Estimating technical efficiencies of Indian IT companies for setting improvement targets for inefficient companies: An empirical analysis with workers’ effort as key input

Abstract

BACKGROUND:

The Information Technology (IT) industry of India has proved its capabilities in delivering both on- and off-shore services to clients globally over the years. However, the technological advances and innovations taking place at the global level not only present a whole new range of growth prospects, but also challenges for this highly competitive industry. Moreover, the IT sector of India also witnessed the economic recession in 2008, which had an adverse impact on the prospects of this industry. In this scenario, it is imperative for Indian IT companies not only to maintain their focus on increasing their technical efficiencies, but also to deal with the increased competition emanating from the Asia Pacific region.

OBJECTIVE:

This study aimed to estimate the relative efficiency of the top 18 selected Indian IT software service companies in order to determine benchmarks, output slacks and target settings.

METHODS:

Data envelopment analysis has been used for achieving the stated objective.

RESULTS:

The paper found mixed trends in efficiency. The top five IT companies exhibited higher efficiency as compared to the rest of the selected IT companies. Tata Consultancy Services, HCL Technologies Ltd. and Tech Mahindra Ltd. are more efficient while Infosys Ltd. and Mphasis have lower efficiency.

CONCLUSION:

The inefficient companies have to increase their workers’ productivity to become more efficient, and have to catch-up and follow the best practices of the benchmark company HCL.

Keywords

Efficiency data envelopment analysis (DEA)Information Technology (IT)Information and Communication Technology (ICT)decision making unit (DMU)

1 Introduction

The Information Technology (IT) or Information and Communication Technology (ICT) industry has a huge impact on all the production sectors such as transportation, logistics, retail, banking, manufacturing, healthcare, communication and media, and education [1 –6]. The global IT services spending stood at 747 billion (USD) in 2007. It is estimated to grow to 987 billion (USD) in 2018 showing a growth of 32.13% over 2007. It is further estimated to increase to 1,034 billion (USD) by 2019 [7, 8]. The major upcoming areas of the IT industry in near future are digitalization, artificial intelligence, cloud computing, human-computer interactions and block chain [7 –11].

The process of globalization, privatization and liberalization started in India in 1991-92. Since then, the IT industry has transformed India’s economic image from an agriculturist state to IT offshoring destination for the world [10, 12]. The role of the IT sector in the Indian economy and its impact on different sectors of the economy can be comprehended from the fact that the IT sector in India has raised its contribution to Gross Domestic Product (GDP) from 1.2% in 1998 to 7.9% in 2018. The IT industry of India has grown considerably from 8.2 billion (USD) in 2000 to 154 billion (USD) in 2017 [10 , 13]. Consequently, the IT sector has become one of the leading sectors of the Indian economy in 2018. However, the Indian IT industry suffered a set-back on the technological front due to the economic downturn of 2008 [12, 14]. Also, the present-day challenges such as stringent visa conditions, technological advancements in the industry and increased global competition, especially from China and the Philippines, had an adverse impact on the efficiency of Indian IT companies. The average annual growth of the Indian IT industry, which was 30% from 2002 to 2008, dropped to 8.10% from 2008 to 2009. Since then, the IT industry of India has not been able to regain its growth momentum [10, 11].

Whilst the demand for automation and skilled talent exists globally, the IT industry of India only accounted for a dismal 17% of the total IT services expenditure worldwide in 2017 [10, 11]. In this scenario, providing low cost semi-skilled talent by Indian IT companies cannot be considered an effective solution. The IT industry is a human capital intensive sector, which is dependent on high skilled workers. The workers’ skills, health, efforts and individual performance have a huge impact on the performance of IT companies [3 , 15–17]. The workers’ effort provides an indication of the health of the company and is thus a key input variable for the IT companies. In the literature, performance analysis of the IT sector has been analyzed by various frontier and non-frontier approaches. Data envelopment analysis (DEA) is a frontier approach that has been widely used to analyze the efficiency of IT/ICT companies [15 , 18–27].

DEA, introduced in 1978, is a non-parametric technique that has evolved as an alternative method to regression analysis for efficiency measurement. It is defined as a technique to measure the performance of similar organizations known as decision making units (DMUs) in terms of multiple inputs and outputs. DMUs may include IT companies, manufacturing units, banks, educational institutions, etc. [16 , 29]. The extensive literature on IT efficiency is listed in Section 2. Furthermore, the literature review suggested that there are only a few studies have tried to measure the efficiency of the IT companies after the financial meltdown of 2008 [20, 30]. However, most of the studies on efficiency measurement of IT companies after the 2008 recession were conducted in foreign countries. Therefore, there is a need to investigate the performance of Indian IT companies, to set the output slacks and target values and to benchmark the IT companies in India after the financial meltdown of 2008. This will enable the companies to manage inefficiency and devise policy measures for improving their overall efficiency. These specific aspects of the Indian IT industry need to be addressed. However, there is very scant literature available in this area. The present paper therefore tries to fill this gap by studying the efficiency of the IT sector using the DEA technique and has utilized workers’ effort as key input.

The rest of the paper is structured as follows: Section 2 presents a brief review of the available literature on DEA and its application on the IT industry. Section 3 delineates the objectives of the study. This is followed by Section 4, which describes the DEA methodology employed in the paper. Section 5 provides information about data collection and selection of input-output variables in the DEA analysis for the study and Section 6 explains the empirical results and discussion. Section 7 describes the findings of the study and Section 8 presents the conclusion and limitations. Finally, Section 9 provides the implications of the study.

2 Literature review

The measurement and improvement of efficiency has remained a topic of great interest to researchers as has lots of implications for organizations in increasing the output and reducing the overall cost of production [31 –35]. There are numerous studies that have investigated and analyzed the efficiency and effectiveness of IT companies. Researchers have also studied the impact of the IT sector on individual as well as organizational productivity [3–5 , 33–36]. The IT automation processes have helped to save time and reduce the overall cost of production in organizations. The increase in productivity has resulted in economic growth and human development [6 , 38]. However, there are also numerous studies which have highlighted the adverse impact of IT on the health of the people both at home [39, 40] and at work [41], thereby reducing the overall organizational efficiency.

Trierweiller et al. (2012) measured the organizational effectiveness of ICT companies from a managerial point of view. In this study, a survey was done taking a sample of 80 managers from ICT companies and item response theory (IRT) was applied. The study found that IRT provided the degree of effectiveness [32]. Another study by Coelho and Lourenço (2018) measured the effectiveness and efficiency of three different pointing devices of computers in the IT industry. They found operational efficiency of alternative pointing devices of computers and ranked them [34]. Furthermore, De Brito et al. (2019) examined the efficiency of students in the IT environment. The study found that the productivity of students can be changed due to changes in air temperature in the air-conditioned IT environment [42]. Furthermore, Sadeghniiat-Haghighi et al. (2016) investigated sleep efficiency of 295 shift workers. In Indian IT companies also, people work in shifts and this has implications on productivity. They documented that working a night shift has an impact on sleep efficiency and consequently also impacts overall efficiency [43]. Thus, it is inferred that the efficiency measurement in the IT industry is a topic of great interest for researchers and has very high implications for policy makers.

