MEASURING EFFICIENCY OF INDIAN PORTS THROUGH DATA

ENVELOPMENT ANALYSIS: A COMPARATIVE ANALYSIS OF

MUNDRA AND KANDLA PORT IN GUJARAT

Dr. JUSTIN JOHN STEPHEN 1 and Dr. PRAKASHCHANDRA PARMAR 2

1, 2
Assistant Professor, Department of Business Economics, The Maharaja Sayajirao University of Baroda,
Gujarat.

Abstract
The purpose of the study is to evaluate the ownership structure of Mundra and Kandla seaports with a objective
to compare the efficiency of these two ports over the period of 2017-2022. The data on various variables of inputs
and output were collected from the port Management. The study applies data envelopment analysis with input
orientation to calculate the relative efficiency score of Gujarat’s most leading ports for the respective reference
period. The investigation carried out both on singular aspects of individual port and on a cross sectional analysis
of the port. The efficiency estimated on individual ports reveal that the Mundra Port has an average efficiency of
0.92 percent on reported periods. And the efficiency representation of the Mundra Port is consistent compared to
its counterpart with 0.73 percent ofscore. As a result Mundra Port outperforms the Kandla Port. However the
analysis of panel data reveals a fluxuation in these two ports efficiency, indicating that the result outcomes of
cross-sectional data may bemisleading. It was also found that ports exhibiting a mixture of decreasing, increasing
and contant returnedto scale. Moreover Data Envelopemnt Analysis results obtained for individual ports can be
more reliable, because each single port could have its specific and unique contexts in place.
Keywords: Ownership Structure, Technical Efficiency, Data Envelopment Analysis
JEL Classification: C01, C33, O13, and Q43

I. INTRODUCTION
Assessing a firm's performance, whether it is evaluating successes against stated goals and
objectives or the competition, is a fundamental notion for any business. Seaports are no
exception, and performance canonly be assessed through comparison. Seaports are complicated
businesses with diverse inputs and outputs, making direct comparisons between seemingly
similar ports problematic. The many forms of port ownership and organisational systems
worldwide further complicate the situation. Over the last two decades, ownership of one of the
essential commerce entry points into any country, the ports, has shifted from national or
municipal governments to private hands, either entirely or partially. This trend, known as
privatisation, and because of its importance in international trade and business, port performance
evaluationhas piqued the interest of policymakers, industry stakeholders, and scholars
Given that India is bordered on three sides by the sea, its economic growth depends on its
seaports, whichhandle 90% of its export and import. In this regard, large ports play an essential
part in the country's exportand import activities. India's biggest ports account for more than
75% of the country's total trade. The commodities are moved through the logistics chain, and
logistics become crucial for trade competitiveness. As a result, monitoring port efficiency
should be prioritised, and necessary efforts should be made to improve inefficient ports'

949 | V 1 8 . I 0 8
performance. Efficiency analysis allows port owners and authorities to make better judgments
about port development and operations. At the same time, it also enables port users to
(particularly shipping lines) examine the relative competitiveness of ports and make better
decisions aboutport usage.
Considering the preceding, this study investigates the issue of whether ownership influences the
efficiencyof India's major ports. In the context of the above, this study will look at how the
ownership structure of seaports in India has any relation to its performance based on the Data
Envelopment Analysis (DEA) to assess the relative efficiency of a sample of ports – Mundra
port and Kandla port, both located in the stateof Gujarat. Thus, the purpose of this study is to
see if the claim that ownership pattern is a contributing factor to port efficiency is valid. To
answer this question, during the study period 2017-2021, the current paper evaluated the
operational efficiency of the two ports in India, as stated above. These two ports were chosen
because they are both on India's west coast, on the Gulf of Kutch. They have equivalent
catchmentregions and serve comparable hinterlands due to their geographical proximity. It is
expected that the results of this paper will help serve as a guide for governments, port
administrators, and port owners on the different ways in which they can structure their ports to
lead to greater efficiency. The remainder of the research is organised as follows. The profiles
of the two ports are described in section 2. The third part is devoted to a review of the literature.
Section 4 describes the data sources and methodology. The results areprovided in section 5.
Finally, section 6 brings the paper to a close.

