Economic Growth, Social Expenditure, Unemployment, and

Inflation: The Impact on Poverty in South Sumatera

Harunurrasyid, Syaipan Djambak, Mardalena, and Putri Adelia

Faculty of Economics, Universitas Sriwijaya, Palembang, Indonesia

Keywords: Regencies/Cities in South Sumatera, Poverty, Model Common Effects

Abstract: This study examines factors that influence and how to accelerate the decline in the number of poor people in

South Sumatera Province from 2009 to 2017. The research used panel data regression, with the Common

Effects Autoregressive (1) as the selected model. The results showed that the level of poverty in South

Sumatra Province experienced an annual average decrease of 1.87 percent, while atsame period, several

regencies/cities in South Sumatera experienced an increase, namely: Ogan Komering Ulu, Palembang, East

OKU, Pagaralam and Prabumulih. The difference in changes in the number of poor people is a result of the

occuring structural inequality, because the transition of dominant economic sectors in GDRP is not followed

by the optimal labors absorption in the field. The structural imbalance contributes to the economic

inequality, poverty, open unemployment and natural resources exploitation in South Sumatera. The test

results show an increase in GDRP growth and social expenditure, and a decrease in the open unemployment

rate and inflation affect the growth of poor people number in South Sumatera.

1 INTRODUCTION

Economic development still leaves a number of

issues that should receive serious attention. One of

the main problems that arises as a result of the

implementation of inconsistent and impartiality

national development programs is the widening

inequality and chronic poverty (Sugema, Irawan,

Adipurwanto, Holis, & Bakhtiar, 2010).

Poverty is one of fundamental problems which

become major concern of Indonesia Government

(Budiantara, Diana, & Darmesto, 2011). Poverty

illustrates the living condition of many developing

countries in the world, which consist of not less than

one billion of the world's population. The reality

shows that most development efforts in poverty

alleviation programs have not been sufficiently able

to suppress the increase the number of poor people

in many countries. The condition was compounded

by demographic bonus events in many developing

countries over the past few decades. This increase in

demographic bonus later increases the number of

poor people, even though the increase in population

does not supposed to increase the number of poor

people.

Badan Pusat Statistik (BPS) publication stated

that the percentage of Indonesia's poor population as

of March 2018 was only 9.82 percent, which, if it

was estimated as many as 25.95 million people, the

fewer of the poor population in September 2017

which were 26.58 million people. The statement is

in line with World Bank publication data in 2017,

where the percentage of poor people in Indonesia is

10.6 percent, with 7.7 percent of poverty comes

from the urban areas and 13.9 percent from rural

areas (World Bank, 2018).

In addition, according to data from the

publication of the Human Development Report

(HDR), Indonesia ranked 116th out of 189 countries

with HDI values of 0.694 percent, along with a

Gross National Product Percapita PPP$ value of $

10,846 in year basic 2011(UNDP, 2018).

Indonesia poverty rate is only 9.82 percent as of

March 2018, as reported by BPS, using Purchasing

Power Parity (PPP) methodology as it base. PPP is

the value of purchasing power of people determined

based on the standards of each country, ignoring the

prevailing international exchange rate (World Bank,

2018).

Poverty of Indonesia is calculated using the

urban poverty line (Rp. 400,995/ capita/ month) and

rural poverty line (Rp. 370,910/ capita/ month). In

626

Harunurrasyid, ., Djambak, S., Mardalena, . and Adelia, P.

Economic Growth, Social Expenditure, Unemployment, and Inﬂation: The Impact on Poverty in South Sumatera.

DOI: 10.5220/0008443306260636

In Proceedings of the 4th Sriwijaya Economics, Accounting, and Business Conference (SEABC 2018), pages 626-636

ISBN: 978-989-758-387-2

other words, every Indonesian citizen is considered

not poor if the income per capita per day of them is

Rp. 13,366 for urban-dweller and Rp. 12,363 for

rural areas-dweller (BPS, 2018). The national

average poverty line determined by BPS is higher

than the World Bank Purchasing Power Parity

standard (Novalia, 2018).

The calculation of poverty using the World

Bank’s PPP standard of US$ 1.9 per day is measured

using 2011 as its base year. The estimated

conversion of 1 US dollar in 2011 is

Rp11,157/capita/day, which then shifted to

Rp13,162/capita/day in 2018 (International

Monetary Fund, 2018).

Even Enny Sri Hartati, the Director of the

Institute for Development of Economics and Finance

(INDEF) stated: “The poverty rate of Indonesia

number could be doubled to 70 million people if the

USD 1.9/capita/day standard is used” (Suryowati

06/03/2018, https://www.jawapos.com/jpg-

today/06/03/2018/).

The statement comes from the result of

calculation using the applicable international

exchange rate. Using the international exchange rate

as of March 2018 (Rp. 13,761), the monthly poverty

standard is Rp. 784,377/capita/month, equivalent to

(Rp. 26,146/ capita/ day). It doubles the standard of

Purchasing Power Parity (PPP) used by BPS for

both urban and rural poverty standards (Suryowati,

Tuesday (6/3) JawaPos.com,

https://www.jawapos.com/jpg-today/06/03/2018/).

