Discussion Paper - 2

Estimates of District Domestic Product

R. P. Katyal, M. G. Sardana
and J. Satyanarayana
Socio-Economic Research Centre,
New Delhi

United Nations Development Programme

55, Lodi Estate
New Delhi - 110003
India
2001

The analysis and policy recommendations of this Paper do not necessarily reflect the views of i
the United Nations Development Programme, its Executive Board or its Member States.
ii
 Prologue
The central and state authorities and research institutions often require data on economic activities and levels of
living of people at the district level for planning and policy purposes. Income of a district is considered one of
the most important indicators/barometers to measure the economic growth/development of a district vis-à-vis
others. Preparation of District Income estimates has recently gained added importance, since it is one of the
three indicators to construct a composite Human Development Index (HDI) for inclusion in the Human Devel-
opment Reports (HDR) being prepared by several states in India.

The present paper has been written in a scenario where the national level data system is rigorous and well
designed, with matching state data but very little information at the district level. The 73rd and 74th constitutional
amendments have brought in a third tier of government – the Panchayats – into India. How does this major
change impact on data systems? The need thus is for district level indicators not only to make the state HDR
more relevant, but also to improve local understanding of the development process.

This background paper was presented at the National Workshop on State HDRs and the Estimation of District
Income & Poverty under the State HDR Project that is executed by the Planning Commission, GOI, with UNDP
support held in Bangalore in July 2001. It is the second in the HDRC’s Discussion Paper Series.

Dr. M.G. Sardana is the Chairperson and Dr. R.P. Katyal and Dr. J. Satyanarayana are Directors at the Socio-
Economic Research Centre, New Delhi

iii
 GLOSSARY

AIDIS All India Debt and Investment Survey

ASI Annual Survey of Industry
CCS Cost of Cultivation Studies
CIF Chief Inspector of Factories
CSO Central Statistical Organisation
DDP District Domestic Product
DES Directorate of Economics and Statistics
DPD Data Processing Division
DTE Directory Trade Establishment
FISIM Financial Intermediary Services Indirectly Measured
FOD Field Operation Division
GVA Gross Value Added
HDI Human Development Index
ISS Integrated Sample Survey
NIC National Industrial Classification
NAD National Accounts Division
NSSO National Sample Survey Organisation
ONGC Oil and Natural Gas Corporation
PPP Purchasing Power Parity
SDP State Domestic Product

iv
 CONTENTS

1. INTRODUCTION 1

2. METHODOLOGY FOR DISTRICT INCOME ESTIMATES 3

Commodity Producing Sectors 4

Non-Commodity Producing Sectors 4

3. COMMODITY PRODUCING SECTORS 5

Agriculture including Animal Husbandry 5

Value of Output of Crop Husbandry 5

Value of Output of Animal Husbandry 6

Value of Inputs for Crops and Animal Husbandry 7

Operation of Government Irrigation System 9

Forestry and Logging 9

Fishing 10

Mining and Quarrying 10

Manufacturing 10

Construction 12

Electricity, Gas and Water Supply 12

4. NON-COMMODITY PRODUCING SECTORS 13

Trade, Hotels and Restaurants 13

Transport, Storage and Communication 14

Banking and Insurance 15

Real Estate, Ownership of Dwellings and Business Services 15

v
 Public Administration 16

Other Services 16

Estimates at Constant Prices 17

5. ISSUES FOR DISCUSSION 18

Income Originating vs. Income Accruing Approach 18

Frequency of Estimates 18

Purchasing Power Parity and HDI 18

DDP and SDP 18

Prices 19

Yield Rates 19

Forest Boundaries 19

Inputs in Forestry and Fishing 19

Manufacturing - Registered 19

Unorganised Segments of Economic Activity 20

Workforce 20

Value Added per Worker 20

BIBLIOGRAPHY 21

ANNEXURE 22

vi
 Introduction

For planning and policy purposes, the of information. The Seventh Conference
central and state authorities and research of Central and State Statistical
institutions often require data on eco- Organisations which was held at
nomic activities and levels of living of Hyderabad in December 1985, discussed
people at the district level. Income of a issues relating to the estimation of in-
district is thought to be one of the most come at district as well as at rural and
important indicators/barometers to mea- urban levels. Keeping in view the demand
sure the economic growth/development from the Planning Commission and the
of a district vis-à-vis others. Preparation state governments for income estimates
of District Income Estimates has re- at the district and rural/urban levels, the
cently gained added importance, since it Conference recommended constitution Income of a district is
is one of the three indicators to construct of a Technical Group to examine the re- thought to be one of
a composite Human Development Index quirements of data and recommend ap- the most important
(HDI) for inclusion in the Human De- propriate methodology for compilation indicators/barometers
velopment Report being prepared by of these estimates. In pursuance of this to measure the
many State Governments in India. The recommendation, the Department of economic growth/
other two indicators for calculating HDI Statistics constituted a Technical Group development of a
are life expectancy and educational attain- for Estimation of Income at District and district vis-à-vis
ment. Rural/Urban Levels in January 1987. The others.
Group in its Report, which was submit-
Some of the State Governments have ted in September 1988 made the follow-
started preparing estimates of district in- ing observations (CSO, 1988):
come to measure income disparity and
to plan for the development of backward l To begin with, attempts should be
districts during the 1970s and early 1980s. made to compile district income es-
For example, the state of Uttar Pradesh timates following the standard meth-
compiled district income estimates for all odology, which is based on the in-
sectors for 1968-69 but later restricted come originating approach. This
the preparation of these estimates to method is similar to the one used for
commodity producing sectors only. compilation of state income esti-
mates. Due to the free flow of goods
The estimation of income at district level and services across district borders
is beset with the problems of availability and the non-availability of net factor
of data as well as collection and analysis income earned by the residents of

1
 other districts/states/countries, the determining the backwardness/de-
income accruing approach is not fea- velopment of a district, and as a con-
sible. sequence for allocating resources, it
would be necessary to make essen-
l Even to compile estimates following tial adjustments in these estimates
the income approach, considerable with regard to significant flow of in-
additional data, as detailed in the Re- come across the territories of dis-
port, needs to be collected. tricts which are rich in minerals and/
or forest resources or where flow of
l For district income estimates, a large daily commuters, migration of labour
amount of information would need and cattle are involved.
to be collected and compiled at the
district as well as at state level. The l With the present availability of data
Group felt that the state governments at the district level, it would be suffi-
should assess the precise require- cient to compile these estimates at
With the present ments for additional resources for an interval of five years.
availability of data at implementing the recommendations
The Group also detailed the data require-
the district level, it of the Group and make necessary
ments for compilation of district income
would be sufficient to provisions in their core schemes.
estimates and annexed the essential items
compile these
l Since estimates at the district level for which data need to be maintained at
estimates at an
would be utilised for the purpose of the district level.
interval of five years.

