S udies Managemen and Finance Economics, o Jou nal
0504-2644 (online): ISSN 0490,-2644 (p in ): ISSN
5202 No embe 11 Issue 80 Volume
8.317 Fac o : Impac ,30-i11-10.47191/je ms/ 8 DOI: A icle
4247-7414 :No Page
JEFMS, Volume 08 Issue 11 No embe 2025 www.ije m.co.in Page 7414
Analysis o Fac o s A ec ing he Numbe o Poo People in Cen al Sulawesi
Abd. Rangga1, MA Dji imu2, Edhy Taqwa3, Yunus Sading4, Musdayan i5
1,2,3,4,5Depa men o Economics and De elopmen S udies, Facul y o Economics and Business, Uni e si as Tadulako, Palu,
Indonesia
ABSTRACT: Po e y is a pe sis en challenge cons aining egional de elopmen in Cen al Sulawesi. The e o e, his s udy
in ends o analyze he main de e minan s a ec ing he numbe o poo people in he ci ies and dis ic s o Cen al Sulawesi om
2015 o 2024. The analysis was conduc ed using h ee econome ic app oaches, namely Panel Da a Reg ession, Geog aphically
Weigh ed Panel Reg ession (GWPR), and Vec o E o Co ec ion Model (VECM). The esul s showed ha REM (Random E ec
Model) was he mos app op ia e speci ica ion, sugges ing li e expec ancy and g oss egional domes ic p oduc (GRDP) pe
capi a ha e a signi ican and nega i e e ec on po e y. GWPR model cap u ed spa ial he e ogenei y mo e e ec i ely,
demons a ing ha he impac o GRDP and educa ion a ies ac oss egions. Meanwhile, VECM esul s con i med bo h sho - un
and long- un ela ionships among a iables, indica ing ha economic, educa ional, and heal h ac o s join ly in luence po e y
educ ion. These esul s emphasize he u gency o economic g ow h, human capi al de elopmen , and employmen
oppo uni ies in educing po e y sus ainably ac oss Cen al Sulawesi. In conclusion, policymake s should design a ge ed
in e en ions in po e y alle ia ion ac oss Cen al Sulawesi.
KEYWORDS: Cen al Sulawesi, Geog aphically Weigh ed Panel Reg ession, Panel Da a Reg ession, Po e y, Vec o E o
Co ec ion Model
1. INTRODUCTION
De elopmen is ans o ma i e p ocess a ge ed a imp o ing he social sys em owa d a mo e ad anced and sus ainable
condi ion. The concep o sus ainable de elopmen unde sco es signi ican mul idimensional po e y educ ion and imp o emen
in quali y o li e (Toda o & Smi h, 2020). Howe e , de elopmen o en encoun e s complex challenges in p ac ice, one o which is
po e y.
Acco ding o S a is ics Indonesia (BPS, 2024), Cen al Sulawesi P o ince is cha ac e ized by a luc ua ing po e y si ua ion,
bu has shown a downwa d end in ecen imes. Mo e speci ically, he numbe o poo people eached 398.73 housand in 2020,
o 12.92% o he o al popula ion, and by 2024, he e was a decline o 379.76 housand, o 11.04%. Al hough he po e y a e has
dec eased, he alue emains abo e he a ge se in he Regional Medium-Te m De elopmen Plan 2021–2026, which aimed o
a educ ion o 7.65% by 2024. The a e age annual decline in he po e y a e has also been ela i ely slow, es ima ed a 0.12%
pe yea .
F om 2020 o 2024, Cen al Sulawesi emained o ha e he second-highes po e y a e on Sulawesi Island (BPS, 2024).
This condi ion places a conside able bu den on bo h local go e nmen and egional economic de elopmen , speci ically when
compa ed wi h o he p o inces. The high po e y a e ac oss he p o ince is la gely shaped by he dispa i ies among he dis ic s
and ci ies. Po e y is no only a inancial issue bu also e lec s mul idimensional dep i a ion such as poo heal h, limi ed
educa ion, exposu e o ulne abili y, inabili y o exe cise au onomy, and lack o access o decision-making powe .
BPS eco ded wide dispa i ies in po e y le els among dis ic s in Cen al Sulawesi. In 2024, Pa igi Mou ong had he highes
numbe o poo people (74.68 housand), while Banggai Lau had he ewes (11.16 housand). This pe sis en in e -dis ic
inequali y indica es ha po e y emains a s uc u al issue ac oss he p o ince. The abili y o local go e nmen s o use and
op imize a ailable esou ces becomes essen ial o educe po e y le els. In gene al, egional economic de elopmen
undamen ally depends on local esou ce u iliza ion, whe e economic po en ial se es as he ounda ion o egional de elopmen
(A mida & Mu nining yas, 2018).
U ban a eas, such as Palu Ci y, end o ha e lowe po e y le els compa ed o u al egions, including Pa igi Mou ong and
Donggala (BPS, 2024). The concen a ion o economic ac i i ies and educa ional ins i u ions in u ban cen e s con ibu es o lowe
Analysis o Fac o s A ec ing he Numbe o Poo People in Cen al Sulawesi
JEFMS, Volume 08 Issue 11 No embe 2025 www.ije m.co.in Page 7415
po e y incidence, as hese ac o s c ea e be e employmen oppo uni ies and imp o ed access o essen ial se ices ha
enhance li ing s anda ds.
Heal h plays a c ucial ole as an in es men in imp o ing p oduc i i y and he quali y o human esou ces. I is gene ally
conside ed a ounda ion o aising li ing s anda ds. Heal h s a us di ec ly in luences well-being and is closely associa ed wi h
po e y, which can be obse ed h ough indica o s such as li e expec ancy (Adhi ya, Pa abawa, & Kencana, 2022). Be ween 2020
and 2024, Palu Ci y eco ded he highes li e expec ancy, ollowed by Poso Dis ic , due o he concen a ion o heal hca e and
educa ional acili ies. Con e sely, Pa igi Mou ong and Banggai Lau we e epo ed o ha e lowe li e expec ancy a es (BPS, 2024).