2. PROFILES OF KANDLA AND MUNDRA PORT

Kandla Port
Kandla Port (also known as Deendayal Port) is a seaport in the public sector under the
administrative superintendence of the Ministry of Shipping, Government of India, which has
been functioning since the 1950s. The geographical location of this port is on the Gulf of Kutch.
It is a natural port. It is India's largestbulk cargo port. The containers imported through Kandla
Port in the Apr-Aug period of FY 2021-22 increased by 52.16% compared to FY 2020-21. The
increase in import of Heavy melting Scrap and Steel Coils (15.71%) and Bitumen was the key
reason behind this jump.
The gross revenue in the Apr-Aug period of FY 2021-22 increased by 163.03% compared to
thecorresponding period of 2020-21; driven by an increase in import of edible oils, primarily
Crude Palm Oil (+384%) and Soyabean Oil (+260%) as well as Chemicals like Ammonia
(+215.27%) and Methanol (+332.52%).
The containers' export increased by 41.88% between April and August of this FY 2021-22
compared to 2020-21. The major commodities driving this surge are Steel Billets (+774.23%
in terms of FOB Value) and Ceramic Glazed Tiles (+306.16% in FOB Value). 33.88% increase
in the number of Bills of entry in Apr-Aug period of current FY 2021-22 compared to FY 2020-
21. Compared to the corresponding period of 2019-20, there's a slight downtick of 6.18% in
the number of BEs in 2021-22. Still, overall revenue has gone up by 90.04% due to the higher

950 | V 1 8 . I 0 8
import volume of Edible Oils and chemicals like Acrylonitrile, Anhydrous Ammonia, and
Methanol. There has been a 93% increase in the number of shipping bills in theApr-Aug period
of current FY 2021-22 compared to the corresponding period of previous FY 2020-21. The
higher count is attributed to an increase in the export of commodities like Indian White Sugar
(+86% of FOB Value), Cup Brand Indian Basmati Rice (+375.49%), Sharbati Golden Shell
Rice (+375.49%), Steel Billets (+774.23%) and Ceramic Glazed Tiles (+306.16%).
Mundra Port
APSEZ Limited, Mundra (also known as Mundra Port) is a multisector SEZ and the country's
largest privately developed port. It is located in the Northern Gulf of Kutch, Gujarat, on India's
western coast, andhas an area of 100 square kilometers. It is a year-round port. It is also a natural
port, like Kandla. The port'scommercial activities began in October 2001. Dry, Bulk, Break
Bulk, Liquid, Crude Oil, Project Cargo, Cars, and Containers are among the cargo types handled
by APSEZ Limited, Mundra. Mundra port has oneof the deepest draughts among India's ports.
Twelve multi-purpose berths, nine container berths, 2 SBM for crude oil import, four coal
import berths, and Ro-Ro facilities for vehicle traffic are all available at theport. Mundra port's
West Basin is likely the world's largest coal import terminal. Mundra port has 30 activeberths
for dry bulk, break bulk, project cargo, crude oil, LPG, LNG, container cargo, and liquid cargo.
Theautomated terminal has a capacity of 4.67 lakh KL and contains 100 tanks. The Mundra
port's dedicated Container Terminals have a capacity of more than 5.8 million TEUs. There is
also a Pure Car Carrier/ PureCar Truck Carrier (PCC / PCTC) berth with adequate parking
space at this port. There is a pre-dispatch yard with a capacity of 2000 cars and a buffer yard
with a capacity of 2000 cars.
Mundra's APSEZ Limited operates a 64-kilometer private rail network that has been expanded
to accommodate double-stacked container trains. APSEZ Limited, Mundra, has a 380-
kilometer coverage area in India's northern hinterland. The NH 8A, NH 15, and other state
highways connect well to Mundra port. Mundra's APSEZ Limited also has a functioning airport
to handle private jets. Mundra is also well connected to two commercial airports located within
60 kilometers of the city. Mundra port's dry cargo facility can handle any form of dry cargo.
Import freight may be handled at 1500 TPH on a 3.6 km long import conveyor system. The
Mundra port's Agri-Complex can handle 2 MMTPA and has rake handling facilities. A
Fertilizer Cargo Complex is also located at this port. There is a mechanical method for bagging
and loading fertiliser bags straight into railway carriages for quick freight evacuation. This
automated fertiliser bagging facility can load 20 rakes each day. The port also has a lot of storage
capacity in the formof open and covered warehouses. The cargo volume in Mundra Port was
115 MMT from December 1 to December 21, with Foreign Transhipment accounting for 30%
of the total.

