Growth without productivity

Growth without productivity

Journal of Development Economics 18 (1985) 25-38. North-Holland G R O W T H W I T H O U T PRODUCTIVITY Singapore Manufacturing in the 1970s Yuan TSA...

721KB Sizes 3 Downloads 25 Views

Journal of Development Economics 18 (1985) 25-38. North-Holland

G R O W T H W I T H O U T PRODUCTIVITY

Singapore Manufacturing in the 1970s Yuan TSAO* National University of Singapore, Singapore 0511 Received August 1983, final version received February 1984 There is often a tendency to associate high total factor productivity (TFP) growth with rapidlygrowing output. The evidence for Singapore manufacturing industries indicates that this has not been. the case. TFP growth has been very low for the majority of Singapore manufacturing industries during 1970-1979. Three hypotheses are suggested as the possible reasons for this result - - the predominance of foreign capital in Singapore manufacturing, the government's lowwage policy combined with the influx of low-skilled foreign labour and the relatively low level of industrial competence in Singapore.

1. Introduction There is more often than not a presumption that total factor productivity (TFP) growth is high in a rapidly-growing economy, and that high TFP growth contributes significantly to output growth as well as increases in labour productivity. Available studies on such high-growth countries as South Korea and Japan by Christensen and Cummings (1981), Nishimizu and Robinson (1983), Nishimizu and Hulten (1978) and Jorgenson and Nishimizu (1981) all point in this direction at the aggregate economy as well as the sectoral level. The positive relationship between output and productivity growth has been postulated in Verdoon's Law; expansion of output enables both static and dynamic economies of scale to come into play, thus resulting in higher levels of productivity. 1 This paper examines the applicability of this predilection to manufacturing industries in one of the fastest-growing economies in the world - - Singapore. The somewhat surprising conclusion is that the rapid growth of output of ,manufacturing industries from 1970 to 1979 has been almost entirely due to growth in factor inputs. There has been little increase in the efficiency of use *I wish to thank Professors D.W. Jorgenson and H.B. Chenery of Harvard University and Mieko Nishirni~ql at the World Bank for their invaluable and insightful comments. My colleagues, Eng-Fong Pang and Basant Kapur also provided much help in the preparation of this paper. 1See Cornwall (1977, ch. 7) for a more detailed discussion of Verdoon's Law. 0304-3878/85/$3.30 © 1985, Elsevier Science Publishers B.V. (North-Holland)

26

Y. Tsao, Growth without productivity

of factor inputs. In other words, TFP growth, which measures such changes as technical progress, improvements in organisational structure and workermanagement relations as well as the diffusion of technology across firms, has been negligible in Singapore's manufacturing industries. The plan of the paper is as follows: section 2 discusses the theoretical framework, section 3 describes the results, while section 4 discusses several hypotheses which may attempt to explain these results. Section 5 is the conclusion. 2. Theoretical framework

In continuous time the Divisia index of TFP growth is calculated as a residual of the Divisia index of output growth less the weighted sum of Divisia indexes of the growth of factor inputs. The discrete time analogue used in this study is based on the translog production function as in Gollop and Jorgenson (1980). Denoting by Zi, Xi, Kt, Li, Ei and T respectively the output, non-energy intermediate, capital, labour and energy inputs and time for industry i, we can write the translog production function as lnZl =g~ + g ~ c l n X i + ~ ln Ki + O~iLln Li + ot~ln Ei + ot~r• T

+ ½fl~cx(ln X,) 2 + fl~cKIn X, In K, + fl~XLIn X, In L, + ffxElnX, lnE,+ flixrlnX, • T + ½~cK(In K3 2 + flKLIn K l In L i "b [~KEIn K, In E, + ,#~r In K,. T

+½fl~L (ln Li) 2 + [3iLEIn E, + fliT In L," T + ½fl~E(In E3 2 + ~ r In E,. T

+½fl~r" T2, 2

i= 1 2. . . . . n.

