Chukyo University Institute of Economics Discussion Paper Series

Transcription

1 Chukyo Unverty Inttute of Economc Dcuon Paper Sere November 2014 No Contructon of a Mult-regonal Input-Output Table for Nagoya Metropoltan Area, Japan Mtuo Yamada

2 Contructon of a Mult-regonal Input-Output Table for Nagoya Metropoltan Area, Japan Mtuo Yamada * Chukyo Unverty, Japan Abtract Japan ha many nput-output table, natonal and regonal. We focu on the Nagoya metropoltan area, n whch Toyota Motor headquartered along wth many ndutre manufacturng tranport equpment, machne tool and other machnery. Th area ncluded n the regon covered by three prefecture; Ach, Gfu, and Me. To contruct a mult-regonal nput-output table from thee prefecture, we frt break down each prefecture table wth 186 ector nto everal maller ub-regonal table. Then we combne each table nto one mult-regonal nput output table, contng of 14 ub-regon. Tranacton value among ub-regon of each ector are etmated by the gravty-ras method, n whch the ntal value are obtaned by the gravty model. Ung the mult-regonal table, we dcu the tructural charactertc of the Nagoya metropoltan area. We then how a way to cope wth the varou ncontence of regonal defnton, admntratve and economc, n the nput-output analy. Key Word: Mult-regonal nput-output table, Gravty-RAS method, Average Propagaton Length, Nagoya metropoltan area The orgnal veron of th paper wa preented at the 22nd Internatonal Input-Output Conference held n Lbon, Portugal, from 15th to 18th July Th tudy wa fnancally upported by a Chukyo Unverty Grant for Specal Reearch Project (No ). * School of Economc, Chukyo Unverty, Yagoto Honmach, Showa-ku, Nagoya, , Japan emal: yamada@mecl.chukyo-u.ac.jp 1

3 1 Introducton In Japan all of 47 prefecture have releaed ther nput-output table (hereafter IOT) for every fve year begnnng wth The prefecture IOT have been effectvely ued for the evaluaton of many event, the polcy evaluaton of local government, and the economc plannng of the regon. Recently, the prefecture table are appled to comple them nto a mult-regonal IOT (hereafter MRIOT). Th tendency tem from the recognton that the mult-regonal nput-output analy become more mportant, becaue the regonal econome are trongly connected dometcally and nternatonally. In th paper, ung the IOT for the three prefecture of Ach, Gfu, and Me, repectvely, we recomple them nto one type of MRIOT for the Toka Regon coverng thoe prefecture. For th purpoe, we frt break down each prefecture table nto the table for everal mall area. Then the commodty flow among the area for each ector are etmated by the Gravty-RAS method, n whch the ntal value for RAS teratve calculaton are reduced from the etmated Gravty model. Fnally, we ntegrate all the mall area table nto a MRIOT, ung the etmated tranacton hare of the commodty flow. Th MRIOT for the Toka Regon cont of 186 ector for each of the 14 area. Breakng down the prefecture IOT nto thoe of everal mall area and the ntegraton of the mall area IOT nto one MRIOT may make t poble to olve the analytcal problem that a regon dentfed n term of economc actvty not necearly the ame a the admntratve regon. Nagoya the thrd larget metropoltan area n Japan, followng Tokyo and Oaka. Nagoya Cty, the captal cty of Ach Prefecture, the economc center of the Toka Regon. However, each urroundng area not equally connected to Nagoya Cty. Generally peakng, the cloer the dtance between area, the tronger ther economc connecton. Thu, ung the MRIOT for the Toka Regon, we nvetgate uch relaton by the Average Propagaton Length (APL) ndex propoed by Detzenbacher et al. (2005). In Secton 2, the related precedng tude n Japan are brefly dcued. Then n Secton 3, we ntroduce a method to recomple our MRIOT of the Toka Regon. Secton 4 dcue the tructural charactertc of the Toka Regon, and n Secton 5 two applcaton are conducted; one the nterregonal mpact of the Motor Vehcle Sector on the Value Added, and the other the charactertc of the Nagoya metropoltan 2

4 area n term of the APL ndex. Fnally, Secton 6 provde concludng remark. 2 Background In Japan, there are everal MRIOT-baed prefecture table. Table 1 how the typcal experence related to uch ue n three group. The frt one the nter-prefectural MRIOT for pecfc regon n Japan, whch nclude Kana Regon MRIOT (Kana Inttute for Socal and Economc Reearch (2008)), Tohoku Regon MRIOT (Tohoku Regon Advancement (2011)), Toka Regon MRIOT (Yamada (2010)), and Chubu Regon MRIOT (Chubu Regon Inttute for Socal and Economc Reearch (2011)). Thee table were made to nvetgate ome partcular regon n Japan, and ther ector ze commonly about nnety. The commodty flow between prefecture were etmated eentally by the RAS method wth dfferent tattcal ource for the ntal value. The econd group the nter-prefectural MRIOT coverng overall the 47 prefecture n Japan, ncludng Ihkawa and Myag (2003) for the 1995 Table, Htom (2008) for the 2000 Table, and Hagwara (2012) for the lnked MRIOT. Becaue the number of prefecture ha ncreaed, the ector ze wa decreaed to around ffty. The etmaton of commodty flow among prefecture alo depend on the RAS method. A the ntal value of the RAS method, the author ued the data of urvey tattc of dometc net freght flow (MLIT 1 ), the commodty flow data of METI 2, employee commutng flow of the Cenu, and the communcaton traffc data of MIC 3. The thrd group the ntra-prefectural MRIOT for the pecfc prefecture. The Ehme Prefecture MRIOT (Tubouch (1991)), Hokkado Prefecture MRIOT (Takahata (1991)), and Me Prefecture MRIOT (Yamada (1996)) were the poneerng tude. After 2000, everal new tude appeared uch a the 1995 Ach MRIOT (Ihkawa (2004)), 2000 Ach Prefecture MRIOT (Nakano and Nhmura (2007)), Yamaguch Prefecture MRIOT (Nomura et al. (2011)), and 2005 Ach Prefecture MRIOT (Yamada and Owak (2012) and Yamada (2013)). 1 Mntry of Land, Infratructure, Tranport and Tourm, Japan 2 Mntry of Economy, Trade, and Indutry, Japan 3 Mntry of Internal Affar and Communcaton, Japan 3

