SBP-Research Bulletin

Volume 1, Number 1, 2005

Rebasing of National Income Accounts in

Pakistan
Sajjad Zaheer∗

The objective of this note is to discuss the rebasing of national income accounts in
Pakistan, mainly according to 1999-2000 prices and coverage. While using the
information provided in the Federal Bureau of Statistics publication, this note
outlines and analyzes the different techniques used in rebasing constant price
estimates, decomposition of change in constant estimates due to rebasing (using
1999-2000 prices) and new coverage, implications of rebasing on sectoral GDP
estimates, and their growth rates. The note uses the data from 1999-2000 to 2002-
03 for the analyses on both old and new base.1

National income accounts constitute a formal framework for the classification of

information about the economic performance of a country. It serves as a powerful
tool for policy formulation and economic analysis. The data provided in these
accounts shows the behavioral trends of various macroeconomic indicators.
However, to get a real picture of the economy, it is imperative that from time to
time these accounts are revised. This is usually achieved through a methodology,
broadly and simply known as “rebasing,” that takes into account those factors,
such as price increase and structural changes, which over time might under- or
over-estimate the national income accounts.

For example, growth rates of some important indicators are not reported exactly
and might cause confusion. In Pakistan, for instance, the real GDP growth rates
according to the rebased series of GDP (base 1999-2000) in the years from 2000-
01 to 2002-03 are less than that of old series (base 1980-81; Table 1 (a)). In 2000-
01, growth rate of real GDP according to old series was 2.2 percent but according
to new base and enhanced coverage it was 1.8 percent.

Moreover, some important ratios, such as budget deficit to GDP ratio and M2 to
GDP ratio, with which fiscal and monetary policies are related, are not stated

∗
The author [[email protected]] is Inspection Officer in the Banking Inspection Department
of the State Bank of Pakistan. He wishes to thank the Editor, Sahar Babar, Riaz Riazuddin, and
Moinuddin for helpful suggestions. Remaining errors are the responsibility of the author.
1
Unless mentioned otherwise, all the data is from Economic Survey of Pakistan (2003 – 2005) and
Annual Reports (2000 – 2003) of the State Bank of Pakistan.

© 2005 by the State Bank of Pakistan. All rights reserved. Reproduction is permitted with the consent of the Editor.
Opinions 60

Table 1 (a). Real GDP Growth

Target Old Series1 New Series2
2000-01 5.0 2.2 1.8
2001-02 4.0 3.4 3.1
2002-03 4.5 5.1P 4.8
1
Base 1980-81
2
Base 1999-2000 ( with enhanced coverage)
P
Provisional

accurately. Hence, the policies based on these indicators can confuse the
realization of desired targets. For example, in 2002-03 the budget target was 4.4
percent of GDP and the realization 4.6 percent (Table 1 (b)). But, according to
new coverage, the budget deficit is 3.7 percent of GDP, which is less than the
target; actually suggesting a contractionary fiscal policy stance when compared to
the target. Other macroeconomic indicators may exhibit similar dichotomy.
Hence, it is extremely important to rebase the national income accounts more
frequently, let us say, once in five years.

In particular, periodic rebasing of national income accounts is essential due to the

following major reasons: (i) over time, the structure of the economy changes in
terms of production and consumption patterns; (ii) introduction of new products
due to technological innovations and developments; (iii) alteration in the relative
prices of commodities; (iv) product wise improvements in the variety of products
and services; and (v) changes in the pattern and classification of sectors/sub-
sectors and their regrouping due to the adoption of latest System of National
Accounts (SNA) of the United Nations (UN), currently SNA-1993.2

