In addition to evaluating the credibility of sources, it is important to evaluate articles by reading critically. Reading critically involves drilling deeper into the subject matter, going beyond the written text to contextualize, question, and reflect upon what you’ve just read.Reading Help (Links to an external site.)Links to an external site.For this assignment, find two articles from the School Global Library(please see articles below) and two articles from sources outside of the library that relate to the topic and purpose you’ve identified for your Portfolio Project. For each article, address the following.
- What is the general topic and main issue or problem?
- What are the author’s statements based on (e.g., fact, opinion)?
- What conclusions were reached and how?
- Identify 5-6 key words that resonate throughout the article. Are those words objective, subjective, or emotional?
- Reflect on the article’s suitability for use in your final Portfolio Project.
Your well-written paper should meet the following requirements:
- Be 4-5 pages in length, not including title page and reference page.
- Include a minimum of four scholarly sources
- Include in-text citations and a reference page as required by APA
Write clearly and logically. You will be graded on content, analysis, and your adherence to the tenets of good academic writing, which should be succinct where possible while also exploring the topics appropriately.
Part 2 Assignment but not to be conjoined with the above. Must be done separately
Using the four articles you selected for this week’s Critical Thinking Assignment, prepare an annotated bibliography using APA format as shown in the sample annotated bibliography (Links to an external site.)Links to an external site.. The Online Research and Writing Lab (Links to an external site.)Links to an external site. is also a good resource. Each of the four annotations should include the full APA reference of the article and a paragraph that provides:
- an overview or description of the content,
- a critical analysis of the article,
- and an evaluation of the relevance or usefulness of the article for your research proposal topic.
Submit your completed annotated bibliography to your instructor for feedback.
Understanding knowledge management capability in business process outsourcing: A cluster analysis
– The purpose of this paper is to investigate trends in the dimensions of low, medium, and high knowledge management (KM) capability of business process outsourcing (BPO) firms. It also explores the trends in BPO performance with different levels of KM capabilities of BPO firms. Moreover, the study determines how firm characteristics, such as size, age, industry, and outsourcing age, affect KM capability.
– A survey was employed to collect data on managers from 605 firms.K -means cluster analysis was performed on the aggregate measures of the four KM capability dimensions and BPO performance to reveal trends. Subsequently, MANOVA was used to evaluate the effects of four firm characteristics on KM capability, and individual ANOVA tests were performed to examine the specific differences among the four dimensions.
– Among the four dimensions of KM capability, knowledge application is the most significant. Knowledge protection is the second highest in terms of expressing the profile for low KM capability firms, but the lowest among the four dimensions of KM capability for medium and high KM capability firms. Each dimension of KM capability affects BPO performance positively. Firm size, age, industry, and outsourcing age differentially affect the dimensions of KM capability.
– This study presents a theoretical model of firm characteristics, KM capability, and BPO performance. Through the model, ideas are offered: firms with high KM capability significantly differ from those with low and medium KM capabilities; different firms exhibit different KM capabilities; developing knowledge application capability should be the priority in managing BPO; and improving KM capability is an effective means to enhance BPO performance.
Business process outsourcing (BPO) plays an increasingly significant role in service outsourcing because of its high value added. The 2012 global service outsourcing development report revealed that the BPO market size in 2011 alone was valued at 176 billion dollars. The compound annual growth rate of BPO market size has exceeded that of information technology outsourcing (Zheng, 2013). However, the outcome of BPO implementation has not been optimistic. According to a survey of 189 BPO enterprise clients conducted by HfS Research, approximately 50 percent of the clients claimed that BPO failed to reduce the costs stipulated in the contract, offer knowledge of specific industry processes, and provide talent that adds value beyond standard operations (Fersht, 2014).
The proponents of knowledge management (KM) argue that the failure of BPO can be attributed to the low KM capability of outsourcers based on the concept of BPO as a process of creating, transferring, integrating, and using knowledge (Willcocks et al. , 2004; Blumenberg et al. , 2009; Beverakis et al. , 2009). Numerous studies have examined the types and effectiveness of KM capability within an organization (Tanriverdi, 2005; Wang et al. , 2007), but minimal research (e.g. Christopher and Tanwar, 2012) has attempted to categorize KM capability and investigate its effect in the context of BPO in at least two organizations. Understanding the dimensions of KM capability and the patterns or trends that these dimensions may follow in different types of organizations is critical because managers can formulate appropriate KM strategies in BPO by taking advantage of potentially high KM capability to improve BPO performance. In addition, previous studies have presented contrasting findings on the relationship between KM capability and firm performance. Although some studies have found that KM capability affects firm performance directly and significantly (Tanriverdi, 2005), other studies have determined that their relationship is insignificant and mediated by innovation or organizational learning processes (Lee and Sukoco, 2007; Easterby-Smith and Prieto, 2008). Investigating how KM capability dimensions change performance is significant because understanding this issue enables managers to identify the most effective KM capabilities in managing BPO. Therefore, the present study investigates trends in KM capability dimensions across low, medium, and high KM capabilities of BPO firms. Moreover, it explores the trends in BPO performance with different levels of KM capabilities of BPO firms. Finally, this study determines the effects of firm characteristics, such as size, age, industry, and outsourcing age, on organizational KM capability in BPO. BPO performance relies on the extent to which the client integrates and uses knowledge. Thus, we focus on the perspective of the clients.
BPO is a mechanism of designating business processes to a service provider that possesses, manages, and administrates selected IT-intensive processes by adopting predefined and measurable metrics (Gartner, 2013). The major area of BPO comprises procurement, finance and accounting, training, human resource, and customer relationship management. The two types of organizations involved in a BPO process are the client and service provider (Malik et al. , 2012). Clients should articulate their requirements to allow the service provider to appropriately manage the outsourced business processes (Mann et al. , 2011). Clients should also acquire, integrate, and use knowledge generated by the service provider to ensure that the BPO is well executed and delivered with high performance (Narayanan et al. , 2011). Therefore, greater KM capability is likely to engender high BPO performance.
Literature defines KM capability as the ability to organize, shift, configure, and arrange knowledge-based resources to achieve the goals of and gain business values from the organization from a knowledge-based view (Kearns and Lederer, 2003; Chuang, 2004). Among the different dimensions of KM capability that have been proposed, two categorizations are noteworthy. On the one hand, Gold et al. (2001) categorized KM capability from the infrastructure view and developed three dimensions of capabilities, namely, technology, structure, and culture. These three variables can be measured by 37 items. Although this categorization was adopted by several studies, KM literature provides no evidence with regard to its widespread use. On the other hand, KM capability was categorized from the process view, a categorization that has been broadly accepted, as proven by KM literature. For example, Tanriverdi (2005) categorized KM capability into knowledge creation, transfer, integration, and leverage provision. Tseng and Lee (2014) posited that KM capability consisted of knowledge transfer and integration. Gold et al. (2001) also examined this variable and categorized it as knowledge acquisition, conversion, application, and protection capabilities. Table I describes the four dimensions. We adopted the categorization from the process view of Gold et al. (2001) because it can accurately capture the nature of KM capability. Moreover, this categorization is widely used as research basis.
Few attempts have been made to understand how KM capability dimensions vary in different kinds of BPO firms. Although previous research has proposed a BPO framework that integrates specific KM practices within each phase of BPO (Mahmoodzadeh et al. , 2009), KM capability has not been explicitly categorized and investigated. The specification of KM capability enables managers to formulate an appropriate strategy to manage knowledge in BPO. Managerial measures can be derived by investigating the differences between BPO firms with high, medium, and low KM capabilities. The nature of KM can be better understood through the insights into the trends of KM capability. The current research addresses this issue by adopting the instrument of Gold et al. (2001) to gather data from numerous BPO firms. High, medium, and low KM capabilities were identified and used to determine whether patterns across the categories and among the dimensions of KM capability exist.
