The Productivity Analysis of Industrial Clusters in India · 2019. 1. 8. · Cluster Development Approach. 2. Literature Survey Pioneering contribution for Cluster Development Approach

indiastat.com March - April, 2013 1 socio - economic voices

The Productivity Analysis of Industrial Clusters in India (Dr.E. Bhaskaran.B.E, M.B.A., Ph.D., M.I.E., (D.Sc)., Deputy Director of

Industries and Commerce, Government of Tamil Nadu) 1. Introduction

A cluster in a district consists of 100 or more registered MSME units which are engaged in

manufacturing the same product as per ASICC 2000 (at 5 digit-level). There were 2443 clusters

covering 321 products in registered MSME sector. These clusters had a share of 45.92% in total

number of units, 34.58% in total employment, 36.12% in Original Value of Plant and Machinery,

33.64% in total Market Value of Fixed Assets and 19.01% in total Gross Output of registered

MSME sectors [1].

For inclusive growth and sustainable development most of the MSMEs has adopted the

Cluster Development Approach.

2. Literature Survey

Pioneering contribution for Cluster Development Approach (CDA) were made by Porter (1990,

1998), Harrison (1992), Humphrey and Schmitz (1995), Nomisma (1996) and Tewari (1997).

International experiences show that micro, small and medium enterprises are unable to face

competitive pressure because of their isolation and ineffective linkages with relevant support

organizations and commercial service providers. In developed countries like Italy, USA, France

and Germany, clusters of SMEs bound together in strong and dynamic networks and relying upon

strong linkages with relevant local / national support institutions have proved competitive in the

global context. Cluster exists in developing countries too like India, Brazil, Peru & Mexico. There

exists an Automotive Industrial Cluster in Chennai where before and after intervention it is found

that there is increase in Infrastructure Interrelationships, Technology Interrelationships,

Procurement Interrelationships, Production Interrelationships and Marketing Interrelationships

[2][3] and Lean Manufacturing techniques also implemented in this cluster[4].According to Fourth

All India Census[1] it is estimated that 2443 Clustersexists in India and share of Clusters in

Registered MSME Sector is given in Table 1.

Gujarat reported the highest number of clusters with 369 clusters. States of Uttar Pradesh and

Tamil Nadu reported 359 and 350 clusters, respectively.

MSME Clusters is talk of the day for the entrepreneurs in India. Hence, there is need for study on

State wise Technical Efficiency of MSME clusters in India for the benefit of MSMEs.


Table 1 : Share of Clusters in Registered MSME Sector

No. of working

Enterprises

Employment Original Value of

Plant and Machinery (Rs. Crore)

Fixed Investment (Rs. Crore)

Gross Output

(Rs. Crore)

2443 Clusters 718140 3244507 37937.11 151096.49 134463.49 Registered MSME Sector

1563974 9309490 105024.60 449138.44 707510.30

Share of 2443 Clusters (%)

45.92 34.85 36.12 33.64 19.01

Source: Fourth All India Census of MSMEs

3. Objectives of the Study

The objectives are

1. To Study on the Correlation Coefficient Analysis of the dependent and Independent

variables.

2. To Study the Technical Efficiency (θ) under Constant Returns to Scale (CRSTE) and

Variable Return to Scale (VRSTE).

3. To study the Peers, Peer Counts and Peer Weights (λi) of MSME Clusters in India

4. To Study the Input Slacks (S-) and Output Slacks (S+) of MSME Clusters in India.

5. To Study the Input and Output Targets of MSME Clusters in India.

4. Methodology

The methodology adopted is collection of data from Final Report, Fourth All India Census of

Micro, Small and Medium Enterprises, 2006-07, registered sector of Ministry of Micro, Small and

Medium Enterprises, Government of India and analyzing with Data Envelopment Analysis of Input

Oriented Banker Charnes Cooper (BCC) Model by taking No of Clusters, No of Working

Enterprises, employment in nos, Original Value of P & M (Rs In Crores), Market Value of Fixed

Assets (Rs In crores) corresponding to 25 DMUs (Clusters) is taken as input data (I) and Gross

Output (Rs In Crores) is taken as output data (O). A DMU is efficient if θ = 1, S- = 0 and S+ = 0.


