IT outsourcing and impacts in Thailand’s financial institutions

Abstract

Information technology outsourcing (ITO) studies have been the focus of researchers for decades. One topic has been the factors and impact of ITO on organizations. Inconclusive findings about relationships between the levels of ITO and ITO success demands further exploration on this issue. In Thailand, the National Electronics and Computer Technology Center (NECTEC) reported that the financial and banking industry has the highest spending on computer services primarily by outsourcing the core banking and service management system. Nonetheless, research on ITO in Thailand’s financial industry has been limited. This study is among the first to empirically investigate the relationship between the levels of ITO, defined as the percentage of IT activities and components selected to be outsourced, and ITO success, defined by the benefits received from outsourcing in Thailand. A three-pronged approach is used in this study. It began with a comprehensive literature review. This was followed by in-depth expert interviews with representatives from financial institutions, academia, and outsource vendors, all of who offered opinions that shaped the development of the research model. A survey was then carried out with 428 IT personnel from the Thai financial sector. The results were analyzed using structural equation modeling (SEM). Contrary to previous researches, this study found that internal uncertainty and internal IT skills were two factors that do not affect the levels of ITO, while transaction costs and specific assets do. In addition, it was found that levels of ITO have impacts on ITO success. In a real world application, this study focuses on the different levels of ITO implementation among Thailand’s top five banks and their effects on ITO success. The findings increase researchers’ understanding regarding the factors that affect the levels of ITO and suggest that practitioners must consider the levels of ITO, in addition to other factors, when making ITO decisions. The framework can help practitioners optimize ITO practice for successful ITO outcomes.

Keywords

information technology outsourcing outsourcing success financial institutions Thailand

Determining whether or not IT outsourcing adds value to the organization is an important management question.

Introduction

Information technology (IT) plays an increasingly significant role in organizational transactions and processes. Thus, IT has become an important resource of existing business strategy (Kaplan and Norton, 2004). However, the complexity of IT management, the rapid changing of IT tools and high IT investment have increased (Udo, 2000; Han et al., 2008; Shi, 2010). Organizations that have a mature IT process are more inclined to use IT outsourcing (ITO) strategies so they can focus on their core business competencies and explore the latest technologies from outside vendors (Gulla and Gupta, 2012). The scale and range of ITO has become a global phenomenon evidenced by the growth from US$184 billion in 2003 to US$246.6 billion by the end of 2011, as reported by Gartner inc. (Lin et al., 2007; Britz et al., 2012). Accordingly, ITO has become a common issue in the organizational management agenda (Aubert et al., 2004; Jiang et al., 2006; Karyda et al., 2006).

In this research, ITO is defined as the practice of transferring IT infrastructure, activities, people and management responsibility for delivery of services and products to an external vendor (Loh and Venkatraman, 1992; Hirschheim and Lacity, 2000). In a general sense, as outsourcing increases, investments in facilities, equipments and manpower are reduced, and costs will decline. Even though ITO practice has increased over the years, there are few organizations that have reported ITO success (Willcocks and Lacity, 1998; Wang and Yang, 2007; Han et al., 2008). Some organizations gained an advantage from ITO while others failed (Lacity and Hirschheim, 1993; Rochester and Douglas, 1993). Lacity et al. (1995) also indicated that nearly 70 percent of the organizations which have undergone ITO deals expressed dissatisfaction and others were not successful in terms of quality, productivity and cost efficiency (Pati and Desai, 2005; Rottman and Lacity, 2009).

ITO increase in Thailand is the same as in other countries. As reported by Thai public organizations, the growth of the ITO market is among the top three in computer services (19.3 percent), with a total market value of THB5,679 million (US$189 million) in 2010 (NECTEC, 2010). Additionally, the financial and banking industry held the highest share in computer services spending through outsourcing its core banking and service management systems (NECTEC, 2010).

The Central Bank of Thailand (2013) organizes 137 financial institutions into 12 types, including Thai commercial banks, retail banks, subsidiaries, foreign bank branches, finance companies, credit financiers, foreign bank representatives, assets management companies, specialized financial institutions, Thai asset management corporations, credit card companies and personal loan companies.

However, since 2008, nearly 60 percent of Thai financial institutions’ total assets, deposits and loans are dominated by the big four domestic Thai commercial banks, namely, Bangkok Bank, Krung Thai Bank, Kasikorn Bank, and Siam Commercial Bank (Sufian and Habibullah, 2010). In order to accurately represent the Thai financial sector, this study focuses on Thai commercial banks and, more specifically, the big four and other financial institutions that heavily use outsourcing strategy in their IT functions.

ITO is a fairly recent option for financial institutions in Thailand. The first of the big four banks that used an ITO strategy signed a 10-year contract with IBM in 2003 (Sufian and Habibullah, 2010). NECTEC reported that the reason for IT market growth in Thailand is based on financial institutions upgrading their IT systems to comply with the Bank of Thailand’s regulations such as Basel II, which focuses on maintaining IT contingency plans, and the implementation of a computer-related crime act (NECTEC, 2010). Additionally, due to intensified competition in the financial industry, financial institutions begun using marketing strategy and advanced IT services by external vendors to improve operational efficiency (Bátiz-Lazo and Wood, 2001).

Though ITO among financial institutions in Thailand is rising, empirical studies and research in this area are limited (Ang and Cummings, 1997; Marcolin and McLellan, 1998; Baldwin et al., 2001; Goolsby et al., 2002). The research on outsourcing in Thailand primarily focuses on material and service outsourcing in the manufacturing industry (Thangavelu and Chongvilaivan, 2011), logistic outsourcing (Kitcharoen, 2003), surgery outsourcing in medical tourism (Bies and Zacharia, 2007; Marlowe and Sullivan, 2007), and relationship outsourcing in the automobile industry (Chiasakul, 2004; Kess et al., 2008).

This research attempts to fill in the context of Thailand’s financial institutions by investigating the relationship between the levels of ITO and ITO success. To address the gap in research regarding ITO, the following questions were formulated:

RQ1: What is the relationship between the levels of ITO and ITO success in Thailand?

RQ2: Are the drivers of the levels of ITO the same as those found in previous research?

To answer these questions, this study adopts the basic concept from the models of Ang and Straub (1998) and Aubert et al. (2004). The paper proceeds in the following way. First, the theoretical background of the study is presented. The model and hypotheses are tested using a field survey. Factor analysis and structural equation model (SEM) are used in data analysis. Finally, results and limitations of the study are discussed, and the conclusion is drawn.

Literature review

Theoretical background

Transaction cost theory (TCT) provides much of the theoretical background for ITO research (Lacity and Hirschheim, 1993; Aubert et al., 1996; Aubert et al., 2004; Whitten and Leidner, 2006). The theory explains that firms will consider all transaction costs and benefits, and choose to outsource when net potential gain is expected (Gottschalk and Solli-Sæther, 2005).

The TCT concept is applicable to firms that have the objective of finding ways to minimize costs associated with economic transactions (Williamson, 1981). Grover, et al. (1996) also apply the TCT in their research, stating that organizations will choose to outsource when outside vendors have greater efficiency than internal resources, thereby reducing an organization’s costs.

The TCT presents a set of principles for analyzing buyer-supplier transactions and determining what functions should be outsourced (Lacity and Hirschheim, 1993). The TCT is based on two assumptions: bounded rationality and opportunism (Williamson, 1985). Bounded rationality refers to a limitation in the way a person thinks or reasons, and leads to making an imperfect assessment. Opportunism occurs when a person is motivated to take advantage of someone else through some form of manipulation, such as by taking advantage of another person’s lack of information. When this happens, a transaction cost results from contractual opportunism.