(1)

The translog index of TFP growth is exact for the translog production function and is therefore a superlative index) Assume constant returns to scale and producer equilibriurn. 4 Using V~x, V~:, V~ and V~ to denote the 2This specification is more general compared to one where value added is specified as a function of capital, labour and time. The existence of a value added function requires the weak separability of a function of capital, labour and time on the one hand and intermediate goods (and energy) on the other. Although this is a convenient assumption to make because data on intermediate goods in constant and current prices are not required for T F P computations, it may lead to biased results unless the conditions for weak separability are empirically justified. SSee Diewert 0976) for a discussion of exact and superlative index numbers. 4Caves, Christensen and Diewert (1982) have shown that under constant returns to scale the output based productivity index used here (differences in maximum output conditional on a

Y. Tsao, Growth without productivity

27

cost shares of Xi, Ks, L~ and E~, respectively, and a bar to indicate the simple average over two successive time periods, the translog index of sectoral T F P growth is given by P~. = In Z~(T) - In Z~(T -- 1) - P~[ln X~(T) - In X , ( T - 1)] -- P~:[ln K~(T) -- In K , ( T -- 1)] - P£ [ l n L,(T) - In L ~ ( T - 1 )]

- P~[ln Ei(T) - i n

E,(T- 1)],

where P~x=½[V~(T) +

V~(T- 1)],

~'~ =½[V~T) +

V~T-

1)2,

P~=½[V~L(T)+VL(T-I)], Pk=½[V~T)+V~T- i)], P~r=½EV~T)+V~T-I)],

i = 1 , 2 . . . . . n.

and

(2)

The translog index of T F P growth is obtained as the growth of output less the weighted average of growth of inputs. Each of the factor inputs can in turn be expressed as a translog function of its components. The rate of growth of each factor input is then calculated as a translog index of the rate of growth of its individual components with their average value shares as weights. For example, for labour input Li, In L, = ~ al In L u + ½~ ~ ill, In L u In L=,, l

I m

i = 1,2,...,n,

l,m= 1,2 ..... q, (3)

so that In LX 73 - In L , ( T - I) = ~ P~,,[In L,,('/3 - In L,,(T- I )], where --i 1 i VLI-'~[VLt( T- 1)+

VLt(T -

1)]

(4)

and V~ is the value share of the Ith component of labour input in the total labour compensation in industry i. 3. T h e results

The study covers 28 manufacturing industries in Singapore during 1 9 7 0 given level of inputs) is equal to the input based productivity index (differencesin minimum input requirements conditional on a given level of output). When returns to scale are either increasing or decreasing however, these two measures differ. See also Caves, Christensen and Swanson (1981).

28

Y. Tsao, Growth without productivity

1979.5 A discussion of the data construction can be found in the appendix, and appendix table A.I gives the average annual value shares of the four inputs. Fairly rapid growth was experienced by the majority of industries. The manufacturing sector as a whole grew at 8.3 percent per year (see table I). 6 The high rates of growth of the electrical machinery, industrial machinery, transport equipment and wearing apparel industries reflect their increasing importance in the manufacturing sector after the petroleum products industry. TFP growth rates however, were low. The TFP growth rate for the manufacturing sector was almost nil (0.08 percent), and 17 out of 28 industries had negative TFP growth rates. The productivity slow-down since 1973 in the industrialised countries has recently been given much attention. 7 The low TFP growth rates in Singapore's manufacturing industries could be part of this apparently worldwide phenomenon. If this were the case, then the average annual TFP growth rates for 1973-1979 would be lower than those for 1970-1973. This was found to be so for only half of the industries. For the manufacturing sector the average annual TFP growth rate for 1973-1979 was 0.71 percent, higher than that for 1970-1973, -1.18 percent. The low TFP growth rates therefore appear to reflect a more fundamental phenomenon than the productivity slow-down alone. The average annual contributions of TFP to output growth for 1970-1979 are shown in table 2. For the manufacturing sector the average contribution of TFP to output growth was 9.6 percent. This is low compared to the figures obtained for the manufacturing sectors of South Korea (1960-1977) and Japan (1953-1973) - - 20.7 and 17.6 percent, respectively.8 Rapid output growth, therefore, has not been accompanied by high TFP growth rates in Singapore's manufacturing industries in the seventies. The hypothesis of this positive relationship was tested in two ways. First, the correlation between the annual averages of the rates of output and TFP growth was computed for each of the 28 industries. Next, the correlation for the pooled cross-section and time-series data on output and TFP growth was also calculated. These correlations were low - - 0.340 and 0.556, respectively. The evidence derived from data on manufacturing industries in Singapore therefore does not support the hypothesis that rapid output growth is associated with high productivity growth. SA longer time period would be desirable but di~culfies due to the change in industrial classifications before and after 1970 make this a hazardous exercise which is best avoided at this stage, 6The manufacturing sector was taken to be the aggregate of the 28 industries listed in table 1. The indexes of output and factor inputs for the sector as a whole were calculated using translog aggregation procedures. ~See, for example, the papers in Economic Journal, Volume 93, March 1983. SNishimizu and Robinson (1983, p. 12, table 1).

