Infrastructure Planning In A Dualist Agricultural Economy:
A Search For A Policy Framework
Khalid Saeed
Social Science and Policy Studies Department
W.P.I., Worcester, MA 01609, USA
Xu Honggang
Human Settlement Planning Division
Asian Institute of Technology, Bangkok, Thailand
May 1998
Abstract
Provision of adequate infrastructure has been a major responsibility of the state in many
developing countries over the course of their development. However, various case studies
indicate that infrastructure policies have not been very effective. Not only has
infrastructure become a major bottleneck for economic development, infrastructure
policies aimed at improving income distribution have also failed to achieve their
objectives. Typically the poor subsidize the rich in the process of public infrastructure
delivery, while such subsidization is more rampant in the agricultural sector that often
dominates the developing country economies before a structural transformation has taken
place, which is often the case in the initial economic development stages.
We attempt in this study to analyze the efficacy of public provision of infrastructure in
the initial economic development stages where the agriculture is the dominant sector. A
system dynamics model of a dual economic system pervasively found in the developing
countries was originally developed by Saeed (1980) to search for fiscal and institutional
policy instruments to affect income distribution. We have extended Saeed’s model to
include decision rules affecting infrastructure provision, so the efficacy of the
infrastructure policy as a lever to improve income distribution could be evaluated.
Computer simulation is used as an experimental process to examine the impact of the
various policy options tested.
1. Introduction
Infrastructure investment is an important instrument employed by the developing
country governments over the past forty years to affect economic development
(Kreguerl992). A World Bank study examining a cross-section of developing
countries shows that infrastructure typically represents about 20 percent of their total
investment and 40 to 60 percent of their public investment (World Bank 1994).
Since the govemments in most developing countries do not have the necessary
institutions to implement fiscal policies to influence income redistribution,
infrastructure policy can be an effective tool to influence income distribution and
affect economic growth (Hirschman 1958, van de Walle and Nead 1995 Boadway
and Marchand 1995). In order to understand the potential for public provision of
infrastructure as an effective policy instrument for development, it is necessary to
understand the dynamic interaction between public infrastructure provision with the
socio-economic structure in place.
In this study, we attempt to develop a framework to analyze the efficacy of public
provision of infrastructure in the initial economic development phases where
agriculture is the dominating sector, which is currently the case in a large number of
developing countries. In fact, many scholars have considered agricultural
development to be the foundation for further economic development, while
governments have also used infrastructure investment as a major tool to achieve
agricultural development (Mellor 1967). World Bank Development Report (1994)
shows that during the early stages of the development, infrastructure resources are
primarily invested in the agricultural sector - in irrigation and transportation. It has
been observed, however, that the spread of the benefits of the infrastructure
investment is limited and large farmers receive more benefit from infrastructure
provision than small farmers in many developing countries (World Bank 1994, van
del Walle and Nead 1995). We feel that the failure of the infrastructure policy to
create a larger spread of benefits arises from the fact that the economic models
underlying policy design do not exist in reality. While the economic models assume
a homogenous socio-economic structure, in reality, the agricultural economy in the
developing countries is characterized by the co-existence of a dual economic system
consisting of a profit maximizing formal sector and a consumption maximizing
peasant sector. Any infrastructure policies implemented by the goverment will
create resource reallocation between these two sectors. Therefore, policies ignoring
the dynamic interaction between these two sectors may not perform according to
expectations.
A system dynamics model of public provision of infrastructure within a dualist
agricultural economy is developed in this study. A formal system dynamics model
offers an opportunity to experiment with the various infrastructure polices proposed
and implemented in the past and to understand their performance under controlled
conditions. Simulation experiments with the model help to understand the dynamics
of reallocation of resources between the formal and the peasant production sectors
of the economy created by the infrastructure policies. Insights into the changing
patterns of infrastructure service performance can be attained only after these
detailed dynamic processes are understood. The model helps explain the variability
in the performance of infrastructure policy experiences in the past, while also
pointing toward critical elements for a successful policy framework. This model also
helps resolve some of the debate on development policy. For example, Saeed (1980,
1994) suggests that the policy of imposing a tax on uneamed income improves
income distribution; whereas Nicholls (1971) argues that discouraging uneamed
income does less to improve income distribution than the expenditure of scarce
public resources for infrastructure improvement. Experiments with our model show,
however, that infrastructure improvement alone -even if targeted to the poor - will
not accomplish income redistribution. Only, when this instrument is combined with
the taxation of uneamed income will the income distribution improve. Policy
guidelines are outlined for an effective infrastructure based intervention. Technical
details of the model, including a machines readable listing, for replicating the
experiments discussed in this paper and for conducting further experimentation, are
available from the authors on request.
2. A system dynamics model of public provision of infrastructure in a dualist
agricultural economy
The information structure of the proposed model is adapted from a model of wage
determination and income distribution developed by Saeed (1980), which he usesfor
an experimental evaluation of development polices (Saeed 1987, 1988, 1994, 1997).
Saeed’s original model draws on neoclassical economics to construct a basic
economic growth and market clearing system, he modifies the model by relaxing its
simplifying assumptions about aggregation of sub-economies, saving and
investment behavior, and wage determination. His model subsumes the concept of
economic dualism first recognized by Boeke (1953) and developed further by Lewis
(1954), Sen (1966), Bardhan (1973) and others to represent the multiple sub-
economies coexisting in developing countries. In such a dual economic system two
sub-economies function side by side. A formal production sector operating on the
premises of profit maximization, and a peasant production sector attempting to
maximize consumption for the labor are internalized. The two sectors interact with
each other in that they bid for the resources of the economy and the surplus labor not
hired by the formal sector is accommodated in the peasant sector, while surplus
capital and land resources not employed by the formal sector can be rented out to the
peasant sector.
Our revised model incorporates additional structure representing the government
provision of economic infrastructure through taxation. Public infrastructure
provision in the model is assumed to be rationally determined by the government
depending on the infrastructure productivity, demand pressure and the financial
capacity of the goverment. The revised model tracks the decisions of the
government conceming infrastructure provision and their impact on resource
allocation and income distribution. An overview of the revised model is shown in
Figure 1, with the newly created infrastructure sector added to the original model
structure.