Table 2: Physical and Financial Performance S.No. State and

UT (I)No of Clusters

(I)No of Working

Enterprises

(I)Employment (I)Original Value of

P & M (Rs In Crores)

(I)Market Value of

Fixed Assets (Rs In

crores)

(O)Gross Output (Rs In

Crores)

1 Haryana 42 8000 97000 717.34 3183.41 8494.98

2 Himachal Pradesh

18 5000 7000 24.52 67.17 271.63

3 Jammu & Kashmir

10 2000 4000 7.91 1353.89 915.69

4 Punjab 82 17000 136000 950.36 4468.41 14536.24

5 Rajasthan 77 14000 83000 1146.46 3098.79 6060.82

6

Andaman & Nicobar Islands

1 10 1000 1.70 3.09 7.89

7 Bihar 81 16000 44000 85.25 1291.56 418.26

8 Jharkhand 26 5000 13000 27.99 54.04 205.68

9 Orissa 16 2000 9000 31.26 74.56 171.59

10 West Bengal

47 8000 35000 224.06 934.65 1504.34

11 Andhra Pradesh

68 15000 97000 1035.45 3028.27 8818.53

12 Karnataka 227 66000 291000 1131.85 2904.35 7839.47

13 Kerala 255 84000 319000 1178.14 6738.12 6682.12

14 Tamil Nadu 350 153000 815000 6041.55 22120.69 29030.16

15 Goa 2 10 100 1.49 8.70 16.93

16 Gujarat 369 129000 609000 21396.28 83837.08 17619.44

17 Maharastra 69 28000 292000 3006.75 12083.26 20582.11

18 Chattisgarh 41 11000 17000 25.00 37.53 75.38

19 Madhya Pradesh

228 56000 86000 75.77 180.65 500.98

20 Uttar Pradesh

359 86000 248000 714.40 5329.30 9155.96

21 Assam 17 3000 10000 23.37 46.34 727.69

22 Manipur 4 1000 3000 3.27 8.95 21.65

23 Meghalaya 4 1000 2000 8.24 12.62 38.67

24 Mizoram 6 1000 7000 24.19 48.40 91.04

25 Uttarakhand 44 10000 18000 54.51 182.65 675.94 Source: Fourth All India Census of MSMEs


5. Productivity Analysis

5.1 Physical and Financial Performance of MSME Clusters The correlation coefficient Analysis is given in the table 3.

Table 3: Correlation Coefficient Analysis Variables (I)No of

Clusters (I)No of Working

Enterprises

(I)Employment (I)Original Value of

P & M (Rs In Crores)

(I)Market Value of

Fixed Assets (Rs In

crores)

(O)Gross Output (Rs In

Crores)

(I)No of Clusters

1.00

(I)No of Working Enterprises

0.96 1.00

(I)Employment 0.85 0.95 1.00 (I)Original Value of P & M (Rs In Crores)

0.60 0.67 0.71 1.00

(I)Market Value of Fixed Assets (Rs In crores)

0.61 0.67 0.71 1.00 1.00

(O)Gross Output (Rs In Crores)

0.64 0.74 0.89 0.58 0.57 1.00

Source: Computed Data There is high degree of relationship exists between the No. of Clustersand No of Working

Enterprises, No of Working Enterprises and Employment and Employment and Gross Output (Rs

In Crores).

5.3 Data Structure

The No of Clusters, No of Working Enterprises, employment in nos, Original Value of P &

M (Rs In Crores), Market Value of Fixed Assets (Rs In crores), corresponding to 25 DMUs

(Clusters) is taken as input data s and Gross Output (Rs In Crores) is taken as output data.