Y. Tsao, Growth without product/v/ty

29

Table 1 Average annual rates of growth of real output, non-energy intermediate, capital, labour and energy inputs and total factor productivity by industry, 1970-1979 (percent).

Industry

Output

Intermediate input

Capital input

Labour input

Energy input

Food Beverages Tobacco products Textiles Wearing apparel Leather products Footwear Timber products Furniture & fixtures Paper products Printing & publishing Industrial chemicals Chemical products Petroleum products Jelutong processing Rubber products Plastic products Pottery & glass products Structural clay products Cement Concrete products Non-metallic mineral products Iron & steel Non-ferrous metals Fabricated metal products Industrial machinery Electdcal machinery Transport equipment

3.07 5.22 1.66 6.34 10.80 -3.41 -9.50 -0.13

2.16 6.00 - 2.92 10.11 12.92 -0.60 -0.15 6.88

4.83 -0.23 1.97 11.78 16.54 -5.20 5.17 10.54

2.10 1.52 2.54 4.23 11.76 6.77 -0.95 0.72

4.01 3.92 7.60 12.44 18.18 3.82 3.58 7.21

0.62 1.73 3.22 - 3.23 -2.11 -3.06 -9.91 -6.57

9.76 13.50

10.65 10.91

19.15 14.31

11.63 7.26

20.16 22.63

-2.44 2.18

5.58

7.06

8.82

4.93

8.16

- 1.36

14.43 9.78 -0.01 -8.19 1.72 17.88

10.34 12.37 12.23 5.63 2.41 17.40

8.92 5.48 5.49 -3.28 0.95 14.44

11.82 14.29 20.24 -4.21 2.01 25.35

-0.24 4.80 1.49 8.32 - 1.57 -3.16

-- O.13

8.03

3.33

-- 7.09

2.45

- 3.03

3.96 7.61 7.89

- 1.40 11.24 15.51

2.13 12.95 11.02

-9.16 3.08 8.23

-0.02 6.71 8.88

5.63 -3.78 -5.36

11.39 9.79 -2.63

10.29 4.09 10.83

13.22 12.57 17.73

3.93 4.89 2.99

6.97 8.24 -2.52

1.44 3.41 - 13.87

3.44

6.57

9.35

5.76

12.44

- 3.59

19.31

20.01

29.38

18.60

27.61

-3.28

24.40 11.84

25.35 11.79

25.54 11.46

17.74 6.90

18.77 9.27

-0.04 1.27

8.34

6.45

14.27

8.41

13.97

0.08

Manufacturing sector

12.53 15.24 3.92 1.30 0.45 14.35

Total factor productivity

4. An in~,'emlion of the resul~ T h e c o n c l u s i o n o f v e r y l o w T F P g r o w t h is a t first g l a n c e s u r p r i s i n g for a r a p i d l y g r o w i n g e c o n o m y . T h e s e results a r e in m a r k e d c o n t r a s t to, for e x a m p l e , t h e h i g h T F P g r o w t h r a t e s o b t a i n e d for K o r e a n m a n u f a c t u r i n g

30

Y. Tsao, Growth without productivity

Table 2 Average annual contribution of total factor productivity and total factor inputs to output growth, 1970-1979 (percent). Industry Food Beverages" Tobacco products" Textiles" Wearing apparel Leather products Footwear Timber products a Furniture & fLxtures Paper products Printing & publishing Industrial chemicals" Chemical products Petroleum products" Jelutong processing" Rubber products Plastic products" Pottery & glass products Structural clay products Cement" Concrete products Non-metallic mineral products Iron & steel" Non-ferrous metals" Fabricated metal products Industrial machinery Electrical machinery Transport equipment Manufacturing sector