South Sumatera Province ranked the third in the

largest poor residents number in Sumatera, and

ranked the seventh in all Indonesia (BPS, 2018). The

number of poor people in South Sumatra Province is

beyond the percentage of the average number of

poor people in Indonesia. BPS reported in Semester

1 of 2018, the percentage of the poor population of

South Sumatra Province was 12.80 percent, or

exceed the average number of poor people in

Indonesia (9.82 percent). The condition has been

continuously occuring for the past four years, where

the percentage target of the number of poor people

in South Sumatra Province is always above the

national average of 7.5 percent. Therefore, further

research is needed in determining the factors

affecting the level of poverty in the regencies/cities

in South Sumatera Province. The results are

expected to provide as a reference in determining the

direction of poverty alleviating policies.

Research on the factors determining the number

of poor people has been widely carried out, both in

national and provincial scale. (Fajriyah & Rahayu,

2016) has conducted a modeling analysis of the

factors that influence poverty in the regencies/cities

in East Java Province with panel data regression,

which then revealed that the significant predictor

variables included literacy, labor force participation,

number of population working in the agricultural

sector, as well as GDRP per capita. Meanwhile,

predictor variables with no effect on the response

variable are residents without health access.

Furthermore, research related to poverty modeling is

carried out by (Zuhdiyati & Kaluge, 2015) with the

results showing that the HDI has a significant

negative effect on poverty and the open

unemployment rate has no significant effect.

However, previous research conducted by Yacoub

(2012) shows that the open unemployment rate has a

significant effect on poverty.

2 LITERATURE REVIEW

2.1 Poverty Definition

Poverty can be illustrated as a situation where

there is a lack of common things related with the life

quality, such as food, clothing, shelter and clean

water. Economically, poverty can be indicated by

the level of lack of resources in fulfilling the needs

of life and improve the welfare of a group of people.

According to (Suryawati, 2005) poverty can be

divided into four forms : 1) absolute poverty, where

the income is below the poverty line or not enough

to fulfill the standard needs of food, clothing, health,

shelter and education needed to maintain live and

gain job; 2) relative poverty, which caused by the

failure of development policies in reaching all the

layer of communities, causing inequality in income

among residents; 3) cultural poverty, refers to the

attitude matter of a person or community caused by

cultural factors, such as the reluctancy to improve

the level of life, wasteful, and not creative; and 4)

structural poverty, a poverty caused by minimum

access to resources that occur in socio-cultural and

political systems that do not support poverty

alleviation.

2.2 Measurement of Poverty

According to BPS, in 2018, the level of poverty

is based on the amount of rupiah consumption spent

for food, precisely 2.100 calories per person per day

(from 52 types of commodities representing the

consumption patterns of residents of lower-class),

and non-food consumption of 45 types foods

commodities in accordance with national agreements

Economic Growth, Social Expenditure, Unemployment, and Inﬂation: The Impact on Poverty in South Sumatera

627

and are not differentiated between rural and urban

areas). The adequacy of 2.100 calories applies to all

ages, genders, level of physical activeness, body

weight, and estimated physiological status of

population size, this measure is often referred to as

the poverty line.

Otherwise, the human development model from

UNDP is also used to measure the poverty in

Indonesia. Periodically every three years since 1991,

UNDP has been publishing the Human Development

Report (HDR). The human development approach is

different from conventional approaches such as

economic growth, human resource development and

community welfare development. The economic

growth approach only pursues an increase in Gross

National Product (GNP) rather than improving the

quality of human life. The human resource

development approach makes humans an input

factor in the production process, making humans are

seen more as tools than as goals. While in the

concept of human development, growth is not seen

as a goal but as an instrument to reacht the goal.

Table 1: Types of Poverty Indexes and Indicators

INDEX

INDICATOR

HDI

 Living expectancy level

 Adult literacy

 Average education length

 Purchasing power rate per capita

HPI

 Number of births unable to live until

 Adult illiteracy

 Percentage of the residents without

access to clean water

 Percentage of residents without

access to healthcare

 Percentage of underfed children

GDI

 Life expectancy of men and women

 Literacy of men and women

 Average education length of men and

women

 Estimated income level of men and

women

GEM

 Percentage of number of DPR

members from men and women

 Percentage of senior level employees,

managers, professionals and technical

positions of men and women

 Estimated income levels of men and

women

Source : (Cahyat, 2004)

The HDR contains an explanation of three

indexes: Human Development Index (HDI), the

Gender Development Index (GDI), Gender

Empowerment Measure (GEM) and the Human

Poverty Index (HPI). In Indonesia, HDR uses BPS

data, especially Susenas data, so it has the same

survey unit as BPS, namely households.

3 RESEARCH METHODOLOGY

3.1 Scope of Research

According to (Sarris, 2001) many results of the

study showed that economic growth has an

important role in reducing poverty, and governments

need more detailed information about it to make a

sufficient decisions in allocating APBN and APBD.