Table 1 (b). Old and New Series (as percent of current market prices GDP)
Total Government Private
sector
Revenue Expenditure Budget Deficit Credit M2
Target B O1 N2 Target B O N Target B O N O3 N O N
1999-00 16.8 16.3 13.5 20.2 22.5 18.7 3.3 6.6 5.4 0.6 0.5 44.5 36.9
2000-01 17.3 16.2 13.3 22 21 17.2 4.6 5.2 4.3 1.4 1.2 44.6 36.7
2001-02 17.3 17.2 14.2 22.3 22.8 18.8 4.9 5.2 4.3 1.5 1.2 48.5 40
2002-03 17.7 17.6 14.9 22.1 22.2 18.6 4.4 4.6 3.7 4.2 3.5 49.3* 43.1
B
Budget
1
Old series (Actual) i.e. with GDP in denominator at 1980-81 coverage
2
New Series (Actual) i.e. with GDP in denominator at 1999-2000 coverage.
3
Values are computed by using the private credit data given in Economic Survey 2004-05,
*
July -March

2
In general, most of the Asian countries, including Bangladesh, China, India, Indonesia, Malaysia,
Hong Kong, Philippines, Sri Lanka, Thailand, and Vietnam have followed the practice of rebasing
their accounts every 10 years or so; while Korea and Singapore revise their base year after every 5
years. Developed countries undertake rebasing even more frequently. On the other hand, Bhutan and
Pakistan on average took 20 and 17 years respectively [Asian Development Bank (2002)].
61 Opinions

1. Rebasing in Pakistan

Pakistan currently follows the SNA3 as put forward by the UN in 19934 to have an
internationally comparable mechanism in place. National income accounts
maintain data at both current and constant prices. Constant price data is used to
obtain a compatible analysis in different periods, as the inflation factor is
eliminated. Data at constant prices employs fixed prices pertaining to a specific
year, called the base year, to obtain monetary values of the different sectors of the
economy, and hence of the GDP. As a result, it reflects the actual economic
performance and growth of the economy.

Pakistan, since 1949-50, has been compiling national income accounts both on
constant and current prices.5 In 1961-62 the country revised its base for national
income accounts for the first time, from the average of five years 1949-53, used as
base, to year 1959-60. The framework and methodology provided by the UN’s
SNA 1953 was followed in rebasing at that time, to make economic and statistical
information more accurate, useful, and internationally comparable. In 1987-88, the
long overdue rebasing exercise was done for the second time. The base year was
changed from 1959-60 to 1980-81 according to the UN SNA 1968 structure.
Pakistan changed its base year for the third time in 2004, declaring 1999-2000 as
the new base, by implementing the concepts and classifications of SNA 1993.

2. Comparative Analysis of New and Old Series and Impacts of

Rebasing

A comparative analysis of the GDP series, both at constant and current prices of
1980-81 and 1999-2000 base presents the difference in absolute estimates and in
their growth rates, along with the factors responsible for causing this variation.
The explicit impact of rebasing is that it has increased the absolute figures of GDP
significantly (Figure 1). Annual Average Percentage Difference (AAPD)6 between
old and new series at current price is 21.3 percent and at constant prices 440.8
percent (Table 2). Due to rebasing, along side enhanced coverage of the real GDP,

3
SNA is given by UN’s Statistics Division to keep the national accounts in a comparable and
standardized form across the world.
4
National accounts estimates were prepared for the first time in 1949; for more details, see Federal
Bureau of Statistics (2004).
5
Technical notes of revaluation of constant price data are given in Appendix 1.
6 n
AAPD = ∑ (1 / n )[(Ε NS − Ε OS ) / ΕOS ] * 100 . Where, Ε NS is the estimate of sector/sub-sector according to
i =1

new series, Ε OS is estimate of sector/sub-sector according to old series, and n is the number of
observations.
Opinions 62

Figure 1. Gross Domestic Product (GDP)

6,000
real old sereis real new series
nominal old series nominal new series
5,000

4,000
(billions of rupees)

3,000

2,000

1,000

0
1999-00 2000-01 2001-02 2002-03

Figure 2. Decomposition of Total Change in Real Estimates

600
due to new coverage/methodology
due to rebasing
500

59
57 67
400
46
(percent)

300

200 383 384 403

342

100

0
GDP Agriculture Industry Services
63 Opinions

there is a magnification of about 441 percent; out of which, 57 percentage points is

due to enhanced coverage and 383 percentage points is due to the use of 1999-
2000 prices (Figure 2).7 The core rationale for this increase is discussed in the
subsequent subsections of GDP.