In addition, evidence has shown that organizational performance can be enhanced by KM capability, but the findings on the relationship between these two factors in previous studies are contradictory (Alavi and Leidner, 2001; Lee and Sukoco, 2007). Some researchers contend that KM capability affects effectiveness significantly and directly. Alavi and Leidner (2001) surveyed 300 senior managers and found that knowledge infrastructure and knowledge process capabilities have significant effects on organizational effectiveness. Tanriverdi (2005) also empirically found that KM capability enhances corporate performance based on the data from 1,000 of the 250 Fortune firms. However, other researchers point out that KM capability affects performance indirectly. Innovation and agility can be mediators. Lee and Sukoco (2007) found that innovation mediates the relationship between KM capability and effectiveness. Easterby-Smith and Prieto (2008) proposed that KM capability affects organizational effectiveness indirectly and through organizational learning process and dynamic capability. Therefore, gaps exist among previous studies with regard to the relationship between KM capability and performance. Further examination of this issue is necessary because the process through which KM capability changes performance can be deeply understood and the relative importance of each KM capability dimension can also be investigated specifically. In this study, we investigate the trends of BPO performance with the change of KM capability to understand the effectiveness of the latter.
Firm characteristics also influence the level of KM capability. In the present research, four firm characteristics, namely, size, age, industry, and outsourcing age, were examined. The former three are the basic characteristics of a firm, whereas the latter one is a significant characteristic of BPO firms. These four factors were selected despite the lack of empirical evidence to support this claim because previous studies have suggested that these affect KM capability (Cui et al. , 2005).
Firm size is the first characteristic that we examined. Cui et al. (2005) proposed that firm size may affect KM capability because a large firm requires high KM capability to enhance its organizational performance. We chose the number of employees to measure firm size because it can appropriately reflect firm size (Lu and Ramamurthy, 2011).
Firm age is the second characteristic that we examined. Old and young firms exhibit different organizational cultures and structures (Kale and Karaman, 2011) that can change the KM process. Nevertheless, firm age affects KM capability both positively and negatively. For instance, young firms may have low knowledge protection capability but high knowledge acquisition capability, whereas old firms have high knowledge protection capability but low knowledge acquisition capability. Young firms may not have complete knowledge protection systems, but these easily accept new knowledge.
Firm industry is the third characteristic examined. Information intensity changes across different industries (Mao et al. , 2014). Thus, information intensity requires different KM capabilities to process information.
Outsourcing age is the fourth characteristic examined. Firms with higher outsourcing experience accumulate richer KM experience in BPO (Rustagi et al. , 2008). Therefore, KM capability varies according to the different outsourcing experiences of the firm.
In sum, better KM strategies to manage BPO can be proposed and developed by understanding common patterns exhibited across KM capability dimensions and how these characteristics affect each dimension of KM capability. The dimensions presented in Table I are used in this study so that the following research questions can be addressed:
RQ1. Which trends in the dimensions of KM capability and BPO performance can be detected across BPO firms with high, medium and low KM capabilities?
RQ2. How do firm size, age, industry, and outsourcing age affect the KM capability of BPO firms?
- Research methodology
3.1. Data collection
A survey was employed to gather data from a number of firms to address the research questions. Managers from the client side who are in charge of BPO business in their firms were identified as ideal subjects, because they were familiar with the situation of BPO in their firms and were also in a good position to report KM capability. We obtained a list of Chinese firms that experienced BPO as clients with the assistance of an outsourcing association in China. Subsequently, we contacted the senior executives in these firms to identify managers who are responsible for BPO management. These managers were contacted to confirm their participation in the survey.
In 2013, we distributed approximately 1,000 questionnaires to the managers. The managers were requested to evaluate the firm characteristics, KM capability, and BPO performance of their firms. A total of 605 usable questionnaires were returned after three months (60.5 percent response rate). The age of respondents ranged from 24 to 59 years (mean=39.4). Previous experience of managing BPO ranged from two to 26 years (mean=12.2). The characteristics of the firms the managers worked for are shown in Table II.
We examined external validity by evaluating non-response bias. A t -test was performed on the latent variables between the questionnaires obtained in the early and late stages of the study (Liu and Wang, 2014). The significant levels of knowledge acquisition, conversion, application, protection, as well as BPO performance, are p =0.36, 0.58, 0.27, 0.65, and 0.31, respectively. Therefore, no significant difference exists between the two samples and non-response bias is also non-existent. We also checked for common method bias by performing Harmon’s single-factor test (Podsakoff et al. , 2003). Exploratory factor analysis was conducted with all latent variables. The results show that no single factor in the samples accounts for the majority of the covariance (maximum <24 percent). Thus, common method bias does not threaten our sample.
3.2. Constructs and measures
Each KM capability dimension includes multiple-item measures, as shown in Appendix. These measures were adapted from Gold et al. (2001). The measures of BPO performance were also included in the instrument. BPO performance is the success of BPO processes and outcomes that the firm undertakes. Four measures were adopted from (2010), as shown in Appendix. Seven-point Likert scale that ranges from “strongly disagree” to “strongly agree” was used to measure the items. In addition, firm size was assessed through the number of employees in the firm (Table II). In particular, respondents were requested to indicate the number of employees in their firms as [< or =, slant]100, 101-500, 501-1,000, or >1,000. Seven levels, namely, 1-5, 6-10, 11-15, 16-20, 20-25, 26-30, and >30 years, were used to measure firm age. Firm industry was identified using ten classifications, namely, information technology, business trading, finance, retailing, manufacturing, utilities, agriculture, engineering, education and training, and others. Lastly, the outsourcing age of the firm was measured with the same seven levels used to measure firm age.
3.3. Measurement evaluation
Measurement validity was assessed by employing several tests. We first evaluated convergent validity and internal consistency. The constructs and their measurement properties, as well as the number of measures of each construct, are shown in Table III. The values of the average variance extracted (AVE) are all higher than 0.5 (Fornell and Larcker, 1981). The reliability was checked by evaluating construct reliability (CR) and Cronbach’s [alpha] for the constructs of KM capability dimension and BPO performance. Table III shows that the Cronbach’s [alpha] and CRs of KM capability dimensions and performance construct are >0.7, which indicates adequate reliability (Nunnally and Bernstein, 1994). Factor loadings (FLs) of each scale are presented in Appendix. The FL of each item is significant ( p <0.000) and exceeds 0.7. Therefore, convergent validity has passed for the constructs and measures. Afterward, discriminant validity was evaluated to check if the construct measured by each scale is significantly different from the others (Liu, 2013). Table IV shows that the correlation between each pair of latent variables is lower than the square root of AVE. Thus, our model also passed this test. Lastly, we conducted confirmatory factor analysis (CFA) on the five latent variables with AMOS 17.0 to check whether good fit exists between the observed data and the measurement model (Yan and Dooley, 2013). The CFA results indicate that [chi] 2 is 859.56 and is significant ( p =0.000). The degree of freedom (df) is 374. The ratio of [chi] 2 to the df is 2.30, which is under the threshold of 3. Moreover, the root mean square error of approximation is 0.048 and the standardized root mean square residual is 0.0495, which are both lower than 0.08 (Hulland et al. , 1996). In addition, the normed fit index is 0.91, the Tucker-Lewis fit index is 0.90, the comparative fit index is 0.92, and the incremental fit index (IFI) is 0.91, which are all higher than 0.90 (Hulland et al. , 1996). The aforementioned goodness-of-fit indices demonstrate good fit between the data and the measurement model. Therefore, the overall measurement model displays satisfactory measurement properties.
3.3. Cluster analysis
We performed K -means cluster analysis on the aggregate measures of four KM capability dimensions. The clusters were subsequently obtained, representing that the number of firms with low, medium, and high KM capabilities was 110, 275, and 220, respectively.
4.1. Trends in the dimensions of KM capability across clusters
The cluster means for four dimensions of KM capability obtained by K -means cluster analysis are shown in Table V. The higher the cluster means of each dimension, the greater the level of KM capability. The profiles of firms with low, medium, and high KM capabilities are shown as a star chart (Figure 1).
