QUEENS UNIVERSITY OF CHARLOTTE

Transcription

1 f QUEENS UNIVERSITY OF CHARLOTTE DISCUSSION PAPER SERIES Discussion Paper OPERATING PERFORMANCE IN THE WAKE OF ERP IMPLEMENTATION: TRIANGULATING RESULTS FOR CHINESE MANUFACTURING COMPANIES by ARLINE SAVAGE JOSEPH CALLAGHAN LI DANG Y. SUN Contact Information: Arline Savage McColl School of Business Queens University of Charlotte 1900 Selwyn Ave. Charlotte, NC

2 OPERATING PERFORMANCE IN THE WAKE OF ERP IMPLEMENTATION: TRIANGULATING RESULTS FOR CHINESE MANUFACTURING COMPANIES By ARLINE SAVAGE Queens University of Charlotte, Charlotte, NC JOSEPH CALLAGHAN Oakland University, Rochester, MI LI DANG Cal Poly San Luis Obispo, CA Y. SUN Beijing Technology and Business University, China August, 2010

3 Operating Performance in the Wake of ERP Implementation: Triangulating Results for Chinese Manufacturing Companies Abstract This paper reflects the first stage of research related to company performance in the aftermath of an ERP implementation. We obtained a unique set of data for 198 public Chinese manufacturing companies involved in enterprise resource planning (ERP) implementations from 1993 to We collected fundamental data from a survey of these companies, and then analyzed their ERP implementations and performance. We use structural equation modeling (SEM) of subjects perceptions to model firm operational performance as a function of effective ERP implementation that, in turn, is a function of both the level of CEO involvement and the extent of Business Process Reengineering (BPR). We find that operational performance is positively associated with ERP effectiveness, and that this effectiveness is positively related to both the level of CEO involvement and the extent of BPR. A subsequent stage of this research program will examine both accounting and market performance measures as possible validating measures of perceived operating performance. Consequently, measures from three levels will be examined: managerial perception, firm performance and market response to these ERP implementations. This three-prong approach enables us to triangulate results regarding the perceived and actual factors affecting the efficacy of the ERP implementations.

4 Operating Performance in the Wake of ERP Implementation: Triangulating Results for Chinese Manufacturing Companies INTRODUCTION Since economic reform in China was initiated by Premier Deng Xiaoping in 1978, the country s economy has grown significantly at an annual rate of 9.0 percent or more for nearly three decades (Brown and He 2007: 132; World Bank 2010: 10). China now plays an increasingly important role in the world economy. While the global economy remains in dire straits, China's economic growth has been strong and its economic outlook remains favorable, with predictions of GDP of 9.5 percent for 2010 and 8.5 percent for 2011 (World Bank, 2010). This makes China the world's second-biggest economy after the United States. It is from this perspective that we consider the effects of the growth in the implementation and integration of enterprise resource planning (ERP) in China. As Martinsons (2005: 46) succinctly states: More than any other country, China is being transformed by its application of IT, from a poor and isolated society to a major force in the global economy. It is in this context that the operating effectiveness of systems implementation and integration in China can be envisaged. The increasing globalization of the Chinese economy is leading to a growing need for modern software to replace traditional legacy systems, which inhibit economic sustainability and competitiveness. The Chinese government supports ERP systems implementations. Prior to 2000, ERP implementations generally had a success rate of about 33 percent, while the success rate in China was much lower than the Western World at only 10 percent. However, since 2001, the Chinese government s policy of using IT to speed up industrialization has been widely implemented (Ge and Voß 2009: 503). According to a Gartner, Inc. (2010) press release, ERP is a priority area for software spending in China. China s ERP software market is forecast to 1

5 maintain strong performance, with an estimated compound annual growth rate (CAGR) of 14.6 percent from 2008 to 2013 and 14.8 percent in 2010 the highest in the world. In comparison, the ERP market in the U.S. had a CAGR for of 5.0 percent (IBISWorld 2010). The top four industries, which account for approximately 60 percent of total software spending in China, are manufacturing, financial services, communications and government. In this study, we focus on Chinese manufacturing companies. New empirical studies are needed to validate any improvement in operating and financial performance after ERP implementation, especially in view of the enormous financial resources expended on ERP implementations and the paucity of research in the accounting literature. That is the purpose of this study. Disparate information systems impede an organization s long-term growth (Xue et al. 2005). Change is driven by economics and competition and made possible by technology (Bancroft et al. 1998: 216). ERP systems were developed to provide the solution by automating and integrating all or many of an organization s business processes and functions into a single software system, using one application, a single database to share data across the enterprise, a unified interface, and real-time access to information (e.g., Bingi et al. 1999; Nah et al. 2001), resulting in integrated information flows (Davenport 1998). Adopting firms expect ERP systems to improve operating performance and ultimately financial performance (Kallunki et al. 2010). The top-tier ERP vendors are SAP, Baan, Oracle and Peoplesoft. ERP was developed to manage and automate all aspects of an organization s business operations, including sales and distribution, materials, human resources, production planning, purchasing, sales, customer relationships, supply chain, product life-cycle, business intelligence and accounting. Each ERP module is business-process specific (Benco and Prather 2008: 145). EDI, Internet EDI, and extranets are used to connect an ERP system to the IT systems of vendors and customers. This 2