The original model incorporated the following behavioral assumptions governing
the roles of its actors: Both the formal and peasant production sectors of the
economy carry out production using capital, workers and land. Capital investment is
driven by profitability, which is given by the marginal revenue product of capital
and the interest rate, as well as the financial capacity. Workers can be wage earners
or sel-employed. Wage workers are hired depending on the marginal revenue
product of workers and the average wage rate. Workers unable to find employment
in the formal sector are absorbed in the peasant production sector. The average wage
rate is set not according to the average marginal revenue product of workers as
postulated in the equilibrium models of economic growth, but according to the
bargaining power of the workers which depends on the opportunity cost for a worker
to leave self-employment, given the average consumption expenditure of workers
(Sraffa 1960; Sen 1966).
Government sector
Resources allocation
gh Tax determination SB
Public infrastructure provision Si
infrastructure facilities
Formal production sector Peasant production sector
Worker Hire-lavoff process Worker
capital Rent - ownership transfer capital
Land Rent ownership transfer Land
Cash balance seamented Cash balance
Tradable goods management Market share Tradable goods management
Figure 1. An overview of the model
The infrastructure sector we have added to the model includes three sub-sectors:
allocation of resources by the government to its various service functions, including
infrastructure; the development of infrastructure facilities consisting of transportation and
irrigation services; and the product mix determination.
In the sector allocating government resource, it is assumed that government collects taxes
and makes decisions to allocate collected funds to public expenditure. The government is
the only provider of the infrastructure services in the model. Financial resources for
infrastructure building are obtained by the government through general tax collection.
Munnel (1990) points out that the price of infrastructure consumption is the tax rate.
Musgrave and Musgrave (1976) also argue that even though resources can be borrowed
from banks, the payback would still be through taxation. However, this does not exclude
building of infrastructure by the private sector. Public infrastructure provision only means
that the public sector makes the investment decision and provides the resources for
infrastructure building (Munnel 1990,Musgrave and Musgrave 1976)
The model also assumes that the resources allocated by the government to infrastructure
building must compete against resources transferred out of the rural area and those
consumed for government services delivery. The resources consumed for government
services delivery are linked to the number of infrastructure projects and the stock of
infrastructure in service as suggested by Hirschman (1967) and UNPAD (1977). The
allocation of infrastructure resources is determined by the financial capacity of the
goverment, infrastructure resources demanded to finance ongoing projects and the
economic retums on infrastructure. Fraction of resources transferred out of the rural area
is a fixed fraction based on the empirical studies conducted by Lipton (1977), Parker
(1995), Shiff and Valdes (1992), and Winters, et al. (1996). The literature on the
determination of infrastructure resource allocation is quite fragmented. The determination
of the allocation of infrastructure resources by the financial capacity of the government is
supported by the work of Dudley and Montmarquette (1992), Musgrave and Musgrave
(1976), Raj (1993). The infrastructure resources allocation to finance ongoing projects is
supported by the work of Hirschman (1967), UNDP (1977) and Mashiyaki 1996.
Allocation based on the economic returns on infrastructure is supported by the work of
Simon (1975), Frederiksen and Looney (1980); Frederiksen (1981), Munasinghe (1987),
Kikuchi and Hayami (1979) and Clements (1995).
There are two categories of infrastructure facilities in the model: irrigation and
transportation. The allocation of infrastructure resources between irrigation and
transportation is assumed to be based on their respective productivity and the demand to
finance the ongoing projects. The infrastructure subsector takes into consideration the
long supply chain for infrastructure planning, design and construction before it is
available for use. This long delay embodied in the supply chain of the infrastructure has
been observed by many researchers (Rondinelli 1977 and Hirschman 1967, Saeed and
Brooke 1996). The infrastructure project startup rates for planning are determined by the
availability of infrastructure resources and the productivity of the infrastructure.
Infrastructure project startup rates for construction are also affected by the availability of
infrastructure resources. The rates for infrastructure projects to be completed are
determined by the available infrastructure resources and the unit cost of infrastructure
facilities. The unit costs of the infrastructure facilities are endogenously determined by
the model. They are increased by resource scarcity, since the scarcity would create
bottlenecks and delays, and decreased by the economy of scale (World Bank 1994,
Hirschman 1967 and Mashiyaki 1996).
As suggested by Biehl (1986), Nadiri and Mamuneas (1994) the stock of public capital
must be adjusted by an appropriate index to demonstrate the degree of their usage by the
producers. In the model, the impacts of infrastructure on production are determined by the
infrastructure service level not by the infrastructure stock. The irrigation service level is
the accessible irrigation facility per cultivated land. The irrigation service levels in the
formal sector and in the peasant sectors can be different and controlled by irrigation
accessibility parameters. The transportation service level is the market accessibility per
unit of demand for tradable products. This structure is suggested by Liang [1981]. The
transportation service levels in the formal sector and in the peasant sectors can be
different and controlled by a transportation accessibility parameter. Literature has well
documented the different accessibility to the infrastructure service for the formal sector
and peasant sector and an the higher accessibility to the infrastructure service in the
formal sector than in the peasant sector (Ahmad and Sampath 1994, FAO 1996, Broersma
1975, Hirschman 1967). In the model a high irrigation service level brings more land into
cultivation as well as bringing a higher return on land, as suggested by Kichuchi and
Hayami (1979), Oshima (1987) and Akino (1979)). A high transportation service level
has an impact on the preference for tradable goods production and improves the
marketability of tradable goods which, in tum, impacts on the total demand and sales in
the two sectors, as pointed out by FAO (1996), Khan (1984) and Richards (1982).
The production mix sector determines the proportions of tradable and non-tradable goods.
Both the formal and peasant sectors produce tradable as well as non-tradable goods. The
non-tradable products are to meet the demand for government service consumption, rural
traditional capital, and rural traditional service consumption. Tradable goods are the
goods for infrastructure building and agricultural commodities.. Each sector maintains its
own tradable products inventory, has it own tradable goods distribution system and its
own market share. The allocation of the capacity to the production of the two types of
goods depends primarily on their respective demands, but the preference for tradable
goods rises with the availability of transportation facilities. The total sales revenue in
each sector is generated by both the sale of the tradable goods and the sale of the non-
tradable goods.
3. Model behavior and the history of infrastructure performance in developing countries
The first experiment is replicated from Saeed 1980 as a starting point for further analysis.
