Measuring the Efficiency of Public Transport Sector in India: An Application of Data Envelopment Analysis Shivi Agarwal Maths Group, Birla Institute of Technology and Sciences, Pilani – 333031, Rajasthan, India Email: [email protected]

This paper measures technical and scale efficiencies of public transport sector in India. The study makes an attempt to provide an overview of the general status of the State Transport Undertakings (STUs) in terms of their productive efficiencies for the years 2004-05 to 2007-08. Efficiencies of the STUs are measured by applying the new slack DEA Model with categorical DMUs (STUs transported in Rural, Hill and Urban area). Fleet size, Total staff and Fuel consumption are considered as inputs and Passenger kilometers as output. The paper concludes that performance of the STUs has not improved over the earlier three years and has improvement in the last year but still very far from the optimal level. Further, results of sensitivity analysis reveal that by and large, efficiency scores are robust and stable. Keywords: DEA, Transport, Efficiency.

Measuring the Efficiency of Public Transport Sector in India: An Application of Data Envelopment Analysis

1. Introduction India as a developing country seems to be riding on the information economy with the potential of being a developed nation in decades to come. However, much progress remains to be achieved in increasing literacy and public awareness, and providing accessible and quality healthcare to the general masses. Education, transport and health, being vital components of human development, play significant role not only in the well being of the people but also contribute substantially to the economic development of a country. Transport sector plays a significant role in the overall development of a nation’s economy. It is a fact that transport is the barometer of economic activity. Road transport is the prime motorized mode of transport linking the remote and hilly areas with rest of the country. Road transport system co-exists in the public and private sector in India. Although, the private sector has an effective participation in the passenger mobility, but its operational activities are very disaggregated and unorganized while the operational activities of public transport sector are well regulated and organized. The State Transport Undertakings (STUs), controlled by the respective state government, are the imperative mode of passenger mobility in public road transport sector. Since STUs are public utility service with a social objective, it is essential to regularly monitor their performance, specifically with a view to identifying appropriate measures including proper investment and pricing policy and to improve their output efficiency. In public transport sector, efficiency measurement is the first step in the evaluation of individual performance of STUs. This study is an attempt in this direction to assess the relative technical efficiency of STUs in India. Passenger road transportation is a “service business” and evaluating the efficiency of a service business is a complex matter. Transport efficiency is often more difficult to evaluate than manufacturing business efficiency, because it is difficult to determine the efficient amount of resources required to produce various service outputs. The manufacturing standard can be used to identify operating inefficiencies through classical

2

cost accounting variance analyses. However, in service organization like road passenger transportation system, it is difficult to identify the specific resources required to provide a specific service output. Since the mathematical relationship between inputs and outputs of STUs is not known clearly, STU efficiency is operationalized using Data Envelopment Analysis (DEA). It is a non-parametric linear programming model that estimates the magnitude of departure from efficiency frontiers for each STU. The DEA model is used to measure the OTE. The DEA is initially proposed by Charnes, Cooper and Rhodes (1978) followed by Banker, Charnes and Cooper (1984). DEA measures the relative technical efficiency of a group of decision-making units (DMUs) by simultaneously evaluating multiple inputs and outputs common to each unit; each DMU is thus assigned an efficiency score [Charnes et al. (1997)]. The efficiency performance of transport sector has been measured by means of ratio analysis and econometric methods. Hensher (1992), Hensher and Daniels (1995) evaluate total factor productivity (TFP) growth of public bus firms while Tretheway et al. (1997) examine the productivity performance of Canadian Railways in transport sector by using translog production function method. Singh (2000) uses index number approach to estimate the growth and relative level of productivity of 21 STUs in India for the period 1983-84 to 1996-97. Data Envelopment Analysis (DEA) has been also used to estimate the productivity performance of transport sector. Hjalmarsson and Odeck (1996) assess the performance of trucks in road construction and maintenance using DEA. However, studies based on bus transport operation using DEA are relatively few. Patankar (1985) assesses the productivity growth of the four largest metropolitan road transport services in India through DEA for the year 1982-83. Ramanathan (1999) applies DEA for the assessment of the productivity of 29 State Transport Undertakings (STUs) in India for the year 1993-94. Odeck (2000) analyses the relative efficiency and productivity growth of 67 units in the Norwegian Motor Vehicle Inspection Agencies for the period 1989-1991. Efficiency measures are calculated by input-oriented DEA and productivity is measured by the MPI. Odeck and Alkadi (2001) evaluate the performance of 47 Norwegian bus companies for the year 1994, within the framework of DEA. Odeck (2005) investigates the target achievements of the 19 operational units of the Norwegian Public Roads Administration (NPRA) charged with traffic safety services. The DEA and

