Human capital spillovers, productivity and growth in the manufacturing sector of Pakistan.
Article Type: Report
Subject: Human capital (Research)
Labor productivity (Research)
Manufacturing industry (Research)
Authors: Hamid, Abdul
Pichler, J. Hanns
Pub Date: 06/22/2009
Publication: Name: Pakistan Development Review Publisher: Pakistan Institute of Development Economics Audience: Academic Format: Magazine/Journal Subject: Business, international; Social sciences Copyright: COPYRIGHT 2009 Reproduced with permission of the Publications Division, Pakistan Institute of Development Economies, Islamabad, Pakistan. ISSN: 0030-9729
Issue: Date: Summer, 2009 Source Volume: 48 Source Issue: 2
Topic: Event Code: 310 Science & research Canadian Subject Form: Labour productivity
Product: Product Code: 9918220 Productivity Improvement
Geographic: Geographic Scope: Pakistan Geographic Code: 9PAKI Pakistan
Accession Number: 242017200
Full Text: Manufacturing is an important sector of Pakistan's economy. The main focus of this paper is to analyse the major factors of value-added growth and productivity in the manufacturing sector by using Translog Production Technology over the period 1971-72 to 2004-05. The empirical findings show that the contribution of productivity and human capital is around one-third of the total value-added growth in manufacturing sector which is less than the contribution attributed to these factors in developed and many other developing countries.

Conventional factors like capital and labour are still the mainstay in the value-added growth of Pakistan's manufacturing sector.

JFL classification: O1, 03, O4, O14, O15, O31

Keywords: Human Capital Spillovers, Total Factor Productivity, Absolute and Relative Shares

I. INTRODUCTION

Manufacturing sector has been playing an important role in the economy of Pakistan. In 2005-06 its contribution to GDP and employment amounted to 18 percent and 14 percent respectively. It also plays a vital role in exports whose composition over time has changed significantly from primary commodities to manufactures and semi-manufactures with their share in total exports having nearly tripled, from 28 percent in 1972-73 to 79 percent in 2004-05. (1) Development therefore of the manufacturing sector will have far reaching impact on exports, employment prospects, development of agriculture (by providing machinery and other inputs like fertilisers, etc.) and other sectors by bringing technological changes and absorbing technological spillovers from abroad.

This paper aims at measurement of contribution of factor inputs, technological change and technical efficiency to value-added growth in the manufacturing sector together with measurement of total factor productivity (TFP) change index.

The layout of the paper is as follow: a review of literature is presented in Section II. Section III discusses the methodology, variables and data sources. Discussion of empirical findings and comparisons with other relevant findings is given in Section IV. Summary and conclusions with relevant policy suggestions are presented in Section V followed by references.

II. REVIEW OF LITERATURE

The following review is intended to provide an overview of the broad aspects of topical literature relating to the subject of this paper on a selected basis.

Abramovitz (1956) did an empirical study for the US labour market for the period 1900-1950 and concluded that almost two-third of the increase in labour productivity could not be explained by increase in factor inputs. Solow (1957), Schultz (1964), Ferguson (1965), Hulten (1973), and Kendrick (1973) reported similar results for subsequent periods.

Perelman (1995) did an analysis for eleven OECD countries for the period 19701987 and measured technical efficiency, technological progress and total factor productivity (TFP) growth for the manufacturing industries. According to his findings TFP growth was 1.6 percent and technological progress was 1.8 percent during the sample period. He found that the main contributor in Japan's productivity growth in manufacturing was the efficiency factor.

Coe, et al. (1997) used data of 77 countries over the period 1971-1990 for measuring technological change and development with spillovers on productivity and growth. They concluded that a developing country can benefit more from the technological progress and innovations occurring in the world and can boost its productivity by importing a larger variety of intermediate products and capital equipments with new technology and innovations.

Kruger (2003) measured TFP for a sample of 87 countries for the period 19601990. Technological progress contributed about 66 percent of TFP growth in OECD/EU/G7 economies, while the share of technical efficiency was one third of TFP.

Kumbhakar (2003) used panel data for 450 manufacturing industries in US for the period 1959-92 to measure TFP and technical change. His results show that capital productivity increased by 6.5 percent.

Romer (1986) proposed that development and growth were driven by the accumulation of knowledge. He termed knowledge as a basic form of capital with investment in knowledge and R&D leading to increasing marginal returns of factor inputs. He held that knowledge had a "natural externality" and positive spillovers within and outside the economy. (2) Romer (1990) assumed four basic factors of production in an endogenous growth model: i.e., capital, labour, human capital and an index for the level of technology. According to him, key to growth and sustained development was an adequate stock of human capital.

He and Liu (2006) measured investment-specific technological change and dynamics of skill accumulation for the US for post Second World War period 19492000. According to their analysis, US placed great importance on skill accumulation and on job training. (3) Investment-specific technological change and technical efficiency contributed about 62 percent to average output growth per hour over the period.