In general, the efficiency and productivity is studied using a regression framework. However, this study used DEA, a non-parametric method which has evolved in recent times for measuring efficiency [19 –30]. There is an enormous amount of literature available on DEA and its applications for efficiency estimation and the IT industry is not an exception [19–27 , 44–46]. The literature review on DEA is listed in Table 1.

Table 1
Literature review on DEA and variables used

Variable Type Variable Name References from Literature Review

Input Shareholder’s Equity [15 , 30]

No. of Workers/Labor [15 , 47]

Expenses [12 , 48]

Assets [21 , 44]

Output Income/Sales/Revenue [15 , 47]

Net Profit [18 , 48]

Variable Type	Variable Name	References from Literature Review
Input	Shareholder’s Equity	[15 , 30]
	No. of Workers/Labor	[15 , 47]
	Expenses	[12 , 48]
	Assets	[21 , 44]
Output	Income/Sales/Revenue	[15 , 47]
	Net Profit	[18 , 48]

Source: Author’s compilation.

As can be seen from Table 1, the majority of the studies on the IT industry are outdated. These studies have investigated the productivity/efficiency trends before the sub-prime crisis of 2008. Shao and Shu (2004) measured productivity growth of ICT in 14 Organization of Economic Cooperation and Development countries from 1978 to 1990. They found that most of the productivity growth was due to technological progress. They also found that the change in scale economies adversely influenced productivity [47]. Another study by Chen and Ali (2004) applied the DEA Malmquist approach on a set of Fortune Global 500 Computer and Office Equipment companies from 1991 to 1997. This study revealed a mixed trend with an average of overall technical efficiency (AOTE) [21]. Furthermore, Liu and Wang (2009) estimated the efficiency of 15 IT packaging and testing companies in Taiwan for the period of 2000 and 2003. They found that none of the selected companies performed efficiently, neither at the stage of production acquisition nor at the stage of profit earnings. They also suggested that the IT manufacturing companies need to address issues relating to product improvements and technological innovations in order to improve the productivity [22]. Dash Wu and Ho (2007) studied 36 integrated circuit IT companies in Taiwan from 2002 to 2004. They established that the relationship between asset size and efficiency is negative [23]. Furthermore, Chou and Shao (2014) investigated the total factor productivity growth for the IT sector of 25 countries from 1995 to 2007 using DEA. They concluded that the technological innovations are a key driver of productivity growth. They also found that efficiency change and scale change have a negative impact on productivity [24].

Mathur (2007) studied the efficiency of 92 Indian software companies for the year 2005-2006 and concluded that only 16 companies were efficient. He found that while the IT sector of Taiwan was the most efficient, the IT sector of India was the least efficient in a sample of 12 countries [25]. Furthermore, Bhattacharjee (2012) measured the technical efficiency of all IT and IT enabled service companies in the eastern region of India between 1993 and 2007 using DEA analysis. He found that the efficiency of big companies registered an increase in their productivity over the sample period. He emphasized the segment specific policies in place of uniform policies in order to improve productivity [26]. Sahoo and Nauriyal (2014) used the input-oriented DEA model to investigate trends of productivity growth of 72 Indian software companies from 1999 to 2008. They found that Indian software companies were not able to utilize their inputs to the extent of almost 35%. They also found that Indian-owned software companies demonstrated high efficiency in comparison to foreign-owned IT companies [27].

The literature review reveals a mixed trend in the changes in efficiency of IT companies at the global level as well as the domestic level in the pre-recession period. Only a few studies have tried to measure the efficiency of IT companies after the financial meltdown of 2008 [20, 30]. However, most of the studies on the efficiency measurement of IT companies after the 2008 recession were conducted in foreign countries. Thus, there is a need to analyze the efficiency trends of the Indian IT sector for the post-recession period from 2008 onwards; and also to analyze the effect of various global challenges on the efficiency of the IT sector of India. This study tries to fulfill this gap.

3 Objectives of the study

The objectives of the paper are as follows: i) To calculate and analyze different efficiency measures, namely the overall technical, pure technical and scale efficiencies for the selected 18 IT software service companies in India from 2010 to 2017 (see Appendix A for a list of IT companies selected for this study). ii) To identify and benchmark the efficient companies for the period 2016-2017. iii) To investigate the factors influencing output slacks and set target values for inefficient companies. iv) To apply sensitivity analysis to check the robustness of results.

4 DEA methodology

The DEA technique is a non-parametric technique widely used to study relative efficiency of different sectors [19 , 44]. In this technique, the efficiency score of homogenous entities also called DMUs, with multiple inputs and multiple outputs are measured. The efficiency of any DMU is measured as the ratio of output to input. However, in case of multiple inputs and outputs, it is the ratio of the weighted sum of outputs to the weighted sum of inputs as shown below [28 , 44–46]. $Efficiency = \frac{weighted sum of outputs}{weighted sum of inputs}$ (1)

To analyze the relative efficiency, there are two popular DEA models, named the Charnes, Cooper and Rhodes (CCR) model and Bankers, Charnes and Cooper (BCC) model. The mathematical formulations of CCR and BCC models are discussed in detail in Appendix B. The efficiency scores are calculated using these two DEA models [28 , 46]. The DEA models provide three types of efficiency measures, namely overall technical efficiency (OTE), pure technical efficiency (PTE) and scale efficiency (SE). The OTE is calculated using the CCR model. The inefficiency arising on account of the input variables, the output variables, and the size of the company are reflected through OTE scores (see Appendix C). The PTE is computed using the BCC model. It represents the managerial efficiency of the DMUs. In order to capture the effect of the size of DMUs, the SE is measured. The SE is basically a ratio of OTE and PTE. With the help of these models, returns to scale can also be estimated. The returns to scale refers to the response of output when all the input variables change in the long-run. There are three types of returns to scale: constant returns to scale (CRS), increasing returns to scale (IRS) and decreasing returns to scale (DRS) [28 , 44–46]. All DMUs that have an OTE, PTE and SE score of one are considered overall efficient. These DMUs are considered as a benchmark for the inefficient units. In this study, an output oriented DEA model is used for the estimation of efficiency, keeping in mind that sales maximization or market share maximization is the guiding principle of IT companies [49].

5 Data and variables selection

We collected data on the top 18 IT software service companies in India, to meet the various objectives of this study. The basis of selecting these companies was their market capitalization. The companies having market capitalization of more than ₹10 billion comprise our sample. The source of the data is the Prowess Database of Centre for Monitoring Indian Economy (CMIE). We collected the data from 2010 to 2017. The IT companies selected for this study are listed in Appendix A.

