JRSSEM 2023, Vol. 02, No. 8, 1759 – 1778

E-ISSN: 2807 - 6311, P-ISSN: 2807 - 6494

DOI : 10.36418/jrssem.v2i08.407 https://jrssem.publikasiindonesia.id/index.php/jrssem/index

IMPACT OF TOLL ROAD DEVELOPMENT ON MODERN

RETAIL GROWTH IN INDONESIA

Arditya Prakoso Putra

Faculty of Economics and Business, University of Indonesia, Indonesia

e-mail: ardityaprakosoputra@gmail.com

*Correspondence: ardityaprakosopu[email protected]

Submitted

: February 24

2023

Revised

: March 12

2023

Accepted

: March 27

2023

Abstract: This study would like to see the impact of toll road development on retail industry growth

especially modern retail by taking the construction of the Trans Java toll road in Indonesia. Using a

difference-in-differences (DID) approach, this study analyzes the effect of the Trans Java Toll Road

in 7.656 villages crossed by the toll road from 2006 to 2018. The result shows that, on average, after

the Trans Java toll road program, the trend of increasing the number of minimarkets in villages

crossed by the Trans Java toll road is 0.53 units larger compared to The villages that don't cross by

the Trans Java toll road. But the results show that there is no significant effect of the Trans Java toll

road on the number of traditional stores, namely the grocery store as traditional retail. It can be

seen that toll road development programs tend to open up access to the growth of the modern

retail industry in Indonesia.

Keywords: Toll Road; Retail Industry; Minimarket; Traditional Stores; Difference-in-differences.

Arditya Prakoso Putra | 1760

DOI : 10.36418/jrssem.v2i08.407 https://jrssem.publikasiindonesia.id/index.php/jrssem/index

INTRODUCTION

The development of toll road

infrastructure has a positive impact on

regional economic growth, which can

simultaneously reduce production costs

and logistics costs. Toll road access

connections in an area can increase

location advantages, attract inflows from

production factors and facilitate trade

implementation to lead to economic

growth. In addition, toll roads also affect

the improvement of domestic trade

efficiency because they can reduce travel

time and logistics travel costs. The

construction of toll roads has a positive

influence on increasing household income

in the manufacturing sector and retail

sectors around the region. This makes it

easier for manufacturers to reach a wider

market at a lower cost (Zhang et al., 2020)

(Yoo & Hoon, 2016) (Shatz, Howard J;

Kitchens, Karin E; Rosenbloom, 2011).

Existing studies say that the pattern

occurs in fulfilling the provision of modern

retail supplies such as supermarkets, one of

which is due to good infrastructure access.

(Reardon & Hopkins, 2006) Infrastructure

development results in massive increases in

connectivity in various regions, especially

for the areas through which it passes which

will eventually open new centers of

economic growth and integrated special

economic zones (Feder, 2018). However,

infrastructure development can cause

negative externalities for residents around

the development area which include air

pollution and noise pollution (Glaeser,

2018). Thus, the extent of the effect of toll

road construction in encouraging the

growth of the retail industry, both modern

retail and traditional retail, is still unknown.

This study wants to see how the impact

of toll road construction on the growth of

the retail industry in the area it passes. To

be able to support the research question,

this study wants to see the effect of toll

road construction on the growth of the

retail industry. In this study, the author tries

to analyze the influence of toll roads on the

retail industry, where the transformation of

the retail industry can be represented of

them by the increase in the number of

supermarkets. In addition, this study also

wants to see the influence of toll roads on

traditional retail. The existence of the Trans

Java toll road is expected to affect the

number of grocery stalls/stores due to the

increase in access and economic activity

generated by toll road construction (Berger

& van Helvoirt, 2018).

Toll roads have a far-reaching impact

on economic growth. The construction of

toll roads significantly increases local

economic growth. (Chandra & Thompson,

2000a) (Zhang et al., 2020) Private

involvement is also needed due to limited

government funding for infrastructure

development (Macário, 2010). The

provision of new infrastructure in the field

of transportation is expected to create new

opportunities for local companies and

workers. This will produce spillover effects

by affecting trading locations, household

income, government tax revenues as well as

general economic performance in the

region. (Yoshino & Abidhadjaev, 2017)

Spillover effects have a positive and

significant influence on capital or public

capital, especially those aimed at

1761 | Impact of Toll Road Development on Modern Retail Growth in Indonesia

transportation infrastructure on regional

productivity (Condeço-melhor et al., 2014).

Investment in infrastructure stimulates

economic activity, increases company

productivity, improves economic

performance, and leads to sustainable

economic development and equal

distribution of income. The retail sector

(Zolfaghari et al., 2020) is one of the main

economic activities in every urban area,

where the retail sector creates jobs, a

source of state income through taxes, and

reflects the viability and ability of its people.

By looking at the impact of toll road

development on the retail industry as a

policy evaluation, we can see how the

process of economic equality between

regions and also the increase in people's

income which is ultimately expected to

reduce poverty levels as a result of a policy,

namely in this case investment in

infrastructure (Erkip & Ozuduru, 2015).

The supermarket revolution is a

process of expansion from traditional forms

of retail to modern forms of retail. From a

macro perspective, the acceleration of

change is influenced by the rapid increase

in people's income and also urbanization.

At first, modern retail was aimed at

consumers with high incomes and later

expanded to middle and lower-class

consumers. The first wave in the

transformation of modern retail initially

occurred in developed countries such as

European and American countries. The next

wave occurred in developing countries in

Asia and was followed by countries in Africa

(Maruyama et al., 2016) (Paper et al., 2016)

(Schipmann et al., 2020). The area traversed

by the toll road experienced an increase in

income in the manufacturing industry, retail

trade industry, and service industry. Based

on this, the author wants to see how the

impact of the construction of the Trans Java

toll road on economic growth, especially

the growth of the retail trade industry, in

this case, can be represented by an increase

in the number of minimarkets and an

increase in the number of stalls/grocery

stores in the area traversed by the toll road

(Chandra & Thompson, 2000b).