5 Table 1 Mult-Regonal Input Output Table n Japan Type of RIO Author Table name Sector Etmaton of Commodty flow Kana Inttute for 2000 Kana 7 prefecture/100 Specal urvey data on commodty Socal and Economc Interregonal IOT ector flow (METI) Reearch (2008) Aa Pacfc Inttute of Reearch (2012) 2005 Kana Interregonal IOT 1) Inter prefectural MRIOT for the Tohoku Regon pecfc regon wthn 2005 Tohoku Regon Advancement Center Japan MRIOT (2011) 2000 Toka Regon Yamada(2010) MRIOT 2) Inter prefectural MRIOT coverng Japan overall 3) Intra prefectural MRIOT for the pecfc prefecture Chubu Regon Inttute for Socal and Economc Reearch (2011) Ihkawa and Myag (2003) Htom (2008) Hagwara (2012) Tubouch (1991) Takahata (1992) Yamada (1996) Ihkawa (2004) Nakano and Nhmura (2007) Nomura et al. (2011) Yamada and Owak (2012), Yamada (2013) 2005 Chubu Regon MRIOT 1995 MRIOT for Japan 2000 MRIOT for Japan lnked MRIOT 7 prefecture/104 ector 6 prefecture/43 ector 3 prefecture/185 ector 10 prefecture/95 ector 47 prefecture/45 ector 47 prefecture/59 ector 47 prefecture/59 ector RAS method wth the urvey tattc of dometc net freght flow (MLIT ) and employee commutng flow of the Cenu (MIC), and other Specal urvey data on commodty flow (METI) RAS method wth the pecal urvey data of commodty flow (METI) RAS method wth the urvey tattc of dometc net freght flow (MLIT ) RAS method wth the urvey tattc of dometc net freght flow (MLIT ) Gravty model ung the freght flow data of MLIT and the commodty flow data of METI Extended RAS method wth the net freght flow of MLIT, employee commutng flow of the Cenu, and the communcaton traffc data of MIC Specal urvey data on the commodty flow n Ehme 1988 mult regonal IOT 6 area/50 ector of Ehme Prefecture 1985 mult regonal IOT of Hokkado Prefecture 4 area/61 ector Stattc on the freght flow and paenger flow 1985 Me Prefecture RAS method wth the urvey data of 5 area/84 ector MRIOT freght flow 1995 Ach Prefecture 3 area/46 ector Locaton Quotent (LQ) method MRIOT 2000 MRIOT for Ach Gravty model etmated ung the 3 area/46 ector Prefecture commodty flow of METI MRIOT 2000 a MRIOT for 3 area/104 ector Locaton Quotent (LQ) method Yamaguch Prefecture 2005 Ach Prefecture MRIOT 4 area/186 ector Gravty RAS method For the etmaton of commodty flow wthn a prefecture, the poneerng work depended on the tattc regardng the freght flow and paenger flow wthn the prefecture, or on pecal urvey conducted on the commodty flow of the good produced f there were no tattc for flow wthn the prefecture. In the cae of Me Prefecture, Yamada (1996) appled the RAS method wth the urvey data of freght flow a the ntal value of the teraton. Nakano and Nhmura (2007) appled the gravty model. Ihkawa (2004) and Nomura et al. (2011) ued the Locaton Quotent (LQ) method, 4

6 whch etmate the net tranacton between two area by the relatve advantage meaure 4. Yamada (2013) developed the Gravty-RAS method, n whch a gravty model wa ued to obtan ntal value of the RAS teraton. Here we extend the IOT to cover the three prefecture of Ach, Gfu, and Me (See Fgure 1). Th called the Toka Regon, whoe core cty Nagoya. Ung th IOT wth 14 area and 186 ector for each area, we are gong to nvetgate the economc tructure of the Nagoya metropoltan area, whch ncluded n the Toka Regon. How cloe each area to Nagoya Cty to be dcued. Fgure 1 Locaton of the Toka Regon (Fourteen Area) Chuno Hda Gfu Tono Prefecture Gfu Seno Nh- mkawa Me Prefecture Hokue Hgahmkawa Iga Chue Nane Owar Nagoya Cty Ach Prefecture Hgah -khu 4 There are everal varaton of the Locaton Quotent model for the etmaton of the export and mport of the regon (See Mller and Blar (2009)). 5

