How useful is a regional SAM in evaluating regional projects in Sri Lanka? An illustration for post-war regional development policy analysis.
Wijerathna, Deeptha ; Bandara, Jayatilleke S. ; Karunagoda, Kamal 等
1. INTRODUCTION
Following the end of nearly three decades of separatist war in May
2009, Sri Lanka is attempting to re-establish the livelihoods of war
torn Northern and Eastern provinces through large scale programs of
resettlement, rehabilitation, reconstruction, reintegration,
reconciliation (known as five 'R's) and investment in new
infrastructure projects. The economic development process within the
country is also being accelerated in response to the delays encountered
from this period of armed conflict. Regional inequality, rural poverty
and exclusion from economic benefits of globalization have long been
recognized as important root causes of political violence in Sri Lanka,
with the unequal regional distribution of economic opportunities in the
country well documented (see for example, Central Bank of Sri Lanka,
2010; UNDP, 1998; Karunanayake, 2001; Dangalle, 2005; Wanasinghe, 2001;
Uduporuwa, 2007). According to estimates from the Central Bank of Sri
Lanka (2011), 44.4 percent of the country's Gross Domestic Product
(GDP) is produced within the urbanized Western province despite it only
housing 28.4 percent of the population. In contrast to these figures,
the other eight provinces which house 71.6 percent of the total
population, produce only 55.6 percent of national GDP. Given this
unequal geographical distribution of economic opportunities, it is not
surprising to see that poverty incidence also varies from region to
region. For example, whilst the poverty head count in urbanized Colombo
is less than 4 percent, the count of rural Batticaloa in the East is
over 20 percent (Department of Census and Statistics, 2010). Even though
the country has implemented a number of rural development and poverty
reduction projects during last six decades of its post-independence era,
9 percent of its population is still living below the absolute national
poverty line. This has led the current Sri Lankan government to launch
large sized infrastructure projects to develop the war torn Northern and
Eastern provinces and to reduce regional inequality and poverty; created
in part by the nearly three decades of conflict between the Sri Lankan
security forces and the separatist group of the Liberation Tigers of
Tamil Eelam (LTTE; known as the Tamil Tigers). In order to achieve the
long-term goals of growth, peace and a regional harmony in post war Sri
Lanka, the current government has recognized the need for strategies
which equitably share the benefits of development projects amongst the
countries different geographical regions.
In order to evaluate these post-war regional development projects
in Sri Lanka, there is a need to undertake an economic impact evaluation
by use of certain economic tools. For the purpose of assessing regional
policies, there are a number of economic techniques which are widely
cited as applicable. These include: cost-benefit (C-B) analysis,
regional input-output (I-O) and social accounting matrix (SAM) models,
regional computable general equilibrium (CGE) models, and econometric
models. However, because of data problems and the costs associated with
creating regional CGE and econometric models in many developing
countries regional policy analysts often use the I-O and SAM models
(Partridge and Rickman, 2010). Although these regional models have been
used to analyse regional and village level issues in many developing
countries (Hartono and Resosudarmo, 2008; Keuning, 1997; Pradhan et al.,
2006; Round, 2003b), few regional level quantitative analyses have been
undertaken for Sri Lanka mainly due to lack of appropriate economic
models, regional economic data and the shortage of regional policy
analysts (Wijerathna and Karunagoda, 2007). In spite of this regional
policy analysis is still paramount in the post-war development
strategies of Sri Lanka.
Based on this evident gap in literature, the main objectives of
this paper are to develop a SAM-based model for quantitative analysis of
a regional development project in Southern Sri Lanka, specifically the
Udawalawe irrigation development project, and to demonstrate how a
SAM-based model can successfully be used in evaluating the impact of
post-war development projects in Sri Lanka, especially in the Northern
and Eastern regions of the country. The SAM developed in this study is
based on a data framework for analysing the structure of a regional
economy. Policy experiments were also conducted to analyse the possible
impacts of potential pro-poor development policies.
The rest of the paper is structured as follows. The next section
will present a brief technical description of SAM followed by an
illustration of the procedure of compiling a SAM for the Udawalawe left
bank region (a small region in the Southern province of Sri Lanka).
Section three will then evaluate the impact of an irrigation project on
a rural economy and discuss the results of the policy experiments. The
final section will summarise the paper and provide any concluding
remarks.
2. DEVELOPMENT OF A SAM FOR THE REGIONAL ECONOMY OF UDAWALAWE IN
SRI LANKA
A social accounting matrix (SAM) is a comprehensive economy-wide
data framework that represents the circular flow of income and
expenditure in the economy of a nation or region, within a given time
(Lofgren and El-Said, 1999; Keuning and De Ruijter 1988). This matrix
was first developed by Richard Stone (in association with Brown) in the
1960s for the Cambridge Growth Project and has subsequently been used by
a number of authors in various studies (see for example, Pyatt and Roe,
1977). Initially SAMs were used only for national level accounting
purposes, but later a demand arose for applying them in regional and
local level activities (Kinlen, 2003; Thorbecke, 2000). They can also be
used in descriptive or prescriptive analysis of an economy (Fannin,
2001). A SAM differs from other economic models since it has explicitly
been designed to depict how income is both generated and distributed
within an economy. A SAM is an important tool for policy analysts to
measure impacts, make predictions and also to examine the link between
social and economic development (Pyatt and Round, 1977). Given the large
body of literature on construction of SAMs, their usefulness and their
applications (see for details: Round, 2003a; Thorbecke, 2000; Sadoulet
and de Janvry, 1995; Subramanian and Sadoulet, 1990), this paper does
not intend to repeat what has already been documented in the literature.
Rather, the ensuing paragraphs delineate a brief history of Sri Lankan
SAMs and the construction of a SAM for a regional economy.
The history of developing SAMs for the Sri Lankan economy dates
back to the 1970s. The SAM constructed by Pyatt and Roe, (1977) for Sri
Lanka was among the first few SAMs constructed for developing countries
and the first well-documented SAM in the world. Following this, however,
a number of attempts to develop new SAMs for Sri Lanka failed (see
Bandara and Kalegama, 2008 for a detailed history of SAMs in Sri Lanka)
until a new SAM was eventually constructed by Naranpanawa and Bandara
(2006). Although a number of policy analyses based on the 1970 Sri
Lankan SAM were undertaken in the late 1970s and the early 1980s (see
for example, Pyatt and Round, 1979; de Melo, 1982), there have been very
few new attempts to undertake SAM-based impact analyses in Sri Lanka. To
our knowledge, there is no single regional or village SAM for Sri Lanka
and this study is the first attempt of such type. This paper will
therefore assist other policy analysts to develop SAMs, which can
similarly be applied to the war-torn Northern and Eastern provinces.
This first attempt has two basic components. First, it develops a SAM
for a selected small region using a set of household data gathered from
a household survey, and secondly it analyses the potential impacts of
selected pro-poor development policies within a SAM based model.
The development phase of a SAM normally starts with the
identification of all economic agents and transactions in a selected
economy (Kiyoshi 2004). Basically, there are two different approaches
for constructing SAMs which are used by the practitioners; (1) top-down
approach and (2) bottom-up approach. In the top down approach, the
process begins with a highly aggregated SAM constructed with already
estimated summary accounts for a country or a region. In contrast, the
bottom up approach starts with disaggregated data sets compiled by
primary data sources such as household surveys and rapid appraisals.
National level SAMs usually follow the top down approach and are mostly
based on national account statistics (Sadoulet and de Janvry, 1995).
Round (2003a and 2003b), supports the top-down approach for a national
SAM when it is impossible to define detailed data needs for compiling a
SAM given the degree of country specificity. He further emphasises that
national accounts of a country should be the starting point of a SAM.