2.1. Gross Value Added (GVA) by Economic Activity8

This section presents a comparison of the levels and growth rates of the sub
sectors of the GDP taking into account the data for two base years. Broadly, the
main sectors are (a) agriculture, (b) industry, and (c) services.

2.1.1. Agriculture

The AAPD between the old and the new series at current prices is 20.6 percent
(Table 3 and Figure 2). The difference at current prices can be explained as a
result of enhanced coverage in 1999-2000 in the sector and improved reworking
methodology. Some new items, absent in the old base, have been added in the
estimation of the agriculture sub-sectors’ output. These items are strawberry,
mushroom, betel leaves, tea, henna (myrtle), condiments, oilseeds, and some non-
reported crops. In year 1999-2000, tea, strawberry, henna, betel leaves, and falsa
leaves added 80.8, 21.2, 19.4, 12.3, and 10.4 million rupees respectively to the
minor crops category.9

Table 2. GDP at Factor Cost

at constant prices at current prices
Years Old Series New Series Percent Old Series New Series Percent
(billion of rupees) Difference (billion of rupees) Difference
1999-00 649.656 3529.345 443.3 2921.988 3529.345 20.8
2000-01 664.048 3594.124 441.2 3166.954 3876.025 22.4
2001-02 686.382 3705.718 439.9 3377.098 4095.212 21.3
2002-03 721.251 3884.952 438.6 3709.67 4481.412 20.8
Annual Average 440.8 21.3

7
The respective formulae of decomposition are given in Appendix 2.
8
Source of the data according to old base year (1980-81) is Economic Survey of Pakistan 2002-2003
and 2004-05. For 2003-2004 old base year data is not available, due to which the comparison
presented is up to 2002-03.
9
The values of the increased items included in national income accounts, due to enlarged coverage,
is available only for the year 1999-2000.
Opinions 64

Table 3. Agriculture Output

at current prices at constant prices
Years Old Series New Series Percent Old Series New Series Percent
(millions of rupees) Difference (millions of rupees) Difference
1999-00 779,692 923,609 18.5 168,459 923,609 448.3
2000-01 800,854 945,301 18.0 163,845 903,499 451.4
2001-02 783,723 968,291 23.6 163,731 904,433 452.4
2002-03 864,828 1,059,316 22.5 170,523 941,942 452.4
Annual Average 20.6 451.1

Moreover, the value of flowers and leaves (horticulture) have also been estimated
for the very first time and included in the crops category. Their total value was Rs.
988.23 million with major share of roses (Rs 300 million), Gladiolus (Rs 280
million) Tube Rose (Rs 181 million), and Rose budded (Rs 110 million) in 1999-
2000. An example of a change in the methodology of estimation is that in the
previous base year, harvest prices were extrapolated using WPI due to the non-
availability of any reliable data for major crops, but in the 1999-2000 data, harvest
prices have been used.10

The difference in the valuation techniques of intermediate inputs, such as seeds,

fertilizer, pesticides, and insecticides leads to a variance in the levels of agriculture
production on current prices between the two base periods. The changes in the

Figure 3. Agriculture Output

curre nt prices constant prices
1,200 Old Series
1,200 Old Series
New Series
New Series
1,000
1,000
(billions of rupees)
(billions of rupees)

800 800

600 600

400 400

200 200

0 0
1999-00 2000-01 2001-02 2002-03 1999-00 2000-01 2001-02 2002-03

10
For details, see Federal Bureau of Statistics (2004). The book does not provide exhaustive list of
new items and their data to know the exact details of increase in nominal GDP due to new coverage
and revised methodology; for instance, one is unable to find the complete information and data about
the increase in nominal GDP (Rs 607,357 million) in year 1999-2000 according to new coverage.
65 Opinions

inputs in the new series are also the result of their redistribution among major and
minor crops. This increase in the new series is offset, albeit slightly, by the
reclassification of slaughtering from livestock to the manufacturing sector.

