OECD-ESCAP WORKSHOP ON ECONOMIC
MAY 22-25, 2000
Consumer Price Index (Macao, China)
Other than the Composite CPI which refers to the whole population, two index
series are compiled to show the price change for the higher and lower expenditure
groups which have different consumption patterns, i.e., the CPI (A) that relates to 44%
of households which have an average monthly expenditure of MOP 2,400 to MOP
7,399, and the CPI (B) that accounts for 32% of households which have an average
monthly expenditure of MOP 7,400 and MOP 14,999.
Selection of Basket
As it is impossible to evaluate the price change of all products consumed by
households, it is necessary to select a representative basket of goods and services for
the reference population, based on the consumption pattern obtained from the
Household Expenditure Survey 93/94. Generally speaking, the CPI basket includes
self-production and gifts that are valued at market prices; while the CPI basket is final
consumption goods of families. It excludes second-hand items, goods acquired for
profit making, investment or saving purposes. Therefore, expenses that do not belong
to consumption expenditure are excluded, such as income and profit tax, donations,
acquisition of valuable assets, cash remittance overseas, etc.. After these exclusions,
the basket of goods and services is selected according to their percentage shares in the
overall expenditure. Based on this assumption, a minimum percentage threshold is
established and under which the products will not be selected. However, for those
products that weighted below the minimum percentage threshold but are widely
recognized and largely consumed by public, they will be included in the observation
panel according to their popularity.
The next phase consists of the characterization/detail specification of the quality
and other physical characteristics of each product to guarantee the information
collected throughout the period simply refers to the same product. Every product has
a variety of brands in the market, thus makes it impossible to collect prices of a
particular brand of goods from all outlets. To overcome this difficulty, only the most
popular brands in each outlet are selected such that observation of price change is
referred to the same brands in the same outlet.
Revision of CPI Basket
Using the consumer price index as an indicator of cost of living has certain
limitations due to the selection of only some of goods and services. Once the panel of
products is defined, it remains constant for a certain period of time; any modification
in the composition of this basket resulting from changes in consumption pattern will
be ignored. Hence, in view of this, our Services conducts a Household Expenditure
Survey (now called Household Budget Survey) every 5 years in order to:
1. Understand the changes in the social-economic composition of the population;
2. Consider the changes in consumer preferences, in terms of the relative share of each
group of goods and services to the overall expenditure of families;
3. Accompany with the emergence of new products and exclusion of others which
have lost importance.
With the information collected from the Household Expenditure Survey, the
consumption pattern of the CPI basket will then be revised accordingly.
Selection of Outlets for Goods and Services
Determination of the number of outlets is based on the weight of each product to
the total consumption expenditure, other factors are considered as well, namely, price
fluctuations and availability of product in the market.
Based on these criteria, every product has a minimum number of 3 outlets for
price collection, foodstuff has a maximum of 9 outlets because of its significant
weight (over 30% in the 3 index series of Composite CPI, CPI (A) and CPI (B)) and
frequent price fluctuations. Attempts have been made to attain a geographic
dispersion of outlets that sell similar products, in such a way as to reflect the price
differences in the territory. Consequently, a sample of approximately 960 outlets is
obtained, including municipal markets, supermarkets, groceries, restaurants and
eating-places, boutiques, etc.. Thus, the outlets selected must be:
- retail outlets;
- fixed outlets;
- large and well-known outlets with significant sales volume;
- outlets located in the central and important outlets of each area.
In addition, because of the special situation in Macao where hawkers represent an
important share of retail trade, this type of outlets is also included in the sample.
Among those who are selected, they have to possess an operation license and stay
regularly in fixed locations, particularly for street hawkers who sell fresh products
adjacent to the municipal markets.