7 Table 2 how the area and populaton of the Toka Regon. More than 11 mllon peron lve n th regon. The populaton of Nagoya Cty telf 2.2 mllon, and t ha the hghet populaton denty, about 6,900 peron per quare km n th regon. Owar urroundng Nagoya Cty ha the econd hghet denty, about 2,000 peron per quare km. In Gfu Prefecture, the Gfu area ha the hghet denty, 814 peron per quare km, and the Hokue area ha 759 peron per quare km n Me Prefecture. Table 2 Area and Populaton of the Toka Regon Area Populaton Populaton n Populaton n Denty n km 2 Thouand Thouand Peron/km 2 Ach Prefecture 5, , , , Nagoya , , , Owar 1, , , , Nh-mkawa 1, , , Hgah-mkawa 1, Gfu Prefecture 10, , , Gfu Seno 1, Chuno 2, Tono 1, Hda 4, Me Prefecture 5, , , Hokue 1, Chue 1, Nane 1, Iga Hgah-khu Toka Regon 21, , , Contructon of a mult-regonal nput-output table In th ecton, we explan the outlne to contruct a mult-regonal IOT of the Toka Regon, whch cont of three prefecture; Ach, Gfu, and Me. Frt, each prefecture 6

8 table dvded nto everal area IOT of 186 ector. Thee table are mperfect n the ene that they nclude ntra-prefectural tranacton a net value. Second, we etmate the tranacton value of commodty flow among the area n two tep; the tranacton wthn each prefecture, and the tranacton among area of dfferent prefecture. The two-tepwe etmaton are adopted for the contency wth the dometc trade of each prefecture table. 3.1 Regonal decompoton of each prefecture IOT Here we explan the method to dvde a prefecture IOT nto everal IOT of the maller area. 1) Output value The output value of each maller area are etmated by multplyng the approprate rato to the prefecture value by ector a follow: X r X, where X the output value of the -th ector n the -th area, r denote the dvdng ndex of the -th ector n the -th area, and X mean the output value of the -th ector n a gven prefecture. In the manufacturng ector, for example, the product value of each muncpal are added to that of each area, the area hare of whch ued a the dvdng ndex. 2) Intermedate nput and value added Aumng that the nput coeffcent and value-added rato to the output are ame for each area wthn the prefecture, the ntermedate nput value and value added of each ector are etmated a follow: X ax, j j V v X, j j 7

9 where X j the ntermedate nput of the -th commodty for the j-th ector n the -th area, a j denote the nput coeffcent of the -th commodty for the j-th ector n the prefecture, V j mean the -th value added of the j-th ector n the -th area, and the -th value added rato of the j-th ector n the prefecture. v j 3) Dometc fnal demand The dometc fnal demand cont of prvate conumpton 5, government conumpton, prvate nvetment, publc nvetment, and change of nventory. The prvate conumpton of each area etmated by multplyng the approprate rato to the prefecture value by ector a follow: C C, where C the conumpton of the -th commodty n the -th area, and denote the dvdng ndex of the -th ector n the -th area, and C mean the conumpton of the -th commodty n the prefecture. The dvdng factor etmated a follow: cn cn, j 1 j1 2 j2 where c1 and c 2 are the conumpton for each houehold of the -th commodty n the prefecture, for two-or-more-peron houehold and one-peron houehold, repectvely, whch are obtaned from the Natonal Survey of Famly Income and Expendture, MIC. n j1 and j2 n mean the number of houehold; two-or-more-peron and one-peron, repectvely. The etmaton of government conumpton, prvate nvetment, and publc 5 Japanee IOT, natonal and regonal, nclude conumpton expendture outde houehold a the exogenou ector, n both value-added part and fnal demand part. We etmate them n the value-added part aumng that the rato to output for each ector the ame a thoe of the prefecture. We are able to acqure the total value n the fnal demand part by ummng the ectoral value n the value-added part. Then each value of the tem n the fnal demand part etmated by multplyng the total value by the correpondng hare value, aumng the hare value are the ame for all area n the prefecture. 8

10 nvetment conducted n almot the ame way o a to dvde them nto thoe of the area by ung the approprate dvdng hare. The change of nventory etmated aumng that the nventory change rato to the output n an area ame n the prefecture. J ( J / X ) X where J the nventory change of the -th ector n the -th area, and J the correpondng value of the prefecture. 4) Internatonal trade Here we mut etmate the export and mport value for each mall area. We aume that the export rato to the output for each ector the ame wthn the prefecture. Then the export for each area reduced a follow: E e X, where E mean the export of the -th commodty n the -th area, and e denote the export rato to output by ector n the prefecture, that e E / X. Here E the export of the -th ector n the prefecture. Aumng that the mport coeffcent for each area the ame wthn the prefecture, the mport of the area reduced a follow: M md, where M mean the mport of the -th commodty n the -th area, m denote the mport coeffcent by ector n the prefecture, and D how the dometc demand of the -th commodty n the -th area. The mport coeffcent of the prefecture defned a m M / D, where M the mport of the -th ector and D the dometc demand of the -th ector n the prefecture. 5) Internal trade wth outde of the prefecture 9