Some other practitioners like Jabara et al. (1992), Taniguchi (2003),
Shiferaw and Holden (2000), Alarcon, J. (2005), Haggblade et al. (1991)
and Thorbecke (2000) have alternatively discussed the importance of
bottom-up approach and use of household data in compiling sub-national
level SAMs. SAMs with the bottom-up approach are advantageous in
sub-national level analysis for two reasons. Firstly, instead of
starting with some conventions on income and expenditure shares based on
national accounts, the construction of a SAM begins with actual data
collected at the ground level and calculation of income and expenditure
shares with collected data. Secondly, they are very flexible since the
analyst can decide the level of disaggregation and accounts according to
objectives and analytical requirements (Adelman et al., 1988; Lewis and
Thorbecke (1992); Round, 2003a; Stats, 2005; Taniguchi, 2003).
Given its role in assisting the regional development goals of the
country, the Udawalwe irrigation project area of the Southern Province
in Sri Lanka was selected for this study. It is one of the major
multipurpose development projects implemented by the government of Sri
Lanka after independence was gained from the British in 1948. The
Udawalawe project is located at the boundary of the wet and dry Zones of
Sri Lanka, around 200 km Southeast from Colombo (see Figure 1). The
government of Sri Lanka initiated this project in 1969, with a plan to
develop 23,000 hectares of low productive and unused lands in the
Southern dry zone into irrigated agriculture (ADB 1969; Nijman 1991).
The improvement of food security in the country, reduction of population
pressure in the wet zone by shifting people to irrigated settlement,
addition of some hydro power to the national grid and reduction of long
lasting rural poverty were the key objectives of this multi-purpose
development project (Wijerathna 2009; Wijerathna and Jayakody 2007,
NIPPON KOEI, 1996, 2005).
[FIGURE 1 OMITTED]
Even though the construction of this planned reservoir was
completed in 1969, the development downstream was carried out in
different phases mainly due to capital constraints. The downstream area
identified for development basically had two divisions which were the
left and right bank. The left and right bank main canals (LBMC and RBMC)
starting from the reservoir and flowing along the ridges of the area
identified for development, were designed to irrigate 12,000 hectares
into the left bank and another 11,000 hectares into the right bank. Both
left and right bank areas were again divided into smaller geographical
areas called blocks, to assist with planning and implementing
development activities. The right bank was given the first priority in
development agenda mainly because of the higher population density in
the area (compared to the left bank). Construction of the left bank main
canal was completed up to Kiriibbanwewa tank in 1969. The command area
of the left bank however, was divided into four blocks with development
scheduled to take place in five steps. The head reach or the first block
was identified for sugarcane cultivation and developed under the
Sevenagala sugar cane project in 1983. Through this project a total of
2,300 households were settled in 2,000 hectares of irrigated lands, with
another 1,200 farm households settled in 2,100 hectares of un-irrigated
lands. Households in the irrigated area were provided an allotment of
0.75 hectares of land to cultivate sugar cane and 0.25 hectares for rice
paddies. Settlers in unirrigated area, however, were provided an
allotment of 1.75 hectares of land for sugarcane cultivation under
rainfed condition. The second block, named Kirriibanwewa, was developed
for paddy cultivation. Construction of the field level canal system of
this block was completed in 1993. About 2,000 farmers were settled in
the area and two acres of irrigable lands and half an acre of highlands
for homestead were provided to each farming household. Development of
the third block, Sooriyawewa, was the next completed in 2000. Around
3000 families were settled in this area within 2,300 hectares of
irrigated lands (Hussain et al. , 2004). While the majority of the land
in this block was developed for paddy cultivation, some land was
conversely developed to grow other field crops (OFC), given the
limitations in water availability experienced by some farmers with
irrigable low lands. Developments in the fourth block, named as
Mayurapura, were planned to be implemented in two phases with the
construction of first phase initiated in 2001. By 2002, however (the
time of household survey), irrigation water was available only up to the
third block, Sooriyawewa. Despite this, some farmers settled in
non-irrigated areas of Mayurapura (the tail end of the left bank area)
and were carried on rain-fed slash and burn (chena) cultivations, and
paddies in small pockets with water received from rain-fed tanks in the
area (Wijerathna and Jayakody, 2007). The total area actually developed
by 2002 (the time of data collection) was 11,000 hectares in the RBMC
and 6,400 hectares in the LBMC.
According to records held by the Mahaweli Authority of the Sri
Lankan government, the Udawalawe left bank project area consisted of
16,567 households (farm and non-farm) and an estimated population of
74,000 during the study period. This area is largely rural, however
there are four semi-urban centres close to the study area; namely
Sooriyawewa, Hambantota, Ambalantota and Embilipitiya. Embilipitiya is
the closest main town/city to the study area, with the right bank main
canal of the reservoir passing through the city. Other infrastructure
facilities such as electricity, roads, schools and hospitals are
reasonably well developed in some parts of the project area (Hussain et
al. 2009). The majority of the first generation settler families depend
on farming for a livelihood, whilst a considerable number of the second
generation family members are instead employed in agricultural,
non-agricultural or service sectors both within and outside of the
project area. Sugar cane is the main crop in the first block or
Sevenagala area since cultivation of this crop is mandatory in lands
developed under the sugarcane project. To maintain subsistence, however,
farmers in the area also cultivate paddy. With regards to those farmers
who cultivate sugarcane, they generally have a forward contract system
with the sugar cane factory located at the vicinity of the Udawalawe
reservoir. This sugarcane factory assists farmers in acquiring inputs
such as planting materials, fertilizer and other chemicals.
Looking now at the second block, or Kirriibbanwewa area, paddy is
the main crop of the majority of farmers, followed by the cultivation of
bananas and OFCs. In the third developed block, Sooriyawewa, both paddy
and banana are dominant crops with some farmers also cultivating OFCs.
The majority of farmers in the area have also adopted modernized
techniques. That is, they are cultivating improved varieties and using a
considerable amount of chemical fertilizer and machinery. As fertilizer
or chemical products are not produced locally, farmers buy these
products from dealers in nearby towns directly or through retailers in
their area. The major share of agricultural output is traded through
retail traders and dealers in the area as primary commodities. The
annual average household income is estimated as Rs 105,000 (1117 AUD)
and about 50 percent of this is attained directly from the cultivation
of crops (Wijerathna 2009).
Data Collection
Data required for the Regional SAM developed in this paper was
gathered from various sources using a multitude of techniques. Primarily
though, it was compiled using a dataset gathered from two household
surveys carried out with a representative sample from the Udawalawe left
bank area. This sample included 712 households selected with a
multistage stratified random sampling technique. The study area,
Udawalawe left bank, was initially stratified into five strata, in order
to take into account the variation in the criteria, namely (1) the
availability of irrigation infrastructure, (2) condition of the
irrigation infrastructure, and (3) the cropping pattern. The existing
project blocks (as described above; Sevanagala, Kiri-ibbanwewa and
Sooriyawewa) were identified as three different strata. The un-irrigated
rain-fed area at the tail end of the project (which is identified for
the development by extending the irrigation and is thus called as
extension area) and the non-irrigated sugarcane cultivated area in
Sevenagala (Sevenagala rain-fed) were identified as another two strata.
In the second stage, one to two clusters representing each of the strata
were selected. A cluster was defined as a command area, under a canal in
the case of irrigated areas, and as a village or division, in case of
rain-fed areas. As a result of the above criteria/characteristics, it
was acknowledged that the clusters within a stratum may not always be
homogeneous, and as such there was the chance of variations arising, in
terms of access to water, within clusters located in the same stratum.
Given this, clusters were chosen to represent the potential differential
access to water within a stratum. For example, in rain-fed areas,
criteria such as the size of village or division, access to markets, and
period of residence of settlers, were used for the selection of
representative clusters.