Number of Goods Priced
The CPI is based on a basket comprised of 712 products (goods and services)
which are aggregated into 167 subgroups, then into 72 groups, 23 classes and finally
into 9 sections (Section 1 - Foodstuff & Beverage; Section 2 - Clothing & Footwear;
Section 3 - Rent, Maintenance & Housing Expenses; Section 4 - Alcoholic Drinks &
Tobacco; Section 5 - Household Goods; Section 6 - Health; Section 7 - Transport &
Communications; Section 8 - Education, Culture & Leisure; Section 9 - Other Goods
Product allocation and its grouping of CPI are based on the Classification of
Consumption by Purpose of the SNA 1993, with reference to the consumption habits
of Macao residents.
On a monthly basis, a small brochure is available with information of the global
index and section index of current month, a retrospective series for the past 12 months
and a brief analysis on the performance of the index series. It is released generally in
the month following the reference month. At the same time, detailed quarterly and
annual reports are also available. The former is available in the month following the
reference quarter and the latter is available 3 months after the reference year.
Some of the goods in CPI basket are considered as seasonal products because of
their specific seasonal characteristics at the production level, sales and price
fluctuations. In Macao, vegetable and fruit are defined as seasonal products so that
the calculation method of their index is different from other products.
The method of calculating the index of seasonal products is as follows :
1. The simple index is calculated with the average prices of the current month and the
same month in the base period;
2. The “seasonal index” is determined by the weighed average of the simple indexes
of the subgroup while the weights are referred to the expenditure structure of the
corresponding month in the base period;
3. The monthly index of the subgroup is obtained from simple arithmetic average of
the “seasonal index” of the last 12 months (method of moving average).
Two types of index series are computed, i.e., one including rents, and the other
excluding rents. For CPI including rents, a global rent index is calculated from a
fixed panel of about 350 private housing units and integrated into CPI, using the
weight (the share of rent in the total expenditure) obtained form the Household
Expenditure Survey 93/94.
When a product lost its popularity and disappeared from the market, or an outlet
is closed, substitution process would be taken. However, the problem we encounter is
how to eliminate or minimize the effect of quality change caused by substitution, to
the price movement of products such as clothing, lighting, furniture, electrical
appliances and automobile, etc..
The formula used for computing the various consumer price indexes is the
Laspeyres' index, calculated first for each product. The product indexes are then
aggregated to form index at the next higher level, i.e. the subgroup indexes. This
process continues until the section indexes are formed, which are then finally
aggregated to give the consumer price index. The general formula used for the
product index is:
Ij(t) = * 100
Ij(t) = index of product j in period t
pmj(t) = average price of product j in period t
pmj(t0) = average price of product j in base period t0
Base year (t0) refers to the 12 months period starting from July 1995 to June 1996.
The average price of the different products in the base year is the arithmetic average of
prices of those 12 months.
Index of subgroup:
(t0) * Ij(t)
Iu(t) = n
Iu(t) = index of subgroup u in period t
Wj(t0) = weight of each product j in the subgroup u, in base period t0
n = number of products in subgroup u
Index of section:
∑Wc (t0) * Ic(t)
Is(t) = c =1
Is(t) = index of section s in period t
Wc(t0) = weight of each class in section s, in base period t0
Ic(t) = index of class c in period t
v = number of classes in section s
Consumer price index:
∑Ws (t0) * Is(t)
IPC(t) = s =1
IPC(t) = consumer price index in period t
Ws(t0) = weight of each section s of CPI in base period t0
z = number of sections in CPI
According to the results of the 98/99 Household Budget Survey, the revision of
CPI is undertaking in our Services, with the objective to update the composition of the
CPI basket. The base year is from October 1999 to September 2000. Publication of
the new series will begin in January 2001.
The sampling frame of all retail outlets in Macau is stratified into 17 groups: 16
major economic activities (represent 86.3% of the retail sales based on the results of
the Wholesale and Retail Census 1997) and one stratum for other economic activities.
These groups are further subdivided into 3 clusters according to the number of
employees: 1 to 9 employees, 10 to 19 employees and 20 or more than 20 employees.
In the quarterly retail sales survey, a rotation sampling method is applied. Every
quarter, two third of the samples will be rotated by new outlets, and one third will be
repeated in the following quarter.
Altogether, there are 376 outlets selected for the quarterly retail sales survey.
