estate
Income
Real estate
approach
development
and management
Final accounts provided
by the Ministry of Finance
and the Ministry of Foreign
Economy and Trade.
Financial statements of
the enterprises above the
cutoff level in wholesale
and retail trade and catering
provided by the NBS
Final accounts of main
banks and insurance
companies
NBS real estate
development statistics,
final accounts of real
estate management
provided by the Ministry
of Construction
Others
Calculation on
total retail sale of
consumer goods
Total retail sale of consumer goods
statistics provided by the NBS
Financial statistics provided by the
Estimation on
balance of deposits People’s Bank of China. Insurance
and loans, premium statistics provided by China Insurance
Regulatory Commission
Owner-occupied Imputed
Rural and urban household survey
houses services depreciation of
provided by the NBS, the data of
of urban and
fixed assets of
construction cost of buildings provided
rural households owner-occupied
by the Ministry of Construction
houses of
households
Other
financial
institutions
X. Xu / China Economic Review 15 (2004) 302–322
State-owned
enterprises
and non-stateowned large
or medium-sized
enterprises
310
Wholesale and
retail trade
and catering
X. Xu / China Economic Review 15 (2004) 302–322
311
The formula for GDP using the income approach is as follows:
X
GDP using the income approach ¼
Value added using the income approach
X
¼
ðCompensation of employees
þ Taxes on production; net
þ Depreciation of fixed assets
þ Operating surplusÞ
The compensation of employees includes all remunerations acquired by the employees
for their participation in production activities. It is composed of three parts: (1) wages and
salaries in cash, (2) wages and salaries in kind, and (3) employers’ social contributions.
Net taxes on production stands for the difference between taxes and subsidies on
production. Taxes on production are the different kinds of taxes, fees paid by producers
to the government for their producing, delivering, sales and transferring goods and
services, and for the use of production factors, for example, fixed assets, land, and labor
for production purposes. They are composed of sales taxes, value-added taxes, operating
taxes, stamp taxes, taxes on real estate, taxes on vehicles, etc. Subsidies on production
comprise government subsidies on prices and operating losses to producers for their
production activities. Depreciation of fixed assets refers to the drawn or imputed value
used to compensate for the loss of fixed asset value during the production process.
Operating surplus is the balance between value added and compensation of employees, net
taxes on production, and depreciation of fixed assets. It mainly refers to business profits.
As data sources vary from industry to industry, the production-side estimation of GDP
at current prices in China adopts both the production and income approaches; that is,
estimation of value added uses the income approach for some industries and the
production approach for others. Table 2 shows the main data sources and basic calculation
methods for the estimation of value added at current price for major industries.
In Table 2, the ‘‘enterprises above cutoff level’’ refers to state-owned enterprises and
non-state-owned enterprises with annual sales of 5 million yuan or more in mining and
quarrying, manufacturing, electricity, gas, and water. The ‘‘enterprises below the cutoff
level’’ are non-state-owned enterprises with annual sales below 5 million yuan of mining
and quarrying, manufacturing, electricity, gas, and water. A comprehensive assessment of
the quality of staff, management level, the amount of capital, and its past record
determines the grading system for construction enterprises.2 Enterprises assigned a rating
of four or higher are qualified to undertake the more complex and expensive engineering
and construction work. Enterprises rated at lower qualifications are all construction
enterprises and units except enterprises assigned a rating of four or higher. The ‘‘enterprises under the ministries’’ in transport, post and telecommunications include national
railway transport, highway transport, and waterway transport carried out by enterprises
under the Ministry of Transport; highway transport and waterway transport carried out by
enterprises under the guidance of local authorities in the system of the Ministry of
Transport; national and local civil aviation transport; national pipeline transport, national
312
X. Xu / China Economic Review 15 (2004) 302–322
post and telecommunications. ‘‘The rest ’’ in transport, post and telecommunications
include enterprises aside from those mentioned earlier. The ‘‘main banks’’ in banking and
insurance industry include the central bank (People’s Bank of China), policy banks, stateowned commercial banks, and other commercial banks; the ‘‘main insurance companies’’
include the People’s Insurance Company of China, China Pacific Insurance Company,
China Ping An Insurance Company, and other insurance companies dealing with life
insurance and nonlife insurance. Other financial institutions include all financial institutions not included in either of these categories. The value of owner-occupied dwelling
services of urban and rural households is equal to the imputed depreciation of owneroccupied dwellings. The imputed depreciation is calculated by the following formula:
The imputed depreciation of rural owner-occupied dwellings
equals the value per single room at the end of the accounting year
times the average number of the rooms owned by the rural households sampled
times the number of households in the rural area
times 2% (depreciation rate)
The imputed depreciation of urban owner-occupied dwellings
equals the construction cost of a square meter of floor space of the dwellings owned by
urban households
times the total floor area of the owner-occupied dwellings of urban households
times 4% (depreciation rate)
5.1.2. Estimation at constant prices
Production-side GDP at constant prices can be calculated in two ways: deflation and
extrapolation. Deflation methods themselves are further divided into double deflation and
single deflation. Double deflation means that both output and intermediate inputs at
current prices are deflated by an output price index and an intermediate input price index,
respectively, to calculate output and intermediate input values at constant prices. The
difference between the output and input values is defined as value added at constant prices.