In this study the test was conducted twice with

different variable components. The first test was

carried out on multidimensional variable and the

second test on economic dimension variable. The

growth of Poverty rates (PM) playing the role as

dependent variable in this study, while the growth of

GDP with Constant Prices Year 2010 (GDP), the

growth of expenditure allocation for social

expenditure (BS), open unemployment rate (TPT) ,

Inflation (I), Literacy (MH), Feasibility Board (KP),

Clean Water Access (AAB), growth of Education

Participation Less than Middle School (PKS) and

Growth of Malnutrition Toddler Number (GB) are

the independent variables in this study. The data

used in this study are pooled data of 15 regencies/

cities in South Sumatra province in year 2008 to

2017, plus additional secondary data from BPS,

IDHS, and Bank Indonesia (BI).

The time-span of variables in this study started

from year 2009 to 2017.

Table 2: Operational and Dimensional Variables

Definition

Research

Variables

Operational

Definition

Dimension

Growth of

Poor People

Number

Percentage of

growth of poor

people (population

with average per

capita expenditure

per month below

the poverty line)

Economy

Growth of

Constant Price

Year 2010

GDRP

Percentage of

GDRP growth of

each regency/ city

SEABC 2018 - 4th Sriwijaya Economics, Accounting, and Business Conference

628

Growth of

Fund Allocated

for Social

Ependiture

Percentage of

growth in social

ependiture

allocation funds

(for social

expenditure and

included in the

regional

government

expenditure

budget)

Economy

Inflation

Percentage of

annual inflation

rate

Open

Unemployment

Rate

Percentage of

unemployment to

the total workforce

Literacy

Number

Percentage of

population aged 15

years and over who

can read and write

latin letters and/or

other letters

Education

Growth of

Education

Participant

Less than

Middle High

School

Percentage of

school-age

population

attending school at

the primary school

level in both

private and public

school

Household

with Clean

Water Access

Percentage of

households who

can access clean

water

Social

Household

with Board

Feasibility

Percentage of

households with

housing facilities in

the form of

permanent walls

The Growth of

Number of

Malnutritioned

Toddler

Percentage of

children under five

suffering from

malnutrition

3.2 Step of Analysis

The steps of data analyzing in this study are as

follows (Baltagi, 2005):

1. Estimating the panel data regression using a

fixed effect model.

2. Perform a Chow test

a) If accepted, then the common effect model is

used (continue step 5).

b) If rejected, then the fixed effect model is used

(continue step 4).

3. Conduct the Hausman test

a) If received, then the random effect model is

used (continue step 5).

b) If rejected, then the fixed effect model is

used (continue step 4).

4. Perform an assumptions test on selected models.

5. Perform a parameter significance test which

includes simultaneous test and partial test with

the revised regression equation

6. Dispose of some research variables that are not

in accordance with the theory.

7. Interpret the final model of panel data regression

with the selected model.

4 ANALYSIS AND DISCUSSION

4.1 Poor People and Poverty Factors in

Regencies/Cities in South Sumatera

The percentage of poor people in South Sumatra

Province over the past decade tends to decrease. It

was 17.67 percent in 2008, 14.80 percent in 2010

and 13.48 percent in 2012. However, in 2015 it

increased to 13.82 percent, followed by 13.19

percent in 2017. In other words,in 2017 there were

1,086,920 poor people out of 8.052.315 total

population in South Sumatra. The average

percentage indicates that there are about 123 poor

people living in every 1000 people in 17 regencies /

cities in South Sumatera Province.

Economic Growth, Social Expenditure, Unemployment, and Inﬂation: The Impact on Poverty in South Sumatera

629

Table 3: The Percentage of Average Growth of Poor People in Regencies/Cities of South Sumatera Province

Year 2009 - 2017

Regency/City

Average Growth

Poor

People

Social

Expenditure

Open

Unemployment

Inflation

Ogan Komering Ulu (OKU)

2,33

-11,03

4,17

4,89

Palembang

1,89

-11,03

4,17

4,89

Ogan Komering Ulu Timur (OKUT)

0,90

0,47

6,14

4,89

Ogan Komering Ilir (OKI)

0,58

-10,27

1,69

4,89

Pagaralam (PGA)

0,44

1,93

-4,73

3,57

Prabumulih

0,11

10,19

0,32

3,57

Lubuklinggau

-0,57

3,25

-6,65

3,57

Ogan Ilir

-1,59

0,39

-4,99

4,89

Lahat

-1,62

2,04

0,38

3,57

Banyuasin

-1,62

2,31

8,43

4,89

Ogan Komering Ulu Selatan

-2,03

-9,23

3,02

4,89

Musi Banyuasin

-2,22

-2,80

-4,84

4,89

Empat Lawang

-2,68

0,00

-4,51

3,57

Muara Enim

-3,83

10,35

-7,37

3,57

Musi Rawas

-7,18

1,14

-1,15

3,57

Sumatera Selatan Province

-1,87

-8,43

-1,31

4,27

Source: South Sumatera in the Number (reprocessed by authors)

Table 4: The Percentage of Average Contribution of Each Regencies/Cities to South Sumatera Province

REGENCIES/

CITIES

The Percentage of Average Contribution of Each Regencies/Cities to South Sumatera