Results show potential trends in the dimensions of KM capability in BPO. First, the mean level of the KM capability of each dension increases significantly as the cluster moves from low to medium level and from medium to high level. This observation has an intuitive appeal and provides further empirical support for the validation of the KM capability dimensions identified by Gold et al. (2001). Second, the use of seven-point scale to indicate the aggregate level of each KM capability dimension reveals that on average, the low KM capability cluster consists of firms with low KM capability (between 2.38 and 3.23) along with three of the four KM capability dimensions. Firms with low KM capability have moderate level of knowledge application capability as indicated by the value of 4.03 that exceeds the midpoint of the seven-point scale. However, the knowledge application capability of the medium KM capability cluster is also higher (5.39) than those of other dimensions with values from 4.43 to 4.63. Knowledge conversion and acquisition capabilities are the second and third highest with respect to describing the profile of firms with medium KM capability. For the cluster of firms with high KM capability, knowledge application capability remains the most prominent dimension of KM capability. The overall results indicate that knowledge application capability is the most significant among the four dimensions of KM capability. Therefore, the effective application of knowledge is critical in managing BPO. Interestingly, knowledge protection capability is the second highest in terms of expressing the profile of firms with low KM capability, but the lowest among the four dimensions of KM capability in firms with medium and high KM capabilities. Thus, protecting knowledge is a necessary requirement although only minimal attention should be paid to it in managing BPO.
We observe another trend in the relationship between the levels of KM capability and BPO performance. This relationship is clearly presented in Table VI. As firms with low KM capability move toward high KM capability, BPO performance significantly increases. Firms with low KM capability exhibit low performance below three on the seven-point scale. Moreover, firms with medium KM capability display medium performance (4.38), and firms with high KM capability show high BPO performance (5.76). Therefore, KM capability has a significant and positive effect on performance in BPO.
4.2. Effect of firm size, age, industry, and outsourcing age on KM capability
We also examined the effect of firm size, age, industry, and outsourcing age on KM capability to address the second research question. The mean values for firm size for low, medium, and high KM capabilities are 3.94, 2.81, and 1.56, respectively. Surprisingly, as firm size increases, the KM capability of the firm decreases significantly. MANOVA was performed to evaluate the effects of firm size on the four dimensions of KM capability. A significant level of the overall model was obtained ( p =0.000 for Hotelling’s and Pillai’s trace statistics). Four individual ANOVA tests were conducted to investigate the specific differences among the four dimensions of KM capability. The results show that three of the four dimensions of KM capability (i.e. knowledge conversion, application, and protection) are significantly affected by firm size ( p =0.002, 0.028, and 0.000), whereas the other dimension (knowledge acquisition) is unaffected ( p =0.092); larger firms may have a larger amount of knowledge to manage, which makes converting, applying, and protecting knowledge difficult.
The relationship between firm age and KM capability dimensions was also examined. The mean values of firm age for low, medium, and high KM capabilities are 6.50, 5.43, and 2.22, respectively. Similar to firm size, the KM capability of these firms decreases significantly as firm age increases. MANOVA was performed to evaluate the effects of firm age on the four dimensions of KM capability. The overall model is significant ( p =0.000 for Hotelling’s and Pillai’s trace statistics) and the results of the four individual ANOVA tests also reveal significant relationships ( p =0.000, 0.000, 0.001, and 0.000). Old firms may accumulate considerable knowledge, but may also lack an effective KM process to utilize this knowledge.
The relationship between firm industry and KM capability dimensions was subsequently examined. MANOVA analysis reveals that the model obtained a significant level ( p =0.000 for Hotelling’s and Pillai’s trace statistics). The four ANOVA tests show that all relationships are significant ( p =0.000 for each dimension). Therefore, as firm industry varies, the overall KM capability and all its four dimensions also differ in BPO.
Finally, we examined the effect of outsourcing age on the KM capability of firms. MANOVA analysis reveals that the model obtained a significant level ( p =0.000 for Hotelling’s and Pillai’s trace statistics). Nevertheless, when the specific differences in the levels of KM capability were analyzed deeply by individual ANOVA tests, the results suggest that the relationships of outsourcing age with knowledge conversion and application are significant ( p =0.028 and 0.000), but the relationships of outsourcing age with knowledge acquisition and protection are insignificant ( p =0.283 and 0.218). Therefore, as firms gain richer outsourcing experience, the knowledge conversion and application capabilities become higher in relation to managing BPO.
4.3. Theoretical model of firm characteristics, KM capability, and performance
The above results indicate that a model of KM capability and BPO performance can be proposed and developed (Figure 2). The results of cluster analysis indicate a pattern where each dimension of KM capability move toward the same direction as KM capability changes from low to high level. A positive relationship between KM capability and BPO performance is also revealed. Further analysis indicates that firm industry and age influence all four dimensions of KM capability, whereas firm size and outsourcing age affect some specific dimensions of KM capability.
- Discussions and implications
5.1. Implications for research
This study is one of the first attempts to integrate KM capability, performance, and firm characteristics in the context of BPO. A number of theoretical implications can be obtained from our results.
First, the results have shown the cluster means of each dimension of KM capability across firms with low, medium, and high KM capabilities. This can provide a basis for the evaluation of a firm’s KM capability when engaging BPO. If the KM capability of the firm is low, the firm should strengthen its KM capability to effectively manage BPO. Furthermore, we find that different KM capability dimensions exhibit various levels across low, medium, and high KM capabilities. This observation indicates that managing BPO requires different levels of KM capabilities. A unified KM practice is unnecessary in managing BPO. This finding is consistent with the argument of Mahmoodzadeh et al. (2009) that KM practices should differ across BPO processes.
Second, our results reveal the relative importance of each KM capability dimension in managing BPO. In particular, high knowledge application capability is the most important aspect in managing BPO, whereas knowledge protection capability appears to receive minimal emphasis because its scores are the lowest in both medium and high KM capabilities. Previous research has only compared the significance of KM capability and other capabilities (Mao et al. , 2014), but failed to focus on the specific dimensions of KM capability. This finding adds new knowledge to literature because no study has yet explicitly analyzed the importance of KM capability dimensions.
Third, the results reveal that each dimension of KM capability has a positive effect on BPO performance. This finding provides additional evidence on the significant effect of KM capability on performance. Moreover, it supports previous result that KM capability is positively and directly associated with performance (Tanriverdi, 2005). Therefore, improving KM capabilities is an effective means to enhance BPO performance. Such capabilities can produce high competitiveness among firms because other firms find them difficult to imitate.
Fourth, firm size, age, industry, and outsourcing age differentially affect the dimensions of KM capability. Although some characteristics (e.g. firm age and industry) significantly influence all four dimensions of KM capability, the relationship between certain characteristics (e.g. outsourcing age) and KM capability dimensions (e.g., knowledge protection) were insignificant. Moreover, the effect of different firm characteristics on KM capabilities was identified as either positive or negative. Therefore, the relationship between firm characteristics and KM capability dimensions exhibits mixed results. Firm size has a negative effect on knowledge conversion, application, and protection capabilities; meanwhile, it affects knowledge acquisition insignificantly. This finding supports the proposition of Cui et al. (2005), who argued that firm size significantly influences KM capability. However, not all KM capability dimensions are affected by firm size (e.g. knowledge acquisition). Large firms are not required to acquire excess new knowledge, but tend to create new knowledge based on existing knowledge. Moreover, large firms also display lower KM capability. Therefore, managers in large firms should recognize this disadvantage and focus on constructing more effective KM capability to manage BPO. Firm age has a negative effect on each dimension of KM capability. As old firms exhibit lower agility (Mao et al. , 2014), which is positively related to KM capability, these manifest deterioration of KM capability. Each dimension of KM capability varies across industries. The intensity of knowledge changes according to the different industry settings. Firms classified under industries with high intensive knowledge should construct greater KM capability to deal with knowledge effectively. Finally, outsourcing age positively affects knowledge conversion and application capabilities. Higher outsourcing experience can strengthen the conversion and application of knowledge in BPO. However, outsourcing age insignificantly affects knowledge acquisition and protection. Therefore, firms with low outsourcing experience should effectively leverage their knowledge acquisition and protection capabilities because these two capabilities do not significantly differ between firms with young and old outsourcing ages.