Table 1 highlights the various input and output variables used in previous literature on DEA. As depicted in Table 1, based on the literature review, we have selected the following input-output variables. This research paper focuses on measuring and comparing the relative efficiency of the Indian IT companies with respect to their financial performance. Therefore, input and output variables relating to financial performance were particularly kept in mind while selecting the input and output variables in this paper. The input variables selected are total assets (I₁), workers’ cost (I₂), total expenses (I₃) and tax paid (I₄). The worker’s cost is taken as proxy of the worker’s effort.

The IT industry is a human capital intensive sector, which is dependent upon high skilled workers. The workers’ skills, health, effort and individual performance have a huge impact on the performance of IT companies [3 , 17]. The IT/ICT companies or the IT facilities within the organizations have implemented ergonomic principles and solutions for well-being of their workers [50, 51]. The ergonomic performance measurement framework [52] models to evaluate worker’s total cost of ownership [53] and importance of individual’s work performance are not only applicable to other industries, but also to the IT industry. The workers’ effort provides an indication of the health of the company and thus is a key input variable for the IT companies. The output variables used in this study are net sales turnover (O₁) and net profit after tax (O₂). The profit after tax is a good measure of profitability of companies. This paper employed the software DEA Program (DEAP) for the estimation purpose [46].

6 Empirical results and discussions

Efficiency measures such as OTE, PTE and SE are computed, corresponding to the respective DEA model, using DEAP software (see Appendix B and C for DEA models and definitions of different types of efficiency, respectively). The efficiency results are further discussed and interpreted in the subsequent sections.

6.1 Efficiency analysis and results for the period 2010-2017

The descriptive statistics of efficiency measures for all the selected Indian IT software service companies for the period 2010 to 2017 are depicted in Table 2. Table 2 shows that the average overall technical efficiency (AOTE) has mixed trends and has reduced from 95.97% in 2010 to 94.86% in 2014. The average pure technical efficiency (APTE) was lowest at 96.37% in 2014 and the lowest average scale efficiency (ASE) of 96.9% was in 2012. In 2014, 3.63% inefficiency is estimated due to inappropriate management practices and processes which is represented as average pure technical inefficiency (APTIE). The remaining part of average overall technical inefficiency (AOTIE) can be attributed to inappropriate size.

Table 2
Descriptive statistics of efficiency measures for all IT companies for the period 2010-2017

Descriptive Statistics All IT companies - Overall Technical Efficiency (OTE)

2010 2011 2012 2013 2014 2015 2016 2017

Average Efficiency (AOTE) 0.960 0.944 0.957 0.952 0.949 0.955 0.953 0.970

Minimum 0.880 0.766 0.790 0.782 0.708 0.838 0.872 0.861

Maximum 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000

Std. Deviation 0.044 0.066 0.060 0.057 0.072 0.054 0.051 0.042

Avg. Inefficiency % = 1-AOTE 4.03 5.61 4.30 4.78 5.14 4.49 4.74 3.05

Descriptive Statistics All IT companies - Pure Technical Efficiency (PTE)

Average Efficiency (APTE) 0.985 0.969 0.987 0.981 0.964 0.967 0.975 0.984

Minimum 0.920 0.787 0.933 0.902 0.790 0.856 0.892 0.868

Maximum 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000

Std. Deviation 0.024 0.056 0.023 0.031 0.056 0.046 0.037 0.034

Avg. Inefficiency % = 1-APTE 1.52 3.09 1.29 1.94 3.63 3.31 2.53 1.65

Descriptive Statistics All IT companies - Scale Efficiency (SE)

Average Efficiency (ASE) 0.974 0.974 0.969 0.971 0.983 0.988 0.977 0.986

Minimum 0.880 0.838 0.790 0.782 0.896 0.926 0.885 0.916

Maximum 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000

Std. Deviation 0.033 0.042 0.054 0.051 0.027 0.019 0.031 0.022

Avg. Inefficiency % = 1-ASE 2.56 2.58 3.06 2.87 1.67 1.24 2.29 1.42

Descriptive Statistics	All IT companies - Overall Technical Efficiency (OTE)
Average Efficiency (AOTE)	0.960	0.944	0.957	0.952	0.949	0.955	0.953	0.970
Minimum	0.880	0.766	0.790	0.782	0.708	0.838	0.872	0.861
Maximum	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000
Std. Deviation	0.044	0.066	0.060	0.057	0.072	0.054	0.051	0.042
Avg. Inefficiency % = 1-AOTE	4.03	5.61	4.30	4.78	5.14	4.49	4.74	3.05
Descriptive Statistics	All IT companies - Pure Technical Efficiency (PTE)
Average Efficiency (APTE)	0.985	0.969	0.987	0.981	0.964	0.967	0.975	0.984
Minimum	0.920	0.787	0.933	0.902	0.790	0.856	0.892	0.868
Maximum	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000
Std. Deviation	0.024	0.056	0.023	0.031	0.056	0.046	0.037	0.034
Avg. Inefficiency % = 1-APTE	1.52	3.09	1.29	1.94	3.63	3.31	2.53	1.65
Descriptive Statistics	All IT companies - Scale Efficiency (SE)
Average Efficiency (ASE)	0.974	0.974	0.969	0.971	0.983	0.988	0.977	0.986
Minimum	0.880	0.838	0.790	0.782	0.896	0.926	0.885	0.916
Maximum	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000
Std. Deviation	0.033	0.042	0.054	0.051	0.027	0.019	0.031	0.022
Avg. Inefficiency % = 1-ASE	2.56	2.58	3.06	2.87	1.67	1.24	2.29	1.42

The reasons for the declining trend in AOTE probably is on account of unfavorable business environment prevailing during 2010 and 2014; IT companies were offering generic services with low innovations and failed to provide specialized services; decline in off-shore business due to unfavorable changes in macro-economic conditions and increase in exchange rate risks; and matured vendor management processes by their clients.

Table 2 also reveals that the OTE of IT companies in India ranges between 0.88 and 1 for the year 2010, whereas this range reduces to 0.86 and 1 for the year 2017. However, the AOTE was 0.96 in the year 2010, which increased to 0.97 in the year 2017. This increase in AOTE could be on account of rigorous focus brought back towards talent management, skill enrichment and automation processes by selected IT companies. The IT companies in India also adopted various strategies to tap and capitalize talents from Mexico and Brazil to meet time zone requirements. They also implemented agile methodologies in IT projects, enriched workers’ skill sets, and improved workers’ utilization rate. Figure 1 depicts the most and least efficient DMUs.

Fig. 1

Efficiency scores for all IT companies for 2010-2017. 2(a) OTE, 2(b) PTE, 2(c) SE.