Province Year 2009 – 2017

Poor People

GDRP of

2010

Constant

Price

Allocation of

Social

Expenditure

Funds

Open Unemployment

Rate

Inflation

Palembang

18,26%

27,35%

3,32%

13,55%

7,62%

Ogan Komering Ilir

11,14%

7,88%

23,20%

7,47%

7,62%

Musi Banyuasin

10,30%

16,53%

0,95%

5,62%

7,62%

Muara Enim

9,19%

12,95%

15,64%

6,54%

5,57%

Banyuasin

8,95%

7,65%

1,56%

5,36%

7,62%

Musi Rawas

8,23%

4,57%

15,00%

2,71%

5,57%

Kab. Lahat

6,50%

4,74%

10,75%

5,13%

5,57%

OKU Timur

6,03%

3,93%

11,01%

5,37%

7,62%

Ogan Ilir

5,22%

2,73%

1,21%

4,65%

7,62%

Ogan Komering Ulu

3,81%

3,04%

8,21%

6,13%

7,62%

OKU Selatan

3,62%

3,07%

0,71%

3,52%

7,62%

Empat Lawang

2,96%

1,27%

2,55%

5,93%

5,57%

Lubuklinggau

2,81%

1,57%

4,33%

11,15%

5,57%

Prabumulih

1,87%

1,82%

0,73%

9,53%

5,57%

Pagaralam

1,11%

0,92%

0,83%

7,33%

5,57%

South Sumatera

Province

100,00%

Source: South Sumatera in the Number (reprocessed by authors)

SEABC 2018 - 4th Sriwijaya Economics, Accounting, and Business Conference

630

The average growth of the number of poor

people in South Sumatra Province in 2009-2017

decreased by 1.87 percent. This decline did not

occur in several regencies/cities in South Sumatra

Province. OKU Regency experienced an increase in

the number of poor people on average 2.33 percent

per year. This was caused by the high average

growth of open unemployment of 4.17 percent in

OKU Regency with its contribution to open

unemployment in South Sumatra Province of 6.13

percent, and caused altogether by high inflation with

7.62 percent average annual contribution to the

province. Furthermore, the average increase in the

number of poor people in Palembang City was 1.89

percent, OKUT Regency 0.90 percent, OKI District

0.58 percent, PGA City 0.44 percent, and

Prabumulih City 0.11 percent. Conversely, there are

several other regencies/cities that experience a

decline in the average number of poor people. The

highest decline was Musi Rawas Regency at 7.18

percent, affected by the decrease in the average

annual open unemployment in Musi Rawas Regency

by 1.15 percent, and 1.14 percent increase in the

average annual allocation of social expenditure.

By its contribution to the number of poor people

in South Sumatra Province during 2009 s.d 2017,

Palembang City is the greatest conributor with 18.26

percent. The highest GDRP of Palembang City was

37.65 percent, which came from the manufacturing

industry sector, which only absorb 11.84 percent of

workers from the labor force, while the largest

labors-absorber sector in Palembang City was the

groceries, retails, restaurants and hotels sector,

which was 33.70 percent,contrary to itssmall

contribution(15.57 percent)to GDRP. This caused

the open unemployment rate in Palembang City to

become the largest contributor to the Province,

reaching 13.55 percent. The position of Palembang

City as the provincial capital also plays a role in

attracting rural communities to urbanize, causing a

massive population growth. Large urbanization

flows lead to an increase in the number of labor

force, but most of the workforce does not have the

ability desired by the market so this actually increase

the number of open unemployment. In addition, the

high inflation factor causes the allocation of social

expenditure funds uneffectivein reducing the number

of poor people in Palembang City.

The lowest contribution of the poor people to

South Sumatera comes from Pagaralam City at 1.11

percent. The condition was happeneddue to the

ability of GDRP in the agriculture, forestry, hunting

and fisheries sectors to absorb the largest workforce

of 50.25 percent, making it the largest GDRP

contributor (25,38 percent) of Pagaralam City over

the past 9 years.

Table 5: The Structure of Average Poverty Contribution of

Regencies/Cities in South Sumatera 2009 s.d 2017

Regency

/ City

Contribution

Percentage

Total

Contributio

to Province

Above

the

Province

Average

Palemba

Ogan

Komerin

g Ilir

Musi

Banyuasi

Muara

Enim

Banyuasi

Musi

Rawas

18,26%

11,14%

10,30%

9,19%

8,95%

8,23%

40,00%

Province

Average

6,67%*

Under

Province

Average

Lahat

OKU

Timur

Ogan Ilir

Ogan

Komerin

g Ulu

OKU

Selatan

Empat

Lawang

Lubuklin

ggau

Prabumu

lih

Pagarala

6,50%

6,03%

5,22%

3,81%

3,62%

2,96%

2,81%

1,87%

1,11%

60,00%

Source : processed data

The average contribution of poor population in

South Sumatra Province over the last 9 years is 6.67

percent. From table 5 above there are six

regencies/cities with the level above the average

contribution of the poor population in South Sumatra

Province: Palembang City (18.26%), Ogan

Komering Ilir Regency (11.14%), Musi Banyuasin

(10.30%),Muara Enim (9.19%), Banyuasin Regency

(8.95%) and Musi Rawas Regency (8.23 percent).