Therefore, before making any decision to outsource, firms will analyze transaction costs that can occur (Ang and Straub, 1998; Grossman and Helpman, 2002). One reason for ITO failure occurs due to the complexity of the ITO decision, because most outsourcing decisions are based on limited information between organizations and outsource vendors (Loh and Venkatraman, 1992; Lacity and Hirschheim, 1993).

Levels of IT outsourcing

Level of ITO in this research is defined as the percentage of transferring IT-related activities and components that are not a part of an organization’s core competency to external vendors (Lacity and Willcocks, 2000; Gonzalez et al., 2005a; Pati and Desai, 2005; Tafti, 2005). Firms typically select IT activities and IT components to be outsourced based on their strategies (Earl, 1996). IT activities consist of the maintenance, operation, development and design of technology (Österle and Winter, 2003), while IT components include IT infrastructure, applications, tasks and processes (Zarnekow et al., 2004).

Additionally, to increase the understanding of the levels of ITO, the researchers reviewed factors in the literature that could affect the levels of ITO. Ang and Straub (1998) focused on the economic determinants of ITO that influenced outsourcing decisions. Their research looked at production cost, transaction costs, and financial slack. They argued that transaction cost plays the main role in the decision-making process. Financial slack and business skills are omitted from their research because prior findings argued that there was no statistically significant effect on the levels of ITO.

Aubert et al. (2004) discussed the characteristics of IT operation activities by developing a transaction cost model of ITO. They examined specific elements of transaction cost which influence the levels of outsourcing. They concluded that high IT asset specificity and high uncertainty act as deterrents in the ITO decision, while high levels of technical skills lead to increased levels of ITO.

Table 1 summarizes the determinants and measurements of levels of ITO that are used in this study from previous literature.

Table 1.

Levels of ITO determinants and measurements.

Determinants	Measurement	Source
Levels of ITO	Percentage of outsourcing activities/components	Lacity et al. (1996), Lacity and Willcocks (2000), Gonzalez et al. (2005a)
Transaction costs	Outsourcing costs (monitoring cost, contracting cost, problem cost, etc.) Perceived transaction costs	Aubert et al. (1996), Ang and Straub (1998), Lacity and Willcocks (1998), Santos and Eisenhardt (2005)
Specific Assets	Differentiation HR and technical specificity Client and supplier investment	Nam et al. (1996), Aubert et al. (2004), Bahli and Rivard (2005)
Uncertainty	Demand uncertainty Job standardization Task complexity	Robertson and Gatignon (1998), Poppo and Zenger (2002), Aubert et al. (2004), Bahli and Rivard (2005)
IT Skills	Technical skill Capability skill	Aubert et al. (2004) Pati and Desai (2005)

IT outsourcing success

ITO success is defined as satisfaction with the benefits from outsourcing gained by organizations (Grover et al., 1996). Lee et al. (2004) defined ITO success as the levels of appropriateness between the organization’s requirements and outsourcing outcomes delivered by the service provider.

However, since general outsourcing is motivated by strategic, economic and technological benefits, many researchers use these benefits to evaluate ITO success (Grover et al., 1996; Lee and Kim, 1999; Lee, 2001; Jiang et al., 2006; Gottschalk and Solli-Sæther, 2006; Dahlberg and Lahdelma, 2007).

Strategic benefits are defined as the ability of a firm to focus on its core business by outsourcing routine IT activities (Lacity and Willcocks, 1998). Technological benefits are defined as the ability of a firm to gain access to leading-edge IT and to avoid the risk of technological obsolescence that results from dynamic changes in IT (Lacity and Willcocks, 1998; Lee and Kim, 1999). And economic benefits are defined as the ability of a firm to use the expertise and economies of scale in human and technological resources of the service provider and to manage its cost structure through unambiguous contractual arrangements (Ngwenyama and Bryson, 1999).

Research framework

The main focus of this study is to investigate the relationship between the levels of ITO and ITO success. As focused by prior researchers based on the TCT (Lacity and Willcocks, 1998; Ang and Straub, 1998; Aubert et al., 2004; Dahlberg and Lahdelma, 2007), the levels of ITO are generally categorized as total outsourcing or selective outsourcing, based on the percentage of IT budget being outsourced.

The studies by Lacity and Willcocks (1998) in the US and UK and a study by Seddon et al. (2002) in Australia found that selective outsourcing presented more successful results. The studies by Dahlberg et al. (2006) in Finland, and Lee et al. (2004) in South Korea, however, found no statistically significant differences between selective and total outsourcing and ITO success. Inconclusive findings about the relationship between levels of ITO and ITO success suggest the need for further exploration on this issue.

To answer these research questions, the researchers developed the research framework shown in Figure 1.

Figure 1.

Main Research Framework.

The first hypothesis explores the relationship between the levels of ITO and its success. This hypothesis is based on the previous study of Grover et al. (1996). In their research, they used IT functionality (such as applications, telecom development, and system operations) as the determinants for the levels of outsourcing. They determined that if outsourcing is undertaken with conscious economic, strategic and technology considerations as conducted by TCT, the relationship between ITO and ITO success would be positive and significant. This research therefore hypothesizes that:

H1: The level of ITO has positive impacts on ITO success.

Moreover, to better understand the factors that influence the levels of ITO based on the literature review, four determinants consisting of transaction costs, specific assets, internal uncertainty and internal IT skills are included. Therefore, this research has expanded the research framework to be followed as shown in Figure 2.

Figure 2.

Complete Proposed Framework.

Transaction costs

The TCT extends the neoclassical economic perspective by recognizing the significance of transaction costs in any market change. Transaction costs refer to all costs that occur in the ITO process, i.e. costs of effort, time, management, coordinating, monitoring, negotiation, communication, and making contracts (Porter and Millar, 1985; Mahoney and Pandian, 1992; Williamson, 1985).

Many scholars have stated that transaction costs are an important factor in making outsourcing decisions (Ang and Straub, 1998; Grossman and Helpman, 2002). When a firm foresees substantial efforts and costs in supervising, coordinating, and monitoring the activities of the vendor, it may decide that external sourcing is too costly (Ang and Straub, 1998). This research therefore hypothesizes that:

H2: The levels of transaction costs have negative impacts on the levels of ITO.

Specific assets

Specific assets are defined as any specialized knowledge or skill developed from experience within an organization, or specialized hardware or software customized for a particular purpose in an organization (Williamson, 1985; Heide, 1994; Ang and Cummings, 1997). Aubert et al. (2004) argued that from the TCT perspective, investments in specific assets lead to additional transaction costs. As the organizations’ assets and processes become more complex, ITO becomes increasingly costly to implement and train. When these costs are too high, organizations will decide to conduct the transaction in-house (Aubert et al., 2004; Malik and Nilakant, 2011; Watjatrakul, 2005).

Specific assets are also found to be the most important dimension of a transaction wherein the vendor provides specific investments in physical (special software) or human (specialized technical skills of personnel) capital. The asset-specific transactions can lead to a tied-up relationship between the transacting parties since the supplier is “locked into” the transactions and the buyers cannot turn to alternate suppliers because of the high cost of supply. This research therefore hypothesizes that:

H3: The levels of specific assets have negative impacts on the levels of ITO.