Single deflation can be dealt with in two ways, one is to deflate value added at current
prices by an output price index; the other deflates value added at current prices by an
intermediate input price index. Extrapolation methods are also divided into double
extrapolation and single extrapolation similar to the case of deflation. Double extrapolation indicates that output and intermediate inputs in the base period are extrapolated by
using an output volume index and an intermediate input volume index, respectively, to get
output and intermediate inputs at constant prices in the current accounting period. The
difference between the former and the latter is defined as value added at constant prices.
Single extrapolation can also be done in two ways, by extrapolating value added in the
base period by an output volume index or by an intermediate input volume index.
In practice, deflation and extrapolation methods are sometimes combined depending on
the data sources. This mixed method can be applied into two ways, either using the
deflation method for output and the extrapolation method for intermediate inputs, or using
the extrapolation method for output and the deflation method for intermediate inputs.
X. Xu / China Economic Review 15 (2004) 302–322
313
Table 3 summarizes calculation methods for value added at constant prices for main
industries in China.
In Table 3, the output volume index of agriculture, forestry, animal husbandry, and fishing
at constant prices equals output at constant prices in agriculture, forestry, animal husbandry,
and fishing in the current year divided by output at constant prices in agriculture, forestry,
animal husbandry, and fishing in the base year. The detailed calculation method of output at
constant prices of agriculture, forestry, animal husbandry and fishing is as follows: output at
constant prices in the base year equals output at current prices in the base year; output at
constant prices in nonbase years (those have same base year) is derived year after year
according to an extrapolation method by using the growth rate of output value at constant
Table 3
Basic calculation methods of value added at constant prices in main industries
Industries
Basic calculation methods
Basic indices
Agriculture, forestry,
animal husbandry,
and fishing
Mixed method: extrapolation
for output, deflation for
intermediate input
Mining and quarrying,
manufacturing, electricity,
gas, and water
Construction
Single deflation
Output: output volume index of
agriculture, forestry, animal husbandry,
and fishing at constant prices
Intermediate input: price indices of
production material and others
corresponding to 13 kinds of
intermediate inputs
Output price indices, adjusted according
to changes in raw material prices
Single deflation
Transport, post and
telecommunications
Single extrapolation
Wholesale and retail
trade and catering
Banking and insurance
Single deflation
Single deflation
Real estate
Single deflation
Construction and installation price index
included in price index of fixed-asset
investment, adjusted according to the
difference between labor cost price
index and raw material cost price index
Transport: conversion turnover index
from passenger and freight transport
Post and telecommunications: index
of business volume of post and
telecommunications
Retail price index
Weighted average price index of
consumer price index and price index
of fixed-asset investment
Value added of real estate development:
price index of selling of houses
Value added of real estate management:
consumer price indices
Depreciation of newly increased
household owner-occupied dwellings:
price index of fixed-asset investment
Depreciation of existing household
owner-occupied dwellings: adjusted
depreciation of owner-occupied dwellings
at constant prices in the last period
314
X. Xu / China Economic Review 15 (2004) 302–322
prices3 in agriculture, forestry, animal husbandry, and fishing specific statistics. The 13 types
of intermediate inputs of agriculture, forestry, animal husbandry, and fishing in Table 3 are
seed, forage, fertilizer, fuel, pesticide, agriculture plastic film, power, small farm tool
purchases, etc. The weights in banking and insurance are the proportions of household
consumption expenditure and gross fixed capital formation in the sum of these two
indicators.