Kim and Lee (2006) measured TFP, technological change and technical efficiency for 49 countries for the period 1965-1990. Their analysis found that East Asian countries led the world in technical efficiency and productivity growth leading to higher economic growth. For example Korea, Taiwan, Hong Kong and Japan grew at the high rates of 8.51 percent, 8.62 percent, 8.03 percent and 5.74 percent respectively during 1965 to 1990. Technological development, human capital accumulation and technical efficiency were the major contributors in their higher and continuous growth rates during the reference period. Hong Kong, Japan, Taiwan and Korea showed the highest TFP growth rates of 3.85 percent, 3.53 percent, 2.85 percent and 2.18 percent respectively during the period 1965-1990.

Robinson (1971) estimated technological change, technical efficiency and spillovers caused by human resource accumulation for 39 developing countries and found that, on average, the share of productivity in total growth was 15 percent in these countries. This is a much smaller percentage attributable to technological change, technical efficiency and human capital accumulation than that in developed countries which is over 50 percent in most of the cases.

Yanrui (1995) estimated technical efficiency for agriculture and manufacturing sectors of China. According to his estimations, technological change and efficiency contributed about 53 percent in the state industrial sector, 58 percent in the rural industrial sector, and 55 percent in agriculture in the Chinese economy.

Zheng, et al. (2003) measured TFP in Chinese state-owned enterprises (SOEs) for the period 1980-1994. Their findings show that technical progress contributed significantly in the TFP growth for Chinese SOEs during the reference period and its annual average growth rate was as high as 10 percent. Technical efficiency ranges between 50 to 80 percent during the reference period. TFP grew at significant rates of 3 percent to 12 percent during 1980-1989, and at 3 percent to 8 percent during 1990-1994. Education was found to play a significant role in technical efficiency.

Ruhul (2006) found that in food manufacturing sector in Bangladesh, efficiency ranged between 60 to 81 percent which could be increased by 19 to 39 percent through human capital accumulation in the form of education and on job training.

Cheema (1978) found high productivity growth rate and significant contribution in the manufacturing sector of Pakistan while Ahmed (1980) who estimated productivity growth for the period 1958-70 found low gains in labour productivity. Kemal (1981) analysed the impact of technological change and technical efficiency for the period 1959-60 to 1969-70. He found overall decreasing returns to scale. Kemal and Ahmed (1992) estimated technological change, efficiency growth and productivity for agriculture and manufacturing as well as for the whole economy of Pakistan, but their studies suffer from certain limitations due to use of various functional forms to get estimates of technical efficiency without determining which form was appropriate for which industry.

Kemal, et al. (2002) analysed technological change, technical efficiency and TFP for Pakistan. According to their estimates, TFP grew at a rate of 1.66 percent for the period 1964-65 to 2000-01 and its share in growth of GDP was roughly one-third during the period. TFP in the manufacturing sector showed an average growth of 3.21 percent during 1964-65 to 2000-01 and 4 percent during the sixties mainly through the process of learning by doing coupled with improved export competitiveness.

Khan (1989) measured elasticity of substitutions between inputs, technical progress and returns to scale in the manufacturing sector of Pakistan by using two-level CES production function. He calculated low elasticity between labour, capital and energy and found that the manufacturing sector was exhibiting decreasing returns to scale having experienced disembodied technical progress at the rate of 3.7 percent per annum.

Mahmood (1989) used Translog Cost Function to estimate derived demand for factors in the large-scale manufacturing sector of Pakistan. The estimations found that capital and energy were complementary, and labour, capital and energy were substitutes. The lifting of any subsidy on energy and capital would tend to reduce the energy and capital intensity and, in turn, would increase the labour-intensity in the large scale manufacturing sector of Pakistan. According to his results, adoption of such a policy could help in reducing the burden of unemployment.

Mahmood (1992) further used Translog Cost Function to estimate the effects of change in government's pricing policy and external price shocks on factors demand for the industrial sector of Pakistan. According to this study, the skills of the labour force improve with technological advancements and growth in income. He found that production and non-production workers were high substitutes in the pre-energy shock period and had become marginally stronger substitutes in the post-energy shock period.

Ali and Hamid (1996) measured technological change, technical efficiency, productivity and their impact on input demand for agriculture and manufacturing in Pakistan. They found that major contributors to value-added growth in both sectors were primarily traditional factors of production.

Tariq, et al. (1997) estimated factor substitution, technical efficiency and employment generation in large scale manufacturing industries of Pakistan and found that technological change was capital intensive and labour saving.

Mahmood and Siddiqui (2000) measured TFP for manufacturing in Pakistan over the period 1972 to 1997. They found that increased expenditure on R&D, growth of scientific and technical manpower and growth in knowledge and human capital had a significantly positive impact on TFP growth in manufacturing. Knowledge and human capital were found to explain 30 percent and 18 percent of the variance in TFP, respectively, They also found positive and significant impact of openness and trade liberalisation on TFP.

This review of relevant literature reveals that human capital, technological change and technical efficiency are important sources of growth in the developed countries but these factors have exhibited less importance in developing countries like Pakistan. As a result, only a few studies on human capital accumulation, technological progress and technical efficiency are available especially on Pakistan, and those few also suffer from certain analytical limitations using e.g., Hicks neutral technological change assuming that technological change is happening at a constant rate. In the present research, technological change, technical efficiency, and productivity growth are measured for the manufacturing sector of Pakistan. Besides this, the study also endeavours to measure the major determinants of growth and productivity and the absolute and relative shares of these determinants of value-added growth and TFP in the manufacturing sector. Empirical analysis with international comparisons will be made in this background with related policy implications and conclusions.

