Jurnal Bina Praja 9 (2) (2017): Jurnal Bina Praja. e-issn: p-issn:

Transcription

1 Jural Ba Praa 9 (2) (2017): Jural Ba Praa e-issn: p-issn: Eergy Goods Demad Tabalog Regecy: Almost-Ideal Demad System Approach Ahmad Mura * BPS-Statstcs of Tabalog Regecy Jala Jaksa Agug Soeprapto No. 82 Taug, Tabalog Regecy, South Kalmata, Idoesa Receved: 18 August 2017; Accepted: 2 November 2017; Publshed ole: 28 November 2017 DOI: /bp Abstract The declg tred of Tabalog Regecy s ecoomc growth recet years adversely affected the poverty rate. Further, the recet eergy subsdy polcy appled by the Idoesa Govermet has pushed the subsdy s budget dow for some eergy goods. Therefore, there should be a awareess regardg the curret eergy polcy ad the mpact o the poverty partcularly Tabalog Regecy. Ths paper vestgates the demad system for the three ma eergy goods; premum fuel, electrcty, ad Lqud Petroleum Gas (LPG) Tabalog Regecy of the South Kalmata Provce. Although the same method was prevously used o the dfferet topcs, ths paper uquely utlzes the combed Lear Approxmato ad Quadratc Almost-Ideal Demad System o the partcular eergy polcy topc. Ths paper utlzes the Natoal Socal Ecoomcs Survey coducted by BPS-Statstcs of Tabalog Regecy The results show that the come elastcty of demad for the top 60% ad the bottom 40% of the come groups were postve; however, slght dffereces could be see. For the top 60% of the come group, the come elastctes of demad were 0.97, 1.02, ad 1.08 for premum fuel, electrcty, ad LPG respectvely. O the other had, the bottom 40% of the come group had 0.99, 1.07, ad 0.91 of come elastcty of demad for premum, electrcty, ad LPG. The prce elastcty of demad for both come groups had egatve sgs, whch s agreeg wth the theoretcal demad fucto. These results dcate that the curret eergy polcy should cotue wth securg the poor households from the possble effect. Keywords: Tabalog, Prce Elastcty of Demad, Icome Elastcty of Demad, Quadratc Almost-Ideal Demad System, Demad Aalyss I. Itroducto Tabalog Regecy expereced a declg tred of the ecoomc growth the perod 2011 to Luckly, the Tabalog s ecoomc growth recovered After eoyg the hghly-sevepercet growth 2011, the growth kept declg to 2.36% Ths tred, ufortuately, brgs a mpact o the poverty rate Tabalog. The headcout dex of Tabalog Regecy decreased from 6.22% to 5.83% the perod 2011 to 2012 but creased the remag perod of 2012 to 2015 (from 5.83% to 6.59%). I 2016, the creasg ecoomc growth of Tabalog Regecy stmulated a good mpact o the poverty rate. The data showed that the 3.06-percet ecoomc growth 2016 affected the socal codto Tabalog whch led to a decreasg headcout rato to 6.35% (Statstcs of Tabalog Regecy, 2017). The above stuato should become a prmary cocer for the polcymakers Tabalog regardg the recet polcy, partcularly the eergy subsdy polcy that was tated by the prevous Idoesa Govermet. Idoesa Govermet has bee gradually decreasg ts eergy subsdy due to the budget defct ad reallocato of fuds from 2014 to 2017 (Drektorat Peyusua APBN & Drektorat Jederal Aggara, 2017). The eergy subsdy comprses of two prmary targets: fuel ad electrcty subsdes. I Idoesa, at the tal phase, there were fve products usg fuel subsdes: gasole, kerosee, automotve desel ol, dustral desel ol, ad fuel ol. As the tme chaged, ad as a part of the eergy reform, * Correspodg Author Phoe : Emal : ahmadmura@gmal.com 2017 Lbrary, Iformato, ad Documetato R&D Agecy - MoHA all rghts reserved. Accredtato Number 735/AU2/P2MI-LIPI/04/

2 from 2005, oly three products receved the fuel subsdy: gasole, kerosee, ad automotve desel ol (Dartato, 2013). I 2007, the fuel subsdy expaded to corporate lqud petroleum gas (LPG) as a part of the govermet s coverso program from kerosee to LPG (Iteratoal Isttute for Sustaable Developmet, 2012). Based o the prevous researches, reducg the eergy subsdy wll lead to a crease the poverty rate. Dartato (2013) utlzed a CGEmcrosmulato method o 64,407 SUSENAS samples to observe the mpact of removg the fuel subsdes o the poverty rate Idoesa. By usg data from the year 2005, he foud that whe the fuel subsdes were reduced by 25%, the poverty rate wll crease by 0.259%. Aother research coducted by Bresger, Egelke, & Ecker (2012) used Dyamc-CGE to exame the mpact of reducg fuel subsdes o GDP ad poverty Yeme. They foud that reducg the fuel subsdes wthout compesato would lead to creases poverty rate both urba ad rural areas by 2.6%. However, the adverse mpact o the poverty could be couterbalaced by drect trasfer to the poorest oe-thrd of all households. Based o these researches, the decelerato of the ecoomc growth ad creasg tred of the poverty rate Tabalog, the focus wll be o the subsdzed eergy goods whch are the ma commodtes the eergy subsdy polcy. There was some research Idoesa whch vestgated the demad system for some goods. Wdaroo (2013) modeled the demad system Idoesa for meat usg Almost-Ideal Demad System model. Sulsto, Wrabhuaa, & Wratama (2017) examed the demad for electrcty Idoesa usg Dyamc Modellg method. Maryoo (2017) employed a Lear Fuctoal Form Model to