3

MPI framework is applied with a unique constant input, or equivalently, with no inputs. Odeck (2006) determines the impact of inputs on operator’s efficiency and investigates operations characteristics associated with inefficient use of inputs in the Norwegian bus industry for the year 1994 by applying DEA. Agarwal et al.(2006) estimate the relative efficiencies of Uttar Pradesh State Road Transport Corporation for the year 2002-2003 by applying DEA-AR model. The purpose of this study is to evaluate the performance of STUs by providing a mathematical technique to analyse the efficiency with which service is rendered. This study estimates OTE and SE of the STUs, for the period from the year 2001 to 2004. The paper is organized as follows: in section 2 DEA approach and methodology is given. Empirical results and discussions are given in section 3, followed by conclusions in the last.

2. DEA Approach and Methodology This paper measures the overall technical efficiency (OTE), and scale efficiency (SE) of the STUs for the period from 2004-05 to 2007-08. The technical efficiency refers to the extent to which a STU can produce the maximum output from its chosen combination of factor inputs and scale efficiency refers to sub optimal activity levels. DEA is a linear programming (LP) based multi-factor productivity analysis model for measuring the relative efficiency of homogenous set of DMUs. It calculates a maximal performance measure for each DMU relative to all other DMUs in the observed population with the sole requirement that each DMU lie on or below the external frontier. Each DMU not on the frontier is scaled down against a convex combination of the DMUs on the frontier facet closest to it [Charnes et al. (1997)]. DEA is a non-parametric linear programming model that estimates the magnitude of departure from efficiency frontiers for each STU. The DEA model is used to measure the OTE, PTE and SE. DEA methodology has several advantages over the traditional regression-based production function approach. DEA is chosen over other methods because ¾ It handles multiple inputs and multiple outputs; ¾ It does not require a prior weights (as in index numbers);

4

¾ It does not require any specific assumptions about the functional form between inputs and outputs; ¾ It emphasizes individual observations rather than statistical estimates (as in regression analysis); ¾ It is a dynamic analytical decision-making tool that not only provides a “snapshot” of the current efficiency of the DMU compared with the group, but also indicates possibilities for improving relative efficiency; ¾ It uses benchmarking approach to measure STU efficiency relative to others in their group. ¾ It can assist in identifying best-practice or efficient STUs and inefficient STUs within the group. ¾ The DEA results can allow policy makers to develop policies that can assist the relatively inefficient STUs to improve their performance. However, the basic CCR model [Charnes et al. (1978)] and BCC model [Banker et al. (1984)] suffer from one shortcoming; they neglect the slacks in the evaluation of efficiencies. To overcome this shortcoming, efficiency scores can be computed using the New slack model (NSM-DEA) [Agarwal S. (2008)]. Step-wise methodology used to compute the technical and scale efficiencies of the STUs are described as: First Step: Selection of the Homogeneous DMUs We have selected to measure the OTE and SE of 29 STUs using data from CIRT (2005-2008) for the period from year 2004-05 to year 2007-08. A list of these 29 selected STUs is given in the Appendix A.1. Second Step: Selection of Input and Output Variables To evaluate the relative efficiency of the STUs, three inputs, viz., Fleet size (FS), Total Staff (TS), and Fuel consumption (FC) and single output, Passenger kilometers (Pass-Kms) are considered. 1. Fleet Size (number of buses in hundred) comprises the average number of buses on road of a STU; it is representative of the capital input.

5

2. Total Staff (numbers in thousand) refers to the total number of employees working in a STU; it is representative of the labour input. 3.