In his original model, the same experiment was conducted to explain how economic
duality functions in the absence of technological differentiation between the two sectors.
Saeed’s model shows that resources are concentrated in an absentee ownership mode
creating the occurrence of what has been described in the literature as feudalism, which
has been quite pervasive in the developing country agricultural economies before
development effort was undertaken. However, the experiment conducted here
incorporates the basic information linkages of tax determination and government
expenditure which is not included in the original model. The experiment with the
extended model should also replicate the occurrence of feudalism as observed in the real
world because in a feudalist society there was an administration system collecting tax and
consumed the taxed income. The implicit assumption about infrastructure in this
experiment is that its supply is completely elastic. The second experiment repeats the first
experiment with the extended model imposing limitations of infrastructure supply
depending on government financial capacity and infrastructure productivity. The behavior
in this experiment is more or less similar to the first experiment except that there is a
slight falling off in output due to the limitations of infrastructure supply. The third
experiment assumes that the access to infrastructure for the producers in the two sectors is
different, peasant sector having limited access as is often the case in reality. This
experiment shows that production in the formal sector increases while production in the
peasant decreases due to unequal access to infrastructure. This experiment also explains
the experience of Japan when it embarked upon expanding rural infrastructure over the
9% century (Akino 1979, Kikuchi and Hayami 1979). The fourth experiment shows the
model behavior with a technological differentiation created between the two sectors by
making modern capital available to the formal sector and maintaining the assumption of
differentiation in the accessibility to infrastructure to the two sectors. This is the
background in which many development policies have been implemented. Due to the
combined influence of differentiation in capital and infrastructure access, the formal
sector production gets a further impetus.
The fourth experiment is served as the base run for the policy design. It is used to
replicate the development of the dual technological and economic systems widely
experienced in developing countries after the process of economic development begins.
The simulation is generated by supplying a certain quantity of modem capital to the rural
area with the same assumptions as in experiment 3. The output elasticity of modern
capital is assumed to be higher than that of traditional capital. To apply the possible
effects of capital differentiation between the sectors, it is also assumed that the peasant
sector is unable to employ any modem capital. Capital differentiation between the sectors
develops as the capitalists sector starts meeting its additional and replacement capital
needs by acquiring a mixture of modem and traditional capital inputs. Capital demand is
met by modem capital to the extent that the fixed supply permits. The balance of the
demand is met by acquiring traditional capital (Saeed 1980). This simulation run is given
as a base run since the development process in many countries is dominated by the
penetration of modem capital even without government intervention. The amount of
modem capital can be effected by the government's development policies. All other
parameters and behavioral relationships remain unchanged. The results of the simulation
are shown in figure 2. From the simulation runs, we can observe that the dualist economic
structure is distinct. The availability of modem technology and the availability of
infrastructure services offer an opportunity to the formal sector to improve its
productivity. Marginal returns on land and capital increase in the formal sector. It is more
profitable for the formal sector to be engaged in production than to rent out the land and
capital. The employment of land and capital in production also increases the marginal
retum on workers in the formal sector. Demand for workers is high in the formal sector.
While production in the peasant sector declines, the average consumption level of the
workers reduces to a lower level than the marginal return on the workers in the formal
sector. Hence, wage rates fall and more workers are hired. Toward the end of simulation,
the total production increases greatly compared to the initial equilibrium level. The
relative size of the formal production sector expands while that of the peasant sector
reduces.
4. Infrastructure provision: A strategy to improve economic efficiency and income
distribution.
A formal system dynamics model provides an experimental tool to explore the
effectiveness of proposed and implemented polices in achieving equitable income
distribution and economic growth. The experiments in this section of the paper attempt to
understand the variability of the performances of selected development policies with the
interaction with public infrastructure provision, and also to identify guidelines for an
effective infrastructure strategy that could serve as an alternative policy for economic
development. The policies selected for the experiments are based on two considerations:
1). The performance of the system with realistic infrastructure constraints added needs to
be understood. 2). The promise of infrastructure supply as a policy lever needs to be
investigated. Three policies are selected for experimentation with the extended model to
address the first consideration. These include taxation of unearned income, improving the
working of the financial markets and organizing peasant sector into cooperatives. The
first policy was proposed as a critical instrument to redistribute income and the later two
as facilitators in Saeed (1980,1994). The second and third are facilitating policies. The
facilitators were effective only when the critical policy was in place but did not help
without it. In the absence of an explicit infrastructure supply process, Saeed’s original
model assumed in default that infrastructure supply is infinitely elastic, hence
infrastructure could not be factored in as a policy lever. In the extended model, we
provide policy space to control the magnitude, the supply delay, the mix and the targeting
of infrastructure, so these attributes could be tested for their implications for meeting
objectives of growth and equity laid out in the original model. Therefore, the next set of
experiments addressing the second consideration explore the explicit infrastructure
polices suggested by researchers or used by the infrastructure policy decision-makers. The
proposed infrastructure policies are mobilization of funds to build more infrastructure
(Hischman 1958, Hansen 1965) and targeting infrastructure supply for the poor (Besley
and Kanbur 1993). The efficacy of each tested policy is evaluated by comparing the
simulation patterns with the experiment 4 designated as base run.
4.1. Analysis of the individual policy
4.1.1.Policy 1 mobilizing resources for infrastructure building
Many policies over the past three decades have attempted to mobilize government funds
for infrastructure investment. Policy makers usually refer to these funds as subsidies to
the rural area. Based on base run at time 50, a policy is introduced to reduce the resources
transferred to the urban area and use these resources to build infrastructure in the rural
area. From the simulation nun, it can be observed that this policy leads to an increase in
output. Average wage rate rises to a higher level than the base run. Yet, the wealth
distribution and the share of revenue in the formal sector representing the inequality in
income distribution do not change. When policy is implemented, the land owned by the
formal sector still grows but at lower rate than the base run. Towards the end of the
simulation run, the share of the land owned by the formal sector reaches a new
equilibrium that is lower than the base run. The revenue share of the formal sector
reaches an equilibrium level slightly lower than the base run. Rent payments from the
peasant sector to the formal sector rise to a much higher level than the base run (see table
1). These changes are explained as follows:
When resources are mobilized to build infrastructure in the rural area, the total demand
for rural products increases while the infrastructure facilities in service increase after a
certain delay. Fueled by demand and increased infrastructure availability, the production
in the peasant sector also increases. Hence, the average consumption expenditure per
worker rises, which fuels the wage rate demanded that reduces the profitability of
production in the formal sector. The formal sector begins to lay off workers, who enter
the peasant sector. Formal sector also begins to sell out or rent out land and capital to the
peasant sector. The demand for land and capital in the peasant sector remains high
because of the crowding of workers into it and the growth in demand. However, its
financial capacity constrains its ability to acquire more land and capital. As more and
more workers are laid off and accommodated in the peasant sector, the financial capacity
of the peasant sector worsens rapidly. The demand for land in the peasant sector is met
more and more through renting land, hence, the total rent payment to the formal sector
from peasant sector rises at an increasing rate. Due to infrastructure improvement, the
income of the peasant sector lightly improves and it is able to buy more land. Therefore,
compared with the base run, land owned by the formal sector is slightly lower towards the
end of the simulation.