At the other hand,there are 11 districts and/or cities,

covering 60 percent of the average contribution of

the poor population of South Sumatra Province,

namely: Lahat Regency (6.50%), East OKU

(6.03%), Ogan Ilir (5.22%), Ogan Komering Ulu

Economic Growth, Social Expenditure, Unemployment, and Inﬂation: The Impact on Poverty in South Sumatera

631

Regency (3.81%), South OKU (3.62%), Empat

Lawang Regency (2.96 %), Lubuklinggau City

(2.81%), Prabumulih City (1.87 %) and Pagaralam

City (1.11%).

Based on BPS data (2018), poverty depth index

(P1) and poverty severity (P2) of South Sumatra

Province tend to increase during the last three years.

In 2017, it sequentially reached 2.35 percent and

0.62 percent. This shows a trend towards an increase

in income inequality and an increase in public

consumption expenditure. That is, the need for

serious attention related to public consumption

patterns due to the direct impact on the difference in

prices of volatile foods, the ability of the purchasing

power of the people and the impact on the rate of

inflation. Public consumption expenditures on food

consumption are concentrated on processed foods

and beverages, grains, tobacco and betel, fish,

shrimp and shellfish, as well as vegetables.

Meanwhile, the concentration of non-food

expenditure concentrates on housing and household

facilities, various goods and services, and durable

goods.

Labor conditions in South Sumatra Province

show the number of Open Unemployment Rate in

2017 is 195,222 people, with the highest number of

Palembang City as many as 81,449 inhabitants.

Based on data from the last level of education

completed by job seekers, both in the city and

village, the highest percentage in 2017 was a

graduate at the senior high school level with a

percentage of cities of 12.84 and a percentage in the

village of 12.03. Thus, the total open unemployment

rate at the high school level in cities and villages in

2017 is 12.46 percent. In other words, in that

year,there were 10,369 unemploymentwho never

had any formal education; 20,392 elementary school

graduates; 677,008 middle high school graduates;

985,974 high school graduates; and 452,511 D/I-

III/Academy/University graduates in every

4,123,669 workforce.The final level of education

also influences job positions of each workforce, and

later affect income earned(BPS, 2016).The

relationship between poverty and education is

particularly important because of the key role played

by education in raising economic growth and

reducing poverty. The better educated have higher

incomes and thus are much less likely to be poor

(World Bank, 2005).

And also, Inflation has affects for the poor more

than the rich. This is especially true in terms of food,

energy, and housing inflation. In fact, a number of

studies on inflation and poverty in developing

countries have shown the effect of inflation on the

poor (Odekon, 2015).

4.2 Research Model Specification Test

4.2.1 Multidimensional Variables Testing

Chow Test

Chow test is the initial stage of model

specification testing to choose the common effect

model and the fixed effect model.

Table 6: Redundant Fixed Effects Tests Results

Effects Test

Statistic

Prob.

Cross-section F

0.517670

0.9185

Cross-section Chi-square

8.538564

0.8594

S.E. of

Regression

0.103188

F-statistic

3.516014

Prob(F-statistic)

0.000654

The chow test results show that:

F = 0.103188≤ F

(14;111;5%)

= 0.517670

Because the value of F ≥ F

(14;111;5%),

and Chi-

Square probability value 0.8594 > 0.05, Ho is

accepted,which means the right model is the fixed

effect model.

The correct model used in this study is the

common effect. If the estimation model chosen is the

common effect, there is no need to do a thirst test

and classical assumption test. The next step is to test

the significance of the parameters.

Parameter Significance Test

A. Simultaneous Test

Simultaneous testing is done to see the effect of

the overall independent variable on the dependent

variable. The test results show the probability value

(F-statistic) of 0.00 <0.05, so that Ho is rejected,

meaning that the independent variables

simultaneously affect the dependent variable.

Table 7: Simultaneous Test Results of Multidimensional

Variables

R-squared

0.402013

Mean dependent var

-0.011393

Adjusted R-squared

0.344558

S.D. dependent var

0.111567

S.E. of regression

0.103188

Akaike info criterion

-1.633337

Sum squared resid

1.330976

Schwarz criterion

-1.418132

Log likelihood

120.2503

Hannan-Quinn criter.

-1.545884

F-statistic

3.516014

Durbin-Watson stat

2.561604

Prob(F-statistic)

0.000654

SEABC 2018 - 4th Sriwijaya Economics, Accounting, and Business Conference

632

B. Partial Test

Partial test is conducted to see the effect of

individual independent variables on the dependent

variable, with the criteria if the probability of t value

or significance is < 0.05, there is an influence

between the independent variables on the dependent

variable partially, and vice versa.

Table 8: Multidimensional Variable Partial Test Result

Variab

Coeffi-

cient

t-number

Prob.

Conclusion

PDRB

-1.542969

-3.642670

0.0004

Significant

-0.001383

-2.940471

0.0039

Significant

0.009761

1.506454

0.1345

Not Significant

TPT

-0.003461

-0.873012

0.3843

Not Significant

-0.003877

-0.970246

0.3338

Not Significant

PKS

0.002914

0.328463

0.7431

Not Significant

AAB

0.001433

2.318121

0.0221

Significant

0.000583

0.17901

0.8582

Not Significant

0.007124

1.699702

0.0917

Not Significant

Based on the test results, it can be seen that in the

significance level of α = 5%, the growth of GDRP,

literacy, open unemployment rate, the number

malnutritioned toddler, the number of people with

education level less than middle school, house with

feasible boarding and inflation do not significantly

affect the growth of poor people number in

regencies/cities in South Sumatra Province during

2009 s.d 2017.