Fuel Consumption refers to the fuel consumed (in ten thousand kilolitres) which is measured by dividing total earned kilometer by fuel average; it is representative of the material input.

4.

Passenger-kilometers (in Billions) is a measure of service utilization which represents the cumulative sum of the distances ridden by each passenger. It is normally calculated by multiplying the passenger load to the distance between individual bus stops. Table 1: Descriptive Statistics of Inputs and Output YEARS

20042005

20052006

20062007

20072008

INPUTS

OUTPUT

FS

TS

FC

Pass Km

Min

28

720

4.474048

185

Max

19105

117400

4395.085

762554

Mean

3189.414

21555

800.6535

140899.45

S.D.

4312.449

27511.81

1012.972

163600.53

Sum

92493

625095

23218.95

4086084

Min

31

680

4.736462

187

Max

19357

115946

4517.647

820459

Mean

3311.345

21650.9

826.7622

151028.86

S.D.

4289.937

27019.72

1006.507

171757.04

Sum

96029

627876

23976.1

4379837

Min

28

680

4.00641

189

Max

19232

115529

4652.662

878557

Mean

3356.483

21669.93

846.6709

157022.36

S.D.

4247.105

26778.37

1030.318

185367.41

Sum

97338

628428

24553.46

4553648.5

Min

28

650

3.89726

301

Max

19558

114699

4846.463

925866

Mean

3465.069

22380.66

893.3138

169512.31

S.D.

4312.45

26288.83

1071.709

194630.8

Sum

100487

649039

25906.1

4915857

6

The descriptive statistics of the observed data for the selected STUs of the input and output variables are shown in Table 1. There is a substantial variation in the inputs and the outputs across STUs, evidenced by standard deviation, minimum and maximum values. The inputs used are in some cases hundred times larger than that used by other STU. Third Step: Selection of the models In this study, New Slack model (NSM) has been employed. Since the set of sample STUs has three categories, viz., Rural, Hill and Urban, So, we combine the categorical DEA model [Banker and Morey (1986)] in the NSM-DEA model. To decompose OTE into PTE and SE, NSM- VRS model is also applied. Descriptive statistics of the results are given in Table 2. Fourth Step: Selection of the category of the STUs The set of the sample STUs can be classified into three categories as follows. Category 1 consists of STUs operated in the rural areas, category 2 in the Hill areas and category 3 in the urban areas. STUs in Category 1 are in the most advantageous situation while STUs in category 2 are in severe situation. So, we evaluate the efficiency of STUs in category 2 only within the category while STUs in category 3 are evaluated with reference to category 2 and 3 and STUs in category 1 are evaluated with reference to all STUs [Banker and Morey (1986)]. The categorization of the STUs is shown in Appendix A.1. Fifth Step: Calculate the OTE of a STU To describe DEA efficiency evaluation, assume that the performance of the homogeneous set of n decision making units (DMU j ; j=1…n) be measured by DEA. The performance of DMU j is characterized by a production process of m inputs (x ij ; i = 1…m) to yield s outputs (y rj ; r = 1…. s). The study use the following NSM-DEA model,

7

θk = θk −

Min

s 1 ⎡ m sik− srk+ ⎤ + ∑ ⎥ ⎢∑ m + s ⎣ i =1 xik r =1 yrk ⎦

subject to n

∑λ j =1 n

∑λ j =1

jk

yrj − srk+ = yrk

∀r = 1,..., s

x + sik− = θ k xik

∀i = 1,..., m

jk ij

λ jk ≥ 0

(1)

∀j = 1,..., n

srk+ , sik− ≥ 0 ; r = 1,..., s, i = 1,..., m

where srk+ = slack in the r th output of the kth STU, sik−

= slack in the i th input of

the kth STU, λ jk ' s = intensity variables. These represent the weights of the jth STU in the

evaluation of OTE of kth STU. If the optimal value λ *jk of λ jk is non zero then j th STU represents the reference set (peers) of the kth STU and the corresponding optimal value is known as the peer weights of the j th STU and θ k (scalar) is the (proportional) reduction applied to all inputs of DMU k to improve efficiency. This reduction is applied simultaneously to all inputs and results in a radial movement towards the envelopment surface. The non-zero slacks and (or) θ k ≤ 1 identify the sources and amount of any inefficiency that may exist in the kth STU. The reference set shows how inputs can be decreased to make the kth STU efficient. OTE score of the kth STU is given by θ k . Sixth Step: Calculate OTE of every sample STU by using the same procedure from the

year 2004-05 to 2007-08. The detailed information of DEA results is given in Table 2. Seventh Step: Estimate PTE for the kth STU through adjoining the convexity constraint