In the third stage, a systematic random sampling procedure was
adopted to select the sample within the selected cluster. This
systematic random sample was drawn from a sampling frame of a complete
list of households in a cluster. The number of households within each
cluster was decided based on the total number of households in the
block, which represented about 4.5 percent of the total households in
the study area. However, factors such as the adequate representation of
the variations within the study area, adequacy of the sample for
statistical validity, and cost and time frame for the completion of
surveys were also considered in selecting an appropriate sample size.
A comprehensive and semi-structured household level questionnaire
was used for data collection. A one year period from October 2001 to
September 2002 was considered as the reference period. This period of
one year was justified as it covers two full agricultural seasons for
seasonal crops led by paddy. The first and second surveys were carried
out in June and November 2002, at the end of the major and minor seasons
respectively.
A series of interviews with farmers and other stakeholders of
private and public sector institutions was conducted to understand
different income generating activities, marketing channels, saving
patterns, credit markets, and expenses such as taxes. Secondary data on
factors such as population, production activities, businesses, taxes
etc. were also collected from published and unpublished reports of the
Mahaweli Authority, Central Bank of Sri Lanka, Department of Census and
Statistics, International Water Management Institute, and Institute of
Policy Studies.
Designing a regional SAM
A SAM for the Udawalawe Project area was constructed by extending
the typical SAM with seven categories of accounts. Figure 2 and Table 1
show the seven major accounts identified for the Macro SAM. Following
this, seven macro accounts were disaggregated into 52 micro accounts to
represent different identified economic activities and agents in the
area (see Figure 3). The study objectives and the availability of data
were considered in deciding the number of accounts in the micro SAM and
classifying them.
[FIGURE 2 OMITTED]
Land, labour and capital assets for renting were identified as the
basic factors. As it is difficult to separate return to unpaid family
labour, and family owned land (either from agricultural activities or
household level small scale enterprises), the returns to family owned
inputs from both agricultural and small scale enterprises were
calculated with the residual method and it was defined as the restricted
profit. Next, the restricted profits were assigned to five accounts
defined for the respective geographical areas that were defined as the
five strata in the sample. Paid labour was separated into two categories
which were skilled and unskilled labour. Labour employed in agriculture,
livestock and fishing was considered as unskilled labour whilst labour
employed in service sector, industry and constructions was considered as
skilled labour. Skilled labour in the service sector included a position
of employment in Government and other public services such as police,
armed forces, security companies, private companies and institutions,
transport sector, financial business, wholesale and retail trade,
telecommunication, hotels and restaurants and other personal services.
Skilled labour in the industry sector included employment in mining,
manufacturing and processing companies. As the industry sector was not
significant in this region, a separate industry account was not
classified and the factor income received from industries in the rest of
the world was added to the service sector. The final skilled labour
sector, constructions, included employment in constructing irrigation
canals, roads and houses. A separate construction sector was identified
in this SAM as the focus of the study is on the impact of project
closure on the construction sector. Land, houses and machinery for rent
were considered as a combined factor of production capital. Finally, as
shown in Figure 3, there were eight accounts of factors of Production.
These were: (1) land and family labour Sevenagala irrigated area, (2)
land and family labour Sevenagala rain-fed area, (3) land and family
labour Kiriibanwewa, (4) land and family labour Sooriyawewa, (5) land
and family labour extension area, (6) skilled workers, (7) unskilled
workers, and (8) capital (assets for rent).
Twelve production activities identified in the region were divided
into two basic categories which were agricultural and non-agricultural
activities. As shown in Figure 3, the agricultural activities included
in the first six production accounts were: (1) paddy, (2) sugarcane, (3)
banana (4) other field crops like onion, chilli, vegetables, pulses,
other grains (e.g. kurakkan), oil crops (e.g. gingelly), (5) livestock
keeping, (6) fishing. The non-agricultural activities included in the
rest of production accounts were; (7) enterprises, (8) processing, (9)
trade, (10) construction, (11) service sector, and (12) renting
machinery.
As also demonstrated in Figure 3, the commodity account was
disaggregated into 19 sub accounts to represent the major household
consumption and input categories in domestic production. Six accounts
were created to record food consumption of households. These included:
(1) staple food rice, (2) other cereals, (3) wheat flour, (4)
vegetables, (5) meat, fish, milk and eggs, and (6) other foods.
Transactions on fuel for lighting, cooking as well as operating vehicles
and machinery at households were recorded in the energy account (7
energy: electricity, gas, fire wood and petroleum). Services consumed by
households were recorded in the services account (8). Other day to day
consumer items such as soap were recorded in the other consumer items
(9). Given that households spent considerable amounts on recreation and
ceremonies, these expenses were recorded in a separate account titled
recreation and ceremony (10). The account for manufactured items (11)
includes expenditure on items manufactured by local industries and
enterprises. The output of crops and enterprises were recorded in
another four accounts (12; paddy, 13; sugarcane, 14; banana, 15; OFC:
Other Field Crops) and services produced by the construction workers
were included in the construction account (16). Seed and cash inputs in
agricultural production were recorded in the accounts of seed (17) and
cash inputs (18). Seed accounts referred to the cost of family owned, or
purchased planting materials, and the cash input included the expenses
on fertilizer and other agrochemicals. All other expenditure items,
which could not be included in any of the above categories, were
included as other expenses (19). The other expenses mainly included the
expenses on fixed and intermediate assets.
Institutions were divided into two groups which were households and
government. Since this analysis does not focus on activities of firms,
as the rural area of Udawalawe has only a few firms which are also
stakes of large scale firms from rest of the country, firms were not
separated into another account. Some other authors such as Adlemen et
al. (1988) and Shiferaw and Holden (2000) have similarly not included
firms in their regional SAMs for comparable reasons. With regards to
households, they were disaggregated according to their per capita income
and divided into ten separate income deciles (See Figure 3). Since there
was no discernible separate local government for this smaller region and
the national government was primarily responsible for financial
allocations for all regions, the government account of this SAM
essentially represents an insignificant fraction of national government
relevant to this area. While government expenditures were estimated
using consumption data and tax rates, government expenditure was
estimated using data on transfer payments and subsidies. Government
consumption is also not estimated mainly due to the limitations of data.
Government transaction with rest of the world corresponds with the
inflow and out flow of money to the region through national government.
The capital account represents the savings, long-term borrowings
and investment in fixed and intermediate assets. The rest of the world
account includes all the transactions that occur outside the regional
economy (i.e. other regions of the country and other countries).
Following the method suggested by Logfran and El-Said (1999) and the
method used by Thobecke (1992), imports of goods and services,
remittances, and income from and expenses for employment (in the rest of
the country or other countries) were included in this 'rest of the
world' account. Further, any transactions that were not recognized
with our comprehensive database, such as operations of black market,
were also treated as operations of rest of the world.
Using the accounting framework described above and accounts shown
in Figure 3, a dummy table for a disaggregated SAM was prepared in Excel
and the sub matrices to be completed were identified. The identified
entries for the SAM were calculated in STATA by using the household
level database and some secondary information gathered from government
accounts and various other sources. Each cell entry was estimated as the
total amount of transaction to that particular account by the entire
population in the study area (region). Ratios of sample from each
stratum (from the population) were used in estimating the population
values from the sample.
Inconsistencies in this disaggregated SAM generated with primary
data (as result of the bottom-up approach) were reconciled by using
secondary data for the country and the region. The final reconciliation
or the balancing of the SAM was carried out by the cross entropy method
(Robinson et al., 2001). A mathematical programme prepared in computer
software GAMS, following a method developed by Robinson and El-Said
(2000), was used for balancing of the SAM by equating all raw and column
totals. A macro SAM for the region was produced by aggregating accounts
in micro SAM (Table 2). It provides a comprehensive picture of the
regional economy including the magnitude of the transactions between the
main sectors of the economy.