Figure 2 depicts the average inefficiency score for all the selected IT companies in India from 2010 to 2017. From Fig. 2, it is clear that the inefficiency of Indian IT companies fluctuated in significant manner over the sample period. It is also evident that the average inefficiency which was 4.03% declined to 3.05% in 2017. This clearly established that the average efficiency improved in 2017 compared to 2010. Likewise, the APTIE was 1.52% in 2010, which rose to 1.65% in 2017 indicating a decline in pure technical efficiency. Similarly, the Average Scale Inefficiency (ASIE) was 2.56% in 2010, which declined to 1.42% implying improvement in scale efficiency.

Fig. 2

Average inefficiency trend (%) for the selected IT companies.

6.1.1 Cluster-wise descriptive statistics of efficiency results

The companies selected for this study are further divided into two clusters for analytical purpose. In the first cluster, top five IT software service companies with a market capitalization greater than ₹100 billion and net profit greater than ₹10 billion were included. All other IT companies were kept in the second cluster. It was found that from 2010 to 2016 period, the OTE of the first cluster was higher than that of the second cluster. This could be attributed to better efficiencies of the top five IT companies on account of large scale, well established processes, and a large and stable client base. Nonetheless, in 2017, the overall technical inefficiency (OTIE) of the first cluster was higher than that of the second cluster. Perhaps during this time, the companies in the second cluster taking cue from the top five companies tried to improve their workers’ productivity, workers’ skills, provided good work environment, implemented ergonomic principles at work and thereby building an overall positive environment. This led the rest of the IT companies to achieve better efficiencies.

In this paper, an attempt is also made to compare the efficiency of the top five IT software service companies with the rest of the selected companies using box-plots as shown in Fig. 3 [45]. In this figure, the substantial tall boxes for the top five companies indicate that there is greater variability in the efficiency scores of the top five companies. The figure shows slightly smaller boxes for the rest of the companies, which indicates a relatively high stability in their efficiency scores.

Fig. 3

Overall technical efficiency comparison for the selected IT companies. 5(a) Top five IT companies, 5(b) The rest of the selected IT companies, 5(c) All selected IT companies.

We found that amongst the top five companies, only one IT company was efficient in 2010. However, 3 out of 5 IT companies were found to be operating efficiently in 2017. Moreover, the study also found that 44.44% of the companies were working efficiently in 2010, whereas 53% of the IT companies were found to be operating efficiently in 2017.

6.2 Efficiency results of selected IT companies with main focus on the period 2016-2017

The main highlight in this section is the discussion on the efficiency scores, returns to scale, reference sets and peer count for the selected IT software service companies. The results have been depicted in Table 3. Since GEOMR got acquired in 2016 by HCL, the total number of the selected IT companies was reduced from 18 to 17.

Table 3
Efficiency scores, returns to scale, peers for all IT companies for 2016–2017

IT Co. OTE PTE SE RTS Peers Peer Count

TCS 1.000 1.000 1.000 CRS 1 4

INFY 0.915 0.953 0.961 DRS 1, 4 0

WIPRO 0.916 1.000 0.916 DRS 4, 5, 7, 14 0

HCL 1.000 1.000 1.000 CRS 4 8

TECHM 1.000 1.000 1.000 CRS 5 5

ORCLF 1.000 1.000 1.000 CRS 6 0

MNDTR 1.000 1.000 1.000 CRS 7 5

MPHSI 0.861 0.868 0.993 IRS 4, 5, 14 0

HEXWR 0.993 1.000 0.993 IRS 1, 4, 7 0

CYENT 0.963 0.980 0.983 IRS 4, 5, 14 0

PRSTN 0.920 0.934 0.985 IRS 1, 4, 5, 7 0

TLXI 1.000 1.000 1.000 CRS 12 0

ZNSR 0.947 0.985 0.961 IRS 1, 4, 5, 7 0

NIIT 1.000 1.000 1.000 CRS 14 4

POLRS 1.000 1.000 1.000 CRS 15 0

HNDG 0.967 1.000 0.967 IRS 4, 7, 14 0

HNDVR 1.000 1.000 1.000 CRS 17 0

0.970 0.984 0.986 < —Average Efficiency

3.00 1.60 1.40 < —Average Inefficiency (% age)

IT Co.	OTE	PTE	SE	RTS	Peers	Peer Count
TCS	1.000	1.000	1.000	CRS	1	4
INFY	0.915	0.953	0.961	DRS	1, 4	0
WIPRO	0.916	1.000	0.916	DRS	4, 5, 7, 14	0
HCL	1.000	1.000	1.000	CRS	4	8
TECHM	1.000	1.000	1.000	CRS	5	5
ORCLF	1.000	1.000	1.000	CRS	6	0
MNDTR	1.000	1.000	1.000	CRS	7	5
MPHSI	0.861	0.868	0.993	IRS	4, 5, 14	0
HEXWR	0.993	1.000	0.993	IRS	1, 4, 7	0
CYENT	0.963	0.980	0.983	IRS	4, 5, 14	0
PRSTN	0.920	0.934	0.985	IRS	1, 4, 5, 7	0
TLXI	1.000	1.000	1.000	CRS	12	0
ZNSR	0.947	0.985	0.961	IRS	1, 4, 5, 7	0
NIIT	1.000	1.000	1.000	CRS	14	4
POLRS	1.000	1.000	1.000	CRS	15	0
HNDG	0.967	1.000	0.967	IRS	4, 7, 14	0
HNDVR	1.000	1.000	1.000	CRS	17	0
	0.970	0.984	0.986		< —Average Efficiency
	3.00	1.60	1.40		< —Average Inefficiency (% age)

Note: OTE: overall technical efficiency, PTE: pure technical efficiency, SE: scale efficiency, RTS: returns to scale, CRS: constant returns to scale, IRS: increasing returns to scale, DRS: decreasing returns to scale.

Table 3 shows that the mean OTE score of the selected IT companies was found to be 0.97 for the year 2016-2017. Furthermore, the results reveal that nine companies, namely TCS, HCL, TECHM, ORCLF, MNDTR, TLXI, NIIT, POLRS and HNDV had an OTE score of one. These efficient IT companies collectively formed the efficient frontier. The inefficient companies compared with the above mentioned companies were INFY, WIPRO, MPHSI, HXWR, CYENT, PRSTN, ZNSR and HNDG. This paper also estimated the PTE and the SE of the selected IT companies using the BCC model. The results depicted that twelve companies formed the variable returns to scale frontier. Three companies, namely WPR, HXWR and HNDG were PTE efficient but scale inefficient. Five companies were neither overall technical efficient nor pure technical efficient. This indicates that these companies required improvement in managerial performance as well as improvement in scale utilization. Our results show that there were nine companies that had SE of one, indicating that these companies were able to utilize their scale effectively. These companies were TCS, HCL, TECHM, ORCLF, MNDTR, TLXI, NIIT, POLR and HNDV. However, there were eight companies that deviated from their optimal size and thus were scale inefficient.