5.3.1 Data Envelopment Analysis- Banker, Charnes And Cooper Model (BCC Model)

5.3.2 Input Oriented BCC Model (BCC-I)

Min Z0 = θ –ε 1 S+ - ε 1 S-

θ, λ, S+ , S-

Subject to Y λ - S+ = Y0 θX0 - X λ -S- = 0

1 λ ≥ 1

λ, S+ ,S- ≥0

5.3.3 Model Description

The Scalar variable θ appears in the primal problem, is the reduction applied to all inputs

of DMUs to improve efficiency. This reduction is applied simultaneously to all inputs and results in

a radial movement toward the envelopment surface. The presence of non-Archimedean

(Infinitesimal constant) ε in the primal objective function effectively allows the minimization overθ

to preempt the optimization involving the slacks. Thus, the optimization can be computed in a

two-stage process with

i) maximal reduction of inputs being achieved first via θ

ii) then in the second stage movement on to the efficient frontier is achieved via the

positive input and output slack variables (S-,S+ )

Here, the constraint

1 λ ≥ 1 is known as convexity constraints, which will admit variable return to scale (VRS).

The above discussion leads to form the following statement.

A DMU is efficient if and only if

a) θ = 1,

b) All slacks are zero. S- = 0 and S+ = 0.

5.3.4 Computing Methodology

Initially we consider Harayana, as the studied DMU and the LP Model is formulated as

given below

Min θ0

Subject to

8494.98 λ1 + 271.63 λ2 +…………675.94 λ25≥ 8494.98 Output Constraints


42 θ0 - 42 λ1- 18λ2- ………………. 44 λ25≥0 Input Constraints

8000 θ0 - 8000 λ1- 5000λ2- …………10000 λ25≥0 Input Constraints

970000 θ0 - 970000 λ1- 7000λ2- … 18000 λ25 ≥0 Input Constraints

717.34θ0 - 717.34 λ1- 24.52 λ2- ………54.51 λ25 ≥0 Input Constraints

3183.41θ0 - 3183.41 λ1- 67.17 λ2- ……182.65λ25 ≥0 Input Constraints

λ1+λ2+… λ25=1.

λ1,λ2,… λ25≥0, θ0 is unrestricted.

By solving the above and continuously changing the studied DMUs we get the values we

get the value of λi’s and θi’s for each DMU.

5.3.5 Efficiency Scores

The efficiency is calculated using DEAP Software. The data entry programme is given in table 4.

Table 4: Input File

eg6-dta.txt DATA FILE NAME

eg6-out.txt OUTPUT FILE NAME

25 NUMBER OF FIRMS

1 NUMBER OF TIME PERIODS

1 NUMBER OF OUTPUTS

5 NUMBER OF INPUTS

0 0=INPUT AND 1=OUTPUT ORIENTED

1 0=CRS AND 1=VRS

0 0=DEA(MULTI-STAGE), 1=COST-DEA, 2=MALMQUIST-DEA

Source: Computed Data

The instruction file programme is given in table 5.

Table 5: Instruction File Instruction file = EG5-ins.txt


Data file = eg5-dta.txt Input orientated DEA Scale assumption: VRS Slacks calculated using multi-stage method


The result obtained namely crste= technical efficiency (θi) from CRS DEA, vrste = technical

efficiency (θi) from VRS DEAand scale efficiency= crste / vrsteis given in table 6.All subsequent

tables refer to VRS results.