While the macro SAM is suitable to provide an overall picture of
the regional economy, the detailed micro SAM can be used for further
elaboration. For example, the macro-SAM can be used in calculating the
regional GDP and understanding the sectoral contributions to the GDP. As
shown in Figure 4, the direct contribution of the agriculture sector to
the regional GDP through crop production was 53 percent. Livestock, and
inland fisheries contributed to 1 and 0.05 percent of GDP respectively.
Rental income and agro-processing at farm gate level contributed to
another 3 percent of GDP and the construction sector induced by the
development project contributed to 33 percent of GDP. As previously
discussed, the construction sector mainly consists of employment income
from irrigation canal construction, road construction and private and
public building constructions. The service sector (made up of skilled
workers) contributed to 7 percent of GDP. Per capita GDP in the regional
economy is estimated at Rs 45,358.57 per annum and is only 47 percent of
the national per capita GDP for the same year.
The average annual household income in the region can be calculated
as the net return to factors of production. Based on this, it is
estimated at Rs 148,019 per average household in the area. Micro SAM
that classifies households according to the income deciles is useful in
understanding the pattern of income distribution and main income
generating activities of different income groups. As is depicted in
Figure 5, the share of income as the gross profits from agriculture is
high in richer deciles and low in poorer deciles. The share of income
from transfers and other unclassified sources is high in lower income
deciles.
[FIGURE 5 OMITTED]
Analysis of household expenditure (by using the commodities
household sub matrix) shows the consumption and expenditure pattern of
the region. An average household spent 56 percent of its income on food
items, 14 percent on staple food (e.g. rice) and 42 percent on other
foods. Other day to day consumer items accounted for seven percent of
the expenditure and other items of an average household account for 30
percent. The composition of household expenditure was also not uniform
(Figure 6) among households. Lower income groups spent 80 percent of
income on food while the richest groups spent less than 40 percent of
income on the same items. The share of food expenditure on staple food
(i.e. rice) for the lowest income group was more than 50 percent, while
that of the richest income group was contrastingly less than 30 percent.
The share of expenditure for purchasing of fixed and intermediate assets
was also high in richer groups, while it was negligible within the
poorest income groups.
[FIGURE 6 OMITTED]
The above descriptive analysis based on our regional SAM
demonstrates a number of key features in a typical regional and rural
economy in Sri Lanka. Firstly, a Sri Lankan rural regional economy is
dominated by agricultural activities. Secondly, a new construction
project induces a large flow of income to this rural economy by
generating employment opportunities in construction related activities.
Finally, a considerable proportion of total income in low income deciles
is coming from transfers.
3. IMPACT ANALYSIS USING THE REGIONAL SAM
Even though a SAM itself is not an economic model, it can easily be
transformed into an economic model by assuming that all relationships
are linear and prices are fixed (at least for short run). This system
can then be used to estimate multipliers that are useful in policy
analysis. The SAM multiplier analysis gives an indication of the
possible resultant effects of an exogenous shock on the functional
(factorial) and institutional distributions of income as well as on the
structure of output. Estimation of multipliers begins with
classification of endogenous and exogenous accounts. Accounts of
factors, activities, commodities and households are considered as
endogenous accounts while accounts of government, capital and rest of
the world are considered as exogenous. After classifying the accounts,
the regional SAM can be represented in a matrix format as seen in Table
3. Transactions within endogenous accounts are represented by matrices
[T.sub.ij]. Endogenous expenditures into exogenous accounts (leakages)
are represented by matrices [L.sub.ij]. Exogenous expenditures into
endogenous incomes are represented in matrices [X.sub.ij] while
exogenous expenditures into exogenous accounts (residuals) are
represented in matrices [Z.sub.ij].
To analyze multipliers concerning exogenous injections on
endogenous accounts, all exogenous accounts are summed up horizontally.
When average propensities of exogenous accounts to spend on endogenous
accounts are given in a separate matrix, [A.sub.ij], the first four rows
of above schematic matrix can be given with matrix algebra as follows.
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (1)
The reduced form of the above is
[Y.sub.n] = [(I - [A.sub.n]).sup.-1] X = [M.sub.a]X (2)
where [(I - [A.sub.n]).sup.-1] or [M.sub.a] is the matrix of
accounting multipliers.
Expression 2 shows how the effects of an exogenous injection on the
income of endogenous accounts ([Y.sub.n]) can be obtained by
pre-multiplying the vector of exogenous income injections(x) by the
accounting multiplier matrix ([M.sub.a]), provided that [M.sub.a]
exists.
Following the method suggested by, Sadoulet and De Janvry (1995),
multipliers calculated from the SAM are used in undertaking some policy
experiments.
If [DELTA] represents the operator "change," and [DELTA]X
as the vector of shocks or exogenous interventions, the impact of those
exogenous interventions on endogenous accounts can be given in vector
[increment of Y]. When matrix of SAM multipliers are given by [(I -
A).sup.-1], the vector of impacts is given by:
[DELTA]Y = [(I - [A.sub.n]).sup.-1] [DELTA]X (3)
Five illustrative policy experiments were carried out by simulating
them as external shocks in the above model. We follow the procedure used
by Sadoulet and Janvry (1995) to explain our experiments. The first four
of the following experiments are related to possible policy
interventions in regional development while the last one is related to
possible negative impact of the closure of the project.
Experiment 1: The impact of irrigating all non-irrigated lands in
the extension area.
Experiment 2: The impact of an increase in output and exports
through increase in productivity of agriculture.
Experiment 3: The effects of a direct income transfer to the
poorest three income deciles of households from any exogenous source to
lift them up to the level just above the poverty line (PL).
Experiment 4: The effects of a re-distribution of income from
highest three income deciles to lowest three income deciles so as to
lift the lowest three deciles just above the PL.
Experiment 5: The effects of the closure of the project.
The above policy experiments were chosen to demonstrate the
possibility of using the regional SAM in analysing some issues relevant
to development both in Udawalawe in the Southern province and war-torn
Northern and Eastern provinces. The first experiment was undertaken to
analyse the impact of irrigating all possible non-irrigated lands in an
agricultural area and to also understand the impact of the ongoing
project to irrigate all irrigable, but currently un-irrigated areas
within the study area. Total direct impact of the intervention is
estimated at the incremental value of net output of irrigated lands,
compared to that of existing non irrigated lands. The second experiment
explores the possibilities for increasing household income through an
increase in output and exports of agriculture, by means of an
improvement in productivity of agricultural sectors by 10 percent. This
analysis was carried out to see the impact of possible productivity
improvement, the variation among yield of farmers and estimated
technical efficiencies on increasing agricultural output. This shock is
performed with the assumption that (a) a 10 percent increase in
productivity will result in a 10 percent increase in output of all crops
and (b) all additional output can be exported to the rest of the world.
Based on this, we use a 10 percent increase in export demand in this
experiment as an indirect method to increase productivity since there
are no productivity variables in this SAM model. In both cases, it was
assumed that there were no supply and demand constraints or price
variations. Assumptions of this nature are commonly applied in other I-O
and SAM models.
The third experiment analyses the result of a cash transfer policy
(from outside the regional economy) to increase the income of the three
lowest income deciles up to the poverty line (using the national poverty
line figure). This idea is consistent with the Sri Lankan
government's attempts to lift low-socio economic groups out of
poverty through central government transfer programmes. In other words,
after gaining independence from the British in 1948, successive
governments which have come to power in Sri Lanka have attempted to
implement different kinds of social safety net programs to protect poor
households for both political and economic reasons. The purpose of this
third experiment policy shock was to therefore evaluate the impact of
such a safety net program in terms of cash transfers. The value of the
transfer payment was estimated by calculating the income gap between the
last three deciles and the poverty line.