5.2. Expenditure-side estimation
5.2.1. Estimation at current prices
The calculation of GDP at the current prices using the expenditure approach is as
follows:
GDP using the expenditure approach ¼ final consumption expenditure
þgross capital information þ net export
where final consumption expenditure consists of household consumption expenditure and
government consumption expenditure; gross capital formation includes gross fixed capital
formation and changes in inventories; net exports are exports of goods and services less
imports of goods and services. Household consumption expenditure is the expenditure on
goods and services for final demand by residential households. Household consumption
expenditure includes both cash and noncash (imputed consumption) purchases. There are
several types of household imputed consumption: acquired goods and services in the
forms of wages and salaries in kind and transfer in kind, self-supplied goods and services
such as self-supplied foodstuff in rural households. Purchases of dwellings and construction of owner-occupied dwellings are included in gross fixed capital formation, not in
household consumption expenditure. Household consumption expenditure is broken down
into seven components: consumption of traded consumer goods; consumption of ownproduced consumer goods; consumption counterpart of income in kind; consumption of
services other than housing; consumption of housing, water, electricity, and gas; consumption of free medical services; and consumption of collective welfare services.
Government consumption expenditure includes expenditure on public services provided
by the government and net expenditures on goods and services provided to the households
free or at reduced prices. Gross fixed capital formation is the value of fixed assets, which
are purchased, transferred in, and self-supplied by residents, deducting the value of sold
and transferred out fixed assets. The value of changes in inventories is equal to the value of
the inventories acquired less the value of the inventories disposed of during the accounting
period.
Table 4 summarizes the main data sources and basic calculation methods of GDP at
current prices using the expenditure approach.
5.2.2. Estimation at constant prices
GDP at constant prices using the expenditure approach equals the sum of its
components at constant prices, which are obtained by deflating the components of GDP
X. Xu / China Economic Review 15 (2004) 302–322
315
Table 4
Main data sources and basic calculation methods of GDP at current prices using the expenditure approach
Components
Main data sources
Household consumption expenditure
1. Consumption of
The NBS statistics on total
traded consumer
retail sales of consumer goods
goods
2. Consumption of
own-produced
consumer goods
3. Consumption
counterpart of
income in kind
The NBS rural household survey
The NBS urban household survey
4. Consumption of
services other than
housing
The NBS rural and urban
household survey
5. Consumption of
housing, water,
electricity, and gas
7. Consumption of
collective welfare
services
The rural and urban household
survey, the NBS population survey,
the statistics on the construction
cost of per square meter of floor
space from the Ministry of
Construction
The statistics on social security
and welfare from the Ministry of
Labour and Social Security
The statistics on social security and
welfare from the Ministry of Labour
and Social Security
Government
consumption
expenditure
The Ministry of Finance annual final
accounting, the NBS Census of the
tertiary industries
6. Consumption of free
medical services
Basic calculation methods
Excluding retail sales to enterprises,
government units, and public utilities,
purchases of building materials by
households for the construction and
major repair of houses
Consumption expenditures by rural
households minus consumption
expenditures in cash by rural households
Average income in kind received by
each urban household that reported
having received in kind the proportion
of urban households receiving income
in kind among the urban households
sampled the number of urban
households in the country
The sum of transport fees, postal and
telecommunication fees, charges for
medical and health care services,
tuition, fees for adult technical training,
fees for cultural and recreational
services, fees for processing and
repairing services, and other services
expenditures
The sum of rentals for dwellings,
expenditures on water, electricity,
and fuel, imputed expenditures on
owner-occupied dwellings services
by rural and urban household
The sum of medical payments by
state-owned units and medical
payments by collective units
The sum of payments by state-owned
units and collective units to provide
recreational and cultural facilities for
their employees
Current expenditures in operating
expenses within the scope of budget,
current expenditures in operating
expenses within the scope of extra-budget,
depreciation of fixed assets of government
administrative units and nonprofit public
utility units, the difference between the
gross output and operating revenue of
urban household commissions and rural
villagers commissions
(continued on next page)
316
X. Xu / China Economic Review 15 (2004) 302–322
Table 4 (continued)
Components
Main data sources
Basic calculation methods
Gross fixed capital
formation
The NBS fixed-asset investment
statistics, the Ministry of Finance
annual final accounting, the Ministry
of Construction real estate final
accounts and statistics, the Ministry
of Land and Resources data on the
area of newly reclaimed land and
cost of reclamation per unit of area,
and the data on the expenses of
geological prospecting