951 | V 1 8 . I 0 8
3. LITERATURE SURVEY
Evaluation of port performance is important for policy suggestions, port destination selection,
and researchavenues for various stakeholders. Port performance measurement, which began
with ideas for indicators and has expanded to include efficiency and productivity measurements,
is now being graded based on manyfactors. Several researchers have used the DEA technique
to evaluate port performances. The use of the DEA approach in the port sector is not a new
concept. In many places around the world, several forms of the DEA technique have been
applied to study port production. Unlike the port performance indicators created by UNCTAD
(1976), the advantage of DEA is that other inputs and outputs may be added to the model,
allowing it to provide an overall evaluation of port performance (Wang et al., 2003).This
section reviews the available literature on port efficiency measurement based on DEA in the
Indian context to identify the research gap/s.
Chudasama and Pandya's (2008) study is the first one to measure the efficiency of Indian ports
by using Data Envelopment Analysis (DEA). Their main objective was to bring out the actual
working and performance of 12 major ports of India for the year 2005-06. They found that "the
DEA-BCC model yieldshigher efficiency estimates than the DEA-CCR model, with average
values of 0.98 and 0.86, respectively.Out of the 12 ports, seven were identified as efficient, and
five were relatively inefficient when the DEA- CCR model was applied. When the DEA-BCC
model was used, all the ports except one turned out to be efficient in the analysis." Empirical
results also showed that a large production scale is more likely to be associated with high-
efficiency scores. For instance, the efficiency score and port output correlation were 0.84 for
the DEA-CCR model.
Similarly, the study by Kamble et al. (2010) used data envelopment analysis on chosen input
and output variables to assess the efficiency of the major Indian seaports. Storage facilities, the
number of berths, andthe quantity of cargo handling appliances were the input variables. The
two output variables included wereaverage total turnaround time and average output per ship
berth day. As per this study, "only six of the twelve ports were verified to be operating
efficiently." In contrast, the study by Haralambideset al. (2011) assessed the efficiency of 26
dry ports in India's Jawaharlal Nehru Port Trust region while also considering the critical
problem of container security. Their findings showed that "the public-sector dry port ownership
arrangements and competition rules significantly impacted the development of container
security methodsand, as a result, dry port operational efficiency."
The study by Munisamy and Singh (2011), on the other hand, examined the technical and scale
efficiencyof major container ports in Asia, including India. This study used the non-parametric
DEA technique. According to the findings, Asian container ports had an average technical
efficiency of 48.4 percent, and the technical inefficiency was due to pure technical inefficiency
rather than scale inefficiency. As per the results of the comparison across nations, Bangladesh,
the Philippines, China, Cambodia, India, and Singapore have the most efficient ports in Asia.
Sekar and Deo (2012) used DEA – Additive models to look at the relative efficiency of India's
major portsfrom 1993 to 2011. They choose the inputs and outputs for the study by evaluating