The implementation of the Trans Java

toll road construction specifically provides

an opportunity for this study to analyze the

impact of the construction of the Trans Java

toll road on the growth of the retail

industry, especially for the areas through

which it passes. But so far, research that

discusses directly the effects of toll road

construction on the growth of the retail

industry is still very limited. So this research

is very important to be able to see the

extent of the impact of toll road

development, especially the Trans Java toll

road run by the Government on the growth

of the retail industry for the area it passes.

One of the challenges to evaluate the

impact of the construction of the Trans Java

toll road on the growth of the retail industry

is that transportation infrastructure is not

the only influencing factor. The

transformation of modern retail is also

driven by increased economic growth

(Berger & van Helvoirt, 2018). To overcome

the problem of endogeneity in the

selection of locations traversed by the Trans

Java Toll Road, this study uses a difference-

in-differences (DID) approach to see how

the impact of the Trans Java Toll Road

development program in the growth of the

retail industry in Indonesia. Through the

DID approach, the villages traversed by the

Arditya Prakoso Putra | 1762

Trans Java Toll Road are analyzed as

treatment groups. While other villages that

are not traversed by the Trans Java Toll

Road are villages in one regency/city other

than villages in one sub-district with the

treatment group used as a group as a

control group.

This research uses Village Potential

data (PODES) from 2006 to 2018 issued by

the Central Statistics Agency (BPS) as well

as map data and location distribution of all

Trans Java toll road sections managed by

the Ministry of Public Works and Public

Housing, especially the Toll Road

Management Agency hereinafter referred

to as BPJT. Podes data is used to obtain data

at the village level because they want to see

the Trans Java Toll Road construction

program carried out at the village level,

where the growth of the retail industry

represented by the existence of

minimarkets and stalls/grocery stores at

the village level has a more even

distribution compared to supermarkets

which are mostly located in the city center.

With the DID approach, the effect of the

Trans Java toll road on the growth of the

retail industry in villages/ ward was

measured by comparing the number of

minimarkets and also the number of

stalls/grocery stores in the treatment group

compared to the control group before and

after the policy.

Literature

Toll road infrastructure can affect the

economy in several ways, almost all of

which are related to increased mobility.

Increased mobility can increase the speed

at which producers reach markets or inputs,

allowing them to lower costs and produce

on time (Shatz et al., 2011). Transport

infrastructure is considered social capital

and basic in economic development.

Because transportation infrastructure is the

main determinant of transaction costs to be

more efficient, economic growth is

influenced by the construction and

improving the quality of transportation

infrastructure (Zhang et al., 2020). In

addition to long-term effects, toll road

infrastructure can also increase economic

activity through direct construction

activities resulting from investment in the

construction of new toll roads (Shatz et al.,

2011).

Improving transportation

infrastructure is crucial in the global

economy, where the flow of people, goods,

and information is a key aspect of a

country's competitiveness. Infrastructure

investment has implications for people

being able to move faster, and more

efficiently and use fewer economic

resources, which increases the flow of

people and goods. Over time, this will result

in (Condeço-melhor et al., 2014) spillover

effects by affecting trading locations,

household income, government tax

revenues as well as general economic

performance in the region (Yoshino &

Abidhadjaev, 2017). Some studies show

spillover effects are defined as the positive

effects of transport infrastructure that

extends beyond the regional boundaries in

which it is located (Condeço-melhor et al.,

2014).

Improving transportation

infrastructure has an impact on the

accessibility of an area by reducing

transportation costs and increasing the

superiority of the local area. Therefore,

1763 | Impact of Toll Road Development on Modern Retail Growth in Indonesia

accessibility is an important factor for

economic development because areas that

benefit from high accessibility are usually

richer and have benefits in economic

activity in the region (Condeço-melhor et

al., 2014). Toll roads can simultaneously

reduce the cost of goods and factors in

transportation, affecting the economy on

both sides of production and consumption

(Zhang et al., 2020). The area traversed by

the toll road experienced an increase in

income in the manufacturing industry, retail

trade industry, and service industry. The

area traversed by the toll road experienced

an increase in trade-related activities such

as trucking and retail sales (Chandra &

Thompson, 2000b) (Michaels, 2008).

The construction of toll roads

significantly increases local economic

growth. The expansion of the toll road

network can reduce transportation costs

which will later result in an increase in

market potential and contribute to local

economic growth. Toll roads influence

economic growth by attracting investment

and promoting the consumption of goods.

(Zhang et al., 2020)Social infrastructure has

a major impact on the economic growth of

the industrial sector and the service sector.

Planning in optimizing the allocation of

government capital expenditure is needed

to reduce inequality in various regions,

especially those with inappropriate

inequality indexes (Zolfaghari et al., 2020).

Since the 1990s, the growth of modern

retail has become a global phenomenon

and spread throughout the developing

world. The demand for modern retail in

developing countries is further accelerated

by structural changes in infrastructure and

people's living conditions. These include

growing urbanization, private vehicle

ownership, better access to public

transportation, and women's participation

in the workforce. Urbanization and the size

of a city is an important keys to economic

growth especially the growth of the retail

industry. The growth of the retail industry

forms a retail transformation where there is

transformation from traditional or informal

retail to modern retail. Retail

transformation generally occurs in

developing countries (Paper et al., 2016)

(Dholakia et al., 2018a) (Berger & van

Helvoirt, 2018).

The supermarket revolution in

developing Asia began in the early 2000s

and continues to this day. The revolution

increased modern retail rapidly, especially

in the food sector at the expense of

traditional stores or wet markets (Reardon

et al., 2012). Supermarkets offer one-stop

shopping where middle-to-upper

consumers through private vehicles can

have a larger storage capacity. In contrast

to most traditional store conditions that

have limited land space the goods sold are

not affordable and not even visible to

buyers. This causes more traditional stores

to use counter service systems or

(Maruyama et al., 2016) services by sellers

rather than self-service/self-service or own

service by buyers (Dholakia et al., 2018b).