4.1.2.Policy run 2 equal access to the infrastructure
Although there are many technical difficulties on the design of infrastructure for the poor,
in the model it is assumed that these difficulties can be overcome so both sectors can have
equal access to infrastructure facilities. Compared with base run (see table 1), the rate for
the formal sector to acquire land slows down. The share of land owned by the formal
sector and revenue shared by the formal sector reach new equilibrium levels slightly
lower than in the base run. However, land owned by the formal sector still increases after
the implementation of the policy. The rent payment to the formal sector keeps on rising to
a level higher than the base run.
It is interesting to compare this run with the previous policy run. In previous run, when
government increases investment in infrastructure it takes a long time for infrastructure
projects are planned and constructed. Therefore, there is a long delay before the
infrastructure facilities are put into service. Production and revenue in the peasant sector
do not increase as rapidly in the former run as in this run. In this policy experiment,
production in the peasant sector rises rapidly since there is a sudden improvement of the
infrastructure service level. Hence the average consumption expenditure per worker in the
peasant sector rises more rapidly in this run than in former run. Also, the building of the
new infrastructure directly benefits the two sectors in the former simulation run,
production in the formal sector also increases. Therefore, marginal revenue return on
workers in the formal sector also increases. Demand for worker in the former sector keeps
high. The laying off of workers in the formal sector proceeds at a slower rate. Overall,
workers are laid off more slowly in the previous run than in this run. Therefore, in the
formal run, demand for rented capital and land rises slowly when at the policy is
implemented initially. Thus financial burden in the peasant sector due to rent burden and
consumption pressure of unemployment workers crowded into the peasant sector is lower
in the former run than in this simulation run. The peasant sector has financial capacity to
buy the land and capital.
4.1.3.Policy 3 Cooperative policy
The cooperative policy allows the peasant sector to compete with the formal sector for
modem capital. The availability of modem capital to the peasant sector re-enforces the
growth feedback of production in the peasant sector. Demand for land and capital for
production in the peasant sector increases. The formal sector begins to transfer land and
capital from production to renting. The rent payment from the peasant sector to the formal
sector rises. Compared with base run, the rate for the formal sector to acquire land is
slower. The share of land owned by the formal sector and the revenue share of the formal
sector reach new equilibrium levels slightly lower than the base run but still higher than
the policy starting point.
There are many similarities between policies 3 and 2. Reynolds (1975) points out that in
agricultural production, infrastructure investment is a kind of modem capital investment.
Therefore, both the cooperative policy and accessibility to infrastructure policy are
intended to improve the production conditions for the peasant sector. The cooperative
policy is a policy to be carried out through market mechanism, while the policy of
improving the accessibility to infrastructure is carried out through the distribution of
public goods. In both cases, production in the peasant sector increases and the wage rate
also increases. In both cases, the asset ownership of peasant sector and the share of
revenue of peasant sector are not improved very much and the total rent payment from
peasant sector to the formal sector is increased. However, it seems that the policy which
improves the infrastructure service access to the peasant sector is more effective in
inducing economic growth in the peasant sector and overall output than the cooperative
policy. Cooperative policy is crippled for the following reasons: (1) The peasant sector is
labor intensive. A labor saving technology is not as effective in the peasant sector as in
the formal sector. (2) Although the production potential in the peasant sector is increased
with modern capital, due to the low infrastructure service level, the production potential
in the peasant sector is not fully realized. On the other hand, since the production in the
peasant sector is low, the average wage rate is also low. Fewer workers are laid off. Less
land and capital are rented and less rent is paid to the formal sector.
Policy 1, policy 2 and policy 3 are effective in promoting economic growth through
promotion of production in the peasant sector. However, they all fail to change the asset
ownership pattern and revenue share between the formal and peasant sectors. All the
three policies lead to increased rent payments from the peasant sector to the formal sector.
When the policies are implemented to promote the production in the peasant sector, the
average consumption expenditure in the peasant sector increases. The average wage rate
rises. It becomes less profitable for the formal sector to produce. The land and capital
which are employed in production in the formal sector will be either sold out or rented
out. However, financial capacity in the peasant sector limits the peasant sector to acquire
land and capital. The financial capacity is largely determined by the unemployment rate in
the peasant sector. When the workers are laid off too rapidly, financial capacity in the
peasant sector are worsened even the overall production and revenue in the peasant sector
are improved. In the end, little land and capital can be acquired by the peasant sector. The
inefficiency of the policy run 2 and 3 are due to the rapid layoff of workers after the
implementation of the policy. In the policy experiment 1 where the layoff workers are
much slower due to long delay of infrastructure supply chain and the general
improvement of production in both sector, the peasant sector can acquire some land and
capital when the policy is implemented initially. However, as more and more workers are
gradually laid off and enter the peasant sector, the financial capacity in the peasant sector
worsens. Demand for land and capital in the peasant sector rises rapidly. The formal
sector rents more and more land and capital to the peasant sector. In each case, the
peasant sector ends up paying a higher rent to the formal sector. Therefore, in order to
change the pattern of asset distribution, renting activity should be discouraged and
financial assistance should be provided with policies to promote production in the peasant
sector.
