Decision making process for Low Cost Country Sourcing based on Total Cost of Ownership: a case study in European plastic pipe and fitting industry

University of Groningen Faculty of Economics & Business

Master Thesis,

MSc Business Administration,

Operations & Supply Chains

Decision making process for Low Cost Country

Sourcing based on Total Cost of Ownership: a case

study in European plastic pipe and fitting industry

Final Version

First supervisor: Prof. Dr. D.J. Kamann

Second supervisor: Dr. S. Brinkman

28

th

September, 2009

Zhida Xu

Abstract

This study about Low Cost Country Sourcing (LCCS) intends to answer

two questions: 1) Which countries offer low cost sourcing opportunities?

And 2) Is it attractive to purchase from that country in the medium and

long term? The case study method is used in one European plastic pipe

and fitting manufacture and it addresses the following sub research

themes: product selection for low cost country sourcing, low-cost

country selection based on Total Cost of Ownership principle and

sustainability of the low cost. These issues are integrated into a

decision process model for LCCS. The main finding of the study is that,

among 21 evaluated low-cost countries, India, Indonesia, Egypt, Philippines, and Tunisia are found to be attractive as destination of LCCS, and their low cost condition are sustainable in the medium term (3 to 5 years). Vietnam, Morocco, Turkey, Argentina, Mexico, and Brazil also have the potential. China is an ideal destination compared with other low-cost countries, but its window of opportunity as a low-cost country is closing. Some organizational problems regarding purchasing practice were identified within the case company during the research process. They are analyzed and recommendations are provided. The decision process model provides management a “one stop shopping” for LCCS analysis, and it can be used for repetitive propose.

Key words: Low Cost Country Sourcing, Total Cost of Ownership,

Acknowledgement

I would like to express my appreciation to my thesis supervisors:

Prof. Kamann and Dr. Brinkman, for their guidance on theory and

methodology throughout this graduation project.

I would like to appreciate Mr. Richard van Delden from SMIT for

giving me this opportunity to do this graduation project in the

company. And I would like to express my special thanks to Mr. Olle

Klaasen, my coach in SMIT, for his support and help during this

project.

Table of Contents

1. Introduction ...5

2. Theoretical Framework ...8

2.1 Defining Low Cost Country Sourcing (LCCS) ...8

2.2 Total Cost of Ownership and Low Cost Country Selection criterion...9

2.3 Product selection for Low Cost Country Sourcing...12

2.4 Sustainability of the low cost ...15

2.5 LCCS decision process model ...19

3. The Case ...21

4. Product selection for LCCS...23

5. Low cost country selection...30

6. Sustainability of low cost country sourcing...40

7. Discussion of results ...44

7. 1 Discussion of LCCS country selection ...44

7.2 Discussion of organization issue...47

8. Conclusion ...51

Reference ...54

Appendix 1 Original score for LCCS country selection ...58

Appendix 2 Sensitivity Analysis of country selection models ...61

Appendix 3 Break-down analysis for each low cost country’s performance...63

Appendix 4 Threshold of assigning scores to demographic bonus study...68

1. Introduction

International purchasing and outsourcing have become pervasive business activities for multinational enterprises (MNEs) thanks to the wave of globalization and the progressive elimination of trade barriers. Given that multinationals are facing more and more cost pressures, it is a tendency that MNEs search for resources and reallocate production in countries with low cost. Research from Aberdeen Group (2005) shows that Low Cost Country Sourcing (LCCS) is becoming a common strategy, and through 2008 the average of the total spending for direct materials with low-cost country suppliers will almost double, from 21% to 39%. Sourcing from low cost countries can benefit the bottom line and improve companies’ competitiveness. According to a report from IBM Global Business Service (2006), total savings up to 40% on purchased goods can be realized.

Regarding the motivations of Low Cost Country Sourcing, besides the advantages of low cost, companies can focus on core business, benchmark competitors in their outsourcing endeavours, own a broad supply base, access to better quality and establish a sales footprint for future market entry (Cater et al, 2005; Andersson et al, 2007). On the other hand, LCCS strategy also involves lots of risks. Some of the commonly identified risks include supply disruption, long lead times due to far transportation, security issues due to political instability in the host country and “hidden cost” such as foreign exchange control, tariffs, duties and other kind of taxes (Fitzgerald, 2005). Besides, some indirect risks come from communication difficulty because of culture difference, insufficient intellectual property protection, spillover effect of technical know-how and corporate social responsibility issues such as use of child labor in the low cost countries. Poor quality is also found to be one of the concerns, which contradicts with previous discussions as an advantage. This shows that the quality level in low cost countries is not consistent.

and Pearl River Delta in China increased significantly since 2006, and some manufactures even had difficulties to recruit workers (Business Week, 2006). In addition, due to fluctuation of oil prices, oversea transportation cost also increased. The question of “is LCCS still viable” emerged.

All of the risks mentioned above give the caution that LCCS need to be carefully managed. When companies choose to do low cost country sourcing, it is important to look beyond the direct price paid to suppliers, because there are lots of “hidden costs”, such as inventory build-up due to long lead time, machine down time caused by bad quality. Also, activities such as supplier selection, negotiation, contracting, communication, sample testing, quality checking are very time consuming. In order to take a comprehensive view to evaluate cost, the framework of Total Cost of Ownership (TCO) provides useful guidance. If management does not take a total cost view to do Low Cost Country Sourcing, it is possible to end up with negative total savings.