III. METHODOLOGY

The following factors are assumed to be the major contributors to value-added growth in manufacturing:

Y = f( A, L, K, H) ... ... ... ... ... ... ... (1)

Y = Value-added growth in manufacturing.

L = Labour employed in manufacturing.

K = Capital stock in manufacturing.

A = Level of technology.

H = Human resources in manufacturing. (4)

Labour and capital have since long been considered as among the most important factors of production in the literature. The more recent studies like Romer (1986, 1990), He and Liu (2006) and Yanuri (2006), etc., have also used the level of technology and human capital as important factors in the analysis of total factor productivity. This paper also assumes both of these factors as key contributors to value-added growth in manufacturing along with traditional factors of labour and capital. In order to measure the major factors contributing to value-added growth, technological change and technical efficiency (which in-builds overtime due to human capital formation), the Translog production function for labour, capital, human capital and technology can be written as:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2)

The following homogeneity constraints are implied in the Translog production function:

([[alpha].sub.LK] = [[alpha].sub.KL], [[alpha].sub.HK] = [[alpha].sub.KH], [[alpha].sub.HL] = [[alpha].sub.LH)]

[summation]([[alpha].sub.K], [[alpha].sub.L], [[alpha].sub.H] = 1

[[alpha].sub.KK], [[alpha].sub.KL], [[alpha].sub.KH] = 0

[[alpha].sub.KL], [[alpha].sub.LL], [[alpha].sub.LH] = 0

[[alpha].sub.KH], [[alpha].sub.LH], [[alpha].sub.HH] = 0

Subject to these homogeneity constraints, the Translog production function will be estimated in conjunction with a cost share function with cross-equation restrictions imposed, a method suggested by Berndt and Christensen (1973). Labour, capital and human capital cost share equations are derived as:

[CS.sub.L] = [partial derivative]lnY/[partial derivative]lnL = [[alpha].sub.L], [[alpha].sub.LL]lnL + [[alpha].sub.LK]lnK + [[alpha].sub.LH]lnH + [u.sub.it] ... ... (3)

[CS.sub.K] = [partial derivative]lnY/[partial derivative]lnK = [[alpha].sub.K], [[alpha].sub.KK]lnK + [[alpha].sub.KL]lnL + [[alpha].sub.KH]lnH + [u.sub.it] ... ... (4)

[CS.sub.L] = [partial derivative]lnY/[partial derivative]lnH = [[alpha].sub.H], [[alpha].sub.HH]lnL + [[alpha].sub.HL]lnL + [[alpha].sub.HK]lnH + [u.sub.it] ... ... (5)

Where

[CS.sub.L], [CS.sub.K], [CS.sub.H] are the labour, capital and human capital shares of total cost, respectively. (5)

The cost share equations will be estimated by applying "Seemingly Unrelated Regressions" [Zellner (1962)]. The Translog production function can also be estimated by using a Stochastic Frontier Approach (SFA), adopting a more flexible approach for restriction conditions. (6)

Following Baltagi and Griffin (1988) and Kumbhakar (2003), single output Translog cost function in the form of time trend model is written as:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

Where

[[beta].sub.jk] = [[beta].sub.kj], [[summation].sub.j][[beta].sub.j] = 1, [[summation].sub.j][[beta].sub.jk] = 0, [[summation].sub.j][[beta].sub.jy] = 0, and [[summation].sub.j][[beta].sub.jt] = 0

The first restriction is due to symmetry; the rest due to the fact that cost function is homogenous of degree one in the input prices.

where

C = total cost

Y = output

[P.sub.j] = jth input price

From the Translog cost function given in Equation 9, technological change is measured as follows (technological change being defined as the percentage change in the total cost over time ceteris paribus):

-[partial derivative ln[C.sub.it]/[partial derivative]t = -[[beta].sub.t] + [[beta].sub.tt]t + [[summation].sub.j][[beta].sub.jt]ln[P.sub.jit] + [[beta].sub.yt]ln[Y.sub.it]] ... ... ... (7)

Measurement of Total Factor Productivity (TFP) Change Index

The TFP change index is defined as the difference between rate of change of output and rate of change of inputs:

[TFP.sup.*] = [y.sup.*-] [x.sup.*] ... ... ... ... ... ... ... (8)

Where

[TEP.sup.*] TFP = total factor productivity change index

[y.sup.*] = rate of change of output

[x.sup.*] = rate of change of inputs

TFP growth can be estimated by subtracting the contribution of measured inputs growth from output growth.

Measurement of Absolute and Relative Contribution

The method for calculation of absolute contribution was introduced by Hicks (1979) and calculation of relative contribution by Hadjimichael, et al. (1995). The absolute share of any factor of production towards growth can be found by multiplying the estimated coefficient of the explanatory variable by standard deviation of the respective explanatory variable. The relative contribution for each independent variable can be measured by dividing its estimated absolute share by the standard deviation of the dependent variable. The relative share of variables will be unit free.