The panel data regression test results also show

that the independent variables of research on the

social and educational dimensions do not have a

significant influence on the dependent variable, so

that the next research step (step 6) was taken to get

the best model with selected variables. Therefore,

the the social and educational dimensions variable

were released from the model, leaving the testing

model with only economic dimension variable,

including: the growth of constant price year 2010

GDRP, the growth of fund allocation for social

expenditure, inflation, and open unemployment

level. The basic assumptions that lead to retesting

only economic dimension is caused bythe national

standard used by BPS, where the calculation is only

based on the ability of poor people infulfillingtheir

basic needs (basic needs approach) measured by the

average spending/ capita/ month according to

poverty line (BPS, 2017).

4.2.2 Economic Dimension Variable Testing

Chow Testis the basic step in model specification

testing, in order to choose common effect model and

fixed effect model.

Table 9: Redundant Fixed Effects Tests Results

Effects Test

Statistic

Prob.

Cross-section F

0.758479

0.7112

Cross-section Chi-square

11.824655

0.6204

S.E. of

Regression

0.104891

F-statistic

5.399550

Prob(F-statistic)

0.000470

The chow test results show thatthe value of

F = 0.104891≥ F (14; 116; 5%) = 0.758479.

Because the value of F ≤ F (14; 116; 5%), and Chi-

Square probability value 0.00> 0.05, then Ho is

accepted, which means the right model is a common

effects model.

As in the testing of multidimensional variables,

because the best estimation model is the common

effect model, the next step of the test is the

significance of the parameters.

Parameter Significance Test

A. Simultaneous Test

Simultaneous testing is performed to see the

effect of the overall effect independent variable has

caysed towards the dependent variable. According to

the probability value (f-statistic) significance value

of 0.00 ≤ 0.05, Ho is rejected, which means that

overall the independent variables simultaneously

affect the dependent variable.

B. Partial Test

Partial testing is performed to see the effect of

individual independent variables on the dependent

variable. If the probability of t-value or the

significance is <0.05, then there is partially an

influence between the independent variables on the

dependent variable partially, and vice versa.

Table 10: Partial Test Results of Economic Variable

Variable

Coeffi-

cient

number

Prob.

Conclusion

GDRP

-1.4631

-3.5208

0.0006

Significant

-0.0079

1.8898

0.0610

Not

Significant

-0.0011

2.5375

0.0123

Not

Significant

TPT

0.0027

0.7910

0.4304

Not

Significant

Economic Growth, Social Expenditure, Unemployment, and Inﬂation: The Impact on Poverty in South Sumatera

633

Based on the test results, it can be seen that in the

significance level of α = 5%, only the GDRP growth

and the significant growth of fund allocation for

social expenditurevariable are significant. The next

step is to improve the model with autoregressive (1)

to acquire the best model.

Table 11: Economic Variable Partial Test Results

withAutoregressive (1)

Variabl

Coeffi-

cient

t-number

Prob.

Conclusion

GDRP

-1.3324

-3.5208

0.0022

Significant

0.0084

1.8898

0.0661

Not

Significant

-0.0022

-2.5375

0.0046

Significant

TPT

0.0077

0.7910

0.0232

Significant

After improving the model with Autoregressive

(1), only the inflation variable was not significant

with a significance value of only 0.06 in the

significance level α = 5 percent.

4.3 Panel Data Final Regression Model

Based on several tests that have been carried out,

the final panel data regression model for the number

of poor people in the regencies/cities of South

Sumatra Province during 2009 s.d 2017 is the

common effects model with Autoregressive (1).

Table 12: Economic Variable Panel Regression Results

Variable

Coefficient

Variabel

Connectivity

GDRP

-1.3324

Negative

0.0084

Positive

-0.0022

Negative

TPT

0.0077

Positive

With the equation of panel data regression,

Growth of PM = -0,0379 -1,3324 GDRP

Growth-0,0022 BS Growth

+0,0077TPT +0,0084I

Where:

Growthof PM = Percentage of the poor people

number in South Sumatera

provincial regencies / cities during

2009 to 2017

GrowthofGDRP = Percentage of Constant Price

Year 2010 GDRP

BS Growth = Percentage of Fund Allocated

for Social Expenditure

TPT = Open Unemployment Rate

I = Annual Inflation

Regression models for each variable can be

interpreted as follows:

1. Decreasing total GDRP production by 1 percent

will increase the number of poor people by

1.3324 percent.

2. Reducing the allocation of funds for social

expenditure by 1 percent will increase the

number of poor people by 0.0022 percent.

Poverty alleviation programs through social

assistance allocation funds are the PKH, KIP,

KIS Program and the formation of the TP2NK.

3. Increasing inflation by 1 percent will increase the

number of poor people by 0.0084 percent.