Internal uncertainty

Uncertainty is defined as a lack of information (Aubert et al., 1996). There are a number of unexpected situations which may occur during the transaction process (Miozzo and Grimshaw, 2005). For example, uncertainty influences the levels of ITO because of bounded rationality and opportunism. Both partners seek to reduce uncertainty by negotiating complex contracts or by extensively renegotiating. Aubert et al. (1996) stated that organizations will try to predict all the possibilities associated with the governance of their particular transaction. The more complex the environment, the more partners will seek to negotiate comprehensive agreements.

Williamson (1985) found that it is difficult for buyers of goods and services to evaluate the suppliers’ actions because of imperfect information and uncertainty. Additionally, Nam et al. (1996) found that uncertainty played a significant role in the decision to outsource. Aubert et al. (2004) also confirmed that increased uncertainty can lead to increased use of internal governance rather external vendors. High levels of uncertainty could lead to decreases in organizations’ decisions to outsource. So this research hypothesizes that:

H4: The levels of internal uncertainty have negative impacts on the levels of ITO.

Internal IT skills

IT skills are defined as technical skills and capabilities that were obtained from training, education, and experiences related to the accomplishment of certain tasks (Medlin et al., 2007; Goo et al., 2000). Although in theory outsourcing provides access to technical skill and IT specialists’ expertise, what often happens in practice is that the outsourcing firm is supported by the organization’s own staff (Fowler and Jeffs, 1998).

Even though the vendor supplies innovative services to the customer, a large part of the new knowledge required remains in the hands of the providers and cannot be transferred to the customers (Gonzalez et al., 2005b; Gonzalez et al., 2005a). Therefore, the customers need to retain certain know-how and internal capacities in both technical and business areas if they want to handle the outsourcing relationship properly (Willcocks et al., 1999). If internal staff have the IT skills to perform the required tasks, it is not necessary to utilize ITO. Accordingly, the higher the skills of the internal staff, the lower the levels of ITO. Thus, the research hypothesizes that:

H5: The levels of internal IT skills have negative impacts on the levels of ITO.

Research methodology

Research methodology started with a review of the literature to identify measures for each construct. Experts were then interviewed to confirm the understanding of organizational ITO perceptions. Next, a pilot study was conducted with IT professionals from both private and public financial institutions. Last, a full scale survey was conducted to evaluate the hypothesis.

Expert interview

The expert interviews were conducted with both practitioners and academicians. As shown in Table 2, there were 25 persons involved from three sectors: 12 persons from the financial sector, 8 persons from academia, and 5 persons from outsource vendors. IT professionals and managers from the financial sector and the outsource vendors represent the practitioner’s perspectives. Their expertise was determined based on their positions and experience with firms.

Table 2.

Profile of experts interviewed.

Position Title ^a	Organization Type ^b	IT experience (years)	ITO experience (years)	Persons
SVP, VP, AVP (in IT department)	Specialized financial institutions	10–25	5–15	6
VP, AVP (in IT department)	Commercial banks	8–20	5–20	6
Associate professor	Academic institutions	18–25	5–10	4
IT Consultant	Academic institutions	10–15	10–12	4
VP, AVP, IT manager	Business IT companies	10–20	8–15	5

Notes: ^a Position titles may be different in each organization (AVP – Assistant Vice President, VP – Vice President, SVP – Senior Vice President); ^bExcept for academic institutions, organization types are categorized by the Bank of Thailand.

The interviews employed semi-structured and open-ended questions. They averaged about 90 minutes and usually involved a tour of the facility and, in some cases, opportunities to meet employees involved in the operations. Most interviews were conducted in Thai, while the remainder were conducted in English for non-native speakers. The purpose of the interviews was to gain an initial understanding of ITO practice in Thailand, confirm definitions, verify measurements, fine-tune the framework, and validate the ITO questionnaire. The interviews were transcribed and coded manually using the four categories based on the recommended theoretical framework to guide analysis in each decision process (Hamzah et al., 2010). After the transcription of interviews, an interpretation and identification of issues and themes was carried out through a process of multiple individual reading of the transcripts, followed by intensive periods of discussion between the researchers. Table 3 shows an example of the interpretation and identification of themes during data analysis.

Table 3.

Example of interpretation and identifications of themes.

Quotation	Interpretation	Themes
“If the cost for outsourcing is lower than insourcing, we should outsource, especially non-core business and non-critical outcomes of the organization” (VP from commercial bank)	They were focused on the costs and benefits before making their decision to outsource.	Transaction costs
“We should be concerned with losing our organization’s specific knowledge and information before implementing an outsourcing strategy” (VP from business IT company)	They considered the implications of losing a competitive advantage by sharing knowledge of their specific assets.	Specific assets
“We had no ITO experience and knowledge of how well we would do. We had no idea how much and which activities should be outsourced. We had no plan to start the ITO process” (SVP from Specialized financial institution)	They were not sure about their internal information, process, method, etc.	Internal uncertainty
“If internal IT staff have the capacity or IT skills to perform the required tasks, it is not necessary to utilize ITO” (IT consultant from academic institutions)	They considered their internal capability before making their decision to outsource.	Internal IT skills

When available, reports, contracts, and any other relevant materials were collected to obtain a complete picture. The discussions were supported by our literature review and the theoretical foundations. The information was classified into variables used in the study: transaction costs, specific assets, uncertainty and IT skills. All data obtained from expert interviews was consolidated to create a framework between the levels of ITO and ITO success, and subsequently integrated into the questionnaire. The final questionnaire was then piloted.

Table 4 presents the summary of determinants from the literature review, expert interviews and pilot study, which provide a guideline for the questionnaire and final framework development.

Table 4.

Determinants of levels of ITO.

Determinants	Literature Review	Expert Interview	Pilot Study	Final Framework
Transaction Costs (TC)
Actual TC	–	–	–	–
Perceived TC	*	*	*	*
Specific Assets (SP)
Firm SP	*	*	*	*
IT SP	–	*	*	*
Uncertainty (UN)
External UN	*	–	–	–
Internal UN	*	*	*	*
IT skills (SK)
External SK	–	*	–	–
Internal SK	*	*	*	*

Pilot study

The measurement instruments were validated by experts in the field, a pilot test, an internal reliability test, and statistical analysis. The initial questionnaire was evaluated for consistency by five IT experts. The pilot study questionnaire was pre-tested by 70 IT professionals from five financial institutions. The questionnaire’s wording, structure, sequence, and overall presentation were then adjusted accordingly.

Field survey

A full scale survey was conducted with IT professionals and executives in Thai commercial banks and specialized financial institutions registered at the Bank of Thailand (BOT), a regulatory body of Thai financial institutions. The questionnaire consists of both 5-point Likert-scale and open-ended questions. The survey was designed to obtain a comprehensive view of ITO practices.

Each question item was developed from a revised framework. There were six parts in the questionnaire: participants and their organizational profiles, IT department profiles, ITO terminologies and issues, outsourcing decision process and factors, and general comments. The levels of ITO can be divided into IT activities and IT components as shown in Table 5. All firms were involved in either partial or full ITO. In total, 670 questionnaires were sent and 428 were completed and collected, yielding a response rate of 64 percent. Table 6 reports survey respondents’ profile.

Table 5.

Levels of ITO^a (IT activities and IT components).