The rapid restructuring of the retail

market increases productivity and provides

a higher quality supply to consumers and

the modernization process is inevitable. But

on the other hand, there is opposition to

liberalization in the modern retail industry

because it causes problems for traditional

retailers who are unable to compete.

Therefore, policies from the government

Arditya Prakoso Putra | 1764

are needed that can be a meeting point of

problems between modern retail and

traditional retail. Government policies

should embrace a new conceptualization of

the retail sector where formal and informal

retailers can coexist (Berger &; Helvoirt,

2018) (Paper et al., 2016).

Modern retail is beginning to adapt by

reducing prices to cover low-income

segments of society. This is also supported

by economic growth and globalization

which causes an increase in household

income. However, modern retail does not

always offer competitive prices compared

to informal traders so the lower class with

limited income cannot access modern

retail. (Schipmann et al., 2020) Hybrid

systems of modern retail and traditional

retail are needed in food fulfillment in all

segments of society. In the absence of

modern retail, consumers face implicit

tariffs in the form of inefficient

transportation, distribution, wholesale, and

retail. Consumer transportation

capabilities, where ownership of personal

transportation is the main reason

consumers shop at stores with modern

retail formats. Spatial considerations about

increasing the size of cities and decreasing

distances to cities have a significant

positive effect on purchasing power when

they have increased access to

markets.(Berger & van Helvoirt, 2018)

(Bronnenberg et al., 2015) (Maruyama et al.,

2016) (Dolislager, 2017).

Toll Road Development Policy in Indonesia

Jagorawi toll road is the first toll road

in Indonesia which began operating in 1978

with a length of 59 km. The toll road

connects 3 (three) regencies/cities namely

Jakarta, Bogor, and Ciawi built by PT. Jasa

Marga. Then from 2004 to 2005, the

government issued changes to regulations

governing toll roads through Law No. 38 of

2004 concerning Roads and Government

Regulation No. 15 of 2005 which regulates

more specifically toll roads. Both

regulations mandate the establishment of

BPJT as a toll road regulator in Indonesia

which replaces the role of regulators that

have been in PT. Jasa Marga. With the

establishment of BPJT as the new regulator,

toll road development in Indonesia has

again entered an acceleration phase. The

government began to invite the private

sector in terms of toll road infrastructure

development through the Build, Operate,

and Transfer (BOT) mechanism. To

accelerate the process of infrastructure

development in Indonesia, the Government

issued Presidential Regulation Number 3 of

2016 concerning the Acceleration of the

Implementation of National Strategic

Projects, which was later amended through

Presidential Regulation Number 58 of 2017.

The government is accelerating several

national strategic projects, which involve

the Central Government, Regional

Governments, and Business Entities. The

construction of the Trans Java Toll Road is

included in the national strategic project

that is the focus of the Government. In

2018, the Trans Java toll road has been fully

operated with the inauguration of 3 (three)

toll roads in Central Java and 4 (four) toll

roads in East Java.

MATERIALS AND METHODS

This study uses the difference-in-

differences (DID) method to see the impact

1765 | Impact of Toll Road Development on Modern Retail Growth in Indonesia

of before and after the Trans Java toll road

policy on the villages traversed on the

number of minimarkets and the number of

stalls/grocery stores in the area. The DID

method requires a treatment group and a

control group to be able to see parallel

trends before the policy shock of a policy.

Using the DID method we can measure the

(Zhang et al., 2020)net effect of a policy by

comparing changes in average outcomes

between treatment groups and changes in

average outcomes of control groups, so

this method is widely used as a policy

evaluation tool. (Card & Krueger, 1994;

Gultom, 2019; Lucas & Mbiti, 2012).

Through the DID approach, villages

traversed by the Trans Java toll road (Toll

Road village) are analyzed as treatment

groups. Sedan villages that are not

traversed by the Trans Java toll road (Non-

Toll Road villages) are villages located in

one district/city with treatment villages

other than villages in one district with

treatment villages used as control groups.

In addition, to clarify the comparison

between the treatment group and the

control group, this study does not include

one district that is still in one sub-district

with the treatment village as a control

village because it is estimated that it is still

getting the effect of the Trans Java toll road

policy on these villages.

The Year Toll variable, which is a

dummy variable, is needed to reflect the

first year of operationalization of Trans Java

toll road sections and the years after

because each toll road section has a

different operational year. In this case, the

toll road sections used are toll road sections

that have the first year of operationalization

in 2006-2018. The main thing in the DID

method is the interaction between Toll

Road and Year Toll which aims to see the

effect or impact before and after

development by the Trans Java toll road on

micro retail modernization in the area. In

addition to the availability of toll road

infrastructure, the growth in the number of

minimarkets and the number of grocery

stalls/stores may be influenced by other

factors, including low conflict and low

crime in the area. Conflict can lead to an

influence on retail(Esmark & Noble, 2016a).

In addition, regions that have higher rates

of violent crime, have fewer food

retailers(Singleton & Cooper, 2019). So to

be able to capture the influence of these

factors, the model in this study uses conflict

or mass fights (conflict) and crime (crime)

as control variables.

Based on the description above, the

basic equation of the DID model is:





 



 







 

















 



















 



 



 



(1)

where 



is the number of

minimarkets in village I in the period of year

t; is 



a dummy application of

the Trans Java toll road development policy

in village I in the period of year t, if the value

is 1 then the village is passed by the Trans

Java toll road, if the value is 0 it is the

opposite; is the 



dummy year of

implementation of the Trans Java toll road

policy and the following years in village I in

the period of year t, if the value is 1 then

the area is passed by the Trans Java toll

road that has been operational in that year,

if the value is 0 it is the opposite; is















a control variable

consisting of dummy conflict in village I in

Arditya Prakoso Putra | 1766

the period of year t, if the value is 1 then

in the area there has been a conflict or mass

fight in the last 1 (one) year, if the value is

0 is the opposite; and dummy crime in

village I in the period of year t, if the value

is 1 then in the area has occurred in the last

1 (one) year, if the value is 0 is the opposite;





is random effects; is 



the year effect,

and 



is the error term for each number

of minimarkets in village I in the period of

year t. The use of year effects is to capture

the average outcome in each village each

year, to ensure that the main explanatory

coefficient (



) can measure variations in

the impact of Trans Java toll road

construction each year.