952 | V 1 8 . I 0 8
variables such as the number of berths, berth length, number of equipment, number of staff,
container throughput, and total cargodirectly associated with port efficiency. According to the
findings, "both larger ports (JNPT, Mormugao) and smaller ports (Ennore, Tuticorin) were
efficient. As a result, it was demonstrated that there is no substantial difference in port size and
efficiency." Similarly, in another study, Sekaret al. (2014) used data envelopment analysis to
look at the operational efficiency of a few major Indian ports from 1993 to 2011. They
concluded that the port size had no bearing on port efficiency as both larger ports - Mormugao
and the Jawaharlal Nehru Port Trust, and smaller ports, such as Ennore and Tuticorin, were
found to have efficient port operations throughout.
Dasgupta and Sinha (2016) attempted to identify the effect of liberalization on the efficiencies
of containerterminals of major ports of India. In their study, the efficiency of privately managed
terminals under majorports was compared with public container terminals with the help of the
output-oriented DEA using DEAP (Data Envelopment Analysis (Computer) Programme)
software. The study results showed that "though theefficiency of container terminals is affected
by privatization to a great extent, they depend on other factorstoo."
Iyer and Nanyam's (2021) study looked at the technical efficiency of 26 container terminals in
India during2015–2018 using a data envelopment analysis approach. The study discovered that
"container terminals onIndia's west coast were more efficient than those on the east coast. The
efficiency of container terminals operating under major ports is dropping compared to that of
minor ports." Besides, it was shown that the size of the terminal, which offers economies of
scale, is the most crucial element impacting its efficiency. Furthermore, it was found that there
was no consistent increase in productivity across all container terminals due to private
participation.
In contrast, Mustafa et al. (2021) attempted to compare the technical efficiency of less-explored
South Asian and Middle Eastern ports to East Asian ports to find strategies to improve
their efficiency and management. For 2018, cross-sectional data were collected for 15
container ports in the South & Middle Eastern and East Asian regions, and input and output
variables were created. The DEA-CCR and DEA- BCC models were used to examine the data.
Only one port in South Asia, JNPT in India (from JNPT, Hazira, Pipavav, and Mundra), was
determined to be efficient on the CCR model. Still, the number of efficient ports on the BCC
model increased by 47 percent.
Adler et al. (2021) devised a set of contextual factors, including an absolute measure of
specialisation anda berth-level measure of ownership structure for understanding the efficiency
drivers of ports in India. Thisformulation was applied to major Indian seaports for21 years,
from 1995 to 2015. The findings indicated that "average seaport efficiency has gradually grown
over time. According to the second stage, fixed effectsregressions, specialization, and external
stakeholder(private) participation significantly influenced seaport performance. Surprisingly,
competition between major and minor seaports was found to impact performance negatively.

953 | V 1 8 . I 0 8
Although several studies on measuring port efficiency have been conducted in India, the major
shortcomingof these studies is that they have not focused on analysing the efficiency of the
ports operating under different administrative control, such as the public sector port and private
port. This is the primary motivating factor for the conduction of this study.

4. SOURCES OF DATA AND METHODOLOGY

Data source
The data on various parameters of the performance of the ports were systematically collected
and collatedfrom the respective port managements, viz, the Deendayal Port Trust and Adani
Ports & SEZ Ltd, Mundra.
Methodology
Every firm is concerned about technical efficiency. Technical efficiency refers to how a firm
reduces input consumption in the production process to achieve a certain output level or
maximises output without changing any input units. Many methodologies have been developed
to quantify technical efficiency in various sectors. The non-parametric DEA methodology is
one of the most often utilised measuring methodologies. It is a type of linear programming used
to evaluate the effectiveness of a decision-making unit (DMU). Two popular models are
extensively used. One DEA-CCR assumes constant returns to scale,while the other DEA-BCC
assumes variable returns to scale, with the DEA-CCR with a constant return toscale receiving
the most significant support.
This investigation uses an input-oriented approach to see the possibility of reduction in input
and the consequent effect of change in output. Let a set of DMUj (j=1,…., n) be in the
measurement system. Let (x1j,…,xmj) be the input vector of DMUj and input weight vector
(v1,…, vm),. Let (y1j,…,yqj) be the output vector of DMUj and output weight vector (u1,…, uq).
Assume that each DMUj consumes xij amount of input i to produce yrj amount of output r
and that the input and output of DMUk(k=1,….,n) being investigated are, respectively,
(x1k,…,xmk) and (y1k,…,yqk), where xik≥ 0 and yrk ≥ 0. Let µr=tur and vi= tvi, where t
-1

(i) The input-oriented CCR-DEA model has the following form

Subject to ≤ 1 (j=1,….,n)

954 | V 1 8 . I 0 8
 (ii) The input-oriented BCC-DEA model has the following form

Subject to

The current study assesses the efficiency of two major ports in Gujarat using the above models.
It comparesprivate and public-sector port performance from 2017 through 2022. On the input
side, total berth, numberof cranes, labour engaged, and berth length are being investigated;
container throughput and total traffic are being considered on the output side. The investigation
was divided into three parts. The first section establishes the framework for the Mundra port
analysis. The DEA model creates an efficient frontier froma data sample of the best-performing
decision-making units. The second part examines Kandla Port, and thethird part constructs DEA
from the Mundra panel and the Kandla Port sample's efficient frontier.