Item	Functions Area	Mean	S.D.
1	IT Activity – Operation	3.10	1.22
2	IT Activity – Design	3.17	1.09
3	IT Activity – Development	3.51	1.11
4	IT Activity – Maintenance	3.24	1.11
5	IT Component – Infrastructure	3.28	1.21
6	IT Component – Task	3.20	1.17
7	IT Component – Application	3.37	1.10
8	IT Component – Process	2.88	1.12

Notes: ^a Levels of ITO (where 1 = outsource < 20%, 2 = outsource 21% - 40%, 3 = outsource 41% – 60%, 4 = outsource 61% – 80%, 5 = outsource 81% – 100%)

Table 6.

Survey respondents’ profile.

Survey Characteristic	Frequency	Percent
Financial organizations
Institution Type Specialized financial institutions	129	28.35
Commercial banks	299	71.65
Employee expertise
IT Experience Less than 1 year	80	18.69
1–7 years	132	30.84
8–14 years	95	22.20
15–20 years	87	20.33
More than 20 years	34	7.94
ITO Experience Less than 1 year	160	37.38
1–7 years	127	29.67
8–14 years	116	27.11
15–20 years	25	5.84

Analysis and results

Exploratory factor analysis

Exploratory factor analysis (EFA) was conducted to examine common aspects and reveal certain determinants of ITO. EFA is data driven (Brown, 2006). EFA was used to identify the structure among the questionnaire items and reduce questionnaire items, by confirming items and structure suitability for each construct. Principle component analysis with varimax rotation method was used to examine the dimensionality.

According to the sample size, a cut-off factor loading of 0.5 was used to screen out items that were weak indicators of the constructs (Hair et al., 2006). The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy was 0.864, which was well above the acceptable 0.50 level. In addition, Bartlett’s Test of Sphericity indicated that there was a correlation among the variables (significant 0.000).

Hair, et al. (2006) suggested that as a minimum requirement the sample size should be at least five times the total number of variables. The sample size in this study met that requirement, as there was over eight times the number of samples (428) as the variables (53). Of the 55 items we started with, two items, one from transaction costs and one from specific assets, were omitted since the factor loading was less than 0.5 and Eigenvalue was greater than 1.0. Based on factor loading and communality results, the collected data was determined to be suitable to conduct factor analysis. In Table 7 sample results are reported.

Table 7.

Sample of factor analysis results.

Items	Factor loading	Communality
Transaction costs
Total cost of IT infrastructure, IT applications and process is lower due to appropriate outsourcing	0. 799	0.792
Organization faced with the lock-in situation problem when they want to change the providers such as switching cost	0.775	0.762
Organization faced with the problem of hidden costs from the outsourcing contract	0.763	0.781
Organization loses of the ability of innovation in the long run	0.717	0.760
Specific assets
Outsource providers need specific knowledge / specific technologies for supporting the requirements of organization	0.801	0.736
Organization products and services are different, outsource providers require times of learning to support	0.714	0.738
Organization management complexity surpasses outsource provider’s ability to handle	0.762	0.654
Internal uncertainty
The contracts written by outsource providers do not cover all requirements (i.e. specific requirements)	0.812	0.736
Organization needs are difficult to understand and predict (i.e. quantity, quality and time-frame of output)	0.792	0.784
The unreliable standard of products and services and outsource time-frame	0.694	0.708
In the period of outsourcing, it is difficult to monitor the trend of changing IT business needs	0.657	0.678
Internal IT skills
Organization has IT staffs who have capability to integrate new IT technical with organization business	0.865	0.744
Organization has IT staffs who have ability to learn and modify IT functions along with the business requirement	0.806	0.840
Organization can effectively and efficiently use IT tools	0.774	0.801
Organization can effectively align IT strategies	0.712	0.704
Strategic benefits
Organization has more time to focus/concentrate on core business	0.817	0.747
Organization gains better performance and quality in business/IT process	0.761	0.675
User satisfaction of products and service increases	0.711	0.783
Economic benefits
Total costs of the IT organization decrease	0.753	0.628
Cost efficiency in IT investment increases (may not be reduced)	0.820	0.784
Organization has increased access to skilled personnel	0.738	0.693
IT expenditure has better control (more effective control)	0.711	0.658
Technical benefits
IT capability increases after outsourcing	0.783	0.791
Outsourcing reduces risk of IT obsolescence	0.636	0.655
Outsource practice improves quality of IT infrastructure, software and application	0.733	0.772

The content validity of each construct was established by certifying that the measurement items were reliable, when compared with results from the literature, the three sectors represented by expert interviews, and the pilot-study testing. Cronbach’s alpha (α), a measure of internal consistency, was calculated to confirm the reliability of each determinant. According to Hair, et al. (2006), an acceptable level is above 0.70, and this may decrease to 0.60 in exploratory research. All variables in the analysis had Cronbach’s alphas above 0.70, ranging from 0.79 to 0.92. Therefore, all variables had an acceptable content validity.

Then, convergent validity was assessed to test the construct reliability. Hair et al. (2006) defined convergent validity as “the extent to which indicators of a specific construct converge or share a high proportion of variance in common”. The composite reliability and the average variance extracted (AVE) were used. The composite reliability values ranged from 0.92 to 0.99. These are higher than the 0.7 threshold commonly used for acceptable reliability. The AVE values for the measurements ranged from 0.51 to 0.64, while the threshold is 0.5. The reliability results are reported in Table 8.

Table 8.

Reliability results.

Measure	Number of	Cronbach’s	Composite	AVE
Measure	Items	alpha (α)	Reliability	AVE
IT activities	10	0.92	0.98	0.59
IT components	10	0.84	0.97	0.51
Transaction costs	7	0.81	0.92	0.50
Specific assets	6	0.87	0.96	0.51
Internal uncertainty	6	0.79	0.96	0.52
Internal IT skills	4	0.85	0.97	0.64
Strategic benefit	3	0.86	0.96	0.55
Economic benefit	4	0.79	0.98	0.54
Technical benefit	3	0.89	0.99	0.50

Finally, discriminant validity was analyzed via AVE by examining the square root of AVE. Hair, et al. (2006) define discriminant validity as “the extent to which a construct is truly distinct from another construct”. The rule is that variables should relate more strongly to their own factor than to another factor. The result of the square root of the AVE for each construct was greater than the correlations between all other constructs, which means that the constructs successfully passed the criteria. Table 9 presents the correlations between constructs; the values range from 0.71 to 0.93.

Table 9.

Correlations between constructs.

Constructs	1	2	3	4	5	6	7	8	9
IT activities (1)	0.78
IT components (2)	0.71	0.93
Transaction costs (3)	0.17	0.15	0.71
Specific assets (4)	0.19	0.21	0.44	0.71
Internal uncertainty (5)	0.08	0.15	0.48	0.59	0.72
Internal IT skills (6)	0.11	0.11	0.41	0.51	0.45	0.80
Strategic benefit (7)	0.32	0.32	0.46	0.46	0.41	0.37	0.74
Economic benefit (8)	0.34	0.37	0.59	0.33	0.30	0.27	0.69	0.73
Technical benefit (9)	0.26	0.24	0.47	0.48	0.46	0.41	0.71	0.65	0.73

Confirmatory factor analysis

Confirmatory factor analysis (CFA) was conducted to determine the level to which alternative models explain the relationships between items in a scale. CFA is theory- or hypothesis driven (Albright and Park, 2009). CFA is most useful when the researchers test a priori models, because more effective decisions can then be made about the possibility of the target model (Thompson and Daniel, 1996). Moreover, CFA is used when the researchers have knowledge of the underlying latent variable structure (Byrne, 2001), allowing them to specify the number of factors or set the effect of one latent variable on observed variables to particular values (Albright and Park, 2009).