(2) in the above LPP (1) n

∑λ j =1

jk

=1

(2)

Then the model is known as NSM-VRS DEA model. Eighth Step: Similarly, estimate PTE for every STU from the year 2004-05 to 2007-08. Ninth Step: Calculate SE of every STU given by

SE = OTE / PTE

8

Model Solution

The solutions of the NSM-DEA models are carried out by using MATLAB. 3. Results and Discussions

Table 2 presents the information on descriptive statistics of OTE and SE for the years 2004-05 to 2007-08. The OTE scores indicate that the STUs having value of the efficiency score equal to 1.00 are on the efficient frontier under CRS technology assumption and those having the value less than 1.00 are less efficient relative to the STUs on the frontier. The lower the efficiency score, higher the scope for the potential increase in outputs (while maintaining inputs) relative to the best practice. Some points emerge from the perusal of Table 2. In the first instance, it is clear that the efficiency measures show a much greater spread, as evidenced by the standard deviations and minimum and maximum values. This implies that there is significant variation in each of the efficiency measure and the efficiency scores are falling within very large efficiency range across sample STUs. Secondly, the average efficiency scores for OTE have steadily decreased over the first three years and then increased but SE does not show any clear trend over the sample periods. During the year 2004-05, on average, sample STUs could have used 32.8% more resources than they actually used to produce the same level of output, this potential efficiency gain has increased, by the year 2007-08, to a mere 36.9%. This implies that average performance of the STUs has turned down between the years 2004-05 and 2007-08. The third point is related to the scale efficiency scores. The DEA analysis evinces that there is no clear trend for SE scores over the sample time period. The average scale efficiency score was 0.860 in 2004-05 which indicates that 14% proportionate contraction in all inputs beyond what was used by eliminating pure technical inefficiency would be feasible if the input and output bundles are suitably altered. For the year 2007-08, the SE score has risen to 0.882, revealing that the scale inefficiency has fallen from 14% in 2004-05 to 11.8 % in 2007-08.

9

Table 2: Descriptive Statistics of OTE, SE for the years from 2004-05 to 2007-08 Overall Technical Efficiency

Scale Efficiency

2004-05

2005-06

2006-07

2007-08

2004-05

2005-06

2006-07

2007-08

Min

0.321

0.314

0.270

0.326

0.502

0.544

0.494

0.555

Max

1.00

1.00

1.00

1.00

1.00

1.00

1.00

1.00

Mean

0.672

0.662

0.587

0.631

0.860

0.883

0.866

0.882

S.D.

0.194

0.210

0.205

0.192

0.145

0.148

0.168

0.140

3

3

3

3

3

4

3

No.

of

efficient

3

STUs

(10.3%)

(10.3%)

(10.3%)

(10.3%)

(10.3%)

(10.3%)

(13.8%)

(10.3%)

No. of STUs

29

29

29

29

29

29

29

29

Returns to Scale CRS

3

3

3

3

IRS

16

19

22

19

DRS

10

7

4

7

Fourthly, it is evident from Table 2 that both pure technical inefficiency and scale inefficiency are responsible for overall technical inefficiency. The information on returns to scale (RTS) in Table 2 reveals that about 10.3% STUs were operating at the Most Productive Scale size (MPSS) i.e., these STUs operated at constant returns to scale (CRS) over the sample period. The remaining STUs don’t show the clear trend of RTS. In the year 2004-05, the majority of the scale-inefficient STUs (61.5%) were operating under increasing returns to scale (IRS) and the remaining (38.5%) STUs under decreasing returns to scale (DRS). The former are small STUs that need to increase their size, whereas the latter are larger STUs which would be better off by reducing their size so that they can operate at optimal scale size. In 2006-07, about 88% scale-inefficient STUs were operating under IRS and remaining 12% STUs under DRS and in the year 2007-08, 73.1% scale-inefficient STUs are operating under IRS and remaining 26.9% STUs under DRS. Finally, the DEA analysis also evaluates the set of STUs which construct the production frontier. The detailed results of OTE and SE for the year 2004-05 reveal that out of 29 STUs, only 3 (10.3%) STUs were operating at MPSS i.e., these STUs were overall technical and scale efficient (OTE and SE Score =1). These STUs were S11, S20,