Analysis using the DID method focuses

on looking at the coefficients of interaction

between variables  and variable

cap. From equation (1) we can

see that the coefficient of the interaction

variabl  



is the

treatment effect, which is the difference in

the average outcomes after and before the

policy in the treatment group minus the

difference in the average outcomes after

and before the policy in the control group.

In this study, to better capture existing

variations, analysis methods were used

using the unbalanced panel data method

and using random effects.

Before identifying the impact of Trans

Java toll road construction on the number

of minimarkets using the DID method, it is

necessary to prove the common pre-

treatment trend assumption. The focus of

the DID method is to identify trends

between treatment and control groups

before policy. To control for potential

confusion in the presence of other factors

that can influence the dependent variable,

a trend of time before the policy is included

to be able to separate the effects of the

policy. Although the purpose of the (Lucas

& Mbiti, 2012) common pre-treatment test

is to overcome other factors that can affect

the dependent variable, this study still uses

control variables as explained above with

the basic equation model (1). This

assumption requires that there is a similar

trend in the number of minimarkets in

control and treatment villages before the

implementation of the Trans Java toll road

development policy.

In checking the common pre-

treatment trend or parallel trend

assumption, this study follows the method

used and uses the basic model of the

regression equation as follows: (Gultom,

2019) (Muralidharan & Prakash, 2013).





  



 





 





 









 



 



(2)

where 



is the number of

minimarkets in the village I in the period of

year t; is a dummy time rescale

variable for each village, where 0 is the first

year of implementation of the Trans Java

toll road policy (in this study 20 06 is used

as the first year of implementation of the

Trans Java toll road policy, the common

pre-treatment test aims to see the similarity

of trends before 2006 ); -1,-2,-3, etc. are the

years before the implementation of the

Trans Java toll road policy in each village,

and 1,2,3, etc. is the opposite; is a dummy

village affected by the Trans Java toll road

policy program, where 1 is the village

traversed by the Trans Java toll road, while

0 is the opposite.

1767 | Impact of Toll Road Development on Modern Retail Growth in Indonesia

For assumption tests in classical

regression, heteroscedasticity tests, and

autocorrelation tests are not performed,

because the panel data sets used in this

study have a considerable number of

observations, but only have a small time

variation. The regression panel data in this

study was estimated using the clustered

standard errors method, which can

overcome the problem of

heteroscedasticity and autocorrelation.

Thus, there is no need for the

heteroscedasticity test and autocorrelation

test (Wooldridge, 2010).

In addition, to strengthen the results of

the study, in this study DID regression was

also carried out to see the impact analysis

of the Trans Java toll road construction

program on the number of Grocery

Stalls/shops. The DID equation used to

measure the impact of the effect of the

Trans Java toll road on the number of

Grocery Stalls/shops is as follows:

 











 







 

















 



















 



 



 



(3)

where 



is the

number of stalls/grocery stores in village I

in the period of year t; is 



dummy application of the Trans Java toll

road development policy in village I in the

period of year t, if the value is 1 then the

village is passed by the Trans Java toll

road , if the value is 0 it is the opposite; is

the 



dummy year of

implementation of the Trans Java toll road

policy and the following years in village I in

the period of year t, if the value is 1 then

the area through which the Trans Java toll

road has been operational in that year, if

the value is 0 it is the opposite; for the same

variable















as used in

equation (1); 



is random effects; is the

year 



effect, and 



is the error term for

each number of stalls/grocery stores in

village I in the period of year t. A common

pre-treatment test is also performed before

analyzing the equation (4).

Data and Sample Characteristics

In this study, an analysis method using

panel data was used. The panel data used

is secondary data obtained from PODES

data in 2006, 2008, 2011, 2014, and 2018

issued by BPS. In addition to data from BPS,

this study also uses data from the Ministry

of PUPR namely maps and locations of all

Trans Java toll road sections that are

operational from 2006 to 2018. The data we

get from the Ministry of PUPR, especially

BPJT is data in the form of a Geographic

Information System (GIS) which is then

processed so that it can see all village data

traversed by the Trans Java toll road.

This study only focuses on villages

traversed by the Trans Java toll road which

was operational from 2006 to 2018. Data on

maps and locations of all Trans Java toll

road sections derived from GIS are

combined using village id and village data

derived from PODES data. The total

number of villages used as treatment

villages is 628 villages. Villages located in

one regency/city other than one sub-

district with villages traversed by the Trans

Java toll road are used as control villages

except for villages that are in one sub-

district with treatment villages. So that the

total number of villages sampled in this

study are 7,656 villages.

Arditya Prakoso Putra | 1768

In Table 4.1 we can see the

characteristics of the sample villages at the

beginning and end of the study year. The

area traversed by the Trans Java toll road or

Toll Road village has a higher average

increase in the number of minimarkets if we

compare it with Non-Toll Road villages.

Then the percentage of villages that

experienced conflicts or mass fights that

occurred in the last 1 (one) year (conflict) in

the first year of the policy, conflicts in Toll

Road villages were higher than those of

Non-Toll Road villages. This becomes

relevant to the assertion that conflict can

cause influence retail (Esmark & Noble,

2016b).