estmate the demad model for toursm Idoesa. The smlar research topc was coducted by Bhakt (2011) usg the Lear-Approxmato of Almost-Ideal Demad System (LA-AIDS). She examed the demad system for eergy goods Java the perod Thus, ths paper ams to vestgate the demad system for three partcular eergy goods Tabalog Regecy usg two models, combg the Lear-Approxmato of Almost-Ideal Demad System (LA-AIDS) ad Quadratc Almost-Ideal Demad System (QUAIDS). Ths paper corporates the demad system for eergy goods Tabalog Regecy wth the curret eergy polcy by the Idoesa Govermet to produce the polcy recommedato for the local govermet of Tabalog Regecy related to the eergy cosumpto. The secod part of ths paper s the methodology, whch comprses the data selecto ad model specfcato. The thrd part of ths paper s results ad dscusso, ad the last part s coclusos. A. Eergy Cosumpto Profle Tabalog Regecy Natoal Socal Ecoomc Survey (Suseas) produces the households' cosumpto data. Based o the data, ths paper tabulates the eergy goods cosumpto to two basc tables (Table 1 ad Table 2) to aalyze the eergy cosumpto profle Tabalog Regecy. Table 1 tabulates the eergy goods spedg patter o partcular cosumpto decso. Hece, ths table provdes formato regardg both the most ad the least eergy goods cosumed by the households regardless the come groups ad locatos (as the percetages are summed to 100% vertcally, the comparso caot be made colum per colum). Alteratvely, Table 2 provdes the colums comparsos for the come groups ad locatos. However, sce the percetages are 100% total horzotally, the table ca't be compared based o cosumpto s decso. 1) Eergy Goods Cosumpto Decso (Table 1 Aalyss) Geerally, 2016, households Tabalog Regecy mostly cosumed electrcty. Table 1 shows that the households that used electrcty reached 99.06%, ad oly 0.94% dd ot have access to electrcty. Ths dcates that Tabalog Regecy almost reached full electrfcato stage. The secod hghest eergy goods cosumed by households Tabalog Regecy was premum fuel. The ease of buyg ew vehcles usg facal lease help to duce the crease premum fuel usage. Not oly that, the premum fuel s oe of the eergy commodty that s stll subsdzed by the govermet utl ow. I Table 1, 88.38% of households used premum fuel, oly 11.62% dd ot cosume t 2016 (or the perod of the survey). The thrd hghest eergy goods cosumed by households Tabalog Regecy was LPG. The coverso program tated by the Idoesa Govermet 2007 successfully reduced the usage of kerosee by 80% the perod 2007 to 2011 (Iteratoal Isttute for Sustaable Developmet, 2012). The declg usage of kerosee meas the households started to shft from kerosee to LPG as a fuel for cookg. Table 1 descrbes that 73.73% households Tabalog Regecy used LPG as a cookg fuel, 26.27% were usg other cookg fuel such as kerosee, frewood, etc. Kerosee became the fourth hghest eergy goods cosumed by the households Tabalog Regecy durg The mpact of successful Jural Ba Praa 9 (2) (2017):

3 Table 1. Eergy Goods Cosumpto by Icome Groups ad Locato Tabalog, 2016 (vertcal summato %) Type of Eergy Goods Cosumed Icome Groups Locato Top 60% Bottom 40% Urba Rural Total Pertamax Fuel yes o Premum Fuel yes o Electrcty yes o LPG yes o Kerosee yes o Automatc Desel Ol (ADO) yes o Source: Statstcs Idoesa, 2016, Author s calculato kerosee-to-lpg coverso program s the low cosumpto of kerosee, whch was oly 16.19% Automotve Desel Ol (ADO) s oe of the subsdzed eergy goods Idoesa. As the ffth hghest eergy goods cosumed Tabalog Regecy, ADO s used mostly by cars. As the ature of these products whch s oly cosumed by the household that ows a car, oly 2.33% of the households Tabalog Regecy cosumed t as show Table 1. Pertamax fuel was the least eergy goods cosumed Tabalog Regecy As descrbed Table 1, oly 2.09% of the households cosumed t durg ) Icome Groups ad Locato (Table 2 Aalyss) I term of locatos, as ca be see Table 2, the households that dd ot have access to electrcty were located both urba areas (13.47%) ad rural areas (86.53%). O the other had, Table 2 also shows that those households wthout electrcty comprsed of 35.11% from the top 60% ad 64.89% of the come groups. For premum fuel, the percetage of households that cosumed premum the rural areas was more tha the urba areas, whch were 61.64% ad 38.36% respectvely. Ths meas the access for premum fuel s ot oly urba areas but also rural areas. Furthermore, ths fgure descrbed that the ease of obtag vehcles reached all areas. A dfferet fgure appeared term of come groups' access to premum fuel show Table 2, 64.47% of households the 60% of the come group cosumed more premum fuel compared wth the households from the bottom 40% of the come group (35.53%). Aga, ths meas the wealther households cosume more premums fuels for vehcles. Ths evdece supports the prevous fdgs that proved fuel subsdy s cosumed mostly by the wealther household Rohac (2013). I Table 2, LPG was mostly cosumed the rural areas (55.57% of households); ad term of come group, 63.56% of the households from the top 60% of the come groups used LPG as a cookg fuel. The fgure mples that the proporto of households that used LPG s more the rural area; however, the top 60% of the come group had a hgher proporto to use LPG. Ths meas, term of proporto, households the rural areas had easy access to LPG, ad the top 60% of the come group also possessed the hgher porto. Table 2 shows that the rural areas shared 70.01% of kerosee cosumpto whereas the urba area had 29.99% of share. Iterestgly, the households from the top 60% of the come Eergy Goods Demad Tabalog Regecy: Almost-Ideal Demad System Approach Ahmad Mura 309