Since the main objective of this study is to confirm the proposed framework and hypothesized model, a measurement part and a path model will be verified separately. Based on the two-step rule, if both measurement part and structural parts of the model are identified, the overall fit of the model ensures that the data are valid enough to represent the entire model; this approach helps researcher to modify the unfitting model more easily (Kline, 2010). The structural equation modeling (SEM) technique is applied in this study because it has ability to handle both confirmative factor analysis and path model. Accordingly, AMOS 18 was employed for this purpose.

SEM is a statistical tool which takes a confirmatory approach to the analysis of a structural theory (Byrne, 2001). Therefore, SEM is a suitable method to test a theoretically derived model (Hair et al., 2006). Structural relations can be modeled pictorially to enable a more precise conceptualization of the theory under study by SEM (Byrne, 2001). The hypothesized model can be tested statistically in a real-time analysis to determine the level which is consistent with the data. If the goodness of fit is adequate, the model argues for the possibility of hypothesized relations among variables. If it is inadequate, the tenability of such relations is rejected (Byrne, 2001).

The observed variables used to predict the latent variables in SEM were obtained by processing the data in the instrument. Three types of goodness-of-fit measurements were examined: absolute fit indices, incremental fit indices, and parsimonious fit indices.

Absolute fit indices are the measurement that provides the most fundamental indication of how well the proposed theory fits the data. This measures the suitability of the previous model in terms of how it fits with the sample data (McDonald and Ho, 2002) and helps to determine if the proposed model has a better fit. To assess the absolute fit of the model, this study used the likelihood ratio Chi-square (χ²), goodness-of-fit index (GFI) and the root mean square error of approximation (RMSEA). Prior research states that RMSEA values less than 0.08 represent reasonable fit (Browne and Cudeck, 1993).

Incremental fit indices are also recommended (Hoyle, 1995; Hu and Bentler, 1998) to assess model fit. These indices compare the chi-square value of the proposed model with the baseline model (McDonald and Ho, 2002). The comparative fit index (CFI) developed by Bentler (Bentler, 1990) is one such index and assesses the improvement to fit of the hypothesized model to the null model.

Finally, parsimonious fit measures are used to diagnose the goodness-of-fit that has been achieved by analyzing the number of coefficients in the model (Mulaik et al., 1989). Table 10 presents the goodness-of-fit measures obtained for this model and illustrates how they compare to the recommended value.

Table 10.

Goodness-of-fit of the model.

Measurement	Results	Recommended Values^a
Absolute fit indices
Chi-square /d.f. (CMIN/DF)	2.13	1.0 – 3.0
Goodness of fit index (GFI)	0.84	> 0.9
Adjusted goodness of fit index (AGFI)	0.82	> 0.9
Root mean square error of approximation (RMSEA)	0.05	< 0.08
Incremental fit indices
Comparative fit index (CFI)	0.90	> 0.9
Normed fit index (NFI)	0.83	> 0.9
Parsimonious fit indices
Parsimony goodness of fit index (PGFI)	0.70	As high as possible
Parsimonious Normed Fit Index (PNFI)	0.74	As high as possible

Notes: ^a Reference recommended values from Browne and Cudeck (1993), Kline (1998), DeCoster (1998), Sharma et al. (2005), Hair (2006), Hooper et al. (2008), and Byrne (2010).

The overall model fit appears to be quite good and acceptable. The absolute fit of the model was assessed by using the CMIN/DF, GFI, AGFI, and RMSEA. The CMIN/DF value of 2.13 was found to be within the recommended range of 1.0-3.0, indicating a reasonable model fit (Segars and Grover, 1993). The GFI, AGFI and RMSEA indices values were 0.84, 0.82 and 0.05 respectively. Among these, the GFI and AGFI values were below the commonly accepted threshold level. However, the values can fall outside recommended values with larger samples (Sharma et al., 2005; Hooper et al., 2008). The CFI was 0.90, which is within the recommended level. The PGFI was 0.70 and PNFI was 0.74. Though the results of the parsimony fit index values are considerably lower than other goodness of fit indices, a result over .50 is acceptable when the model achieves a values over 0.90 on other goodness of fit indices (Mulaik et al., 1989; Hooper et al., 2008). From the statistically derived results, it is shown that the proposed model is acceptable.

The hypotheses were estimated by using the critical ratio (C.R.) that represents the parameter estimate divided by its standard error. It operates as a z-statistic in testing that the estimate is statistically different from zero. Based on a level of confidence at 0.05, the critical ratio of the test must exceed ±1.96 in order for the relationship to be significant (Byrne, 2001; Ullman and Bentler, 2003). The statistically significant relationships among the study variables and hypothesis are shown in Figure 3 and Table 11 respectively.

Figure 3.

Results of the model analysis using AMOS.

Table 11.

Result of hypothesis testing.

Path	Hypothesis	Estimate	C.R.^*	P	Result
Levels of ITO → ITO success	H1	0.40	2.59^*	.00	Support
Transaction costs → Levels of ITO	H2	-0.16	2.11^*	.00	Support
Specific assets → Levels of ITO	H3	0.18	2.10^*	.03	Not Support
Internal uncertainty → Levels of ITO	H4	0.01	0.02	.98	Not Support
Internal IT skills → Levels of ITO	H5	0.02	0.42	.67	Not Support

^* C.R. = Critical Ratio (C.R. > ±1.96).

Discussion

In summary, this article achieved its two main objectives. The finding confirms the positive relationship between the levels of ITO and ITO success, which is the primary objective of the study, and determined whether the drivers of the levels of ITO are the same as those in prior research, which is the secondary objective.

Figure 3 presents the research results from the SEM analysis. Three links were found to be statistically significant (C.R. > ±1.96). Firstly, the link between the levels of ITO and ITO success is statistically significant at p < 0.05 (with C.R. = 2.59). The result supports the hypothesis that the levels of ITO have a positive impact on ITO success (H1). Secondly, the link from levels of transaction costs to the levels of ITO is statistically significant at p < 0.05 (with C.R. = 2.11). The result supports the hypothesis that the levels of transaction costs have a negative impact on the levels of ITO (H2). Thirdly, the link from levels of specific assets to the levels of ITO is statistically significant at p < 0.05 (with C.R. = 2.10) but in the opposite direction from the hypothesis (H3). Lastly, two results in H4 and H5 revealed that levels of internal uncertainty and levels of internal IT skills do not have any statistically significant impacts on the levels of ITO.

As hypothesized in the main objective, the study confirmed a positive relationship between the levels of ITO and ITO success, which has not been confirmed before in previous literature. The levels of ITO are measured by percentage of IT activities and IT components to be outsourced (Gonzalez et al., 2005a; Pati and Desai, 2005).

Organizations must understand the levels of IT activities and components that should be outsourced. The appropriate levels of ITO could increase ITO success as it involves increased cost efficiency in IT investments and control and decreased total costs of IT and obsolescence risk. On the other hand, an excessive level of ITO could result in decreased ITO success as it leads to unnecessary coordination and costs.

IT activities consist of the maintenance, operation, development and design of technology (Österle and Winter, 2003). IT components include IT infrastructure, applications, tasks and processes (Zarnekow et al., 2004). It is also found that there are statistically significant positive links between IT activities and components, and levels of ITO, which indicate that organizations should consider the levels of their IT activities and components when deciding to outsource.