In comparison, the annual average percentage difference at constant prices

between the two series is 451 percent. This difference is due to two main reasons:
the enhanced coverage and reworking methodology and due to change of base
year from 1980-81 to 1999-2000. Since the old base year prices were much lower
than that of the new base year, this change resulted in higher AAPD. Figure 3
shows the decomposition of the total change in the real estimates of the
agriculture, industry and services sector respectively.

Disaggregating the causes of this increase, it is found that 384 percent of the
AAPD of the agriculture sector is attributed to the rebasing effect. This effect
shows the increase in agriculture estimates due to the difference in prices of the
new base year and the old base year. On the other hand, 67 percent of the increase
in AAPD is because of enhanced coverage and the new classification, reworking,
or estimation methodology.

Annual Average Percentage Point Difference (AAPPD)11of growth rates at

constant prices between the two series is 0.24 (Table 4 and Figure 4). This shows
that the new data series has grown faster than the old series, albeit insignificantly,
at less than one percent point over these four years. At current prices, likewise, the
annual percentage point difference is about 1.1 showing relatively faster growth
than the old series. In 2000-01 and 2002-03 the new series has lower growth rates
than that of the old series, contrary to 2001-02.

Table 4. Agriculture Growth Rates

at current prices at constant prices
Years Old Series New Series % point Old Series New Series % point
(percent) difference (percent) difference
2000-01 2.71 2.35 -0.37 -2.74 -2.18 0.56
2001-02 -2.14 2.43 4.57 -0.07 0.10 0.17
2002-03 10.35 9.40 -0.95 4.15 4.15 0.00
Annual Average 1.09 0.24

11 n
AAPD = ∑ (1 / n)[(GNS − GOS ) . Where, GNS is the growth of sector/sub-sector according to new
i =1

series, GOS is growth of sector/sub-sector according to old series, and n is the number of
observations.
Opinions 66

Figure 4. Agriculture Growth

nominal re al
12 o ld s eries 6 o ld s eries
new s eries
new s eries
8 4
percent

percent
2
4
0
0
-2

-4 -4
2000-01 2001-02 2002-03 2000-01 2001-02 2002-03

The difference in the growth rates between the old and the new series at current
prices has arisen due to variation in the volumes of items added in new coverage
and of their yearly respective prices over 1999-2000 to 2002-03. On the other
hand, in the case of constant prices this difference is only due to the change in
quantities of increased items in the new coverage of base year 1999-2000.
Consequently, the fluctuations in new series at current prices which is causing the
difference between the old and the new series, is higher than that of the series at
constant prices. Figure 3 shows this fact through higher gap between the old and
new series at current prices as compared to constant prices.

2.1.2. Industry

In line with SNA-93, the industrial sector has gone through several changes in
terms of coverage and estimation methodology. In addition, sub sectors have also
been reclassified according to the prescribed guidelines. Consequently, the
estimates of industry according to the new base are higher than the previous base.

AAPD on current prices is, coincidentally, again 20.2 percent, which shows
increase in the industry estimates owing to the enhanced coverage and changed
estimation methodology. Slaughtering, which was part of livestock category, has
been re-categorized as a sub sector of manufacturing in the revised classification.
In the mining and quarrying sub sector surface minerals, allied services and
mineral exploration establishments categories, which were entirely missing in the
1980-81 base, have been included in the revised estimation. Surface minerals and
67 Opinions

Figure 5. Industrial Output

current price s constant price s
1,200 1,000 o ld s e rie s
o ld s e rie s ne w s e rie s
1,000 ne w s e rie s 800
(billions of rupees)

(billions of rupees)
800
600
600
400
400

200 200

0 0
1999-00 2000-01 2001-02 2002-03 1999-00 2000-01 2001-02 2002-03

allied services mineral exploration jointly augmented Rs 3,128 million value

addition to the sector.