4 Table 2. Eergy Goods Cosumpto by Icome Groups ad Locato Tabalog, 2016 (horzotal summato %) Type of Eergy Goods Cosumed Icome Groups Locato Top 60% Bottom 40% Urba Rural Total Pertamax Fuel yes o Premum Fuel yes o Electrcty yes o LPG yes o Kerosee yes o Automatc Desel Ol (ADO) yes o Source: Statstcs Idoesa, 2016, Author s calculato group cosumed more kerosee % of the households from the bottom 40% of the come group cosumed less whch proves that kerosee subsdy s ot fully absorbed by the ma target of the subsdy. Table 2 depcts that 100% of the households that cosumed ADO came from the top 60% of the come group, whch was mostly located the rural areas by 75.12%. I term of the recpet of the ADO subsdy, ths subsdy was mostly absorbed by the hgher come households. For the Pertamax fuel, most of the cosumers were located the urba areas (89.14%). I term of the come groups, the cosumpto was about the same, but the hgher cosumpto wet to the top 60% of the come group by 54.31%. B. Demad System: Theoretcal Backgroud 1) Effect of Icome ad Prce Chages to the Cosumer s Demad Ncholso & Syder (2011) explaed the possble effect regardg the chages cosumer s come ad goods prce. The frst effect was the demad chages regardg the come. For the ormal goods, the hgher come would brg the hgher demad for the goods because the cosumers would maxmze ther spedg (the hgher budget, the more budle of goods cosumed). However, ths assumpto was dfferet the case of feror goods. The hgher come would create the shftg prorty for the cosumers. The cosumers pcked the hgher qualty goods rather tha spedg more o the feror goods. The secod effect was the demad chages regardg the prce of the goods. Geerally, the hgher prce of partcular goods would decrease the demad for t. The stuato could be dfferet some cases. For the ecessty ad addctve goods, the demad for t would ot chage so sgfcatly whe the prce chages. O the other had, the goods wthout ay close substtutes would follow the ormal goods demad behavor to the prce chages. The chagg prce of the complemetary goods also determed the demad for the pared goods, for example: whe the prce of sugar decreased, the demad for tea would crease (the lower prce of sugar, the more demad for sugar ad tea). 2) Nature of Demad The utlty s a ecoomcs cocept that explas the cosumers satsfacto from cosumg goods ad servces. A ormal cosumer wll choose a basket of goods ad servces to acheve Jural Ba Praa 9 (2) (2017):

5 the hghest utlty level. Based o these cocepts, the researcher ca derve a demad system ad estmate the demad fucto to obta the estmato parameters. Evetually, the researcher ca predct the cosumers' behavor term of demad. Next equatos are the ature of demad fucto that explas the households' behavor cosumg goods ad servces (Wdaroo, 2016). Mathematcally, cosumers maxmze ther utlty related to the prces ad the quattes of goods ad servces. The relato ca be wrtte as: Max p q = I...(2.1) U= =1 where U s utlty level, p s the prce, q s the quatty cosumed, s the goods ad servces, ad I s the amout of Icome. Usg the Lagraga, the equato (2.1) produces Marshalla demad fucto: q = q( pi, )...(2.2) The demad fucto above descrbes the relatoshp betwee the quattes of goods ad servces, prces, ad come. The quatty demaded wll be determed by the prces ad the households come. Usg the equato (2.1) ad (2.2), the parameters of demad fucto ca be obtaed. The frst ature of demad fucto s addg up. Ths ature explas that cosumers wll sped all the come they have, ad ca be wrtte as: pq + pq + + pq = I pq( p, I) = I =1...(2.3) where p s the prce, q s the quatty cosumed, ad I s the cosumers come. The equato (2.3) ca be decomposed wrtte to: pq( pi, ) + pq( pi, ) + + pq( pi, ) = I..(2.3) The frst order dfferetato of equato (2.4) cocerg I produces the equatos: q q q p + p + + p = q I q q I q p + p q I q I q I q I q1 + p = 1 q I q I 1 1 pq q I pq q I + + I q I q pq q I + = 1 I q 1 w w w η + η + = η = w η 1 =1 Where w s the budget share, ad η s the come elastcty. The equato (2.5) s commoly kow as the Egel Aggregato. The secod ature of demad fucto s homogeety. If all prces of all goods ad servces crease whle the same tme, comes also crease wth costat level (θ); therefore, there wll ot affecto the quattes of all goods ad servces cosumed. The demad fucto has a homogeousof-degree-zero property. It ca be expressed as: q( θθ) = q( pi, )...(2.6) p I Usg Euler s theorem, the frst order dfferetato of equato (2.6) satsfes the equato: p p p I q1 q2 q q = p1 p2 p I If equato (2.7) s dvded by q 1, the t producess the followg equato: p p p I = 0 q q q q 1 q1 2 q2 q q 1 p1 1 p2 1 p 1 I e e e η = = e η 0 k =1 k... (2.5)...(2.7)...(2.8) Both sdes are multpled by Iq qi The thrd ature of demad fucto s symmetry. It meas the cross dfferetato of the demad fucto s symmetrc. Ths ca be wrtte as: Eergy Goods Demad Tabalog Regecy: Almost-Ideal Demad System Approach Ahmad Mura 311