Data and Variables Description

The above model for the measurement of major factors contributing to value-added growth and productivity, technological change, technical efficiency and relative and absolute share of factors in the manufacturing sector of Pakistan is based on the following variables and data sources (data series covers the period from 1972-73 to 200607). All data is converted on 1980-81 constant market prices. Real value-added in manufacturing sector on constant market prices is used and data sources for value-added include the Pakistan Economic Survey (1990-91 and 2007-08) and Pakistan (1999) 50 Years of Pakistan in Statistics: Volume I-IV. The number of employed workers in manufacturing is used as a labour input and data on the variable is taken from Pakistan Economic Survey (1990-91 and 2007-08) and Pakistan (1999) 50 Years of Pakistan in Statistics: Volume I-IV. Enrolment in the secondary, higher and other categories like professional, vocational colleges, universities and other institutes as a ratio to total employed labour force in the manufacturing is used as a proxy to measure the impact of human capital (7) in the manufacturing; data sources include Annual Education Statistics (Various Issues); Pakistan Statistical Years Book (Various Issues); Pakistan Economic Survey (1990-91 and 2007-08); Pakistan (1999) 50 Years of Pakistan in Statistics: Volume I-IV; Human Development Report, UNDP (2007) and World Development Report (2007).

Capital Stock

Capital stock is measured by using perpetual inventory method as per following equation:

[K.sub.t] = [I.sub.t] + (1 - [phi])[K.sub.t-1] ... ... ... ... ... ... (9)

Where

[K.sub.t] = Capital Stock in current year.

[K.sub.t-1] = Capital Stock in previous year.

[I.sub.t] = Current Year Investment or Gross Fixed Capital Formation.

[phi] = Depreciation rate.

For estimating the initial capital stock K(0), the method used by Nehru and Dhareshwar (1993) and Khan (2006) is being followed. The capital stock series will be generated in the following way:

[K.sub.t] = [I.sub.t] + [(1 - [phi]).sup.t] K(0) + [summation][I.sub.t-i] [(1 - [phi]).sup.t] ... ... ... ... (10)

Where

I = o to t-1

K(0) = initial capital stock in year zero.

Nehru and Dhareshwar (1993) and Khan (2006) used a modified Harberger (1978) method to estimate K(0). The value of investment for the first year is estimated by way of a linear regression equation of the log of investment against time. The estimated value of investment for the base or zero year is used to calculate K(0) as per following equation:

K(0)= [I.sub.t] /(gr + [phi]) ... ... ... ... ... ... (11)

Where

gr = Compound value-added growth rate

[phi] = Depreciation rate

Various depreciation rates have been used in empirical studies. Here 5 percent capital depreciation rate is assumed. Several other studies used 4 percent depreciation rate e.g., Nehru and Dhareshwar (1993), Collins and Bosworth (1997), Khan (2006), etc.

IV. EMPIRICAL FINDINGS

The measurement of major determinants of value-added growth and the contribution of factor inputs, technological change and technical efficiency to value-added growth in the manufacturing sector of Pakistan is presented in this section. It also includes calculation of relative and absolute shares of factor inputs in the value-added and the measurement of total factor productivity change index (TFPI).

Translog Production Function Estimates for the Manufacturing Sector

Table 1 presents the estimated results for Translog Production Function in manufacturing. Zellener's Seemingly Unrelated Regression Equations (SURE) technique has been used to find the estimations. All the results are according to theoretical expectations. The estimated coefficient for technology is 0.02, a positive contribution towards the value-added growth in manufacturing. The estimated t-value shows that the coefficient is significant at 1 percent level of significance. The estimated coefficient for capital stock (K) is 0.65 and has a 1 percent level of significance. The estimated coefficient for labour and human capital are 0.15 and 0.20 respectively. (8) The contribution of factor inputs, technological change and technical efficiency to the value added in manufacturing is presented in Table 2. The estimated share for capital (K), Labour (L) and Human Capital (HK) are taken from Translog production function estimations presented in Table 1 and these are 0.65, 0.15 and 0.20 respectively. The weighted growth rates for K, L and HK are measured by multiplying average growth rates of these variables by their respective estimated coefficients in Table 1. The calculated weighted growth rates for K, L and HK are 3.83 percent, 0.48 percent and 1.00 percent respectively. The share of total factor inputs growth in value-added of manufacturing accounted for 5.31 percent. The share of technological change and technical efficiency in value-added is the difference between the average value-added growth (6.6 percent) and total factor inputs weighted growth rates (5.31 percent). This estimated TFP contribution accounted for 1.29 percent. The estimated shares as percentage of total value-added growth in manufacturing show that capital stock contributes the maximum (58 percent) while labour and human capital contributes 07 percent and 15 percent respectively. One reason for this insignificant role of human capital is that technical and vocational training is given low priority and is not of that quality in Pakistan which enables human capital to bring new significant technological changes and can compete internationally and absorb technological spillovers from the advanced world. The total contribution of input factors accounts for 80 percent. The contribution of technological change and technical efficiency (TFP) is 20 percent and the contribution of TFP and human capital in value-added growth rate of manufacturing accounts for 35 percent.