Our analysis established that MPHSI and INFY are amongst the lowest efficient companies with an OTE score of 0.861 and 0.91 respectively. Among the top five companies, TCS and HCL are peers of INFY. This implies that INFY was approximately 10% less efficient than its peer companies. INFY was found to be the least efficient on every parameter. Moreover, the production system of INFY exhibited decreasing returns to scale over the sample period.

The study furthermore revealed that nine companies, namely TCS, HCL, TECHM, ORCLF, MNDTR, TLXI, NIIT, POLRS and HNDV showed constant returns to scale. In addition, the six scale inefficient companies, namely MPHSI, HXWR, CYENT, PRSTN, ZNSR and HNDG, exhibited increasing returns to scale. This has implications on productivity enhancement. These companies can increase their efficiency by expanding their workers by the internal process of routine recruitments or through mergers and acquisitions.

6.2.1 Benchmarking –overall technical efficient companies

The results of this study pointed out that there were nine IT software service companies which had an efficiency score of one. These nine companies become the benchmark for the remaining eight companies. The efficient frontier formed by these companies was used as the basis for performance improvement of the inefficient companies. The peer count of DMUs, as shown in Table 3, was used to determine the robustness of the DMUs. The higher the score of peer count, the more robust is the company. For 2016-2017, HCL had the largest peer count, and it was identified as the most robust company. The efficient DMUs with smaller peer count, such as ORCLF, TLXI, POLRS and HNDV, were likely to be less robust. On the basis of the peer count, the companies were classified as high robust, medium robust and low robust. HCL, TECHM and TCS were respectively high robust, medium robust and low robust companies.

6.2.2 Sensitivity analysis (post-DEA analysis)

This paper also conducted sensitivity analysis to check the robustness of the DEA results. In this analysis, the robustness of the results is measured by removing one of the efficient companies at a time, and then analyzing the results [44, 46]. Thus, five models were developed for the sensitivity analysis. The results are shown in Table 4.

Table 4
Results of sensitivity analysis based on all five models

IT co. removed from the analysis Model AOTE Effect of removing of an IT co. on AOTE Number of efficient IT cos.

Initial 0.97 – 9

HCL Model -I 0.98 Increase 10

TECHM Model -II 0.98 Increase 8

MNDTR Model -III 0.97 No change 8

TCS Model –IV 0.97 No change 9

NIIT Model -V 0.97 No change 9

IT co. removed from the analysis	Model	AOTE	Effect of removing of an IT co. on AOTE	Number of efficient IT cos.
Initial	0.97	–	9
HCL	Model -I	0.98	Increase	10
TECHM	Model -II	0.98	Increase	8
MNDTR	Model -III	0.97	No change	8
TCS	Model –IV	0.97	No change	9
NIIT	Model -V	0.97	No change	9

Source: Author’s calculation. Note: AOTE – Average Overall technical efficiency.

An efficient company is considered as an outlier if its removal from the selected DMUs significantly changes the average OTE of the companies. The results show that the average OTE was not altered significantly in any of the five models, thereby establishing that the initial results were stable.

6.2.3 Factors impacting the efficiency of IT companies: Output slacks and target setting for inefficient companies

This study employed an output oriented DEA model, so it analyzed only output slacks. The output slacks provided information about the actual output and potential output. We analyzed the data to determine how to achieve the desired efficiency by varying only the outputs, and holding the inputs constant. Based on CCR model, the optimal output slacks for inefficient DMUs were computed for the year 2016-2017 (see Appendix B).

It was found that ten companies had no output slack. It is important to note that nine companies out of these ten were overall technically efficient. These companies made maximum utilization of the skills of their workers. It was also found that there were no slacks in the net sales turnover of any company. However, the average slack in the net profit after tax of all the seventeen companies was found to be around ₹4046.22 million. Among the top five companies, INFY had the maximum slack in profit after tax to the tune of ₹38747 million. This implies that INFY has to reduce its inputs by 8.5 percent. Not only this, it has to augment its profit after tax also by ₹38747 million to become efficient. In case of medium-sized companies, MPHSI had the maximum slack in profit after tax to the extent of ₹2491 million. CYENT was found to be the only company that was overall technically inefficient but it had zero slack. This indicates that there are other factors responsible for inefficiency; optimal scale efficiency may be one of the reasons.

The DEA analysis furthermore helps in setting up target for the inefficient DMUs in order to become efficient. The input-output target values provide information on how much input can be reduced or how much output can be increased for a particular DMU to become efficient [44, 46]. The actual and targeted values for input and output variables for the inefficient companies are estimated using Equation 5 (See Appendix B) and reported in Table 5.

Table 5
Peers and target values (in Rupee million) of input and output variables under CCR output-oriented model

Inefficient IT co. Target values of input variables Target values of output variables

Peers I₁ I₂ I₃ I₄ O₁ O₂

INFY IT1, IT4 805940 297845 490750 45686 648383 189695

(0) (3.75) (0) (10.77) (9.24) (37.28)

WIPRO IT4, IT5, IT7, IT14 636486 218544 403692 25254 498195 113245

(0) (0) (0) (0) (9.16) (38.75)

MPHSI IT4, IT5, IT14 52601 13791 25786 2013 34269 9746

(4.2) (0) (0) (0) (16.08) (55.94)

HXWR IT1, IT4, IT7 16819 6654 10799 943 14030 3885

(0) (8.32) (0) (0) (0.72) (9.1)

CYENT IT4, IT5, IT14 18119 6781 11113 501 13429 2459

(32.75) (0) (3.74) (0) (3.81) (3.81)

PRSTN IT1, IT4, IT5, IT7 21020 8732 15344 1097 18845 4376

(0) (0) (0) (0) (8.74) (48.85)

ZNSR IT1, IT4, IT5, IT7 13463 7116 11253 728 13546 2804

(0) (0) (0) (0) (5.62) (55.62)

HNDG IT4, IT7, IT14 14100 10051 14688 545 16515 2065

(0) (3.33) (0) (0) (3.38) (103.49)

Inefficient IT co.	Target values of input variables	Target values of output variables
INFY	IT1, IT4	805940	297845	490750	45686	648383	189695
(0)	(3.75)	(0)	(10.77)	(9.24)	(37.28)
WIPRO	IT4, IT5, IT7, IT14	636486	218544	403692	25254	498195	113245
(0)	(0)	(0)	(0)	(9.16)	(38.75)
MPHSI	IT4, IT5, IT14	52601	13791	25786	2013	34269	9746
(4.2)	(0)	(0)	(0)	(16.08)	(55.94)
HXWR	IT1, IT4, IT7	16819	6654	10799	943	14030	3885
(0)	(8.32)	(0)	(0)	(0.72)	(9.1)
CYENT	IT4, IT5, IT14	18119	6781	11113	501	13429	2459
(32.75)	(0)	(3.74)	(0)	(3.81)	(3.81)
PRSTN	IT1, IT4, IT5, IT7	21020	8732	15344	1097	18845	4376
(0)	(0)	(0)	(0)	(8.74)	(48.85)
ZNSR	IT1, IT4, IT5, IT7	13463	7116	11253	728	13546	2804
(0)	(0)	(0)	(0)	(5.62)	(55.62)
HNDG	IT4, IT7, IT14	14100	10051	14688	545	16515	2065
		(0)	(3.33)	(0)	(0)	(3.38)	(103.49)

Source: Author’s compilation. Note: The figure in parenthesis are the percentage augmentations in the corresponding outputs and percentage reduction in the corresponding inputs to make the IT company efficient.