4.1.4. Policy 4 financial policy
The financial policy requires that the constraints of cash balance on the acquisition of
land and capital be reduced. The financial policy reduces the dependence of the bidding
for land and capital on internal savings in each sector. However, the segmentation of the
financial market is still existing. Each sector has to finance the acquisition of land and
capital. We find that financial policy improves income distribution slightly. After
implementing the policy, land owned by the peasant sector shows a slight trend to rise,
and land owned by the formal sector shows a slight trend to decline. Rent payment to the
formal sector reduces. Y et, the improving rate is very slow. Overall, land ownership in
the peasant sector improves but only slightly. There are two reasons for the inefficiency
of the financial policy. First of all, the formal sector acquires land and capital mainly
because it is profitable to produce. The profitability to produce in the peasant sector is
low due to the low productivity of traditional capital and to the limitation of infrastructure
service. The desire for the peasant sector to invest on land and capital acquisition is low.
Improving the financial capacity of the peasant sector does not help the peasant sector to
gain more land. Second, the financial policy only reduces the dependence of the peasant
sector on internal savings. However, the segmentation of the financial market is still
existing and the financial capacity in the peasant sector is still weak. The ability to
acquire land and capital is limited in the peasant sector.
4.1.5.Policy 5 Rent tax policy
Experimenting with the original model, Saeed (1980, 1994) has found that a fiscal policy
to levy tax on rent income will be a critical policy in changing the income distribution
pattern. A rent income tax depresses renting activities. The formal sector has to either
transfer the rented land and capital for self-production or to sell the surplus land and
capital to the peasant sector. When there is no capital differentiation, the formal sector
tends to sell the land and capital to the peasant sector. Income distribution is improved in
terms of asset ownership and rent payment. He also shows that this policy is diluted when
modem capital differentiation is incorporated into the model. When the formal sector can
adopt highly productive capital, a profitable alternative to renting has been already
created. Further discouragement of renting activity does not help to change the income
distribution pattern.
A rent income tax is simulated in the model by deducting a tax equal to 20% of the rent
income at time 50. Rent tax policy is not effective in changing the wealth distribution
pattern (see table 1). However since this experiment is conducted with modem capital
differentiation, it is difficult to understand whether the reason for the dilution of
effectiveness is only due to modern capital differentiation or if it is also effected by the
public provision of infrastructure. Therefore, a comparative policy is conducted with the
assumption of equal access to infrastructure in both sectors. The assumption is relaxed at
the beginning of the simulation run. The other parameters stay the same as in the
experiment conducted above. In this comparative run, since the production of the peasant
sector is not especially constrained by infrastructure service, the bargaining power of the
worker is higher and wage rate is higher accordingly. Workers are laid off to the peasant
sector. The profitability for the formal sector to produce is reduced. Meanwhile, the
demand for renting land and capital in peasant sector are higher. Capital and land rent are
active in this comparative run. Therefore, the policy which reduces rent income depresses
renting activities and changes the asset distribution pattern radically. It can be concluded
that the rent tax policy is effective when renting is an important economic activity in the
economy.
But implementation of the rent tax policy also brings a side effect. The overall output is
reduced. The reduction of the output is due to two reasons: (1) Allocation inefficiency.
When the rent tax is implemented, in order to maintain the same level of allocation
efficiency, all available land should be bought by the peasant sector to be put into
production. However, due to lack of financial capacity, peasant sector can not afford to
buy the land. The land and capital, which are formerly rented out, are thus forced to be
employed in the formal sector production. This creates a low efficiency of production. (2)
Constraint on production in the peasant sector. The production in the peasant sector is
heavily constrained due to the lack of availability of modern capital and accessibility to
infrastructure, the total production can not be much improved.
5.2. Analysis of the policy package
Many infrastructure researchers and policy makers have realized that individual policies
are not effective. Policy packages are suggested. Typically, there are two kinds of policy
packages. The first category of infrastructure policy package is to mobilize government
funds for infrastructure investment implemented along with the other policies. The
second category of infrastructure policy package is to improve the accessibility of
infrastructure targeting to the poor implemented along with the other policies. However,
many policy packages have failed to recognize the dynamic change in the role of renting
activities when the policies are implemented. The effectiveness of these policies is not
satisfactory.
5.2.1 Policy package 1: Resource mobilization of the government fund and targeting of
the poor through equal access to infrastructure.
The first package is to mobilize the government funds for rural infrastructure building and
design the infrastructure facilities in such a way that equal access is obtained by the two
sectors. The simulation result is the same as policy one. Compared with base run, policy 1
and policy 2, it can be observed that the output increases at a higher level. The policy
package has little effect on the improvement of asset ownership distribution. The average
wage rate rises to a higher level. The production in the peasant sector increases and the
production in the formal sector reduces. However, the share of revenue received in the
peasant sector rises only lightly compared with the base run but stays at the same level as
the policy starting point. This infrastructure policy benefits the formal sector who
receives a higher level of rent payment.
5.2.2. Policy package 2: Implementing resource mobilization policy with financial policy
The individual policy analysis in the former sector demonstrates that financial constraint
is also one of the limiting factors for the peasant sector to buy land when the formal
sector decides to transfer land either for self-cultivation or sell it. Therefore, this policy
package suggests that when resources are mobilized for infrastructure investment, a
financial policy to improve the financial capacity of the peasant sector should be
implemented at the same time. When the policy package is implemented, revenue in the
peasant sector increases. The average wage rate also increases. The formal sector begins
to lay off workers who then enter the peasant sector. The formal sector also begins to sell
or rent out the land and capital formerly employed in production. The improved financial
capacity in the peasant sector allows it to buy land. Within a short time, production
increases because of the general improvement of infrastructure, less rent payment to the
formal sector and a higher wage rate paid by the formal sector. Therefore, revenue
received by the peasant sector increases. More workers are laid off to the peasant sector.
The financial capacity in the peasant sector worsens. The capacity for the peasant sector
to buy land is reduced. Toward the end of simulation run, the improvement of land
ownership is small.
5.2.3. Policy package 3: Implementing resource mobilization policy with the rent tax
From previous individual policy analysis, it is clear that when renting is an important
economic activity in the dualistic economy, a rent tax is an effective tool to radically
change the income distribution pattem. This policy suggest that when resources are
mobilized for infrastructure investment and the peasant sector has a chance to increase its
production, a rent tax should be implemented at the same time to enable the peasant
sector to get a larger share of benefits during the course of the development effort. This
policy is satisfactory in suppression of the renting activity and radically changes the
income distribution. After implementing the policy, not only the formerly rented land and
capital are gradually transferred to the peasant sector, the formal sector even begins to sell
the land which has been employed in the formal sector production to the peasant sector
when average wage rate rises. Y et, when the policy is implemented, the total output does
not increase as much as when the resource investment policy is implemented alone.