5. RESULTS AND DISCUSSION

Mundra Port
The constant return to scale-input-oriented data envelopment analysis results is shown in Table
1. Where, for the year 2017-18, the optimal result of efficiency score is denoted by (theta),
reference weights are denoted by (lambda A, lambda B, lambdaC, lambdaD, lambdaE), and
slack is denoted by (i.number of berths, i.number of cranes, i.number of labour, i.berth length,
throughput in TEC The Mudtra port performed well this year, as evidenced by its efficiency
score of one. Similarly, for the year 2018-2019, theta score, reference weights, and slack imply
0.923668, (0.111404, 0,0,0,0.812265), and (0,0,0,0,1156967, 0.0434), respectively. This year,
we estimate that the port's efficiency was 92.36 percent.In this reference period, the port could
have boosted throughput by 1156967 and improved traffic by 0.0434 to reach the efficient
border. Put another way, and the port should have boosted container traffic by 1.31 percent and
increased container volume by 25% to be more efficient.
Whereas theta score, reference weights and slack for the reference period 2019-2020 suggest
0.831111, (0,0,0,0,0.831111), and (0,0,220.244,0, 544612, 0.0000792) respectively. This year,
the port's
Source: Author's calculation
Note: R=Ranks, Theta=Efficiency Score, Lamda=reference weights, TB=Total Berth, C= No
of Cranes, L= Labor, BL= Berth Length, CTP= Container Throughput, TT= Total traffic.

955 | V 1 8 . I 0 8
 Table 1: Constant return to scale- input-oriented DEA efficiency Results
Year Rank theta λA λB λC λD λE TB C Input and output slacks
L BL CTP TT
2017-18 1 1 1 0 0 0 0 0 0 0 0 0 0
2018-19 4 0.92 0-111 0 0 0 0.81 0 0 0 0 1156967 0.0434
2019-20 5 0.83 0 0 0 0 0.83 0 0 220.2 0 544612 0-0792
2020-21 3 0.96 0-23 0 0 0 0.72 0 0 0 0 0 198.065
2021-22 1 1 0 0 0 0 1 0 0 0 0 0 0
performance was up to 83.11% efficient. And to move to the efficient frontier that is 1, the port
could have reduced the labor by 220.244 units, increased the throughput by 544612, and
improved the traffic by 0.0000792 in this reference period. In other words, to have been
efficient, the port should have increased the containers by 11.29% and improved the traffic by
2.64%.
Subsequently the theta score, reference weights and slack for the reference period 2020-2021
derive 0.960092, (0.231593,0,0,0,0.728499), and (0,0,0,0,0,198.065) respectively. This year,
we obtained that the port's performance was efficient up to 96%. And it could have been
efficient fully if the port could have improved the traffic alone by 198.065 in this reference
period. In other words, to have been efficient, the port should have improved the traffic by
3.50%.
Table 2: VRS input-oriented DEA Efficiency Results
Year CRS-TE VRS-TE SCALE RTS
2017-18 1.000000 1.000000 1.000000 -
2018-19 0.923668 1.000000 0.923668 IRS
2019-20 0.831111 1.000000 0.831111 IRS
2020-21 0.960092 1.000000 0.960092 IRS
2021-22 1.000000 1.000000 1.000000 -
calculation Source: Author's
For the reference period 2021-2022, the data has been extrapolated from October 2021 to
March 2022. Accordingly, the result of theta, reference weights, and slack have been worked
out. The obtained result values denote the fully efficient performance of the Mudhra port in
this particular period. Hence port efficiency analysis of referred years reveals that performance
has been good in 2017-18 & 2021-22. And these years can be a benchmark for subsequent
years to come.
In the table-2, variable return to scale is specified. This gives additional information about the
efficiency of the port in the referred years. For instance, in the year 2017-18 and 2021-22, the
port enjoys constant return to scale, and in the rest of the year, it is faced with increasing return
to scale.