Total factor productivity/ output

Total factor inputs/outputs

- 56.3 47.6 155.2 33.8 37.6 53.6 28.7 - 9.7 39.5 - 54.5 77.1 -49.6 17.5 84.9 -271.1 141.3 11.6

156.3 52.4 --55.2 66.2 62.4 46.4 71.3 109.7 60.5 154.5 22.9 149.6 82.5 15.1 371.1 - 41.3 88.4

- 24.2

124.2

36.3 - 94.6 76.8

63.7 194.6 23.2

168.6 -21.3 - 103.6

- 68.6 121.3 203.6

80.7 - 18.6 -0.7 - 14.6

19.3 118.6 100.7 114.6

9.6

90.4

"For eight years only. i n d u s t r i e s d u r i n g 1 9 6 0 - 1 9 7 7 as m e n t i o n e d earlier. 9 T h e q u e s t i o n r e m a i n s as t o w h y T F P g r o w t h in S i n g a p o r e m a n u f a c t u r i n g i n d u s t r i e s h a s b e e n negligible. One reason could be because of data inaccuracies. If output and inputs are n o t m e a s u r e d a c c u r a t e l y , t h e n T F P g r o w t h r a t e s w o u l d be biased. F o r 9The comparison with Korea seems to be particularly relevant as the manufa~uring sectors of both countries have grown rapidly and are export-oriented. However, low rates of TFP growth have also been obtained for the manufacturing sector in Thailand - - 0.69 percent for 1963-1976 [see Wiboonchutikula (1982)].

Y. Tsao, Growth without productivity

31

example, Berndt and Fuss (1981) have shown that changes in capacity utilisation can be taken into account by valuing capital (the quasi-fixed input) at its shadow rental price rather than its market price. If this were not done then a decline in capacity utilisation such as in 1974/1975 would lead to a downward bias in TFP growth rates. 1° Due to lack of data on the shadow price of capital services, however, this was not taken into account in this study. Nevertheless, inspite of these possible biases, the results might reflect some more fundamental characteristics of the manufacturing sector in Singapore. Three hypotheses, not necessarily independent, are put forward as possible explanations of low TFP growth and discussed below.

4.1. Hypothesis 1: Predominance of foreign investment One of the most striking features of the Singapore manufacturing sector is the predominance of foreign capital. Table 3 shows the number of workers, output, value added and capital expenditure for local and foreign establishments in the manufacturing sector. During the first decade of industrialisation since 1968, foreign firms have expanded both in number and in size much more rapidly than local firms. By 1979, foreign firms in manufacturing employed 57 percent of the manufacturing work force, produced 74 percent of manufacturing output and 67 percent of manufacturing value added and accounted for 73 percent of capital expenditure. Singapore's manufacturing sector therefore, can be said to be 'dominated' by foreign firms. Although transnational companies may bring with them superior technology, there may be reason to expect that the prevalence of foreign capital in an industry may lead to a smaller TFP growth rate than when the industry is largely indigenous. At the firm level, to the extent that transnational companies are orientated towards global and not local profit goals and rely more on research and development effort in the parent company, there may be less room for adaptation of technology to the local environment and for the learning which is involved in making minor innovations in the local plant. At the industry level, as the nature of the technology which is used in developing countries is usually mature, the foreign companies in that industry may have levels of TFP which are relatively close to that embodied in the best practice frontier. Over time there may be less opportunity for within-industry diffusion of technological know-how and learning by doing. In contrast, one would expect that such factors which would lead to a reduction in technical inefficiency would be significant in accounting for TFP growth in a largely indigenous industry. l°Berndt and Fuss, using Tobin's q as an empirical measure of the shadow value of capital, obtained a higher measure of TFP growth during 1973-1977 for the US manufacturing sector. 25 percent of the traditionally measured decline in TFP growth during this period could be attributed to a decline in capacity utilisation.