5.2. Implications for practice
Several managerial implications can be drawn from our findings. First, given that different levels of each KM capability dimension are required to manage BPO, investments on their development should be unequal. Knowledge application capability is the most important dimension for managing BPO. Thus, establishing knowledge application capability should be prioritized. Knowledge conversion and acquisition capabilities should also receive adequate investments and should be fairly developed. However, managers should not overlook knowledge protection capability because it still exhibits high cluster means in low KM capability. Managers must selectively exchange information or knowledge with external providers, as well as predefine scopes and rules to clarify which type of knowledge should be protected. Second, the abilities to acquire, convert, apply, and protect knowledge are critical to improve BPO performance. Therefore, managers can optimize KM structure; develop the culture to acquire, convert, apply, and protect knowledge; and reward employees who exhibit behaviors and activities that manage knowledge efficiently. Members of organizations or groups should also positively and autonomously communicate with one another, as well as acquire and process information from external providers, partners, and markets. Third, different types of firms should develop various KM strategies. Given that large and old firms display low KM capability in managing BPO, they should develop an effective culture with rules and an enhanced infrastructure for KM. They can also divide large groups into small teams such that information and knowledge can be exchanged and used efficiently. Given that firms with young outsourcing age exhibit low knowledge conversion and application capabilities, these two KM capabilities should be considerably developed among them. Firms can strengthen capabilities to integrate different types of knowledge and learn to effectively accumulate and use such knowledge to address problems in BPO.
This study pioneered in revealing the trends of KM capability in BPO and in developing a theoretical model of firm characteristics, KM capability and BPO performance. Empirical evidence has shown that firms with high KM capability significantly differ from those with low and medium KM capabilities. Knowledge application is the most prominent capability of firms with low to high KM capabilities. However, knowledge acquisition, conversion, application, and protection, affect BPO performance positively and significantly. In addition, firm characteristics influence the dimensions of KM capability.
- Limitations and directions for future research
The present study has several limitations. First, KM capability is a complex variable, and some of its aspects may not have been captured. Second, this study adopted convenience samples. Therefore, generalizability of the results should be further investigated. Third, four firm characteristics were selected because these have been found to influence KM capability. However, other characteristics (e.g. firm structure) may also affect the dimensions.
Future research can employ the theoretical model developed in the present study to examine whether each dimension of KM capability complements or substitutes one another in the process of affecting BPO performance. Such examination is particularly significant in designing effective KM strategies and portfolios in BPO. In addition, KM capability positively affects BPO performance. In this regard, understanding the potential moderators that can change the relationship between KM capability and performance is an interesting proposition. Finally, KM capability is affected not only by firm characteristics, but also by other factors, such as resources. Therefore, future research can explore more factors that can improve KM capability.
In my firm, we are able to apply knowledge learned from mistakes and experiences (FL=0.92).
In my firm, we are able to use knowledge to solve new problems (FL=0.87).
In my firm, we have the ability to use knowledge in the development of new products/services (FL=0.88).
In my firm, we use knowledge to improve efficiency (FL=0.86).
In my firm, we are able to quickly apply knowledge to critical competitive needs (FL=0.89).
In my firm, we are able to convert knowledge into the design of new products/services (FL=0.94).
In my firm, we are able to transfer organizational knowledge to individuals (FL=0.83).
In my firm, we have the ability to absorb knowledge from individuals and service providers (FL=0.89).
In my firm, we are able to integrate different sources and types of knowledge (FL=0.87).
About the authors
Dr Shan Liu is an Assistant Professor at Economics and Management School in the Wuhan University. His research interests focus on mobile commerce and IT project management with particular emphasis on software risk management. He has published more than ten refereed publications including papers that have appeared or been accepted in Information Systems Journal , European Journal of Information Systems , Management Decision , International Journal of Project Management , Information Development and International Journal of Medical Informatics .
Dr Zhaohua Deng is an Associate Professor of Medical Information Management at the Huazhong University of Science and Technology. Her research focusses on mobile business and mobile health. Her research has appeared in Information Systems Journal , International Journal of Information Management ,International Journal of Medical Informatics , Electronic Markets , International Journal of Mobile Communications , International Journal of Services Technology and Management and International Journal of Information Technology and Management . Dr Zhaohua Deng is the corresponding author and can be contacted at: firstname.lastname@example.org
This work was supported by grants from the National Natural Science Foundation of China (Nos 71101060, 71471141, and 71201063). Shan Liu appreciates Zhaohua Deng to work as the corresponding author.
Alavi, M. and Leidner, D.E. ( 2001 ), ” Review: knowledge management and knowledge management systems: conceptual foundations and research issues “, MIS Quarterly , Vol. 25 No. 1, pp. 107 – 136 .
Beverakis, G., Dick, G.N. and Kecmanovic, D.C. ( 2009 ), ” Taking information systems business process outsourcing offshore: the conflict of competition and risk “, Journal of Global Information Management , Vol. 17 No. 1, pp. 32 – 48 .
Blumenberg, S., Wagner, H.T. and Beimborn, D. ( 2009 ), ” Knowledge transfer processes in IT outsourcing relationships and their impact on shared knowledge and outsourcing performance “, International Journal of Information Management , Vol. 29 No. 5, pp. 342 – 352 .
Christopher, D. and Tanwar, A. ( 2012 ), ” Knowledge management in outsourcing environment: people empowering people “, IUP Journal of Knowledge Management , Vol. 10 No. 2, pp. 61 – 86 .
Chuang, S. ( 2004 ), ” A resource-based perspective on knowledge management capability and competitive advantage: an empirical investigation “, Expert Systems with Applications , Vol. 27 No. 3, pp. 459 – 465 .
Cui, A.S., Griffith, D.A. and Cavusgil, S.T. ( 2005 ), ” The influence of competitive intensity and market dynamism on knowledge management capabilities of multinational corporation subsidiaries “, Journal of International Marketing , Vol. 13 No. 3, pp. 32 – 53 .
Easterby-Smith, M. and Prieto, I.M. ( 2008 ), ” Dynamic capabilities and knowledge management: an integrative role for learning? “, British Journal of Management, Vol. 19 No. 3, pp. 235 – 249 .
Fersht, P. ( 2014 ), “BPO will continue to fail miserably without a mindset to embrace change, develop talent and tech-enable processes”, available at: www.enterpriseirregulars.com/72079/bpo-will-continue-fail-miserably-without-mindset-embrace-change-develop-talent-tech-enable-processes / (accessed April 2, 2014).
Fornell, C. and Larcker, D.F. ( 1981 ), ” Evaluating structural equation models with unobservable variables and measurement error “, Journal of Marketing Research, Vol. 18 No. 1, pp. 39 – 50 .
Gartner. ( 2013 ), “Business process outsourcing (BPO)”, available at: www.gartner.com/it-glossary/business-process-outsourcing-bpo (accessed January 6, 2014).
Gold, A.H., Malhotra, A. and Segars, A.H. ( 2001 ), ” Knowledge management: an organizational capabilities perspective “, Journal of Management Information Systems , Vol. 18 No. 1, pp. 185 – 214 .
Hulland, J., Chow, Y.H. and Lam, S. ( 1996 ), ” Use of causal models in marketing research: a review “, International Journal of Research in Marketing , Vol. 13 No. 2, pp. 181 – 197 .
Kale, S. and Karaman, E.A. ( 2011 ), ” Evaluating the knowledge management practices of construction firms by using importance – comparative performance analysis maps “, Journal of Construction Engineering and Management , Vol. 137 No. 12, pp. 1142 – 1152 .
Kearns, G.S. and Lederer, A.L. ( 2003 ), ” A resource-based view of strategic IT alignment: how knowledge sharing creates competitive advantage “, Decision Sciences , Vol. 34 No. 1, pp. 1 – 29 .
Lee, L.T. and Sukoco, B.M. ( 2007 ), ” The effects of entrepreneurial orientation and knowledge management capability on organizational effectiveness in Taiwan: the moderating role of social capital “, International Journal of Management , Vol. 24 No. 3, pp. 549 – 572 .