Table 1. Characteristics of Sample Villages

Desa Toll Road

Desa Non Toll Road

2006

(n=578)

2018

(n=628)

2006

(n=6.593)

2018

(n=6.462)

Average Number of Minimarkets

0.28 Unit

1.46 Unit

0.26 Unit

1.17 Unit

Average Number of

Stalls/Grocery Stores

38.56 Unit

43.64 Unit

35.25 Unit

43.63 Unit

Percentage of villages with

Conflict in the last 1 (one) year

3,96%

4,61%

2,81%

4,03%

Percentage of villages with Crime

in the last 1 (one) year

59,68%

67,19%

52,51%

65,11%

Source: BPS processed, 2020

In table 1. we look at the number of minimarkets in Toll Road villages and the number

of minimarkets in Non Toll Road villages from 2006 to 2018. There is a difference in the

number of villages in the treatment group and the control group, of course, we cannot

compare directly about the comparison of the number of minimarkets.

Table 2. Number of Minimarkets in Toll Road Village and Non Toll Road Village Source: BPS

processed, 2020

Tahun

2006

2008

2011

2014

2018

Desa Toll Road

166

(n=578)

285

(n=624)

707

(n=627)

880

(n=627)

923

(n=628)

Desa Non Toll Road

1.724

(n=6.593)

2.353

(n=6.593)

5.710

(n=6.438)

6.173

(n=6.468)

7.613

(n=6.462)

Sumber: BPS diolah, 2020

1769 | Impact of Toll Road Development on Modern Retail Growth in Indonesia

RESULTS AND DISCUSSION

Toll Road Effect on the Number of

Minimarkets

The first step to be able to see how

much effect the construction of the Trans

Java toll road has on the growth of the retail

industry in the villages it passes, will be a

pre-treatment test using minimarkets as

the dependent variable. The test was

conducted to be able to see the similarity

of trends at the time before the

implementation of the policy. The purpose

of the common pre-treatment test is to

address other factors that may affect the

dependent variable. If there is a common

trend between the treatment group and the

control group before the implementation

of the policy, then the use of toll road

development policy variables is considered

sufficient and no control variables are

needed in the equation model. However,

this study still uses control variables in the

equation model to be able to see the

consistency of the regression results of the

equation.

Before analyzing the impact of the

construction of the Trans Java toll road on

the number of minimarkets, we need to

conduct a common pre-treatment test to

be able to see the similarity in the trend of

the number of minimarkets between the

treatment group and the control group

before the policy.

Table 3. Common Pre-treatment Test Results Number of Minimarkets

Dependent Variable : Number of Minimarkets

Pre-treatment coefficient

Independent Variable:



0.0227***

(0.00521)

  

0.00230

(0.00645)

Observations

23,177

Number of Villages

7,647

Note: confidence level 99% (***), 95% (**), 90% (*).

Clustered standard errors based on villages in parentheses.

Source: data processed, 2020

Based on the results in Table 3. it can

be seen that the interaction between

  



, indicated by the

coefficients 



shows insignificant results.

From these results, the null hypothesis

cannot be rejected in this common pre-

treatment test. Theresults of the common

pre-treatment test showed a similar trend

between the number of minimarkets in the

treatment group and the control group

before the implementation of the Trans

Java toll road development policy

operating in the treatment group. These

results show that there is no need for

Arditya Prakoso Putra | 1770

control variables in the equation model.

Showing trends during the pre-treatment

period or the period before the policy (time

rescale < 0) shows no difference in trends

or commons regarding the average market

amount between the treatment group and

the control group.

Figure 1. Average Trend in the Number of Minimarkets Source: data processed, 2020

From table 3. using the treatment

model group (1) we can see that the value

of the coefficient of the interaction variable

() is positive, which is 0.53 in column (1) and

  



0.53 in column

(2) and significant at the level of 5% in both.

The interaction coefficients in column (1)

and column (2) show that on average, after

the construction of the Trans Java toll road,

the trend of increasing the number of

minimarkets in villages traversed by the

Trans Java toll road is higher by 0.53 units

when compared to villages that are not

traversed by the Trans Java toll road and is

significant at 5% level. This result remains

consistent even if some control variables

are added to column (2

Table 4. Effects of Trans Java Toll Road on the Number of Minimarkets

Group Treatment Model (1)

Basic Model

Models with

Control

Variables

Dependent Variable : Number of

Minimarkets

(1)

(2)

Independent Variable:

  

0.530**

0.531**

(0.226)

(0.228)



0.560***

(0.108)



0.316***

(0.0360)

Minimarket

-10 -5 0 5 10

Time Rescale

Toll Road Villages Non Toll Road Villages

1771 | Impact of Toll Road Development on Modern Retail Growth in Indonesia

Observations

35,419

R-squared

0.016

0.013

Number of Villages

7,656

Note: confidence level 99% (***), 95% (**), 90% (*).

Clustered standard errors based on villages in parentheses.

Variables and variables are included in the regression but are not

reported.









Source: data processed, 2020

This finding shows that the Trans Java

toll road program has an impact on

increasing the average number of

minimarkets in villages traversed by the

Trans Java toll road when compared to

villages not traversed by the Trans Java toll

road, where the increasing number of

minimarkets can illustrate the growth of the

retail industry in the region. So we can

conclude that after the construction of the

Trans Java toll road, the number of

minimarkets in the area traversed on

average has increased when compared to

areas that are not traversed by toll roads.

The hope of the growth policy of the

modern retail sector in the area traversed

by the Trans Java toll road, will later create

jobs, increase sources of state revenue

through taxes, and ultimately will increase

the feasibility and eliminate economic

inequality between regions (Erkip &;

Ozuduru, 2015; Zolfaghari et al., 2020).

In the PODES data used, several

villages do not have minimarkets, so the

dependent variable is 0 (zero). To be able

to anticipate data administration problems,

considering that several villages do not

have minimarkets, to be able to test the

consistency of the estimation results from

equation (1), this study is the first

robustness check. The estimation made

does not make village I in period T a sample

in the calculation when the value of the

number of minimarkets in Village I in

period T value 0 (zero). Each regression

result in this study uses the clustered

standard errors method. The main model

in this study uses villages traversed by the

Trans Java toll road as a treatment group

and does not include villages in the same

sub-district with villages traversed by the

toll road as a treatment group because it is

estimated that these villages still get the

effects of the Trans Java toll road policy.