956 | V 1 8 . I 0 8
Kandla Port
The table-3 presents the review of five years' performance analyses of Kandla port. In 2020-21
the port achieved a strong performance index for the rest of the years. The result record of data
envelopment revealsfully efficient scores of 1, and all slacks, both input, and output, are zero.
Since dual conditions are satisfied,i.e., efficiency being 1 and all slacks being zero, the firm has
functioned efficiently in terms of radial, technical and Pareto-efficient.
Kandla port performed efficiently in 2020-21 given the input and output combinations whose
theta is one and slacks are zero. Assessing other years' performance relative to the technically
efficient frontier in 2020- 21 reveals that in 2017-18, 2018-19, 2019-20, and 2021-22, the
deviation from the efficient frontier has been 0.28, 0.53, 0.94, and 0.90 respectively. However,
it can be noticed that the two years, 2019-20 and 2021-22, theta are close to the efficient
frontier.
The optimal result of efficiency score theta, reference weight lambda and slack of input &
output was obtained for the reference period of 2017-18. Theta, reference weights, and slacks
obtained for the year 2017-18 confirm 0.28, (0,0,0.22,0), and (0.10,0.422,0,28.80,0,76180.2)
respectively. Thus, if the port had made an effort to reduce the total number of berths by
0.105693 and the number of cranes by 0.422772, berth length by 28.8013 units, or maximise
the total quantity of traffic by 76180.0, the port's performancein the year 17-18 may have been
improved. This could have enhanced the port's efficiency, allowing it to achieve theta equal to
one. However, because there is no slack in labour, it has been used to its full potentialover the
referred period. Similarly, the port's economic activity in 2018-19 yielded an efficiency score
of 0.54, which is significantly lower than the efficient frontier score of one. The recorded
reference weights and input-output slacks support the following (0,0,0,0.4745,0),
(0.12,0.4802,0,32.71,200.343 ,0). The port'sperformance might have been improved if the total
number of berths was lowered by 0.12, cranes were reduced by 0.4802, berth length was
reduced by 32.71 units, or the throughput was increased
Table 3: Constant return to scale- input-oriented DEA efficiency Results of Kandla Port
Year Theta i-TB i-C i-L i-BL o-CTP o-TT
2017-18 0.282351 0.105693 0.422772 0 28.8013 0 7680.2
2018-19 0.534569 0.120064 0.480255 0 32.7173 200.343 0
2019-20 0.942192 0.151154 0.604615 0 41.1894 0 215075
2020-21 1 0 0 0 0 0 0
2021-22 0.906075 0 0 15.4033 0 29810.3 0
Source: Author's calculation
Note: Theta=Efficiency Score, TB=Total Berth, C= No of Cranes, L= Labor, BL= Berth
Length, CTP= Container Throughput, TT= Total traffic.
to 200.343 units. Such input-output management may have resulted in a port's scale and technical
efficiencyscore in this period.
The theta value for the evaluation period 2019-20 is 0.9421, which is closer to the efficient
frontier. On reference weights and input-output slacks, the results are (0,0,0,0.8666,0),

957 | V 1 8 . I 0 8
(0.15,0.6046,0,41.18,0,215075), respectively. Although the port operated efficiently closer to
the frontier, it might have been on the frontier if it loweredinputs such as total berths by 0.15,
cranes by 0.6046, and berth length by 41.18 units or raised total trafficby 215075. The Kandla
port's efficiency scores for 2020-21 may be seen using theta, reference weights, and input-
output slacks. Theta value of 0.906075 is slightly off the efficient frontier, implying a strong
response to full efficiency. The input-output slacks can be stated in the following way: (0,
0,15.40,0,29810.3,0). According to the input slacks, the port could have achieved the scale and
technical efficiency if it had saved 15.40 units of labour or boosted throughput by 29810.3.
Table-4: Variable Return to Scale Frontier for Kandla Port
Year CRS-TE VRS-TE SCALE RTS
2017-18 0.282351 1.000000 0.282351 IRS
2018-19 0.534569 1.000000 0.534569 IRS
2019-20 0.942192 1.000000 0.942192 IRS
2020-21 1.000000 1.000000 1.000000 -
2021-22 0.906075 1.000000 0.906075 IRS
Source: Author's calculation
The VRS DEA values differ from the CRS DEA estimated efficiency score values in table-4.
Although theVRS DEA efficiency ratings demonstrate total efficiency across the reference
period more than the CRS DEA, optimal benchmarking efficiency evaluations are based on the
CRS frontier (Coelli,2008). If CRS istaken into account, Kandla port has fared well only in
2020-21.

