The Convergence of Percapita Income in Central Sulawesi Province

This research aims at analyzing the influence of economic growth and human development index on the poverty rate of 13 regencies/cities in Central Sulawesi Province. The data were analyzed using panel data which employed Eviews version 11 software. The result of fixed effect model regression shows that economic growth and mean years of schooling negatively and significantly affect poverty rate, life expectancy and per capita expenditure positively and significantly affect poverty rate, and expected years of schooling does not affect poverty rate in regencies/cities of Central Sulawesi Province. To mitigate the poverty rate constant reference to pro-poor, pro-job, and pro-growth policies should be maintained and job vacancy need to be expanded and regency/city government could increase the realization of school and teacher facility spending budget allocation, administer nonformal education which can increase workers’ productivity, provide hospital, public health center, secondary public health center and health personnel facilities, public infrastructure and education assistance such as providing training on creativity in cooperatives and micro-, small, and medium-enterprises (MSMEs).


INTRODUCTION
Poverty has been a central issue in every country, including Indonesia.The poverty phenomenon serves as a real portrait between regions which can be depicted in terms of geographical location, income level, low human resources quality, poorly managed natural resources, rapidly increasing number of population, limited knowledge on developing economic sectors such as agriculture, processing industries, and mining and energy, annually increasing umemployment trend, and decreasing economic growth from one year to another.The poverty portrait can also be found in regencies/cities in Central Sulawesi Province.From 2018 to 2020, its poverty rate percentage mean ranges from 6% to 7%.In 2020, the highest poverty rate was found in Parigi Moutong Regency at 78.78% which was actually a regency with abundant natural resources and located close to the capital of Central Sulawesi Province.Meanwhile, the lowest poverty rate was found in Banggai Laut Regency at 11.09%.Based on the obtained data, the poverty rate in rural (regency) areas is higher than in urban areas from 2018 to 2020 with the number of population being higher in regency areas than in urban areas.This implies that many people in Central Sulawesi Province live under the poverty line.

Figure 1. Proverty Rate Of Regency/Cities
In Central Sulawesi Province Sources: BPS Sulteng, 2021 One of the indicators used to see whether or not a development and poverty rate alleviation are successful is economic growth level.This economic growth level is measured from the magnitude of regencies/cities' gross regional domestic product (GRDP).
The data (BPS Kabupaten Banggai, 2021) shows that the economic growth value in Central Sulawesi Province can be said as constantly decreasing as can be seen from the data in Morowali Utara Regency (-0.22%), followed by Banggai Laut Regency (-1.97%), and Morowali Regency (28.93%).Tambunan (2011) suggests that economic growth and poverty has a tight correlation, which at the start of a development the poverty rate tends to be high and at the end of the development the poverty rate will gradually decrease.In addition, Kuznets argues that the development process will be accompanied by substantial increase of inequality so that the poor will receive a portion of economic growth.The research conducted by Knowles indicates the negative and significant correlation between economic growth and poverty.
Just as the case with GRDP, human development index (HDI) can also influence every increase or decrease of poverty rate.Some components of human development index according to UNDP (2007) include life expectancy, expected years of schooling, mean years of schooling, and per capita expenditure.HDI level will affect population productivity.The lower the HDI the lower the population productivity level would be.In turn, this will result in low income.On the contrary, the higher the HDI the higher the population productivity would be, resulting in higher income.The problem is that the HDI in each region is varied.This makes HDI one of the factors which affects poverty (Maulana & Alamsyah, 2021).High-quality human resources is an important requisite for a development to run sustainably.Also, Sharp suggests that economically speaking the cause of poverty is the low quality of human resources.Jhingan (2012) mentions that shaping human capital is an attempt to obtain and increase the number of people with skill, education and experience which determine the economic development in a country.2018 18,38 33,73 17,03 41,75 54,28 31,8 25,4 83,66 27,78 29,78 11,97 19,4 25,26 2019 17,54 29,3 16,61 39,92 55,83 30,79 24,51 81,36 26,36 30,82 11,46 19,25 26,62 2020 16,7 28,16 16,5 40,02 53,17 30,51 22,93 78,76 25,43 30 11,09 18,38

LITERATURE REVIEW
Poverty is the inability of a person economically based on food and non-food which is calculated based on expenditure.To measure the level, the Poverty Line is used which consists of two categories, namely the Food Poverty Line (GKM) and the Non-Food Poverty Line (GKNM).
The Human Development Index (HDI) is a comparative measure of life expectancy, education, and living standards for all countries.HDI is used as an indicator to assess the quality aspects of development and to classify whether a country is a developed country, a developing country, or an underdeveloped country and also to measure the effect of economic policies on the quality of life.(Directorate of Statistical Analysis and Development of BPS, 2015).Economic growth is development economic activity that causes the production of goods and services to increase and the prosperity of society increases.(Sukirno, 2011)