Table 6: Technical Efficiency Summary S.No. State crste(θi) Vrste(θi)

scale

1 Haryana 1.00 1.00 1.00 -

2 Himachal Pradesh 0.49 0.49 1.00 irs

3 Jammu & Kashmir 1.00 1.00 1.00 -

4 Punjab 1.00 1.00 1.00 -

5 Rajasthan 0.67 0.67 1.00 drs

6 Andaman & Nicobar Islands 0.88 1.00 0.88 irs

7 Bihar 0.13 0.13 0.98 irs

8 Jharkhand 0.24 0.28 0.87 irs

9 Orissa 0.29 0.29 0.98 irs

10 West Bengal 0.43 0.43 1.00 -

11 Andhra Pradesh 0.87 0.87 1.00 drs

12 Karnataka 0.50 0.80 0.63 drs

13 Kerala 0.30 0.35 0.84 drs

14 Tamil Nadu 0.44 1.00 0.44 drs

15 Goa 1.00 1.00 1.00 -

16 Gujarat 0.27 0.35 0.76 drs

17 Maharastra 1.00 1.00 1.00 -

18 Chattisgarh 0.13 0.18 0.68 irs

19 Madhya Pradesh 0.21 0.21 0.99 irs

20 Uttar Pradesh 0.47 0.81 0.57 drs

21 Assam 1.00 1.00 1.00 -


22 Manipur 0.23 0.65 0.35 irs

23 Meghalaya 0.26 0.52 0.50 irs

24 Mizoram 0.28 0.37 0.77 irs

25 Uttarakhand 0.47 0.48 0.99 drs

mean 0.54 0.64 0.85


Figure 1: Constant / Variable Return to Scale Efficiency and Scale Efficiency


The States like Haryana, Jammu and Kashmir, Punjab, Andaman and Nicobar Island,

Tamil Nadu, Goa, Maharastra and Assam has got variable returns to scale as 1 (θ =1 ) and is

highly efficient.For a long time DEA models were based on Constant Returns to Scale (CRS) and

it has been criticized as a limiting factor for the application of DEA. Many economists viewed CRS

assumption as over restrictive and preferred alternative methodologies. Banker et al (1984) for

the first time introduced the VRS in DEA models through convexity constraints and thereafter

remarkable change has led to make changes in CCR DEA models of MSME Clusters proud to

say that the RTS of 10States has increasing Variable Return to Scale. For other 7States it has

constant RTS and there is 8 states havingdecreasing RTS.


5.3.6. Input and Output Slacks of MSME Clusters in India

The Input and Output Slacks is given in table 7.

Table 7 :Summary of Input Slacks: (S- ) & Output Slacks:(S+ )

DMU Output Slacks: (S+

)

Input Slacks: (S- )

Output 1 Input 1 Input 2 Input 3 Input 4 Input 5

1 0.00 0.00 0.00 0.00 0.00 0.00 2 0.00 2.17 1515.70 0.00 2.00 0.00 3 0.00 0.00 0.00 0.00 0.00 0.00 4 0.00 0.00 0.00 0.00 0.00 0.00 5 0.00 15.81 2132.17 0.00 384.30 0.00 6 0.00 0.00 0.00 0.00 0.00 0.00 7 0.00 1.36 598.27 1024.52 0.00 0.00 8 0.00 1.78 552.10 143.98 0.00 0.00 9 0.00 0.33 0.00 0.00 1.60 0.00 10 0.00 6.34 972.36 0.00 6.91 0.00 11 0.00 4.14 1860.25 647.37 335.21 0.00 12 0.00 131.13 42592.20 157920.27 405.24 0.00 13 0.00 47.11 20848.60 50219.07 0.00 0.00 14 0.00 0.00 0.00 0.00 0.00 0.00 15 0.00 0.00 0.00 0.00 0.00 0.00 16 0.00 55.23 23048.82 0.00 5573.94 21322.01 17 0.00 0.00 0.00 0.00 0.00 0.00 18 0.00 5.07 1738.48 1294.70 0.65 0.00 19 0.00 35.57 9687.66 11196.17 0.00 0.00 20 0.00 237.01 58532.48 116869.21 0.00 1075.28 21 0.00 0.00 0.00 0.00 0.00 0.00 22 0.00 0.99 594.15 1141.74 0.00 0.00 23 0.00 0.11 392.77 0.00 1.68 0.00 24 0.00 0.00 127.51 650.91 2.97 0.00 25 0.00 7.35 2426.46 0.00 0.00 0.00

mean 0.00 22.06 6704.80 13644.32 268.58 895.89 Source: Computed Data

The states which has θ< 1 needs improvement according to the input slacks whereas the

output slacks is constant.10 Clusters, are efficient by having θ = 1, S- = 0 and S+ = 0. The


remaining cluster needs improvement by either decreasing the no. of working enterprises or

employment or to increase gross output. It is impossible to reduce the no. of enterprises or

employment, so the clusters can go in for increase in sales or exports to become efficient like

other 10 clusters.