2 ESTIMATES OF DISTRICT DOMESTIC PRODUCT

 Methodology for District Income Estimates

With a view to giving impetus to the The methodology for compiling the
compilation of district income estimates DDP, given in the following paragraphs,
by the state governments and to takes into consideration the broad guide-
standardise the methodology of the Na- lines given in the Report of the Techni-
tional Accounts Division (NAD), the cal Group on District Income and also
Central Statistical Organisation (CSO) its recommendations and the methodol-
convened a meeting with the State Di- ogy which was developed jointly by the
rectorates of Economic and Statistics DES of Karnataka and Uttar Pradesh in
(DES) in 1995. The Meeting entrusted August 1996. The proposed methodol-
the development of the methodology of ogy takes into account the available data Broadly, the
compilation of District Domestic Prod- at the district level for the commodity methodology of
uct (DDP) jointly to the DES of the producing sectors and the results of the computation of
states of Karnataka and Uttar Pradesh surveys, both relating to socio-economic sectoral estimates is
which was completed in August 1996. aspects and unorganised sectors of the the same as adopted
The state governments have been fol- economy, conducted by the National for estimates of State
lowing this methodology for preparation Sample Survey Organisation (NSSO) and Domestic Product
of district income estimates. states DES. For many sectors of the (SDP).
economy, the proposed methodology
A number of state governments have al- avoids allocation of state estimates to
ready initiated the process of prepara- districts in proportion to the districtwise
tion of DDP. While Andhra Pradesh, workforce, as this does not take into ac-
Bihar, Haryana, Karnataka, Kerala, count the income disparity at the district
Madhya Pradesh, Maharashtra, level.
Rajasthan, Tamil Nadu and Uttar
Pradesh are compiling estimates for all Broadly, the methodology of computa-
the sectors of the economy, Arunachal tion of sectoral estimates is the same as
Pradesh and Assam have restricted this adopted for estimates of State Domestic
to the commodity producing sectors. Product (SDP). The database for estima-
Specific comments on the methodology tion of DDP is still not satisfactory for
adopted for compiling the estimates of any of the states. Data for compilation
DDP by Karnataka, Maharashtra, and of DDP for commodity producing sec-
Tamil Nadu are given in the Annexure. tors viz., primary sector and manufactur-
ing (registered) sector, are mostly avail-

Methodology for District Income Estimates 3

 able on a regular basis but are very scanty For compiling DDP, the economy may
in respect of the remaining sectors. As be divided into following 14 broad sec-
such, wherever districtwise basic data are tors (8 commodity producing and 6 non-
available, they may be utilised to compute commodity producing sectors) as done
the DDP estimates, following the meth- in the case of SDP.
odology adopted at the state level. In the
case of non-commodity producing sec- Commodity Producing Sectors
tors, where districtwise basic data are not l Agriculture
available, the state level estimates may be l Forestry and Logging
allocated to the districts on the basis of l Fishing
suitable districtwise indicators. Further, l Mining and Quarrying
in some of the commodity producing l Manufacturing – Registered
sectors, though districtwise production l Manufacturing – Unregistered
data are available, the corresponding l Construction
prices are not available. In such cases, l Electricity, Gas and Water Supply
districtwise production may be evaluated
using prices prevailing in the adjoining Non-Commodity Producing
district/region. Sectors
l Trade, Hotels and Restaurants
For the commodity producing sectors, l Transport, Storage and Communication
the estimates of DDP may be prepared l Banking and Insurance
using the production approach i.e., gross l Real Estate, Ownership of Dwellings
value added is equal to value of output and Business services
minus value of inputs, whereas for ser- l Public Administration and Defence
vice sectors income approach i.e., gross l Other services
domestic income is equal to compensa-
tion of employees plus gross operating A detailed sector-wise methodology
surplus, may be used. follows.

4 ESTIMATES OF DISTRICT DOMESTIC PRODUCT

 Commodity Producing Sectors

Agriculture including Animal worked out by using districtwise

Husbandry production and prices.

As in the case of SDP, the estimates of l Crops for which districtwise data on
value of output of agriculture and ani- production is available but corre-
mal husbandry may be prepared sepa- sponding districtwise prices are not
rately. Since farmers feed the animals to- available. For this category of crops,
gether and do not keep separate account districtwise value of output may be
of inputs for agricultural crops and ani- computed by using production data
mal products, the estimates of inputs can- of the district and available price data
not be prepared separately. As such, the from a neighbouring district. In case
value added estimates may be prepared neighbouring district prices are also Value added
for the overall activity of agriculture. not available, regional prices may be estimates may be
used. If these are also not available, prepared for the
Value of Output of Crop the only alternative is to use state overall activity of
Husbandry prices. Efforts should be made to agriculture
collect prices in respect of all such
As in the case of SDP, the estimates at crops for which area and yield esti-
district level can also be worked out us- mates are available.
ing the ‘production approach’. But dis-
trict level data on production and prices l Crops for which districtwise data,
are generally not available uniformly for both on production and prices, are
all the crops. Data is compiled by State not available but crop area figures
directorates for “Major Crops” as listed are available. In such cases,
by CSO. However the category “Minor districtwise value of output of these
Crops” varies according to specific states. crops may be worked out by allocat-
Hence, for estimation purposes, crops ing the state level value of output on
may be classified in the following the basis of districtwise crop area.
categories:
l Miscellaneous and unspecified crops
l Crops for which districtwise data on for which production and price data
production and prices (primary mar- are not available even at the state
ket or farm harvest prices) are avail- level but districtwise crop area is
able. In such cases, districtwise value available. In these cases, the value
of output for each crop may be of output may be worked out by mul-