RESEARCH METHODOLOGY
This research is a quantitative one of descriptive analysis nature, using a combination of time series and cross section secondary data or panel data.The research was conducted in 13 regencies/cities in Central Sulawesi Province for 2016-2020 period.The data in this research were collected by combining the secondary data which had been officially documented and published by BPS of Central Sulawesi Province and other relevant institutions.The variables used in this research are poverty rate (Y), economic growth (X1), life expectancy (X2), expected years of schooling (X3), mean years of schooling (X4), and per capita expenditure (X5)

Operating Definitions of Variables
The definitions of variables to be used in this research are as follows.1. Poverty (Y) means economic inability in fulfilling basic needs as measured from expenditure which is proxied with percentage of the poor population in regencies/cities.2. Economic growth (X1) which is proxied with gross regional domestic product on the basis of 2010 constant price in regencies/cities means the total added value generated by all business units in a region as expressed in billion rupiah unit.

Data Analysis Technique
The data in this research were analyzed using quantitative analysis.In its calculation, the statistic method uses Eviews Version 11 software.The multiple regression analysis used in this research is panel data regression model.The general form of panel data multiple regression equation is as found in Gujarati (2012) and Baltagi ( 2005

Model Determination
Two techniques were used to obtain the right model in estimating the panel data regression, namely:

Chow Test
Chow test is the test to compare the common effect model with the fixed effect model (Widarjono, 2009).The Chow test in this research used Eviews program.The hypotheses formed in the Chow test were as follows.H0: Common effect model H1: Fixed effect model H0 is rejected if the P-value is less than a value.On the contrary, H0 is accepted if the P-value is greater than a value.The a value used is 5%.

Hausman Test
This test compares the fixed effect model with the random effect model in determining the best model to be used as a panel data regression model (Gujarati, 2012).The Hausman test uses similar program as the Chow test, i.e., Eviews.The hypotheses formed in the Hausman test are as follows.H0: Random effect model H1: Fixed effect model H0 is rejected if the P-value is less than a value.Conversely, H0 is accepted if Pvalue is greater than a value.The a value used is 5%.

RESULT AND DISCUSSION Panel Data Model Selection Test: Chow Test
Chow test is used to select the most appropriate model between the common effect model and the fixed effect model.This test uses an assumption that if the probability value < 0.05, then the right research model is the fixed effect model and if the probability value > 0.05, then the right model is the common effect model.This test uses Eviews version 11, and the following results are obtained.
From the result of Chow test in the table above, a probability cross section random is obtained at 0.0000.The error rate used is 0.05.Thus, the obtained result indicates that the probability value is 0.0000 < 0.05, meaning that the appropriate model to be used is the fixed effect model.

Hausman Test
Hausman test is used to select the most appropriate model by comparing the fixed effect model with the random effect model.This test uses an assumption that if the probability value < 0.05, then the right model is the fixed effect model.And if the probability value > 0.05, then the most appropriate model is the random effect model.This test uses Eviews10 application and the following result is obtained.

Fixed Effect Model Regression Test
Based on the panel regression estimation using the fixed effect model, the following result is obtained.
The regression equation of the fixed effect model is as follows. Y

Coefficient Determination
The coefficient value is determined to discover the contribution that independent variables (X) can make in affecting the dependent variable (Y) as measured by percentage.Based on the result of test which has been conducted, the R-Squared value obtained is 0.558275.This means that around 55.83% of poverty can be explained by independent variables in this research, namely human development index and economic growth.Meanwhile, the remaining 44.17% is explained by other variables beyond the model or not included in this research.

Hypothesis Testing of t-Statistic Test
The t-statistic test is conducted to discover the partial influence of independent variables on the dependent variable by comparing the t-statistic with the t-table value.
To figure out the t-table value, α=0.05 is sought for with a degree of freedom (df) = n-k-1.With a significance test of 0.05, the t-table value is obtained at 1.895.Therefore, the results of hypothesis testing are as follows: 1. Economic growth, its t-score is -4.764181,where t-score ≥ t- 3. Life expectancy (X2) means the average estimation of number of years spent by an individual since birth.4. Expected years of schooling (X3) means the duration of schooling (in years) expected to be spent by children at certain ages in the future.5. Mean years of schooling (X4) means the number of years used by a population to complete a formal education.6. Per capita expenditure (X5) means the per capita expenditure as determined from the value of per capita expenditure and purchasing power parity of a community in rupiah unit.
26,89human development index of City of Palu is the highest at 80.91% increases to 81.5% and this is relatively high since it is within the 70-80 percent range.

Table 2 .
Hausman table (-4.764181 ≥ 1.895), meaning that partially there is negative and significant influence between economic growth and poverty rate in regencies/cities in Central Sulawesi Province.2. Life expectancy, its t-score is 2.795778, where t-score ≥ t-table (2.795778≥ 1.895), meaning that partially there is significant influence between life expectancy and poverty rate in regencies/cities in Central Sulawesi Province.3. Expected years of schooling, its t-score is 0.448695, where t-score ≤ t-table (0.448695≤ 1.895), meaning that partially there is no significant influence between