Then the authors performed a

robustness check using 2 different

additional treatment group models. First,

the author uses model (2) where the

equation used is the same as the basic

equation of model (1) but with a different

treatment group. The treatment group in

the model (2) only included villages located

near the Trans Java toll gate to be able to

capture variations in the sample area and

see the consistency of the results. The

treatment group in the model (3) included

all villages in one sub-district with villages

traversed by toll roads to be able to capture

variations in the sample area and see the

consistency of the results.

Arditya Prakoso Putra | 1772

Table 5. Comparison of Robustness Check Effect of Trans Java Toll Road on Number of

Minimarkets in 3 (three) Treatment Group Models

Model (1)

Village Treatment

Group Passed by Toll

Road

(n=628 villages)

Model (2)

Village Treatment

Group located at the

Toll Gate

(n=87 villages)

Model (3)

Treatment Group of All

Villages located in one

sub-district with villages

traversed by toll roads

(n = 2440 villages)

Basic

Model

Models

with

Control

Variables

Basic

Model

Models

with

Control

Variables

Basic

Model

Models

with

Control

Variables

Dependent Variable :

Number of

Minimarkets

(1)

(2)

(3)

(4)

(5)

(6)

Independent Variable:

  

0.530**

0.531**

1.747*

1.754*

0.370***

(0.226)

(0.228)

(0.962)

(0.968)

(0.0928)

(0.0935)



0.560***

0.595***

0.476***

(0.108)

(0.117)

(0.0902)



0.316***

0.330***

0.302***

(0.0360)

(0.0347)

(0.0339)

Observations

35,419

32,760

43,831

R-squared

0.016

0.013

0.017

0.014

0.017

0.015

Number of Villages

7,656

7,115

9,468

Note: confidence level 99% (***), 95% (**), 90% (*).

Clustered standard errors based on villages in parentheses.

Columns 1, 3, and 5 are the base model; and columns 2, 4 and 6 are the base model with

controls

Variables and variables are included in the regression but are not

reported.









Source: data processed, 2020

Table 5. We can compare the results of

the estimated basic equation of model (2)

by using only villages located at the Trans

Java toll gate as a treatment group. The

estimation results from the robustness

check model (2) which only uses villages

located at the Trans Java toll gate as a

treatment group shows consistent results.

From Table 5. In model (2) we can see that

the value of the coefficient of the

interaction variable ()  

is positive, which is 1.74 in

column (1) with a significant level of 10%,

and of 1.75 in column (2) with a significant

value at the level of 10%. 



We can see that the results of the

robustness check model (2) remain

consistent with the main results of the

study in Table 3, so this can strengthen the

main results of the study, namely the Trans

Java toll road construction program has an

impact on increasing the average number

1773 | Impact of Toll Road Development on Modern Retail Growth in Indonesia

of minimarkets in villages traversed by the

Trans Java toll road when compared to

villages that are not traversed by the Trans

Java toll road. From these results, we can

see that the effect of the Trans Java toll road

with a treatment group that only uses

villages located at toll gates has a higher

effect if we compare it with treatment

groups that use villages with locations

along the Trans Java toll road. This shows

that the closer the village is to the toll gate,

the more it has a strong effect on the

number of modern retailers in the village.

Then from table 4. We can also

compare the results of the estimation of the

basic equation of model (3) using all

villages traversed by the Trans Java toll road

including villages in one sub-district as a

treatment group. The estimation results

from the robustness check model (3) which

uses all villages traversed by the Trans Java

toll road including villages in one sub-

district show consistent results. From Table

4. in model (3) we can see that the value of

the coefficient of the interaction

variabl  



() is

positive, which is 0.37 in column (1) with a

significant at level 1 %, and 0.37 in column

(2) with a significant at level 1%.

We can see that the results of the

robustness check model (3) remain

consistent with the main results of the

study in Table 3. So that this can strengthen

the main results of the study, namely the

Trans Java toll road construction program

has an impact on increasing the average

number of minimarkets in villages

traversed by the Trans Java toll road when

compared to villages that are not traversed

by the Trans Java toll road. The results of

Table 4. model (3) show that the use of

treatment groups using all villages located

in one sub-district with villages traversed

by the Trans Java toll road has a smaller

coefficient value than model (1). The results

show that the effect of the construction of

the Trans Java toll road using villages with a

wider scope with toll roads as the treatment

group, shows a smaller effect when

compared to the treatment group in Table

3. Thus, we can conclude that the farther

the location of the village from the Trans

Java toll road, the smaller the effect of the

Trans Java toll road.

Although model (2) and model (3) in

Table 3. show consistent results, this study

still use the estimation results from Table 3.

as the main model because estimation

using the model is considered to have the

number of observations and variations in

control and treatment areas that are more

precise which only use villages traversed by

the Trans Java toll road as a treatment

group. From the estimation results of the

three equation models in table 4. all of

them show positive and significant effects.

This finding certainly supports the

statement of Chandra & Thompson who

said that the area traversed by toll roads

experienced an increase in income in the

manufacturing industry, retail trade

industry, and service industry, whereas in

this study the villages traversed by toll

roads experienced an increase in the retail

trade industry through an increase in the

number of minimarkets (Chandra &;

Thompson, 2000).

Toll Road Effect on Number of Grocery

Stalls/Stores

Then testing equation (3) was carried

out to see the impact of the Trans Java toll

Arditya Prakoso Putra | 1774

road on the number of stalls/grocery

stores, with the test steps being the same

as equation (1). The common pre-

treatment test is obtained by regressing

equation (2) but using the dependent

variable of the number of grocery stores.

The results of the common pre-treatment

test are presented in Table 6.