Liu, S. ( 2013 ), ” How team risk and planning and control risk moderate the effects of clan and self control on the process performance of IT projects: the perspective of user liaisons “, Information Development , pp. 1 – 13, in press, doi:10.1177/0266666913501682.
Liu, S. and Wang, L. ( 2014 ), ” User liaisons’ perspective on behavior and outcome control in IT projects: role of IT experience, behavior observability, and outcome measurability “, Management Decision , Vol. 52 No. 6, pp. 1148 – 1173 .
Lu, Y. and Ramamurthy, K. ( 2011 ), ” Understanding the link between information technology capability and organizational agility: an empirical examination “, MIS Quarterly , Vol. 35 No. 4, pp. 931 – 954 .
Mahmoodzadeh, E., Jalalinia, Sh. and Yazdi, F.N. ( 2009 ), ” A business process outsourcing framework based on business process management and knowledge management “, Business Process Management Journal , Vol. 15 No. 6, pp. 845 – 86 .
Malik, A., Sinha, A. and Blumenfeld, S. ( 2012 ), ” Role of quality management capabilities in developing market-based organisational learning capabilities: case study evidence from four Indian business process outsourcing firms “, Industrial Marketing Management , Vol. 41 No. 4, pp. 639 – 648 .
Mani, D., Barua, A. and Whinston, A. ( 2010 ), ” An empirical analysis of the impact of information capabilities design on business process outsourcing performance “, MIS Quarterly , Vol. 34 No. 1, pp. 39 – 62 .
Mann, A., Kauffman, R.J., Han, K. and Nault, B.R. ( 2011 ), ” Are there contagion effects in information technology and business process outsourcing? “, Decision Support Systems , Vol. 51 No. 4, pp. 864 – 874 .
Mao, H., Liu, S. and Zhang, J. ( 2014 ), ” How the effects of IT and knowledge capability on organizational agility are contingent on environmental uncertainty and information intensity “, Information Development , pp. 1 – 25, doi:10.1177/0266666913518059.
Narayanan, S., Jayaraman, V., Luo, Y. and Swaminathan, J.M. ( 2011 ), ” The antecedents of process integration in business process outsourcing and its effect on firm performance “, Journal of Operations Management , Vol. 29 Nos 1-2, pp. 3 – 16 .
Nunnally, J.C. and Bernstein, I.H. ( 1994 ), Psychometric Theory , McGraw-Hill, New York, NY .
Podsakoff, P.M., MacKenzie, S.B., Lee, J.Y. and Podsakoff, N.P. ( 2003 ), ” Common method biases in behavioral research: a critical review of the literature and recommended remedies “, Journal of Applied Psychology , Vol. 88 No. 5, pp. 879 – 903 .
Rustagi, S., King, W.R. and Kirsch, L.J. ( 2008 ), ” Predictors of formal control usage in IT outsourcing partnerships “, Information Systems Research , Vol. 19 No. 2, pp. 126 – 143 .
Tanriverdi, H. ( 2005 ), ” Information technology relatedness, knowledge management capability, and performance of multi-business firms “, MIS Quarterly , Vol. 29 No. 2, pp. 311 – 334 .
Tseng, S.M. and Lee, P.S. ( 2014 ), ” The effect of knowledge management capability and dynamic capability on organizational performance “, Journal of Enterprise Information Management , Vol. 27 No. 2, pp. 158 – 179 .
Wang, E., Klein, G. and Jiang, J.J. ( 2007 ), ” IT support in manufacturing firms for a knowledge management dynamic capability link to performance “,International Journal of Production Research , Vol. 45 No. 11, pp. 2419 – 2434 .
Willcocks, L., Hindle, J., Feeny, D. and Lacity, M. ( 2004 ), ” IT and business process outsourcing: the knowledge potential “, Information Systems Management , Vol. 21 No. 3, pp. 7 – 15 .
Yan, T. and Dooley, K.J. ( 2013 ), ” Communication intensity, goal congruence, and uncertainty in buyer-supplier new product development “, Journal of Operations Management , Vol. 31 Nos 7-8, pp. 523 – 542 .
Zheng, X. ( 2013 ), “Global service outsourcing report”, available at: http://tradeinservices.mofcom.gov.cn/i/2013-01-28/93824.shtml (accessed September 2, 2014).
Shan Liu School of Economics and Management, Wuhan University, Wuhan, China
Zhaohua Deng School of Medicine and Health Management, Huazhong University of Science & Technology, Wuhan, China
Word count: 6505
© Emerald Group Publishing Limited 2015
Jenni, Corinne.Alliant International University, ProQuest Dissertations Publishing, 2015. 3701192.
Akbari, Naser; Ghaffari, Ali.Journal of Knowledge Management; Kempston Vol. 21, Iss. 5, (2017): 1120-1141.
Dearstyne, Bruce W.Information Management Journal; Lenexa Vol. 38, Iss. 2, (Mar/Apr 2004): 28-36.
The Role of Management in International Business Theory1: A Meta-Analysis and Integration of the Literature on International Business and International Management
Buckley, Peter J.Management International Review, suppl. Special Issue 1; Wiesbaden Vol. 36, (1996): 7-54.
IT capability, uncertainty and organizational performance: Development of measures and empirical examination
Lu, Ying.The University of Wisconsin – Milwaukee, ProQuest Dissertations Publishing, 2006. 3222385.
The Trouble with Offshoring: Static and Dynamic Losses in the Presence of Unemployment.
Brecher, Richard A.1
World Economy. Jan2013, Vol. 36 Issue 1, p1-11. 13p. 4 Diagrams.
*Outsourcing & the economy
*Business process outsourcing
523930 Investment Advice
523999 Miscellaneous Financial Investment Activities
This study provides theoretical support for the popular objection to offshoring, whereby firms at home employ services of labour located abroad. In the presence of unemployment, our analysis highlights welfare losses from offshoring – not only for the static case of a fixed stock of capital, but also for the dynamic one of optimal saving and investment. We compare these static and dynamic losses to the gains that would instead arise under full-employment conditions, assumed by most of the theoretical literature on offshoring. Our results suggest that public concerns over offshoring are justified when unemployment is taken explicitly into account. [ABSTRACT FROM AUTHOR]
Copyright of World Economy is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder’s express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
1Department of Economics, Carleton University, Ottawa, ON, Canada
Full Text Word Count:
Choose Language الإنجليزية/العربية английски език/български 英语/简体中文 英語/繁體中文 angličtina/čeština Engelsk/dansk Engels/Nederlands Anglais/Français Englisch/Deutsch Αγγλικά/Ελληνικά English/Hausa אנגלית/עברית अंग्रेज़ी/हिंदी angol/magyar Inggris/bahasa Indonesia Inglese/Italiano 英語/日本語 영어/한국어 Engelsk/Norsk انگليسی/فارسی angielski/polski Inglés/Português English/Pashto Engleză/română Английский/Русский Inglés/Español English/Serbian Angleški/Slovenski Angličtina/Slovenčina Engelska/svenska อังกฤษ/ไทย İngilizce/Türk Англійська/Українська انگریزی/اردو
The Trouble with Offshoring: Static and Dynamic Losses in the Presence of Unemployment.
- Static Analysis
- Full Employment
- Dynamic Analysis
- Full Employment
Listen American Accent Australian Accent British Accent
This study provides theoretical support for the popular objection to offshoring, whereby firms at home employ services of labour located abroad. In the presence of unemployment, our analysis highlights welfare losses from offshoring – not only for the static case of a fixed stock of capital, but also for the dynamic one of optimal saving and investment. We compare these static and dynamic losses to the gains that would instead arise under full employment conditions, assumed by most of the theoretical literature on offshoring. Our results suggest that public concerns over offshoring are justified when unemployment is taken explicitly into account.