11 We etmate the export and mport value wth the ret of Japan for each area. Aumng that the export rato to the output for each ector ame for the nternal trade wth the ret of Japan, the export for each area reduced a follow: E e X, * * where * E mean the -th commodty export to the ret of Japan n the -th area, and * e denote the correpondng export rato to output by ector n the prefecture, e E X * * /. Here * E the export of the -th ector n the prefecture. For the mport, we aume that the mport coeffcent from the ret of Japan the ame for each area wthn the prefecture. Then the mport of the area reduced a follow: M md, * * where * M mean the -th commodty mport n the -th area, and * m denote the mport coeffcent by ector n the prefecture. The mport coeffcent of the prefecture defned a m M D * * /, where * M the -th commodty mport from the ret of Japan. 6) Balancng equaton After etmatng each tem by ector, the total demand mut be equal to the total upply. Then the followng equaton hold for each ector n each area: D E E M M N X, * * where D Xj F. Here D denote the total dometc demand of the -th commodty n the -th area, whch cont of the ntermedate demand X and the fnal demand j F. The R varable N mean the net export to the other area wthn the prefecture, whch work a the balancng factor after etmatng all the other tem ndependently. 10

12 3.2 Etmaton of commodty flow between area Here we ue the gravty-ras method by whch the etmate from the gravty model are ued a the ntal value for the RAS teraton to obtan the commodty flow among area. Table 3 how the llutratve flow of the -th commodty among four area, two of whch are ncluded n each of two prefecture, repectvely 6. Two regon, the ret of Japan and the ret of the world, appear n the table to capture the dometc and nternatonal tranacton. how the tranacton of the -th commodty from the r-th area to the -th area, denote the total value of tranacton of the -th commodty from the r-th area, and mean the total value of tranacton of the -th commodty to the -th area. E and M are repectve export and mport to and from the world, and *o E and *o M are export and mport to and from the ret of Japan, whch defned a the regon excludng the two prefecture n Japan n th cae. Table 3 Inter-regonal flow of the -th commodty Detnaton Prefecture-1 Prefecture-2 Ret Ret Total \ Area-1 Area-2 Area-3 Area-4 of of the Orgn Japan World Prefecture-1 Area-1 T Prefecture-2 Area-3 T T T T E E Area-2 T T T T E E T Area-4 T T T T E E T T T T T E E T Ret of Japan M M M M - - M Ret of the World M M M M - - M Total T T T T E E T The followng equaton mut be condered. r * ro r r for r=1,2,3, and 4, (1) T E E T 6 Actually, there are not two area for two prefecture but fourteen area for three prefecture n all. 11

13 r * o for =1,2,3, and 4. (2) r T M M T In th table, the value of the followng varable are known from each of the decompoed IOT. 1) Export E and mport M are known for =1, 2, 3, and 4. 2) Dometc tranacton E * and relaton, * M wth the ret of Japan are known wth the Prefecture-1: Prefecture-2: r * ro * r T 1 E E for r=1, 2 (3) r * ro * r T 2 E E for r=3, 4. (4) 3) The total value uppled from the r-th area to all area and regon r T equal to the output of the area r X, and the total value receved from all area and regon to the r-th area T ame a the total demand of the -th area, D, that, T r X, (5) r T D. (6) Here, r X and D are obvouly obtaned from the dvded IOT. Conderng the above retrcton, we mut determne the tranacton value r T n an approprate way. To olve th model, we ue a two-tep etmaton trategy that dtnguhe the ntra-prefectural tranacton from the tranacton between prefecture n the etmaton Etmaton of ntra-prefectural flow Table 4 how the ntra-prefectural flow of the -th commodty, and the row-um value and column-um value are obtaned from each decompoed IOT. Holdng the followng 12

14 equaton, we are able to determne the value of ntra-prefectural tranactont. r r r * r r T 1,2/3,4 X E E (7) r * T 1,2/3,4 D r M M (8) Then we apply the RAS method wth approprate ntal value. In th table, we apply the gravty model, whch to be explaned n ecton 3.3.3, to obtan the ntal value. Table 4 Intra-prefectural flow of the -th commodty Detnaton Prefecture-1 Prefecture-2 Total Orgn\ Area-1 Area-2 Area-3 Area-4 Prefecture-1 Prefecture-2 Total Area-1 T T - - X -E -E Area-2 T T Area T - - X -E -E T X -E -E Area T T X -E -E D -M -M D -M -M D -M -M D -M -M Etmaton of nter-prefectural flow After etmatng the ntra-regonal commodty flow, we mut determne the value of nter-prefectural commodty flow, a hown n Table 5. Table 5 Inter-prefectural flow of the -th commodty Detnaton Prefecture-1 Prefecture-2 Total \ Area-1 Area-2 Area-3 Area-4 Orgn Prefecture-1 Area T Area T Prefecture-2 Area-3 T Area-4 Total T T T T T E -E E -E - - E -E - - E -E M -M M -M M -M M -M In the followng equaton, both export and mport to and from the regon outde of both prefecture *ro E and *o M are not known, though export and mport wth the 13