Commodity Producing Sectors 5

 tiplying the districtwise crop area stock products and by-products and
with state level value per hectare of poultry meat.
respective crops.
In general, the results of Integrated
For by-products or miscellaneous type Sample Survey (ISS) conducted by State
of products, wherever possible, Cost of Animal Husbandry Department provide
Cultivation Studies (CCS) data on yield state/district level data on production of
rates, inputs, etc .may be used. In the ab- milk, meat, wool and eggs. The survey
sence of data of CCS for the district, at- also provides estimates of the number
tempts may be made to undertake type of slaughtered animals category wise.
studies to fill this gap. Till such time as Wherever, ISS results are available only
the results of studies or data from other for the state, the estimates for districts
sources at the district level become avail- may be obtained by allocating the state
able, state level ratios, whether in rela- estimate to the districts on the basis of
tion to output or in relation to crop area, relevant livestock population. To esti-
as used in current series of estimates of mate districtwise production of all other
SDP, may be adopted uniformly for all items (i.e. other than milk, meat, eggs and
the districts. wool), state level yield rates and ratios
may be utilised along with district esti-
For the quantity of food grains procured mates of the number of relevant ani-
on Government account, the prices are mals/poultry.
different from the prevailing market
prices. As such, the quantity procured by Districtwise value of output may then be
the Government may be evaluated at pro- worked out by multiplying the produc-
curement prices and the rest of the pro- tion obtained as above, by the corre-
duce at the primary market/farm harvest sponding district prices, wherever avail-
prices. able. In the absence of region or district
level prices, state prices may be utilised
Value of Output of Animal for evaluating district production. Total
Husbandry value of output thus worked out may
then be adjusted to the state level esti-
A procedure similar to the one adopted mates.
for compiling SDP may be followed to
estimate the value of output from ani- With regard to the items for which price
mal husbandry for the district. data are not available, wholesale/retail
Districtwise estimates of a number of prices of allied items should be utilised.
different categories of animals and poul- The price data can be used in various
try may be worked out using the results cases as indicated in the table.
of the latest two/three livestock cen-
suses, assuming linear/compound In the case of other animal products, for
growth rates. These may be used along which neither the yield rates nor the
with relevant yield rates to obtain the prices are available at the district level,
estimates of production of various live- the state level value may be allocated to

6 ESTIMATES OF DISTRICT DOMESTIC PRODUCT

 Item Price to be used
l Milk Rural price/cooperative’s collection price
l Meat (buffalo) Slaughter house price
l Goat, sheep and pig meat Slaughter house price
l Hides (cow and buffalo) Slaughter house price
l Skins (goat and sheep) Urban wholesale price of goat skin
l Bones (of various animals) Urban wholesale price of bone
l Dung Rural prices of dung cake
l Eggs (hen and duck) National Egg Coordination Committee prices

the districts in proportion to the number the Animal Husbandry sector, the value
of animals of each category separately. of various input items are to be first de-
ducted from the gross value of output
Districtwise increment in stock for each of these sectors to which the gross prod-
category of animal/poultry may be uct from operation of government irri- To arrive at the
worked out as at the state level and these gation system is added. The deductible districtwise gross
may be evaluated by corresponding dis- inputs are the same as used in compila- value added (GVA)
trict prices. tion of SDP viz. from agriculture,
Seed including the Animal
The value of silkworm cocoons compiled l
Husbandry sector, the
for the SDP may be allocated to districts l Organic manure (cattle and buffalo
value of various input
in proportion to the area under mulberry dung)
items are to be first
in each district, after adjusting for the cost
l Chemical fertilisers deducted from the
of rearing silkworm cocoons.
l Feed of livestock gross value of output
Keeping in view the data gaps, it is sug- of these sectors to
l Pesticides & insecticides
gested that the coverage of ISS may be which the gross
l Diesel oil consumption product from
enlarged by the respective state govern-
ments so as to generate the estimates of l Electricity operation of
production of each of the animal prod- l Irrigation charges government irrigation
ucts at the district level. Further, prices system is added
l Market charge
of agricultural and animal products at the
l Repairs and maintenance of fixed
district level need to be collected and
assets and other operational costs and
compiled from representative centres so
that for each of the products price data l Financial intermediary services
are available and utilised for district in- indirectly measured (FISIM).
come estimates. To work out the value of these input items at
the district level, the state level estimates ar-
Value of Inputs for Crops and
rived at independently under the current
Animal Husbandry
series of state income estimates may be dis-
To arrive at the districtwise gross value tributed to the districts on the basis of the
added (GVA) from agriculture, including procedure indicated below:

Commodity Producing Sectors 7

 l Seed: Seed rates and area sown for universities that are engaged in such
each of the important crops at the studies in the State and analysed. In
district level are generally available the absence of such study results,
with the district agricultural depart- the rates of concentrates as adopted
ments. These data should be col- for state level estimates for differ-
lected and the quantity of seed for ent types of animals and poultry
each of the crops may be worked may be used for working out dis-
out as a product of seed rate and trict level estimates.
area under each crop. This may be
multiplied by the respective prices l Pesticides and insecticides: State
at the district level to arrive at the level value may be allocated to the
value of seed input. For other crops districts in proportion to area
for which seed rates are not avail- treated by chemical pesticides in
able at the district level, the input different districts. Efforts may be
of seed for each of the crops may made to obtain the results of CCS
State level value of be worked out by allocating the in a state at the district level and the
consumption by value at state level in proportion to value of pesticides and insecticides
tractors and oil the area under each crop at the dis- districtwise may be worked out.
engines may be trict level.
distributed to districts l Diesel oil consumption: State level
in proportion to the l Organic manure: State level value value of consumption by tractors
district weighted of output may be allocated to the and oil engines may be distributed
totals of number of district in proportion to the to districts in proportion to the dis-
tractors and oil districtwise estimates of value of trict weighted totals of number of
engines/oil engine output of dung manure, as worked tractors and oil engines/oil engine
pumps used for out in the Animal Husbandry sec- pumps used for irrigation as per lat-
irrigation as per latest tor. est livestock census, weight being
livestock census the per unit consumption of diesel
l Chemical fertilisers: State level oil based on the data collected from
value may be distributed to districts the schedules of CCS. If district
in proportion to total quantity of level consumption of diesel per unit
chemical fertilisers distributed, as is not available, then state level
obtained from the Agriculture De- norms may be adopted.
partment.
l Electricity: State level value may be
l Feed of livestock: Districtwise value distributed on the basis of
of roughage may be estimated by districtwise number of private and
adopting the same methodology as government electric tube-wells and
followed for estimating the rough- energised pump-sets.
age in the case of SDP. For estimat-
ing the quantity of concentrates l Irrigation charges: Districtwise ir-
consumed, the returns of CCS may rigation charges, as collected from
be collected from the agricultural the district administration may be