Educa ion is equally c ucial in he e o o educe po e y, as e lec ed by he a e age leng h o schooling. Thahi , Semmaila, and
A ah (2021) epo ed ha one majo cause o po e y is he low quali y o human capi al due o limi ed educa ion. Highe
educa ional a ainmen enhances he capaci y o indi iduals o secu e be e employmen and income, he eby educing po e y.
The ypical du a ion o schooling is a c ucial measu e o educa ional success and indica es he o al yea s indi iduals spend in
o mal educa ion. Acco ding o BPS (2024), om 2020 o 2024, Palu Ci y and Poso Dis ic consis en ly eco ded he highes
a e age yea s o schooling, while Pa igi Mou ong and Donggala eco ded he lowes . This educa ional dispa i y signi ican ly
in luences po e y le els, as egions wi h lowe schooling a e ages ace limi ed job oppo uni ies. Toda o in Hasanah & P iyadi
(2021) also men ioned ha educa ion is a key de e minan o income and economic mobili y.
Unemploymen is conside ed ano he c ucial ac o a ec ing po e y in Cen al Sulawesi. Open unemploymen e e s o
he p opo ion o he labo o ce ac i ely in sea ch bu unable o ob ain employmen . Despi e ha ing a ela i ely high educa ion
le el, Palu Ci y consis en ly eco ded he highes open unemploymen a e (TPT) be ween 2020 and 2024. The a e declined om
8.38% in 2020 o 5.63% in 2024 (BPS, 2024). Ac oss dis ic s, he e is a pa adox o “educa ed unemploymen ,” whe e egions wi h
highe schooling a e ages expe ience highe unemploymen a es. This phenomenon may be a ibu ed o a misma ch be ween
he skills o educa ed wo ke s and a ailable job oppo uni ies, leading o an excess supply o educa ed labo (BPS, 2024)
As s a ed by Balasub amanian, Bu chi, and Male ba (2023), a socie y is conside ed o ha e a decen s anda d o li ing when
i can independen ly ul ill basic needs. On he o he hand, when indi iduals ail o mee hese needs, po e y pe sis s. Economis s
and policymake s o en use GRDP (G oss Regional Domes ic P oduc ) pe capi a as a p oxy o measu ing wel a e le els. Da a om
BPS (2024) showed ha Mo owali Dis ic had he highes GRDP pe capi a, ollowed by No h Mo owali, while Banggai Islands
had he lowes in Cen al Sulawesi. Ch is iani and Nainupu (2021) also emphasized ha GRDP pe capi a is a mo e accu a e
indica o o economic g ow h compa ed o o al GRDP, as i be e e lec s he wel a e o local popula ions.
The complex in e ac ion o ac o s such as heal h, educa ion, unemploymen , and egional economic pe o mance
unde sco es he need o an in-dep h analysis o he simul aneous and pa ial e ec s on po e y ac oss dis ic s and ci ies in
Cen al Sulawesi. This s udy inco po a ed bo h spa ial and empo al dimensions, acknowledging a ia ions in egional
cha ac e is ics and esou ce po en ial. By examining he de e minan s o po e y in Cen al Sulawesi, he esul p o ides aluable
insigh s o policymake s o design a ge ed and e idence-based po e y alle ia ion s a egies. I also con ibu es o he ongoing
academic discou se on egional po e y educ ion by in eg a ing economic, social, and spa ial pe spec i es.
2. LITERATURE REVIEW
Po e y is de ined as a low s anda d o li ing cha ac e ized by a lack o inancial and ma e ial esou ces (Ha dian o, 2021). I is
widely ecognized as an undesi able condi ion ha gene a es a ious socio-economic p oblems, namely de e io a ing heal h,
unemploymen , and declining human esou ce quali y (Is yawan, 2025). The capabili y app oach in oduced by Ama ya Sen
concep ualizes po e y no me ely as a lack o income bu a dep i a ion o undamen al human capabili ies o lead a aluable li e.
These capabili ies comp ise access o educa ion, heal hca e, decen employmen , and a sus ainable en i onmen . Va iables such
as mean yea s o schooling, li e expec ancy, and GRDP pe -capi a a e dimensions o human capabili y ha de e mine oppo uni ies
o escape po e y. As s a ed in a p e ious s udy, imp o ing he quali y o educa ion, heal h, employmen , and access o basic
se ices is an essen ial ins umen o po e y alle ia ion (Lundahl, Rauhu , & Ha i, 2021).
Based on he Theo y o Human Capi al, heal h and educa ion a e long- e m in es men s ha imp o e labo p oduc i i y.
The a e age yea s o schooling (RLS) e lec s he quali y o educa ion, while UHH (li e expec ancy) unc ions as an indica o o
heal h. An inc ease in hese wo indica o s s eng hens he compe i i eness o he labo o ce and con ibu es o po e y
educ ion. A p e ious s udy by Fi daus, Dawood, and Ab a (2021) ound ha educa ion signi ican ly educed po e y le els in
Indonesia du ing he 2016–2020 pe iod. The co ela ion be ween educa ion and po e y is pa icula ly s ong, as educa ion gi es
people he skills and knowledge necessa y o economic ad ancemen (Hasanah, Syapa uddin, & Rosmeli, 2021). Li e expec ancy,
which e lec s a highe quali y o li e and be e heal h condi ions, is also associa ed wi h lowe po e y le els ( Nias u i &
Se yowa i, 2024).
Analysis o Fac o s A ec ing he Numbe o Poo People in Cen al Sulawesi
JEFMS, Volume 08 Issue 11 No embe 2025 www.ije m.co.in Page 7416
Unemploymen se es as a key indica o o labo ma ke condi ions. In gene al, highe unemploymen a es ypically lead o an
inc ease in po e y due o income loss and inadequa e social p o ec ion (P adip a & Dewi, 2020). The e o e, TPT is conside ed a
c i ical ac o in luencing po e y. The in o mal sec o plays a s a egic ole in imp o ing he income o low-wage wo ke s and
educing po e y by inc easing ea nings abo e he po e y h eshold.