Table 6. Common Pre-treatment Test Results Number of Grocery Stalls/Stores

Dependent Variable : Number of Grocery Stalls/Stores

Pre-treatment coefficient

Independent Variable:



-0.646***

(0.222)

  

-0.189

(0.270)

Observations

23,177

Number of Villages

7,647

Note: confidence level 99% (***), 95% (**), 90% (*).

Clustered standard errors based on villages in parentheses.

Source: data processed, 2020

Based on the results in Table 6. it can

be seen   that the

interaction between, represented by the

coefficient 



, is not statistically

distinguishable from zero (insignificant), so

we cannot reject the null hypothesis of this

common pre-treatment test. Thus, the

results of the common pre-treatment test

show a similar trend between the number

of stalls/grocery stores in the treatment

group and the control group before the

implementation of the Trans Java toll road

development policy. With the similarity in

trends before the implementation of the

Trans Java toll road development policy.

Empirical tests are carried out by

estimating by regression of panel data

using equation (4). From table 5.5 we can

see that the coefficient values of the

interaction variables are negative, which are

-3.613 and -3.566 in the base model in

column (1) and the model with the control

variable in column (2). However, the results

of both models show no signs of the Trans

Java toll road to the number of grocery

stalls/stores. 



  

Table 7. Effects of Trans Java Toll Road on the Number of Grocery Stalls/Stores

Basic Model

Models with

Control

Variables

Dependent Variable: Number of Grocery

Stalls/Stores

(1)

(2)

Independent Variable:

  

-3.613

-3.566

1775 | Impact of Toll Road Development on Modern Retail Growth in Indonesia

(2.543)

(2.557)



6.535**

(2.756)



9.342***

(0.742)

Observations

35,419

R-squared

0.009

0.007

Number of Villages

7,656

Note: confidence level 99% (***), 95% (**), 90% (*).

Clustered standard errors based on villages in parentheses.

Variables and variables are included in the regression but are not

reported.







Source: data processed, 2020.

Although there is no significant

influence from the impact of the

construction of the Trans Java toll road on

the number of stalls/grocery stores in the

village traversed by the Trans Java toll road,

the increase in the number of minimarkets

is considered to have been able to

represent the growth of the retail industry

as a result of the construction of the Trans

Java toll road. This is in line with the

statement that the transformation of

modern retail can be represented of them

by the increase in the number of

supermarkets (Berger & van Helvoirt, 2018).

However, the government still needs to pay

attention to support for the traditional

retail sector because the segmentation of

traditional retail is a middle to lower-class

community and requires attention to

fulfilling their consumer goods. Support for

traditional retail can be provided in the

form of programs or assistance that can

stimulate traditional retail to remain

competitive with modern retail.

CONCLUSIONS

This research shows that the

implementation of the Trans Java toll road

development policy carried out in areas

traversed by the Trans Java toll road can

encourage modern retail growth. This study

shows that on average after the Trans Java

toll road program operates, the trend of

increasing the number of minimarkets in

villages traversed by the Trans Java toll road

is higher by 0.53 units when compared to

villages that are not traversed by the Trans

Java toll road and this result is significant at

the level of 5%. The increase in the number

of minimarkets in the treatment group is

the influence of the existence of the Trans

Java toll road. The presence of the Trans

Java toll road can provide opportunities for

the development of the modern retail

industry.

In addition, this study also found that

there was no significant influence of the

impact of the construction of the Trans Java

toll road on the number of stalls/grocery

stores in villages traversed by the Trans Java

toll road even though the coefficient of the

interaction variable showed a negative

value. The toll road development program

policy can be one solution to encourage

economic improvement through the

growth of the retail industry in the areas it

Arditya Prakoso Putra | 1776

passes. Referring to this, the government

needs to continue the integrated toll road

development program to have an impact

on the growth of the retail industry. Toll

road construction must still pay attention to

a balance that not only supports modern

retail, but the government still needs to pay

attention to traditional retail by providing

stimulants that can encourage traditional

retail growth to be able to encourage

competitiveness from traditional retail to

modern retail.

REFERENCES

Berger, M., & van Helvoirt, B. (2018). Ensuring

food secure cities – Retail modernization

and policy implications in Nairobi,

Kenya.

Food Policy

(February), 12–22.

https://doi.org/10.1016/j.foodpol.2018.

04.004

Bronnenberg, B. J., Ellickson, P. B.,

Bronnenberg, B. J., & Ellickson, P. B.

(2015).

Adolescence and the Path to

Maturity in Global Retail Adolescence

and the Path to Maturity in Global

Retail1 Distributing of overall economic

activity. Using the United Nations

National Accounts

(4), 113–134.

Card, D., & Krueger, A. B. (1994). Minimum

Wages and Employment : A Case Study

of the Fast-Food Industry in New Jersey

and Pennsylvania.

The American

Economic Review

, 772–793.

Chandra, A., & Thompson, E. (2000a).

Does

public infrastructure affect economic

activity ? Evidence from the rural

interstate highway system

, 457–490.

Chandra, A., & Thompson, E. (2000b).

Does

public infrastructure affect economic

activity ? Evidence from the rural

interstate highway system

, 457–490.

Condeço-melhorado, A., Tillema, T., Jong, T.

De, & Koopal, R. (2014).

Distributive

effects of new highway infrastructure in

the Netherlands : the role of network

effects and spatial spillovers

, 96–105.

https://doi.org/10.1016/j.jtrangeo.2013.

11.006

Dholakia, R. R., Dholakia, N., &

Chattopadhyay, A. (2018a). Indigenous

marketing practices and theories in

emerging economies : Consumer

behavior and retail transformations in

India.

Journal of Business Research

(October 2016), 406–415.

https://doi.org/10.1016/j.jbusres.2017.0

9.030

Dholakia, R. R., Dholakia, N., &

Chattopadhyay, A. (2018b). Indigenous

marketing practices and theories in

emerging economies : Consumer

behavior and retail transformations in

India.