As shown in Table 5, MPHSI was found to be the least efficient company in the sample. It can improve its efficiency by reducing its total assets by 4.2%, by increasing its sales turnover by 16% and net profit by 56%. In general, all the inefficient companies can increase their sales turnover by an average of 9%, and their net profit by an average of 38% to meet the efficiency of the benchmark companies. This clearly implies that companies must focus on increasing their sales output and net profit to become more efficient.

7 Results

The productivity of organizations is central to economic growth and workers’ development [6 , 38]. The IT sector helps to improve the efficiency of individual as well as organizational productivity [5, 36]. The IT sector, however, is itself passing through difficult times following the recession of 2008, and due to increase in competition at domestic and global levels. This prompted us to examine the efficiency trends of Indian IT sector. Following are the key findings of the study for the benefit of policy makers for strategizing for the way forward.

It was found that over the sample period from 2010 to 2017, MNDTR, TLXI and HNDV were overall technical efficient across all the years, with an efficiency score of one. However, ORCLF, CYENT and GEOMR were amongst the less efficient companies for this duration. During this period, the OTE has increased for TCS, HCL, TECHM, CYENT, NIIT, POLRS, HNDG and GEOMR. However, the OTE of INFY, WIPRO, MPHSI, HXWR, PRSTN and ZNSR has decreased over the same duration. The AOTE of the selected IT companies has increased from 95.97% in 2010 to 96.95% in 2017. This implies that in order to operate at higher efficiencies, these companies should continue to implement the strategies adopted in 2017. Furthermore, for the period 2010 to 2017, the top five IT companies were overall technically efficient than the rest of all selected IT companies. Thus, the rest of the selected IT companies should follow the strategies and best practices followed by the top five IT companies to improve their efficiency. However, it was observed that the top five selected companies showed more variability in the efficiency score as compared to the rest of the companies.

The results for the year 2016-2017 revealed that nine companies, namely TCS, HCL, TECHM, ORCLF, MNDTR, TLXI, NIIT, POLRS and HNDV, had an OTE score of one. These efficient IT companies collectively define the efficient frontier of all the selected companies. Our analysis established that MPHSI, INFY, WIPRO and PRSTN are amongst the companies that have low OTE values; MPHSI, INFY, PRSTN and ZNSR are pure technical inefficient; whereas WIPRO, INFY, ZNSR and HNDG are amongst the companies with low scale efficient scores. It was observed that OTIE of WIPRO, HXWR and HNDG is primarily due to SIE rather than PTIE. Thus, these companies can be said to be facing problems pertaining to the number of workers rather than managerial problems. Their efficiency could probably increase by reducing the number of un-utilized workers and optimizing the scale.

It is important to highlight that INFY, which is one of the leading IT software service companies in India, and has remained the corporate face of India, was found to be the least efficient on every parameter. Moreover, the production system of INFY exhibited decreasing returns to scale over the sample period. This implies that policy makers of INFY should first focus on addressing the technical inefficiency before fixing the scale of operations. The co-ordination between the management and the workers and workers’ skills enhancement could be the primary issues causing inefficiency that needs to be fixed.

It was furthermore found that HCL can be considered as the benchmark company followed by TECHM and MNDTR. INFY had maximum slack. CYENT was the only company that was overall technically inefficient but had zero slack. This indicates that there are other factors responsible for inefficiency; optimal scale efficiency may be one of the reasons. MPHSI was found to be the least efficient company.

8 Conclusion

This paper investigated the relative efficiency of the top 18 Indian IT software service companies from 2010 to 2017 using the DEA technique with workers’ effort as key input. As is evident from our research findings the average efficiency of the companies is very high i.e. 96%, indicating there is small room for improvement of about 4% in their outputs, at the existing input levels. The positive government policies towards IT sector and availability of skilled talent have helped the IT industry to continue performing at high level of efficiencies, however some companies are not able to efficiently use their inputs resulting in managerial or scale inefficiency.

The results of this study established that AOTE, APTE and ASE showed mixed trends. It was found that AOTE decreased from 0.960 to 0.953 for the period 2010 to 2016. The reasons for declining trend in AOTE is probably on account of IT services being offered as generic services rather than value added services by Indian IT service companies. During the same period, the APTE decreased from 0.985 to 0.975 which could be attributed towards implementation of inappropriate management practices and processes. At the same time, the ASE increased from 0.974 to 0.977 for the duration 2010 to 2016, indicating that on an average the IT companies have marginally shown improvement in managing their size. However, the results of the study also pointed that overall efficiency of all the IT companies increased in 2017. This increase may be on account of rigorous focus on talent management and automation processes by the selected IT companies.

Our analysis established that in 2017, the top five IT companies exhibited higher efficiency as compared to the rest of the companies. It was furthermore discovered that three amongst the top five companies, i.e. TCS, HCL and TECHM, were overall efficient. MPHSI and INFY are the amongst the less efficient IT companies. In this paper, an attempt is also made to conduct the benchmarking exercise. The results pointed out that HCL had the largest peer count, and it can be considered as the most robust company. Also, sensitivity analysis was carried out and the results were found to be stable.

On average, it is recommended that all the inefficient companies can increase their sales turnover by an average of 9%, and their net profit by an average of 38% to meet the efficiency of the benchmark companies. This implies that companies must focus on increasing their sales output and net profit to become more efficient. The results furthermore suggest that IT companies should additionally focus on improving workers’ utilization rate and workers’ skills; thus increasing workers’ productivity.

The paper also has certain limitations. Due to time and money constraints, the scope of the study was restricted to the examination of 18 selected IT companies in India. However, this study can be extended to include more number of IT companies. Also, one can conduct an efficiency comparison of Indian IT companies for pre-recession and post-recession periods. Moreover, the study can also be extended to examine the performance of IT companies of developed as well as new emerging economies.

9 Implications

This study has immense implications for the IT sector of India. The findings of the paper can be usepful for policy makers and managers in formulating policies and strategies to improve efficiency of IT companies and how to be competitive in the global IT industry.