The present study provides theoretical support for the popular objection to offshoring, whereby firms at home employ services of labour located abroad.1 This objection is based largely on well known fears of rising unemployment, caused by the transfer of domestic jobs to offshore workers. Clearly, such fears cannot be evaluated under the traditional full employment assumption, adopted by most of the pertinent literature, including the prominent work by Bhagwati et al. (2004). Moreover, it is not obvious that offshoring will be harmful whenever job loss is taken properly into account. Indeed, in the small number of existing models of offshoring with unemployment, the use of offshore labour typically has uncertain implications for employment and/or welfare, as in Davidson et al. (2008), Egger and Kreickemeier (2008), Keuschnigg and Ribi (2009), Brecher and Chen (2010, 2012), Koskela and Stenbacka (2010), Mitra and Ranjan (2010), Kohler and Wrona (2011) and Zhang (2011).
Is it possible, nevertheless, to develop a coherent theoretical argument that provides a clear cut rationale for the popular opposition to offshoring? This question is important because unless the opponents’ logic can be unambiguously formalised in a model, economists will tend to remain skeptical and dismissive of the anti offshoring position. The development and analysis of such a model will thus help to polarise and hence invigorate the professional debate over the employment and welfare consequences of offshoring, while permitting a scientific assessment of the competing assumptions involved.
In the spirit of fostering this debate, we construct a model in which offshoring definitely reduces the home levels of aggregate employment and national welfare. Our construction starts with Bhagwati et al.’s (2004) single good model 1, which provides the clearest statement of the pro offshoring position based on full employment.2 Then, to allow for the presence of persistent unemployment, we depart from them by making a polar opposite assumption about the labour market. In particular, rather than maintaining their flexible wage assumption, our framework assumes instead that the economy is subject to downward inflexibility of the real wage.3
We address not only the static case – without capital accumulation, as in the above cited work – but also the dynamic one with capital investment determined by optimal savings. For both of these cases, our analysis demonstrates that offshoring unambiguously reduces the home levels of employment and welfare. We also compare the static and dynamic losses of welfare to the gains that would occur instead within a full employment version of our model. These results suggest that explicit consideration of unemployment is critical for understanding the welfare consequences of offshoring.
Section 2 analyses the static case, which provides a straightforward demonstration of our basic insight, regarding the importance of wage rigidity/flexibility for an offshoring country. This insight is further developed in Section 3, which broadens the theory of offshoring to cover the dynamic case, for both flexible and rigid wage economies. To assess the robustness of our results, Section 4 briefly examines various extensions of the basic framework. Section 5 concludes with a summary of the paper’s main contributions.
For simplicity, let the world produce only a single homogeneous product. This simplification removes any incentive for international trade in goods,4 thereby allowing us to concentrate on the employment and welfare implications attributable directly to offshoring per se. Suppose that the preferences of home consumers can be aggregated into the utility function U(C) of a representative agent, where C denotes the aggregate level of consumption and U′ >0 > U′′ (for positive but diminishing marginal utility). The representative agent is endowed with one unit of labour, which is in perfectly inelastic supply.5 When the real wage is downwardly rigid, the supply of labour may exceed its demand, in which case there is involuntary unemployment. To be consistent with the representative agent framework, assume that this unemployment is equally shared among all home workers (e.g. via job sharing or income sharing within large families).
Home output is given by F(K, L), where F is a neoclassical production function characterised by constant returns to scale, positive but diminishing marginal products, and the Inada conditions; K denotes the fixed aggregate stock of fully utilised capital; and L stands for total employment, comprising N (≤1) units of home labour plus S units of foreign labour. Since all foreign workers remain physically located abroad, S represents the level of offshoring.
The incentive to offshore arises because the wage is higher at home than abroad.6 To simplify the analysis, assume initially (until section 4) that S is exogenously fixed by a binding ceiling, say due to government policy or technological considerations. Also suppose at first (here and in Section 3) that offshore labour used in home production receives the same wage w as home workers, which would be the case if (for example) S is subject to a quota given without charge to foreign workers.7 Thus, national income is F(K, N + S) − wS. Since perfect competition ensures zero profits, an equivalent expression for national income is rK + wN, where r denotes the rental rate of capital. According to the usual marginal productivity conditions for profit maximisation, r = FK(K, N + S) and w = FL(K, N + S), where subscripts of functions indicate partial differentiation (e.g. FK ≡ ∂F/∂K).
In this static model, the entire national income is spent on the single consumption good, which means that C = F(K, N + S) − wS. Because the marginal utility of consumption is positive, home welfare is a monotonically increasing function of national income.
To establish a benchmark for our results with rigid wage unemployment, first review the standard flexible wage full employment case.8 Since the supply of home labour is perfectly inelastic, N (like K) is constant in this case. Then, in the event of a small increase in S, the resulting change in national income is d[F(K, N + S) − FL(K, N + S)S]/dS = −SFLL ≥ 0 as S ≥ 0, where FLL < 0 by diminishing marginal productivity. Thus, more offshoring is better than less. For example, starting from autarky (with S = 0), the introduction of offshoring (S > 0) increases home welfare. In terms of Figure 1, as total employment increases from OA units (= N) to OB units (= N + S), output rises by area ABCD, while offshore wages rise by the smaller area ABCE, leaving a net gain equal to area CDE.
Next, turn to the polar opposite case in which w is exogenously fixed by a binding floor , set by some institutional arrangement (such as government regulation, labour unions or social custom).9 Because of this floor, home labour is partially unemployed. Given that the total endowment of labour is normalised to 1, N is both the rate and quantity of home employment. (Similarly, 1 − N represents both the rate and level of home unemployment.) Note also that N = K/k − S, where k ≡ K/(N + S). Since this capital/labour ratio is uniquely determined by the given (under constant returns to scale), an increase in S causes an equal decrease in N, leaving N + S unchanged. Thus, , which implies that more offshoring is now worse than less, in contrast to the full employment case in section 2a. In particular, the offshoring equilibrium is inferior to autarky for the home country, given the presence of unemployment due to a downwardly rigid wage.10
In terms of Figure 2, the fixed wage () equals OE units, yielding OA units of employment in autarky. When AB units of offshoring are introduced, home employment (N) falls by the same amount to OB units, leaving total employment (N + S) and output unchanged at their autarkic levels. However, part of this constant output must now be used to pay offshore wages in the amount of area ABCD, which represents the loss in national income and (hence) welfare.
Our above analysis follows the previous literature in abstracting from the possibility of capital accumulation induced by offshoring. Such accumulation would tend to raise the demand for labour, with important implications for the steady state levels of employment and welfare. These implications are addressed in the present section, which also derives some interesting results about the transitional path to the new equilibrium after an increase in offshoring.
In our dynamic model, the representative agent maximises (the present discounted value of lifetime utility), subject to the wealth accumulation constraint , and the initial condition on wealth (assets); where t is a (continuous) time index, which is suppressed whenever not needed for clarity; ρ represents the fixed rate of time preference and X stands for the stock of wealth. The agent takes r, w, and N as given. (The value of N is either identical to 1 under full employment or equal to whatever firms demand in the case of unemployment.) Thus, C is the single control variable of the agent, whose only state variable is X.
The current value Hamiltonian for this optimal control problem is H = U(C) + μ(rX + wN − C), where μ(t) is a co state variable, which represents the shadow price of wealth. The necessary conditions for a maximum include the following equations (in addition to the constraint as well as the initial and transversality conditions):
where θ ≡ U′′(C)/U′′(C) < 0.
Under full employment, with the wage adjusting to clear the labour market, the dynamic system of the economy is given by the following two conditions:
The first of these equations follows directly from the agent’s wealth accumulation constraint, after we set X ≡ K (and N ≡ 1), note that F(K, 1 + S) rK + w(1 + S), and replace w by the marginal product of labour. To obtain ( 5) from ( 3), simply replace r by the marginal product of capital.
In steady state equilibrium,. Then, ( 5) implies that FK(K, 1 + S) = ρ, which uniquely determines K (given the exogenously specified S). Thus, in Figure 3– the phase diagram for the dynamic system ( 4) and ( 5)– the schedule for is a horizontal line, at a height equal to the steady state value of K. As the horizontal arrows show, a higher (lower) value of K would make the optimal. negative (positive), by lowering (raising) FK via diminishing marginal productivity. (Recall that θ < 0.) On the other hand, the (generally nonlinear) curve for has a slope of 1/(FK − SFLK), which is positive because we assume that the denominator of this expression is greater than zero. (In other words, a rise in K raises national income, although this assumption could be relaxed without qualitatively changing our main results.) Thus, the vertical arrows indicate that a rise or fall in K (given C) would make the optimal positive or negative, respectively. It is clear from these (vertical and horizontal) arrows that the steady state equilibrium, denoted by point A, is saddle path stable. Given any initial value of K, C jumps to the saddle path (generally nonlinear), and the economy follows this path to point A.