15 ret of Japan, *r E and * M, are obtaned from the IOT. r * r * ro T 1,2/3,4 E E (9) r * * o T r 1,2/3,4 M M (10) To obtan ther value, we multply the predcted hare by the gravty model to the total value a follow: Eˆ Eˆ * ro * ro * r E ˆ r ˆ * ro T E, (11) Mˆ Mˆ * o * o * M ˆ r ˆ * o T r M, (12) where the varable wth hat denote the predcted one by the gravty model wth ome modfcaton to fulfll the condton that the um of the total column equal to that of the total row 7. Th condton mean that the um of the export wthn prefecture concerned equal to the um of the mport wthn the ame prefecture. Here we apply the gravty-ras method to obtan the tranacton value r T, that keep the condton expreed by the equaton (9) and (10), wthn the concerned prefecture The modfed gravty model The gravty model how that the volume of tranacton from regon r to regon r T proportonally related to the total volume of both the orgn r T and the detnaton T and dproportonally to the dtance between the regon 8 r L a follow: 7 Here we adopted the adjutng rule that the larger um adjuted to the maller um by decreang each tem proportonally. A a reult, ether export hare or mport hare are modfed from the etmated value. 8 There are ome varaton to how the dtance between regon: geographcal dtance, tme dtance, and cot dtance, and o on. Yamada and Owak (2012) adopted the geographcal dtance. r They meaured the dtance L between two pont, each of whch belong to the dfferent regon to be meaured, by the root earchng n the Google webte. Then they calculated the average dtance 14

16 T r r r T T r L k, (13) where,, and are parameter of the two volume varable and dtance varable, repectvely. Takng the logarthm on both de of the equaton (13), the parameter are etmated from the tranacton data among 9 regon of the 2005 METI MRIOT for each of 186 ector. Becaue ome ector have no tranacton wth the other regon, 153 ector are etmated. Generally peakng, the tranacton wthn each regon are not ncluded n the gravty model. However, we etmated the model wth the tranacton data not only between regon but alo wthn each regon, becaue we needed ome etmate for the tranacton wthn regon. Table 6 decrbe the charactertc of the etmated parameter. Almot all of the parameter are etmated to be tattcally gnfcant. The average value of the volume parameter for the orgn and the detnaton are 0.99 and 0.74, repectvely. The average of dtance parameter The dtance average of the commodty ector 1.12, whch lower than that of the ervce ector, 1.77, a expected. Fgure 2 how the dtrbuton of each etmated parameter. We are able to confrm that the dtance parameter are cattered wder than the two volume parameter, and that dtance parameter of ervce ector are more wdely dpered than thoe of commodty ector. Table 6 Charactertc of the etmated parameter Total volume of the Total volume of the orgn detnaton Dtance Maxmum Value Mnmum Value Average of Commodty ector of Servce Sector Standard Devaton No. of Equaton No. of Sgnfcant Value RS L wth the weght of the employment number a follow: r r LEE RS rr S L r E E rr S 15

17 Fgure 2 Dtrbuton of the etmated parameter Dtance, commodte Dtance, ervce Vol. of Orgn Vol. of Detnaton Dtance For our purpoe, we appled the gravty model aymptotcally to nduce the tranacton coeffcent t ˆr between area of our IOT, and the etmate were ued a ntal value for the RAS teratve method to etmate the tranacton value r T. tˆ r Tˆ r ˆ r T, ˆ r T kˆ r r ˆ ˆr ˆ T T L ˆ ˆ r ˆ where ˆ, ˆ, ˆ, and k ˆ r are the etmated parameter. The varable Tˆ r, ˆ T, and ˆr L 9 are value correpondng to the MRIOT compled here. 9 The average dtance were calculated from the dtance between muncpal that belonged to each 16

18 Table 7 The etmated tranacton matrx Unt: bllon Yen, % Tranacton Flow Export Export, Nhmkawmkawkhu Hgah- Hgah- (->) Nagoya Owar Gfu Seno Chuno Tono Hda Hokue Chue Nane Iga to the Internatonal ROJ Total 1 Nagoya Owar Nh-mkawa Hgah-mkawa Gfu Seno Chuno Tono Hda Hokue Chue Nane Iga Hgah-khu The Ret of Japan Import Total Export Export, Share of Orgn Nhmkawmkawkhu Hgah- Hgah- Nagoya Owar Gfu Seno Chuno Tono Hda Hokue Chue Nane Iga to the Internatonal ROJ Total 1 Nagoya Owar Nh-mkawa Hgah-mkawa Gfu Seno Chuno Tono Hda Hokue Chue Nane Iga Hgah-khu The Ret of Japan Import Total Export Export, Share of the Detnaton Nhmkawmkawkhu Hgah- Hgah- Nagoya Owar Gfu Seno Chuno Tono Hda Hokue Chue Nane Iga to the Internatonal ROJ Total 1 Nagoya Owar Nh-mkawa Hgah-mkawa Gfu Seno Chuno Tono Hda Hokue Chue Nane Iga Hgah-khu The Ret of Japan Import Total Table 7 how the etmated tranacton matrx among area for all commodte. In th table, there are export and mport, nternatonal and dometc, of the Toka regon. The export hare and mport hare are alo calculated. The hadow cell mean the area whoe tranacton hare ha amount to 5 percent and more. Lookng at thee hare, we note that four area n Ach Prefecture have trong relaton among them. Alo there appear to be trong nterdependence among Gfu, Seno, Chuno, and Tono area n Gfu Prefecture, though the Hda area eem to have a lghtly weak relaton to them. In Me Prefecture, the Hokue and Chue area have relatvely trong lnk. area wth employment weght. 17