Measurement of Absolute and Relative Contribution to Manufacturing Value-added and Calculation of TFP Change Index

The estimated results are depicted in Tables 3 and 4 respectively. The absolute shares for capital, labour and human capital are 0.398, 0.039 and 0.076 respectively. The relative shares for these explanatory variables follow the same pattern which for capital, labour and human capital are 0.621, 0.061 and 0.119 respectively. The measurement of absolute and relative shares show that value-added growth in manufacturing depends more on physical factors of production and less on human capital.

Table 4 presents the calculations for total factor productivity (TFP) change index in the manufacturing sector from the estimations based on Translog production function given in Table 1. Column (2) in Table 4 shows overtime rate of change of value-added in the manufacturing sector, while columns (3), (4) and (5) show overtime weighted rate of change of inputs i.e. weighted rate of change in capital, labour and human capital. The aggregated weighted rate of change of manufacturing inputs is presented in column (6). The difference between column (2) and column (6) i.e. difference between rate of change of value-added and the rate of change of aggregated weighted inputs is given in column (7) which is the overtime TFP change. The three years moving average growth counts for the sample period are measured at 0.0644, 0.0496 and 0.0147 for value-added, weighted aggregated inputs and TFP respectively. The last column in Table 4 shows the TFP Index which has changed from 100 in 1972-73 to 147.11 in 2004-05.

V. SUMMARY AND CONCLUSIONS

The findings of this study show that conventional factors of production are still the mainstay for value-added growth in manufacturing, contributing about 65 percent of the total value-added growth, while the shares of human capital, technological changes and technical efficiency were measured at 14 percent and 22 percent respectively. TFP along with human capital was contributing around 35 percent to the total growth in the manufacturing sector. This is significant but not up to the required level as in case of developed and in some developing countries its share has been reported at over 50 percent. Based on the empirical findings, the following recommendations and conclusions may be derived:

* Human capital should be given top priority by allocating more resources to education, training, health and to other measures along with physical factors so that human capital can properly be used for enhancing growth of the economy.

* Education policies should be devised according to the requirement of the economy and technical, vocational and professional education must be given top priority as the manufacturing sector has very high need for technical and vocational labour force.

* As openness of the economy is important for raising value-added growth and attracting more technological innovations and spillovers, there is need to search for new markets for Pakistani products through international publicity and interaction with other countries, especially with less developed and neighbouring countries.

REFERENCES

Abramovitz, Modrd (1956) Resource and Output Trends in the United States since 1870. American Economic Review 46:2, 5-23.

Aghion, P. and Peter Howitt (1992) A Model of Growth through Creative Destruction. Econometrica 60:2, 323-351.

Ahmed, M. A. (1980) Factors Which Hinder or Help Productivity Improvements in Asian Region. Country Paper for Pakistan. Asian Productivity Organisation, Tokyo.

Aigner, A., C. A. K. Lovell, and P. Schmidt (1977) Formulation and Estimation of Stochastic Frontier Production Function Models. Journal of Econometrics 86, 21-37.

Ali, K. and Abdul Hamid (1996) Technical Change, Technical Efficiency, and Their Impact on Input Demand in the Agricultural and Manufacturing Sectors of Pakistan. The Pakistan Development Review 35:3, 215-228.

Baltagi, B. H. and J. M. Griffin (1988) A General Index of Technical Change. Journal of Political Economy 96, 20-41.

Batra, Geeta and Hong Tan (2003) SME Technical Efficiency and its Correlates: Cross-National Evidence and Policy Implications. World Bank Institute (Working Paper).

Berndt, E. R. and L. R. Christensen (1973) The Translog Function and the Substitution of Equipment, Structures, and Labour in US Manufacturing 1929-68. Journal of Econometrics 1, 81-114.

Cheema, Aftab Ahmad (1978) Productivity Trends in the Manufacturing Industries. The Pakistan Development Review 17:1, 55-65.

Coe, D., E. Helpman, and A. Hoffmeister (1997) North-South R&D Spillovers. Economic Journal 107, 134-149.

Collins, S. and B. P. Bosworth (1997) Economic Growth in South Asia: Accumulation Versus Assimilation. Brookings Papers on Economic Activity 2, 135-203.

Ferguson, C. (1965) Time Series Production Functions and Technological Progress in American Manufacturing Industry. Journal of Political Economy 73:2, 135-47.

Grossman, G. and E. Helpman (1991) Quality Ladders in the Theory of Growth. Review of Economic Studies 58, 43-61.

Hadjimichael, M. T., D. Ghura, M. Muhleisen, R. Nord, and E. M. Ucer (1995) Sub-Saharan Africa Growth, Savings and Investment, 1986-93. Washington, DC. International Monetary Fund.

Harberger, A. (1978) Perspectives on Capital and Technology in Less Developed Countries. In M. Artis and A. Nobay (eds.) Contemporary Economic Analysis. London: Croom Helm.

He, Hui and Zheng Liu (2006) Investment-Specific Technical Change and the Dynamics of Skill Accumulation and Wage Inequality. University of Minnesota and Emory University, February.

Hicks, N. L. (1979) Growth vs. Basic Needs: Is there a Trade-off? World Development 7, 985-994.

Hulten, C. R. (1973) The Measurement of Total Factor Productivity in United States Manufacturing: 1948-66. PhD Dissertation, University of California, Berkeley.