6 h ( u, p) h ( u, p) x / x x py x py =, where... (2.9) e = =. =. x, py... (2.12) p p p / p p x p x y y y y f h h S = = =S p p ad h= C p thus, by the Youg s theorem 2 2 C S = = = S pp pp h p q p I I = ew p q p I p pp S = S we = we 1 1 e + η = e + η w w w e = e + w ( η η ) Elastcty s utlzed to measures the prces ad comes chages to the quattes of goods ad servces demaded. The elastcty used ths paper refers to Marshalla elastcty that ca be expressed as: e x, px... (2.10) x / x x px x px = =. =.... (2.11) p / p p x p x x x x x Where e x,px s the prce elastcty of demad, x/x s the chage quatty demaded of goods x, p x /p x s the chage prce of goods x. e x,px s expected to be egatve; e x,px = 0 s perfectly elastc; 0 < e x,px < 1 s elastc; e x,px = 1 s ut elastc; e x,px > 1 s elastc. Where e x,py s the cross-prce elastcty of demad, x/x s the chage quatty demaded of goods x, p y /p y s the chage the prce of goods y. e x,py could be postve or egatve, e x,py > 0 meas both goods x ad y are substtuto goods, e x,py < 0 meas both goods x ad y are complemetary goods. e x, I x / x x I x p y = =. =....(2.13) I / I I x I x Where e x,i s the come elastcty of demad, x/x s the chage quatty demaded of goods x, I/I s the chage come. e x,i could be postve or egatve, e x,i > 0 meas goods x s a ormal goods, e x,i < 0 meas goods x s a feror goods, e x,i > 1 dcates the goods x s a luxury goods. The fourth ature of the demad fucto s egatvty. Matrx wth x dmeso that has q p elemets should have egatve semdefte property. Aother demad fucto s ature s addtvty ad separablty. Addtvty meas f the utlty from cosumg certa goods s depedet of other goods, the the utlty level ca be summed up. It ca be expressed: U = U + U + + U... (2.14) 1 2 U f ( q, q,, q ) = 1 2 m m Where m s umber of goods. Separablty the demad fucto explas that the cosumers ca dvde ther come to some parts (groups). For example, a household s cosumpto ca be categorzed to some baskets of goods ad servces regardg the come ad preferece. Ths also meas the utlty of oe basket of goods s depedet to other baskets of goods. 3) Demad System Models I the welfare aalyss, the modelg of demad fuctos s a very mportat step to exame the households' behavor. The well-kow demad fucto modelg was developed by Stoe (1954) based o the cosumers theory. However, there were may demad models that produced by may ecoomsts. Kle & Rub (1948) tated a demad model that geerated from the cost of lvg dex that socalled The Lear Expedture System. Smlar to Jural Ba Praa 9 (2) (2017):

7 other demad models, ths model uses addg-up, homogeety, ad symmetry restrctos. Thel (1975) modfed the demad model created by Stoe (1954). The demad model developed by Hery Thel s kow as the Rotterdam Model. Ths model dfferetates the Stoe Model ad serts the budget share the dfferetated equato. Trascedetal Logarthmc Demad System a.k.a Tras Log Model s a model that employs the drect utlty fucto, drect utlty fucto, ad the cost fucto. Developed by Chrstese, Jorgeso, & Lau (1975), ths model s kow for ts flexblty. Ths model uses ormalzed prces to come ts equato. Deato & Muellbauer (1980) developed the demad model based o expedture fucto called Almost-Ideal Demad System (AIDS). Ths model clams that t has some advatages compared wth other demad models such as: havg frstorder approxmato; accommodatg homogeety, symmetry, ad addg up restrctos; ad havg a fucto that cosstet wth cosumer's budget. I ts developmet, ths model had modfed to Lear-Approxmato Almost-Ideal Demad System (LA-AIDS) cocerg the u-learty of the prce dex. I AIDS model, the Egel's curve s assumed to be lear whch meas the goods cosumed by a household are ormal goods. However, realty, there are some feror goods. Thus, wth the exstece feror goods' cosumpto, the Egel's curve s o loger lear. Baks, Bludell, & Lewbel (1997) developed the Quadratc Almost- Ideal Demad System (QUAIDS) to allow the Egel's curve assumpto to be quadratc. C. Demad System for Eergy Goods: Emprcal Evdeces The AIDS model s wdely used for estmatg the demad system of eergy goods. The chose AIDS model s based o varous cosderatos across the coutres, for example, the eergy goods cosumpto patter, eergy goods producto, the avalablty of substtuto goods, etc. Gudmeda & Köhl (2008) vestgated the come elastcty of demad for eergy goods Ida usg LA-AIDS demad model. They utlzed the mcrodata of approxmately 100,000 households all over Ida, ad they also dvded the households to dfferet locatos, urba ad rural areas by dfferet come groups. Y, Geetha, & Chadra (2017) utlzed the LA-AIDS to estmate the elastcty of petrol, desel, electrcty, ad LPG Sabah Malaysa. They foud that the low-come households are the group that possesses the hghest