6 allows different functional areas within the organization to communicate with one another and with customers and vendors more effectively. The successful implementation of an ERP system should lead to significant benefits such as reduced labor, better cash flow management, increased accounts receivable turnover, fewer bad debts, more timely reporting, reduced inventory, tightened supply chain links, improved production scheduling, reduced manufacturing costs, reduced order processing time, improved customer service, and consequently, an increase in competitiveness, profitability and stock prices. However, history shows that the successful implementation and integration rate remains fairly low and many enterprises that have gained some benefits from ERP implementation have not exploited the software s full potential (Ge and Voß 2009: 501; Zabjek et al. 2009). Research into ERP effectiveness and performance is important from an accounting perspective. By its nature, accounting should provide the answer of whether ERP systems are worth the money that firms invest in them. The focus of this paper is on ERP implementation and integration in the manufacturing industry in China, specifically post-implementation ERP effectiveness and post-implementation operational performance. This paper is the first part of a stream of research examining operating performance in the wake of an ERP implementation using subjects perception measures. In the next part of this research stream, we will use accounting and market measures of performance respectively. The next section of this paper reviews the relevant literature. The third section proposes a theoretical model, while the fourth section discusses the research methodology. The fifth section discusses the results, while the final section summarizes the paper and lays the foundation for the next stage of this research stream. 3

7 LITERATURE REVIEW ERP Implementation and Integration Critical Success Factors The most extensive category of ERP implementation research is the critical success factor (CSF) stream (Brown and He 2007: 134), based on the concept of critical success factors popularized by MIT s Rockhart (1979). This research appears to be based mainly on survey research of perceptions of successfulness and case studies of ERP implementations. Dezdar and Sulaiman (2009) provide the most recent extensive investigation into ERP implementation CSFs in the literature from and then formulate a taxonomy and rankings of 17 CSFs identified from 95 articles by doing frequency and comparative analyses. The highest ranking CSFs (with frequency out of 95 articles in parenthesis) were (1) top management support (72%); (2) effective project management and evaluation (70%); (3) business process reengineering and minimum customization (62%); and (4) ERP team composition and competence (56%). Motwani and Akbulut (2008: 162) identify almost identical CSFs from the literature as Dezdar and Sulaiman (2009). We have decided to focus on only two top-cited CSFs, CEO sponsorship/involvement and BPR, for our model for the following reasons: (1) our methodology s focus on operating performance, (2) the abundance of research into ERP CSFs, (3) our need to limit the length of the questionnaire, (4) enterprise systems are essential IT enablers of BPR (Hammer and Champy 1993); and (5) findings that the top management sponsorship and involvement variable is a prerequisite for the other CSFs (Akkermans and van Helden 2002: 42; Bancroft et al.1998: 136). CEO Sponsorship and Involvement The literature contends that top management support is key to securing the necessary conditions (i.e., the other critical success factors) to successful implement an ERP system (e.g., 4

8 Aldwani 2001: 272; Bingi et al. 1999; Cheng et al. 2006; Dong 2001; Kamhawi and Gunasekaran 2009: 699; Reimers 2003; Umble and Umble 2002; Yusuf et al. 2006). Partially overlapping with Desai and Sulaiman s (2009) synthesis of 95 articles, Ngai et al. (2008) present a review of 48 studies in the CSF literature up to and including 2005 of 12 groups of countries and regions, including South America, the USA, Europe, the UK, Denmark, Australia, China, the Middle East, and the Arab Gulf States. Six of the studies they identify include China, with four solely focused on China. These 48 studies identify 18 CSFs and all of them list top management support as a one of the most critical success factors, with many investigations finding this CSF to be the single most important factor internationally to secure the necessary conditions for introducing the organizational changes that will result from a successful ERP implementation (e.g., Akkermans and van Helden 2002: 36; Aladwani 2001; Al- Sehali 2000; Brown and He 2007; He 2004; Jafari et al. 2009; Yusuf et al. 2006, Zabjek et al. 2009). Law and Ngai (2007) and Reimers (2003: 347) also find that top management support of BPI (a consequence of successful ERP) is positively associated with the extent of BPI. CEO sponsorship and commitment provides a symbolic indication of priority for an ERP project, which should in turn promote organization-wide commitment, user involvement and a successful ERP system (Jafari et al. 2009: 399; Zhang et al. 2005: 78). Business Process Reengineering BPR followed by ERP (Koch 2001: 261) is necessary for a successful and integrated ERP system (e.g., Ip and Chen 2004; Koch 2001; Yusuf et al. 2006: 1332; Zabjek et al. 2009). A business process is a set of logically related tasks or activities performed to achieve a particular business outcome. For example, the business outcome of the cash receipts process is cash deposited in the bank. BPR is the analysis and redesign of an organization s workflows and 5