8 ESTIMATES OF DISTRICT DOMESTIC PRODUCT

 used for this purpose. In the ab- Forestry and Logging
sence of such data, the state level
value may be distributed to the dis- Major Forest Products
tricts in proportion to the area irri-
gated by government canals. The forest circles/divisions/sub-divi-
sions are generally not coterminous with
l Market charges: Market charges district boundaries. The records of the
may be ascertained from the pri- forest/revenue departments at the divi-
mary markets within the district and sion/sub-division/forest range/forest
used. In the absence of such data, coupe levels will have to be looked into,
state level norms may be used for so as to demarcate forest areas within a
district income estimates also. district. For example, in case of Madhya
Pradesh, the boundaries are now cote-
l Repair and maintenance of fixed rminous with one or more forest
assets and operational costs: divisions.
Districtwise data on fixed assets as
available from latest All India Debt Districtwise value of output of timber
and Investment Survey (AIDIS)/ may be worked out using districtwise pro-
Livestock Census may be used to duction and prices for different varieties.
Districtwise value of
allocate the state total to the districts. output may be
Alternatively, the districtwise value In the case of firewood, consumption of
worked out using the
of output may be multiplied with firewood may be estimated at the district
districtwise
the state level ratios of repairs and level using the pooled results of central
production and prices
maintenance etc. to the value of and state samples of the latest available
of different products
output. data (55th Round) of NSSO relating to
household consumer expenditure. The
l FISIM: State level value of FISIM value of firewood at the state level may
may be distributed on the basis of be allocated to the districts in propor-
total value of output of the district. tion to the districtwise consumption of
firewood.
Operation of Government
Irrigation System Minor Forest Products

Data on salaries and wages are likely to Districtwise value of output may be
be available with the district agriculture worked out using the districtwise produc-
departments. These should be collected tion and prices of different products
and used for allocation of the state level wherever available from the Minor For-
value of contribution by government est Products Corporation etc. For other
irrigation system to the districts in pro- products, in the absence of districtwise
portion to the districtwise salaries and data, the state level value of other minor
wages/area irrigated by government forest produce may be allocated to dis-
canals. tricts in proportion to the forest area in
the districts.

Commodity Producing Sectors 9

 Repairs, Maintenance and other Mines and Geology Department. If the
Operational Costs data are not available, the value of out-
put of minor minerals may be worked
State level ratio may be utilised for work- out in proportion to the royalty value. For
ing out districtwise estimates of repairs, working out input costs, the state level
maintenance and other operational costs. proportions may be adopted.

Fishing Manufacturing
In the case of marine fishing, the Registered
districtwise value of output may be
worked out by multiplying the The Annual Survey of Industry (ASI)
districtwise production by the corre- being conducted by the Field Operation
sponding district prices. A similar pro- Division (FOD) of the NSSO provides
cedure may be followed in the case of the requisite data for compiling estimates
inland fish also. The district wise data of GVA for manufacturing-registered
In case the on production of marine and inland fish sector (units registered under the Facto-
districtwise prices are is likely to be available from the district ries Act, 1948) both at the national and
not available, then fishery officer. In case the districtwise state levels. The sampling strategy
districtwise value may prices are not available, then districtwise adopted by ASI till 1996-98 aimed at pro-
be worked out using value may be worked out using viding data to generate district level esti-
districtwise districtwise production and adjoining mates by the DES.
production and district/region prices. Data on opera-
adjoining district/ tional costs, including repairs and main- On the recommendation of a Steering
region prices tenance are not available even for esti- Committee constituted by the Depart-
mating SDP and the national level norms ment of Statistics and Programme Imple-
are used for SDP. Thus, there is no other mentation, Government of India in May
alternative but to use the state level 1999, a revised sampling strategy has been
norm/ratios for DDP estimates also. put into operation from ASI 1998-99 so
as to enable the FOD of the NSSO to
Mining and Quarrying manage the field work of about 45000
units with the sanctioned field resources.
For major minerals including petroleum Under the new sampling strategy, ASI
(crude) and natural gas (utilised), the would cater to the generation of estimates
districtwise value of output as available only at all-India and State levels. All fac-
with Indian Bureau of Mines, Nagpur tories engaging 200 or more workers are
may be used. District wise value of pro- being covered on a complete enumera-
duction of coal can also be obtained tion basis (called census sector) and the
from Coal India Ltd. and that of natural remaining factories on a sample basis
gas and crude oil from Oil and Natural (called sample sector). However, in the
Gas Corporation (ONGC). For minor case of the small states /union territo-
minerals, the districtwise value of out- ries of Manipur, Meghalaya, Nagaland,
put may be collected from the State Tripura and Andaman and Nicobar

10 ESTIMATES OF DISTRICT DOMESTIC PRODUCT

Islands, all factories are being covered on the CIF contains information relating to
a complete enumeration basis. the number of workers in each registered
factory. This list can be used for estimat-
At present, the states of Assam, Orissa, ing the workforce engaged in the sample
Rajasthan, Uttar Pradesh, and West Ben- sector of factories covered in the ASI in
gal cover the residual units belonging to any of the districts, as an alternative to
the sample sector of their states with their using the data collected under the Fourth
own resources. Districtwise estimates can Economic Census.
be computed for these states by appro-
priately pooling the results of units cov- Unregistered
ered by the state field staff and the ASI
field work (census sector and sample sec- For the manufacturing unregistered and
tor) conducted by FOD of the NSSO. the unorganised segments of the remaining
Other state governments who are inter- sectors of the economy i.e. unorganised
ested in generating district level estimates trade, unorganised hotels and restaurants,
may take steps to cover additional unorganised road transport,
samples with their own resources. Till unrecognised educational institutions, State governments
such time the states DES are able to unorganised medical and health services, who are interested in
mount resources to participate in the ASI etc. the estimates of GVA are worked out generating district
programme, the GVA for the sample sec- as a product of workforce and GVA per level estimates may
tor may be allocated to the districts us- worker. While the GVA per worker is take steps to cover
ing an appropriate indicator. The value derived from the follow-up surveys of additional samples
added per worker derived from the re- economic censuses carried out by the with their own
sults relating to the census sector of the NSSO, the workforce is estimated from resources
ASI for the district multiplied by an esti- the quinquennial surveys of employment
mate of the workforce engaged in the and unemployment of the NSSO.
sample sector of ASI from the district
may be taken as the indicator for allocat- The Fifty-sixth round (2000-01) of the
ing the GVA for the sample sector of NSS covered the unorganised manufac-
the ASI for the state to the districts. The turing sector, (i.e. not registered under
data collected under the Fourth Eco- the Factories Act, 1948). Central and state
nomic Census 1998-99 can be utilised to samples may be pooled to generate the
work out an estimate of the workforce districtwise GVA per worker at 2-digit
engaged in the sample sector of facto- level of National Industrial Classification
ries covered in the ASI of the district. (NIC) as done for the SDP. GVA per
worker may be multiplied by the corre-
It is understood that the Chief Inspec- sponding estimates of workforce (dis-
tor of Factories (CIF) in most of the cussed in the following paragraphs) to
states maintains an up-to-date comput- arrive at the GVA at 2-digit level of NIC.
erised list of factories registered under The total GVA for the sector at the
the Factories Act, 1948. The field staff district level may be arrived at by
of FOD of NSSO is using it as a frame aggregation.
for ASI fieldwork. The list available with