The ickle-down e ec heo y assumes ha economic g ow h will e en ually bene i lowe -income g oups. Howe e ,
s uc u alis pe spec i es emphasized ha g ow h does no au oma ically educe po e y due o unequal income dis ibu ion.
The e o e, GRDP pe capi a becomes an essen ial indica o o assess he ex en o which economic g ow h con ibu es o po e y
educ ion. Sapu i, Se yonug oho, and Ha ono (2024) epo ed ha economic g ow h accompanied by equi able access leads o
po e y educ ion a he household le el.
Empi ical s udies ha e p o ided s ong suppo o he ela ionships among hese a iables and po e y. Fi daus, Dawood, and
Ab a (2021) con i med ha educa ion and heal h play a signi ican ole in alle ia ing po e y in Indonesia. Simila ly, Sapu i,
Se yonug oho, and Ha ono (2024) men ioned ha inclusi e economic g ow h p omo es po e y educ ion, al hough he e ec s
depend on he accessibili y and dis ibu ion o esou ces. Acco ding o heo e ical and empi ical insigh s, his s udy in eg a es he
capabili y app oach as he p ima y amewo k, suppo ed by human capi al, economic g ow h, and s uc u alis heo ies. The
in eg a ion o e s a comp ehensi e analy ical ounda ion o examining how li e expec ancy, a e age yea s o schooling, TPT, and
GRDP pe capi a a ec po e y ac oss ci ies and dis ic s in Cen al Sulawesi.
The e iewed heo ies and empi ical esul s unde sco e he need o examine po e y de e minan s wi h g ea e a en ion
o spa ial and empo al a ia ions. Mos p e ious s udies ha e analyzed po e y de e minan s a he na ional o p o incial le el
using con en ional panel da a me hods (P adip a & Dewi, 2020; Fi daus e al., 2021). Howe e , limi ed a en ion has been paid o
he spa ial he e ogenei y o po e y de e minan s ac oss dis ic s in Cen al Sulawesi using he Geog aphically Weigh ed Panel
Reg ession (GWPR) app oach. The e o e, his s udy aims o in eg a e panel da a, GWPR, and VECM models o comp ehensi ely
cap u e bo h spa ial and dynamic aspec s o po e y de e minan s.
3. STUDY METHODS
Seconda y da a used in he o m o panel da a, comp ising c oss-sec ional obse a ions om 12 egencies and 1 ci y in Cen al
Sulawesi P o ince, combined wi h ime-se ies da a co e ing he pe iod om 2015 o 2024. Da a p ocessing was conduc ed using
EViews 13, R S udio, and Mic oso Excel. Mo eo e , a quan i a i e app oach was used, including seconda y da a collec ed om
he BPS (Cen al Bu eau o S a is ics) o analyze he de e minan s o po e y ac oss egencies and ci ies in Cen al Sulawesi. This
app oach allows o a comp ehensi e analysis ha cap u es bo h empo al and spa ial a ia ions in he selec ed a iables. The
independen a iables, including UHH, RLS, TPT, and GRDP pe capi awe e chosen due o heo e ical conside a ions and empi ical
e idence indica ing a connec ion o po e y le els.
Th ee analy ical models we e u ilized o analyze he in luence o he a iables on he numbe o poo people, namely he
Panel Da a Reg ession, Geog aphically Weigh ed Panel Reg ession (GWPR) model, and Vec o E o Co ec ion Model (VECM).
The andom e ec s app oach was applied o ensu e mo e obus and eliable pa ame e es ima ion.
The combina ion o Panel Da a Reg ession, GWPR, and VECM was used o ensu e a comp ehensi e examina ion o po e y
de e minan s. The panel eg ession p o ides a global es ima ion o he ela ionships among a iables, GWPR cap u es spa ial
he e ogenei y ac oss egencies and municipali ies, and VECM iden i ies bo h sho - un and long- un dynamics. This in eg a ed
app oach enables a s onge unde s anding o he spa ial– empo al ac o s in luencing po e y in Cen al Sulawesi.
The gene al panel eg ession model can be speci ied as ollows:
Ln JPMi = β0 + β1UHHi + ꞵ2RLSi + ꞵ3TPTi + ꞵ5Ln PDRBpki + 𝑒i
GWPR model is o mula ed as:
Ln JPMi = ꞵ0(𝜇𝑖𝑡, 𝜗𝑖𝑡) + ꞵ1(𝜇𝑖𝑡, 𝜗𝑖𝑡)UHHi + ꞵ2(𝜇𝑖𝑡, 𝜗𝑖𝑡)RLSi + ꞵ3(𝜇𝑖𝑡, 𝜗𝑖𝑡)TPTi + ꞵ5(𝜇𝑖𝑡, 𝜗𝑖𝑡)Ln PDRBpki + 𝑒i
Whe e:
Symbol
Desc ip ion
JPM
:
Numbe o poo people
UHH
:
Li e expec ancy
RLS
:
A e age yea s o schooling
TPT
:
Open unemploymen a e
PDRBpk
:
G oss Regional Domes ic P oduc pe capi a
ꞵ0
:
Cons an
ꞵ1ꞵ2ꞵ3ꞵ4ꞵ5ꞵ6
:
Reg ession coe icien s
Analysis o Fac o s A ec ing he Numbe o Poo People in Cen al Sulawesi
JEFMS, Volume 08 Issue 11 No embe 2025 www.ije m.co.in Page 7417
Symbol
Desc ip ion
𝑒
:
E o e m
i
Regency/Ci y
:
Yea
(𝜇𝑖𝑡 ,𝜗𝑖𝑡)
Geog aphic coo dina es (la i ude and longi ude)
4. RESULTS AND DISCUSSION
The selec ion o he mos app op ia e model o examine he aspec s a ec ing he o al o po e y in he egencies/municipali ies
o Cen al Sulawesi P o ince was ca ied ou h ough a sequence o es s, namely Chow, Hausman, and Lag ange Mul iplie . Chow
es showed a ejec ion o H₀ wi h a p alue o 0.000, which is less han he 5% signi icance le el (α = 0.05). This sugges s ha FEM
(Fixed E ec Model) is mo e app op ia e han CEM (Common E ec Model). Hausman es ailed o ejec H₀ wi h a p- alue o
0.7588, which is g ea e han α, sugges ing REM (Random E ec Model) is mo e sui able han FEM. Lag ange Mul iplie es
ejec ed H₀ wi h a p- alue o 0.000, smalle han α, implying ha REM is p e e able o CEM. Based on he h ee es s, REM is he
mos sui able model o apply. These indings align wi h he s udies ca ied ou by He mawan, Sukmawa y, and Les ia (2022) as
well as Hidaya i, F ansiska, and Agwil (2023).