19 There are relatvely trong relaton between the Owar area n Ach Prefecture and three area n Gfu Prefecture; Gfu, Seno, and Chuno. The Hokue area n Me Prefecture and Nh-mkawa area n Ach Prefecture are alo lnked. 3.3 The compled mult-regonal nput-output table We dvde three prefecture IOT wth 186 ector nto fourteen area IOT, and etmate the tranacton flow among area and the outde regon, dometc and nternatonal for each ector. Here we are able to combne the fourteen table nto one mult-regonal IOT ung the ectoral tranacton flow matrce. Table 8 how the MRIOT compled for the Toka Regon wth fourteen area, n whch 186 ector are ntegrated nto one ector. Table 8 The etmated IOT for Toka Regon 14 area and one ector Unt: bllon Yen Intermedate Demand Nagoya Owar Intermedate Nhmkawmkawa Hgah- Hgahkhu Demand, Total 1 Nagoya Owar Nh-mkawa Hgah-mkawa Gfu Seno Chuno Tono Hda Hokue Chue Nane Iga Hgah-khu The Ret of Japan Import Intermedate Input Value Added Total Input Fnal Demand Import, Nagoya Owar Export Export, Dometc and Output Nhmkawmkawkhu Hgah- Hgah- Gfu Seno Chuno Tono Hda Hokue Chue Nane Iga to the Internatonal Inter- ROJ natonal 1 Nagoya Owar Nh-mkawa Hgah-mkawa Gfu Seno Chuno Tono Hda Hokue Chue Nane Iga Hgah-khu The Ret of Japan Import Total

20 4 Comparon of output tructure In th ecton, we ntegrate the MRIOT wth 14 area and 186 ector for each area to that wth 14 area and 40 ector, and dcu the charactertc of the output tructure n Table 9, 10, and 11. Thoe table contan ectoral output and the Revealed Comparatve Advantage (RCA) ndce of output n the Toka Regon. Table 9 Output Structure of Ach Prefecture Output (Bllon Yen) Revealed Comparatve Advantage Nagoya Owar Nhmkawa Hgahmkawa Ach Pref. Nagoya Owar Nhmkawa Hgahmkawa 01 Agrculture Foretry Fhery Mnng Beverage and Food , Textle product Pulp, paper and wood product Chemcal product , Petroleum and coal product Platc product , Pottery, chna and earthenware Other ceramc, tone and clay product Iron and teel , , Non-ferrou metal Metal product , General machnery , , Electrcal machnery , Informaton and communcaton electronc equpment Electronc component , Motor vehcle , , , , Arcraft and repar of arcraft Other tranportaton equpment Precon ntrument Mcellaneou manufactured product , Contructon 1, , , Electrcty, ga and heat upply , Water upply and wate dpoal bune Commerce 5, , , , Fnance and nurance 1, , Real etate 1, , , Tranport 1, , , Informaton and communcaton 1, , Publc admntraton , Educaton and reearch , Medcal ervce, health, ocal ecurty and nurng care 1, , Other publc ervce Bune ervce 2, , Peronal ervce 1, , Offce upple Actvte not elewhere clafed Sector, total 23, , , , , Table 9 how that Nagoya Cty ha relatve advantage n Informaton and communcaton, Commerce, Fnance and nurance, Real etate, and Bune ervce. Thoe ector are trongly related to the core ndutre n the metropoltan area. On the other hand, the Owar area ha advantage n Petroleum and coal product, Pottery, chna and earthenware, Arcraft and repar of arcraft, and Mcellaneou manufactured product. The Nh-mkawa area, where Toyota Motor Co. headquartered along wth t factore and group compane, ha advantage 19

21 n Informaton and communcaton electronc equpment, Electronc component, and Motor vehcle. The Hgah-mkawa area ha advantage n Agrculture, Motor Vehcle, Other tranportaton equpment, and Precon ntrument. Table 10 how the output tructure of fve area n Gfu Prefecture. The Gfu area, where the captal cty of the prefecture located, ha trong ndutre n Arcraft and repar of arcraft, and Textle product. Htorcally, textle ndutre propered n th area. The Seno area, located wet of the Gfu area, ha advantage n t manufacturng ndutre a Other ceramc, tone and clay product, Metal product, Electrcal machnery, Precon ntrument, Textle product, and Platc product n addton to the Mnng. The Tono area, n eatern Gfu Prefecture, ha advantage n manufacturng ndutre lke Pulp, paper and wood product, Metal product, General machnery, and Informaton and communcaton electronc equpment, along wth Foretry. On the other hand, the Hda area, located n northern Gfu Prefecture ha much natural reource, uperorty n Foretry, Fhery, Mnng, Pulp, paper and wood product, and Electrcty, ga and heat upply, that generated by hydropower taton. Table 10 Output Structure of Gfu Prefecture Output (Bllon Yen) Revealed Comparatve Advantage Gfu Seno Chuno Tono Hda Gfu Pref. Gfu Seno Chuno Tono Hda 01 Agrculture Foretry Fhery Mnng Beverage and Food Textle product Pulp, paper and wood product Chemcal product Petroleum and coal product Platc product Pottery, chna and earthenware Other ceramc, tone and clay product Iron and teel Non-ferrou metal Metal product General machnery Electrcal machnery Informaton and communcaton electronc equpment Electronc component Motor vehcle Arcraft and repar of arcraft Other tranportaton equpment Precon ntrument Mcellaneou manufactured product Contructon , Electrcty, ga and heat upply Water upply and wate dpoal bune Commerce , Fnance and nurance Real etate Tranport Informaton and communcaton Publc admntraton Educaton and reearch Medcal ervce, health, ocal ecurty and nurng care Other publc ervce Bune ervce Peronal ervce Offce upple Actvte not elewhere clafed Sector, total 5, , , , , , Table 11 Output Structure of Me Prefecture 20