The purpose of the fourth experiment was to analyse the possible
impact of boosting the poor with an income transfer programme within the
region (i.e. income redistribution within the region). Although a
transfer payment by the national government can protect the poor, it is
a burden to the government and to the national economy. As such, it may
be alternatively possible to do some income transfer within regional
economies. This policy experiment thus has the purpose of analysing the
impact of such a redistribution of income within the regional economy.
In this case, the shock was calculated using the same methodology as the
third experiment. Instead of providing cash transfer from an external
source however, the income gap between the poverty line and the last
three deciles is filled by transferring income from the three richest
(8th , 9th and 10th) deciles to the three poorest (1st, 2nd and 3rd)
deciles in the region.
The fifth and last experiment is used to understand the possible
negative shock resulting from the end of a large investment project. The
reverse of this is the positive temporary impact created by the project
on this regional economy. The analysis of employment data shows that 60
percent of semi-skilled and unskilled labourers employed in the
construction sector are likely to lose their income at the end of the
project. The possible impact of this scenario is analysed in the last
experiment by applying a shock of 60 percent decline in factor income
from the construction sector.
Table 4 provides a summary of the results of all five policy
experiments. According to the first policy experiment, the provision of
irrigation water to all non-irrigated lands in the extension area can
lead to an increase in output of paddy, banana and OFC sectors in the
region by 9.10, 13.53 and 21.58 percent, respectively. This process
increases the total gross value of production in the region by 6.90
percent and income of households by 3.88 percent. The level of output of
other sectors is also expected to increase due to multiplier effects (as
shown in Table 4). Since all the newly irrigated lands created by the
intervention are going to be in the extension area, all incremental
factor income on land and major parts of the incremental factor income
earned by labour, should be accrued to the factors and consequently to
the households within the extension area. As a result, the factor income
in the regional economy is expected to increase by 4.87 percent. As
shown in Table 4, the impact of this experiment will increase the income
of all income deciles, except for the lowest one. This is mainly because
the lowest income decile does not have land ownership. If this
incremental income is equitably distributed among households in the
extension area, existing high levels of poverty in the area (head count
ratio of 50 percent as estimated by Wijerathna (2009)) can be totally
eliminated with this intervention. The overall results of this
experiment therefore demonstrate the importance of an irrigation project
on a rural agricultural economy within a developing country.
The results of the second policy experiment suggest that the
increase in productivity of all agricultural sectors by 10 percent will
lead to an increase in output of agricultural sectors of more than 10
percent as a result of multiplier effects. Output of other sectors will
also increase as result of the increase in demand for other products.
This shock again leads to an increase in household income for all income
deciles, with the exception of the lowest decile (see Table 4). It is
also positive to note that the overall regional output increases by a
notable 7.79 percent. If the same level of productivity improvement is
achieved by all the farmers, the rich income deciles receive more
benefits in comparison with the lower income deciles because of the land
ownership of rich households. With the higher possibilities for
efficiency improvement in lands of poor income deciles, a productivity
improvement programme may be able to increase income in poor deciles at
a higher rate, so that the poor will receive more benefits with the
increase in net exports of the region resulting from productivity
increase. Therefore, programmes that support an increase in net exports
from agriculture by improving the efficiency of production are one of
the possible pro-poor policy options which can work to reduce the
existing poverty in this regional economy.
According to the results of the third experiment, income of poor
households in the last three deciles increases by 0.75, 0.49 and 0.36
percent, respectively. As a result of multiplier effects, the other
households also receive an average net increase in their income of
between 0.02 and 0.03 percent. These outside transfers thus stimulate
the whole economy and increase the output level of all production
activities and sectors as a result of an increase in demand (see Table
4). It is also worth noting that the regional output increases in this
case as well. This kind of cash transfer is only a temporary solution,
as the government or any other external body has to bear an annual
expenditure if this process is continued. Given this, the transferring
of income from the top three income deciles to the bottom three income
deciles may be a more suitable means of alleviating regional extreme
poverty and redistributing income within the region itself. As described
above, the fourth experiment is related to this type of
re-distributional income policy. The results of this redistributive
policy suggest that it is not as effective as the policy of external
income transfers. As expected the rich income groups will lose and poor
income groups will benefit from this policy. The overall regional output
also only increases by a neglible 0.02 percent as a result of an
increase in demand from lower income deciles. The output of production
activities also experiences only a moderate increase under this policy
scenario (see Table 4).
In order to finally analyse the possible income loss from the
closure of the irrigation project, the fifth policy experiment was
carried out. During the survey period the households in this regional
economy received nearly 15 percent of their total income from
construction activities stimulated by the construction project. It has
been estimated that if this project were to close, a 60 percent decline
in regional construction activities would result. On this basis, a 60
percent decline in factor income in the construction sector was
introduced. The results shown in Table 4 demonstrate that this has an
immense negative shock to the regional economy. That is, the regional
output will decline by about 21 percent as a result of the closure of
the project and all household income deciles are expected to lose
(assuming that there are no compensatory policies). These results
clearly demonstrate the important role that a new construction project
plays in developing a rural regional economy. This is the reason why we
argue that new construction projects are needed to revive war affected
Northern and Eastern provinces and provide employment for the youth who
took arms during the war.
4. CONCLUDING REMARKS
In this paper, an attempt was made to assess the appropriateness of
applying the SAM technique to analyse the structure of a regional
economy and to evaluate the impact of a large-scale project on a
regional economy in Sri Lanka. The paper demonstrates that there are a
number of advantages in developing a SAM for a rural economy in Sri
Lanka. Firstly, a rural regional SAM is useful in better understanding
the structure of the regional economy. Households in a rural
agriculture-based economy act as the producers as well as the consumers
and their economy is not necessarily similar to the national economy.
Secondly, the adoption of a bottom up approach in the construction of a
regional SAM, as opposed to following a top down approach starting from
aggregated national accounts, is important in identifying the most
relevant disaggregated accounts by considering the actual situation
depicted by household data. The disaggregation of the household sector
into deciles based on per capita income is important in understanding
the inequality in distribution and analysing pro-poor policy
interventions for convergence. The disaggregation of commodity accounts
according to grass root level commodity demand is also useful in better
understanding the structure of household expenditure. Thirdly, regional
SAM-based multiplier models are useful tools in evaluating alternative
regional development policies. Finally, the paper demonstrates that the
SAM technique can be beneficial in evaluating the effects of new
projects to be implemented in the North and East of Sri Lanka as part of
the post-war development and reconciliation process.
Before concluding this paper it is important to note that although
they are very useful in analysing the issues related to a regional
economy, there are certain limitations of SAM based models. It is
well-known that SAM models are based on a number of assumptions such as
fixed prices, linear production functions and unitary elasticity's
of demand. In other words, it is assumed that there are no supply side
constraints and all prices are exogenous. These limitations can be
eliminated by developing a regional CGE model based on SAM data
(Devarajan and Robinson 2002). Finally it is recommended that future
research undertake a Path-Analysis technique so that outcomes can be
explained more explicitly. This task was beyond the scope of the current
research.
Based on the study outcomes, this paper can be considered as the
first step in compiling some disaggregated regional SAMs for Sri Lanka
and using them in analysing sub-national level development issues.
Development of similar SAMs for economically backward regions, including
war torn areas, and building regional CGE models based on such newly
developed SAMs will be important in analysing the appropriate policy
interventions. Furthermore, until some regional level SAMs for other
provinces are developed, the SAM for the Udawalawe region should be used
as a proxy in understanding both the structure of similar agricultural
economies and in analysing the impact of similar agricultural projects.