In Figure 3, a small rise in S would cause the line for to shift up by –FKL/FKK (> 0). At the same time, the curve for. would shift down by SFLL/(FK − SFLK) at each C. Thus, an expansion of offshoring causes the following changes to the steady state equilibrium: a rise in K, an increase in C, and hence an improvement in welfare. As we now show, moreover, welfare also improves even when the transition to the new equilibrium is taken into account.
For example, suppose that the economy is initially in steady state with S = 0 (autarky) and is suddenly confronted with S > 0 (offshoring). Because national income instantaneously rises (in accordance with section 2), it is feasible to raise C by the same amount, thereby keeping. Maintaining consumption at this new elevated level, moreover, would leave the capital stock constant. This feasible path is clearly a welfare improving one, since it provides more consumption than autarky at each point in time. Nevertheless, such a path (with constant C) is inconsistent with optimality condition ( 5), which requires instead that , because the rise in S (with constant K) raises FK(K, 1 + S). Thus, as the economy follows instead the optimal (saddle) path to the new steady state equilibrium, the welfare gain is even greater.
Turning now to the unemployment case, assume that the wage floor is indexed in terms of utility. To be more specific, suppose that wU′(C) ≥ ω, where ω (the floor) is a positive constant. According to this inequality, the wage multiplied by the marginal utility of consumption – to get the real wage in terms of utility – cannot fall below some minimum value (ω).11 Setting this floor high enough to be binding, rewrite the minimum wage constraint as
Under constant returns to scale, the capital/labour ratio k[≡K/(N + S)] is a monotonically increasing function of w. Thus, in the light of ( 6), , where.
where f(k) ≡ F(k,1) = F(K, L)/L (by constant returns to scale), f ′(k) = FK(K, L), and f ′′(k) = LFKK (<0 by diminishing marginal productivity). Equations ( 7) and ( 8) describe the dynamic system of the economy in the case of minimum wage unemployment.
Consider Figure 4, the phase diagram for the dynamic system ( 7) and ( 8). The schedule for is a vertical line, since there is a unique value of C that makes the right hand side of ( 8) equal to zero. The horizontal arrows point towards this line because. From ( 7), the slope of the curve for is ( , where ( 6) has been used to eliminate ω. The vertical arrows point away from this upward sloping curve, since. These arrows (for both curves) indicate that the steady state equilibrium is saddle path stable.
Let point B in Figure 4 correspond to the steady state with a given level of offshoring (S > 0). Since S does not enter ( 8), the line for would not be affected by a reduction in offshoring, say (without loss of generality) to the autarkic level (S = 0). However, this reduction would cause a downward shift in the curve for , because (at each C) dK/dS = wk/f (> 0) to keep the right hand side of ( 7) unchanged. Thus, the autarkic steady state equilibrium is at (say) point A, vertically below B.
Now, suppose that the economy begins in autarkic equilibrium at point A in Figure 4 and that offshoring is suddenly introduced. Instantaneously, the optimal C falls to point A′ on the saddle path, and then gradually returns to its initial level, as the economy moves along this (saddle) path to the new steady state equilibrium at B. Throughout the transition, C is lower than in autarky. Thus, offshoring leads to an overall (lifetime) loss in welfare.
Intuitively, at each point in time, the continuing use of offshore labour tends to depress national income, for reasons discussed in Section 2. The welfare impact of this negative shock can be mitigated through optimal saving and investment (which reduces the rise in unemployment), but not completely eliminated, since the resulting accumulation of capital comes at the expense of current consumption.
Offshoring also lowers the steady state equilibrium value of N, while raising N + S. To verify these employment effects, set in ( 7), fix C at its equilibrium level, replace by its equivalent N + S, and differentiate the resulting equation with respect to S. Thereby obtain dN/dS = w/f − 1 (<0 because marginal is less than average product), implying that d(N + S)/dS = w/f (> 0). Notably, the fall in N occurs despite the capital accumulation from point A to B in Figure 4.
The transitional paths of home and total employment are not so clear cut. At point A in Figure 4, the rise in S lowers N by the same amount to keep N + S constant (given the autarkic levels of C and K), with w (and hence k) determined by ( 6). Via the jump to point A′, N increases (at constant K) – thus raising N + S by the same amount – to keep ( 6) satisfied. (This increase may be more or less than the above decrease at point A.) Then, as the economy moves up the saddle path from point A′ to B, employment is stimulated by the rise in K but dampened by the rise in C (and hence k), which implies ambiguous changes in N and N + S.
On the other hand, the transitional path of the real wage in terms of output is well defined. As C falls instantaneously (from point A to A′ in Figure 4) and then returns gradually to its initial level (reached at point B), so does w, to keep ( 6) satisfied.
So far, we have used a very simple model to keep the analysis sharp and focused. To evaluate the robustness of our findings, this section briefly considers the following four extensions of the basic model.
The main results would still hold if offshore workers receive less than w per unit of services rendered to home firms.12 For this case, let such workers receive a wage equal to only αw, where 1 > α, which is a positive constant. Under this specification, national income becomes F(K, N + S) − αwS. Then, area ABCD in Figure 2 is replaced by a smaller (but still positive) rectangular area (whose height is instead of ), and the schedules in Figure 4 are qualitatively unchanged. The same comments apply to area ABCE in Figure 1 and the schedules in Figure 3.
Although the level of offshoring is taken to be exogenous, we could readily treat it as endogenously determined, without affecting our main results. More specifically, suppose that offshoring proceeds to the point at which w = w*(S), where w* gives the (wage plus transaction) cost per unit of offshore labour, as a strictly increasing function of S. This equation can be rewritten as ω/U′(C) = w* (S) under minimum wage constraint ( 6), in which case , with. For this specification, the curves and analysis in Figure 4 are qualitatively the same as before, except that comparative dynamics are now induced by a downward parametric shift in the w* function at each S. (Such a shift could be due to an exogenous reduction in the transaction cost of using offshore workers.) Similarly, the curves and analysis in Figures 1–3 would be qualitatively preserved, under their respective assumptions about the flexibility or rigidity of the home wage.
Our analysis assumes that there is no disutility from work. To relax this assumption, generalise the utility function to take the form U(C − δN), where δ is a non negative constant that represents the opportunity cost of effort. When δ = 0, the results of Sections 2 and 3 apply as derived. Even with a positive δ, these results still qualitatively hold as long as w > δ (i.e. employment is preferable to leisure). For example, in the static case illustrated by Figure 2, the loss represented by area ABCD would be reduced – but not eliminated – to a smaller area with height instead of. The dynamic analysis corresponding to Figure 4 would be similarly preserved, as we could readily show.
The previous sections compare two polar cases – full employment with a perfectly flexible wage, versus unemployment with a completely rigid wage (in terms of output or utility) – which have diametrically opposite implications for welfare. For intermediate cases, in which both national employment and the real wage account for some of the economy’s adjustment, our results on offshoring would presumably be ambiguous.13 Formally, letting with in our dynamic model, we would approach the full employment or rigid wage case as tends to ∞ or 0, respectively. The same comment applies (mutatis mutandis) to the static model if is re specified as an increasing function of N.
This study makes three main contributions. First, it highlights welfare losses from using offshore labour in the presence of unemployment. These losses occur despite the assumption that workers own the capital stock, and regardless of whether this stock rises in response to offshoring. Thus, our analysis suggests that public concerns over potential job losses from offshoring are justified. Second, in both the static and dynamic cases, we demonstrate the crucial importance of incorporating unemployment into the theoretical framework. When unemployment is assumed away, the analysis yields completely opposite conclusions about the welfare consequences of hiring offshore. Third, we extend the theory of offshoring to incorporate optimal saving and investment, for both rigid and flexible wage economies.