Keller, Wolfgang and Stephen R. Yeaple (2003) Multinational Enterprises, Intentional Trade and Productivity Growth: Firm Level Evidence from the United States. Brown University. (Working Paper).

Kemal, A. R. (1981) Substitution Elasticities in the Large Scale Manufacturing Industries of Pakistan. The Pakistan Development Review 20:1, 1-36.

Kemal, A. R. (1993) Sources of Growth in Pakistan. Report on Economic and Social Well-being for the Eighth Five Year Plan. Islamabad, Pakistan Institute of Development Economics.

Kemal, A. R. and Islam Ahmed (1992) Report of the Sub-Committee on Sources of Growth in Pakistan, Islamabad. Pakistan Institute of Development Economics, Islamabad, Pakistan.

Kemal, A. R., Musleh ud Din, and Usman Qadir (2002) Global Research Project: Pakistan Country Report, Global Research Project Workshop, Dhaka.

Kendrick, John W. (1973) Postwar Productivity Trends in United States: 1948-66. New York: National Bureau of Economic Research.

Khan, Ashfaque H. (1989) The Two-level CES Production Function for the Manufacturing Sector of Pakistan. The Pakistan Development Review 28: 1, 1-12.

Khan, Safdar Ullah (2006) Macro Determinants of Total Factor Productivity in Pakistan. State Bank of Pakistan. (Working Papers, February, No. 10.)

Kim, Sangho and Young Hoon Lee (2006) The Productivity Debate of East-Asia Revisited: A Stochastic Frontier Approach. Applied Economics 38, 1697-1706.

Kruger, Jens J. (2003) The Global Trends of Total Factor Productivity: Evidence from the Nonparametric Malmquist Index Approach. Oxford Economic Papers 55, 265-286.

Kumbhakar, Subal C. (2003) Factor Productivity and Technical Change. Applied Economics Letters 10, 291-297.

Mahmood, Z. (1989) Derived Demand for Factors in the Large-scale Manufacturing Sector of Pakistan. The Pakistan Development Review 28: 4, 731-742.

Mahmood, Z. (1992) Factor Price Shocks, Factor Substitution and Their Implications for Policy. International Economic Journal 6: 4, 63-73.

Mahmood, Z. and Rehana Siddiqui (2000) State of Technology and Productivity in Pakistan's Manufacturing Industries: Some Strategic Directions to Build Technological Competence. The Pakistan Development Review 39:1, 1-21.

Meeusen, W. and J. Van Den Broeck (1997) Efficiency Estimation from Cobb-Douglas Production Function with Composed Error. International Economic Review 18, 435-444.

Nehru, V. and A. Dhareshwar (1993) A New Database on Physical Capital Stock: Source, Methodology and Result. Revista De Analisis Economico 8, 37-59.

Nikaido, Yuko (2004) Technical Efficiency of Small Scale Industry Application of Stochastic Production Frontier Model. Economic and Political Weekly February, 7.

Pakistan, Government of (1999) 50 Years of Pakistan in Statistics. Volume I-IV. Federal Bureau of Statistics, Islamabad.

Pakistan, Government of (Various Issues) Pakistan Education Statistics. Academy of Education Planning and Management, Ministry of Education, Islamabad.

Pakistan, Government of (Various Issues) Census of Manufacturing Industries. Statistics Division, Federal Bureau of Statistics.

Pakistan, Government of (Various Issues) Labour Force Survey. Statistics Division, Federal Bureau of Statistics.

Pakistan, Government of (Various Issues) Pakistan Economic Survey. Finance Division.

Pakistan, Government of (Various Issues) Pakistan Statistical Year Book. Statistics Division, Federal Bureau of Statistics.

Perelman, S. (1995) R&D, Technological Progress and Efficiency Change in Industrial Activities. Review of Income and Wealth 41, 349-366.

Perli, R. and P. Sakellaris (1998) Human Capital Formation and Business Cycle Persistence. Journal of Monetary Economics 42, 67-92.

Rattso, J. and H. Stokke (2003) Learning by Doing and Domestic and Foreign Technology Spillovers in Thailand: Some Empirical Evidence. Nordic Journal of Political Economy 29:1, 47-66.

Robinson, Sherman (1971) Sources of Growth in Less Developed Countries: A Cross-Section Study. Quarterly Journal of Economics 85:3, 391-408.

Romer, P. M. (1990) Endogenous Technical Change. Journal of Political Economy 98, S71-S102.

Romer, P. (1986) Increasing Returns and Long-run Growth. Journal of Political Economy 94, 1002-1037.

Ruhul, A. Salim (2006) Measuring Productive Efficiency Incorporating Firms' Heterogeneity: An Empirical Analysis. Journal of Economic Development 31:1, 135-147.

Schultz, T. W. (1962) Reflections on Investment in Man. Journal of Political Economy 70, 1-8.

Schultz, T. W. (1964) Transforming Traditional Agriculture. New Haven: Yale University Press.

Solow, R. M. (1957) Technical Change and Aggregate Production Function. Review of Economics and Statistics 39:3, 312-320.

State Bank of Pakistan (Various Issues) Annual Reports. Karachi: State Bank of Pakistan.