32

Y. Tsao, Growth without productivity

Table 3 Principal statistics of manufacturing establishments by local and foreign ownership, 1968, 1975 and 1979.*

Number of

Number of

Output

efltablishmfnts

workers

(S$ million)

Value added (S$ million)

Capital expediture (S$ million)

1968

Local 1,400(88.3%) Foreign 186(11.7%) Total 1,586

55,254(73.8%) 1,171.8(53.9°/0) 19.579(26.2%) 1,003.9(46.1~/0) 74,833 2,175.7

342.8(56.0°/0) 268.9(44.0%) 611.7

51.4(57.4%) 38.1(42.6%) 89.5

91,941(48.0%) 3,622.6(28.70/o) 1,272.7(37.3%) 99,587(52.0%) 8,987.5(71.3%) 2,138.5(62.70/0) 191,528 12,610.1 3,411.2

220.3(35.4%) 402.4(64.6%) 622.7

1975

Local 1,860(78.0°/0) Foreign 525(22.0°/0) Total 2,385 1979

Local 2,376(76.1%) Foreign 746(23.9%) Total 3,122

114,355(42.5%) 6,639.1(26~°A) 2,189.5(32.7%) 383.7(26.9%) 154,979(57.5%) 18,657.6(73.8%) 4.513.9(67.3%) 1,040.7(73.1%) 269,334 25,296.7 6,703.4 1,424.4

~Source:. Singapore Department of Statistics, Report on the Census of Industrial Production, various years (Singapore National Printers, Singapore). Local ownership re£~s to establishments which are wholly local and more than half local. Foreign ownership triers to establishments which are less than half local and wholly foreign. The figures in parentheses are percentages of the respective totals. The data are for establishments for ten or more workers.

For these two reasons - - the constraints faced by transnational companies and the clustering of foreign firms near the best practice frontier - - the rate of TFP growth in an industry with a large ratio of foreign capital may be less than that in a largely indigenous one. The best practice frontier could, however, move substantially over time reflecting technological progress and hence TFP growth. Hypotheses 2 and 3 may help to explain why very little of this seems to have occurred.

4.2. Hypothesis 2: The 'low-wage' policy and the availability of low-skilled

labour in Singapore In contrast to Hypothesis 1 which emphasises the nature of the firm as an explanation of the lack of TFP increase, this hypothesis is concerned with the environment external to the ~irm, in particular wages and the availability of low-skilled labour. In response to tight labour market conditions in the early seventies, the National Wages Council (NWC) comprising representatives from employees, employers and the government was formed in 1972 to recommend orderly wage increases on an annual basis. While the wage increases recommended in 1973 and 1974 were relatively high, a moderation in wage increases was the policy adopted during 1975-1978. This was in

Y. Tsao, Growth withoutproductivity

33

order to preserve Singapore's competitive edge during the period of recession and recovery. As a result, nominal increases in average weekly earnings in manufactured averaged 16.5 percent during 1973-1975 and only 6.6 percent during 1975-1979. ~1 Using the wholesale price index of Singapore manufactured products as deflator, real wages actually fell by 0.25 percent during the latter period. At the same time there were two additional sources of low-skilled labour in the seventies in the economy. Not only did the female labour force participation rate rise from 25 to 39 percent between 1970 and 1980 but there was also a rapid and significant influx of relatively low-skilled labour from Malaysia and other neighbouring countries. The proportion of females to the total number of citizens working in manufacturing rose from 34.3 to 46.6 percent between 1970 and 1980, and the proportion of non-residents in the manufacturing work force increased from 3.5 to 11.3 percent during the same period. 12 The labour policy pursued by the Singapore government may therefore have encouraged the preservation of the labour-intensive, assembly-type operations characteristic of the manufacturing sector at the start of the industrialisation process. There may have been less incentive to upgrade labour productivity such as by improvements in organisational efficiency which would otherwise have resulted in a higher TFP growth rate. 13 4.3. Hypothesis 3: A relatively low technological capability and scarcity of domestic industrial entrepreneurship in Singapore