9 processes. According to Hammer and Champy (1993) to benefit from BRP, ERP is the essential [IT] enabler because the objective is that all processes in an organization conform to the ERP model, which consists of generic best practices embedded in templates in the software in areas like management, human resources, accounting, production, etc (Gattiker and Goodhue 2005: 560). Both the costs and benefits of realigning business processes with an ERP model can be substantial. However, some of an organization s business processes may be unique and to redesign them to conform to the ERP model would result in a loss of competitive advantage. Therefore, the ERP software would have to be customized for those processes to better suit the needs of the organization. But more customization results in higher implementation costs (Bingi et al. 1999: 10). Consequently, the goal of BPR would be problem identification to conform to the ERP model, with minimum ERP customization (Dezdar and Sulaiman 2009; Ip and Chen 2004: 81). Research shows that organizations with better BRP abilities usually experience smoother ERP implementations (Kamhawi and Gunasekaran 2009: 691; Zabjek et al. 2009). Because extensive process redesign results in significant organizational changes, top management support for BPR would logically to be needed. Consequently, one would expect a significant relationship between CEO sponsorship and involvement and the successful operationalization of BPR by the ERP system, resulting in business process improvement (BPI). Other Factors The other CSFs indentified by Dezdar and Sulaiman (2009) in their synthesis of the literature (ranked in order of importance) are: (1) project management and evaluation; (2) ERP team composition, competence and compensation; (3) change management program; (4) user training and education; (5) business plan and vision; (6) enterprise-wide communication and cooperation; (7) organizational culture; (8) ERP vendor support; (9) software analysis, testing 6

10 and troubleshooting; (10) project champion; (11) careful selection of ERP software; (12) use of consultants; (13) appropriate business and IT legacy systems; (14) system quality; and (15) user involvement. Performance in the Wake of ERP Implementation An implicit assumption often appears to be that successful ERP implementations will somehow lead to business change and improved operational performance. However, this is not always supported by the extant research (e.g., Law and Ngai 2007: 419). For example, Martin s (1998) title is: An Electronics Firm Will Save Big Money by Replacing Six People with One and Lose All This Paperwork Using ERP Software [emphasis added]). As previously stated, the success rate of ERP implementations has been low, with researchers reporting up to a 90 percent failure rate (see Zabjek et al for an in-depth discussion). We organize this literature stream into two categories: (1) post-implementation ERP and integration effectiveness and BPI; and (2) post-implementation operational performance. Theoretically, ERP implementation affords an opportunity for BPR, hopefully resulting in systems integration and BPI. We next consider two categories of variables (1) ERP Effectiveness and Postimplementation Systems Integration, and (2) Post-implementation operational performance. We consider variables from the first category to be independent variables of the second category. ERP Effectiveness and Post-implementation Systems Integration Using an Analytic Hierarchy Process, Tsai et al. (2006) assess the relative importance weights of ERP measures and identify the even most important measures of success as: (1) accuracy of output; (2) precision of input; (3) data accuracy; (4) frequency of report requests; (5) timeliness of output; (6) user s confidence in the system; and (7) system accuracy. 7

11 Gattiker and Goodhue (2002) found that the adoption of ERP software was positively associated with business process change. Delone and McClean (1992) identified the variable Information Systems Success, which has since been generally accepted as a suitable foundation for further research (Zhang et al. 2005: 59). Delone and McClean identify the following dimensions to measure, which Tsai et al. (2007) use to measure perceived ERP success: system quality, information quality, system use, user satisfaction, individual impact on information recipient s behavior, and organizational impact. Tsai et al. (2007) further propose the Balanced Scorecard (BSC) method to measure the performance improvements in organizational impact from four traditional BSC perspectives, namely financial, customer, internal business process, and learning and growth. Logically, the BSC internal business process and learning and growth dimensions relate to this category, while the other two relate to the following category. Post-implementation Operational Performance The purpose of investments in ERP systems is to improve firm organizational efficiency and effectiveness, resulting in improved operational performance (non-financial) and ultimately, improved financial performance (profitability and stock prices). While Gattiker and Goodhue (2002) found a positive association between ERP adoption and business process change, they did not find support for a relationship between business process change and organizational impact a counter-intuitive finding. For 96 companies internationally and from diverse industries, Law and Ngai (2007) later found that the perceived extent of BPI as measured of five items used for perceptual ratings: errors/defects prevention, process improvement standards, ease of use of new processes, intra- and extra-organization and coordination, was positively associated with ERP success as measured by user satisfaction (an average computed of the 12 8

12 items of the Doll and Torkzadeh (1988) user satisfaction instrument on five dimensions: information content, accuracy of information, output format, ease of use, and timeliness of information). Law and Ngai (2007) later found that both ERP success and the perceived extent of BPI were positively related to perceived organizational performance, as measured by an average of calculated from five items obtained from respondents perceptions: customer satisfaction, customer retention rate, sales growth rate, firm profitability, and overall competitive advantage. O Leary (2004) confirms the Deloitte Consulting (1998) study of key ERP benefits realized, which are inventory reduction, personnel reduction, productivity improvements, order management improvements, financial close cycle reduction, IT cost reduction, procurement cost reduction, cash management improvement, revenue/profit increases, transportation/logistics cost reductions, maintenance reductions and on-time delivery. Results from a study by Wieder et al. (2006) are contradictory. Using perception-based measures to capture both business process and financial performance measures, they found no support for their prediction that ERP adopters will have significantly higher performance scores at both the core business process level and at firm level. However, they did find that longer experience with ERP has a positive impact on firm performance. The financial performance measures used were return on investment, operating profits, sales growth rate, cost reduction programs, and cash flow. The 22 business process measures include relative levels of various logistics, manufacturing and inventory costs, speed at filling orders, percentage of on-time deliveries, backorder level, production time, shipping errors, and customer complaints (Wieder et al. 2006: 28-29). 9