Commodity Producing Sectors 11

 The estimates of the workforce for 1993- Electricity, Gas and Water Supply
94 at the state level, based on the 50th
round (1993-94) Survey of Employment Electricity
and Unemployment of NSS, along with
the methodological note, at 3-digit level Districtwise data on salaries and wages
of NIC were supplied by the National paid to employees are available with the
Accounts Division (NAD) of the CSO State Electricity Boards (SEBs). Special
to the states for use in the state income efforts may be made to collect/extract
estimates in March 1999. In what follows, data from the administrative records avail-
the estimates of workforce will refer to able with SEBs. The state level value
the workforce based on the surveys of added from this sub-sector may be dis-
NSSO mentioned above. tributed to the districts in proportion to
districtwise salaries and wages. The use
For any activity (2-3 digit level of NIC), of workforce alone as an indicator will
the estimate of workforce as provided not take into account the disparity in
by the NAD of the CSO, may be esti- wages and salaries at the district level.
For any activity (2-3 Although the pay scales are the same for
mated by allocating the state total to vari-
digit level of NIC), the all districts, the distribution of workers
ous districts in proportion to districtwise
estimate of workforce by different pay scales may vary signifi-
workforce, in that activity, as available
as provided by the cantly for developed and under developed
from the latest population census.
NAD of the CSO, may districts. Thus, to capture this disparity,
be estimated by From the total workforce for manufac- the indicator based on salaries and wages
allocating the state turing at the district level the correspond- will be more meaningful.
total to various ing workforce of manufacturing - regis-
districts in proportion tered based on ASI may be subtracted to Gas
to districtwise arrive at the workforce for the manufac-
workforce, in that turing - unregistered sector at 2-digit level State level value added from biogas may
activity of NIC. be distributed in proportion to the num-
ber of biogas plants in each district. For
Construction the remaining gas sector, the state level
estimates may be allocated to the districts
Data on wage rates of construction on the basis of workforce engaged in this
workers can be easily collected at the dis- activity.
trict level. The estimates of state income
from this sector may be allocated to the Water supply
districts on the basis of indicator of
workforce engaged in construction ac- State level value added may be allocated
tivity multiplied by the wage rate of con- to districts in proportion to the
struction workers at the district level. In districtwise salaries and wages collected
respect of local bodies, direct districtwise from local bodies in the district.
data may be collected and utilised.

12 ESTIMATES OF DISTRICT DOMESTIC PRODUCT

 Non–Commodity Producing Sectors

For the unorganised segments of the administrative records of the relevant

non-commodity producing sectors, public sector undertakings engaged in
districtwise GVA for a particular activ- trading activity.
ity, may be estimated as a product of
workforce and the GVA per worker. For Private Sector
the non-commodity producing sectors,
districtwise total workforce at two or The 53rd round (1997) NSS survey cov- SDP relating to public
three digit of NIC may be based on the ered unregistered trading activities by sector may be
NSS surveys of employment and unem- non-directory establishments (enter- allocated to districts
ployment as discussed under manufac- prises engaging not more than five work- on the basis of data
turing–unregistered sector. For public ers including at least one hired worker) on salaries and
sector undertakings or for the organised and own account enterprises and the wages available from
segments, the workforce will be based on states participated in this programme as the administrative
the administrative records of the under- in any other round of NSS. The direc- records of the
takings/enterprises. The workforce for tory trading establishments (enterprises relevant public sector
the unorganised segments or for the pri- engaging six or more workers including undertakings
vate sector may be derived by subtract- at least one hired worker) were surveyed engaged in trading
ing from the total workforce of an activ- by special staff of FOD of NSSO of the activity
ity, the corresponding workforce in the same rank as those engaged for conduct-
organised segment or in the public sec- ing the field work of ASI during the same
tor as the case may be. The value added year. The survey is called Directory Trade
per worker may be derived from the Establishment (DTE) Survey. The states
pooled results of central and state did not participate in the DTE survey.
samples of quinquennial surveys of The estimates at the district level for the
NSSO for the unorganised activities. benchmark year 1997 may be worked out
by pooling the results of central and state
Trade, Hotels and Restaurants samples relating to non-directory enter-
prises and own account enterprises. For
Trade other years, SDP relating to the private
Public Sector segment of the economy relating to this
sector may be allocated to each district
SDP relating to public sector may be al- in proportion to the gross value of out-
located to districts on the basis of data put of commodity producing sectors in
on salaries and wages available from the the district.