4. Increasing the open unemployment rate by 1

percent will increase the number of poor people

by 0.0077 percent.

Based on the model equation, the coefficient of

determination is 72.14 percent, meaning the constant

GDRP growth in 2010, inflation rate, growth of fund

allocation for social expenditure and open

unemployment rate can explain the variability of the

growth of poor people in regencies/cities in South

Sumatra Province at 72.14 percent, while the

remaining 21.86 percent is explained by other

variables that have not included in the model.

Based on the significance value, it can be seen

that the Constant Price GDRP in 2010 has the least

significance value of the reference probability value

(0.05 percent), which is 0.0022. As well as from the

table of sectoral contribution to employment

absorption in South Sumatera Province, it can be

seen that the growth of GDRP variable has a high

contribution value in reducing the growth of the

number of poor people through sectoral employment

of GDRP. Thus, the connectivity between GDRP

and the number of poor people is negatively

directed.

From table 13, it is shown that the high

contribution of the mining and quarrying sector

(21.55 percent) percent is not in line with the

absorption of the workforce which only reaches 1.51

percent. It is not comparable to the magnitude of the

exploitation of natural resources carried out in the

production process.

On the other hand, the agricultural, forestry,

huntings and fisheries sector which contributed

18.78 percent turned out to be able to absorb 54.73

percent of the workforce. This indicates that there is

an imbalance in the formation of the primary sector

GDRP into the manufacturing industry sector and

public, social and private services sector.

transformation of the absorption of its workforce.

The permanent workforce is dominated by the

SEABC 2018 - 4th Sriwijaya Economics, Accounting, and Business Conference

634

agricultural, forestry, hunting and fisheries sectors

and is slowly transforming into the public, social and

private services sector as well as groceries, retails,

restaurants and hotels sector. This structural

imbalance contributes to economic inequality,

poverty, open unemployment and exploitation of

natural resources in South Sumatra Province.

Table 13: Ranks of Average Contribution of Constant

Price GDRP Year 2010 and Job Opportunity of Each

Sector in South Sumatera Province During 2009-2017

Sectoral Job

Opportunity in

South Sumatera

Province

Average

Contribution

of GDRP

Average

Contribution

of Job

Opportunity

Mining and

Quarrying

21,55%

1,31%

Agricultural,

Forestry, Hunting

and Fishery

18,78%

54,73%

Manufacturing

Industry

18,19%

5,24%

Groceries, Retails,

Restaurants and

Hotels

11,99%

15,91%

Construction

10,70%

4,08%

Public, Social and

Private Services

7,87%

12,97%

Transportation,

Warehousing and

Communication

5,32%

4,09%

Financial,

Insurance, Building

Rent, Land and

enterprises Services

5,31%

1,49%

Eelctricity, Gas and

Clean Water

0,29%

0,18%

Total

100,00%

Source : BPS Sumsel Dalan Angka (processed data)

In connection with the sectoral contribution of

GDRP towardthe relevant employment absorption

with the results of a study conducted by (Suryahadi,

Suryadarma, & Sumarto, 2009)the declinesof

poverty in Indonesia can be done by increasing the

acceleration of the growth of the agricultural sector

in the village and at the same time increasing the

growth of the service sector in the city. In addition,

there is a need to develop a credible industrial

relations system that can represent the interests of

both formal sector and marginalized workers and

thus contribute to alleviating the sense of

voicelessness and powerlessness of the poor (Islam,

2015).

5 CONCLUSION

1. Poor people in South Sumatra Province during

2008-2017 experienced an average annual

decline of 1.87 percent. However, in the same

period, several regencies/cities experienced an

increase in the number, namely: Ogan Komering

Ulu Regency, Palembang City, East OKU

Regency, Pagaralam City, and Prabumulih City.

2. Palembang City is the highest contributor of the

poor population in South Sumatra by 18.26

percent. This is as a result of economic structural

imbalances that occur due to changes in the

dominant economic sector in GDRP fromthe

sector of Agriculture, Forestry, Huntings and

Fisheries to the Manufacturing Industry sector

and the Groceries, Retails, Restaurants and

Hotels sector, but not followed by optimization

of labor absorption. Labors remain dominant in

the sectors of Agriculture, Forestry, Huntings

and Fisheries and are slowly transforming into

the Groceries, Retails, Restaurants and Hotels

sector and the public, social and private services

sector.

3. The structural inequality between employment

absorption and GDRP production value is one of

the causes of the low decline in the average

growth of open unemployment in South

Sumatera, which is only 1.31 percent, is not

comparable with the average annual growth of

GRDP of 2.89 percent, and thus, causing a low

decline in the average growth of the poor

population at only 1.87 percent. This structural

imbalance contributes to economic inequality,

poverty, open unemployment and exploitation of

natural resources in the province of South

Sumatra.

4. The results of the specification test with the

Common Effect Autoregressive (1) model with a

variability value of 72.14 percent variables that

affect poverty in South Sumatra Province are the

GDP growth variables, the growth of social

expenditure and the open unemployment rate.