Sample coverage of the quarterly retail sales survey includes Macao peninsula,
Taipa and Coloane islands. The sampling units are outlets or places that sell the
(2) Motorcycle and accessories;
(3) Fuels for motor vehicle;
(5) Department store;
(6) Fruit and vegetable;
(7) Meat and meat products;
(8) Dry sea products;
(10) Household textiles;
(11) Adult clothing;
(13) Electrical household appliances;
(14) Furniture and lighting;
(15) Watch and jewellery;
(16) Fuels for domestic use.
Frequency of Data Collection and Publication
Data is collected and published on a quarterly basis.
Currently, our Services does not compile any retail sales index.
(1) The respondents tend to underreport their sales because they are afraid that the
tax will be adjusted according to the data reported to our Services. In fact,
there are regulations in Macao stipulating the information collected is used
exclusively for statistical purpose.
(2) Certain information of the sample is not updated. This problem is related to
the special business environment in Macao, where most of the shops are small
businesses and family based, and the cost of opening and closing a shop is not
very high. Therefore, the shops are opened and closed so frequently that even
administrative record or field visit cannot provide an up-to-date situation of
the retail sector.
International trade statistics record all goods entering (imports) and leaving
(exports) the territory of Macao, except those that are transported through the territory
(goods in transit), exported temporarily or re-imported. In addition, the results of
foreign trade statistics do not include the following classes of goods:
- External trade operations with the value not exceeding 5,000 MOP;
- Personal articles of travellers;
- Parcels posted for personal use or consumption by non-traders;
- Gifts of a personal nature where no payment is or is to be made by the receiver
- Manufactured tobacco and alcoholic drinks for traveller’s consumption, and
subject to a quantitative limit;
- Advertising material or sample which is supplied free of charge;
- Goods temporarily exported and imported for exhibition and sports competition
- Goods temporarily exported and imported for repair and maintenance purposes;
- Goods in transit;
- Monetary gold;
- Banknote and coins in circulation, being legal tender in any country;
- Goods under operational lease.
Unrecorded Border Trade
Trade statistics record goods on the basis of crossing border under the customs
control; hence, we do not produce any statistics about unrecorded border trade.
Classification of Commodities
The Harmonized Commodity Description and Coding System (referred to as
Harmonized System in the later of this text) is used throughout the process of
compiling international trade statistics. However, for analytical purposes, correlation
tables between this harmonized system and other classifications such as Classification
of Economic Activities of Macau (CAM) - Rev. 1 (the industrial classification that is
currently used in Macau), Standard International Trade Classification (SITC) - Rev.3
and the Broad Economic Categories (BEC) are also established.
Classification of Partner Countries
Imports are recorded according to the country of origin, which refers to the
country or territory where the agricultural products were grown, the minerals were
mined or where the manufactured articles were wholly or partly manufactured. For
manufactured articles that are partly processed, the country or territory that is
responsible for the last phase of the manufacturing operation in turning the
merchandise into its final form before entering the territory is considered as the
country of origin. However, repacking, reclassifying and mixing are not considered as
a transformation operation.
Exports are recorded according to the country of destination, i.e. the final country
or territory to which the goods are delivered, whether or not the transportation has
been interrupted, as long as there is not any commercial transaction in the
intermediate country or territory.
Frequency of Publication
Our Services compiles trade statistics on a monthly basis for aggregate data and
for data by major trading partners and commodity groups.
Time Lag of Release
After the whole process of compilation (data collection, codification, validation,
analysis and preparation of publication) of trade statistics, our Services disseminate
trade statistics that become publicly available on the 25th of month N+2.
The External Trade Statistics Division of the Statistics and Census Services of
Macao is responsible for compiling trade statistics.