Tariq, S., Karamat Ali, and Abdul Hamid (1997) Factor Substitution, Technical Efficiency and Employment Generation Prospects in the Large Scale Manufacturing Industries of Pakistan. Journal of Research (Hum.) 14, 141-152.

Taymaz, E. and Bulin Saatci (1997) Technical Change and Efficiency in Turkish Manufacturing Industries. Journal of Productivity Analysis 8, 461-475.

UNDP (2006) Human Development Report 2006. New York: Oxford University Press.

World Bank (Various Issues) World Development Reports. World Bank.

Yanrui, Wu (1995) Productivity Growth, Technological Progress, and Technical Efficiency Change in China: A Three Sector Analysis. Journal of Comparative Economics 21:2, 207-229.

Zellner, A. (1962) An Efficient Method of Estimation Seemingly Unrelated Regressions and Tests for Aggregation Bias. Journal of American Statistical Association, June.

Zheng, J. Xiaoxuan Liu and Arne Bigsten (2003) Efficiency, Technical Progress, and Best Practice in Chinese State Enterprises (1980-1994). Journal of Comparative Economics 31:1, 134-152.

(1) Pakistan Economic Survey 2005-06 and Labour Force Survey 2005-06, Government of Pakistan.

(2) Knowledge is a non-rival good as knowledge of one thing can simultaneously be used by others without additional cost.

(3) Perli and Sakellaris (1998) estimated that expenditures related to on the job training (OJT) in 1987 were about $165 billion, while total educational expenditure was about $331 billion. These numbers suggest that OJT expenditure may account for as much as half of total educational expenditure in the US.

(4) Human resource development activities like education, professional and vocational training, R&D activities etc.

(5) The coefficients of Translog function can also be estimated from Equation (2) by using OLS technique. However, there may occur multicollinearity problem (as labour and capital increase with a specific ratio which results in the existence of a relationship between two explanatory variables and this specific relationship causes multicollinearity problem). In order to overcome this problem we have estimated cost share equations by applying SURE. Parameters of variable H can be estimated from cost share of labour and cost share of capital by using equality constraints.

(6) Stochastic Frontier Approach (SFA) was introduced by Aigner, et al. (1977) and Meeusen and Van den Broeck (1977).

(7) Five years lag enrolment was used for secondary education and four years lag for enrolment in other categories like professional and vocational institutions and enrolment in universities etc.

(8) The value of coefficient for human capital is calculated from the constraints as explained in methodology.

Abdul Hamid is Income Distribution Specialist in the Centre for Poverty Reduction and Social Policy Development (CPRSPD). A joint project of UNDP and Planning Commission, Government of Pakistan. J. Hanns Pichler is Professor of Economics at Vienna University of Economics and Business, Austria.
Table 1
Translog Production Function Estimates for the
Manufacturing Sector
(1971-72 to 2000-01)

                       Dependent Variable
Independent
Variables                 Value-Added

Constant                 3.52 ***
                       (77.68)
A                        0.02
                        (7.95)
In [K.sub.it]            0.650 ***
                        (2.82)

In [L.sub.it]            0.15
                        (0.84)

In [H.sub.it]            0.20 (c]
(In [K.sub.it]) (2)      0.023
                        (1.18)

(In [L.sub.it]) (2)     -0.0008
                        (0.60)

(In [H.sub.it]) (2)      0.0038 (e)
lnL*InK                 -0.0055
                        (0.38)

InL*1nH                 -0.0031
                        (0.22)

InK*lnH                 -0.0007
                        (0.04)

[R.sup.2]                0.93

Adj-[R.sup.2]            0.92

SER                      0.005

DW- stat                 1.66

N                       30

Notes: Values in parenthesis are t-ratios.

*** Significant at 1 percent level.

** Significant at 5 percent level.

* Significant at 10 percent level.

The Method of Estimation is Zellner's Seemingly .
Unrelated Regression Technique (SURE).

e= Values of parameters derived from constraints .

Where:

SER= Standard Error of Regression;
DW-stat= Durbin-Watson stat; N= Number of Observations;
K= Capital Stock; L= Labour Employed; H= Human Capital

Table 2
The Contribution of Factor Inputs, Technological Change and
Technical Efficiency (TFP) to Value-added Growth in Manufacturing
Sector (Calculated frona Translog Production Function Estimates
in Table 1)

                                      Shares

                                  Estimated Based
Variables                         on Translog PF

Value-added Growth Rate               6.60
Labour Growth Rate                    3.17
Capital Growth Rate                   5.89
HK Growth Rate                        5.02
Share of Capital in Value-added       0.65
Share of Labour in Value-added        0.15
Share of HK in Value-added            0.20
Weighted Capital Growth Rate          3.83
Weighted Labour Growth Rate           0.48
Weighted HK Growth Rate               1.00
Total Factor Input Growth
Share in Value-added Growth           5.31
Technological and Technical
Growth (TFP) Share in                 1.29
  Value-added Growth

Major Determinants of Growth as
Percentage  of Value-added Growth

Capital Contribution                   58
Labour Contribution                     7
HK Contribution                        15
Total Factor Inputs Contribution       80
Technological Change and
Technical Efficiency (TFP)             20
  Contribution
Total                                 100
Share of HK, Technological Change
  and Technical Efficiency             35

Source: Estimations is Table 1.