Table 1. Es ima ion Resul s o Random E ec Model (REM) – Panel Reg ession
Va iable
Coe icien
S anda d
E o
-S a is ic
p- alue
C
12,5647
0,50687
24,78876
0,0000
***
UHH
-0,0237
0,00887
-2,67028
0,0086
**
RLS
-0,0175
0,02425
-0,72314
0,4709
TPT
0,0027
0,00504
0,53518
0,5935
PDRBpk
-0,0573
0,02213
-2,58852
0,0108
**
R-Squa ed: 0,404467
F-s a is ic
: 21,22397
P- alue
: 0,0000
No es
:
*** signi ican a 1%, ** signi ican a 5%, * signi ican a
10%
Sou ce
:
Au ho ’s calcula ion, 2025
Pa ially, he a iables o UHH and PDRB pe capi a had signi ican and nega i e e ec s on le el o po e y, wi h coe icien s o
−0.0237 and −0.0573, espec i ely, a he 5% signi icance le el. He mawan, Sukmawa y, and Les ia (2022) also ound ha
inc easing li e expec ancy can educe po e y le els in Sou h Kaliman an P o ince. Simila ly, Az iyansyah (2022) epo ed ha an
inc ease in PDRB pe capi a signi ican ly con ibu es o educing po e y in Cen al Ja a. This sugges s ha imp o emen s in heal h
and economic g ow h e ec i ely educe po e y.
RLS a iable shows a nega i e di ec ion, implying ha he highe he educa ion le els, he lowe he po e y. Howe e , he
e ec was s a is ically insigni ican , indica ing ha he con ibu ion o educa ion o po e y educ ion is no op imal. This esul is
consis en wi h Hidaya i, F ansiska, and Agwil (2023), who ound ha educa ion signi ican ly a ec s po e y in Sou h Suma a.
TPT a iable had a posi i e bu insigni ican coe icien , implying ha an inc ease in unemploymen ends o aise po e y, hough
no su icien ly s ong o be signi ican . P e ious s udies also sugges ed ha unemploymen in luences po e y le els, hough he
deg ee o signi icance a ies (Kasus e al., 2022; He mawan, Sukmawa y, and Les ia, 2022).
The es ima ion esul s show ha R-squa ed sco e o 0.404467 shows ha 40% o he a ia ion in po e y le els could be
accoun ed o by he independen a iables, while he emaining 60% is in luenced by o he ac o s ou side he model. The F- es
was signi ican a he 1% le el, indica ing ha he independen a iables join ly a ec po e y le els in Cen al Sulawesi
egencies/municipali ies du ing 2015–2024. In gene al, he eg ession model is conside ed easible o u he spa ial modeling
using GWPR (Azkiya, Ang aini, & Anisa, 2024; Ka ika, Su i, & Kholijah, 2020; Ninda, Dinayan i, & Be liana, 2022).
The B eusch–Pagan c oss-sec ion es yielded a p alue o 0.0000 (>0.05), leading o he ejec ion o H₀ a he 5% le el. This
indica es he p esence o spa ial he e ogenei y in he panel eg ession model. The esul con i ms ha he cha ac e is ics o
po e y ac oss egencies/municipali ies in Cen al Sulawesi om 2015–2024 demons a e spa ial dependence, consis en wi h
Ln JPMi =
12,56 – 0,02 UHHi – 0,18 RLSi + 0,003 TPTi + 0,06 LnPDRBpki + 𝑒𝑖𝑡
(24,79)*** (− 2,67)** (− 0,7hsd2) (0,54) (− 2,59)**
Analysis o Fac o s A ec ing he Numbe o Poo People in Cen al Sulawesi
JEFMS, Volume 08 Issue 11 No embe 2025 www.ije m.co.in Page 7418
Ninda, Dinayan i, & Be liana (2022). To add ess his issue, a local modeling app oach conside ing spa ial di e si y, namely GWPR,
was applied.
Table 2. Ke nel Func ion Selec ion
Table 2 shows ha he bes GWPR model uses Adap i e Exponen ial ke nel, cha ac e ized by he lowes AIC and CV alues along
wi h he highes R-squa ed.