In defence of offshoring, Mankiw and Swagel (2006, p. 1041) argue (in part) that ‘An open economy can just as easily be fully employed as an autarkic one’. But what if autarky itself suffers from unemployment? Then, our analysis suggests that the use of offshore labour tends to exacerbate this pre existing condition, and thereby lower social welfare. These troublesome tendencies provide a rationale for the popular objection to offshoring.
1 On the definition of ‘offshoring’, see Blinder (2009). The same phenomenon is defined as ‘outsourcing’ by Bhagwati (2009) and Bhagwati et al. (2004).
2 Results in their multiple good models 2 and 3 are subject to a qualification stemming from possible deterioration in the terms of trade.
3 Whereas Zhang (2011) also adopts this type of wage rigidity, the other related papers cited above examine unemployment caused by alternative labour market imperfections. Specifically, Egger and Kreickemeier (2008), Brecher and Chen (2010, 2012) and Kohler and Wrona (2011) incorporate efficiency wage considerations; the latter study further considers search and matching frictions, as do Davidson et al. (2008), Keuschnigg and Ribi (2009) and Mitra and Ranjan (2010), while a monopoly labour union is assumed by Koskela and Stenbacka (2010).
4 Nevertheless, units of the one good are shipped internationally to pay for the use of offshore labour.
5 Here we assume that there is no disutility from work. Nevertheless, as discussed in Section 4, our main results would still hold if this assumption is relaxed.
6 This wage gap could be due to technological or factor abundance differences between countries, or to the wage floor specified below.
7 By well known reasoning, home competition for the quota restricted units of foreign labour would drive up their wage to the same level as earned by home workers. (Of course, labour remaining in foreign production receives the lower foreign wage.) Our main results would still hold if S is endogenously determined, or offshore workers used by home firms receive less than the home wage, as discussed in Section 4. In any case, because home firms pay the same amount for each unit of (offshore or domestic) labour, Grossman and Rossi Hansberg’s (2008)‘productivity effect’ of offshoring does not arise in the present paper.
8 See Bhagwati et al.’s (2004) model 1 for this case.
9 Such a floor could arise even if the wage were endogenously determined, as in Brecher and Choudhri (1994), Romer (2006) and Solow (1979).
10 For an analytically similar treatment of international migration, where migrants are ineligible for unemployment benefits in the host country, see Brecher and Choudhri (1987).
11 Assuming instead that (as in Section 2) would here rule out the existence of steady state equilibrium, except for one particular (knife edge) value of. Incidentally, our static results in the previous section would be qualitatively unchanged by use of the present (utility based) version of the minimum wage constraint (i.e. wU′(C) ≥ ω) , as we could readily show.
12 For instance, suppose that offshore workers are charged for the rights to a quota on S, and that the home government redistributes the resulting revenues to its nationals in lump sum fashion.
13 Examples of such cases include the efficiency wage, union bargaining and search and matching models of Shapiro and Stiglitz (1984), McDonald and Solow (1981) and Pissarides (1985), respectively.
14 We gratefully acknowledge helpful comments and suggestions from Alan S. Blinder and anonymous referees. The research of Chen and Yu was financially supported by the Social Sciences and Humanities Research Council of Canada.
Bhagwati, J. ( 2009 ), ‘ Don’t Cry for Free Trade ’, in J. Bhagwati and A. S. Blinder, Offshoring of American Jobs: What Response from U.S. Economic Policy? (edited and with an introduction by Benjamin M. Friedman) (Cambridge, MA: MIT Press ), 1 – 18.
Bhagwati, J., A. Panagariya and T. N. Srinivasan ( 2004 ), ‘ The Muddles Over Outsourcing ’, Journal of Economic Perspectives, 18, 4, 93 – 114.
Blinder, A. S. ( 2009 ), ‘ Offshoring: Big Deal, or Business as Usual? ’, in J. Bhagwati and A. S. Blinder, Offshoring of American Jobs: What Response from US Economic Policy? (edited and with an introduction by Benjamin M. Friedman) (Cambridge, MA: MIT Press ), 19 – 60.
Brecher, R. A. and Z. Chen ( 2010 ), ‘ Unemployment of Skilled and Unskilled Labor in an Open Economy: International Trade, Migration, and Outsourcing ’, Review of International Economics, 18, 5, 990 – 1000.
Brecher, R. A. and Z. Chen ( 2012 ), ‘ Unemployment and Welfare Consequences of International Outsourcing Under Monopolistic Competition ’, April, manuscript, Carleton University; original version, October 2011, Economic Papers CEP 11 07 (Ottawa: Carleton ).
Brecher, R. A. and E. U. Choudhri ( 1987 ), ‘ International Migration Versus Foreign Investment in the Presence of Unemployment ’, Journal of International Economics, 23, 3/4, 329 – 42.
Brecher, R. A. and E. U. Choudhri ( 1994 ), ‘ Pareto Gains from Trade Reconsidered: Compensating for Jobs Lost ’, Journal of International Economics, 36, 3/4, 223 – 38.
Davidson, C., S. J. Matusz and A. Shevchenko ( 2008 ), ‘ Outsourcing Peter to Pay Paul: High Skill Expectations and Low skill Wages with Imperfect Labor Markets ’, Macroeconomic Dynamics, 12, 4, 463 – 79.
Egger, H. and U. Kreickemeier ( 2008 ), ‘ International Fragmentation: Boon or Bane for Domestic Employment? ’, European Economic Review, 52, 1, 116 – 32.
Grossman, G. M. and E. Rossi Hansberg ( 2008 ), ‘ Trading Tasks: A Simple Theory of Offshoring ’, American Economic Review, 98, 5, 1978 – 97.
Keuschnigg, C. and E. Ribi ( 2009 ), ‘ Outsourcing, Unemployment and Welfare Policy ’, Journal of International Economics, 78, 1, 168 – 76.
Kohler, W. and J. Wrona ( 2011 ), ‘ Offshoring Tasks, Yet Creating Jobs? ’, Working Papers in Economics and Finance, 12 (Tübingen: University of Tübingen ).
Koskela, E. and R. Stenbacka ( 2010 ), ‘ Equilibrium Unemployment with Outsourcing and Wage Solidarity under Labour Market Imperfections ’, European Economic Review, 54, 3, 376 – 92.
Mankiw, N. G. and P. Swagel ( 2006 ), ‘ The Politics and Economics of Offshore Outsourcing ’, Journal of Monetary Economics, 53, 5, 1027 – 56.
15 McDonald, I. M. and R. M. Solow ( 1981 ), ‘ Wage Bargaining and Employment ’, American Economic Review, 71, 5, 896 – 908.
16 Mitra, D. and P. Ranjan ( 2010 ), ‘ Offshoring and Unemployment: The Role of Search Frictions Labor Mobility ’, Journal of International Economics, 81, 2, 219 – 29.
17 Pissarides, C. A. ( 1985 ), ‘ Short run Equilibrium Dynamics of Unemployment, Vacancies, and Real Wages ’, American Economic Review, 75, 4, 676 – 90.
18 Romer, D. ( 2006 ), Advanced Macroeconomics, 3rd edn (New York: McGraw Hill/Irwin ).
19 Shapiro, C. and J. E. Stiglitz ( 1984 ), ‘ Equilibrium Unemployment as a Worker Discipline Device ’, American Economic Review, 74, 3, 433 – 44.
20 Solow, R. M. ( 1979 ), ‘ Another Possible Source of Wage Stickiness ’, Journal of Macroeconomics, 1, 1, 79 – 82.
21 Zhang, T. ( 2011 ), ‘ International Outsourcing and Unemployment in a Minimum wage Economy ’, Review of International Economics, 19, 4, 776 – 86.
Graph: 1 Static Case with Full Employment
Graph: 2 Static Case with Unemployment
Graph: 3 Dynamic Case with Full Employment
Graph: 4 Dynamic Case with Unemployment
By Richard A. Brecher; Zhiqi Chen and Zhihao Yu