cost of electrcty's overcosumpto. Moreover, petrol ad desel were prove to be prce elastc whereas electrcty ad LPG were elastc. Bazzaza, Ghasham, & Mousav (2017) examed the electrcty for electrcty Ira usg AIDS model. The result was that electrcty Ira, from 1991 to 2012, was the ecessty eergy goods. Ths result was foud both rural ad urba areas Ira ad ecouraged households to be more effcet the electrcty cosumpto. J & Zhag (2013) used the mothly mcrodata from households Beg from 2002 to 2009 to calculate the elastcty of resdetal electrcty demad for dfferet come groups. By usg combed AIDS ad the Lear Double-Logarthmc (LDL) method, they foud that the prce elastcty of electrcty was almost equal to 1 (ut elastc). They suggested that tarff adustmet s mportat as a polcy recommedato. Therefore, the codtoal electrcty subsdy should be appled. I Idoesa, Bhakt (2011) vestgated the prce elastcty, come elastcty, ad cross elastcty of eergy goods Java. By applyg the Lear- Approxmato AIDS (LA-AIDS) 2007 to 2010 Suseas data, she foud that all aalyzed eergy goods had a come elastcty of more tha oe whle all of the prce elastcty was egatve. More research that uses AIDS model as well as a combed model to estmate the demad fucto for eergy goods could be foud outsde Idoesa, but oly a few papers employed ths method Idoesa. Therefore, to fll the lack of avalable research o ths partcular topc, ths paper ams to vestgate the elastcty of eergy goods Tabalog Regecy usg combed LA-AIDS ad QUAIDS model II. Method A. Data Selecto Ths paper utlzed the Suseas data a scope of Tabalog Regecy I 2016, BPS of Tabalog Regecy coducted Suseas March ad September for regecy ad atoal estmato respectvely. BPS Tabalog surveyed 555 households March 2016 as samples. However, some households dd ot cosume some eergy goods the perod of the survey. Therefore, to avod udefed value the demad model (whch s usg l), we choose three ma eergy goods cosumed 2016 (electrcty, premum, ad LPG). Evetually, the umbers of households the model are 387. To expad the aalyss, 387 households wll be dvded to two groups, the top 60% (255 households) ad the bottom 40% (132 households) of the come groups. The groups come derved from groupg the households come from the hghest to the lowest, ad sorted t to te decles, Eergy Goods Demad Tabalog Regecy: Almost-Ideal Demad System Approach Ahmad Mura 313

8 the households the upper decles (frst to the sxth) are grouped to the top 60% of the come class. The rest are combed to the bottom 40% of the come class. B. Model Specfcato Ths paper employs both the AIDS ad QUAIDS model, depeds o the prmary result of the aalyss. The value of λ determes the form of the model. If λ s statstcally equal to zero, the the QUAIDS model becomes AIDS model, ad vce versa (Wdaroo, 2016). The QUAIDS model ths paper ca be wrtte as: X w = α + = 1 γlp + β l ap ( ) λ X bp ( ) ( ) ap 2 + (l ) + u...(3.1) where ad are types of goods, w s the budget share allocated for goods, p s the prce for goods, X s the total expedture of household, a( P) s prce dex that comes from the equato: ( ) = l a P α = αlp = = γ lp lp...(3.2) 1 1 b(p) s the aggregate prce of Cobb-Douglass expressed as: ( ) = = 1 b P p β If λ = 0 for all, the QUAIDS model becomes AIDS model that ca be wrtte as: X w = α + =1 γlp + β l + ( ) u... (3.4) a P To accommodate the demographc varables, whch also affect the cosumpto decso for such eergy goods, the demad model s modfed ts tercept such as: α = ρ + η d...(3.5) m 0 k = 1 k k... (3.3) where d k s k-th demographc varable. All all, the varables cluded ths paper comprse of: Expedtures of premum, electrcty, ad LPG to calculate the budget share the model; The prce of premum, electrcty, ad LPG faced by each household; Total expedture of the households; sze of the households, educato level of the households head s, locato of the households, geder of the households head s, ad age of the households head s. For the levels of educato ca be classfed as: 1 = droppg out from elemetary school 2-5 = elemetary school graduates 6-9 = uor hgh school graduates = seor hgh school graduates = havg a udergraduate degree = havg a postgraduate degree The data processg ths paper uses SPSS 15.0 for the cross-tabulato ad the Stata package for QUAIDS developed by Po (2012) to estmate the demad model as well as the elastctes. III. Results ad Dscusso A. Eergy Cosumpto Tabalog Regecy As descrbed chapter 1, the three ma eergy goods cosumed Tabalog 2016 were premum, electrcty, ad LPG. The percetage of households cosumed for the respectve eergy goods were 88.38%, 99.06%, ad 73.73%. The summary statstcs for the eergy goods' demad model ca be see Table 3. Table 3 shows the summary statstcs for the varables used the eergy goods demad model. Based o the fltered data, the households had a average of aroud four famly members. The age of households head was aroud 45 wth the educato level of uor hgh school. Examed from the budget share perspectve, the households used the hghest proporto of buyg premum for ther vehcles (by 0.57). The smallest budget share from the three eergy goods was for LPG (by 0.14). Ths dcates that o average, the households had more prorty trasportato tha cookg, wth the electrcty became the mddle prorty. B. The Demad Model for Eergy Goods Tabalog Regecy The decso whether the demad model forms LA-AIDS or QUAIDS depeds o whether the Egel curve s lear or ot. Ths ca be see the sgfcace of λ. I the tal test, the model dcated that for the top 60% of come group, the λ was statstcally sgfcat for most of the λ (the Egel curve s ot lear form). Ths meas the top 60% of the come group wll be aalyzed by QUAIDS. O the other had, the bottom 40% of come group dd ot provde sgfcat λ, all λ were ot statstcally sgfcat (the Egel curve Jural Ba Praa 9 (2) (2017):