Table 3. Es ima ed GWPR Coe icien s by Regency/Municipali y in Cen al Sulawesi
Regency/
Municipali y
Coe icien and P ob
R2
C
UHH
RLS
TPT
PDRBpk
Banggai
1,74e-17
0,001
-0,175
0,015
0,009
0,825
0,906
0,000
0,008
0,605
Banggai Kepulauan
-9,79e-16
-0,008
-0,182
0,026
0,024
0,845
0,490
0,000
0,002
0,323
Banggai Lau
-1,10e-15
0,014
-0,137
0,020
0,004
0,798
0,211
0,000
0,002
0,817
Buol
-3,45e-16
-0,028
0,029
0,001
-0,087
0,617
0,001
0,339
0,981
0,000
Donggala
3,87e-16
-0,031
0,047
-0,002
-0,170
0,604
0,033
0,095
0,772
0,097
Mo owali
1,22e-16
-0,026
-0,034
0,010
-0,037
0,783
0,002
0,094
0,032
0,031
Mo owali U a a
7,37e-17
-0,031
-0,005
0,006
-0,058
0,783
0,006
0,847
0,242
0,003
Pa igi Mou ong
5,18e-16
-0,014
-0,014
0,001
-0,140
0,637
0,259
0,597
0,833
0,026
Poso
2,77e-16
-0,023
-0,004
0,002
-0,069
0,706
0,010
0,858
0,560
0,001
Sigi
-2,80e-16
-0,026
0,036
-0,015
-0,148
0,597
0,009
0,141
0,018
0,001
Tojo Una-Una
-4,60e-17
-0,018
-0,040
0,004
-0,051
0,743
0,039
0,048
0,327
0,005
Toli-Toli
-6,38e-16
-0,029
0,047
-0,004
-0,110
0,528
0,001
0,042
0,416
0,000
Palu
6,27e-16
-0,025
0,023
0,017
0,006
0,543
0,106
0,436
0,063
0,964
Sou ce: Au ho ’s calcula ion, 2025
Table 3 shows ha he e ec s o independen a iables on po e y di e by loca ion. Some a iables show nega i e e ec s on
po e y, while o he s demons a e posi i e e ec s depending on egional cha ac e is ics. GWPR analysis con i med spa ial
a ia ion ac oss egencies/municipali ies, wi h each ha ing dis inc pa ame e alues. In gene al, he independen a iables
signi ican ly a ec po e y le els in Cen al Sulawesi.
Fo ins ance, he model es ima ed o Palu Ci y is as ollows:
Model Ka nel
CV
AIC
R-Squa e
Adap i e Bisqua e
0,1257
-557,7708
0,7497
Adap i e Gaussian
0,1239
-553,3923
0,7223
Adap i e Exponen ial
0,1204
-559,4871
0,7384
Sou ce: Au ho ’s calcula ion, 2025
Ln JPMi =
12,56 – 0,02 UHHi – 0,18 RLSi + 0,003 TPTi + 0,06 LnPDRBpki + 𝑒𝑖𝑡
(24,79)*** (− 2,67)** (− 0,72) (0,54) (− 2,59)**
Analysis o Fac o s A ec ing he Numbe o Poo People in Cen al Sulawesi
JEFMS, Volume 08 Issue 11 No embe 2025 www.ije m.co.in Page 7419
Table 4. Sho - and Long-Run Es ima ion o VECM (Panel Da a)
Va iable
Coe icien
S anda d
E o
-s a is ik
- abel(1,979)
Sig.
ECT (JMP)
-0,241868
0,12142
1,99192
**
ΔJPM(-1)
-0,618846
0,15847
3,90514
**
ΔJPM(-2)
-0,334913
0,16048
2,08693
**
ΔUHH(-2)
-0,087246
0,02186
3,99049
**
ΔUHH(-3)
0,055965
0,02619
2,13672
**
ΔTPT(-1)
0,035537
0,01406
2,52810
**
ΔTPT(-3)
0,014387
0,00687
2,09475
**
R-Squa ed: 0,7964
F-s a is ic: 11,73541
**
AIC: -3,66
ECT (TPT)
-15,14709
2,04933
-7,39124
**
ΔJPM(-1)
15,81339
2,67455
5,91254
**
ΔJPM(-2)
10,53571
2,70850
3,88987
**
ΔJPM(-3)
6,793848
2,74435
2,47557
**
ΔUHH(-2)
1,136999
0,36900
3,08132
**
ΔTPT(-1)
0,802621
0,23724
3,38315
**
ΔTPT(-2)
0,445630
0,17217
2,58831
**
ΔTPT(-3)
0,350439
0,11592
3,02317
**
ΔPDRB(-1)
-3,327579
1,36882
2,43098
**
R-Squa ed: 0,8447
F-s a is ic: 16,32014
**
AIC: 1,99
No es
:
** signi ican a 5%
Sou ce
:
Au ho ’s calcula ion, 2025
VECM es ima ion esul s show a long- un ela ionship among po e y, UHH, RLS, TPT, and PDRB pe capi a. The nega i e and
signi ican ECT coe icien (−0.241868) con i ms an e o -co ec ion mechanism, indica ing a speed o adjus men owa d long-
un equilib ium o 24.18% pe pe iod. Changes in po e y le els in he i s and second lags nega i ely and signi ican ly a ec
cu en po e y le els. This implies ha inc eases in pas po e y end o co ela e wi h a dec ease in he po e y le els.
Changes in UHH demons a e mixed e ec s ac oss pe iods, wi h a nega i e e ec in he second lag bu posi i e in he hi d.
This pa e n sugges s ha imp o emen s in public heal h ha e dynamic e ec s on po e y. Al hough highe li e expec ancy
educes po e y in he sho un, i may inc ease economic p essu e in he longe e m, due o highe li ing and heal hca e cos s.
Changes in TPT in he i s and hi d lags show posi i e and signi ican e ec s on po e y, indica ing ha ising unemploymen
leads o highe po e y. R² alue o 0.7964 and a signi ican F-s a is ic demons a e s ong explana o y powe o he model o
po e y dynamics in Cen al Sulawesi om 2015 o 2024.
The unemploymen equa ion shows a long- un ela ionship wi h he o he a iables. The ECT coe icien o −15.14709,
which is signi ican a he 5% le el, indica es a high adjus men speed owa d long- un equilib ium. Changes in po e y in he i s
o hi d lags ha e signi ican and posi i e e ec s on unemploymen , implying ha highe pas po e y le els end o inc ease
unemploymen in subsequen pe iods. UHH a iable in he second lag has a posi i e and signi ican e ec on unemploymen ,
indica ing ha imp o emen s in public heal h po en ially in luence labo ma ke dynamics h ough inc eased pa icipa ion. PDRB
pe capi a a iable has a signi ican and nega i e e ec on unemploymen , sugges ing ha highe economic g ow h helps educe
unemploymen le els.