This point can perhaps be best brought out by a comparison of Singapore with South Korea. Westphal, Rhee and Pursell (1981) have argued that the Koreans possess considerable industrial competence (technological mastery and marketing sawy) gained largely by indigenous effort through various forms of learning,. In contrast, indigenous industrial entrepreneurship has been lacking in Singapore. Although commercial and financial expertise has been forthcoming in the country, this seems to have not been easily transferable into industry where the returns are more risky, the gestation period .is longer, and the initial outlay larger. This is in fact one main reason why Singapore has relied principally on foreign investment whereas Korea's manufacturing has largely been an indigenous effort. ~The data on average weekly earnings were taken from Singapore Ministry of Labour, Singapore Yearbookof Labour Statistics, 1978and 1980. 12This was calculated from data found in Singapore Department of Statistics, Report on the Census of Population, 1970and 1980. 13This argument was first put forward by Pang (1980) as the most important explanation of the poor performance of labour productivity in manufacturing as compared to other sectors. Using national accounts data, he computed the average annual rate of growth of labour productivity, 1970-1978, as 2.3 percent in manufacturing. It averaged 4.8 percent in the tertiary sector.

34

Y. Tsao, Growth without productivity

Indicators of human capital in Singapore and Korea also point towards the lower level of industrial competence in Singapore. Not only has the literacy rate in Singapore been lower than that in Korea (75 percent as opposed to 93 percent in 1975) but the percentages enrolled in secondary schools and in higher education have also been lower, implying that the level of education in Singapore has been lower than that in Korea. The number of scientists and engineers per million population was 5,466 in Singapore in 1977; it was 21,550 in Korea in the same year. The number of scientists and engineers engaged in research and development per million population in Singapore was about half that in Korea in 1978.14 It might be expected that the level of industrial competence would have consequences on the rate of growth and level of TFP in a country's manufacturing sector. The relatively low level of industrial competence of Singaporeans in the seventies may, therefore, be one reason for the depressed rates of TFP growth in Singapore manufacturing industries. 5. Conclusion

These three hypotheses remain as hypotheses at this stage. The data required for testing, such as details regarding capital ownership by industry, are unfortunately not available. The conclusion of low TFP growth for Singapore's manufacturing industries is predicated on a theoretical framework and data base which are both not ideal. The dissatisfaction with the neoclassical paradigm stems from the inadequate treatment of the important questions relating to the causes of productivity growth [see Nelson (1981)]. TFP computations require an enormous and detailed data base in which there is plenty of room for improvement. In the case of Singapore's manufacturing industries, the small size of many of the industries leads to the possibility of discrete jumps in the data. Skill/educational categories for labour are not available by industry. Changes in the composition of each industry are ignored, an assumption which may not be justified in rapidlygrowing industries. The causes of TFP growth remain a challenging area for research which has enormous implications for policy-making. The conclusions obtained in this study, for example, underscore the recent emphasis by the Sil~gapore government on productivity. Singapore's manufacturing has m a n a ~ d to achieve rapid growth without TFP increases i n the first decade of industrialisation. In a land-and-labour-scarce economy, however, this pattern of growth cannot be sustained without detrimental effects on output growth. It is to be hoped that a deeper understanding of the low rates of growth of 14Data on literacy rates are available in The World Bank, World Development Report, various years. Data on scientists and engineers are taken from the UNESCO Statistical Yearbook, various years.

Y. Tsao, Growth without productivity

35

TFP in manufacturing industries will contribute towards appropriate policies to raise productivity in the eighties decade. Appendix This appendix discusses the construction of the data base. Most of the data were taken from the annual censuses of industrial production (CIP) for 10 or more workers. All real input and output series were in constant 1974 prices.

Capital input. Seven categories of capital assets were distinguished - land, buildings and structures, machinery and equipment, office equipment, vehicles, stocks of output and stocks of materials. The stocks of output and materials at constant prices were obtained by using the implicit output price deflator (nominal output divided by the index of real gross output) and the computed material input price deflator, respectively. For the other five assets, the perpetual inventory method was used with net fixed assets in 1970 as the benchmark. Capital expenditure in constant prices was computed as nominal capital expenditure (net of assets sold) divided by the gross domestic capital formation deflators from the national accounts. A separate price deflator was computed for land. The depreciation rates estimated by Hulten and Wykoff for the U.S. [see Jorgenson and Sullivan (1981)] were used on the assumption that most of the capital equipment used in Singapore's manufacturing industries was imported; no local estimates are available. The depreciation rates used were: industrial buildings, 0.0361; office equipment, 0.2729; mac~nery and equipment, 0.1047; vehicles, 0.2935; and land, zero. The prices of capital services were computed as in GoUop and Jorgenson (1980), taking into account the income tax, property tax and depreciation allowances in force during the period. These were then used to compute the value shares for each asset category and hence capital input. Labour input. Labour input in man-hours was classified by occupational status and sex. Data from other sources were used to supplement the CIP data on workers by occupational status and sex and remuneration of workmen and other employees. The assumptions that (i) the average remuneration of working proprietors was equal to that of other employees, an¢~ (ii) the average remuneration of unpaid family workers was equal to that of workmen, were used to impute remuneration of these two categories of labour. The ratio of male to female income was imputed from data on income earned by sex in manufacturing from labour force surveys after estimating mean income by a lognormal distribution method. This ratio was then used to allocate remuneration to males and females for each occupational category. Weekly hours worked were computed for each category by