10

and S24. These STUs are on the best-practice frontier within the group and thus form the “reference set” i.e., these STUs can set an example of good operating practice for the remaining inefficient STUs of the group to emulate. MZST (S21) was the most technical and scale inefficient STU. The mean technical efficiency of the sample STUs implies that on average, the STUs may be able to reduce all their inputs by 32.8% to produce the same amount of output. Mean scale efficiency was 0.860, implying that the average size of STUs was not far from the optimal size, although an additional 14% productivity gain would be feasible assuming no other constraining factors, provided they adjusted their STUs operation to an optimal scale. In 2005-06, there is no change in number of the overall technical and scale efficient STUs. The most overall technical and scale inefficient STU was S21. This indicates that the overall technical inefficiency of the STU was due to the disadvantageous scale-size. The mean OTE is showing decline while SE score is improving from 0.86 to 0.883. This clearly implies that, on average, the OTE of STUs have been declined due to inefficient conversion of inputs into output, not due to scale size. In 2006-07, the same 3 STUs were operating at MPSS, i.e., yet again the set of overall technical and scale efficient STUs constitute 3 STUs. NBSTC (S17) was rated as the most overall technical inefficient (OTE = 0.27) STU and MZST (S21) was again rated as the most scale inefficient (SE = 0.49) STU. These results conclude that the OTE of NBSTC have deceleration only due to the inefficient conversion of inputs into output, not due to impact of scale size while the scale size has the large

impact on the

performance of MZST. The mean values of OTE and SE scores imply that STUs have regress in their performance as compare to the last year. In 2007-08, yet the set of overall technical and scale efficient STUs constitute 3 STUs. NBSTC (S17) is again found the worst performer in terms of OTE while its scale efficiency is high (SE = 0.853). This indicates that the performance of NBSTC is poorer only due inefficient utilization of inputs. TMTU (S29) is rated as the most scale inefficient STU and the SE of this STU has gradually decreased over the whole study period.

11

14

Number of STUs

12 10 8 6 4 2 0 1 2004-05

2005-06

0.80-0.99 2006-07

2007-08

0.60-0.79

0.40-0.59

below 0.40

OTE Scores

Figure 1: Distribution of OTE scores across years

The comparative distribution of STUs according to OTE and SE scores across years is shown by Figure 1 and Figure 2, respectively. Figure 1 shows that number of STUs, whose OTE scores are falling in the range from 80 to 99 per cent, have decreased from 5 in 2004-05 and 2005-06 to 1 in 2006-07. The number of STUs with OTE scores falling in the range from 60 to 79% has decreased from 9 in 2004-05 to 6 in 2007-08, while the number of STUs has increased from 9 in 2004-05 to 13 in 2007-08 whose OTE scores are falling in the range from 40 to 59%. This result clearly indicates that performance of STUs is decline. Figure 2 does not show any clear trend of scale efficiency scores over the sample period.

12

20 18 16

Number of STUs

14 12 10 8 6 4 2 0 1 2004-05

2005-06

2006-07

0.80-0.99 2007-08

0.60-0.79

below 0.60

SE Scores

Figure 2: Distribution of SE scores across years

The results show that through out the study period, only three STUs (S11, S20 and S24) have remained relatively efficient. These STUs are efficient with respect to a large number of factors and are probably good example of “global leader” or STUs with a high robustness. Hence, these STUs can be considered the best-practice STUs to be followed by the inefficient STUs to improve their performance.