The specific statistics from the NBS,
the consolidated final accounts of
state-owned enterprises from the
Ministry of Finance, the final
accounts from the Ministry of
Transport, the Ministry of Railways,
the National Bureau of Post and
Telecommunications, the Ministry
of Foreign Trade and Economy
The balance of payments from the State
Foreign Currency Administration
The balance of payments from the
State Foreign Currency Administration
Fixed-asset investment with a value of
500,000 yuan or more—expenses for
the purchase of land, existing buildings
and structures, and existing equipment +
fixed-asset investment below 500,000
yuan + the margins charged on new
dwellings sold + costs of the transfer
of ownership of dwellings + increase of
produced intangible assets + costs of land
improvement
Calculated by ownership, the value of
inventories at the end of the accounting
period, minus the value of inventories at
the beginning of the accounting period,
then adding them up
Changes in
inventories
Export
Import
Export of goods and services
Import of goods and services
at current prices using price indices. Table 5 summarizes the basic methods and main price
indices used for calculating GDP at constant prices under the expenditure approach.
6. Some measures taken by the NBS to improve GDP estimation
In recent years, the NBS has taken a series of measures to improve GDP estimation and
the accuracy of GDP data (Xu, 2000a).
6.1. The reform of statistical survey methods
The traditional statistical surveys, which relied on overall reporting, worked relatively
well in the central planned economy, but showed significant shortcomings when exposed
to the socialist market economy. In the socialist market economic system, private firms,
shareholding enterprises, foreign-invested enterprises, and those invested from Hong
Kong, Macao, and Taiwan account for a certain proportion in the economy. The old
statistical surveys relying on overall reporting ask them to report to the government
statistical agencies frequently, which creates a significant burden, making the enterprises
reluctant to respond to those surveys positively. Hence, these kinds of surveys do not aid
in the smooth implementation of statistical surveys and improvement in the quality of
statistical data. In the socialist market economic system, numerous small-sized enter-
X. Xu / China Economic Review 15 (2004) 302–322
317
Table 5
The basic methods and main price indices used for calculating GDP at constant prices by expenditure approach
Components
Household consumption
expenditure
Consumption of traded
consumer goods
Consumption of own-produced
consumer goods
Consumption counterpart of
income in kind
Consumption of services
other than housing
Consumption of housing,
water, electricity, and gas
Basic methods
Main price indices
Deflation
Retail price index
Deflation
Deflation
The price index of the gross output of agriculture,
forestry, animal husbandry, and fishing
The goods subindex of the urban consumer price index
Deflation
The services subindex of the consumer price index
Deflation
Consumption of housing renting services, water,
electricity, and gas: the rental subindex and the
subindex of water, electricity, and gas in the
consumer price index; the services provided by
new dwellings in owner-occupied dwelling services:
the price index of fixed-asset investment; the
services provided by existing dwellings in owner-occupied
dwelling services: adjusted the owner-occupied dwelling
services at constant prices in the previous year
The simple average of the subindex of articles for
medicine and health care and the subindex for medical
and healthcare services in the consumer price index
The service subindex of the consumer price index
Consumption of free
medical services
Deflation
Consumption of collective
welfare services
Government consumption
expenditure
Deflation
Gross fixed capital formation
Changes in inventories
Deflation
Deflation
Export of goods and services
Deflation
Import of goods and services
Deflation
Deflation
Expenditures on goods: retail price index; expenditures
on services: the services subindex of the consumer price
index; expenditures on wages and salaries: the consumer
price index; depreciation of new fixed assets: the price
index of fixed-asset investment; depreciation of existing
fixed assets: adjusted depreciation of fixed assets at
constant prices in the previous year
The price index of fixed-asset investment
Production material: ex-factory price index of
production material
Living material: ex-factory price index of
living material
Agricultural products: producer price index of
agricultural products
The goods export price index, adjusted according to
relevant service price indices in home
The goods import price index, adjusted according to
relevant service price indices of some countries
which are main exporters of services to China
prises and individual businesses appear, but they often have no sound records of
production, income and expenditure, balance sheets, and so on, so it is impossible to ask
all of them to fill in the statistical questionnaires and to guarantee the reliability of their
reports. Over the past years, the NBS has spent much effort implementing the sampling
318
X. Xu / China Economic Review 15 (2004) 302–322
surveys in an attempt to relieve the burden for grassroot units and improve the quality of
statistical data. For example, sampling surveys have been conducted continuously for
output of grain and cotton and for output, taxes, wages, and profits of non-state-owned
enterprises with annual sales of less than 5 million yuan in mining and quarrying,
manufacturing, electricity, gas, and water. These data played a role in improving the
quality of GDP estimation.