Since both the Northern and Eastern provinces claim a vast potential for
new irrigation development and agricultural projects, the kind of SAM
developed in this study may be very useful in planning and analysing
development policies for those areas.
ACKNOWLEDGEMENTS: Authors wish to express sincere thanks to both
the management of the International Water Management Institute (IWMI),
and the first author's former supervisors for providing assistance
in carrying out the initial study that paved the way for this paper. The
assistance provided by the Japan International Cooperation Agency (JICA)
and the Japan Bank for International Cooperation (JBIC) in the original
data collection phase is also appreciated. The contribution of Prof.
Yasuyuki Sawada of the University of Tokyo in preparation of the
original version of the questionnaire is gratefully acknowledged.
Constructive comments provided by Professor Jogartnam, the Senior
Supervisor of the first author at the Post-Graduate Institute of
Agriculture (PGIA) and the support of all other staff members are
gratefully noted. Our gratitude is also extended to sample households
from the Udawalawe project area, officers of the Mahaweli Authority and
all other institutions that provided valuable secondary data.
The authors would also like to thank two anonymous reviewers and
the editor-in-charge of this journal for valuable constructive comments
and feedback
REFERENCES
ADB. (1969). Appraisal of Walawe Development Project in Ceylon.
Unpublished report, Asian development bank.
Adelman, I., Taylor, J. E., and Vogel, S. (1988). Life in a Mexican
Village: A SAM Perspective. The Journal of Development Studies, 25(1),
pp. 5-24.
Alarcon, J. (2005). Social Accounting Matrix-Based Modelling,
Extensions Wellbeing and Environment: Applications Using The SAMs For
Ecuador 1975 and Bolivia 1989. Hague: Institute of Social Studies.
Bandara, J.S. and Kalegama, S. (2008). Compiling Input-Output
Tables and Social Accounting Matrices (SAMs) for Sri Lanka: Review of
the Progress, Problems and Challenge. In S.P. Senanayake, W.
Wimalaratana and A. de Silva, (Eds) Development Perspectives: Growth and
Equity, Trade and Aid. Colombo: VijithaYapa Publications, pp.19-41.
Central Bank of Sri Lanka (2010). Annual Reports 2009, Colombo:
Central Bank.
Central Bank of Sri Lanka (2011). Annual Reports 2010, Colombo:
Central Bank.
Dangalle, N. (2005). Spatial Disparities in Development in Sri
Lanka. In M.M. Karunanayake, and A. Narman (Eds) Regional Development In
Sri Lanka: Resetting the Agenda. Sri Lanka: University of Sri
Jayawardenepura
De Melo, M.H. (1982). A Simulation of Development Strategies in an
Economy-wide Model. Economic Development and Cultural Change, 30(2), pp.
334-49.
Department of Census and Statistics (2010). Poverty Indicators
2009-10. Columbo: Department of census and Statistics.
Devarajan, S. and Robinson, S. (2002). The Impact of Computable
General Equilibrium Models on Policy. TMD discussion paper No. 98.
International Food Policy Institute.
Fannin, M. (2001). Construction of a Social Accounting Matrix for
County Fermanagh, Northern Ireland. Paper presented to Southern Regional
Science Association, USA, April 11-13th 2001 (unpublished).
Haggblade, S., Hammer, J., and Hazell, P. (1991). Modeling
Agricultural Growth Multipliers. American Journal of Agricultural
Economics, 73(2), pp. 361-374.
Hartono, D., and Resosudarmo, B. P. (2008). The Economy-Wide Impact
of Controlling Energy Consumption in Indonesia: An Analysis Using a
Social Accounting Matrix Framework. Energy Policy, 36(4), pp. 1404-1419.
Hussain, I., Hanjra, M., Thrikawala, S., Wijeratne, D., Shinkai,
N., and Sawada, Y. (2007). Impact of irrigation infrastructure
development on dynamics of incomes and poverty: Econometric evidence
using panel data from Sri Lanka. JBIC Institute Research Paper 32.
Tokyo, Japan: Japan Bank for International Cooperation (JBIC).
Hussain, I., Thrikawala, S., Rewgassa, N. E., Wijerathna, D., and
Shinkai, N. (2004). Impacts of Irrigation on Chronic and Transient
Poverty Alleviation: The Case of Uda Walawe System in Sri Lanka: IWMI
Research Paper, Colombo, Sri Lanka, International Water Management
Institute (IWMI).
Jabara, C., Lundberg, M.K., and Jallow A. S (1992). A Social
Accounting Matrix for The Gambia. (Working Paper 20). Washington, D.C.:
Cornell Food and Nutrition Policy Program, CFNPP Publications
Department.
Karunanayake, M.M. (2001). Introduction. In M.M. Karunanayake (Ed),
Poverty, Spatial Imbalance and Regional Development in Sri Lanka:
Policies and Issues. Sri Lanka: Department of Geography, University of
Sri Jayawardanepura.
Keuning S.J. and De Ruijter W. A. (1988). Guidelines to the
Construction of a Social Accounting Matrix. Review of Income and Wealth,
34(1), 71-100.
Keuning, S. J. (1997). Sesame: An Integrated Economic and Social
Accounting System. International Statistical Review, 65(1), pp. 111-121.
Kinlen L. (2003). The Development of Regional Social accounting
Matrix policy Analysis system for the border, midland and western region
of Ireland. The Border, Midland and Western (BMW) regional assembly,
June 2003.
Kiyoshi, T. (2004). Data Requirements for Village Level Social
Accounting Matrices. In J. Dixon, K. Taniguchi, H. Wattenbach, and A.
Tanyeri-Arbur (Eds) Smallholders, Globalization And Policy Analysis,
AGSF Occasional Paper 5. Rome: Agricultural Management, Marketing and
Finance Service (AGSF) Agricultural Support Systems Division, Food and
Agriculture Organization of The United Nations.
Lewis, B. D., and Thorbecke, E. (1992). District-Level Economic
Linkages in Kenya: Evidence Based on a Small Regional Social Accounting
Matrix. World Development, 20(6), 881-897.
Lofgren, H. and El-Said M. (1999). A General Equilibrium Analysis
of Alternative Scenarios for Food Subsidy Reform in Egypt. Washington
D.C.: International Food Policy Research Institute.
Naranpanawa, A. and Bandara, J.S. (2006). A Framework for Social
Accounting Matrices (SAMs) for Sri Lanka, Research Studies:
Macroeconomic Policy Series No.17. Colombo: Institute of Policy Studies
Nijman, C. (1991). Irrigation Management Processes and Conditions:
A Case Study of Sri Lanka's Walawe Irrigation Improvement Project.
Colombo, Sri Lanka: International Water Management Institute.
NIPPON KOEI (1996). Detailed design report on Walawe left bank
Irrigation Upgrading and Extension Project, Japan, Tokyo : Nippon Koei
Co. LTD.,
NIPPON KOEI (2005). Walawe left bank Irrigation Upgrading and
Extension project, Inception report on the Consultancy service, Japan,
Tokyo : Nippon Koei Co. LTD
Partridge, M. D., and Rickman, D. S. (2010). Computable General
Equilibrium (CGE) Modelling for Regional Economic Development Analysis.
Regional Studies, 44(10), pp. 1311-1328.
Pradhan, B. K., Saluja, M.R., and Singh, S. K. (2006). Social
Accounting Matrix for India: Concepts, Construction and Applications,
New Delhi: Sage Publications Pvt. Ltd.
Pyatt, G. and Roe, A. (1977). Social Accounting for Development
Planning with special reference to Sri Lanka. London: Cambridge
University Press.