VECM es ima ion esul s show a mu ual ela ionship be ween po e y and unemploymen . Mo e impo an ly,
unemploymen se es as an in e ening a iable ansmi ing he e ec s o heal h (UHH), educa ion (RLS), and economic g ow h
(PDRB pe capi a) on po e y. Imp o emen s in human capi al and economic g ow h do no di ec ly educe po e y, bu a he ,
pa o he e ec ope a es h ough labo ma ke imp o emen s. These esul s indica e he complex ansmission mechanisms
among egional economic a iables, whe e communi y wel a e is de e mined by s uc u al in e ac ions be ween heal h,
educa ion, employmen oppo uni ies, and income.
Table 5. Compa ison o Global and GWPR Models
Model
RSS
AIC
R-Squa e
Analysis o Fac o s A ec ing he Numbe o Poo People in Cen al Sulawesi
JEFMS, Volume 08 Issue 11 No embe 2025 www.ije m.co.in Page 7420
REM
0,2147
-
820,7821
0,4044
VECM
0,0579
-
3,6612
0.7964
GWPR Model
0,0896
-
293,9880
0,7383
Sou ce: Au ho ’s calcula ion, 2025
The model compa ison shows ha VECM model ou pe o med REM and GWPR, wi h he lowes RSS and highes R-squa ed alue
(79.64%). Meanwhile, REM and GWPR explained 40.44% and 73.83% o he dependen a iable, espec i ely. REM had he
smalles AIC, bu VECM model demons a ed supe io explana o y powe and p edic i e accu acy.
REM es ima ion esul s show ha UHH signi ican ly and nega i ely impac s po e y, wi h a coe icien o –0.0237 (p- alue
0.0086). This implies ha a 1% inc ease in UHH educes po e y by 0.0237%, p o ided o he a iables emain cons an .
Theo e ically, imp o ed public heal h enhances indi iduals abili y o wo k p oduc i ely, con ibu ing o po e y educ ion. Rambe
e al. (2024) as well as Thahi , Semmaila, and A ah (2021) ound ha be e heal h indica o s we e nega i ely associa ed wi h
po e y.
GWPR model shows ha he e ec o UHH on po e y was spa ially une en ac oss egions. In egions such as Tojo Una-
Una, Toli-Toli, Poso, No h Mo owali, Mo owali, Sigi, Buol, and Donggala, UHH coe icien was nega i e and signi ican . This
ein o ces he REM esul s showing ha public heal h plays a i al ole in alle ia ing po e y. P adip a and Dewi (2020) also s a ed
ha inc easing he quali y o li e and communi y heal h con ibu es o po e y educ ion. Howe e , in se e al egions such as
Banggai and Banggai Lau , UHH coe icien was posi i e bu insigni ican , indica ing ha highe li e expec ancy does no always
immedia ely educe po e y. The dispa i y in esul s may be due o ac o s such as labo p oduc i i y o unequal access o public
se ices.
VECM model u he added a dynamic pe spec i e. In he sho un, changes in UHH a lag-2 had a signi ican nega i e
impac , while lag-3 had a signi ican posi i e e ec on po e y. The pa e n indica es a complex dynamic, whe e imp o emen s in
public heal h in he pas 2 yea s appea o supp ess po e y, bu in he longe e m, he e ec may e e se. This could be
associa ed wi h highe li ing and heal hca e cos s as UHH inc eases.
In gene al, he h ee models p o ide complemen a y insigh s. REM con i ms a signi ican a e age nega i e ela ionship,
GWPR unde sco es spa ial a ia ions ac oss egions, and VECM cap u es sho - e m luc ua ions. Enhancing public heal h h ough
equi able access o heal hca e emains a key s a egy o educing po e y. Howe e , policy in e en ions should accoun o local
condi ions and he long- e m implica ions o heal h imp o emen s.
REM es ima ion shows ha RLS had a nega i e bu s a is ically insigni ican coe icien (–0.0175; p- alue 0.4709). This
implies ha , on a e age, highe educa ion le els a e associa ed wi h lowe po e y, bu he e ec is no su icien ly s ong
empi ically a he agg ega e le el in Cen al Sulawesi. The esul may be due o une en educa ional quali y, implying ha addi ional
yea s o schooling a e no necessa ily accompanied by imp o ed skills o income.
GWPR esul s demons a ed in-dep h spa ial a ia ion. In egions such as Buol, Donggala, Sigi, Toli-Toli, and Palu, RLS
coe icien was posi i e, sugges ing ha highe schooling le els a e co ela ed wi h inc eased po e y. This phenomenon may be
ela ed o a misma ch be ween educa ion and a ailable employmen oppo uni ies, whe e indi iduals wi h highe educa ion
s uggle o ind sui able jobs. Simila ly, Su bak i, Much a , and Sihombing (2023; Jannah & Sa i, 2023; Hasanah and P iyadi, 2021)
epo ed a posi i e ela ionship be ween educa ion and po e y. This unexpec ed ela ionship could be in e p e ed as an
indica ion o educa ed unemploymen , a condi ion in which indi iduals wi h highe educa ion ail o be abso bed in o he labo
ma ke (A sani, A io, and Ramadhan, 2020). Con e sely, in mos o he egions, such as Banggai, Banggai Kepulauan, Banggai Lau ,
Mo owali, No h Mo owali, Pa igi Mou ong, Poso, and Tojo Una-Una, RLS coe icien was nega i e, sugges ing ha educa ion
emains an impo an ins umen o educing po e y, al hough he le el o signi icance a ies.