The S > 0 obtained for other 16 MSME clusters reveals the excess no. of units (S-) or

employment (S-) in the MSME Clusters or shortage in annual sales or exports (S+) as constant.

The summary of Peers, Peer weights, and Peer Counts is given in table 8.

Table 8: Summary of Peers, Weights and Counts

S.No. States / UT peers peer weights (λi)

peer count

1 Haryana 1 1.00 0 2 Himachal Pradesh 15 4 21 0.70 0.00 0.30 0 3 Jammu & Kashmir 3 1.00 6 4 Punjab 4 1.00 11 5 Rajasthan 3 4 15 0.20 0.40 0.39 0 6 Andaman & Nicobar

Islands 6 1.00 6

7 Bihar 15 21 3 0.46 0.43 0.11 0 8 Jharkhand 15 6 21 0.02 0.70 0.28 0 9 Orissa 21 15 6 4 0.18 0.30 0.52 0.00 0 10 West Bengal 21 4 15 0.34 0.09 0.57 0 11 Andhra Pradesh 21 4 0.41 0.59 0 12 Karnataka 21 4 0.49 0.52 0 13 Kerala 4 3 21 0.43 0.33 0.24 0 14 Tamil Nadu 14 1.00 0 15 Goa 15 1.00 10 16 Gujarat 4 17 0.49 0.51 0 17 Maharastra 17 1.00 1 18 Chattisgarh 21 6 0.09 0.91 0 19 Madhya Pradesh 15 3 21 0.32 0.00 0.68 0 20 Uttar Pradesh 4 3 0.61 0.40 0 21 Assam 21 1.00 14 22 Manipur 15 21 6 0.37 0.02 0.62 0 23 Meghalaya 15 21 6 0.35 0.04 0.62 0 24 Mizoram 21 6 4 0.06 0.94 0.00 0 25 Uttarakhand 4 3 21 15 0.01 0.01 0.72 0.27 0



Peer count summary i.e., no. of times each firm is a peer for another, clearly reveals that

Assam is first, Punjab is second and Goa is third even though all of them has got θ =1.

The summary of Input and Output Targets is given in table 9.

Table 9:Summary of Input and Out Put Targets

Sl.No. States / UT Output Targets

Input Targets:

1 1 2 3 4 5 1 Haryana 8495.00 42.00 8000.00 97000.00 717.00 3183.00 2 Himachal Pradesh 272.00 6.72 953.63 3457.06 10.35 33.09 3 Jammu & Kashmir 916.00 10.00 2000.00 4000.00 8.00 1354.00 4 Punjab 14536.00 82.00 17000.00 136000.00 950 4468.00 5 Rajasthan 6061.00 35.92 7272.74 55757.70 385.56 2081.85 6 Andaman & Nicobar

