機械産業主導の時代を考える : IO表による1935-73年期の一解釈, Machine-driven Japanese growth, 1935-73, as explored by historical IO tables

1935、1951、1960、そして1970 年のIO表を使い、この期間における日本工業化の歩みを観察する。このため上記のIO 表は基礎概念を共通化し、以下の21 産業分類を共用する（農林漁業 [以下、農業]、鉱業、食料品、繊維・身廻品 [以下、繊維]、木材・木製品、パルプ・紙製品 [以下、パルプ]、出版・印刷、皮革・ゴム、化学、窯業・土石製品、金属、機械、その他の製造品、建築・土木 [以下、建築]、電力・ガス・水道 [以下、公益業]、商業・金融・保険・不動産 [以下、商業]、運輸・通信、公務、その他のサービス、事務用品・梱包、分類不明）。論稿I は問題意識と観察過程を記述し、論稿II は基礎データとその推計工程を説明する。論稿I の観察結果は次のように要約される。（１）産業構造比（総生産額中または付加価値総額中の割合）：[a] 商業は終始上位を占め、金融等サービス業の大きな貢献を示唆；[b] 機械と金属は敗戦後いったん繊維にリードを許した後、高度成長期に急上昇；[c] （典型的な最終需要的商品である）繊維は敗戦直後まで上昇（付加価値は敗戦直後まで高値継続）の後、一貫して下降；[d] 食料品は（1951 年以降、低下趨勢の典型的素材産業である）農業に支えられ意外と高め；[e] 建築は高度成長期以降、機械、金属に次いで上位。（２）貿易比（輸出[もしくは輸入]額に占める割合、または国内需要に占める輸出[もしくは輸入]額の割合）：[a] 輸出額シェアは繊維が趨勢的に低下、代って金属と機械が増大；[b] 輸入額シェアは（原料供給産業としての）鉱業で趨勢的に上昇、1951 年の農業で一時的に高値、高度成長期の機械と運輸･通信で上昇；[c] 輸出産業としての役割は初期の繊維が高度成長後期の機械によって代替；[d] 輸入品依存度は（敗戦後急上昇の）鉱業に次ぎ農業が高め。（３）社会経済の体質変化：[a] 最終需要への影響（影響度）が大きいのは（どの産業でも必須の）事務・梱包作業、パルプ、金属、繊維、機械、建築など；[b] 最終需要の変化に対する鋭敏性（感応度）が高いのは商業、化学、金属、パルプ、農業、機械など；[c] 技術変化（レオンティエフ行列中のaij の変化）は[c-1] 終始大なのが化学、[c-2] 高度成長期に顕著なのが建築、（エネルギー生産増大と環境汚染防止のための）鉱業、および（意外だが）食料品、[c-3] 高度成長期末に低速化したのが金属、繊維、農業、公益、そして機械（ただし高水準を維持）、[c-4] （総合すれば）機械と化学が技術変化の主役；[d] （三角化計算によれば）、[d-1] (素材など製品の流れの上流に位置する)基礎型産業は運輸・通信、商業、公益、それに敗戦後の(とりわけ高度成長期末期の)化学、[d-2] (製品の流れの下流に位置する)最終型産業は繊維、食料品、[d-3] 最終型から基礎型に転向したのが金属と機械、[d-4] 最終型から中間型に移動したのが出版・印刷。以上を総括すれば、1930 年代以降の約40 年間における日本の工業化は、商業等のサービス産業の貢献に支えられて進み、生産面では機械産業に牽引されるところが大きかった。だが高度成長期末には、工業のリーダーは機械から化学に移りつつあった。, Historical IO tables of Japan for 1935, 1951, 1960 and 1970 have been used to observe the characteristics of the nation’s industrialization process during the period covered by the tables. For this purpose the tables have been adjusted, in as much as possible, to share the identical concepts as well as the same industrial classification as listed below: agriculture, forestry and fishery (Agr), mining (Mng), foods (Fds), textile and clothing (Txtl), woods and wood products (Wds), pulp and paper products (Plp), publishing and printing (Pbl), leather and rubber products (Lthr), chemicals (Chm), ceramics (Crm), metal products (Mtl). machinery (Mchn), other manufacturing (Oth-mfg), building and civil engineering (Bldg), electricity, gas and water supply (Elct), commerce, financing, insurance and real estate (Cmr), transportation and communication (Trsp), government (Gvt), other services (Oth-srv), stationery goods supply and packaging (Stry), and unknown (Unkw) Section I describes the issues concerned by the present paper as well as its major findings, whereas section II explains in detail how the basic data were prepared. A summary of main findings of Section I follows. (1) Industrial structure (respective industry’s share in total product value or in total value added) [a] Highest in Cmr, indicating high contribution of service sector such as financing; [b] Mchn and Mtl, having once fell behind Txtl, went up remarkably in the high-growth era; [c] Txtl, a typical final demand product, continued to grow until the end of WWII, the stating point of its landslide; [d] Fds recorded high scores, having been supported by Agr, a typical upstream industry; [e] Bldg scored high since the high-growth era, closely following Mtl and Mchn (2) Foreign trade (industry’s export [or import] share in total export [or import], and proportion of industry’s export [or import] in total domestic demand)[a] Export share of Txtl declined consistently, while that of Mtl and of Mchn went up;[b] Import share of Mng increased consistently, while that of Mchn and of Trsp increased during the highgrowth era, and that of Agr registered a high share temporarily in 1951;[c] Txtl gave up its role as the leader of export by the end of the high-growth era, having been replaced by Mchn;[d] Import dependency of Mng increased significantly after WWII, being followed by Agr.(3) Effects, if any, of technological change[a] Rasmussen’s index of power of dispersion（influence of an increase in industry’s unit output on the final demand) is topped by Stry, followed by Plp, Mtl, Txtl and Bdg;[b] Rasmussen’s index of sensitivity of dispersion （degree of responsiveness of industry’s output to a change in the final demand) is topped by Cmr, followed by Chm, Mtl, Plp, Agr and Mchn;[c] Extent of technological change (changes in aij in the Leontief matrix) ―― [c-1] was constantly high in Chm, [c-2] became significant during the high-growth era in Bdg, Mng (due to increase in energy demand as well as rising need for minimizing environmental disruption) , and (unexpectedly!) in Fds, [c-3] declined by the end of high-growth era in Mtl, Txl, Agr, Pbl and Mchn (while Mchn keeping its relatively high level) ;[d] Triangularization results of IO tables indicates industry’s relative position in the flow of production process:[d-1] upstream were Trsp and Cmr, and also Chm in the post-WWII (especially post-high-growth) era, [d-2] downstream was Txtl, [d-3] shifted from downstream to upstream in Mtl and Mchm, [d-4] shifted from downstream to in-between in Pbl. All in all, one may summarize that Japanese industrialization during the 1935-70 years was contributed much by the service sector, while it was, in production sphere, led significantly by machinery production, whose role however was increasingly challenged by chemicals toward the end of the high-growth era.