VECM esul s showed bo h sho - and long- un dynamics. In he long un, ECT componen indica ed he p esence o
equilib ium ela ionships, al hough RLS con ibu ion was no signi ican . In he sho un, he lag coe icien s luc ua ed wi h lag-
1 being nega i e, lag-2 posi i e, and lag-3 nega i e, bu all we e insigni ican . This emphasizes ha educa ion does no ha e an
immedia e e ec on po e y educ ion. I equi es ime and suppo ing policies such as imp o ed labo quali y, job access, and
s eng hening o p oduc i e sec o s. In gene al, all h ee app oaches show ha educa ion, as measu ed by RLS, has he po en ial
o educe po e y bu emains weak wi hou pa allel e o s o imp o e educa ion quali y, align wi h labo ma ke demands, and
ensu e equi able educa ional access ac oss egions. These esul s also unde sco e he po en ial o educa ed unemploymen ,
whe e highe educa ional a ainmen is no ma ched by adequa e job c ea ion (A sani, A io, & Ramadhan, 2020).
Analysis o Fac o s A ec ing he Numbe o Poo People in Cen al Sulawesi
JEFMS, Volume 08 Issue 11 No embe 2025 www.ije m.co.in Page 7421
REM esul s show ha TPT has a posi i e bu insigni ican coe icien (0.0027; p- alue 0.5935). This sugges s ha highe
unemploymen ends o inc ease he numbe o poo people. Howe e , he e ec is no su icien ly s ong a he agg ega e le el
in Cen al Sulawesi. P e ious s udies by P adip a and Dewi (2020), Bandha, Moelyono, and Supa man (2022), and Faj iah (2021)
also epo ed simila esul s. This is consis en wi h gene al expec a ions ha highe unemploymen a es con ibu e o ising
po e y le els. The condi ion occu s because po e y in he egion is in luenced no only by open unemploymen bu also job
in o mali y, wo k quali y, and income ins abili y.
GWPR model p o ides a deepe spa ial pe spec i e. In egions such as Mo owali, No h Mo owali, Poso, Sigi, and Tojo Una-
Una, TPT coe icien was posi i e and signi ican , indica ing ha open unemploymen con ibu es o wo sening po e y. The
dependence o he local economy on speci ic sec o s causes weak labo ma ke s o di ec ly impac po e y le els. Con e sely,
egions such as Banggai, Donggala, and Toli-Toli showed nega i e o insigni ican TPT coe icien s, sugges ing ha people in hose
egions s ill ha e access o subsis ence employmen e en when o mally eco ded as unemployed.
VECM es ima ion shows he dynamic ela ionship be ween unemploymen and po e y in bo h he sho and long un. In
he sho un, lag coe icien s o TPT we e posi i e and pa ly signi ican , including lag-1 (0.0355, signi ican ), lag-2 (0.0196, nea
signi ican ), and lag-3 (0.0143, signi ican ). This indica es ha an inc ease in unemploymen in he p e ious one o h ee pe iods
di ec ly aises cu en po e y le els, e lec ing he apid labo ma ke e ec s on communi y wel a e. In he long un, he
signi ican ECT alue indica es a mechanism o adjus men owa d a new equilib ium, whe e highe unemploymen hinde s
po e y educ ion unless accompanied by inclusi e job c ea ion.
The esul s s eng hen he assump ion ha highe open unemploymen may inc ease po e y le els, as he loss o main
income sou ces leads o s uc u al po e y, pa icula ly in u ban a eas (Agin a e al., 2025). In gene al, all h ee models con i m
ha open unemploymen is one o he ac o s wo sening po e y in Cen al Sulawesi, al hough he in luence is no uni o m ac oss
egions. These esul s emphasize he impo ance o s a egies o c ea ing p oduc i e jobs, empowe ing he in o mal sec o , and
imp o ing wo k o ce skills o mee ma ke demands.
REM esul s show ha GDP pe capi a (PDRB) has a nega i e and signi ican e ec on po e y, wi h a coe icien o –0.0573
(p- alue 0.0108). In o he wo ds, an inc ease in GDP pe capi a will educe po e y in Cen al Sulawesi. Theo e ically, economic
g ow h, e lec ed in ising GDP pe capi a, enhances pu chasing powe , expands employmen oppo uni ies, and imp o es income
dis ibu ion, he eby educing po e y. These esul s a e consis en wi h Balasub amanian, Bu chi, and Male ba (2023), who ound
ha highe GDP pe capi a imp o es wel a e and lowe s po e y le els.
F om a spa ial pe spec i e, GWPR esul s show mo e di e se pa e ns. Almos all dis ic s show nega i e and mos ly
signi ican coe icien s o GDP pe capi a, such as Buol (–0.087), Mo owali (–0.037), No h Mo owali (–0.058), Poso (–0.069), Sigi
(–0.148), and Toli oli (–0.110). This demons a es ha economic g ow h in hese a eas uly con ibu es o po e y educ ion.
Howe e , excep ions we e ound in Banggai (0.009) and Banggai Lau (0.004), whe e he coe icien s we e posi i e bu
insigni ican , indica ing ha economic g ow h in hese egions has no been inclusi e. In o he wo ds, GDP g ow h emains
concen a ed in ce ain sec o s such as ex ac i e indus ies, which do no ully bene i he poo .
VECM esul s p o ide a long- e m pe spec i e. In he sho un, mos GDP pe capi a lag coe icien s a e nega i e bu
insigni ican (lag-1: –0.0975; lag-2: –0.0280; lag-3: –0.0039). Al hough insigni ican , he nega i e di ec ion is consis en wi h he
hypo hesis ha highe GDP pe capi a ends o educe po e y. In he long un, he signi ican ECT e m indica es ha economic
g ow h ac s as a s abilizing ac o capable o co ec ing imbalances owa d lowe po e y le els.
Ac oss he h ee models, GDP pe capi a plays a cen al ole in educing po e y le els in Cen al Sulawesi. Howe e , GWPR
esul s emphasize he impo ance o inclusi e g ow h, since in some egions, economic expansion has no eached poo
popula ions. The e o e, egional de elopmen s a egies should no me ely aim o inc ease GDP pe capi a bu also ensu e ha
he bene i s o g ow h a e e enly dis ibu ed among all communi y g oups.