13 Financial and Stock Market Performance Velcu (2007) finds that, while ERP adopters report a number of post-implementation operational benefits, these adopters are unable to quantify the impact of ERP on financial performance (Kallunki et al. in press:17). Poston and Grabski (2001) examine the financial performance for a group of firms for three years before and after ERP implementation and find no general financial improvement. Despite this finding, the employees-to-revenue ratio significantly decreased in all three years and the sales-to-cost ratio improved in the third year. The measures they used were residual income, SG&A expenses/revenues, COGS/revenues and number of employees/revenues. Hunton et al. (2003) investigate longitudinal financial performance effects of ERP implementation by matching adopters to non-adopters. They find that financial performance, measured by return on assets, return on investments and asset turnover, was not significantly different for adopters when examining pre-to-post ERP implementation effects over a three-year period. However, non-adopters significantly underperformed adopters by the third year, suggesting that ERP helps adopters gain a competitive advantage over non-adopters. Similarly, Nicolaou et al. (2003) compare financial data of ERP system adopters with a control group. They used return on assets, return on sales and asset turnover measures. Their results show that firms with ERP systems had significantly higher performance in the second year after implementation than the control group. Matolcsy et al. (2005) track financial ratios (net profit margin, current ratio, fixed asset turnover, sales days outstanding, accounts payable days, inventory turnover, sales change) for two years for companies that adopted ERP systems versus companies that did not adopt ERP. The result was that the adoption of ERP systems led to sustained operational efficiencies and improved 10

14 overall liquidity. In addition, they found some support for increased profitability two years after the adoption of ERP and for improvements in accounts receivable management. Research studies on ERP systems in the accounting literature also tend to focus on stock price or financial analyst reaction to announcements of plans to invest in ERP systems (Benco and Prather 2008; Hayes et al. 2001, Hunton et al. 2002). These studies show that investors and financial analysts believed that the ERP implementations would enhance firm value. Wier et al. (2007) also use return on assets and stock returns to measure corporate performance. A study that considers the long-term stock price and profitability effects of 186 ERP investment announcements appears in the operations management literature (Hendricks et al. 2007). The authors examine performance effects for both the implementation and postimplementation phases and find some evidence of improvements in profitability, but not in stock returns. A possible explanation for this apparent anomaly could be that the market had previously incorporated the anticipated benefits. Despite high ERP implementation costs, they find no consistent evidence of negative performance associated with enterprise system investments. Generally, empirical results of performance improvement are mixed. A subsequent stage of this research program will examine both accounting and market performance measures of our sample firms as possible validating measures of perceived operating performance. THEORETICAL MODEL AND HYPOTHESES Our research focus for this paper is on post-implementation operational performance of ERP systems for public Chinese manufacturing companies. The literature suggests that this dependent performance construct and its related measures are a positive function of both effective ERP implementation and post-implementation systems integration. These intervening 11

15 factors in turn are posited to be a positive function of the critical success factors, CEO sponsorship and involvement and BPR. Post-implementation operational performance measures, as well as measures related to other factors, are based on respondents scaled responses to questions deemed related to operational performance (and possibly business process improvement (BPI) measures). The scale used was a 5-point Likert-like scale, ranging from Strongly Disagree (1) to Strongly Agree (5). All factor categories and their related Questionnaire items used in this study are shown in Appendix A. Intervening factors, ERP effectiveness and post-implementation systems integration, are also based on respondents perceptions, as are the two critical success factors, CEO sponsorship and involvement and BPR. Figure 1 reflects our framework. Working right to left, the underlying CSFs for ERP effectiveness and post-implementations systems integration are CEO sponsorship and involvement, and the extent to which BPR is undertaken. The extent to which BPR is undertaken is also related to CEO sponsorship and involvement. If an ERP implementation is viewed strategically as more than merely a software installation, albeit a big one, and as an opportunity to reengineer underlying business processes, conforming to the implied business process of the ERP system that are perhaps reflective of best business practices, then CEO sponsorship and involvement should positively affect BPR. In addition, BPR independent of the ERP implementation will likely affect how the ERP is configured and customized, and how postimplementation systems integration is effectively accomplished. Thus CEO sponsorship and involvement affects BPR and both intervening factors, ERP effectiveness and postimplementation systems integration. But ERP effectiveness and post-implementation systems integration are not business ends in themselves. Rather, they are intervening factors to BPI and the payoffs in terms of improved operational performance, however measured. 12

16 [Insert Figure 1 about here] Thus, the hypotheses for this study, expressed in the alternative form, are: Hypotheses (alternative form, one-tail) H a 1 The extent of BPR is positively related to CEO sponsorship and involvement (must still be tested). H a 2 ERP effectiveness is positively related to CEO sponsorship and involvement. H a 3 H a 4 H a 5 Post-implementation systems integration is positively related to CEO sponsorship and involvement. ERP effectiveness is positively related to BPR measures. Post-implementation systems integration is positively related to BPR measures. H a 6 ERP effectiveness is positively related to post-implementation systems integration (must still be tested). H a 7 Post-implementation operational performance is positively related to ERP effectiveness. H a 8 Post-implementation operational performance is positively related to post-implementation systems integration. RESEARCH METHODOLOGY We used a self-administered, 5-point Likert scale multi-item survey instrument (see Appendix A for categories and items) because the complexity inherent in an investigation into ERP implementation and performance cannot be adequately measured with a single item. Multiitem scales can reduce measurement error of complex, latent variables by combining several individual items (Yang and Su 2009). The questionnaire was designed after a cross-disciplinary literature review (e.g., accounting, operations management, information technology, information systems and management) to identify ERP implementation factors and performance constructs for post-implementation operational performance. From this review, we determined questionnaire categories and items. Then practitioner experts, namely ERP consultants, CIOs and ERP project managers, who were qualified to assess the content validity of the proposed items and constructs were interviewed to refine the questionnaire. This was followed by a pilot test 13