Pyatt, G. and Round J.I. (1979). Accounting and Fixed-Price
Multipliers in a Social Accounting Matrix Framework. The Economic
Journal, 89, pp. 850-873.
Pyatt, G., & Round, J. I. (1977). Social Accounting Matrices
for Development Planning1. Review of Income and Wealth, 23(4), 339-364.
Robinson, S., and El-Said. (2000). GAMS Code For Estimating A
Social Accounting Matrix (SAM) Using Cross Entropy (CE) Methods. TMD
discussion Paper No. 64. Washington D.C.: Trade and Macroeconomics
Division, International Food Policy Research Institute.
Robinson, S., Cattaneo, A., and El-Said, M. (2001). Updating and
Estimating a Social Accounting Matrix Using Cross Entropy Methods.
Economic Systems Research, 13(1), pp. 47-64.
Round, J. (2003a). Constructing SAMs for Development Policy
Analysis: Lessons Learned and Challenges Ahead. Economic Systems
Research, 15(2), pp. 161-183.
Round, J. (2003b). Social Accounting Matrices and SAM-based
Multiplier Analysis. The impact of economic policies on poverty and
income distribution: Evaluation techniques and tools, 301324.
Sadoulet, E. and de Janvry, A. (1995). Quantitative Development
Policy Analysis. Baltimore: Johns Hopkins University Press.
Shiferaw, B. and Holden, S. (2000). A Social and Environmental
Accounting Analysis: An Extended Village SAM for Household-Farms in
Ethiopia. Diskusjonsnotater, no. D-11/2000. Norway: Department of
Economics and Social Sciences, Agricultural University of Norway.
Stats, S.A. (2005). Constructing a Social Accounting Matrix:
Comparisons across Eleven Countries. Statistics South Africa.
Subramanian, S. and Sadoulet, E. (1990). The transmission of
production fluctuations and technical change in a village economy: A
social accounting matrix approach. Economic Development and Cultural
Change, 39(1), pp. 131-173.
Taniguchi, K. (2003). Village Level Social Accounting Matrices:
Data Requirements. In J. Dixon, K. Taniguchi, and H. Wattenbach (Eds)
Approaches to Assessing the Impact of Globalization on African
Smallholders: Household and Village Economy Modeling, Proceedings of
Working Session, Globalization and the African Small Holder Study.
Thorbecke, E. (2000). The Use of Social Accounting Matrices in
Modelling. Proceedings of the 26th General Conference of the
International Association for Research in Income and Wealth Cracow,
Poland, 27 August to 2 September 2000.
Uduporuwa, R.M. (2007). Regional Dimensions of Development of Sri
Lanka, Sabaragamuwa University Journal, 7(1), pp. 22-36.
UNDP. (1998). Regional Dimensions of Human Development in National
Human Development Report-Sri Lanka-1998. Colombo: United Nations
Development programme.
Wanasinghe, Y. (2001). Regional Imbalances in Sri Lanka: A Critique
of Regional Development Policies and Strategies. In M. M., Karunanayake
(Ed) People, Space and Resources: Perspective on Development Issues in
Rural Sri Lanka. Sri Lanka: Department of Geography, University of Sri
Jayawardanepura.
Wijerathna, D (2005). Spatial Dimensions of Poverty within an
irrigated agricultural setting: Case of UdaWalawe Left Bank Irrigation
Development Project. Proceedings of the Second National Symposium on
Geo-Informatics Sri Lanka (GISSL), Peradeniya, 26th August 2005.
Wijerathna, D. (2009). Impact of Irrigation Investments in Regional
Development: UdaWalawe Development Project in Southern Sri Lanka. M.
Phil Thesis., Sri Lanka: Post Graduate Institute of Agriculture,
University of Peradeniya.
Wijerathna, D. and Jayakody, P. (2007). High Tank Dual Canal
System--An Innovative Approach of Water Allocation for a Water Scarce
Region, Proceedings of the South Asia Water Conference on Water Access
and Conflicts: Implications for Governance in South Asia, Chennai, 21-23
March 2007.
Wijerathna, D. and Karunagoda, K. (2007). A Social Accounting
Matrix for a Rural Agricultural Economy of Sri Lanka: An Analysis of
Sectoral Income Flows. Tropical Agricultural Research, 19, pp. 307-320.
Deeptha Wijerathna
PhD Candidate, Department of Accounting, Finance and Economics,
Nathan Campus, Griffith University, Nathan, Qld, 4111, Australia. Email:
deeptha.wijerathna@griffithuni.edu.au
Jayatilleke S. Bandara
Associate Professor, Department of Accounting, Finance and
Economics, Nathan Campus, Griffith University, Nathan, Qld, 4111,
Australia.
Kamal Karunagoda
Lecturer, Post Graduate Institute of Agriculture, University of
Peradeniya, Peradeniya, Sri Lanka/ Economist, Socio Economics and
Planning Centre, Department of Agriculture, Peradeniya, 20400, Sri
Lanka.
Table 1. SAM Accounts
1 2 3 4
Factors of Production Commodities Households
Production Activities
1 Factors of Factor Income
Production / Value Added
2 Production Production
Activities
3 Commodities Intermediate Consumption
Consumption Expenditure
on Goods
4 Households Allocation Inter
of Factor Household
Income to Transfers
HH
5 Government Indirect Import Direct Tax-
Taxes Tariff Income Tax
6 Institutions Household
--Capital Savings
Account
7 Rest of The Leakage of Import of Raw Imports Household
World (Row) Factor Materials Expenditure
Income on Imports
Total Factor Total Aggregate Total Current
Income Cost Of Supply Outlays Of
Production Households
5 6 7
Government Institutions Rest of The Total
-Capital World (Row)
Account
1 Factors of Factor Total
Production Income Factor
from Row. Income
2 Production Gross Value
Activities of Output
3 Commodities Export of Aggregate
Goods & Demand
Services
4 Households Transfer Borrowings & Transfers Total
Payments Withdrawals from Row Current
to HH from Savings Receipts to
Household
5 Government Total
Current
Receipts to
Government
6 Institutions Net Capital Total
-Capital Transfers Capital
Account From Row Receipts
7 Rest of The Import of Leakages
World (Row) Capital Payments
Goods to Abroad
Total Total Current Total Total Total
Outlays Of Capital Receipts
Government Payment Of From Abroad
Households
Source: adapted from Pyatt and Roe (1977), Sadoulet and de Janvry
(1995) and Wijerathna (2009)
Table 2. Macro SAM for the Regional Economy of Udawalawe Left
Bank Region (2001-2002).
Factors Activities Commodities Households
Factors -- 2,930.9 -- --
Activities -- -- 3,529.6 --
Commodities -- 289.8 -- 2,770.5
Households 3,744.5 -- -- 17.3
Government -- 31.4 22.5 4.7
Capital -- -- -- 357.4
Rest of the 78.2 159.7 1,930.3 1,456.8
world
Total 3,822.7 3,411.9 5,482.4 4,606.7
Government Capital Rest of Total
the
world
Factors -- -- 890.5 3,821.5
Activities -- -- -- 3,529.6
Commodities -- -- 2,304.3 5,364.6
Households 93.9 419.2 333.1 4,607.9
Government -- -- 93.9 152.5
Capital -- -- 419.2 776.6
Rest of the 58.7 357.4 -- 4,041.0
world
Total 152.5 776.6 4,041.0 22,293.7
Source: the Authors
Note: 1. Reference period--October 2001 to September 2002 (Maha
2001-02 and Yala 2002); The one year period was selected so that it
covers two full agricultural seasons.