The most commonly used measurement of ITO success in the literature has three components: strategic benefits, economic benefits, and technical benefits (Grover et al., 1996; Lee and Kim, 1999; Lee, 2001; Jiang et al., 2006; Gottschalk and Solli-Sæther, 2006; Dahlberg and Lahdelma, 2007). Based on these three measurements, the survey results indicated strong and statistically significant ITO success shown in Figure 3.

For example, in the sense of strategic benefits, organizations could have more time to focus or concentrate on their core business and gain better performance and quality in business/IT process. Moreover, they can increase user satisfaction and productivity. Likewise, for economic benefits, total costs of the organizations’ IT could decrease and IT expenditures could be more effectively controlled in order to increase IT cost efficiency. With respect to technical benefits, organizations could increase technical know-how, IT capabilities, IT qualities (infrastructure/software/ application) while reducing the risk of IT obsolescence.

The findings based on the second objective found three results that differ from previous studies. Firstly, this study found that the link between the levels of specific assets and the levels of ITO is statistically significant but in the opposite direction from prior research. Prior research indicated that activities involving more specific assets were less likely to be outsourced (Anderson, 1988; Aubert et al., 2004). In contrast, the survey results indicate a significant positive linkage between levels of specific assets and levels of ITO. From the results, it can be implied that more differentiation on human resources and technical specificity, and greater investment from both the client and supplier will lead to greater levels of ITO. This result also suggests that firms generally choose to outsource non-core activities. Furthermore, it suggests that the activities associated with the technical skills may be carried out more effectively by outsourcing services (Aubert et al., 2004). Thus, more specific assets tend to increase the levels of ITO in the financial sector.

Secondly, the results indicate that the levels of internal uncertainty do not influence the levels of ITO. Contrary to prior research, the results suggest that firms are not much concerned about lack of information before outsourcing. These results are reinforced by the expert interviews, as their opinions showed that most decisions came from top management and were based on organizational policies, rather than actual research around IT processes, which is often performed at the department levels. Lacity and Hirschheim (1993) stated that political motivations play a significant role in the ITO decision. The interviews indicated that top management is largely driven by a number of political motives, wherein they react to the efficiency needs, and outsourcing media reports, and perceived industry trends, rather than their individual organizational IT process needs. Thus, in the case of Thailand, the uncertainty issues are found to be not statistically significant.

Thirdly, contrary to previous research, the results did not confirm a statistically significant relationship between the levels of internal IT skills and the levels of ITO. Our expert interview results indicate that most technology is constantly evolving and organizations tend to upgrade their IT platforms to keep up with the changing market. Most organizations rely on expertise from external suppliers because they do not want to invest in the education and training of new technologies (Aubert et al., 2004). Therefore, the levels of internal IT skills could have less impact on the levels of ITO.

However, our study is consistent with previous studies in that a decrease in transaction costs tends to support the increase in the levels of ITO. In a study of the US banking industry, Ang and Straub (1998) strongly support the effect of transaction costs on levels of ITO. The transaction costs analysis usually occurs before organizations decide to outsource. Transaction costs are costs incurred while gathering information during the planning stage, and continue to rise until the outsourcing process is complete. Most organizations are faced with the problems of hidden costs from the outsourcing contract (Ang and Straub, 1998; Tafti, 2005; Gewald and Dibbern, 2009). Based on the survey results and the interviews of IT experts, most organizations in Thailand are skeptical of the returns they would get from outsourcing. The overall results and implications of the study are summarized in Table 12.

Table 12.

Summary of findings.

Hypothesis	Result	Implication
H1: The levels of ITO have a positive impact on ITO success	Supported^*	The levels of ITO have impacts on the outcome of the ITO success. Organizations must understand what percentages of activities and components of IT should be outsourced (New finding)
H2: The levels of transaction costs have a negative impact on the levels of ITO	Supported^*	The levels of transaction costs have a negative impact on the levels of ITO. Only internal resources that have low transaction costs should be selected to be outsourced. (Consistent with previous studies)
H3: The levels of a specific assets have a negative impact on the levels of ITO	Not supported	The link between the levels of specific assets and the levels of ITO is significant but in the opposite direction. More specific assets tend to increase the levels of ITO. (Contrast to previous studies)
H4: The levels of internal uncertainty have a negative impact on the levels of ITO	Not supported	The study finds that the levels of internal uncertainty have no impacts on the levels of ITO. Most decisions came from top management and were based on organizational policy. (Contrast to previous studies)
H5: The levels of internal IT skills have a negative impact on the levels of ITO	Not supported	The study finds that the levels of internal IT skills do not have impact on the levels of ITO. Our expert interviews supported that most technology is fast changing and many IT platforms are moving to common standards. Therefore, the internal IT skills will have less impact in the future. (Contrast to previous studies)

*support at critical ratio (CR > ±196).

Implications

ITO is an organizational strategy that can be applied to many business environments but it is difficult for top management to find a balance between outsourcing and insourcing. Determining the levels of ITO can be a challenge because it is difficult to predict the return. Thus, determining whether or not ITO adds value to the organization is an important management question. Researchers may use the measurements and results proposed in this study to conduct further investigation with other types of organizations or in other countries.

From a practical perspective, the findings contribute to a more thorough understanding of the ITO phenomenon for top management in an organization. ITO allows organizations to take advantage of knowledge and skills that can be obtained and transferred from the outsourcers. However, the appropriate levels of ITO have important implications for the organization (Gulla and Gupta, 2012). If organizations have high levels of ITO, it brings the risk of overdependence on outside vendors and IT security concerns. Conversely, if they decide to use low levels of ITO, it leads to poor utilization of IT resources. It is therefore necessary to understand the process of ITO before making a decision to outsource.

To illustrate this point, the top five commercial banks in Thailand have adopted different ITO procedural approaches. Bank A relied completely on internal resources. They tried to have their own staff taking care of all IT functions. Their priority was data security and did not consider any other IT issues. This approach leads to this bank’s IT functions becoming dated relative to its competitors. Bank B takes the opposite approach by outsourcing nearly all IT processes. They decided to adopt ITO, despite a lack of both ITO knowledge and ITO experience, and faced lock-in problems. This means the bank depends solely on the particular vendor’s competency to operate and is unable to find other vendors that have a similar competency. The bank has experienced limited success because of the risk of high costs associated with switching vendors.

The remaining three banks, Bank C, Bank D, and Bank E considered different levels of ITO. Each bank studied their internal IT activities and components, to see if IT activities and components are involved with their organizations’ strategies and how much should be outsourced. The activities and components that require business skills should be kept in-house. Moreover, they set specific goals and objectives for each outsourcing decision. They effectively indicated the levels of ITO that later influenced ITO success. Therefore, their success supported sustainable IT growth.

Limitations

This study provides empirical evidence from the context of a developing country, which fills a significant gap in the literature. The research found a significant relationship between the levels of ITO and ITO success. Furthermore, this study has supported the extension and further development of ITO research models.

Although the response rate among financial institutions was high and the sample distribution came close to representing Thai commercial banks and financial institutions, the samples did not represent all institution types. This is due to the fact that there are many financial institution types in Thailand, including foreign bank branches, foreign bank representatives and personal loan companies. This study focuses only on Thai-owned financial institutions, consisting of private and public financial institutions. The results may not be directly applicable to the remaining foreign financial institutions or representatives of other industries. Additionally, the result may not be applicable to other countries. Thus, replicating this model across countries and industries would help confirming and generalizing the findings.