6.2. The reform of statistical aggregating methods
The traditional statistical survey data were aggregated from the grassroots level to the
top. Although this method is advantageous for meeting the needs of economic management agencies at all levels, it cannot eliminate certain drawbacks. Typically, the statistical
data were prone to interference from human factors, and the aggregation process tends to
undermine quality. In recent years, the NBS has introduced some effective aggregating
methods that ensure the reliability and objectivity of aggregated data, especially the
superaggregating method. Since 1999, instead of going through regional statistical offices,
the NBS directly aggregates ‘‘the production and sales value report of mining and
quarrying, manufacturing, electricity, gas, and water enterprises’’ and ‘‘the financial
statements of mining and quarrying, manufacturing, electricity, gas, and water enterprises’’
in the Annual Statistical Report, and directly aggregates the sampling-surveyed data of
grain output and farmers’ income, etc.
6.3. The increased scope of statistical surveys
In recent years, the NBS had continued to enlarge the scope of statistical surveys and
fill in the gaps. For example, in the past rural production in mining and quarrying,
manufacturing, electricity, gas, and water were beyond the statistical surveys of the NBS
system, whose data were collected by a related management agency. The third census of
the industry in 1995 discovered obvious overestimation in the data of the rural part of
the industry. To improve its data quality, the NBS began applying the sampling surveys
to it beginning in 1999. There are some gaps in the coverage of China’s traditional
services statistical surveys. To solve the problem and to find benchmarks for the services
statistical data, the NBS established a regular census system of tertiary industries in
1993. The coverage of total fixed-asset investment statistics was also incomplete,
excluding fixed-asset investment from urban private firms and individual businesses
(but including private investment on building dwellings in the plant and mine district in
urban areas). After having constructed the pilot statistics with this part of fixed-asset
investment, the NBS has incorporated them into the coverage of total fixed-asset
investment statistics.
6.4. The reform of calculation methods for value added at constant prices of mining and
quarrying, manufacturing, electricity, gas, and water
Value added at constant prices in mining and quarrying, manufacturing, electricity,
gas, and water has been derived on the basis of the traditional gross output value at
X. Xu / China Economic Review 15 (2004) 302–322
319
constant prices4 of the industry since 1985. What follows is the main idea of
calculating traditional gross output value at constant prices of the industry. First, the
NBS, together with some related ministries of the State Council, worked out a
catalogue of constant prices for various kinds of products in the industry, using base
year prices, which were supplemented by provincial statistical bureaus, together with
related departments at the same level. Then all enterprises in the industry were required
to use this catalogue to calculate their gross output value at constant prices. Pooling the
gross output value at constant prices over enterprises from bottom to top would result
in regional or national gross output value at constant prices of the industry. This
method has the advantage of meeting the needs of economic management agencies at
different levels simultaneously, but also possesses disadvantages. First, many new
products that were introduced after the base year have no constant prices, making it
difficult to avoid using current prices as a replacement for constant prices without fully
removing the influence of price changes. Current prices are commonly used as constant
prices in measuring gross output value at constant prices of the industry for the rapidly
developing town and township-run enterprises due to the restriction of data sources and
lower qualification of accountants. Second, the level-by-level aggregation of gross
output value at constant prices increases the odds of interference. Some leaders of stateowned enterprises and local governments, motivated by personal political interests,
intervene in the statistical data directly or indirectly. Third, economic theory and
practice show that product shares will shift in favor of products with declining relative
prices. Thus, the growth rate of gross output value derived from fixed price weights is
usually overestimated and this overestimation increases each year. Since 1990, the gross
output value at constant prices of mining and quarrying, manufacturing, electricity, gas,
and water in China has always used 1990 as the base year. As the product structure and
prices change during this period, the growth rates of gross output value at constant
prices of the industry derived from fixed price weights will inevitably affect the
accuracy of growth rates of value added at constant prices of the industry and GDP at
constant prices. Taking this into account, NBS has been experimenting with reforming
the calculation method of the value added at constant prices of the industry in recent
years. Now, the NBS tries to adopt the price index deflation method to calculate value
added at constant prices for mining and quarrying, manufacturing, electricity, gas, and
water.