2. All values are in 2007 prices and in million rupees
Table 3. A thematic SAM for Udawalawe Left Bank region
1 2 3 4
Factors of Production commodities Households
Production Activities
1 Factors of [T.sub.1,2]
Production
2 Production [T.sub.2,3]
Activities
3 Commodities [T.sub.3,2] [T.sub.3,4]
4 Households [T.sub.4,1] [T.sub.4,4]
5 Government [L.sub.5,2] [L.sub.5,3] [L.sub.5,4]
6 Capital [L.sub.6,4]
Account
7 Rest of the [L.sub.7,1] [L.sub.7,2] [L.sub.7,3] [L.sub.7,4]
World (RoW)
Total [E.sub.1] [E.sub.2] [E.sub.3] [E.sub.4]
5 6 7
Government Capital Rest Total
Account of the
World
(RoW)
1 Factors of [X.sub.1,7] [Y.sub.1]
Production
2 Production [Y.sub.2]
Activities
3 Commodities [X.sub.3,7] [Y.sub.3]
4 Households [X.sub.4,5] [X.sub.4,6] [X.sub.4,7] [Y.sub.4]
5 Government [Z.sub.5,7] [Y.sub.5]
6 Capital [Z.sub.6,7] [Y.sub.6]
Account
7 Rest of the [Z.sub.7,5] [Z.sub.7,6] [Y.sub.7]
World (RoW)
Total [E.sub.5] [E.sub.6] [E.sub.7]
Source: adapted from Shiferaw and Holden (2000), Stats (2005) and
Wijerthna (2009)
Table 4. Impact of Possible Pro-Poor Policy Interventions and Shocks
on Production Activities Factor Income and Household Welfare
Base value (Rs.
Million - 2007 Prices)
Factors Land and family labour 282.64
Land and family labour 48.93
Land and family labour 303.83
Land and family labour 771.97
Land and family labour 34.77
Skilled workers 1,960.97
Unskilled workers 286.19
Capital Assets (for rent) 132.17
Activities Paddy 1,064.97
Sugarcane 299.58
Banana 445.96
OFC 298.50
Livestock 37.46
Fishing 1.63
Enterprises 37.81
Processing 12.55
Trade 93.80
Constructions 962.45
Service sector 218.10
Renting machinery 56.77
Households HH in 1st income decile 179.07
(HH) HH in 2nd income decile 240.61
HH in 3rd income decile 245.58
HH in 4th income decile 286.34
HH in 5th income decile 342.54
HH in 6th income decile 460.35
HH in 7th income decile 508.44
HH in 8th income decile 602.78
HH in 9th income decile 704.26
HH in 10th income decile 1,037.95
Gross Value of
Production
Factor Income
Total HH income
Exp 1 Exp 2 Exp 3
Percentage change
Factors Land and family labour 4.75% 10.42% 0.03%
Land and family labour 3.39% 9.33% 0.02%
Land and family labour 11.81% 11.22% 0.05%
Land and family labour 11.99% 11.12% 0.05%
Land and family labour 17.99% 12.06% 0.09%
Skilled workers 0.56% 0.66% 0.01%
Unskilled workers 7.04% 10.89% 0.05%
Capital Assets (for rent) 4.01% 5.25% 0.02%
Activities Paddy 9.10% 12.29% 0.06%
Sugarcane 0.18% 10.19% 0.00%
Banana 13.53% 11.01% 0.02%
OFC 21.58% 13.05% 0.10%
Livestock 3.72% 4.18% 0.12%
Fishing 3.72% 4.18% 0.12%
Enterprises 3.63% 4.08% 0.13%
Processing 3.09% 3.44% 0.22%
Trade 3.51% 3.93% 0.15%
Constructions 0.00% 0.00% 0.00%
Service sector 4.37% 5.21% 0.10%
Renting machinery 9.33% 12.22% 0.06%
Households HH in 1st income decile 0.00% -0.16% 0.75%
(HH) HH in 2nd income decile 2.16% 2.38% 0.49%
HH in 3rd income decile 2.58% 2.81% 0.36%
HH in 4th income decile 4.21% 4.65% 0.03%
HH in 5th income decile 3.74% 4.15% 0.02%
HH in 6th income decile 3.81% 4.40% 0.02%
HH in 7th income decile 4.26% 4.85% 0.03%
HH in 8th income decile 4.05% 4.65% 0.03%
HH in 9th income decile 3.99% 4.49% 0.03%
HH in 10th income decile 4.89% 5.51% 0.03%
Gross Value of 6.90% 7.79% 0.04%
Production
Factor Income 4.87% 5.47% 0.10%
Total HH income 3.88% 4.37% 0.03%
Exp 4 Exp 5
Percentage change
Factors Land and family labour 0.02% -2.98%
Land and family labour 0.01% -2.20%
Land and family labour 0.03% -4.97%
Land and family labour 0.03% -5.60%
Land and family labour 0.06% -9.08%
Skilled workers 0.01% -31.28%
Unskilled workers 0.03% -5.44%
Capital Assets (for rent) 0.01% -2.61%
Activities Paddy 0.04% -6.22%
Sugarcane 0.00% -0.09%
Banana 0.01% -2.17%
OFC 0.06% -9.65%
Livestock 0.05% -15.07%
Fishing 0.05% -15.07%
Enterprises 0.07% -14.95%
Processing 0.18% -13.22%
Trade 0.10% -14.52%
Constructions 0.00% -60.00%
Service sector 0.04% -13.62%
Renting machinery 0.03% -6.07%
Households HH in 1st income decile 0.75% -1.77%
(HH) HH in 2nd income decile 0.48% -6.83%
HH in 3rd income decile 0.35% -15.20%
HH in 4th income decile 0.01% -16.45%
HH in 5th income decile 0.01% -18.88%
HH in 6th income decile 0.01% -17.93%
HH in 7th income decile 0.01% -16.45%
HH in 8th income decile -0.13% -17.25%
HH in 9th income decile -0.15% -17.43%
HH in 10th income decile -0.11% -13.55%
Gross Value of 0.02% -21.03%
Production
Factor Income 0.01% -15.24%
Total HH income 0.02% --
Source: the Authors
Figure 3. Accounts in Disaggregated SAM
1 Factors Land and family labour Sevenagala
2 Land and family labour Sevenagala
3 Land and family labour Kiriibanwewa
4 Land and family labour Sooriyawewa
5 Land and family labour Extension
6 Skilled workers
7 Unskilled workers
8 Capital Assets (for rent)
9 Activities Paddy
10 Sugarcane
11 Banana
12 Other field crops (OFC)
13 Livestock
14 Fishing
15 Enterprises
16 Processing
17 Trade
18 Constructions
19 Service sector
20 Renting machinery
21 Commodities Rice
22 Other cereals
23 Wheat flour
24 Vegetables
25 Meat, fish, milk and eggs
26 Other foods
27 Energy
28 Services consumed
29 Other consumer items
30 Recreation and ceremony
31 Manufactured items
32 Paddy
33 Sugarcane
34 Banana
35 Other field crops
36 Constructions service
37 Seed
38 Cash inputs (Fertilizer, Pesticides etc)
39 Other
40 Households Household in 1st income decile
41 Household in 2nd income decile
42 Household in 3rd income decile
43 Household in 4th income decile
44 Household in 5th income decile
45 Household in 6th income decile
46 Household in 7th income decile
47 Household in 8th income decile
48 Household in 9th income decile
49 Household in 10th income decile
50 Government
51 Capital
52 Rest of the world
Source: the Authors
Figure 4. Sectoral Contribution to GDP.
Paddy 25%
Sugarcane 8%
Banana 12%
OFC 8%
Constructions 33%
Service Sector 7%
Renting Machinery 2%
Trade 2%
Livestock 1%
Fishing 0%
Enterprises 1%
Processing 1%
Source Data: Sub matrices 1,2 and 5,2 of the micro SAM
Note: Table made from pie chart.
