Economic Analysis Series No.149
Economic Growth and the Characteristics of Technology in Japan

March, 1997
Hikaru Sakuramoto (Senior Visiting Fellow)
Kazushige Shimpo (Visiting Fellow)
Mikio Suga
Naoyuki Kainuma
Katsumi Hirashita
Yoshihito Urashima
Hirofumi Nishu

The full text is written in Japanese.


1  Objectives

The aim of this study is to examine empirically the characteristics of economic growth in Japan for the period from 1960 to 1990, from the perspective of technological progress and the spread of productivity growth across sectors. This study is motivated by two main reasons. Firstly, it is initiated by the recent development in the 'New Growth Theory' highlighting the endogenous growth factor. The second reason lies in the need to re-examine the results of past studies using economic observations of the Japanese economy after the oil crisis.

The first objective of this study is to calculate productivity growth by sector and its contributing factors using the input-output table. The second objective is to estimate the spillover effect of productivity growth from one sector to another through inter-sectoral dependency.

2  Main Characteristics of the Analysis
a  Triangularization of Sectors

By rearranging the sectors producing intermediate inputs in the input-output table topdown from products requiring a high degree of processing (final goods) to products requiring alow degree of processing (intermediate goods, in particular basic materials), it is found that final goods are seldom used as intermediate inputs in the production of others goods. On the other hand, while intermediate goods are frequently used as inputs in the production of other goods, they are less likely to be demanded as final goods. Hence, the sectoral spillover of productivity growth tends to lie in one direction. This implies that productivity growth in products requiring a high degree of processing will have no effect on products requiring a low degree of processing, whereas productivity growth in products requiring a low degree of processing will have an effect on products requiring a high degree of processing.

b  Interdependency Between Products

Products could be rearranged into several blocks via triangularization. A kind of block interdependency is found to exhibit through the use of intermediate inputs among the blocks. As shown in Table 5 on page 45, the 12 major product blocks could be further divided into 50 product blocks. Further, the formation of hierarchies through the structure of the intermediate goods was observed within the 12 major blocks and within the 50 smaller blocks (ordering among blocks). In addition, the ordering of products with a low degree of processing at the bottom and products with a high degree of processing at the top according to the ordering in triangularization is also observed within each block.

c  Measurement of the Rate of Technological Progress

By defining the rate of technological progress as the growth in total factor productivity (TFP), TFP of 50 sectors are measured for the period between 1960 and 1990.

3  Results
a  Growth Characteristics of the Japanese Economy
  • i. Capital input had the largest contribution to growth, with the rate of increase(7.6%) in capital input being greater than the rate of output increase(6.8%).
  • ii. Labour input had a 13% contribution to growth, and an annual growth rate of 1.7%. This could be broken down into the 1.0% annual growth of labour force in man-hour base and the 0.7% annual growth rate of improvement in labour quality.
  • iii. Technological progress recorded an average annual growth rate of 2.2%, and its contribution to growth was large at 33%. The rapid slowdown in the rate of technological progress after 1975 constituted one of the reasons for the slow economic growth.
b  Static Spillover Effects of Technological Progress and Productivity Growth

The rate of price increase of an output from a particular sector is measured by subtracting the rate of technological progress in that sector from the product of the share of material inputs and the rate of price increase in material inputs. For instance,

Rate of output price increase = Share of the respective material inputs
* Rate of price increase of the respective material inputs
+ Share of labour input * Rate of wage increase
+ Share of capital input * Rate of price increase in capital.

Therefore, by the above relationship, we could measure the effect of technological progress in one sector in reducing the price of output of another sector. This is termed the static spillover effect.

c  The Relationship Between Static Spillover Effects and Structural Changes

The static spillover effect has a close relationship with structural changes in the Japanese economy via the changes in resource allocation among the various systems of basic materials. In an historical perspective, an increasing amount of resources was allocated to sectors in which there was productivity growth in the sectors themselves, and to sectors which contributed to the price reduction in product sector requiring a high degree of processing via spillover obtained through the basic materials used. For the period from 1960 to 1985, large output increases were observed in sectors with productivity increase such as ore mining (h1) and secondary energy products (I). On the other hand, stagnation was observed in sectoral blocks belonging to basic material sectors such as agricultural products (h3), fisheries (h4), livestock products (h5) and materials for natural textiles (h6), in which there had been no productivity improvements. Hence, resources were allocated from sectors with low productivity to sectors with high productivity.

d  Dynamic Spillover Effects of Productivity Growth

The static spillover effect resulted in the price reduction in the final goods, in particular investment goods. While the price reduction in the investment goods resulted in price reduction of capital services through the static spillover effects, we termed the ultimate price reduction effect exerted on the other sectors as the dynamic spillover effects.

On the other hand, the employment coefficients had decreased rapidly in all sectors while labour productivity had increased. Although the capital coefficients had decreased in a few sectors, they had increased in most other sectors. This phenomenon was in particular distinguishing during the high growth era.

e  Spillover Effect of the Computer Sector (b5)

Sectors that had experienced continuous productivity growth were mainly found in the industrial machinery (B) block. In particular, the computer sector (b5) had recorded rapid productivity growth in recent years, and it was contributing both in static and dynamic terms to cost reduction in other sectors. The every 5-year annual productivity growth from 1960 to 1990 in the computer sector were 6.1817%, 2.1687%, 1.2742%, 6.0535%, 0.7910% and 4.5262% respectively. As for the static spillover effect, the 5-year average since 1960 were 0.0833%, 0.0431%,0.0091%, 0.0922%, 0.0447% and 0.1658% respectively. The dynamic spillover effects for the same period were 0.2725%,0.1352%,0.0294%,0.3105%,0.1618%,0.4526% respectively. For the period after 1985, differences in the static spillover effects and dynamic spillover effects had been enlarging,and the computer sector was playing an increasingly important role as a provider of investment good.

  • 1-6-1 Nagata-cho, Chiyoda-ku, Tokyo 100-8914, Japan.
    Tel: +81-3-5253-2111