Non-Commodity Producing Sectors 13

 Hotels and Restaurants Transport, Storage and
Communication
The Fifty-seventh round (2001-2002) of
the NSS is the first attempt to cover com- Railways
prehensively all the unorganised services
State level estimates, as provided by CSO,
except trade and finance. In particular it
may be allocated to districts in propor-
covered the unorganised segments of
tion to the track length located within the
hotels and restaurants; transport, storage
district.
and communications; real estate, renting
and business activities; education; health Mechanised Road Transport
and social work; and other community,
social and personal service activities. For Public Sector
the purpose of the survey, the
SDP relating to the public sector may be
unorganised segment of any sector of the
allocated to districts on the basis of data
economy is defined to comprise those
on salaries and wages available from the
enterprises whose activities or collection
administrative records of the relevant
of data are not regulated under any legal
The Fifty-seventh public sector undertakings engaged in
provision and/or which do not generally
round (2001-2002) of this activity.
maintain any regular accounts.
the NSS is the first
attempt to cover Private Sector
Public Sector
comprehensively all
the unorganised The SDP relating to these activities in
SDP relating to public sector may be al-
services except trade the private sector may be allocated to dis-
located to districts on the basis of data
and finance tricts on the basis of district wise private
on salaries and wages available from the
workforce and the value added per
administrative records of the relevant
worker as obtained from the pooled re-
public sector undertakings engaged in
sults of central and state samples of 57th
hotel and restaurant activities.
round of NSS for the unorganised sec-
Private Sector tor. Districtwise private workforce may
be estimated by subtracting from the to-
The SDP relating to these activities in the tal workforce for this activity, the corre-
private sector may be allocated to districts sponding workforce in the public sector
on the basis of district wise private undertakings.
workforce and the value added per worker
as obtained from the pooled results of Non-Mechanised Road Transport and
central and state samples of 57th round Other Unorganised Transport
of NSS for the unorganised sector.
The value added from this activity may
Districtwise private workforce may be
be derived at the district level as a prod-
estimated by subtracting from the total
uct of district wise workforce engaged
workforce for this activity, the corre-
in this activity and the value added per
sponding workforce in the public sector
worker for this activity obtained from the
undertakings.
pooled results of central and state

14 ESTIMATES OF DISTRICT DOMESTIC PRODUCT

samples of 57th round of the NSS for Communication
the unorganised sector. Till such time
the bench mark estimates of value added Public Sector
from 57th round are derived by pooling
the central and state samples, the SDP Data on salaries and wages may be col-
relating to the private segment of this lected from the public sector units located
sector may be allocated to districts in pro- within the district to allocate the SDP
portion to districtwise gross value of from this activity. Alternatively SDP may
output of commodity producing sectors. be allocated using indicators of
districtwise revenue generated by post
Water Transport and telegraph offices, telephones, etc.,
data for which may have to be collected
Districtwise estimates may be derived from administrative records of the con-
using the same procedure as for the non- cerned offices.
mechanised road transport.
Private Sector
Air Transport
Districtwise estimates may be prepared Data on salaries and
The SDP from this sub-sector may be using the same procedure as for wages may be
allocated to the districts in proportion mechanised road transport sector. collected from the
to data on salaries and wages collected public sector units
from the local offices of the airlines. Banking and Insurance located within the
district to allocate the
Storage Data on salaries and wages can be col- SDP from this activity
lected from the regional/district offices
Warehousing (State and Central) of banks and insurance companies. The
SDP of this sector may be allocated to
State level estimates may be allocated to
districts on the basis of districtwise sala-
the districts using the same procedure as
ries and wages.
for the mechanised road transport.

Cold Storage
Real Estate, Ownership of
Dwellings and Business
Cold Storage is covered under ASI. The Services
same procedure as used for manufactur-
Real Estate and Business services
ing (registered) may be adopted.
Districtwise estimates may be prepared
Storage not elsewhere classified
using the same procedure as for non-
Districtwise estimates may be prepared mechanised road transport sector.
using the same procedure as for the non-
Ownership of Dwellings
mechanised road transport sector.
State level estimates may be allocated to
districts in proportion to number of

Non-Commodity Producing Sectors 15

 dwellings (Rural/Urban) in each district be collected and may be used to allocate
based on census data multiplied by state totals to districts.
districtwise rent per dwelling worked
from the 55th round of NSS by pooling Unrecognised Institutions
the results of central and state samples.
Till such time districtwise rent per dwell- For this activity the district wise estimates may
ing is estimated from 55th round survey, be prepared using the procedure used for
rent per municipal house may be ascer- private mechanised road transport. Instead
tained from the municipalities within the of public sector, the workforce for the
district. recognised institutions may be subtracted
from the total workforce for education to
Public Administration derive the workforce for the unrecognised in-
stitutions. Till such time the bench mark esti-
In the case of central government admin- mates of value added from 57th round sur-
In the case of central istration, data on central government vey are derived by pooling the central and
government employees are being collected by the Di- state samples, the SDP relating to this sector
administration, data rectorate General of Employment and may be allocated to districts in proportion to
on central Training (DGET). Efforts may be made districtwise workforce multiplied by average
government to collect similar data at the district level tuition fees. Average tuition fees may be as-
employees are being to allocate the state totals. Alternately, till certained from a few unrecognised institu-
collected by the such time these data are collected, state tions and private tutors.
Directorate General total may be allocated in proportion to
of Employment and districtwise workforce. In the case of state Medical and Health Services
Training (DGET). government employees, data on salaries
and wages at the district level may be col- Public Sector
Efforts may be made
to collect similar data lected from concerned departments to
Data on salaries and wages may be col-
at the district level to allocate the state total. For local bodies,
lected from the concerned units within
allocate the state district wise data may be collected from
the district and used to allocate State level
totals local bodies and used
estimates to districts.
Other Services
Private Sector
Education, Research and Scientific Services
Districtwise estimates may be prepared
Recognised Institutions using the same procedure as for the
mechanised road transport.
Data on salaries and wages of teaching
and non-teaching staff are being com- Legal Services
piled by the state education departments
Districtwise estimates may be prepared
for recognised institutions, universities
using the same procedure as for the non-
etc. in connection with their publication
mechanised road transport.
“Education in States” by aggregating the
data at the district level. These data should

16 ESTIMATES OF DISTRICT DOMESTIC PRODUCT

Sanitary Services facturing (unregistered), Construction,
Real Estate, Ownership of Dwellings and In the sectors where
District wise data (Rural/Urban) on sala- Business services and Trade, Hotels and estimates are worked
ries and wages may be collected from the Restaurants. out by ‘Production
local bodies and used to allocate the state Approach’, the
level estimates. In the sectors where estimates are worked current year
out by ‘Production Approach’, the cur- production may be
Rest of Other Services rent year production may be evaluated on evaluated on the
the basis of base year prices to obtain basis of base year
Districtwise estimates may be prepared the value of output. For the remaining prices to obtain the
using the same procedure as for the non- sectors, the District Income estimates at value of output. For
mechanised road transport. constant prices may be worked out using the remaining
the state level deflator of current to con-
Estimates at Constant Prices sectors, the District
stant prices. In the Mining and Quarry- Income estimates at
The methodology for working out esti- ing sector, the districtwise price differ- constant prices may
mates at constant prices will be more or entials may be introduced by multiplying be worked out using
the district wise gross value added at cur-
less the same as that for current prices in the state level deflator
the sectors of Agriculture and Animal rent prices by the ratio between the wages of current to constant
Husbandry, Forestry and Logging, of non-agricultural and rural labour for prices
the base year and for the current year.
Fishing, Mining and Quarrying, Manu-