Moreover, in industry’s gas sub-sectors Compress Natural Gas (CNG) and water
supply, absent in old coverage, contributed Rs 282 and Rs 16,978 million
respectively, according to new coverage. Similarly, instead of deducting a lump
sum 20 percent fixed input cost from the total gross output of this sub sector, as
done previously under the income/cost approach, separate input cost ratios have
been calculated in the 1999-2000 base. Similar changes have been made in the
categories of construction and electricity and gas distribution. These changes have
collectively altered the volume of industry as compared to the old base.

AAPD on constant prices between the new and old series is about 390 (Table 5
and Figure 4). As explained earlier for Agriculture output, this difference is due to
the price hike from 1980-81 to 1999-2000, enhanced coverage, and revised
estimation methodology. The break up of AAPD shows that out of 388 percent,
342 percent increase in industrial estimates is due to the change in base year and
the remaining is attributed to the modification in coverage and methodology. At
current prices, only the change in coverage and methodology has affected the data.

Table 5. Industry Output

at current prices at constant prices
Years Old Series New Series Percent Old Series New Series Percent
(millions of rupees) Difference (millions of rupees) Difference
1999-00 676,369 798,190 18.0 162,457 798,190 391.3
2000-01 721,426 895,044 24.1 166,447 827,229 397.0
2001-02 787,018 938,394 19.2 175,377 849,139 384.2
2002-03 871,250 1,031,404 18.4 184,913 889,031 380.8
Annual Average 19.9 388.3
Opinions 68

Table 6. Industry Growth Rates

at current prices at constant prices
Years Old Series New Series % point Old Series New Series % point
(percent) difference (percent) difference
2000-01 6.66 12.13 5.47 2.46 3.64 1.18
2001-02 9.09 4.84 -4.25 5.37 2.65 -2.72
2002-03 10.70 9.91 -0.79 5.44 4.70 -0.74
Annual Average 0.14 -0.76

The growth rate of industry has also changed in the revised estimates of the new
base (Table 6 and Figure 6). There is also a difference in the AAPD between the
new and old series, at both current and constant price data. The growth rate of new
series, at current prices, is higher than that of the old series in fiscal year 2001 and
2003, contrary to 2002. However, AAPD is about 0.14 percent, which is
insignificant. These percentage point differences are the result of both changes in
the volumes and prices of the increased items at current price estimates. On the
other hand, using constant price data, the average growth rates of the new series
are less than the old series. The core reason of this outcome is the variation in the
relative quantities of increased items between any two years in the new coverage.

2.1.3. Services Sector

This sector consists of transport, storage and communication, wholesale and retail
trade, insurance, ownership of dwellings, public administration and defense,
community, social, and personal services. It is mainly the services sector in which
many structural developments and changes have taken place in the last two
decades. In order to incorporate all these structural changes it was all the more

Figure 6. Growth Rates

nominal 6 re al
16

12
4
(percent)
(percent)

8
2
4 o ld s e rie s o ld s e rie s
ne w s e rie s ne w s e rie s
0 0
2000-01 2001-02 2002-03 2000-01 2001-02 2002-03
69 Opinions

Table 7. Services Output

At current prices At constant prices
Years Old Series New Series Old Series New Series
Percent Percent
(millions of rupees) Difference (millions of rupees) Difference
1999-00 1,465,927 1,807,546 23 318,740 1,807,546 467
2000-01 1,644,674 2,035,680 24 333,756 1,863,396 458
2001-02 1,806,357 2,188,527 21 347,274 1,952,146 462
2002-03 1,973,592 2,390,988 21 365,815 2,053,979 461
Annual Average 22 462

important to rebase the national income accounts.

