اندازه گیری بهره وری نسبی شرکت های طراحی مدار یکپارچه با استفاده از تابع فاصله جهت دار و یک رویکرد فرا مرزی
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|4632||2012||11 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Decision Support Systems, Volume 53, Issue 4, November 2012, Pages 881–891
This paper presents an alternative approach for evaluating the efficiency of integrated circuit (IC) design firms. In doing so, it accounts for differences between technology groups containing one or more design firms, and input and output factors to prevent influences of scale (e.g., firm size). Specifically, we employed a directional distance function approach to data envelopment analysis in order to evaluate inefficiency scores and differences among groups based on input and output factors. We found the efficiency of Taiwan's IC design firms to be dependent not only on firm size but also on R&D expenditure and patent revenue. Our findings suggest that these factors significantly influence the technical efficiency of Taiwan IC design. Furthermore, by focusing on technology gaps, we offer some suggestions for the different groups based on group-frontier and meta-frontier analyses. Finally, using the results of these analyses, we extend the global results of this study, presenting ways to further improve efficiency.
The integrated circuit (IC) design industry,2 also called the “fab-less industry,” plays a key role in the overall semiconductor industry, through its production of highly “intelligent” properties (i.e., patents). Within the semiconductor industry, IC design industry acts as a knowledge-intensive business service division. The IC design industry in Taiwan includes four types of designers: the independent professional design house, the design department within an integrated device manufacturer (IDM), the IC design center in a system vendor, and the design unit of an overseas company. Due to the diverse nature of these design firms, it is difficult to evaluate industry efficiency in Taiwan, particularly because each designer upholds different goals. Moreover, the products produced by these firms also have diverse characteristics, making it difficult to measure the knowledge capabilities of any particular firm. Regarding IC design industrial cost structures, research and development (R&D) expenditures sometimes account for approximately 9–15% of total costs, while manufacturing outsourcing accounts for approximately 50–90%. In addition, the industry sometimes has a winner-takes-all attitude, which means higher gross profits may be concentrated in one or two large companies that hold the patents to a wide range of techniques to avoid the diffusion of necessary knowledge. As such, competition is fierce among IC design firms dealing with the same application, and lax elsewhere. High technology barriers exist for various applications, and firms find it difficult to break into new unfamiliar territories. IC design firms also need to have a good relationship with their downstream partners, because they do not have factories. This allows them to reduce their manufacturing costs and focus all their attention on IC design. Six of the thirty largest IC design firms in the world can be found in Taiwan. In 2006, Taiwan's IC design firms accounted for the second-largest market share (18%) in the world, coming just after the United States (72%), with a production value of nearly US $9.8 billion. In addition, some of Taiwan's leading firms have captured more than 50% of the world market share in specific application domains, including network chipsets (e.g., RealTek), optical storage chipsets (e.g., MediaTek), and consumer market chips (e.g., SunPlus), to name but a few. The ascendancy of these firms on the world market makes it crucial to evaluate their operational efficiency and better understand the feedback processes needed to improve operational performance. As Taiwan's IC design firms vary in terms of size and scope, we cannot use only one variable group (e.g., variables representing different countries) to evaluate efficiency. Chen and Chen  assume that the groups within different countries had the same technology set. When the frontier was applied to each country, the performance of each individual decision-making unit (DMU) was compared against the best practices in the same country, but the efficiency results between two different countries are incomparable. In practice, it is rare for the estimated frontiers of two countries/regions to be similar enough to facilitate the use of a single frontier.3 As such, it is necessary to analyze these firms in several groups to ensure and improve evaluation accuracy. Most previous studies ,  and  have examined the efficiency of various firm groups by employing different approaches that use a single frontier to compare group efficiency. These studies assume that the different groups possess the same technology.4 However, IC design firms within different groups have different available resources (e.g., investment capability), size, scope, and characteristics (e.g., operation philosophies and managerial modes); thus, they have different technology sets.5 As a result, we used grouping techniques (i.e., fuzzy c-mean) to cluster Taiwan's IC design firms, rendering the similarities of the groups of IC design firms in Taiwan easier to observe. In addition, we used a meta-frontier and group-frontier6 to examine the efficiency between the different groups of IC design firms and the IC design industry. More precisely, we not only measured the inter-groups of IC design firms with similar scales and/or scopes, but also the intra-group relations between IC firms with different scales and/or scopes. Furthermore, previous studies have employed an input or output model to analyze the operations of IC design firms . However, these models are not really suitable for analyzing the efficiency of an IC design firm because an IC design firm should simultaneously maximize outputs and minimize inputs instead of focusing on only one action. While evaluating the efficiency of an IC design firm produces projections by only considering input- (output-) oriented perspective, the efficiency score is still affected by the output (input) factor. Moreover, these studies have neglected the (quasi) fixed inputs that may overstate a firm's capacity for adjustment, thus producing misleading results . Fixed (or, quasi-fixed) inputs prevail in all sectors of the economy, so their optimal value cannot be adjusted within a given period. To remedy this problem, we employed the directional distance function in this study. The directional distance function is an analytical tool used for measuring the input and output-orientated technical efficiency of different firms. Furthermore, the directional distance function programming model is linear, and the direction in which performance is measured can be specified to accommodate different analytical purposes. The contributions of this research to IC design firm evaluation are threefold. First, we use the directional distance function to assess the operational efficiency of IC design firms in Taiwan. Using this function, we consider factors related to output slack, input slack and (quasi) fixed input constraints. Second, we use data envelopment analysis (DEA) concepts with a meta-frontier to estimate efficiency in the IC design industry, whereas previous studies have used conventional DEA and other methods (e.g., analytic network process (ANP) and analytical hierarchy process (AHP)) that did not take into account the group concept. Third, we adopt the technology gap7 concept to measure the difference between IC design firms and the IC industry. With these perspectives, we provide directions to specific IC design firms (or groups) on how to adjust their resources to reach maximal efficiency. The remainder of this paper is organized as follows. Section 2 presents our review of the literature on the IC industry's performance evaluations. Section 3 describes our methodology and explains the rationale behind it. Section 4 reports the empirical results from a study of 87 IC design firms in 2008. Section 5 summarizes our findings and considers their theoretical and managerial implications. Finally, Section 6 offers our conclusions and our suggestions for future research.
نتیجه گیری انگلیسی
This study describes a DEA application for evaluating IC design firms using technology grouping. A four-phase approach was proposed and applied to the Taiwan IC design industry, offering a number of suggestions for the different groups of IC design firms. Our results show that the different groups of IC design firms should use different management policies. Furthermore, in our empirical studies, the meta-technology of IC design firms can be viewed as industrial technology, while the group-technologies are regarded as revealed technology. In other words, we measure group efficiency based on a similar scale of IC design firms, while evaluating meta-efficiency based on industrial technology. This implies that existing technology can be upgraded through spillover and/or mutual learning among groups, as shown in our discussion. However, our analysis has certain limitations. First, according to the accrual accounting rule, leading and lagging variables should be considered. Second, the issue of how many groups should be included was not adequately addressed. Researchers with different perspectives may disagree on whether or not the groups would result in four cases. This study showed that a group effect exists in homogeneous decision-making units, as proven by our empirical evidence. We also did not consider scope variables in our empirical study, although we have elaborated on such a concept. These limitations can be considered in future works. Moreover, for future work, our research design could be applied to different contexts, such as financial institutions, educations resources, or the global IC design industry. We hope this study provides value for further research on efficiency evaluations of IC design firms whom face concerns regarding their scale or scope.