Non-Commodity Producing Sectors 17

 Issues for Discussion

Income Originating vs. Income Frequency of Estimates

Accruing Approach
Since the main purpose of the estimates
For capturing the disparity in income and is for planning purposes and also in view
development, income accruing to the of the present availability of data for the
residents needs to be captured. Due to unorganised sectors at an interval of five
free flow of goods and services across years, it would be sufficient to compile
the borders of the districts and the non- these estimates at an interval of 5 years.
For capturing the availability of net factor income earned
disparity in income by the residents from other districts/ Purchasing Power Parity
and development, states/countries, the income accruing and HDI
income accruing to approach is not feasible. Thus, the in-
come originating approach is used for Per capita DDP in real terms expressed
the residents needs to
compilation of estimates of state in- in purchasing power parity dollars (PPP$)
be captured.
come/district income. is used for construction of HDI. PPP$
are not available for the states/districts.
The district level estimates would be Hence use of all India PPP$ is resorted
utilised for the purpose of determining to. The purchasing power of the rupee
the backwardness/development of a dis- varies from state to state, district to dis-
trict, and as consequence for allocation trict and within districts in rural and ur-
of resources by the planners and policy ban areas. Work on the estimation of
makers as well as for construction of a purchasing power of rupee was initiated
composite index HDI in connection with for the state of Maharashtra and two
the Human Development Report. It other states under the plan schemes dur-
would be necessary to make adjustments ing the 1980s but not much progress was
in these estimates in regard to significant made. If the HDI at the district level has
flows of income from and to, across the to have some meaning, then studies on
territories of typical districts which are purchasing power of the rupee need to
rich in minerals and/or forest resources taken up.
or where significant flow of daily com-
muters (particularly the districts having DDP and SDP
state capitals or concentration of indus-
The estimates of SDP for the commod-
trial units), migration of labour, cattle and
ity producing sectors are generally built
net factor income from abroad are
up by aggregating the DDP of all dis-
involved.
tricts by the state DES using data on pro-
18 ESTIMATES OF DISTRICT DOMESTIC PRODUCT
duction, inputs and prices at the district Yield Rates
level. However, for the non-commodity
producing sectors, SDP is compiled us- Yield rates for some of the agricultural
ing state level data from public sector/ products are neither available at the state/
private corporate sector and GVA per district levels in the publications on agri-
worker and workforce for the cultural production. The agricultural de-
unorganised segments of the economy. partments at the district/state levels are
For proper synchronisation, it is neces- expected to have these yield rates based
sary that the SDP estimates are aggre- on eye estimates in their records. These
gated from the DDP estimates for all the data need to be collected from their
sectors. This requires generation of req- records.
uisite districtwise data for the non-com-
modity producing sectors at least to be- Forest Boundaries
gin with, on a quinquennial basis. State
The forest circles/divisions/sub-divi-
DES have to evolve a suitable sampling
sions are generally not coterminous with The price data for all
strategy for the state sample so that
district boundaries. The records of the products, which are
districtwise estimates could be generated
forest /revenue departments at the divi- important for a
with adequate precision from the follow
sion/sub-division/forest range/forest particular district,
up surveys of economic censuses.
coupe levels will have to be looked into, need to be collected
Prices so as to demarcate forest areas within a and compiled on a
district. For example in the case of regular basis for use
Producer prices prevailing in the primary Madhya Pradesh, the district boundaries in district income
markets/farm gates for agricultural prod- are now coterminous with one or more estimates.
ucts/collection centres for forestry prod- forest divisions.
ucts and at the landing centres for the
fishery products at the district level are Inputs in Forestry and Fishing
being collected through market surveys.
Data on inputs for forestry and logging
However these data are not available for
and fishing are not available even at the
many of the products. For example in
national/state level. Ratios developed in
case of Karnataka, state level prices are
the past are being used. Suitable studies
used to value the district production for
in this regard need to be initiated.
animal husbandry, forestry and fishing.
Manufacturing - Registered
The price data for all products, which are
important for a particular district, need The sampling strategy in operation from
to be collected and compiled on a regu- ASI 1998-99 provides serviceable esti-
lar basis for use in district income esti- mates at the all-India and state levels. The
mates. In this regard, it would be neces- states of Assam, Orissa, Rajasthan, Uttar
sary for the state income estimation units Pradesh and West Bengal, which are par-
to first draw up specific lists of such ticipating in the programme of ASI
products for each of the districts before should take steps to augment the size of
initiation of collection of prices. the sample for the sample (non-census)

Issues for Discussion 19

 sector, so that by pooling the data from Karnataka and DES Uttar Pradesh) has
the two samples districtwise estimates are been used to allocate state totals to dis-
generated. For this purpose, the Indus- tricts in respect of firewood, unregistered
trial Statistics Wing of the CSO and the manufacturing, construction, road trans-
FOD of the NSSO should extend full port, real estate and business services and
cooperation to these states in the selec- other services except education, medical
tion of the sample, collection of field data and health and sanitation. Thus the DDP
and processing of data of pooled of Karnataka and the HDI for the dis-
samples. Other states, that are not pres- tricts constructed on the basis of the
ently participating in the ASI programme same, ignore the disparity at the district
and are interested in generating level and do not reflect the extent of
districtwise estimates for the registered development/under-development of the
manufacturing sector to meet the needs districts of Karnataka.
of state HDR, may take steps to create
The allocation of state resources for augmenting the size of the The allocation of state domestic prod-
domestic product for central sample for the sample sector of uct for a particular sector in proportion
a particular sector in ASI. to workforce has the implicit assumption
proportion to that the value added per worker is the
workforce has the Unorganised Segments of same for all the districts. Such a proce-
implicit assumption Economic Activity dure makes the rich districts less rich and
that the value added the poor districts less poor. Thus the dis-
per worker is the Pooling of central and state samples is a parity in income gets reduced and does
same for all the must if the states have to produce not reflect the proper picture. As such,
districts. districtwise estimates of value added for its use, to the extent possible, may be
the unorganised segments of different avoided.
sectors of non-agricultural economy. The
state DES and the Data Processing Divi- Value Added per Worker
sion (DPD) of NSSO need to develop
and maintain an excellent working rela- As in the case of SDP, for compilation
tionship. This would help the DPD and of DDP for a particular year, the district
the state DES to devise efficient identi- level value added per worker, derived
cal data processing procedures for the from the pooled results of central and
central and state samples so that pooling state samples of the bench mark surveys
of the two samples can be undertaken of NSSO for manufacturing – unregis-
on a scientific basis. tered (56th round), unregistered trade
(53rd round), informal non-agricultural
Workforce enterprises (55th round), and unorganised
segments of various sectors (57th round),
In the case of Karnataka, districtwise will have to be carried forward or
workforce (as suggested in the joint meth- backward using appropriate district level
odology developed for DDP by DES price indices.