17 with ten subjects. The feedback was that some of the questions were too complicated. Therefore, we simplified the language to make those questions more understandable. The response rate was 70 percent, with 198 usable responses. We use Structural Equation Modeling (SEM) to evaluate the conceptual model and Cronbach s alpha to measure the reliability of the Likert scales. Although SEM has become one of the most widely-used multivariate statistical tools used in disciplines such as education, psychology and the behavioral sciences, there is a paucity of research using this tool in accounting research (see (Henri 2007; Smith and Langfield-Smith 2004). SEM is a statistical technique that allows for the simultaneous analysis of a series of structural equations with two aspects: (1) the estimation of multiple and interrelated dependent relations between variables; and (2) the ability to represent latent variables 1 in these relations while also accounting for estimated measurement error associated with the imperfect measurement of variables (Smith and Langfield-Smith 2004: 50). We use AMOS (software for SEM) to measure the relations between the variables of our conceptual model. This software allows the error variance to be specified within the statistical analysis. For purposes of this study, we found the SEM model-generating strategy (Jöreskog 1993) particularly useful. With this strategy, the researchers first specify an initial conceptual model. If this initial model does not fit the data, then the model is modified and tested again. This process can be repeated as many times as is necessary. Modifications to the model(s) are driven by either the theory or the data. The objective is to generate a model that provides a statistical fit for the data (Smith and Langfield-Smith 2004: 70). 1 A latent variable is a variable that is not directly measured, but is inferred through a mathematical model from a series of observed indicators that are directly measured. In this paper, an example of a latent variable is ERP Effectiveness, which is indirectly measured through scaled responses to a series of items (the observed variables). 14

18 In the accounting literature, Smith and Langfield-Smith (2004) provide a comprehensive analysis of the advantages of SEM over multiple regression and path analysis. Regardless, before applying SEM, we did perform an ordinary least squares (OLS) regression analysis to provide preliminary evidence that the underlying constructs predicted by the literature, theory and logic would prevail. The results of this regression analysis confirmed our SEM results. DISCUSSION OF RESULTS Descriptive Statistics Subjects Background Information We surveyed 198 publicly-traded manufacturing companies is China in mid-2007, all of which invested in the ERP software of one of the major players in the ERP market. The respondents positions in their firm were CIOs (50%), other IT or systems-related employees (18%), the ERP project manager (5%), other (10%), unclear (9%), and missing values (8%). The average ERP experience of respondents was 6.23 years. Sixty-six respondents were the ERP project team leaders, while 83 were team members. Seven respondents were not members of the project team, while 11 companies did not have ERP project teams. Thirty-one subjects didn t respond to this question. ERP Implementation Of the 198 surveyed companies, 138 had completed ERP implementation and 60 had not completed it. Of the 138 companies that had completed ERP implementations, the average time taken to complete the implementations (the gap between the time when ERP was initiated and the time when it was completed) was months (there are six missing values). Of the 60 companies that had not completed implementations, the average completion rate was 63.21% (13 companies did not respond to this question). 15

19 The average time taken to do pre-implementation research was 9.49 months. The average time taken to implement the ERP system was 8.86 months. The average time taken to test the system was 3.79 months. 150 companies purchased ERP modules from only the one provider; the remainder purchased modules from more than one provider. 72 companies hired a consulting company for module integration, while 125 companies did not. The total implementation cost, on average, was 25,417,719.3 RMB (there were 27 missing values). The average total cost for those companies that had completed implementations was 27,942,177.4 RMB (with 14 missing values). The average total cost for companies that had not completed implementations was 18,757,446.8 RMB (with 13 missing values). The choice of an ERP vendor was decided on the following factors: Reasonable price (41 responses); brand name (65); leading technology (53); perfect functions (39); wide/flexible applications (7); good customer support (37); fits business needs (77); perfect solutions (37); good training (6); user friendly (20); easy for customization (34); good communications (26); prior successful implementations (60); required by a supervising company (1); government recommendations (0); headquarter decision (10); other (60). Sixty-four companies had some modules before ERP implementations. For these companies there were, on average, two pre-erp modules. On average, about 7.59 modules had been or were in the process of being implemented for all surveyed companies. One hundred and three companies had customization, which is not surprising given that they are manufacturing companies. The average customization cost was 4,238, RMB. Of the 138 companies that completed ERP implementations, 90 had gone through customization and the average cost was 4,820, RMB (with 35 missing values). 16