Institutional Arrangements for Data Collection
The international trade statistics are based on information given by the importers
and exporters in the trade licenses/declarations. Each license/declaration contains 4 to
6 copies; one of them is exclusively for statistical purpose. Upon finishing the
transaction process of imports and exports, the Marine Police and Customs will send
the statistical copy to the External Trade Statistics Division of our Services on a daily
Unit Value Indexes & Quantity indexes
The international trade indexes practiced in our Services are quarterly Laspeyres ,
Paasche and Fisher index numbers that are computed for the reference quarter based
on the average quarterly figures of the previous year. This means that a changing base
year method is adopted. Value index, unit value index and quantity index are
computed for the exports and imports of goods. Unit value index, which is the most
basic index, is first computed for each type of good that is classified according to the
Harmonized System. For the exports of goods, unit value indexes (links) are
computed for each 5-digit CAM group by using the unit value indexes of the goods
that belong to the group, and the resulting indexes (links) are then aggregated to form
indexes (links) for each 3-digit CAM group and finally into the overall export unit
value index. On the other hand, for the imports of goods, unit value indexes (links)
are first computed for each 3-digit BEC group by using the unit value indexes of the
goods belonging to that group, and the resulting indexes (links) are then aggregated to
form indexes (links) for 4 broad BEC categories which are then finally combined to
form the import unit value index. Value indexes (links) are first computed at the 5-
digit CAM group and 3-digit BEC group for the exports and imports respectively.
Quantity indexes are then obtained from the value and unit price indexes.
1. Unit Value Index (UVI) for a good classified according to the Harmonised System:
u (y, q,i)
UVI = * 100.............................................................................................(1)
U (y - 1,i)
u (y,q,i) = unit value of item i for quarter q of year y, obtained as the total value
divided by the total quantity in the quarter
U (y-1,i) = average unit value of item i for year y-1, obtained as the total value
divided by the total quantity in the year
2. For the 5-digit CAM group (for exports) and the 3-digit BEC group (for imports),
the formulae used for:
a) Laspeyres Unit Value Link (UVL)
∑ V(y - 1,i) * UVI
UVL = i =1
∑ V(y - 1,i)
V(y-1,i) = average quarterly trade value of item i for year y-1
b) Value link
∑ v (y, q,i)
= i =1
∑ V(y - 1,i)
v (y,q,i) = trade value of item i for quarter q of year y
V(y-1,i) = average quarterly trade value of item i for year y-1
c) Paasche Volume Link = Value Link/Laspeyres Unit Value Link……………..2c)
3. For the 3-digit CAM group (for exports), the 4 broad BEC categories (for imports),
as well as the overall import and export indexes, the formulae used for:
a) Laspeyres Aggregate Unit Value Link (AUVL)
∑ V * UVL *100.................................................................................3a)
V0 = quarterly trade value of the base year
b) Value Link, same as 2b).
c) Paasche Volume Link, same as 2c).
For the Paasche type Unit Value Link and hence the Laspeyres Volume Link at
different aggregate levels are computed in a similar manner, except that the weight
used for the Unit Value Link is the current quarterly trade value. The Fisher Unit
Value and Volume Links are obtained from the Laspeyres and Paasche Links.
Fisher Unit Value Link = Laspeyres Unit Vaue Link * Paasche Unit Vaue Link
Fisher Volume Link = Laspeyres Volume Link * Paasche Volume Link
In addition, for the annual
Value Link at diff. aggregate levels = Trade value for yr. y / Trade value for yr. y-1
Unit Value Link at diff. aggregate levels = Simple average of the 4 quarterly links of
Volume Link at diff. aggregate levels = Value link for yr. y / Unit value link for yr. y
Once the annual links have been found, they can be chained backwards to the
reference year of 1996 by using the following method:
Chain Index for year 1997 = Annual Link for yr. 1997
Chain Index for year 1998 = Chain Index for yr. 1997 x Annual Link for yr. 1998
Chain Index for year 1999 = Chain Index for yr. 1998 x Annual Link for yr. 1999
Chain Index for year 2000 = Chain Index for year 1999 x Annual Link for yr. 2000
Chain Index for current year = Chain Index for yr. y-1 x Annual Link for current yr. y
Chain Index for current quarter = Chain Index for yr. y-1 x Quarterly Link for current
quarter of yr. y
In computing the foreign trade indexes, the products chosen should constitute a
major proportion of the 5-digit CAM group and 3-digit BEC group for exports and
imports respectively. However, there might be cases that some important products are
missing in the base year or in the reference period and hence it makes the selection of
products difficult. Besides, like computing other indexes, quality change in the
products causes the resulting indexes to be less representative of the respective
The Employment Survey collects information on situations of the active
population, employment, unemployment, underemployment, etc.., with the purpose of
analyzing the characteristics of the active population, the employed and the
unemployed by sex, age, industries, professions, etc., and other data such as
remuneration and number of hours worked.