9 Table 3. Summary Statstcs of Varables the Model (=387), 2016 Varable Mea Std. Dev. Household Sze Age of Household Head Educato Level of Household Head Use of Eergy Goods ( Percet) * Premum Electrcty LPG Premum Prce ( Rupah/ltter) 7, Electrcty Prce ( Rupah/Kwh) LPG Prce ( Rupah/kg) 7, , Total Expedture ( Rupah) 3,477, ,385, Budget Share of Premum Budget Share of Electrcty Budget Share of LPG Note: * dcates the data used come from the actual Suseas data (=555) Source: Author s Calculato Table 4. Model Estmato for the Top 60% of Icome Group Tabalog, 2016 Parameter Estmato Depedet Varable Share the Model Premum Electrcty LPG costat 1.109** Prce of Premum fuel Prce of Electrcty ** Prce of LPG ** *** Households Sze Educato Level * * Households Locato ** ** Households head geder Households head age Households Expedture * 0.136* Note: *, **, *** dcate sgfcat at 10%, 5%, ad 1% respectvely Source: Author s Calculato Eergy Goods Demad Tabalog Regecy: Almost-Ideal Demad System Approach Ahmad Mura 315

10 Table 5. Model Estmato for the Bottom 40% of Icome Group Tabalog, 2016 Parameter Estmato Depedet Varable Share the Model Premum Electrcty LPG costat 0.505*** 0.382*** 0.114** Prce of Premum fuel 0.117** *** Prce of Electrcty Prce of LPG *** *** Households Sze Educato Level ** * Households Locato 0.105** *** Households head geder * ** Households head age e Households Expedture Note: *, **, *** dcate sgfcat at 10%, 5%, ad 1% respectvely Source: Author s Calculato s assumed lear). Therefore, the bottom 40% of come group wll be examed usg LA-AIDS. O the top 60% of come group, amog all varables volved the premum fuel cosumpto, oly educato level, households locato, ad households expedture that sgfcatly affected demad for premum fuel. For electrcty demad, the same varables as the premum fuel demad volved. The prce of LPG became the addtoal varable that also affected the demad for electrcty. For the LPG demad, oly two varables that sgfcatly affected: the prce of electrcty ad the prce of LPG. Table 4 shows that for the households the top 60% of come group, the demographc varables such as geder ad age of households head dd ot affect the demad for all eergy goods. Compared wth the Table 4, whch provdes QUAIDS model estmato, table 5 exhbts the LA- AIDS model demad estmato partcularly for the households the bottom 40% of come group. Table 5 shows that for the demad for premum fuel, fve varables were sgfcat: the prce of premum, the prce of LPG, educato level, households locato, ad the geder of households head. However, for electrcty demad, oly oe varable affected sgfcatly, whch was households' locato. For LPG, the varables that sgfcatly affected the demad were smlar to the premum s Table 6. Prce ad Icome Elastctes of the Top 60% of Icome Group for Eergy Goods Tabalog, 2016 Parameter Elastcty Premum Electrcty LPG Prce of Premum fuel Prce of Electrcty Prce of LPG Households Expedture Source: Author s Calculato 316 Jural Ba Praa 9 (2) (2017):

11 Table 7. Prce ad Icome Elastctes of the Bottom 40% of Icome Group for Eergy Goods Tabalog, 2016 Parameter Elastcty Premum Electrcty LPG Prce of Premum Prce of Electrcty Prce of LPG Households Expedture Source: Author s Calculato case wth the absece of households locato as the sgfcat varable. Regardg the demographc varables the LA-AIDS model, households sze ad households head age were ot sgfcat the demad for all eergy goods the model. Table 6 provdes the Marshalla prce ad come elastctes for the three eergy goods Tabalog 2016 partcularly for the top 60% of come group. The ow-prce elastctes were all egatve as expected. Amog three eergy goods table 6, the premum fuel had the bggest elastcty followed by electrcty ad LPG. For the come elastcty of demad, all three eergy goods were postve wth electrcty ad LPG had a value more tha oe. Regardg these fgures, the electrcty ad LPG were cosdered to be the secodary/ luxury goods for the households the top 60% of the come group. Ths also dcates that the households ths come group ted to cosume more electrcty ad LPG wth the crease ther come (for example by buyg more electroc goods or raw foods to cook). Premum fuel s cosdered as the ecessary goods sce ts come elastcty was 0.97 (less tha 1). Table 7 shows the Marshalla prce ad come elastctes for the three eergy goods Tabalog 2016, partcularly for the bottom 40% of come group. The