20 During implementation, 184 companies had an ERP project team. On average, there were 68 (16) members (technical members) in the project team. In 124 companies, the CEO was one of the leading members in the project team. Regression analysis Ordinary least squares (OLS) regression results are presented below. The measurement for each variable is achieved by taking a simple average of those responses to related questionnaire items. The items in turn were selected based first on their questionnaire category and then on their reliability and Confirmatory Factor Analysis (CFA) measures. (See discussion below under the SEM Measurement Model). The OLS regressions proceeded in two stages: 1) Dependent Factor on Intervening Factors with controls for organization structural changes and degree of competition in the ERP client output market, and 2) each intervening factor was regressed on the two CSFs, extent of CEO involvement and BPR. Stage 1 OLS: Post-Implementation Operational Performance (A) = ERP Effectiveness (C) + Postimplementation systems integration (D) + G + H (where G and H are control variables). G and H were removed from the structural model because including them causes a poor fit of the structural model. Unstandardized Coefficients Standardized Coefficients Model B Std. Error Beta t Sig. 1 (Constant) ERP Effectiveness (C) Post-Implementation Systems Integration (D) Structural Changes (G) <.001 Competition (H)

21 Consistent with SEM results, post-implementation systems integration (D) positively impacts operational performances. However, the coefficient on ERP effectiveness (C) is not significantly related to operational performance. The control variables are significant and correctly signed. Stage 2 OLS: (1) ERP Effectiveness (C) = CEO (E) + BPR (F) Unstandardized Coefficients Standardized Coefficients Model B Std. Error Beta T Sig. 1 (Constant) CEO (E) BPR (F) <.001 Also consistent with SEM results, both CEO involvements and BRP significantly and positively impact ERP effectiveness. (2) Post-Implementation Systems Integration (D) = CEO (E) + BPR (F) Unstandardized Coefficients Standardized Coefficients Model B Std. Error Beta T Sig. 1 (Constant) CEO (E) <.001 BPR (F) Additionally, both CEO involvements and BRP positively and significantly impact postimplementation systems integration. Thus, generally, the OLS regression results provided fragmented evidence that underlying constructs predicted by the literature, theory and logic prevail. The one surprise was the lack of evidence supporting the notion that effective ERP implementations necessarily lead to improvements in perceived operational performance. 18

22 SEM Measurement Model 1. Reliability and Confirmatory Factor Analysis Results The following table summarizes the results related to the measurement model of the SEM. In each case the selected set of questions provided a reasonable measure of reliability associating them to their related factor based on a Cronbach alpha cutoff of.65 to.70 (Cheng et al. 2006; Nunnally 1978). The only exception is the CEO involvement factor whose raw alpha measure was only.53. Improvements could be possibly made by dropping questions, but since there were only four questions and the factor was found impactful under OLS, we judged that all questions be included for this factor. Further, CFA found the factor to be a good fit in the proposed model. In general, the CFA results confirmed that the underlying factors should be included in the model. Factor Operational Performance Measured Variables Selected Cronbach's Alpha (standardized) Reliability Confirmatory Factor Analysis (CFA) A1-A (0.81) All loadings on the factor are significant. The loading on A1 is a little low but removing it would lower the reliability score and the CFA model fit. ERP Effectiveness C1-C (0.807) All loadings are significant. Model fit is a little low. Post-Implementation Systems Integration CEO Sponsorship & Involvement D1-D (0.849) All loadings are significant. CFA model has a good fit. E1-E (0.532) All loadings are significant and the model shows a good fit. BPR Measures F1-F (0.697) All loadings are significant. Model fit is a little low. Organization Factor Post-ERP Structural Changes G1-G (0.653) All loadings are significant. Model fit is OK. Competition H1-H (0.770) All loadings significant. Model fit is still a little low. 1 This score is low against the traditional standard of 0.7 or higher. 19

23 2. Factor Loadings The results of the factor loading of the individual questions are presented in the following tables. Specifically, the loading estimate, standard error, critical ratio (C.R.) and two-tailed p- values are detailed. All questions are significantly loading on their associated factor. Dependent Associated Estimate S.E. C.R. p-value Factor Factor A5 > A(Performance) A4 > A(Performance) <.001 A3 > A(Performance) <.001 A2 > A(Performance) <.001 A1 > A(Performance) <.001 C6 > C(Effectiveness) C5 > C(Effectiveness) <.001 C4 > C(Effectiveness) <.001 C3 > C(Effectiveness) <.001 C2 > C(Effectiveness) <.001 C1 > C(Effectiveness) <.001 D1 > D (Integration) D2 > D (Integration) <.001 D3 > D (Integration) <.001 D4 > D (Integration) <.001 D5 > D (Integration) <.001 D6 > D (Integration) <.001 D7 > D (Integration) <.001 D8 > D (Integration) <.001 E1 > E(CEO) E2 > E(CEO) <.001 E3 > E(CEO) <.001 E4 > E(CEO) <.001 F1 > F(BPR) F2 > F(BPR) <.001 F3 > F(BPR) <.001 F4 > F(BPR) <.001 F5 > F(BPR) <.001 F6 > F(BPR) <

24 The measurement model results provide evidence that the underlying measures and associated factors should be included in the structural model that we proposed. We now turn to the main structural model results. Structural Model The results of the structural model are shown below (p-values are based on a two-tail test). Hypothesis Dependent Associated Factor Factor Estimate S.E. C.R. p-value H8-reject A (Performance) < D (Integration) H7-fail to reject A (Performance) < C (Effectiveness) H2-reject C (Effectiveness) < E (CEO) <.001 H4-reject C (Effectiveness) < F (BPR) <.001 H3-reject D (Integration) < E (CEO) H5-reject D (Integration) < F (BPR) Our main findings are: (1) Post-implementation systems integration significantly and positively affects a company s operational performance and therefore supports the rejection of H8. (2) ERP effectiveness does not show a significant effect on operational performance at traditional alpha levels. The one-tail p-value is 6.35% and would be deemed significant by some. However, we conclude that we fail to reject H7. (3) CEO s involvement in ERP implementations positively and significantly affects both ERP effectiveness and post-implementation systems integration. Thus we reject both H2 and H3 in favor of the one-tail alternative hypothesis. (4) BRP factors positively and significantly impact both ERP effectiveness and postimplementation systems integration. Thus we reject both H4 and H5 in favor of the onetail alternative hypothesis. 21