5.2.4. Policy package 4: Resource mobilization, targeting of the poor through equal
access to infrastructure, and implementation of rent tax policy
Implementation of policy package 1 has little effect on the improvement of asset
ownership distribution. When the package 1 is implemented, rent payment from the
peasant to the formal sector rises very rapidly. The improvement of income in the peasant
sector is very limited. When the rent tax policy is implemented with package 1, there are
radical changes in the land ownership pattern. Rent is depressed. Production is mainly
carried out in the peasant sector and the share of revenue of the peasant sector in the
economy rises.
5.2.5. Policy package 5: Implementing resource mobilization policy with financial policy
and rent income tax policy.
This policy attempts to reduce the inefficiency of resource relocation induced by the rent
tax policy; it is suggested that when the rent tax policy is implemented, a financial policy
should also be implemented simultaneously. Simulation runs show that the patterns are
the same as the policy package 3, with further improvement of the total output,
improvement of wealth distribution in the peasant sector and a reduced rent payment to
the formal sector.
5.2.6. Policy package 6: Implementing the resource mobilization policy with the financial
policy, equal access to the infrastructure, cooperative policy and rent tax policy.
When these police are implemented at the same time, simulation runs show the patterns
are the same as the policy package 5, only higher economic growth and a better income
distribution can be achieved. The rent tax policy eliminates the source for the unequal
share of income. The financial policy improves the reallocation efficiency. When the
production in the peasant sector is encouraged, a higher level of infrastructure service and
accessibility to modern technology will sustain the economic growth in the peasant sector
and the overall economy.
5.2.7. Alternative policy packages without mobilizing resources to build infrastructure
Usually mobilizing financial resources is difficult and involves a long time delay. A
policy package without mobilizing resources for infrastructure investment can also
achieve the development goal. Both cooperative policy and equal access to infrastructure
can promote production in the peasant sector. When either of the policies is implemented
with the rent tax policy, an equal share of development effort can be achieved. If the
policy can be implemented with the financial policy the inefficiency during the
transaction period will also be reduced. Three more experiments are conducted. Policy
package 7 is the combination of rent tax policy and equal access to infrastructure policy.
Policy package 8 is the combination of tax rent policy, equal access to infrastructure
policy and financial policy. Policy package 9 is the combination of rent tax policy, equal
access to infrastructure policy, financial policy and cooperative policy. The three policy
packages lead to a pattern of radical changes of asset distribution pattern and increasing
wage rates. However, the total production reduces slightly in the three simulation runs
(see table 1). Comparing policy package 7, the additional financial policy in policy
package 8 does not contribute much to the improvement of land ownership since the rent
activities are already depressed. Compared with package 7 and package 8, package 9
increases the production and revenue received in the peasant sector at the fastest speed
and reaches the highest level.
5.3. Policy conclusion
The formal model provides us a tool to explore the major policy efforts for development
in the dualistic agricultural economy. We have found that a single development policy
either targeting the economic growth or targeting income distribution is not effective to
achieve the development goal which is to improve income distribution pattern and
promote economic growth. While there are many policy instruments to promote economic
growth, the rent tax policy is a critical policy to change the income distribution pattem
especially effective when it is implemented with a policy which is targeted to improve the
production of peasant sector, such as the policy to increase infrastructure investment and
the policy to improve infrastructure accessibility to the poor.
We also make comparisons between the policy packages without and with mobilizing
resources for infrastructure building. All the policy packages which mobilize resources
for infrastructure building result in higher output and a better income distribution. When
infrastructure resources are mobilized in the rural area, both sectors benefit from the
policy. The marginal return on workers in the formal sector is higher. Demand for
workers in the formal sector stays high. Therefore, more workers remain in the formal
sector engaged in formal production. The peasant sector can steadily increase its
production and accumulate more financial capacity and gain more land. This shows that
reducing urban subsidy and earmarking these resources for infrastructure investment
helps the rural development .
5. Conclusion
The economies of the developing countries have been observed to incorporate two
equally significant production modes: a profit-maximizing formal sector and a
consumption-maximizing peasant sector. Most policies, especially the infrastructure
polices described in this paper, have not differentiated between the two production
sectors, therefore, their performance is varied. This paper has attempted to explore the
efficacy of infrastructure policy for developing countries dominated by a dualist
agricultural economy by using a system dynamics model of economic growth, income
distribution and public infrastructure provision based on an earlier model developed by
Saeed (1980 1994). The focus of the study is to show the impact of the dualist economic
structure on the efficacy of public infrastructure policies. Experimentation with this
model shows that public infrastructure provision can be a promising facilitating policy to
achieve economic growth and income distribution when it is implemented in conjunction
with other policies, especially with a rent tax. A policy package which reduces the urban
subsidy, relieves the constraints to the peasant sector through the establishment of
cooperatives, allows equal access to infrastructure, and discourages rent income can bring
the fastest growth with equality.
In the developing countries, government provision of infrastructure has been a traditional
instrument to deliver social welfare. Although the empirical evidence has provided many
counterintuitive behavior patterns, this policy can still be effective when it is
implemented with other policy instruments. Since the developing countries have limited
tools to improve their income distribution patterns, the authors argue that policies in favor
of privatization of infrastructure delivery only attempt to shift the burden to achieve
equitable income distribution with growth away from the government. Whether an
equitable share of economic benefit can be achieved through privatization is unknown.