Islands 8.00 1.00 10.00 1000.00 2.00 3.00

7 Bihar 418.00 9.28 1503.46 4755.23 11.17 169.71 8 Jharkhand 206.00 5.42 831.45 3453.25 7.75 14.94 9 Orissa 172.00 4.37 588.08 2646.34 7.51 22.05 10 West Bengal 1504.00 13.96 2484.19 15122.42 89.87 403.99 11 Andhra Pradesh 8819.00 55.09 11203.51 83831.55 566.19 2637.14 12 Karnataka 7839.00 50.47 10209.88 74888.91 500.40 2323.29 13 Kerala 6682.00 42.42 8642.84 61778.19 413.58 2365.64 14 Tamil Nadu 29030.00 350.00 153000.00 815000.00 6042.00 22121.00 15 Goa 17.00 2.00 10.00 100.00 1.00 9.00 16 Gujarat 17619.00 75.37 22609.16 215548.13 1998.91 8351.07 17 Maharastra 20582.00 69.00 28000.00 292000.00 3007.00 12083.00 18 Chattisgarh 75.00 2.49 288.24 1837.50 3.95 7.00 19 Madhya Pradesh 501.00 12.18 2040.26 6814.55 15.92 37.91 20 Uttar Pradesh 9156.00 53.56 11074.89 83859.03 577.90 3237.95 21 Assam 728.00 17.00 3000.00 10000.00 23.00 46.00 22 Manipur 22.00 1.60 54.45 804.05 1.95 5.84 23 Meghalaya 39.00 1.97 125.80 1037.16 2.47 6.74 24 Mizoram 91.00 2.20 238.74 1912.84 5.82 17.58 25 Uttarakhand 676.00 13.61 2336.84 8573.93 26.20 87.17



As per table 9, input and output targets is given for each States to achieve and become highly efficient (θi = 1 and rank=1).

The overall technical efficiency of Clusters in India is64%, so the inefficient Clusters in the States

should work hard to reach 100%.

6. Findings And Conclusions

To conclude,the Productivity analysis of State wise Industrial Clusters in India gives

interesting results. The efficient clusters should be role model for other clusters and for

inclusivegrowth and sustainable development, the inefficient MSME clusters in States should

increase their turnover or decrease in fixed assets as decrease in no. of units and employment is

not practicallypossible. Moreover the MSME Clusters should strengthen infrastructure

interrelationships, technology interrelationships, procurement interrelationships, production

interrelationships and marketing interrelationships to become highly efficient and should make

use of the benefits announced by Government of India on Micro, Small Enterprises Cluster

Development Progaramme (MSECDP), National Manufacturing Competitiveness Programme and

act according to MSMED Act, 2006. The soft and hard intervention on MSE-Cluster Development

Programme, Lean Manufacturing Programme, ISO Certification Programme, and other ICT

programs like Cloud Computing for MSMEs of Government of India will help Industrial Clusters in

India to increase their productivity and efficiency.

7. References

[1] Final Report Fourth All India Census of Micro, Small and Medium Enterprises, 2006-07, Registered Sector, Ministry of Micro, Small and Medium Enterprises, Government of India, pp. 47-48. [2] Bhaskaran. E., “The Technical Efficiency of Auto Cluster in Chennai”, Society of Automotive Engineering (SAE International) and Asia Pacific Automotive Conference, SAE No. 2011-28-0100, pp: 1-8, http://papers.sae.org/2011-28-0100. [3] Bhaskaran.E.,“Technical Efficiency of Automotive Industry Cluster in Chennai”- Journal of Institution of Engineers (India): Series C, Springer Journal No. 40032, ISSN: 2250-0545 (Print Version), Volume 93,no.3,pp.243-249, ISSN: 2250-0553, http://www.springerlink.com/doi:10.1007/s40032-012-0029-x.

[4] Bhaskaran. E, “Lean Manufacturing Auto Cluster at Chennai”, Journal of Institution of Engineers (India): Series C, Springer Journal No. 40032, ISSN: 2250- 0545 and ISSN: 2250-0553, October 2012, Volume 93, Issue 4, pp 383-390 http://www.springerlink.com/openurl.asp?genre=article&id=doi:10.1007/s40032-012- 0035-z. [5] Tim Coelli, A Guide to DEAP version 2.1: A Data Envelopment Analysis (computer)


program, Centre for Efficiency and Productivity Analysis, Australia. http://www.une.edu.au/econometrics/cepa.htm.

The Productivity Analysis of Industrial Clusters in India · 2019. 1. 8. · Cluster Development Approach. 2. Literature Survey Pioneering contribution for Cluster Development Approach

Documents