Sensitivity Analysis

To investigate the robustness of the efficiency scores, a sensitivity analysis has also been conducted. For this, we used new sensitivity analysis [Agarwal S. (2008)]. In order to perform the new sensitivity approach, the inefficient STUs are to be re-evaluated as follows:

13

Min

s 1 ⎡ m sik− srk+ ⎤ + ∑ ⎥ ⎢∑ m + s ⎣ i =1 xik r =1 yrk ⎦

η a ,b = θ k −

subject to

∑λ

j∈ J −{b}

∑λ

j∈ J −{b}

jk

yrj − srk+ = yrk

∀r = 1,..., s

x + s = θ k xik

∀i = 1,..., m

(6.3)

− ik

jk ij

λ jk ≥ 0

j ∈ J − {b}

srk+ , sik− ≥ 0 ; r = 1,..., s, i = 1,..., m

where J = {1,2,…,n}, a ∈ Jn, b ∈ Je, Jn is the set of inefficient STUs and Je is the set of efficient STUs. It is defined as the efficiency of the ath STU to be evaluated under the condition that j = J-{b}, i.e., the bth STU is excluded from the sample set of STUs. This model verifies the robustness of efficiency scores obtained by the NSM, Table 3: Results of the Sensitivity Analysis 2004-05

2005-06

Efficient STUs

Efficient STUs

S11

S20

S24

S11

S20

S24

0.322

0.379

0.322

0.314

0.314

0.354

0.314

1.00

1.00

1.00

1.00

1.00

0.677

0.675

0.674

0.210

0.203

0.203

Min

0.321

Max

1.00

Mean

0.672

0.706

0.684

0.695

0.662

S.D.

0.194

0.200

0.185

0.188

0.210

No. of efficient STUs

1.00

1.00

3

3

(10.3%)

(10.3%) 2006-07

2007-08

Efficient STUs

Efficient STUs

S11

S20

S24

0.338

0.270

0.270

S11

S20

S24

0.326

0.366

0.326

0.326

1.00

1.00

1.00

Min

0.270

Max

1.00

1.00

1.00

1.00

1.00

Mean

0.587

0.658

0.594

0.621

0.631

0.669

0.630

0.671

S.D.

0.205

0.208

0.204

0.217

0.192

0.190

0.195

0.202

No. of efficient STUs

3

3

(10.3%)

(10.3%)

Source: Author’s calculations

14

Table 3 presents the result of the sensitivity analysis with changing the reference set of the inefficient STUs. The results obtained by the NSM reveal that VPM (S11), TRPTC (S20) and CNI (S24) are efficient STUs. So we exclude these STUs one by one from the set of STUs categorical wise by the above stated sensitivity model. These results indicate that NSM-DEA efficiency scores are robust and stable in the sense that removal of any efficient STU from the set does not have any high influence on the mean technical efficiency. The results show that all the three STUs are not the outliers. 4. Conclusions and Policy Implementations

This paper measures technical and scale efficiencies (OTE and SE) of 29 STUs in India through DEA methodology. The detailed information on input and output data reveals that there exist disparities among the STUs The results of DEA models confirm that performance of the STUs has not improved over the earlier three years and has improvement in the last year but still very far from the optimal level. The results of sensitivity analysis indicate that NSM-DEA efficiency scores are robust and stable in the sense that removal of any efficient STU from the set does not have any high influence on the mean technical efficiency. The results show that all the three efficient STUs are not the outliers. Average OTE scores have gradually decreased over the earlier three years and then improved in the last year of the sample period, whereas average SE scores does not have any clear trend. Moreover, number of STUs at MPSS has no change from 2004-05 to 2007-08. The study finds that only 10.3% STUs have remained overall technical and scale efficient for the entire period. These STUs can be considered as best-practice STUs to be followed by the inefficient STUs to improve their efficiency scores within the group. Their performance can be improved by recruiting motivated and trained workers and capacity building of the existing staff through training and orientation programmes. In addition, to improve the productivity and commitment level and work environment of personnel, some motivational policies such as promotions and performance based reward systems should be introduced.