The inefficient companies must catch up and follow the best practices of the benchmark company HCL, in terms of better project management practices, improving workers’ productivity and their utilization, reducing expenses, managing workers’ costs and organizational assets. The overall inefficiency was attributed due to both poor utilization of inputs (i.e., managerial inefficiency) and failure to operate at an optimized scale (i.e., scale inefficiency). To improve scale efficiency, companies exhibiting DRS can optimize the workers’ strength force either by retrenching or by the skill enrichment of the workers. However, those companies showing IRS can increase their efficiency by expanding their workers’ strength either by routine recruitments or by mergers and acquisitions.

In general, the IT companies in India can develop strategies that could improve utilization of inputs that include investment in software tools that increase workers’ collaborations, enrich workers’ competency, improve workers’ work environments, increased focus on workers’ health and their well-being. Moreover, enhanced focus on inclusivity and diversity also have implications for overall productivity.

The Government of India has recently taken many initiatives such as the introduction of Goods and Services Tax (GST), demonetization to curb black money, and various other reforms to promote the various industries in India. In particular, the they slashed its corporate tax by nearly 10%, which is a revolutionary step and will certainly promote start-ups and the whole corporate sector. The IT sector of India will also get a boost from this bold economic reform initiative as most of the start-ups are from the technology sector. These recent initiatives by the government are certainly going to address the issues faced by the IT sector and enhance the reputation of the corporate face of India.

Conflict of interest

None to report.

Footnotes

Appendices

Acknowledgments

The authors are thankful to the reviewers for their constructive comments and suggestions.

References

Tarutė

, Gatautis

. ICT impact on SMEs performance. Procedia-social and behavioral Sciences. 2014;110:1218–25.

Theis

, Schäfer

, Bröhl

, Schäfer

, Rasche

, Wille

, Brandl

, Jochems

, Nitsch

, Mertens

A Brandl

et al. Predicting technology usage by health information need of older adults: Implications for eHealth technology. Work. 2019;62(3):443–57.

Arvanitis

, Loukis

. Information and communication technologies, human capital, workplace organization and labour productivity: A comparative study based on firm-level data for Greece and Switzerland. Information Economics and Policy. 2009;21(1):43–61.

Ahmed

. ICT and human capital role in achieving knowledge-based economy: applications on Malaysia’s manufacturing. Journal of Information & Knowledge Management. 2006;5(2):117–28.

Abdullahi

, Shehu

, Usman

. Impact of Information Communication Technology On Organizational Productivity In The Nigeria Banking Industry: Empirical Evidence. Noble International Journal of Business and Management Research. 2019;3(1):1–9.

Sandberg

, Gardelli

, Stubbs

. The use of Information and Communication Technology (ICT) in the rehabilitation of individuals with severe functional impairments in a municipal care service system. Work. 2005;24(3):229–38.

Gartner. Gartner Perspective: IT Spending 2010. Gartner;2009.1-24.

Gartner. Gartner Says Global IT Spending to Grow 3.2 Percent in 2019 [Internet]. 2018. Available from: https://www.gartner.com/en/newsroom/press-releases/2018-10-17-gartner-says-global-it-spending-to-grow-3-2-percent-in-2019

IDC. IDC FutureScape: Worldwide IT Industry 2018 Predictions [Internet]. MA(USA): IDC; 2017. Available from: https://www.idc.com/getdoc.jsp?containerId=US43171317

10.

National association of software and services companies(IN). Impact of the IT-BPO industry in India: a decade in review. New Delhi (IN): NASSCOM, New Delhi; c2010. pp. 147.

11.

National association of software and services companies(IN). IT-BPM Industry in India: Sustaining Growth and Investing for the Future [Internet]. Hyderabad (IN): NASSCOM. Annual Report; 2017. pp. 13. Available from: https://www.nasscom.in/sites/default/files/NASSCOM_Annual_Guidance_Final_22062017.pdf

12.

Kumar

, Charles

, Mishra

. Evaluating the performance of Indian banking sector using DEA during post-reform and global financial crisis. Journal of Business Economics and Management. 2016;17(1):156–72.

13.

Kite G . The impact of information technology outsourcing on productivity and output: New evidence from India. Procedia Economics and Finance. 2012;1:239–48.

14.

Adler

, Duval

, Furceri

, Sinem

, Koloskova

, Poplawski-Ribeiro

. Gone with the headwinds: Global productivity. IMF Staff Discussion Note, SDN/17/04.International Monetary Fund; 2017.

15.

Kuo

, Yang

. Does intellectual capital matter? Assessing the profitability and marketability of IC design companies. Quality & Quantity. 2012;46(6):1865–81.

16.

Chang

, Hung

, Huang

. Evaluating the operational performance of knowledge-based industries: the perspective of intellectual capital. Quality & Quantity. 2013;47(3):1367–83.

17.

Harriss

. Book Review [ Upadhya C, Vasavi AR (Eds.), In an Outpost of the Global Economy: Work and Workers in India’s Information Technology Industry. Routledge, New Delhi, IN;2008]. Journal of South Asian Development. 2010;5(2):293–5.

18.

Chen

, Wang

, Wu

, Zhang

. Analysing firm performance in Chinese IT industry: DEA Malmquist productivity measure. International Journal of Information Technology and Management. 2011;10(1):3–23.

19.

Lampe

, Hilgers

. Trajectories of efficiency measurement: A bibliometric analysis of DEA and SFA. European Journal of Operational Research. 2015;240(1):1–21.

20.

Liu

, Huang

. China’s high-tech industry efficiency measurement with virtual frontier data envelopment analysis and Malmquist productivity index. Expert Systems. 2019:e12450.

21.

Chen

, Ali

. DEA Malmquist productivity measure: New insights with an application to computer industry. European Journal of Operational Research. 2004;159(1):239–49.

22.

Liu

, Wang

. Efficiency measures of PCB manufacturing firms using relational two-stage data envelopment analysis. Expert Systems with Applications.. 2009;36(3):4935–9.

23.

Dash Wu

, Ho

. Productivity and efficiency analysis of Taiwan’s integrated circuit industry. International Journal of Productivity and Performance Management. 2007;56(8):715–30.

24.

Chou

, Shao

. Total factor productivity growth in information technology services industries: A multi-theoretical perspective. Decision Support Systems. 2014;62:106–18.

25.

Mathur

. Indian IT industry: a performance analysis and a model for possible adoption [Internet]. Munich Personal RePEc Archive Discussion Paper No.2368. 2007. Available from: https://core.ac.uk/download/pdf/6799223.pdf

26.

Bhattacharjee

. Efficiency dynamics and sustainability of the Indian IT-ITeS industry: An empirical investigation using DEA. IIMB Management Review. 2012;24(4):203–14.

27.

Sahoo

, Nauriyal

. Trends in and determinants of technical efficiency of software companies in India. Journal of Policy Modeling. 2014;36(3):539–61.