While our analysis deals with economic duality in the agricultural sector, our conclusions
and their policy implications can be extended to other contexts of dualist economies, for
example, between the agricultural sector and industrial sector, between economically
growing areas and lagging areas, and between industrialized countries and developing
countries. The underlying structure in other forms of duality is not fundamentally
different from that elaborated in this model. Saeed and Prankprakma (1997) used a
system dynamics model to explore technological policy in the industrial sector in
developing countries. Saeed (1998) considers the global economic structure as having the
characteristics of a dual economy in which free movement of production factors and
commodities might be poised to create global feudalism. Further exploration with the
dualist structure can result in more insights.
totalproduction tax
20B product unitime unit 4
20B product unittime unit
20B product unitstime unit
325
10B product uniytime unit
10B product uniytime unit |
10B product unitstime unity 25
0 product unittime unit
0 product unittime unit 175
0 product units/time unit
0 8 3 76 01
time a
0 B ET 76 io.
PrndPeasant - BASE4 product unittime unit time
ProdFommal - BASE4 product unittime unit
TotalProd - BASE4 product unitstime units TaxRate- BASE4 dimensionless
worker allocation land distribution
6M_ persons 40M land units
6M_ persons 40M land units ———
3M_ persons L Po} as tend units
3M_ persons rN = 25M land units
0 persons 10M land units ‘a |
0 persons 10M_ land units
0 % EN 76 01 t) 35 31 76 01
time Time (YEAR)
WorkersFormal - BASE4 persons LandOwnedFormal - BASE4 land units
WorkersPeasant - BA SE4 persons LandOwnedPeasant- BASE4 and units
rent share landworkerratio
4 dimensionless 20. dimesnionless
4 dimensionless 20. dimensionless i
2 dimensionless 10 dimesnionless | M_|
2 dimensionless ———— 10. dimensionless
|
_—— a H
0 dimensionless 0 dimesnionless
0 dimensionless 0 dimensionless
t) % 31 76 i01 0 25 3 76 01
Time (YEAR) Time (YEAR)
ShareRentinRevFomnal - BASE4- dimensionless land worker ratio formal dimesnionless
ShareRentPeasant- BASE4 dimensionless and worker ratio peasant dimensionless
imigation supply chain transportation supply chain
1.1.B facility units 138 facility units
800M _ facility unit 800M facility unit
15B facility units 8B facility units
750M facility units 750M fecitiy its] 9,
500.M_ facility unit 500M _ facility unit
125B facility units 6B facility units ——
Vy
400.M_ facility units 200M _ facility units
200M. facility unit \F 200M facility unit Lp
10B facility units 4B facility units
ty 25 31 76 01 0 25 BL 76 io1
Time (YEAR) Time (YEAR)
InfraUnderPlanningfirr] - BA SE4 facility units
InfraUnderConstructionfirr] - BA SEA facility unit
InfralnUsefimr - BASE4 ————______________ facility units
Figure 2. Base run
InfraUnderPlanning{ trans] - BASE4 facility units
InfraUnderConstruction{ trans] - BASE¢——— facility unit
InfralnUse{trans] - BASE4 —————_________— facility units
Table 1 Simulation results of altemative policies
18
[rota [Production [Production [Fraction [Land [Rent [Effective [Average
[Production [Formal [Peasant = [Revere [Owned by |Burdenof [Land jage Rate
[Formal = [Formal ~—_ peasant joker
lsectior lector lsector [Ratio in
peasant.
\sector
[BASE y yl yl qf y qf yl qf y
1.09 0.74 Tes 0.94 0.94 44] 13] 14q 1q 134
|More resource for infrastructure
linvestment
1.07 0.83 154 0.98 0.94 3.34 137 1.0 q yl
lEqual access to infrastructure
- y 0.9] 1.24 0.94 0.98 1d 1q 0.98 0.94 0.98
Cooperative policy
5 0.97 0.87 12d 0.84 08d "0 0.99 10g 0.98 0.94
IRent tax policy
q 0.54 223) os] Osa 0.17 La 15] 1.04] 1.07
IRent tax policy with assumption of
lequal access to infrastructure
y 0.94 Lo] 0.99 0.94 0.94) q yl y y
IFinancial policy
1] oO 2.44) 0.93 0 857 1.64 154 12d 1.44)
Resources for infrastructure
linvestment -+equal access
1.04) 0.57 2.43 0.48 0.57 05d Lad 19g 1q 14
Resources for infrastructure
linvestment equal access +rent
Itax
Los a7 Led 0g OX 4.07 13 144 Lig 134
Resources for infrastructure
linvestment -#inancial assist
y 0.67 iq o o.6d 0.27 1g 1.67 11) 1.34
Resources for infrastructure
linvestment ‘rent tax
Lo] 06d Lod 05d 06d oid Lid 165 117 1.34
Resources for infrastructure +rent|
ltax +financial assist
1.05 Oo 3.0d 0.34 0.35 1.34 1d 187 1iq 1.37
Resources for infrastructure +rent|
ltax +financial assist +
lcooperative + equal access
0.98 0.6d Led 0.63 0.67 0.04 129 124 0.98 0.97
Rent tax +equal access
0.99 06d 18d oe 0.67 0.05 1.24 124 0.98 0.97
IRent tax + financial assist +equal
laccess
0.98 0.34 259 0.44) 0.4d 057 Las 1.24 0.94 0.97
Rent tax + financial assist +
|cooperative +equal access
References:
Ahmad, M and Sampath, R.K. 1994. Irrigation Inequalities In Pakistan 1960-1980: A
District Level Analysis. The Pakistan D evelopment Review 33:1 (spring 1994) pp.53-74
Akino, M. 1979. Land Infrastructure Improvement In Agricultural Development: The
Japanese Case, 1990-1965. Economic Development And Cultural Change 28(1): 97-117
Besley, T and Kanbur, R. 1993. Principles of Targeting. In. Including the Poor. Edited by
Michael Lipton and Jacques van der Gaag The World Bank, Washington, D.C.
Biehl, D. (ed), 1986. The Contribution Of Infrastructure To The Regional Development.
Commission of the European Community
Boadway, R and Marchard, M. 1995. The Use Of Public Expenditure For Redistributive
Purposes. Oxford Economic Papers 47:45-59.
Boeke, J.H. 1976. Economics and Economic Policy Of Dual Societies. In Leading Issues
In Economic Development, edited by G.M. Meier, Third Edition, pp.130-121. New Y ork:
Oxford University Express
Broersma, K. 1975. Planning And Design Consideration For Lower And Middle-Tier
Roads In Developing Countries, In High-Way Design In Developing Countries,
Proceedings Of PARCH Seminar E, Summer Annual meeting, University of Warwick,
UK
Clements, P. 1995 A Poverty Oriented Cost Benefit Approach To The Analysis Of
Development Projects. World Development 23(4): 577-592
Dudley, L. and Montmarquette, C. 1992 Is Public Spending Determined By Voter Choice
Or Fiscal Capacity? The Review Of Economics And Statistics LX XIV (3):522-529
FAO, 1996. Technical Background Documents. World Food Summit 1996
Frederikensen, P. C. 1981. Further Evidence On The Relationship Between Population
Density And Infrastructure: The Philippines And Electrification. Economic Development
And Cultural Change, 29(1981): 749-58.