Conclusion

This study empirically investigates the relationship between the levels of ITO and ITO success. Based on a review of the literature, expert interviews and surveys conducted, the qualitative results were confirmed by a quantitative analysis. The result fills a gap in the literature by confirming the relationship between the levels of ITO and ITO success. Moreover, the researcher proposes a model that includes the drivers of levels of ITO. Transaction costs, specific assets, percentage of IT activities and IT components all play a role in determining the levels of ITO. However, internal uncertainty and internal IT skills were not found to have significant impacts in determining the levels of ITO, and these two results differ from prior research. It is possible for future IT researchers to explore the levels of ITO and ITO success in other industries or across countries.

Footnotes

Appendix

References

Albright

Park

(2009) Confirmatory Factor Analysis using Amos, LISREL, Mplus, SAS/STAT CALIS. The Trustees of Indiana University 1: 1–85.

Anderson

(1988) Transaction costs as determinants of opportunism in integrated and independent sales forces. Journal of Economic Behavior & Organization 9: 247–264.

Ang

Cummings

(1997) Strategic response to institutional influences on information systems outsourcing. Organization Science 8: 235–256.

Ang

Straub

(1998) Production and transaction economies and IS outsourcing: a study of the US banking industry. MIS Quarterly 22: 535–552.

Aubert

Rivard

Patry

(1996) Development of measures to assess dimensions of IS operation transactions. Omega 24: 661–680.

Aubert

Rivard

Patry

(2004) A transaction cost model of IT outsourcing. Information & Management 41: 921–932.

Bahli

Rivard

(2005) Validating measures of information technology outsourcing risk factors. Omega 33: 175–187.

Baldwin

Irani

Love

(2001) Outsourcing information systems: drawing lessons from a banking case study. European Journal of Information Systems 10: 15–24.

Bank of Thailand (2013) Bank of Thailand: financial institutions. Available at: http://wwwbotorth/English/FinancialInstitutions/WebsiteFI/Pages/instListaspx

10.

Bentler

(1990) Comparative fit indexes in structural models. Psychological Bulletin 107: 238.

11.

Bies

Zacharia

(2007) Medical tourism: Outsourcing surgery. Mathematical and Computer Modelling 46: 1144–1159.

12.

Britz

Tramacere

Blackmore

. (2012) Market share analysis: IT outsourcing services, worldwide, 2011, 30 April 2012. Available at: http://wwwgartnercom/it/pagejsp?id=2021215

13.

Brown

(2006) Confirmatory factor analysis for applied research: Guilford Press.

14.

Browne

Cudeck

(1993) Alternative ways of assessing model fit. CA: Sage Publications, Inc.

15.

Byrne

(2001) Structural equation modeling with AMOS: Basic concepts, applications, and programming. Mahwah, NJ: Lawrence Erlbaum Associates Publishers.

16.

Byrne

(2010) Structural equation modeling with AMOS: Basic concepts, applications and programming, NY: Taylor & Francis Publishers.

17.

Chiasakul

(2004) Production networks, trade and investment policies, and regional cooperation in Asia: a case study of automotive industry in Thailand. Manila, Angelo King Institute for Economic and Business Studies - De La Salle University. 46p.

18.

Dahlberg

Lahdelma

(2007) IT governance maturity and IT outsourcing degree: an exploratory study. HICSS 2007. 40th Annual Hawaii International Conference IEEE, 236a–236a.

19.

Dahlberg

Nyrhinen

Santonen

(2006) The success of selective and total outsourcing of firm-wide IT-infrastructure: an empirical evaluation. In Proc. ECIS, 2006. pp. 835–846.

20.

DeCoster

(1998) Overview of factor analysis. Retrieved May 24: 2006.

21.

Earl

(1996) The risks of outsourcing IT. Sloan Management Review 37: 26–32.

22.

Fowler

Jeffs

(1998) Examining information systems outsourcing: a case study from the United Kingdom. Journal of Information Technology 13: 111–126.

23.

Gewald

Dibbern

(2009) Risks and benefits of business process outsourcing: A study of transaction services in the German banking industry. Information & Management 46: 249–257.

24.

Gonzalez

Gasco

Llopis

(2005a) Information systems outsourcing reasons in the largest Spanish firms. International Journal of Information Management 25: 117–136.

25.

Gonzalez

Gasco

Llopis

(2005b) Information systems outsourcing success factors: a review and some results. Information Management & Computer Security 13: 399–418.

26.

Goo

Kishore

Rao

(2000) A content-analytic longitudinal study of the drivers for information technology and systems outsourcing. ICIS '00 Proceedings of the twenty first International Conference on Information Systems. Association for Information Systems, 601–611.

27.

Goolsby

Furniss

Whitlow

(2002) An assessment of the outsourcing market in the banking sector of Japan’s financial institutions industry. Executive paper, Everest Group.

28.

Gottschalk

Solli-Sæther

(2005) Critical success factors from IT outsourcing theories: an empirical study. Industrial Management & Data systems 105: 685–702.

29.

Gottschalk

Solli-Sæther

(2006) Maturity model for IT outsourcing relationships. Industrial Management & Data systems 106: 200–212.

30.

Grossman

Helpman

(2002) Integration versus outsourcing in industry equilibrium. The Quarterly Journal of Economics 117: 85–120.

31.

Grover

Cheon

Teng

JTC

(1996) The effect of service quality and partnership on the outsourcing of information systems functions. Journal of Management Information Systems 12: 89–116.

32.

Gulla

Gupta

(2012) Impact of information systems outsourcing: a study of Indian banking sector. International Journal of Business Information Systems 10: 131–150.

33.

Hair

Black

Babin

. (2006) Multivariate data analysis Sixth ed.: New Jersey: Prentice-Hall.

34.

Hamzah

Aman

Maelah

. (2010) Outsourcing decision processes: A case study of a Malaysian firm. African Journal of Business Management 4: 3307–3314.

35.

Han

Lee

Seo

(2008) Analyzing the impact of a firm's capability on outsourcing success: A process perspective. Information & Management 45: 31–42.

36.

Heide

(1994) Interorganizational governance in marketing channels. The Journal of Marketing: 71–85.

37.

Hirschheim

Lacity

(2000) The myths and realities of information technology insourcing. Communications of the ACM 43: 99–107.

38.

Hooper

Coughlan

Mullen

(2008) Structural equation modelling: guidelines for determining model fit. Articles: 2.

39.

Hoyle

(1995) Structural equation modeling: Concepts, issues, and applications: Sage Publications, Incorporated.

40.

L-t

Bentler

(1998) Fit indices in covariance structure modeling: Sensitivity to underparameterized model misspecification. Psychological Methods 3: 424.

41.

Jiang

Frazier

Prater

(2006) Outsourcing effects on firms' operational performance: An empirical study. International Journal of Operations & Production Management 26: 1280–1300.

42.

Kaplan

Norton

(2004) Strategy maps: Converting intangible assets into tangible outcomes: Harvard Business Press.

43.

Karyda

Mitrou

Quirchmayr

(2006) A framework for outsourcing IS/IT security services. Information Management & Computer Security 14: 402–415.

44.

Kess

Phusavat

Torkko

. (2008) External knowledge: sharing and transfer for outsourcing relationships. International Journal of Business and Systems Research 2: 196–213.

45.