25 CONCLUSION AND FURTHER RESEARCH This investigation makes a contribution by providing empirical evidence to support theories on ERP implementations, post-implementation ERP success, integration, process change, and operational performance of these systems for public manufacturing companies in China. One limitation of this study is that it relies on perceptual inputs from survey respondents. Another is that by not limiting the survey to a single ERP vendor s implementations in China, we do not limit any variation in the ERP systems themselves (Reimers 2003: 339). A further limitation relates to the use of SEM. This methodology may lead researchers to only report those fit measures which fall within the acceptable range, and not those that show an unacceptable fit (Smith and Langfield-Smith 2004: 57). By using the SEM modeling generating strategy, we mitigated this limitation by revising our conceptual model. The next stage of this research is to test the revised conceptual model shown in Figure 2 to include items B1 to B6 (Business Process Improvement) of Appendix A, which currently do not show an acceptable fit. The next two stages of this research program will examine both accounting and market performance measures as possible validating measures of perceived operating performance. Consequently, measures from three levels will be examined: managerial perception, firm performance and market response to these ERP implementations. This three-prong approach enables us to triangulate results regarding the perceived and actual factors affecting the efficacy of the ERP implementations. 22

26 REFERENCES Akkermans, H. and K. van Helden (2002). Vicious and Virtuous Cycles in ERP Implementations: A Case Study of Interrelations Between Critical Success Factors. European Journal of Information Systems 11: Aladwani, A.M. (2001). Change Management Strategies for Successful ERP Implementation. Business Process Management Journal 7(3): Al-Selali, S. (2000). The Factors That Affect the Implementation of Enterprise Resource Planning (ERP) in the International Arab Gulf States and United States Company With Special Emphasis on SAP Software. Dissertation, University of Northern Iowa. Bancroft, N.H., H. Seip and A. Sprengel (1998). Implementing SAP R/3: How to Introduce a Large System into a Large Organization (2 nd Ed.). Greenwich, CT: Manning. Benco, D.C. and L. Prather (2008). Market Reaction to Announcements to Invest in ERP Systems. Quarterly Journal of Finance and Accounting 47(4): Bingi, P., M.K. Sharma and J.K. Godia (1999). Critical Issues Affecting an ERP Implementation. Information Systems Management (Summer): Brown, D.H. and S. He (2007). Patterns of ERP Adoption and Implementation in China and Some Implications. Electronic Markets 17(2): Cheng, D., F. Deng and H. Li (2006). Critical Success Factors for Successful Implementation of ERP in China. IEEE International Conference on e-business Engineering (ICEBE 06). China: Shanghai. Davenport, T.H. (1998.) Putting the enterprise Into the Enterprise Systems. Harvard Business Review 76(4): Deloitte Consulting (1998). ERP s Second Wave: Maximizing the Value of ERP-Enabled Processes. New York, NY: Deloitte Consulting. Delone, W.H. and E.R. McLean (1992). Information Systems Success: The Quest for the Dependent Variable. Information Systems Research 3: Dezdar, S., and A. Sulaiman (2009). Successful Enterprise Resource Planning Implementation: Taxonomy of Critical Factors. Industrial Management & Data Systems 109(8): Doll, W.J. and G. Torkzadeh (1988). The Measurement of End-user Computing Satisfaction. MIS Quarterly 12(2): Dong, L. (2001). Modeling Top Management Influence on ES Implementation. Business Process Management Journal 7(3):

27 Gartner, Inc. (2010, March 18). Gartner says China will be World's Fastest Growing Enterprise Software Market Through Press Release: Hong Kong, SAR China. [Accessed June 29, 2010]. Gattiker, T.F., and D.L. Goodhue (2002). Software-driven Changes to Business Processes: An Empirical Study of Impacts of Enterprise Resource Planning (ERP) Systems at the Local Level. International Journal of Production Research 40(18): Gattiker, T.F., and D.L. Goodhue (2005). What Happens After ERP Implementation: Understanding the Impact of Interdependence and Differentiation on Plant-level Outcomes. MIS Quarterly 29(3): Ge, L. and S. Voß (2009). ERP Application in China: an Overview. International Journal of Production Economics 122: Hammer, M. and J. Champy (1993). Reengineering the Corporation: A Manifesto for Business Revolution. New York: Harper Business Books. Hayes, D.C., J.E. Hunton amd J.L. Reck (2001). Journal of Information Systems 15(1): He, X. (2004). The ERP Challenge in China: A Resource-based Perspective. Information Systems Journal 14: Hendricks, K.B., V.R. Singhal and J.K. Stratman (2007). The Impact of Enterprise Systems on Corporate Performance: A Study of ERP, SCM, and CRM System Implementations. Journal of Operations Management 25: Henri, J-F. (2007). A Quantitative Assessment of the Reporting of Structural Equation modeling Information: The Case of Management Accounting Research. Journal of Accounting Literature 26: Huang, S.Y., S-M. Huang and T-H. Wu (2009). Process Efficiency of the Enterprise Resource Planning Adoption. Industrial Management & Data Systems 109(8): Hunton, J.E., A. McEwan and B. Wier (2002). The Reaction of Financial Analysts to Enterprise Resource Planning Implementation Plans. Journal of Information Systems (Spring): Hunton, J.E., B. Lippincott and J.L. Reck (2003). Enterprise Resource Planning Systems: Comparing Firm Performance of Adopters and Nonadopters. International journal of Accounting Information Systems 4: IBISWorld Inc. (2010, February 16). IBISWorld Industry Report: Software Publishing in the US. Ip, W.H. and B. Chen. (2004). An Enterprise Model and the Organisation of ERP. International Journal of Computer Applications in Technology 21(3):