Special study was made to incorporate the estimates of travel agencies, courier
services, tour operators, and inland water transport. The coverage in this sector has
also been broadened through the inclusion of mobile phone, internet, and courier
services. The value of courier services and mobile phone services were Rs 5,797
and Rs 4,717 million in the year 1999-2000. In addition, the estimation
methodology of the services sector has also been improved. Data on non-
mechanized transport, trade margins in wholesale and retail trade and finance and
insurance has also been readjusted and reclassified.

In the revised base, data on investment companies (Rs 271 million) and exchange
companies (168), discount and guarantee houses (69), venture capital and
insurance companies (15), and Postal Life Insurance has been compiled for the
first time along with the change in methodology to estimate the insurance and
pension funds. Consequently, all these factors have contributed to the increase in
the volume of the services sector as compared to the old base data (Table 7 and

Figure 7. Services Output

3,000 current prices 3,000 constant price s
o ld s eries
o ld s eries
(millions of rupees)

new s eries
(millions of rupees)

new s eries

2,000 2,000

1,000 1,000

0 0
1999-00 2000-01 2001-02 2002-03 1999-00 2000-01 2001-02 2002-03
Opinions 70

Table 8. Services Growth Rates

At current prices At constant prices
Years Old Series New Series % Point Old Series New Series % Point
(percent) difference (percent) difference
2000-01 12.19 12.62 0.43 4.71 3.09 -1.62
2001-02 9.83 7.51 -2.32 4.05 4.76 0.71
2002-03P 9.26 9.33 0.07 5.34 5.30 -0.04
Annual Average -0.61 -0.32

Figure 7). The AAPD is 22.4 percent and 462.4 percent at current and constant
prices respectively between the new and old series. AAPD at constant factor cost
is 403 percent due to rebasing, and 59 percent because of enhanced coverage and
new methodology.

Similarly, growth rate of this sector in different years has changed that result in the
difference in percentage points of growth between new and old series both at
current and constant prices (Table 8 and Figure 8).Growth rate of the new series is
slower than that of the old series both at current and constant prices. AAPD levels
are negative 0.61 percent and negative 0.32 percent at current and constant prices
respectively. These differences are not substantial, as they are lower than one
percentage point. The rationale for these differences in growth rates is the change
in estimation methodology, enhanced coverage, and rebasing from previous base
year to new base year.

Figure 8. Services Growth Rates

nominal real
13 6 o ld s eries
o ld s eries new s eries
new s eries
(percent)

11 5
(percent)

9 4

7 3
2000-01 2001-02 2002-03P 2000-01 2001-02 2002-03P
71 Opinions

Figure 9. Sectoral Shares (constant prices) a gric ulture

indus try
s e rvic e s
O ld Series New Se rie s
25% 25%

50%
52%

25% 23%

2.2. Sectoral Shares

By looking at the sectoral shares of agriculture, industry, and services in GDP, it

can be seen that the share of agriculture has increased in the new series at constant
prices as compared to the old series on average from year 2000 to 2003. It is due
to the comparatively more increased coverage than industrial sector (Figure 9). On
the other hand, its share has decreased at current price series. This decrease can be
explained on the basis of the non availability of harvest prices in the old series
(Figure 10). Therefore, WPI that was higher than the harvest price was used for
major crops. The share of the services sector according to the new series is greater
than that of the old series both at current and constant basis as a whole due to

Figure 10. Sectoral Shares (current prices) a gric ulture

indus try
Ne w Series s e rvic e s
O ld Se rie s
25% 24%

52% 53%

23% 23%
Opinions 72

comparatively more increased coverage than industrial sector. On the other hand,
share of the industry sector of the new series is less than that of the old series both
at current and constant prices.

3. Final Remarks

To sum up, rebasing national income accounts along with its enhanced coverage,
which was long overdue, has provided reliable and accurate data incorporating
many structural developments and changes in the economy’s relative prices since
1980-81. Moreover, preparing national accounts according to the SNA-93
standards has made the data internationally comparable. But it is very important,
as mentioned before, to rebase the national income accounts along side enhancing
coverage after, at least, every five years to incorporate the structural changes that
take place in the economy.