The c oss-sec ional e ec s o he model demons a e di e ences in po e y le els among dis ic s and ci ies in Cen al
Sulawesi. Banggai Lau had he lowes po e y a e, while Pa igi Mou ong had he highes . This spa ial a ia ion can be a ibu ed
o ac o s such as economic oppo uni ies, se ice accessibili y, and local economic s uc u es. Based on he GWPR model
es ima ion, he dis ibu ion o signi ican a iables in each dis ic /ci y o Cen al Sulawesi is p esen ed in Table 6.
Table 6. Signi icance o Va iables in GWPR Model
G up
Va iable
Signi icance
Regency/Municipali y
1
RLS
Donggala
2
PDRBpk
Pa igi Mou ong
Analysis o Fac o s A ec ing he Numbe o Poo People in Cen al Sulawesi
JEFMS, Volume 08 Issue 11 No embe 2025 www.ije m.co.in Page 7422
3
RLS and TPT
Banggai
Banggai Lau
Banggai Kepulauan
4
UHH
and PDRBpk
Buol
Mo owali U a a
Poso
5
UHH, RLS,
and PDRBpk
Tojo Una-una
Toli- oli
6
UHH, TPT,
and PDRBpk
Mo owali
sigi
Sou ce: Au ho ’s calcula ion, 2025
Table 6 shows he dis ibu ion o signi ican independen a iables a ec ing po e y ac oss six g oups de i ed om GWPR model.
The analysis demons a es spa ial a ia ions in he in luence o independen a iables on po e y ac oss dis ic s and ci ies in
Cen al Sulawesi P o ince. These esul s sugges ha po e y educ ion s a egies should be spa ially a ge ed, wi h conside a ion
o he egional a ia ions in economic and educa ional e ec s. Policymake s should design localized in e en ions ocusing on
human capi al imp o emen and inclusi e economic g ow h o achie e sus ainable po e y alle ia ion.
5. CONCLUSION AND SUGGESTION
In conclusion, Pa igi Mou ong, Donggala, and Poso ha e he highes a e age numbe o poo people in Cen al Sulawesi. In gene al,
po e y emains high, wi h signi ican in e - egional dispa i ies. REM esul s show ha UHH and pe capi a GRDP signi ican ly
educe po e y, unde sco ing he impo ance o heal h and economic wel a e. GWPR p esen s spa ial a ia ions in he in luence
o socio-economic indica o s ac oss dis ic s, while VECM con i ms alid long- e m ela ionships and dynamic sho - e m e ec s.
These esul s sugges ha heal h imp o emen s, economic g ow h, and employmen oppo uni ies a e key d i e s o po e y
educ ion in Cen al Sulawesi.
Po e y alle ia ion equi es comp ehensi e a en ion om he go e nmen , p i a e sec o , and ci il socie y. P io i y
policies should ocus on equi able heal hca e se ices, inclusi e job c ea ion, and human esou ce de elopmen . S eng hening
he in o mal sec o h ough easie licensing, inancing, and aining can enhance employmen and local economic capaci y.
Educa ion quali y and access mus also be imp o ed o build adap i e and compe i i e human esou ces ac oss he p o ince. An
in eg a ed app oach ha conside s local cha ac e is ics is essen ial o achie e sus ainable po e y educ ion.
REFERENCES
1) Adhi ya, B., P abawa, A., & Kencana, H. (2022). Analisis Penga uh Pendidikan, Keseha an, Sani asi dan Ra a-Ra a Jumlah
Anggo a Kelua ga Pe Rumah Tangga e hadap Kemiskinan di Indonesia. Ekonomis: Jou nal o Economics and Business,
6(1), 288. h ps://doi.o g/10.33087/ekonomis. 6i1.501
2) Agin a, C., Tampubolon, M., Manu ung, M. P., Tobing, R. D., & Rahayu, T. (2025). Analisis penga uh ingka penganggu an
e buka dan indeks pembangunan manusia e hadap jumlah penduduk miskin di Suma e a U a a pada ahun 2000–2023.
EKOMA: Ju nal Ekonomi, Manajemen, Akun ansi, 4(4), 70–88. h ps://doi.o g/10.56799/ekoma. 4i4.8648
3) A mida, S. A., & Mu nining yas, E. (2018). Tujuan pembangunan be kelanju an di Indonesia: Konsep, a ge , dan s a egi
implemen asi. Bandung, Indonesia: Unpad P ess.
4) A sani, A. M., A io, B., & Ramadhan, A. F. (2020). Impac o educa ion on po e y and heal h: E idence om Indonesia.
Economics De elopmen Analysis Jou nal, 9(1), 87–96. h ps://doi.o g/10.15294/edaj. 9i1.34921
5) Azkiya, A. Al, Ang aini, Y., & Anisa, R. (2024). Pene apan Geog aphically Weigh ed Panel Reg ession dan Da a
En elopmen Analysis dalam Pemodelan Kemiskinan di Kaliman an Timu . Jou nal o Regional and Ru al De elopmen
Planning, 8(1), 41–53. h ps://doi.o g/10.29244/jp2wd.2024.8.1.41-53
6) Az iyansyah, Z. (2022). Analisis penga uh IPM, PDRB pe kapi a dan ingka penganggu an e hadap ingka kemiskinan
di Indonesia pe iode 2017–2021. Ju nal Ekonomi, Bisnis dan Manajemen, 1(3), 225–238.
h ps://doi.o g/10.58192/ebismen. 1i3.308
7) Badan Pusa S a is ik. (2024a). Jumlah penduduk miskin Kabupa en/Ko a di P o insi Sulawesi Tengah. BPS.
h ps://sul eng.bps.go.id/s a is ics- able/3
8) Badan Pusa S a is ik. (2024b). P oduk domes ik egional b u o (PDRB) pe kapi a Kabupa en/Ko a di P o insi Sulawesi
Tengah. BPS. h ps://www.bps.go.id/id/s a is ics- able/3
9) Badan Pusa S a is ik. (2024c). Ra a- a a lama sekolah Kabupa en/Ko a di P o insi Sulawesi Tengah. BPS.