Kitcharoen

(2003) The potential of logistics outsourcing industry in Thailand. MBA thesis, Assumption University, Bangkok. 182 p.

46.

Kline

(1998) Software review: software programs for Structural Equation Modeling: Amos, EQS, and LISREL. Journal of Psychoeducational Assessment 16: 343–364.

47.

Kline

(2010) Principles and practice of structural equation modeling: The Guilford Press.

48.

Lacity

Hirschheim

(1993) Information systems outsourcing: Myths, metaphors and realities: Wiley.

49.

Lacity

Willcocks

(1998) An empirical investigation of information technology sourcing practices: Lessons from experience. MIS Quarterly: 363–408.

50.

Lacity

Willcocks

(2000) Survey of IT outsourcing experiences in US and UK organizations. Journal of Global Information Management (JGIM) 8: 5–23.

51.

Lacity

Willcocks

Feeny

(1995) IT sourcing: maximize flexibility and control. Long Range Planning 28: 128–128.

52.

Lacity

Willcocks

Feeny

(1996) The value of selective IT sourcing. Sloan Management Review 37: 13–25.

53.

Lee

(2001) The impact of knowledge sharing, organizational capability and partnership quality on IS outsourcing success. Information & Management 38: 323–335.

54.

Lee

Kim

(1999) Effect of partnership quality on IS outsourcing success: conceptual framework and empirical validation. Journal of Management Information Systems: 29–61.

55.

Lee

Miranda

Kim

(2004) IT outsourcing strategies: Universalistic, contingency, and configurational explanations of success. Information Systems Research 15: 110–131.

56.

Lin

Pervan

McDermid

(2007) Issues and recommendations in evaluating and managing the benefits of public sector IS/IT outsourcing. Information Technology & People 20: 161–183.

57.

Loh

Venkatraman

(1992) Determinants of information technology outsourcing: a cross-sectional analysis. Journal of Management Information Systems: 7–24.

58.

Mahoney

Pandian

(1992) The resource–based view within the conversation of strategic management. Strategic Management Journal 13: 363–380.

59.

Malik

Nilakant

(2011) Extending the “size matters” debate: Drivers of training in three business process outsourcing SMEs in India. Management Research Review 34: 111–132.

60.

Marcolin

McLellan

(1998) Effective IT outsourcing arrangements. Published by the IEEE Computer Society, 654.

61.

Marlowe

Sullivan

(2007) Medical tourism: the ultimate outsourcing. Hum Res Plan 30: 8–10.

62.

McDonald

MHR

(2002) Principles and practice in reporting structural equation analyses. Psychological Methods 7: 64.

63.

Medlin

Schneberger

Hunsinger

(2007) Perceived technical information technology skill demands versus advertised skill demands: An empirical study. Journal of Information Technology Management 18: 14–23.

64.

Miozzo

Grimshaw

(2005) Modularity and innovation in knowledge-intensive business services: IT outsourcing in Germany and the UK. Research Policy 34: 1419–1439.

65.

Mulaik

James

Van Alstine

. (1989) Evaluation of goodness-of-fit indices for structural equation models. Psychological Bulletin 105: 430.

66.

Nam

Rajagopalan

Rao

. (1996) A two-level investigation of information systems outsourcing. Communications of the ACM 39: 36–44.

67.

NECTEC (2010) Exceutive summary: Thailand ICT Market & Industry 2009. National Electronics and Computer Technology Center Thailand.

68.

Ngwenyama

Bryson

(1999) Making the information systems outsourcing decision: A transaction cost approach to analyzing outsourcing decision problems. European Journal of Operational Research 115: 351–367.

69.

Österle

Winter

(2003) Business Engineering: Auf dem Weg zum Unternehmen des Informationszeitalters. Business Engineering Auf dem Weg zum Unternehmen des Informationszeitalters 2.

70.

Pati

Desai

(2005) Conceptualizing strategic issues in information technology outsourcing. Information Management & Computer Security 13: 281–296.

71.

Poppo

Zenger

(2002) Do formal contracts and relational governance function as substitutes or complements?. Strategic Management Journal 23: 707–725.

72.

Porter

Millar

(1985) How information gives you competitive advantage. Harvard Business Review 63: 149–160.

73.

Robertson

Gatignon

(1998) Technology development mode: a transaction cost conceptualization. Strategic Management Journal 19: 515–531.

74.

Rochester

Douglas

(1993) New wrinkles in IS outsourcing. IS Analyzer 31: 1–18.

75.

Rottman

Lacity

(2009) A US client's learning from outsourcing IT work offshore. Information Systems Outsourcing: 443–469.

76.

Santos

Eisenhardt

(2005) Organizational boundaries and theories of organization. Organization Science: 491–508.

77.

Seddon

Cullen

Willcocks

(2002) Does Domberger's theory of the contracting organization explain satisfaction with IT outsourcing?' 23rd International Conference on Information Systems.

78.

Segars

Grover

(1993) Re-examining perceived ease of use and usefulness: a confirmatory factor analysis. MIS Quarterly 17: 517–525.

79.

Sharma

Mukherjee

Kumar

. (2005) A simulation study to investigate the use of cutoff values for assessing model fit in covariance structure models. Journal of Business Research, 58: 935–943.

80.

Shi

(2010) The role of IS architecture planning in enhancing IS outsourcing's impact on IS performance: Its antecedents and an empirical test. Journal of Enterprise Information Management 23: 439–465.

81.

Sufian

Habibullah

(2010) Developments in the efficiency of the Thailand banking sector: a DEA approach. International Journal of Development Issues 9: 226–245.

82.

Tafti

MHA

(2005) Risks factors associated with offshore IT outsourcing. Industrial Management & Data Systems 105: 549–560.

83.

Thangavelu

Chongvilaivan

(2011) The impact of material and service outsourcing on employment in Thailand's manufacturing industries. Applied Economics 43: 3931–3944.

84.

Thompson

Daniel

(1996) Factor analytic evidence for the construct validity of scores: A historical overview and some guidelines. Educational and Psychological Measurement 56: 197–208.

85.

Udo

(2000) Using analytic hierarchy process to analyze the information technology outsourcing decision. Industrial Management & Data Systems 100: 421–429.

86.

Ullman

Bentler

(2003) Structural equation modeling. Handbook of psychology. New York: Wiley.

87.

Wang

Yang

(2007) Using a hybrid multi-criteria decision aid method for information systems outsourcing. Computers & Operations Research 34: 3691–3700.

88.

Watjatrakul

(2005) Determinants of IS sourcing decisions: A comparative study of transaction cost theory versus the resource-based view. The Journal of Strategic Information Systems 14: 389–415.

89.

Whitten

Leidner

(2006) Bringing IT back: An analysis of the decision to backsource or switch vendors. Decision Sciences 37: 605–621.

90.

Willcocks

Lacity

Kern

(1999) Risk mitigation in IT outsourcing strategy revisited: longitudinal case research at LISA. The Journal of Strategic Information Systems 8: 285–314.

91.

Willcocks

Lacity

(1998) Strategic sourcing of information systems: Wiley.

92.

Williamson

(1981) The economics of organization: The transaction cost approach. American Journal of Sociology: 548–577.

93.

Williamson

(1985) The economic institutions of capitalism: Firms, markets, relational contracting: The Free Press.

94.

Zarnekow

Scheeg

Brenner

(2004) A survey of the life cycle costs of IT application systems. Wirtschaftsinformatik 46: 181–187.