ow-prce elastctes were all egatve I table 7, electrcty had the bggest elastcty followed by premum fuel ad LPG. For the come elastcty of demad, all the three eergy goods were postve wth oly electrcty had a value more tha oe. Electrcty was cosdered to be the secodary/luxury goods for the households the bottom 40% of the come group whereas premum ad LPG are the ecessary goods. IV. Cocluso Ths paper vestgates the demad system for three ma eergy goods Tabalog Regecy 2016 usg Suseas data. The tal step was to determe the sutable demad model for two come groups. The secod step was to buld the demad model usg ether LA-AIDS or QUAIDS. Evetually, the elastctes could be calculated from the demad model. Ths paper produced the followg coclusos. Amog fve eergy goods cosumed Tabalog Regecy 2016 such as Pertamax fuel, premum fuel, electrcty, LPG, kerosee, ad ADO; most of the households Tabalog cosumed premum fuel, electrcty, ad LPG. The percetage of households cosumed those eergy goods were 88.38%, 99.06%, ad 73.73% respectvely. For the premum fuel, the hghest cosumpto was the rural area by 61.64% whle the hghest cosumers were the top 60% of the come group (64.47%). For electrcty, most cosumpto was the rural area by 61.36% whereas the hghest cosumers were the top 60% of the come group (64.61%). LPG cosumpto maly the rural area by 55.57% ad the most cosumers came from the top 60% of the come group. These fgures dcate that the top 60% of the come group was the ma cosumer of the eergy goods Tabalog I other words, ths come group receved the eergy subsdy for the most. However, for the electrcty, the govermet already adusted the subsdy scheme by removg the subsdy for the households that usg 900VA (volt-ampere) or more ther houses. For the households that use 450VA voltage ther house, the govermet cotues to provde the subsdy as they are cosdered as a poor household. The Govermet also eeds to be cocered wth adustg the other eergy goods prce such as premum fuel ad LPG sce the ma beefcares come from the top 60% of the come group. However, ths polcy should be take carefully cosderg the households the bottom 40% of the come group used these eergy goods as the ecessty goods (could be see the come elastcty of demad). To obta the come ad prce elastctes of Eergy Goods Demad Tabalog Regecy: Almost-Ideal Demad System Approach Ahmad Mura 317

12 demad for the eergy goods, the estmato results of the demad system model were utlzed ths paper. Ths paper dvded the aalyss to two parts based o the households' come group. The frst part was about the demad system aalyss usg QUAIDS for the top 60% of the come group. The result provded the evdece that the households ths come group cosdered the electrcty ad LPG as secodary/luxury goods. A 1% crease ther come wll be followed by more tha 1% crease those goods' cosumpto. Although the come elastcty of demad for premum fuel was almost oe (0.97), these eergy goods were classfed as the ecessary goods. For the ow-prce elastcty, premum fuel became the eergy goods that had the hghest elastcty by Ths hgh elastcty also dcates that the top 60% of the come group possessed a ablty to shft to other substtuto goods such as Pertalte fuel ad Pertamax fuel whe the prce of premum fuel creases. However, t eeds more cosumpto s data from Pertalte fuel ad Pertamax fuel to aalyze ths hypothess usg cross-prce elastcty. The secod part of the demad aalyss was focusg o the bottom 40% of the come group. Ths come group utlzed the LA-AIDS model. The estmato result usg LA-AIDS model suggested that the households ths come group cosdered the electrcty as secodary/luxury goods whle premum ad LPG were cosdered as ecessty goods. For the ow-prce elastcty, electrcty had the bggest elastcty for ths come group (by -0.87). Ths meas, creasg 1% the electrcty prce wll decrease the electrcty demad (cosumpto) by 0.87%. Despte ts bggest elastcty, the owprce elastcty of electrcty the bottom 40% of come group was smaller tha the elastcty the top 60% of the come group (-0.90). Overall, the ow-prce elastcty the bottom 40% of the come group had smaller values compared wth the top 60% of the come group. Ths realty dcates that the demads of eergy goods the poorer households were less elastc. The creasg prce of eergy goods (reducg of the eergy subsdy) wll be burdeed more by the poorer households. Thus, a wder scope, the govermet eeds to formulate the codtoal eergy subsdy to protect the poor households. Ths paper also accommodated the mpact of the demographc varables o the demad for the eergy goods (o the budget share). Based o the sgfcace, oly the educato level of households head ad households locato affected the households decso to cosume eergy goods both come groups. The terpretatos were mxed sce the sg of coeffcets were dfferet for both come groups. Educato level egatvely affected the premum cosumpto the top 60% of the come group. The hgher educato level of the households head, the lower the possblty to cosume premum fuel. Ths ca be assumed that hgher educato would brg hgher come that ca lead to decreasg cosumpto of