15

Another point requiring attention is the quality of the outputs because facilities apparently efficient are not necessarily producing high quality outputs, and this could render them ultimately inefficient. Our sample is relatively small so we adopted simplistic production model which does not include the quality variable. This perhaps has resulted in missing some cause of quality inefficiency. These inefficiencies may be occurred due to some issues such as the lack of awareness of available facilities and services offered by STUs, lack of consumer awareness, the absence of advocacy groups, lack of adequate equipment, the lack of training institutes and poor transport facilities. The STUs for which the value of efficiency score has been less than one, improved performance could result from diffusion of new technical knowledge, improved managerial practices, and better use of inputs. In order to reduce input slacks in an inefficient STU, downsizing of employee strength are recommended. There is great potential for inefficient STUs to alter their practice patterns making it similar to the best practice STUs in order to significantly reduce their aggregate expenditures. The fleet utilisation has to be increased to become scale efficient for the scale inefficient STUs. Regular training should be imparted to the drivers in new technological buses. Incentives should also be offered to consistently better performing employees. They may be rewarded in public to motivate other employees. Systematic assessment of their periodical work must be made and considered at the time of promotion. Efforts are required to be made in the direction of replacement of old buses and induction of modern buses for raising the efficiency. The development of the road system is not only inadequate but also the available roads are not of quality, particularly in heavy monsoon areas. Road improvements would bring significant savings in fuel consumption, apart from savings in tyre and spare part consumption, reduction in road accidents, and saving in travel time. Use of more advanced fuel-efficient models of buses, better maintenance of fleets and roads and replacement of worn out buses are likely to help in improving the performance of the STUs. Fuel conservation is important factor to improve the efficiency and productivity of the transport services. To attain the better fuel efficiency, rationalise the number of stops, maintenance of the fleet, maintenance of proper tyre pressures and traveling at a steady speed are recommended. It is extremely necessary to create consciousness among the

16

operating staff so that they may appreciate the value of fuel conservation. For this purpose, reward and promotion schemes are recommended. Keeping in view the global competition, it is recommended that STUs should make yearly self-assessment. It should develop a mechanism to evaluate its performance on yearly basis. References

1. Agarwal S., (2008), “Assessment of Efficiency and Productivity of Public Transport Sector using DEA Techniques” Ph.D. Thesis, Indian Institute of Roorkee, Roorkee, India. 2. Agarwal S., Yadav S.P. and Singh S.P. (2006), “A Data Envelopment Analysis Based Efficiency Assessment of Public Transport Sector of Uttar Pradesh in India”, Indian J. Transport Management, 30(1), 5-30. 3. Banker, R.D., A. Charnes and W.W. Cooper, (1984), “Some Models for the Estimation of Technical and Scale Inefficiencies in Data Envelopment Analysis”, Management Science, 30, 1078-1092. 4. Banker R.D. and Morey R.C., (1986), “The use of Categorical Variables in Data Envelopment Analysis”, Management Sciences, 32(12), 1613-1627. 5. Charnes A., Cooper W.W. and Rhodes E., (1978), “Measuring the Efficiency of Decision Making Units”, European Journal of Operational Research, 2, 429-441. 6. Charnes, A., W.W. Cooper, A.Y. Lewin, and L.M. Seiford, (1997) “Data Envelopment Analysis: Theory, Methodology and Application”, Kluwer Academic Publishers, Massachusetts USA,. 7. CIRT, Pune, “Indian Journal of Transport Management”, 29-32, (2005-2008). 8. Hensher D.A., (1992), “Total Factor Productivity Growth and Endogenous Demand: Establishing a Benchmark Index for the Selection of Operational Performance Measures in Public Bus Firms”, Transport Research, 26B (6), 435448. 9. Hensher D.A. and Daniels R., (1995), “Productivity Measurement in the Urban Bus Sector”, Transportation Policy, 2(3), 179-194.