28 Jafari, S.M., M.R. Osman, M.Y.Rosnah and S.H.Tang (2009). A Consensus on Critical success Factors for Enterprise Resource Planning Systems Implementation: An Experience of Malaysian Firms. International journal of Manufacturing Technology and Management 17(4): Jöreskog, K.G. (1993). Testing Structural Equation Models. In K.A. Bollen and J.S. Long, Testing Structural Equation Models: Newbury Park: Sage Publications. Kallunki, J-P., E.K. Laitnen and H. Silvola (2010, in press). Impact of Enterprise Resource Planning Systems on Management Control systems and Firm Performance. International Journal of Accounting Information Systems. Kamhawi, E.M. and A. Gunasekaran (2009). ERP Systems Implementation Success Factors: IS and Non-IS Managers Perceptions. International Journal of Business Information Systems 4(6): Koch, C. (2001). BPR and ERP: Realising a Vision of Process with IT. Business Process Management Journal 7(3): Law, C.C.H. and E.W.T. Ngai (2007). ERP Systems Adoption: An Exploratory Study of the Organizational Factors and Impacts of ERP Success. Information & Management 44: Martin, M.H. (1998). An Electronics Firm Will Save Big Money by Replacing Six People With One and Lose All This Paperwork Using ERP Software. Fortune 132: Martinsons, M.G. (2005). Transforming China: Introduction. Communications of the ACM 48(4): Matolcsy, Z., P. Booth and B. Wieder (2005). The Economic Benefits of Enterprise Resource Planning Systems: some Empirical Evidence. Journal of Accounting and Finance 45: Motwani, J. and A.Y. Akbulut (2008). Organisational Factors for Successful Implementation of ERP Systems. International Journal of Business Information Systems 3(2): Nah, F.F., J.L. Lau and J. Kuang (2001). Critical Factors for Successful Implementation of Enterprise Systems. Business Process Management Journal 7(3): Ngai, E.W.T., C.C.H. Law, and F.K.T. Wat (2008). Examining the Critical Success Factors in the Adoption of Enterprise Resource Planning. Computers in Industry 59: Nicolaou, A.I., T. Stratopoulos and B. Dehning (2003). Financial Analysis of Potential Benefits of ERP Systems Adoption. Journal of Business and Information Technology 2(1):

29 Nunnally, J.C. (1978). Psychometric Theory (2 nd Ed.). New York: McGraw-Hill. O Leary, D.E. (2004). Eneterprise Resource Planning (ERP) Systems: An Empirical Analysis of Benefits. Journal of Emerging Technologies in Accounting 1: Poston, R. and S. Grabski (2001). Financial Impacts of Enterprise Resource Planning Implementations. International Journal of Accounting Information Systems 2: Reimers, K. (2003). Implementing ERP Systems in China. Communications of the Association for Information Systems 11: Rockhart, J.F. (1979). Critical Success Factors. Harvard Business Review (March-April): Smith, D. and K. Langfield-Smith (2004). Structural Equation Modeling in Management Accounting Research: Critical Analysis and Opportunities. Journal of Accounting Literature 23: Tsai, W-H., Y-W. Fan, J-D. Leu, L-W. Chou and C-C. Yang (2007). The Relationship Between Implementation Variables and Performance Improvement of ERP Systems. International Journal of Technology Management 38(4): Tsai, W-H., P-Y Hsu, J.M-S. Cheng and Y-W. Chen (2006). An AHP Approach to Assessing the Relative Importance Weights of ERP Performance Measures. International Journal of Management and Enterprise Developments 3(4): Umble, E.J. and M.M. Umble (2002). Avoiding ERP Implementation Failure. Industrial Management 44(1): Wieder, B. P. Booth, Z.P. Matolcsy and M-L. Ossimitz (2006). The Impact of ERP Systems on Firm and Business Process Performance. Journal of Enterprise Information Management 19(1/2): Wier, B., J. Hunton and H.R. HassabElnaby (2007). Enterprise Resource Planning Systems and Non-Financial Performance Incentives: The Joint Impact on Corporate Performance. International Journal of Accounting Information Systems 8: World Bank (2010). China Quarterly Update June World Bank Office, Beijing, China. Xue, Y., H. Liang, W.R. Boulton and C.A. Snyder (2005). ERP Implementation Failures in China: Case Studies with Implications for ERP Vendors. International Journal of Production Economics 97: Yang, C. and Y. Su (2009). The Relationship Between Benefits of ERP Systems Implementation and its Impacts on Firm Performance of SCM. Journal of Enterprise Information Management 22(6):