22 Output (Bllon Yen) Revealed Comparatve Advantage Hokue Chue Nane Iga Hgahkhu Me Pref. Hokue Chue Nane Iga Hgahkhu 01 Agrculture Foretry Fhery Mnng Beverage and Food Textle product Pulp, paper and wood product Chemcal product 1, , Petroleum and coal product 1, , Platc product Pottery, chna and earthenware Other ceramc, tone and clay product Iron and teel Non-ferrou metal Metal product General machnery Electrcal machnery Informaton and communcaton electronc equpment Electronc component , Motor vehcle 1, , Arcraft and repar of arcraft Other tranportaton equpment Precon ntrument Mcellaneou manufactured product Contructon , Electrcty, ga and heat upply Water upply and wate dpoal bune Commerce Fnance and nurance Real etate Tranport Informaton and communcaton Publc admntraton Educaton and reearch Medcal ervce, health, ocal ecurty and nurng care Other publc ervce Bune ervce Peronal ervce Offce upple Actvte not elewhere clafed Sector, total 9, , , , , Table 11 how the output tructure of fve area n Me Prefecture. Hokue area, the northern part of Me Prefecture and next to the Owar area, ha advantage n t Chemcal product, Petroleum and coal product, Iron and teel, Non-ferrou metal, Informaton and communcaton electronc equpment, and Motor vehcle. Contrarly, the Chue area, located n the central part of Me Prefecture, ha advantage n not only manufacturng ndutre lke Electronc component, Mcellaneou manufactured product, and Other ceramc, tone and clay product, but alo the tertary ector uch a Informaton and communcaton, and Publc admntraton. In the Nane area, Arcraft and repar of arcraft ector one of the compettve ndutre, though t producton not o large. Fhery alo ha a large ector n th area. The Iga area, deep n Me Prefecture, relatvely cloe to the Kana Regon and boat of Pottery, chna and earthenware, Platc product, Pulp, paper and wood product, and General machnery. In the Hgah-khu area, Foretry and Fhery ector have relatve advantage, and the Pulp, paper and wood product ector alo ha relatvely larger producton. In th area there are thermal power plant, o the producton of Electrcty, ga and heat upply ector alo extenve. 21

23 5 Some applcaton 5.1. Interregonal mpact of the motor vehcle ector on value added In th ecton, we dcu the nterregonal mpact of the motor vehcle ector, calculatng the nduced value added, by ector and regon, of one-unt ncreae of the motor vehcle ector fnal demand for each area. Toka MRIOT wth 40 ector by 14 area ued. Table 12 how the reult. Nagoya, Owar, and Nh-mkawa are the area where the motor vehcle ector concentrated. Toyota Motor Co. headquartered n the Nh-mkawa area. We oberve that thoe area gan a relatvely hgh hare n value added. The Gfu area n Gfu Prefecture and the Hokue area n Me Prefecture alo obtan a hgh hare of value added. The Honda Motor Co. man factory located n the Hokue area. Table 12 Induced VA and mport of one-unt ncreae of fnal demand, motor vehcle ector Nagoy a Ach Prefecture Gfu Prefecture Me Prefecture Hgah Nhmkawa Hokue Owar - Gfu Seno Chuno Tono Hda Chue Nane Iga mkawa Nagoya Owar Nh-mkawa Hgah-mkawa Gfu Seno Chuno Tono Hda Hokue Chueu Nane Iga Hgah-khu Hgah -khu Ret of Japan Import Total Fgure 3 ndcate the geographcal dffuon of the nduced value added for one-unt ncreae of the motor vehcle ector n the Nh-mkawa area, whch ha the larget gan n the value added. The Owar area, adjacent to the Nh-mkawa area, ha a relatvely hgher value added, and the other neghborng Hgah-mkawa and Hokue area are not far behnd. Gfu, Seno and Chuno n Gfu Prefecture and Chue n Me Prefecture how ome gan, though the effect for each not o much. Roughly peakng, the gan n value added are nverely correlated wth the dtance from the Nh-mkawa area. Fgure 3 Geographcal dffuon of nduced value added, 22

24 motor vehcle ector n the Nh-mkawa area Fgure 4 how a mlar geographcal dffuon for the one-unt ncreae of the motor vehcle ector n the Hokue area, whch another ntegrated area of the motor vehcle ndutry. The Hokue area connected trongly to Nagoya, Owar, and Nh-mkawa, where many motor vehcle ndutre are located. However, t ha lttle economc connecton to area n Gfu Prefecture, at leat va the motor vehcle ndutry. The mlar geographcal dffuon for one-unt ncreae of the motor vehcle ector n the Gfu area appear n Fgure 5. In th cae, the Nh-mkawa area gan are the larget among the neghborng area, and Owar and Nagoya are cloe behnd. Chono, Seno, and Hgah-mkawa obtan ome gan. Tono and Hokue have a lttle hare. The Gfu area connected more cloely to the Nh-mkawa area than the adjonng area. Alo, we are able to oberve the relaton wth the Hokue area n Me Prefecture, 23