5. Some policies that can be taken to accelerate the

decline in the number of poor people are :

encouraging investment and increasing the

productivity of sectors that absorb a lot of labor

such as agriculture, forestry, huntings and

fisheries; manufacturing industry (especially

agricultural products); andpublic, social and

private services sector. In addition, the

government can utilize demography bonuses by

improving the quality of human resources

through education and community empowerment

Economic Growth, Social Expenditure, Unemployment, and Inﬂation: The Impact on Poverty in South Sumatera

635

in supporting the processing of industrialization

of agricultural, forestry, hunting, fishery, mining

and quarrying products, as well as developing the

tourism services sector to encourage growth

synergies in the agro-industry sector, creative

economy and other services.

REFERENCES

Badan Pusat Statistik. (2017). Provinsi Sumatera Selatan

Dalam Angka 2017, 258. Retrieved from

http://sumsel.bps.go.id/index.php/publikasi/index?Pub

likasi%7B%25%7D5BtahunJudul%7B%25%7D5D=2

016%7B&%7DPublikasi%7B%25%7D5BkataKunci

%7B%25%7D5D=%7B&%7Dyt0=Tampilkan

Baltagi, B. H. (2005). Econometric Analysis of Panel Data

- Third Edition. John Wiley & Sons, 2008.

https://doi.org/10.1017/S0266466600006150

BPS. (2016). Indikator Kesejahteraan Rakyat 2016.

Publikasi BPS. https://doi.org/10.1007/s00717-009-

0309-x

Budiantara, I. N., Diana, R., & Darmesto, S. (2011).

Relationship Pattern of Poverty and Unemployement

in Indonesia. International Journal of Basic & Applied

Sciences, 11 No 6(December), 119–127.

Cahyat, A. (2004). Bagaimana kemiskinan diukur ?

Beberapa model penghitungan kemiskinan di

Indonesia. Governance Brief, 2, 1–8.

https://doi.org/10.17528/cifor/001641

Fajriyah, N., & Rahayu, S. P. (2016). Pemodelan Faktor-

Faktor yang Mempengaruhi Kemiskinan Kabupaten

Kota di Jawa Timur Menggunakan Regresi Data

Panel. Jurnal Sains Dan Seni ITS, Vol. 5, No(2337–

3520), (2301-928XPrint)D-45.

Ghozali, I. (2009). Aplikasi Multivariate dengan Program

SPSS edisi III. Semarang: Badan Penerbit UNDIP.

https://doi.org/10.1016/j.oooo.2017.11.003

International Monetary Fund. (2018). World Economic

Update, An Upadte of The Key WEO Projections.

Retrieved from

https://www.imf.org/en/Publications/WEO%0A/Issues

/2018/07/02/world-economic-%0Aoutlook-update-

july-2018%0A

Islam, I. (2015). Poverty , Employment and Wages : An

Indonesian Perspective. In ILO - JMHLW -

Government of Indonesia Seminar on Strengthening

Employment and Labour Market Policies for Poverty

Alleviation and Economic Recovery in East and

Southeast Asia.

Novalia, T. (2018). di Balik Angka Kemiskinan BPS, 1–8.

Odekon, M. (2015). The SAGE Encyclopedia of World

Poverty

Inflation.https://doi.org/http://dx.doi.org/10.4135/9781

483345727.n415

Saris, Alexander H. 2001. The Role of Agricultural in

Economic Developmentand Poverty Reduction: An

Empirical and Conceptual Foundation. Paper

preparedfor the rural Development Departement of

The World Bank. University of Athens, Athens

Sugema, I., Irawan, T., Adipurwanto, D., Holis, A., &

Bakhtiar, T. (2010). The Impact of Inflation on Rural

Poverty in Indonesia : an Econometrics Approach, (58

(2010)).

Suryahadi, A., Suryadarma, D., & Sumarto, S. (2009). The

effects of location and sectoral components of

economic growth on poverty: Evidence from

Indonesia. Journal of Development Economics.

https://doi.org/10.1016/j.jdeveco.2008.08.003

Suryawati, C. (2005). MEMAHAMI KEMISKINAN

SECARA MULTIDIMENSIONAL =

UNDERSTANDING MULTIDIMENSION OF

POVERTY. Jurnal Manajemen Pelayanan Kesehatan,

JMPK, Vol.

UNDP. (2018). Human Development Data (1990-2015) |

Human Development Reports.

https://doi.org/10.1007/s00338-005-0033-1

World Bank. (2005). INTRODUCTION TO POVERTY

ANALYSIS, 90288.

World Bank. (2018). June 2018 Indonesia Economic

Quarterly: Learning more, growing faster. Retrieved

from

http://documents.worldbank.org/Curated/En/30536152

8210283009/Pdf/126891-WP-PUBLIC-On-6-5-18.Pdf

Yacoub, Y. (2012). Pengaruh Tingkat Pengangguran

terhadap Tingkat Kemiskinan Kabupaten / Kota di

Provinsi Kalimantan Barat. Jurnal Ekonomi Sosial.

Zuhdiyati, N., & Kaluge, D. (2015). KEMISKINAN DI

INDONESIA SELAMA LIMA TAHUN TERAKHIR

( Studi Kasus Pada 33 Provinsi ), (Atalay), 27–31.

SEABC 2018 - 4th Sriwijaya Economics, Accounting, and Business Conference

636