20 ESTIMATES OF DISTRICT DOMESTIC PRODUCT

 Bibliography

Central Statistical Organisation (CSO), Centre for Research and Development, 1999,
1974, First Report of the Committee on “Estimation of Regional Income-Esti-
Regional Accounts, CSO, New Delhi. mate of NDP of Greater Mumbai”, pub-
lished in The Journal of Income and Wealth,
_________________________, 1976, Vol. 21, No. 1, January.
Final Report of the Committee on Regional
Accounts, CSO, New Delhi. Prabhu, K. Seeta and Suraj Kumar, 2001,
Data Base for Human Development As-
_________________________, 1988, sessment: Some Key Issues, Seminar on
Report of the Technical Group for Recom- Human Development Through Na-
mending Methodology for Estimation of tional Surveys, NSS Golden Jubilee
Income at the District and Rural/Urban Celebration, 6-7, April, National
Levels, September, CSO, New Delhi. Sample Survey Organisation, Pune.

_________________________, 1996, Viswanathan, Renuka, 2001, Preparation of

Methodology for Preparation of Estimates Human Development Report for Karnataka,
of District Domestic Product, prepared Seminar on Human Development
jointly by Directorate of Econom- Through National Surveys, NSS Golden
ics & Statistics, Karnataka and Uttar Jubilee Celebration, 6-7 April, National
Pradesh, August, CSO, New Delhi. Sample Survey Organisation, Pune.

21
ANNEXURE

Comments on Methodology Adopted by Karnataka, Maharashtra

and Tamilnadu for Compiling Distict Domestic Product

Main comments on the methodology area. For Tamil Nadu, state level esti-
adopted for compiling the estimates of mates of forestry are allocated to dis-
DDP by the DES of Karnataka, tricts in proportion to districtwise area
Maharashtra and Tamil Nadu are set out under forest. Such a procedure does
below. not take into account the variation in
Price data for all yield rates as well as prices.
products, which are Agriculture, Forestry and Fishing
important for a The agricultural departments at the dis-
Districtwise Prices trict/state levels are expected to have
particular district,
need to be collected these yield rates based on eye estimates
In the absence of district level prices DES in their records. These data need to be
and compiled on a Karnataka uses state level prices to value collected from their records. Data on for-
regular basis for use the district production for animal hus- est products also need to be collected at
in district income bandry, forestry and fishing sectors. Such district level
estimates. a procedure does not take into account
the variation in prices across districts. Eye estimates are no doubt subjective but
in the absence of yield rates based on
Price data for all products, which are im- objective criteria, these may be good
portant for a particular district, need to enough for allocation purposes.
be collected and compiled on a regular
basis for use in district income estimates. Fuelwood
In this regard it would be necessary for
the state income estimation units at the State level estimates are allocated to
state levels to first draw up specific lists districts in proportion to rural workforce
of such products for each district before for Karnataka and in proportion to
initiation of collection of prices. rural/urban population for Tamil Nadu.

Yield Rates Inputs in Forestry and Fishing

For crops for which yield rates are not Data on inputs for forestry and logging
available for Karnataka, Maharashtra and fishing even at the national/state
and Tamil Nadu, state level estimates level are not available. DES of
are allocated to districts in proportion to Karnataka, Maharashtra, Tamil

22 ESTIMATES OF DISTRICT DOMESTIC PRODUCT

Nadu as well as DES of other states use the districts constructed on the basis
ratios developed in the past even for state of the same, ignore the disparity in
level estimates. Input rates vary from income per worker at the district level.
state to state.
The allocation of state domestic prod-
Suitable studies in this regard need to be uct for a particular sector in proportion
initiated by DES in collaboration with to workforce has the implicit assumption
the Chief Conservators of Forest. that the value added per worker is the
same for all the districts. Such a proce-
Communication and Public dure makes the rich districts less rich and
Administration the poor districts less poor. Thus the dis-
parity in income gets reduced and does
The states of Karnataka, Maharashtra not reflect the proper picture. As such
and Tamil Nadu allocated state level its use to the extent possible may be
income in respect of public administra- avoided. Instead it should be used in com-
tion using districtwise workforce. Tamil bination with value added per worker or
Nadu and Maharashtra also used the appropriate wage rate. Districtwise rent per
workforce to allocate state level income dwelling from NSSO
from communication to various districts. Ownership of Dwellings consumption surveys
Such a procedure ignores disparity in needs to be
income per worker at the district level. State level estimates have been allocated estimated by pooling
to districts in proportion to number of the central and state
Manufacturing – Unregistered dwellings by Karnataka, Maharashtra and samples.
and Unorganised Segments of Tamil Nadu. Such a procedure ignores
Economy variation in rent per dwelling at the
district level.
In case of Karnataka, Maharashtra
and Tamil Nadu, districtwise workforce Districtwise rent per dwelling from
has been used to allocate state totals to NSSO consumption surveys needs to be
district in respect of unregistered manu- estimated by pooling the central and state
facturing, construction, road transport, samples. Alternatively, rent per munici-
real estate and business services, and pal house can be collected from the mu-
other services except education, medical nicipalities.
and health and sanitation. Workforce has
also been used as an indicator by Districtwise number of dwellings multiplied
Maharashtra and Tamil Nadu for by rent per dwelling or rent per municipal
trade, hotels and restaurants. Thus, the house would be a better indicator instead of
DDP of these states and the HDI for only number of dwellings.

23