28.

Cooper

, Seiford

, Tone

. Data Envelopment Analysis: A Comprehensive Text with Models, Applications, References and DEA-Solver Software. 2nd ed. New York, NY: Springer Science+Business Media, LLC; 2007.

29.

Emrouznejad

, Yang

. A survey and analysis of the first 40 years of scholarly literature in DEA: 1978–2016. Socio-Economic Planning Sciences. 2018;61:4–8.

30.

Oruç

, Altın

. The Effect of the Financial Crisis on the Information Technologies Sector: Application of Malmquist Productivity Index Method. International Journal of Business and Social Research. 2015;5(6):1–11.

31.

Sickles

, Zelenyuk

. Measurement of Productivity and Efficiency. Great Britain. Cambridge University Press; 2019.

32.

Trierweiller

, Peixe

, Tezza

, do Valle Pereira

, Pacheco

Jr , Bornia

, de Andrade

. Measuring organizational effectiveness in information and communication technology companies using item response theory. Work. 2012;41(Supplement 1):2795–802.

33.

Hajiagha

, Mahdiraji

, Tavana

. A new bi-level data envelopment analysis model for efficiency measurement and target setting. Measurement. 2019;147:106877.

34.

Coelho

, Lourenço

. A tentative efficiency index for pointing device use in computer aided design: A pilot study. Work. 2018;61(1):157–70.

35.

Lefever

, Decuman

, Perl

, Braeckman

, Van de Velde

. The efficacy and efficiency of Disability Management in job-retention and job-reintegration. A systematic review. Work. 2018;59(4):501–34.

36.

Draca

, Sadun

, Van Reenen

. Productivity and ICT: A Reviewof the Evidence. In: Mansell

, Avgerou

, Quah

, Silverstone

(Eds.), The Oxford Handbook of Information and Communication Technologies, Oxford University Press UK, 2016.

37.

Iqbal

, Hassan

, Peng

. Analyzing the role of information and telecommunication technology in human development: panel data analysis. Environmental Science and Pollution Research. 2019;26(15):15153–61.

38.

. ICT as a source of economic growth in the information age: Empirical evidence from the 1996–2005 period. Telecommunications Policy. 2011;35(4):357–72.

39.

Ciccarelli

, Straker

, Mathiassen

, Pollock

. ITKids Part I: Children’s occupations and use of information and communication technologies. Work. 2011;38(4):401–12.

40.

Palmer

, Ciccarelli

, Falkmer

, Parsons

. Associations between exposure to Information and Communication Technology (ICT) and reported discomfort among adolescents. Work. 2014;48(2):165–73.

41.

Robertson

, Boiselle

, Eisenberg

, Siegal

, Chang

, Dainoff

, et al. Examination of computer task exposures in radiologists: a work systems approach. Work. 2012;41(Supplement 1):1818–20.

42.

De Brito

, Da Silva

, De Souza

, Barros

. Study of globe temperature relative to air temperature during cognitive activities in information technology laboratories. Work. 2019;62(3):393–410.

43.

Sadeghniiat-Haghighi

, Bahrami

, Aminian

, Meysami

, Khajeh-Mehrizi

. Melatonin therapy in shift workers with difficulty falling asleep: A randomized, double-blind, placebo-controlled crossover field study. Work. 2016;55(1):225–30.

44.

Puri

, Yadav

. Performance evaluation of public and private sector banks in India using DEA approach. International Journal of Operational Research. 2013;18(1):91–121.

45.

Puri

, Yadav

. Productivity and profitability of public sector banks in India: a multi-component DEA approach. International Journal of Operational Research. 2017;30(2):233–76.

46.

Avkiran

. Productivity analysis in the service sector with data envelopment analysis. 3rd ed. Queensland: Avkiran NK, university of Queensland; 2006.

47.

Shao

, Shu

. Productivity breakdown of the information and computing technology industries across countries. Journal of the Operational Research Society. 2004;55(1):23–33.

48.

Toloo

, Nalchigar

, Sohrabi

. Selecting most efficient information system projects in presence of user subjective opinions: a DEA approach. Central European Journal of Operations Research. 2018;26(4):1027–51.

49.

Avkiran

. Investigating technical and scale efficiencies of Australian universities through data envelopment analysis. Socio-Economic Planning Sciences. 2001;35(1):57–80.

50.

Adeyemi

. ICT facilities: Ergonomic effects on academic library staff. Library Philosophy and Practice [internet]. 2010 Mar 1 [cited 14 October 2019].. Available from: http://digitalcommons.unl.edu/cgi/viewcontent.cgi?article=1351&context=libphilprac

51.

Goodman

, Landis

, George

, McGuire

, Shorter

, Sieminski

, Wilson

. Effectiveness of computer ergonomics interventions for an engineering company: a program evaluation. Work. 2005;24(1):53–62.

52.

Ray

, Tewari

. Design and implementation of ergonomic performance measurement system at a steel plant in India. Work. 2012;41(Supplement 1):5943–9.

53.

Onkham

, Karwowski

, Ahram

. Economics of human performance and systems total ownership cost. Work. 2012;41(Supplement 1):2781–8.

Descriptive Statistics	All IT companies - Overall Technical Efficiency (OTE)
	2010	2011	2012	2013	2014	2015	2016	2017
Average Efficiency (AOTE)	0.960	0.944	0.957	0.952	0.949	0.955	0.953	0.970
Minimum	0.880	0.766	0.790	0.782	0.708	0.838	0.872	0.861
Maximum	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000
Std. Deviation	0.044	0.066	0.060	0.057	0.072	0.054	0.051	0.042
Avg. Inefficiency % = 1-AOTE	4.03	5.61	4.30	4.78	5.14	4.49	4.74	3.05
Descriptive Statistics	All IT companies - Pure Technical Efficiency (PTE)
Average Efficiency (APTE)	0.985	0.969	0.987	0.981	0.964	0.967	0.975	0.984
Minimum	0.920	0.787	0.933	0.902	0.790	0.856	0.892	0.868
Maximum	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000
Std. Deviation	0.024	0.056	0.023	0.031	0.056	0.046	0.037	0.034
Avg. Inefficiency % = 1-APTE	1.52	3.09	1.29	1.94	3.63	3.31	2.53	1.65
Descriptive Statistics	All IT companies - Scale Efficiency (SE)
Average Efficiency (ASE)	0.974	0.974	0.969	0.971	0.983	0.988	0.977	0.986
Minimum	0.880	0.838	0.790	0.782	0.896	0.926	0.885	0.916
Maximum	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000
Std. Deviation	0.033	0.042	0.054	0.051	0.027	0.019	0.031	0.022
Avg. Inefficiency % = 1-ASE	2.56	2.58	3.06	2.87	1.67	1.24	2.29	1.42