7. Some problems with China’s GDP estimation and proposals and ideas of its reform
Some problem in China’s GDP estimation should receive attention (Xu, 2002).
First, the industry classification and expenditure categories are too aggregated to meet
the needs of analyzing the production and demand structure or to deliver necessary data to
international organizations. The industry of mining and quarrying, manufacturing,
electricity, gas, and water is an excellent example. So far, although the value added of
the industry accounts for about 44% of GDP estimates, it is only shown as one industry.
This system cannot meet the needs of analyzing the development trends of its various
subindustries and its structural change, primarily because the sample survey of non-state-
320
X. Xu / China Economic Review 15 (2004) 302–322
owned enterprises with annual sales of less than 5 million yuan and individual business
does not provide detailed subindustry information.
Second, a regular (annual) survey system for nonmaterial services has not been
established since the first census of tertiary industries. The value added of many rapidly
developing new services such as accounting, legal services, information consultancy,
non-public-run education, and so on are estimated mainly based on the wages and taxes
of these industries. Therefore, GDP estimation does not properly incorporate these
activities.
Third, quarterly GDP estimation has suffered from data limitations, allowing for neither
any production measurement on quarterly basis, nor quarterly expenditure measurement.
Quarterly GDP measurement, whether production or expenditure based, will better reflect
the most recent national economic developments and provide a better foundation for shortterm macroeconomic analysis and decision making, compared with cumulative quarterly
production-side GDP measurement implemented in China since 1992. These short-term
data are crucial as the State implements an active fiscal policy and robust monetary policy,
increases infrastructure investment, and expands exports.
Fourth, there are some weak points in GDP estimation at constant prices due to the lack
of price indices. So far, China has not compiled a producer price index for services; most
of the value added at constant prices of service industries is derived from the
corresponding service subindices of the consumer price index (CPI). However, for some
services such as advertising, whose consumers are not households, there is no
corresponding CPI item. In this case, the possible alternative is to use the related price
indices, which affects the accuracy of their value added at constant prices. Statistics on the
import and export of services at constant prices will remain unsatisfactory until price
indices for the import and export of services are constructed.
Fifth, the nonobserved economy (NOE) is not shown fully in GDP estimation.
According to the OECD definition, the NOE includes illegal production, underground
production, production activities of the informal sector, and so on, which are easily
missed by the basic data collection system. In line with the information provided by the
OECD, the percentage of value added created by those production activities in GDP
estimates is up to 3% in Australia, 15% in Italy, and 25% in Russia. Although China’s
GDP estimation covers part of the NOE (e.g., the production activities of own-produced
agricultural products by farmers), a comprehensive, deep-going and systematic study on
it has not been conducted. Besides, some necessary effective statistical means and
measures are still unavailable; it is unavoidable to miss some production activities in
GDP estimation.
Finally, local leaders tend to interfere with some of the statistical data due to a lack of
independence for the statistical system. Although the NBS has taken many measures to
evaluate and adjust the statistical data, it is very difficult to eliminate malfeasance and
remove all unnecessary involvement.
Taking these aforementioned problems with China’s GDP estimation into consideration, this paper proposes the following changes. First, the data should reflect the
internationally accepted classification standard. In particular, the classification of mining
and quarrying, manufacturing, electricity, gas, and water should be further disaggregated.