premum fuel ad hgher cosumpto to ether Pertalte fuel or Pertamax fuel. O the other sde, the hgher educato the bottom 40% of the come group would brg the possblty of hgher come that ca lead to the hgher access for havg vehcles. Dfferet from the top 60% of the come group, the bottom 40% households ted to cosume premum fuel (ths ca be see o the hgher come elastcty of demad for premum fuel for the bottom 40% of the come group). The households locato had the same coeffcet sgs the demad model for both come groups. The postve sg of the coeffcet o the premum meat that the urba area (code 1) had the smaller probablty to allocate more budgets to cosume premum fuel whle the rural area (code 2) had a bgger chace. LPG had a egatve sg o the coeffcet whch meat the urba area had the bgger probablty to allocate more budgets to cosume LPG. The sgfcat effects of those demographc varables o the demad for eergy goods proved that the eergy polcy should cocer ot oly the come groups but also the locatos ad educatoal backgroud of households head. I a bgger scheme, the eergy polcy should cosder the geographcal ad households backgroud. Of course, there are more demographc varables that should be aalyzed the demad model, so the result would be more comprehesve sce Tabalog Regecy s oly a small porto of samples from Idoesa. Ths paper has a lmtato term of ether the sample sze or the umber of observed eergy goods. Other papers utlze the AIDS-Cesored Model (Cesored Model s the model that cludes the mssg values of cosumptos for ts model) to overcome the complete cosumpto data; however, ths paper does t accommodate ths method sce the sze of samples are relatvely small, thus, t s better to obta more observatos to mprove the paper. Ackowledgemet I thak the BPS-Statstcs of Tabalog Regecy for the access to the Suseas data. I also exted my grattude for all of the Statstcas BPS-Statstcs Idoesa who volved the dscusso to esure that ths paper s well-costructed. I hghly apprecate the feedbacks from the aoymous revewers. All of the mstakes ths paper are my ow resposblty. Jural Ba Praa 9 (2) (2017):

13 V. Refereces Baks, J., Bludell, R., & Lewbel, A. (1997). Quadratc Egel Curves ad Cosumer Demad. Revew of Ecoomcs ad Statstcs, 79(4), Bazzaza, F., Ghasham, F., & Mousav, M. H. (2017). Effects of Targetg Eergy Subsdes o Domestc Electrcty Demad Ira. Iteratoal Joural of Eergy Ecoomcs ad Polcy, 7(2), Retreved from Bhakt, D. (2011). Permtaa Eerg Rumah Tagga d Pulau Jawa. Bogor Agrcultural Uversty. Retreved from hadle/ /51435 Bresger, C., Egelke, W., & Ecker, O. (2012). Leveragg Fuel Subsdy Reform for Trasto Yeme. Sustaablty, 4(11), do.org/ /su Chrstese, L. R., Jorgeso, D. W., & Lau, L. J. (1975). Trascedetal Logarthmc Utlty Fuctos. The Amerca Ecoomc Revew, 65(3), Retreved from stor.org/stable/ Dartato, T. (2013). Reducg Fuel Subsdes ad the Implcato o Fscal Balace ad Poverty Idoesa: A Smulato Aalyss. Eergy Polcy, 58, epol Deato, A., & Muellbauer, J. (1980). A Almost Ideal Demad System. The Amerca Ecoomc Revew, 70(3), Retreved from Drektorat Peyusua APBN, & Drektorat Jederal Aggara. (2017). Iformas APBN Jakarta: Mstry of Face of the Republc of Idoesa. Gudmeda, H., & Köhl, G. (2008). Fuel Demad Elastctes for Eergy ad Evrometal Polces: Ida Sample Survey Evdece. Eergy Ecoomcs, 30(2), org/ /.eeco Iteratoal Isttute for Sustaable Developmet. (2012). Padua Masyarakat tetag Subsd Eerg d Idoesa: Perkembaga Terakhr Iteratoal Isttute for Sustaable Developmet. J, Y., & Zhag, S. (2013). Elastcty Estmates of Urba Resdet Demad for Electrcty: A Case Study Beg. Eergy & Evromet, 24(7 8), Kle, L. R., & Rub, H. (1948). A Costat-Utlty Idex of the Cost of Lvg. The Revew of Ecoomc Studes, 15(2), Retreved from Maryoo, J. (2017). Determats of Demad for Foreg Toursm Idoesa. Jural Ekoom Pembagua, 18(1), org/ /ep.v Ncholso, W., & Syder, C. (2011). Mcroecoomc Theory: Basc Prcples ad Extesos. Oho: Nelso Educato. Po, B. P. (2012). Easy Demad-System Estmato wth Quads. Stata Joural, 12(3), Retreved from artcle.html?artcle=st0268 Rohac, D. (2013). Solvg Egypt s Subsdy Problem. Polcy Aalyss. Cato Isttute. Retreved from Statstcs Idoesa. (2016). Idoesa - Surve Sosal Ekoom Nasoal Statstcs Idoesa. Statstcs of Tabalog Regecy. (2017). Tabalog Regecy Fgures BPS-Statstcs of Tabalog Regecy. Stoe, R. (1954). Lear Expedture Systems ad Demad Aalyss: A Applcato to the Patter of Brtsh Demad. The Ecoomc Joural, 64(255), org/ / Sulsto, J., Wrabhuaa, A., & Wratama, M. G. (2017). Idoesa s Electrcty Demad Dyamc Modellg. I IOP Coferece Seres: Materals Scece ad Egeerg (Vol. 215). IOP Publshg. Thel, H. (1975). Theory ad Measuremet of Cosumer Demad. Amsterdam: North-Hollad Pub. Co. Wdaroo, A. (2013). Food Demad Yogyakarta: Suseas KINERJA, 17(2), kera.v Wdaroo, A. (2016). Modelg Sstem Permtaa utuk Peelta Ekoom dega SAS. Yogyakarta: UPP STIM YKPN. Y, K.-J., Geetha, C., & Chadra, V. V. (2017). Estmatg the Elastcty of Eergy Over Cosumpto at Mcro Level: A Case Study Sabah, Malaysa. Eergy Proceda, 105, Eergy Goods Demad Tabalog Regecy: Almost-Ideal Demad System Approach Ahmad Mura 319