17

10. Hjalmarsson L. and Odeck J., (1996), “Efficiency of Trucks in Road Construction and Maintenance: An Evaluation with Data Envelopment Analysis”, Computers and Operations Research, 23(4), 393-404. 11. Odeck J. (2000), “Assessing the Relative Efficiency and Productivity Growth of Vehicle Inspection Services: An Application of DEA and Malmquist Indices”, European Journal of Operational Research, 126, 501-514. 12. Odeck J. (2005), “Identifying Traffic Safety Best Practice: An Application of DEA

and

Malmquist

Indices”,

Omega,

Available

online

at

www.sciencedirect.com. 13. Odeck J. (2006), “Congestion, Ownership, Region of Operation, and Scale: Their Impact on Bus Operator Performance in Norway”, Socio Economic Planning Sciences, 40(1), 52-69. 14. Odeck J. and Alkadi A. (2001), “Evaluating efficiency in the Norwegian Bus Industry Using Data Envelopment Analysis”, Transportation, 28, 211-232. 15. Patankar P.G., (1985), “Productivity Improvements in Passenger Road Transport Services in Metropolitan Areas”, Journal of Transport Management, 9(6), 13-28. 16. Ramanathan R., (1999), “Using Data Envelopment Analysis for Assessing the Productivity

of

State

Transport

Undertakings”,

Journal

of

Transport

Management, 23(5), 301-312. 17. Singh S.K., (2000), “State Road Transport Undertakings, 1983-84 to 1996-97: A Multilateral Comparison of Total Factor Productivity”, Journal of Transport Management, 24(5), 363-388. 18. Tretheway M.W., Waters W.G. and Fok A.K., (1997), “The Total Factor Productivity of the Canadian Railways, 1956-91”, Journal of Transport Economics and Policy, 31(1), 93-113.

18

Appendix A.1

STUs selected for the study are as follows: STU STU No. Acronym

STU Name

S1 APSRTC Andhra Pradesh State Road Transport Corporation S2 MSRTC Maharashtra State Road Transport Corporation S3 GSRTC Gujarat State Road Transport Corporation S4 UPSRTC Uttar Pradesh State Road Transport Corporation S5 RSRTC Rajasthan State Road Transport Corporation S6 KnSRTC Karnataka State Road Transport Corporation S7 MDU Madurai Division S8 NWKnSRTC North West Karnataka State Road Transport Corporation S9 NEKnSRTC North East Karnataka State Road Transport Corporation S10 STHAR State Transport Haryana S11 VPM Villuparam Division 1 S12 CBE-1 Coimbatore Division S13 SLM Salem Division S14 KUM Kumbakonam Division 1 S15 PRTC Pepsu Road Transport Corporation S16 TN Tamil Nadu State Express Transport Corporation Limited S17 NBSTC North Bengal State Road Transport Corporation S18 SBSTC North Bengal State Road Transport Corporation S19 KDTC Kadamba Transport Corporation Limited S20 TRPTC Tripura Road Transport Corporation S21 MZST Mizoram State Transport S22 BEST Brihan Mumbai Electric Supply & Transport Undertaking S23 DTC Delhi Transport Corporation S24 CNI Chennai Metropolitan Transport Corporation Limited S25 BMTC Bangalore Metropolitian Transport Corporation S26 CSTC Calcutta State Transport Corporation S27 PMT Pune Municipal Transport S28 AMTS Ahmedabad Municipal Transport Service S29 KMTU Kohlapur Municipal Transport Undertakings

State of Operation

Nature of Org

Category of STU

Andhra Pradesh Maharashtra Gujarat Uttar Pradesh Rajasthan Karnataka Tamil Nadu Karnataka

Corp Corp Corp Corp Corp Corp Com Corp

Rural Rural Rural Rural Rural Rural Rural Rural

Karnataka

Corp

Rural

Haryana Tamil Nadu Tamil Nadu Tamil Nadu Tamil Nadu Punjab Tamil Nadu

GD Com Com Com Com Corp Com

Rural Rural Rural Rural Rural Rural Rural

West Bengal West Bengal Goa Tripura Mizoram Mumbai city

Corp Corp Com Corp GD MU

Rural Rural Rural Hill Hill Urban

Delhi Chennai city

Corp Com

Urban Urban

Bangalore city Kolkatta city Pune city Ahmedabad city Kohlapur city

Corp Corp MU MU MU

Urban Urban Urban Urban Urban

Corp. stands for Corporation; Com stands for Company; GD stands for Government Department; MU stands for Municipality Undertakings

19