Second, the authorities should establish a regular (annual) survey system for the services
X. Xu / China Economic Review 15 (2004) 302–322
321
industry as soon as possible to improve these estimates. This is most important for those
newly created services. Third, creating and perfecting the quarterly survey system for each
specific statistics will lay a solid foundation for both quarterly production and expenditureside GDP estimation. Fourth, accurate accounts require complete and accurate price
indices, for example, the producer price index of services and the price index of traded
services, in order to perfect GDP at constant prices. Fifth, a full account of economic
activity requires estimates on the NOE. Finally and most importantly, believable statistics
require a relatively independent statistical management system to be free from interference
by local leaders.
References
Administration Office of State Council (1988, July). Transmission notice of Administration Office of State
Council on report of National Bureau of Statistics on establishing tertiary industry statistics (Guowuyuan bangongting zhuanfa guojia tongjiju guanyu jianli disan chanye tongji de baogao de tongzhi,
). A collection of selected document on statistical system
and methods (1950 – 1987) ( pp. 181 – 185). Beijing, China: China Plan Press.
National Bureau of Statistics (1997a, May). The estimation methods of China’s annual gross domestic product
(Zhongguo niandu guonei shengchan zhongzhi jisuan fangfa,
). Beijing, China:
China Statistical Press.
National Bureau of Statistics (1997b, May). The estimation methods of China’s quarterly gross domestic product
(Zhongguo jidu guonei shengchan zongzhi jisuan fangfa,
). Beijing, China: China
Statistical Press.
National Bureau of Statistics (2001, May). Manual of China’s gross domestic product estimation (Zhongguo
guonei shengchan zongzhi hesuan shouce,
). Beijing, China: National Bureau of
Statistics.
National Bureau of Statistics (1990, January). Introduction to the main indicators of national income and gross
national product estimation (Guomin shouru, guomin shengchan zongzhi tongji zhuyao zhibiao jieshi,
). Beijing, China: National Bureau of Statistics.
National Bureau of Statistics (1992, December). Introduction to the indicators and measure scheme of gross
domestic product and national income (Guonei shengchan zongzhi, guomin shouru zhibiao jieshi ji cesuan
fangan,
). Beijing, China: National Bureau of Statistics.
National Bureau of Statistics (1993, October). Introduction to the indicators and measure scheme of gross
domestic product (Guonei shengchan zongzhi zhibiao jieshi ji cesuan fangan,
).
Beijing, China: National Bureau of Statistics.
National Bureau of Statistics (1985). Gross national product measure scheme (Guomin shengchan zongzhi jisuan
fangan,
). Beijing, China: National Bureau of Statistics.
Xu, X. (1998a, July 10). Why does gross domestic product data change regularly (Guonei Shengchan zongzhi
shuju weishenme hui dingqi fasheng biandong,
). Economics News, 1.
Xu, X. (1998b, September 25). Important supplements and adjustment of historical data of China’s gross
domestic product (Wongguo guonei shengchan zongzhi lishi shuju de jici zhongda buchong he tiaozheng,
). Economics News, 1.
Xu, X. (1999a, July 2). Establishment and development of China’s gross domestic product estimation (Zhongguo
guonei shengchan zongzhi hesuan de jianli he fazhan, ). Economics News, 3.
Xu, X. (1999b, October 22). Main data sources and calculation methods of China’s gross domestic
product (Zhongguo guonei shengchan zongzhi hesuan de zhuyao ziliao laiyuan he jisuan fangfa,
). Economics News, 3.
Xu, X. (2000a, January 14). How does National Bureau of Statistics improve gross domestic product
estimation (Guonei shengchan zongzhi hesuan, guojia tongjiju zhengzhi jinxing naxie gaijin,
). Economics News, 3.
322
X. Xu / China Economic Review 15 (2004) 302–322
Xu, X. (2000b, September). China’s gross domestic product estimation (Zhongguo guonei shengchan zongzhi
hesuan,
). Beijing, China: Peking University Press.
Xu, X. (2001a, February). The adjustments of industry classification in China’s gross domestic product estimation
(Woguo guonei shengchan zongzhi hesuan hangye fenlei de tiaozheng,
).
Statistics and Information Forum, 19 – 20.
Xu, X. (2002, April). The problems and reform of China national economic accounting (Wogguo guomin jingji
hesuan mianli de wenti ji gaige fangxiang,
). Statistical Research, 8 – 13.