Journal of Business Research

(October 2016), 406–415.

https://doi.org/10.1016/j.jbusres.2017.0

9.030

Dolislager, M. J. (2017).

Food Consumption

Patterns In Light Of Rising Incomes,

Urbanization, And Food Retail

Modernization : Evidence From Eastern

And Southern Africa By Submitted to

Michigan State University in partial

fulfillment of the requirements for the

degree of Agricultural, Food, and

Resource Economics — Doctor of

Philosophy

Erkip, F., & Ozuduru, B. H. (2015). Retail

development in Turkey : An account

after two decades of shopping malls in

the urban scene.

Progress in Planning

102

, 1–33.

https://doi.org/10.1016/j.progress.2014.

07.001

Esmark, C. L., & Noble, S. M. (2016a). Bad

behavior and conflict in retailing

1777 | Impact of Toll Road Development on Modern Retail Growth in Indonesia

spaces : Nine suggestions to ease

tensions.

Business Horizons

(1), 95–

104.

https://doi.org/10.1016/j.bushor.2015.0

9.004

Esmark, C. L., & Noble, S. M. (2016b). Bad

behavior and conflict in retailing

spaces : Nine suggestions to ease

tensions.

Business Horizons

(1), 95–

104.

https://doi.org/10.1016/j.bushor.2015.0

9.004

Gultom, Y. (2019). Governance structures and

efficiency in the U . S . electricity sector

after the market restructuring and

deregulation.

Energy Policy

129

, 1008–

1019.

Lucas, A. M., & Mbiti, I. M. (2012). Access,

Sorting, and Achievement : The Short-

Run Effects of Free Primary Education in

Kenya.

American Economic Journal:

Applied Economics

(4), 226–253.

Maruyama, M., Wu, L., & Huang, L. (2016). The

modernization of fresh food retailing in

China: The role of consumers.

Journal of

Retailing and Consumer Services

33–39.

https://doi.org/10.1016/j.jretconser.201

5.12.006

Michaels, G. (2008).

The Effect of Trade on the

Demand for Skill : Evidence from the

Interstate Highway System Author ( s ):

Guy Michaels Source : The Review of

Economics and Statistics, Nov ., 2008,

Vol . 90, No. 4 (Nov ., 2008 ), Published

by : The MIT Press Stable URL :

http://www.jstor.com/stable/40043108

THE EFFECT OF TRADE ON THE

DEMAND FOR SKILL : EVIDENCE FROM

THE INTERSTATE HIGHWAY SYSTEM

(4), 683–701.

Muralidharan, K., & Prakash, N. (2013).

Cycling to School Increasing Secondary

School Enrollment for Girls in India

Paper, W., Paper, D., & Paper, D. (2016).

www.econstor.eu

Reardon, T., & Hopkins, R. (2006).

The

Supermarket Revolution in Developing

Countries : Policies to Address Emerging

Tensions Among Supermarkets,

Suppliers and Traditional Retailers

(4), 522–545.

https://doi.org/10.1080/095788106010

70613

Reardon, T., Timmer, C. P., Minten, B., Reardon,

T., Timmerc, C. P., & Mintend, B. (2012).

farmers Supermarket revolution in Asia

and emerging development strategies

to include small farmers

109

(31),

12332–12337.

https://doi.org/10.1073/pnas.10031601

Schipmann, C., Qaim, M., Schipmann, C., &

Qaim, M. (2020).

Agricultural & Applied

Economics Association Modern Food

Retailers and Traditional Markets in

Developing Countries : Comparing

Quality, Prices, and Competition

Strategies in Thailand Modern Food

Retailers and Traditional Markets in

Developing Countries : Comparing

Quality Prices, and Competition

Strategies in Thailand

(3).

https://doi.org/10.1093/aepp/ppr018

Shatz, Howard J; Kitchens, Karin E;

Rosenbloom, S. M. (2011). The Effects of

Highway Infrastructure on Economic

Activity.

Highway Infrastructure and the

Economy

, 15–42.

Singleton, C., & Cooper, M. (2019). P34

County-Level Associations Between

Food Retail Outlet Availability and

Violent Crime Rate P33 An Exploration

of Health Intermediaries Competencies

with Nutrition Labelling Information on

Supplemented Foods P36 Examining

Arditya Prakoso Putra | 1778

the Healthiness of Food and Beverage

Options at Highway Rest Areas in North

Carolina P35 Examining the Feasibility of

Selling Whole Grain Snack Packs in New

York City Corner Stores Participating in a

Healthy Retail Program.

Journal of

Nutrition Education and Behavior

(7),

S47–S48.

https://doi.org/10.1016/j.jneb.2019.05.4

Wooldridge, J. M. (2010).

Econometric

Analysis of Cross Section and Panel

Data

. The MIT Press.

Yoo, J., & Hoon, J. (2016). Straw effects of new

highway construction on local

population and employment growth *.

Habitat International

, 123–132.

https://doi.org/10.1016/j.habitatint.201

5.11.009

Yoshino, N., & Abidhadjaev, U. (2017). An

impact evaluation of an investment in

infrastructure : The case of a railway

connection in Uzbekistan $.

Journal of

Asian Economics

, 1–11.

https://doi.org/10.1016/j.asieco.2017.02

.001

Zhang, X., Hu, Y., & Lin, Y. (2020). The

influence of highway on the local

economy: Evidence from China’s

Yangtze River Delta region.

Journal of

Transport Geography

(October

2019), 102600.

https://doi.org/10.1016/j.jtrangeo.2019.

102600

Zolfaghari, M., Kabiri, M., & Saadatmanesh, H.

(2020). Impact of socio-economic

infrastructure investments on income

inequality in Iran.

Journal of Policy

Modeling

https://doi.org/10.1016/j.jpolmod.2020.

02.004

for possible open-access

publication

under the terms and conditions of the Creative

Commons Attribution (CC BY SA) license

(https://creativecommons.org/licenses/by-sa/4.0/).