Frederikensen, P. C. and Lonney, R. 1980. Road Investment And Regional Development:
A Philippine Case Study, Naval Postgraduate School Working Pager No. 80-3.
Hirchman, A. O. 1958. The Strategy Of Economic Development. New Haven,
Connecticut: Y ale University Press.
Hischman, A.O. 1967. Development Projects Observed Washington: The Brookings
Institution
Khan, M.L.R 1984. Rural Road Planning In Developing Countries: A Network Modeling
Approach AIT Dissertation 1984.
Kikuchi, M and Hayami, Y (ed.). 1979. Agricultural growth in Japan, Taiwan, Korea, and
the Philippines : (proceedings) Honolulu : Published for the East-West Center by the
University Press of Hawaii
Krueger, A., O. 1992. Economic Policy Reform In Developing Countries: The Kuznets
Memorial Lectures At The Economic Growth Center Oxford UK: Y ale University Press
Lewis, A. 1954. Economic Development With Unlimited Supplies of Labor. Manchester
School. May 1954; Reprinted in The Economics of Underdevelopment. Edited by
Argarwala, A. and Singh, S. Oxford University Press. 1958.
Liang, Emest, P.L 1981 Market Accessibility and agricultural development in Prewar
China Economic Development And Cultural Change 30(1): 77-105
Lipton, M. 1967. Why Poor People Stay Poor, Harvard University Press
Mashayekhi, A. N. 1996. Development Projects Cost Dynamics W orking paper, MIT
Mellor, J. M 1967. Toward a Theory of Agricultural Development. In Agricultural
Development and Economic Growth edited by Southworth, H. M and Johnson, B. Ithaca,
New Y ork: Cornell University Press
Munasinghe, M. 1987. Rural Electricity For Development Policy Analysis And
Application London: Westview Press
Munnell, A. H. 1990. with the assistance of Leah M. Cook, . How Does Public
Infrastructure Affect Regional Economic Performance? New England Economic Review
Sept./Oct. (1990): 11-32
Munnell, A. H. 1992. Policy watch Infrastructure Investment and Economic Growth
Journal of Economic Perspectives 4(6): 189-198
Musgrave, R. A., and Musgrave, P.B. 1976. Public Finance In Theory And Practice
McGraw-Hill Book Company
Nadiri, M.I. and Mamuneas, T.P 1994. The Effect Of Public Infrastructure And R&D
Capital On The Cost Structure And Performance Of US Manufacturing Industries.
Review Of Economics And Statistics. 76 (1): 22-37
Nicholls, W. H 1971. Discussion of Land Tenure Institutions. in Institutions in
Agricultural Development edited by Blase, M.G. Iowa, USA: The Iowa State University
Oshima, H. 1987. Economic Growth In Monsoon Asia : A Comparative Survey. Tokyo:
University of Tokyo Press
Parker, A. N.1995, April Decentralization: The Way Forward For Rural Development.
This Is A Research Under The Project “ Decentralization, Fiscal Systems and Rural
Development” RPO 679-68.
Raj, M. 1993. Urbanization, Infrastructure And Besieged Growth Potential Third World
Planning Review 15(2): 159-173
Reynolds, L.G. 1975. Agriculture in development Theory: An Overview. In Agriculture
In Development Theory edited by Reynolds, L.G New Haven and London: Yale
University Press.
Ricards, P. J. 1982. The Economic Context Of Rural Roads, in Ricards, P.J., M.A. EL-
Hawary, T. Y. EL-reedy, J. D. Howe, and J. H. Johns, Basic Needs And Employment:
Rural Roads And Poverty Alleviation: An Introduction And Four Country Studies:
Working paper No. Wep 2-32/Wp-41, International Labor organization, Geneva
Saeed, K and Prankprakma, P. 1997 Technological Development In A Dual Economy:
Altemative Policy Levers For Economic Development. World Development 25(5): 695-
712.
Saeed, K. 1987 A Re-Evaluation Of The Effort To Alleviate Poverty And Hunger. Socio-
Economic Planning Sciences 21(5): 291-304
Saeed, K. 1994. Development Planning and Policy Design a System Dynamics Approach.
Ashgate, England: Avebury
Saeed, K. 1998 Trade Relations In A Dualist Global Economy. System Dynamics Review
(forth coming).
Saeed, K. and Brooke, K. 1996. Contract Design For Profitability In Macro-Engineering
Projects. System Dynamics Review 12(3): 235-246
Saeed, K., 1980. Rural Development and Income Distribution: The Case of Pakistan.
Ph.D. Dissertation (Cambridge, MA: MIT 1980).
Saeed, K., 1988. Wage Determination, Income Distribution and the Design of Change
Behavioral Science 3(33): 161-186
Schiff, M and Valdes, A 1992. The Plundering Of Agriculture In The Developing
Countries Washington DC: World Bank
Sen, A. K., 1966. Peasants and Dualism With or Without Surplus Labor. Journal of
Political Economy, 74(5): 425-450
Simon, J. L. 1975. The Positive Effect Of Population Growth On Regional Savings In
Irrigation Systems Review Of Economics And Statistic 57(1975): 71-79
Sraffa, P. 1960. Production of Commodities By Means of Commodities Cambridge,
England: Cambridge University Press
United Nations Public Administration Division 1977 Development Planning as a
framework for Project Administration in Planning Development Projects edited by
Dennis A. Rondinelli. Stroudesberg, Pennsylvania: Dowden.
Van de Walle, D and Nead, K. 1995. Public Spending and the Poor: Theory and
Evidence. Baltimore and London: The Johns Hopkins University Press for the World
Bank.
Winters, P; Janvry, A; Sadoulet, E and Stamoulis, K. 1996. The Role Of Agriculture In
Economic Development: Visible And Invisible Surplus Transfers. 1996. Working Paper
No.814. California Agricultural Experiments Station. Giannini Foundation of
Agricultural Economics.
World Bank, 1994. World Development Report Infrastructure for Development Oxford
University Press.