25 whch aymmetrcal to the cae of Fgure 4. Fgure 4 Geographcal dffuon of nduced value added, motor vehcle ector n the Hokue area 24

26 Fgure 5 Geographcal dffuon of nduced value added, motor vehcle ector n the Gfu area 5.2 Average Propagaton Length The Average Propagaton Length (APL) ndex, propoed by Detzenbacher at al. (2005), ued to meaure the economc dtance between ndutre. Th ndex defned a follow. When we expre the Leontef nvere matrx L for the tandard Leontef quantty model denoted a follow: 25

27 L=(I-A) = I+A+A +A + where A mean the nput coeffcent matrx of the model, the overall ndrect effect are expreed a L-I=A+A +A +A +. On the other hand, we can defne the weghted um of the ntermedate nput, wth the number of each dffuon tage a the weght, a follow: Φ = A + 2A +3A + 4A + = L(L - 1) Then the average propagaton length (APL) APL j between the j-th ector and the -th ector n the model defned a APL j = [L(L - I)] j / [L - I] j where [L - I] j denote the -j element of the matrx L-I. Calculatng the APL between the j and for the model wth 40 ector for each of 14 area, we have ntegrated them a the ndex between the r-th area and -th area, r APL, APL = v APL f, r r j j r j where r v the hare vector of the value added n regon r and f j the hare vector of the fnal demand n regon. The APL ndex uually nterpreted the dtance between ndutre, a Dezenbacher et al. (2005) orgnally appled to a regonal IOT. Here we are able to obtan APL wth repect to the regon, and examne t a the ndex to meaure the patal dtance. The regonal comparon of the APL ndex appear n the analy ung the nternatonal nput-output model (Detzenbacher et al. (2007)). However, we compare the APL wth repect to area to the geographcal dtance to meaure the relaton between the core cty and the urroundng area n the MRIOT. Fgure 6 how the relaton between the APL and the logarthm of the geographcal dtance. The potve relaton oberved wth the correlaton coeffcent The trangle pont n Fgure 6 denote the APL dtance of Nagoya Cty to the other area n the forward drecton, and the quare pont how them n the backward drecton. Both how that the APL dtance of Nagoya Cty become larger a the geographcal dtance become longer. The area cloer to Nagoya Cty ha deeper connecton wth the cty and the tranacton between them become larger, whch make the APL dtance lower. 26

28 Fgure 6 Average propagaton length and dtance APL 1.5 APL Nagoya (B) Nagoya (F) Dtance log (km) We etmate the regreon of the APL on the logarthm of the dtance wth ome prefectural dumme from 196 ample. The etmated equaton APL ln(d ) D D r r F F B B G G G G D D D D, F F B B F F B B M M G G N N N N where D r the geographcal dtance between r and. F DG and B D G denote dummy varable of Gfu Prefecture, n the forward and backward drecton, repectvely. F D M 27

29 and B D M are alo dummy varable for Me Prefecture, and D F N and B D N are for Nagoya Cty. The etmated reult are hown n Table 13. The relaton gnfcant n the ene of a relatvely hgh determnant coeffcent, The APL ndex potvely correlated to the logarthm of the geographcal dtance. Among the prefectural dummy varable, the dummy of Gfu Prefecture n the forward drecton not gnfcant, though the other are. The prefectural dumme are negatve, whch mean that the ndutre of Ach Prefecture are more cloely concentrated than the other two prefecture. Nagoya Dummy alo not gnfcant, whch mean there no dfference to Ach prefecture. Table 13 Regreon reult of the APL on the logarthm of the dtance Dependent Varable: APL Method: Leat Square Sample: Included obervaton: 196 Varable Coeffcent Std. Error t-stattc Prob. C LOG (DISTANCE) GIFUF GIFUB MIEF MIEB NAGOYAF NAGOYAB R-quared Mean dependent var Adjuted R-quared S.D. dependent var S.E. of regreon Akake nfo crteron Sum quared redual Schwarz crteron Log lkelhood Hannan-Qunn crter F-tattc Durbn-Waton tat Probablty (F-tattc) Fgure 7 how the APL dtance from Nagoya Cty, whch here defned a the 28

30 average of two varaton; the forward and backward drecton: APL = APL APL 2 1 * r r r In th fgure, relatvely wde area urroundng Nagoya Cty are lnked to Nagoya Cty. Among them, the Hda area n Gfu Prefecture ha lttle connecton to Nagoya Cty. The Owar area trongly connected to Nagoya Cty, and Nh-mkawa, and Hgah-mkawa n Ach Prefecture alo have connecton. Chue and Iga n Me Prefecture alo have relatvely tronger lnkage to Nagoya Cty than Hokue, though the geographcal dtance are n revere order. The APL ndex ueful to meaure the degree of lnkage among area. Fgure 7 APL dtance from Nagoya Cty 29