Table 3
Absolute and Relative Contributions of Major Determinants of
Growth to Manufacturing Value-added

                                            Absolute        Relative
              Estimated                   Contribution    Contribution
               Standard     Estimated          to              to
Variables     Deviations   Coefficients    Value-added    Value-added

Ln(K)           0.612         0.650           0.398          0.621
Ln(L)           0.260         0.150           0.039          0.061
Ln(H)           0.381         0.200           0.076          0.119
SD of
  Dependent
Variable
  (Y)           0.641           --             --              --

Where:

Y= Value-added in Manufacturing;

K= Capital Stock in Manufacturing;

L= Labour Employed in Manufacturing;
HK= Human Capital; and

SD= Standard Deviation; Estimated co-efficients are taken from
Table 1.

Table 4
Manufacturing TFP Change Index (1972-73 to 2004-05)
Based on Translog PF Estimates (3-Years Moving Average)

                                               Aggregated
Year       Y *      K *       L *      HK *      Inputs

1971-72
1972-73
1973-74   0.0501   0.0232    0.0157   0.0041     0.0430
1974-75   0.0269   0.0233    0.0422   0.0041     0.0697
1975-76   0.0125   0.0464    0.0110   0.0082     0.0656
1976-77   0.0452   0.0713    0.0117   0.0126     0.0957
1977-78   0.0675   0.0880    0.0119   0.0156     0.1154
1978-79   0.0949   0.0858    0.0079   0.0152     0.1089
1979-80   0.0970   0.0781    0.0043   0.0138     0.0963
1980-81   0.1166   0.0621    0.0004   0.0110     0.0735
1981-82   0.1031   0.0484    0.0009   0.0086     0.0578
1982-83   0.0935   0.0410    0.0032   0.0073     0.0515
1983-84   0.0769   0.0417    0.0052   0.0074     0.0543
1984-85   0.0783   0.0426    0.0016   0.0075     0.0518
1985-86   0.0771   0.0424    0.0087   0.0075     0.0585
1986-87   0.0841   0.0340   -0.0004   0.0060     0.0396
1987-88   0.0704   0.0259    0.0043   0.0046     0.0348
1988-89   0.0643   0.0200   -0.0031   0.0035     0.0205
1989-90   0.0516   0.0225   -0.0004   0.0040     0.0261
1990-91   0.0673   0.0268    0.0008   0.0047     0.0323
1991-92   0.0654   0.0355    0.0012   0.0063     0.0430
1992-93   0.0637   0.0400   -0.0085   0.0071     0.0386
1993-94   0.0478   0.0426   -0.0092   0.0075     0.0410
1994-95   0.0461   0.0309   -0.0132   0.0055     0.0233
1995-96   0.0322   0.0237    0.0149   0.0042     0.0428
1996-97   0.0143   0.0172    0.0067   0.0030     0.0269
1997-98   0.0112   0.0192    0.0071   0.0034     0.0297
1998-99   0.0128   0.0186    0.0074   0.0033     0.0293
1999-00   0.0464   0.0177    0.0114   0.0031     0.0322
2000-01   0.0505   0.0148    0.0332   0.0026     0.0505
2001-02   0.0698   0.0142    0.0188   0.0025     0.0355
2002-03   0.0908   0.0169    0.0219   0.0030     0.0418
2003-04   0.1156   0.0194    0.0041   0.0034     0.0270
2004-05   0.1151   0.0206    0.0072   0.0036     0.0314

Average   0.0644   0.0361    0.0072   0.0064     0.0496

Year       TFP *     TFPI

1971-72
1972-73             100.00
1973-74    0.0071   100.71
1974-75   -0.0428    96.43
1975-76   -0.0531    91.12
1976-77   -0.0504    86.08
1977-78   -0.0479    81.29
1978-79   -0.0140    79.89
1979-80    0.0008    79.96
1980-81    0.0431    84.28
1981-82    0.0453    88.81
1982-83    0.0420    93.01
1983-84    0.0227    95.28
1984-85    0.0265    97.93
1985-86    0.0185    99.78
1986-87    0.0445   104.24
1987-88    0.0356   107.80
1988-89    0.0438   112.18
1989-90    0.0255   114.73
1990-91    0.0350   118.23
1991-92    0.0225   120.48
1992-93    0.0251   122.99
1993-94    0.0068   123.67
1994-95    0.0229   125.96
1995-96   -0.0106   124.90
1996-97   -0.0126   123.64
1997-98   -0.0185   121.79
1998-99   -0.0165   120.14
1999-00    0.0142   121.56
2000-01   -0.0001   121.55
2001-02    0.0342   124.97
2002-03    0.0490   129.87
2003-04    0.0887   138.74
2004-05    0.0837   147.11

Average    0.0147

Notes:

Y= Manufacturing Value-added; K= Capital Stock;

L= Labour Employed; HK= Human Capital;

TFP= Total Factor Productivity;

TFPI= Total Factor Productivity Index;

Where over dotes show the change over time;

Growth rates for K, L and HK are weighted growth rates;

Weights are taken from estimated Translog Production Function in
Table 1.
Gale Copyright: Copyright 2009 Gale, Cengage Learning. All rights reserved.