Administrative Sources for Unemployment Data
In Macao, the Labour and Employment Bureau provides intermediation service
between the unemployed and the employers. This means that the unemployed can
register in the Bureau for purpose of job finding. However, not all jobless people
make their registrations and the registrations may not exactly reflect the current status
of the unemployed. Hence, the number of registrations cannot be used as a source for
complete information on the number of unemployed.
Frequency of Data Collection
Because the administrative records in the Labour and Employment Bureau relate
to persons who have made their registrations, and since it is impractical to actually
count every unemployed person each month, the Statistics and Census Services
conducts a monthly sample survey called the Employment Survey to measure the
extent of unemployment and data on related areas of the territory. Data of the
previous month, together with the data of the prior 2 months, are then used to compile
the various indicators of the employment situation of Macao. In other words, data on
the latest 3-month period will finally be released.
For each month, respondents from a sample of 1,200 housing units will be
visited. Hence, the survey can be regarded as a household type. The housing units
include the general housing units, provisional housing units and housing units that are
located within non-residential buildings. The respondents are residents of Macao.
Respondents are regarded as Macao residents when one of the following 2 conditions
is satisfied: (1) had lived in Macao for most of the time during the past 3 months or
(2) will live in Macao for most of the time in the coming 3 months. The reference
period for the employment situation is the 7 day-period before the day he/she is
interviewed. However, since the interviews are the basic source of data, respondents
are never asked specifically if they are unemployed, nor are they given an opportunity
to decide their own labor force status because many of them are unsure of how our
Services defines unemployment or employment. In addition, in order to follow the
same procedures to obtain comparable results, employees of this Services are given
intensive training, including classroom lectures, discussion, practice and on-the-job
Based on responses to a series of questions on work and job search activities,
each respondent is classified as employed, unemployed or not in the active population.
The concepts adopted in this survey follow closely the recommendations of the
International Labour Organisation (ILO). The basic concepts in defining the employed
and unemployed are the following:
Refers to all individuals aged 14 or over that could participate in the production
of goods and services in the reference period, i.e., it comprises the employed
population and the unemployed population.
Refers to all individuals aged 14 or over that were at work for at least 1 hour in
the reference period for pay, profit or family income, in cash or in kind, including
those individuals who were employed and had a formal relationship with their
employers but were not at work, as well as those individuals who had a company, but
were not temporarily at work for some specific reason.
Refers to all individuals aged 14 or over that did not have a job or any relation
with the employers in the reference period, but were available for work that provided
remuneration or available for business and had sought work for the past 30 days.
Institutional Arrangement for Data Collection
The Employment Survey targets at persons aged 14 or over who is not in
institutions such as barracks, hospitals, prisons, students' dormitory, institutions for
the old, refugee camps, etc.. However, persons that reside in housing units with 5 or
more families and dormitory for employees are included.
Owing to the construction of plenty of housing units in Macao during recent
years, a higher and higher vacancy rate is the main problem encountered in this
survey. In fact, the vacancy rate jumped from 7.8% in 1992 (annual average) to
19.0% in 2000 (first quarter’s average). As a result, this will finally transform into a
negative impact on the response rate of the survey.
Vacancy rate (annual average)
Year 1992 1993 1994 1995 1996 1997 1998 1999 2000 (1st Q average)
% 7.8% 10.8% 13.4% 13.9% 15.8% 17.2% 17.3% 18.7% 19.0%