Y. Tsao, Growth without productivity

36

assigning the weekly hours worked in manufacturing by occupational groups, available in the Singapore Yearbook of Labour Statistics, to each occupational status. The ratio of male to female hours worked was then applied in a similar manner to obtain weekly hours worked by sex for each occupational status. Man-hours worked per year was obtained by multiplying the number of workers by weekly hours worked and by 52.

Intermediate inputs. Non-energy intermediate inputs comprised materials and work given out, water and transport. The components of energy input were electricity, fuel off, gas and other fuels. The material input price deflator was computed by using 1973 input-output coefficients as weights for output price indexes (for locally purchased inputs) and import unit value indexes (for imported inputs). The main source of energy for most of the industries Table A.1 Average annual value shares of non-energy intermediate, capital, labour and energy inputs by industry, 1970-1979 (percent). Intermediate input

Capital

Labour

Energy

Electrical machinery Transport equipment

81.7 53.2 74.9 63.9 68.3 76.3 62.9 67.8 57.7 62.4 45.6 58.1 50.7 83.5 87.8 50.4 62.6 40.7 9.3 71.9 64.8 52.8 47.6 72.9 63.4 54.1 63.0 50.8

11.0 28.7 18.8 16.0 12.2 11.1 13.4 16.1 16.2 21.3 29.3 27.3 37.6 12.6 5.4 28.6 20.1 22.5 40.1 21.2 19.9 29.5 33.9 17.0 19.5 26.7 23.5 25.2

5.8 16.3 6.0 16.2 18.7 12.2 22.6 14.1 25.1 13.4 24.0 9.7 10.8 1.7 5.7 17.6 14.3 24.1 27.6 3.6 14.3 15.8 9.3 8.0 15.4 17.5 12.5 22.2

1.5 1.7 0.4 3.9 0.8 0.4 1.2 2.0 1.1 2.9 1.0 5.0 0.9 2.2 1.1 3.4 2.9 12.7 23.0 3.3 1.0 2.0 9.3 2.1 1.7 1.7 1.1 1.8

Manufacturing sector

70.3

18.1

9.6

2.0

Industry Food Beverage Tobacco products Textiles Wearing apparel Leather apparel Footwear Timber products Furniture & fixtures Paper products Printing & publishing

Industrial chemicals Chemical products Petroleum products Jelutong processing Rubber products Plastic products Pottery & glass products Structural clay products

Cement Concrete products Non-metallic mineral products Iron & steel

Non-ferrous metals Fabricated metal products Industrial machinery

Y. Tsao, Growth without productivity

37

was electricity, for which nominal and volume data were available. The other components of energy input were deflated by using appropriate price indexes.