Source: Author's calculation

Note: M=Mundra Port, K=Kandla Port, R=Ranks, Theta=Efficiency Score, Lamda=reference
weights, i=input, o=output, TB=Total Berth, C= No of Cranes, L= Labor, BL= Berth Length,
CTP= Container Throughput, TT= Total traffic.
A combined efficiency measure gives a different picture compared to individual efficiency
tests. The efficiency test for both Mundra and Kandla ports is shown in Table -5. When the two
ports are combined,Kandla port has an efficient benchmarking frontier in 2020-21, with a score

958 | V 1 8 . I 0 8
of one. Using this as a benchmark for the rest of the year's performance reveals the various
levels of efficiency. The completed analysis shows no efficiency for Mundra port in any year,
but efficiency for Kandla port in 2020-21, whereas Mundra port had shown efficiency for two
consecutive periods in separate analyses (2017-18 & 2021-22).
Table-6: Variable Return to Scale Frontier for Mundra and Kandla Port
DMU Year Ranks CRS-TE VRS-TE SCALE RTS
2017-18 8 0.576368 1 0.576368 DRS
2018-19 7 0.6368 0.71 0.6368 DRS
Mundra port 2019-20 6 0.675415 0.74 0.675415 DRS
2020-21 5 0.792398 0.95 0.792398 DRS
2021-22 4 0.9044 1 0.9044 DRS
2017-18 10 0.2823 1 0.2823 IRS
2018-19 9 0.5345 1 0.5345 IRS
Kandla Port 2019-20 2 0.9421 1 0.9421 IRS
2020-21 1 1 1 1 CRS
2021-22 3 0.906075 1 0.906075 IRS
Source: Author's calculation
A combined examination of variable returns to scale in table-6, on the other hand, shows
Mundra Port to be in the opposite situation. For example, in 2017-18 and 2021-22, the Mundra-
port experienced a constantreturn to scale, while the rest of the year saw an increasing return
to scale. However, the VRS frontier of Mundra port demonstrates a declining return to scale
over time in a combined analysis. Except for 2020- 21, Kandla port operates on the principle
of rising returns to scale.

6. CONCLUSION
In this paper, the Mundra and Kandla ports were evaluated for efficiency. The evaluation's
findings suggestthat the efficiency of these private and public-owned ports differs. Using DEA
in the capacity of individual efficiency study, the Mundra port has an average efficiency of
0.9275 percent for the reference period. Theefficiency representation of the Mundra port is
consistent when compared to its counterpart. As a result, Mundra Port's performance is
exceptional on its own. This was proven by comparing the average efficiencyscore of Mundra
and Kandla ports over the same period. Mundra's average efficiency score is 0.92 percent,which
is higher than the 0.73 percent of its counterpart. This evidence emphasizes a close examination
of the Kandla port and learning about the many input-output combinations.
Following that, compelling inference can be taken. On the other side, when DEA is worked on
the cross-section panel, DEA sets an efficient frontier from Kandla Port in 2020-21. In this case,
the yearly efficiency and average efficiency of Mundra Port reboundeddownward, and Kandla
port remained the same. Mundra port's average efficiency dropped from 0.92 percent to 0.71
percent. Mundra Port is ten times larger than Kandla Port in container handling capability.
Mundra port's efficiency rankings in this category are excellent. Furthermore, both ports have
a lot of infrastructure asymmetries, despite their enormous potential. The unique characteristic
of Mundra port, incontrast to Kandla port, is that it is located within a special economic zone.