Furthermore, another primary task ahead for the Federal Bureau of Statistics is the
need to link the data series of different base periods by using the reworking
methodology since 1948-49. Because, linking data through splicing method does
not depict the most accurate figures. Certainly this requires an uphill task and
involves many practical difficulties. But once the task is done it would provide
very useful information for analytical purposes. The practice has already been
done in linking the data since 1960 for the base year 1980-81.

Appendix 1: Revaluation Techniques of Constant Price Data

Constant price estimates are obtained through three techniques: revaluation,

deflation, and volume extrapolation.12

(i) Revaluation: In this method, quantity of each item is valued by employing the
base-year prices: Qo,t = ∑i pi ,o * qit . Where, pio is the price of ith item in the base
year and qit is its quantity in period t . As a result, the estimates obtained are at
constant (base) year prices.

(ii) Deflation: This method requires a deflation of each period’s current price
value with a suitable price index. For this, current value of each period is divided
by a suitable price index: Qot = Vt / Pot Where Vt is ∑i p i ,t * q it . That is value in
period t and Pot is a price index.

12
For details of gross value added at constant prices, see Asian Development Bank (2002).
73 Opinions

(iii) Volume Extrapolation: In this method, base year values are revised by taking
the product of the values with an appropriate volume index that yields the result as
Qo = ∑ (qi ,t / qi ,o ) * pi ,o qi ,o . Generally, constant price estimates obtained through the
deflation method are considered more accurate than those derived through
extrapolation given the fact that price relatives ( pt / po ) observe less variation than
quantity relatives (qt / q o ) , except in periods of hyperinflation.

Rebasing of national income accounts in Pakistan from 1980-81 to 1999-2000 has

employed a combination of the following three methods:

- Using the revaluation method, a change of base year involves replacing 1980-81
prices with 1999-2000 prices for identical items; that is, replace Q81, t = ∑ i pi ,81 * qit
with Q2000,t = ∑i p i , 2000 * q it .

- Items for which the constant price estimates are derived through deflation a
change of base year requires changing the reference year from 1980-81 to 1999-
2000 for the deflators, used at detailed level. It involves division of the original
index with its level in 2000; that is, Q2000,t = Vt /( Po,t / Po, 2000 ) .

- For the items and aggregates for which constant price estimates are obtained
through volume extrapolation, a change in base year involves changing the
reference period of the volume index from 1980-81 to 1999-2000 and multiplying
the re-referenced volume index with the current price level in 1999-2000
[Q2000,t = V 2000 * ( I 81,t / I 81, 2000 )] .

Appendix 2: Decomposition of Total Change in Real Estimates

RE1999 − 2000 − RE1980 −81 NE1980 −81 − RE1980 −81

TotalChange = ; Pr iceEffect = ;
RE1980−81 RE1980−81
RE1999 − 2000 − NE1980 −81
CoverageEffect =
RE1980 −81

Where, RE1999− 2000 is real estimate at 1999-2000 base, RE1980−81 is real estimate at
1980-81 base, and NE1980−81 is nominal estimate at 1980-81 base.

These formulae are for individual years. The average of all years has been taken
for the whole period since 1999-2000 to 2002-03, as is shown in Figure 3.
Opinions 74

References

Asian Development Bank (2002). RETA 5874: Compiling, Rebasing, and Linking
National Accounts in the Asian and Pacific Region. Manila:ADB
Federal Bureau of Statistics (2004). National Accounts of Pakistan Rebasing From
1980-81 to 1999-2000. Islamabad: Statistics Division.
Government of Pakistan (2003-2005). Economic Survey. Islamabad: Planning
Division, Ministry of Finance.
State Bank of Pakistan (2000-2003). Annual Reports. Karachi: SBP.