Output. Indexes of real gross output were available at the three-digit industry level on a quarterly basis. The average over the year was taken to be the output index for that year. The earlier series with base year 1970 was spliced to correspond with the series with base year 1974. References Berndt, E.R. and M.A. Fuss, 1981, Productivity measurement using capital asset valuation to adjust for variations in utilization, Paper presented at the Econometric Society Summer Meetings (San Diego, CA) June. Caves, D.W., L.R. Christensen and W.E. Diewert, 1982, The economic theory of index numbers and the measurement of input, output, and productivity, Econometrica 50, 1393-1414. Caves, D.W., L.R. Christensen and J.A. Swanson, 1981, Productivity growth, scale economies, and capacity utilization in U.S. railroads, 1955-1974, American Economic Review 71, 9941002. Chen, E., 1979, Hyper-growth in Asian economies: A comparative study of Hung Kong, Japan, Korea, Singapore and Taiwan (Holmes and Meier, New York). Christensen, L.R. and D. Cummings, 1981, Real product, real factor input, and productivity in the Repubfic of Korea, 1960-1973, Journal of Development Economies 8, 285-302. Christensen, L.R., D. Cummings and D.W. Jorgenson, 1978, Productivity growth, 1947-1973: An international comparison, in: W. Dewald, ed., The impact of international trade and investment on employment (U.S. Government Printing Office, Washington, DC). Cornwall, J., 1977, Modern capitalism: Its growth and transformation (St. Martin's Press, New York). Denison, E.F., 1969, Some major issues in productivity analysis: An examination of estimates by Jorgenson and Griliches, Survey of Current Business 49, Part II, 1-27. Denison, E.F., 1972, Final comments, Survey of Current Business 52, 95-110. De~'son, E.F, 1983, The interruption of productivity growth in the United States, Economic Journal 93, 56-77. Diewert, W.E., 1976, Exact and superlative index numbers, Journal of Econometrics 4, 115-146. Giersch, H. and F. Wolter, 1983, Towards an explanation of the productivity slowdown: An acceleration-deceleration hypothesis, Economic Journal 93, 35-55. Gollop, F.M. and D.W. Jorgenson, 1980, U.S. productivity by industry, 1947-1973, in: J.W. Kendrick and B.N. Vaccara, eds., New developments in productivity measurement and analysis (University of Chicago Press, Chicago, IL). Huiten, C.R., 1973, Divisia index numbers, Econometrica 6, 1017-1025. Jorgenson, D.W. and Z. Griliches, 1967, The explanation of productivity change, The Review of Economic Studies 34, 249-283. Jorgenson, D.W. and Z. Griliches, 1972a, Issues in growth accounting: A reply to Edward F. Denison, Survey of Current Business 52, 65-94. Jorgenson, D.W. and Z. Griliches, 1972b, Issues in growth accounting, Final reply, Survey of Current Business 52, 11. Jorgenson, D.W. and M. Nishimizu, 1981, International differences in levels of technology: A comparison between U.S. and Japanese industries, International Roundtable Congress proceedings (Japan Statistical Society, Tokyo) 43-77. Jorgenson, D.W. and M.A. Sullivan, 1981, Inflation and corporate capital recovery, in: C.R. Hulten, ed., Depreciation, inflation and the taxation of income from capital (The Urban Institute Press, Washington, DC) 171-238. Lindbeek, A., 1983, The recent slowdown of productivity growth, Economic Journal 93, 13-34. Morris, D.J. and SJ. Prais, 1983, The recent slowdown in productivity growth: Comments on the papers, Economic Journal 93, 78-88.

38

Y. Tsao, Growth without productivity

Nelson, R.R., 1981, Research on productivity growth and productivity differences: Dead ends and new departures, Journal of Economic Literature XIX, 1029-1064. Nishimizu, M. and C.R. Huiten, 1978, The sources of Japanese economic growth: 1955-1971, Review of Economics and Statistics 60, 351-361. Ni~himi~t, M. and J.M. Page, 1982, Total factor productivity, growth, technical progress and technical eflic/ency change: Dimensions of productivity change in Yugoslavia, 1965-1978, Economic Journal 92, 920-936. Nishimizu, M. and S. Robinson, 1984, Trade policies and productivity change in semiindustrialized countries, Journal of Development Economics 16, 177-206. Pang. E.F., 1980, Factors which hinder or help productivity improvement in the Asian region - A review and the prospect: Singapore (Asian Productivity Organization, Nordica International, Hong Kong). Tsao, Y., 1982, Growth and productivity in the Singapore economy:. A supply side analysis, Ph.D. dissertation (Harvard University, Cambridge, MA). Westphal, L.E., Y.W.' Rhee and G. Pursell, 1981, Korean industrial competence: Where it came from, World Bank Staff Working Paper no. 469, July. Wiboonchutikula, P., 1982, The total factor productivity growth of the manufacturing industries in Thailand, 1963-1976, Ph.D. dissertation (University of Minnesota, Minneapolis, MN).