959 | V 1 8 . I 0 8
Mundra port's average efficiency dropped from 0.92 percent to 0.71 percent. Mundra Port is ten
times larger than Kandla Port in container handling capability. Mundra port's efficiency
rankings in this category are excellent. Furthermore, both ports have a lot of infrastructure
asymmetries, despite their enormous potential. As a privately operated port in a special
economic zone, Mundra port has excellent infrastructure and draws more ships. As a
government-owned port, Kandla port could consider leveraging its traditional strengths in
handling bulk cargo through modernization efforts by utilising the policy focus of the
Sagarmala Project of the Government of India and building capacity more quickly. Other policy
clearances and tendering-related difficulties must be addressed to improve Kandla port's
performance.
In conclusion, our findings suggest that incorporating external stakeholders improves technical
efficiency in ownership structure. The landlord seaport concept reduces public sector budget
demands and improves port performance. As a result, the environment should be more
conducive to public-private partnerships, with terminal operators being provided with an
environment that allows concession contracts to last for theproject's duration. On the other hand,
the landlord port model can only be successfully implemented if port authorities have good
corporate governance and capacity.
Based on the foregoing, the following recommendations are made: (a) promote private sector
participation through a well-developed institutionalframework, including continued adoption
of the landlord port model; (b) improve governance of port authority boards; and (c) promote
competition between and within ports, in part through transparent and competitive concession
bidding. Needless to say, good corporate governance and transparent business practices are a
sine-qua-non, not only for publicly owned port entities but also for privately owned port entities
to perform efficiently and effectively and deliver optimal outcomes.
The future study could involve a thorough examination of each port's enabling and inhibiting
aspects affecting its operational and financial efficiencies. Models could be built to
understand the quantitative impact of these factors on container port efficiency. This research
can be developed to estimate the profitability and sustainability of ports once the proper input
and output variable data are readily available.
There are certain limitations to this study. Instead of the second stage DEA, one could use the
distance function approach, Bayesian techniques, and Monte Carlo techniques. In addition, the
impact of qualitative elements such as unstable situations, government reforms, hinterland
connectivity, management, and so onmight be examined. Future research should evaluate the
Indian port's cost efficiency, as this study estimatestechnical efficiency.

960 | V 1 8 . I 0 8
References
1) Adler, N., Hirte, G., Kumar, S., &Niemeier, Hans‑Martin. (2021). The impact of specialization, ownership,
competition, and regulation on efficiency: A case study of Indian seaports, Maritime Economics & Logistics,
Retrieved from https://doi.org/10.1057/s41278-021-00200-y
2) Chudasama. K. M.,& Pandya, K. (2008). Measuring Efficiency of Indian Ports: An Application of Data
Envelopment Analysis, The IUP Journal of Infrastructure, June, 22-34.
3) Coelli, T.J. (2008) A Guide to DEAP Version 2.1: A Data Envelopment Analysis (Computer) Program.
Working Paper of the University of New England/Center for Efficiency and Productivity Analysis.
4) Dasgupta, M. K., & Sinha, D. (2016). Impact of privatization of ports on relative efficiency of major portsof
India, Foreign Trade Review, 51(3), 225–247.
5) Haralambides, H., Gujar, G., & M. Jain, M. (2011). Dry Port Efficiency and Container Security, paper
presented at the International Association of Maritime Economists (IAME)At: Santiago, Chile, October,
Retrieved from https://www.researchgate.net/publication/299430025Dry_Port_Efficiency_and Container
Security
6) Iyer, K.C, &Nanyam, V.P.S. (2021). Technical efficiency analysis of container terminals in India, The Asian
Journal of Shipping and Logistics, 37 (1), 61-72.
7) Kamble, S.S., Raoot,A. D., &Khanapuri, V. B. (2010). Improving port efficiency: A comparative study of
selected ports in India. International Journal of Shipping and Transport Logistics, 2 (4), 444-470
8) Monteiro, J. G. R. (2018). Measuring productivity and efficiency of seaports in Indiausing DEA technique.
The Central European Review of Economics and Management,2(3). 155-174.
9) Munisamy, S.,& Singh, G. (2011). Benchmarking the efficiency of asiancontainerports. African Journal of
Business Management, 5(4), 1397–1407.
10) Mustafa, F.S., Khan, R.U., & Mustafa, T. (2021). Technical efficiency comparison of container ports in
Asian and Middle East region using DEA, Asian Journal of Shipping and Logistics, 37(1), 12-19.
11) Sekar, R. T., & Deo, M. (2012). The Size Effect of Indian Major Ports on its Efficiency Using Dea-Additive
Models, International Journal of Advances in Management and Economics, 1(5), 12-18
12) Sekar, R.T., P.S. Ashraf., & Deo. P M. (2014). Measurement of efficiency of major ports in India – A data
envelopment analysis approach. International Journal of Environmental Sciences, 4 (5), 926–936.
13) Wang, T., Song, D., & Cullinane, K. (2003) Container Port Production Efficiency: A Comparative Studyof
DEA and FDH Approaches, Journal of the Eastern Asia Society for Transportation Studies, 5, 698-713.
14) UNCTAD (1976) Port Performance Indicators, TD/B/ C.4/131/Supp.1/Rev.1. United Nations Conference
on Trade and Development, New York.

961 | V 1 8 . I 0 8