There has been lots of intensive discussion about Total Cost of Ownership in the literature (Ellram, 1993, 1995, 1996; Ellram and Maltz, 1995; Olsen and Ellram, 1997; Ferrin and Plank, 2002). But limited study applies this view in the situation of low cost country sourcing. Especially, it is necessary to study how the low cost countries should be selected. Therefore, the objective of this paper is to develop a model for LCCS country selection based on Total Cost of Ownership. Secondly, given that LCCS should be a long term strategy, it is also important to investigate for how long the low cost sourcing could be sustainable. The main research questions are:

1) Which countries offer low cost sourcing opportunities?

2) Is it attractive to purchase from that low cost country in the medium and long term?

of selection criterion are summarized based on current literatures, and a matrix calculation method is applied by using data from a questionnaire survey to purchasing managers. Finally, these different steps of evaluation are integrated into a decision process model. By applying this model, companies can select the proper product for low cost country sourcing, identify attractive low-cost countries for those products and assess for how long can it is beneficial to purchase from that country. This is a generic model which can be used for repetitive purpose and long term decision making.

The main finding from this study is that, among 21 evaluated low-cost countries from different areas of the world, India, Indonesia, Egypt, Philippines, and Tunisia are found to be attractive as destination of LCCS, and their low cost condition are sustainable in the medium term (3 to 5 years). Vietnam, Morocco, Turkey, Argentina, Mexico, and Brazil also have the potential. China is indeed an ideal destination compared with other low-cost country, but it is losing the low cost attractiveness gradually in the medium term. During research, certain organizational problems concerning purchasing practice emerged and they are analyzed and discussed separately.

The General recommendation to management is that it is better to purchase from big countries (in the respect of territory and scale of economy) because (1) big countries have relatively abundant resources and (2) the social-economic development in big countries are normally unbalanced, which provide more possibilities to prolong the process of cost increase.

2. Theoretical Framework

2.1 Defining Low Cost Country Sourcing (LCCS)

Global sourcing, international purchasing and low cost country sourcing, as names of business practices, are interchangeable for business practitioners. However, to be rigorous for the research, they should be clearly distinguished, because these terms do differ from scale and scope. Starting with the term “sourcing”, according to Kotabe and Murray (2004, page 9), sourcing is used to describe management the flow of component and finished products in serving foreign and domestic markets. Global sourcing is a broad term. It is a discussion of sourcing on a more strategic level. Trent and Monczka (2003) defines global sourcing as the worldwide integration of engineering, operations and procurement centres within the upstream portion of a firm’s supply chain. They also provide a continuum model of different stages of worldwide sourcing (Table 1), and according to this continuum, international purchasing, as a general term, belongs to level 2 and 3, while a true global sourcing practice (level 5) cannot be done only by the purchasing function.

Level 1 Level 2 Level 3 Level 4 Level 5

Engage in domestic purchasing only Engage in international purchasing as needed International purchasing as part of sourcing strategy Integration and coordination of global sourcing strategies across worldwide locations Integration and coordination of global sourcing strategies with other functional

groups

Table 1. A continuum of different levels of worldwide purchasing practice

(Trent and Monczka, 2003)

outsourcing and offshoring are confused. Based on the classification above, outsourcing is a make-or-buy decision, while offshoring is outsourcing from a foreign country. Javalgi, dixit and Scherer (2009) categorized offshoring business models into: “captive offshoring”, “offshore outsourcing” and “offshore development centres”. Among these, the strategy behind offshore outsourcing is to create value though the low cost, most of the time in low-cost emerging economies such as China and India. This idea is consistent with the Low Cost Country Sourcing practice.

There are some discussions about Low Cost Country Sourcing in business literatures (Fitzgerald, 2005; Timmermans,2005; IBM global business services, 2006; Spekman, 2008) and some discussions about China sourcing in academic literatures (Nassimbeni and Sartor, 2004; Asta, 2005, 2006; Millington, Eberhardt & Wilkinson, 2006; Bankvall and Fredriksson, 2007), but there is not a consistent definition for Low Cost Country Sourcing. Based on literature review, in this study, it is defined as A business practice that

multinationals in developed economies purchase materials, components, products or services from low-cost emerging economies. It will be

referred as LCCS in the rest of the paper.

2.2 Total Cost of Ownership and Low Cost Country Selection criterion

To assess which low cost country to purchase from on a macro level involves a lot of considerations. Besides the purchasing price of product from that country, factors such as the capability of industry sector, product quality and transportation time/cost are important. This comprehensive way of evaluation complies with the modern principle of purchasing which is called Total Cost of Ownership.

objectives besides cost price, such as quality, speed, dependability and flexibility (Slack and Lewis, 2001), since bad performance on these aspects will also cost money. Since 1980s’ there has been some development and evolution of methods to find the “total price” for purchasing. Some contribution include the concept of Total Cost (Cavinato 1991, 1992), Life cycle costing (Jackson and Ostrom, 1980), Product life cycle costs ( Shields and Young, 1991) and Total Cost of ownership (Ellram 1993, 1994, 1995). These concepts are all related and share the same belief that purchasers should not only focus on the face price, but they should take into account other factors in the purchasing activities in order to derive the total price. Among those concepts, Total Cost of Ownership (TCO) has been widely accepted and Ellram did a serial of contribution in this field.

According to Ellram and Siferd (1993, page 164), Total Cost of Ownership implies that all costs associated with the acquisition, use and maintenance of an item be considered in evaluating that item and not just the purchase price. Ellram (1993) classified the total cost of purchasing into three categories: 1) Pre-transaction cost, such as investigating and qualifying sources, selecting and educating suppliers; 2) transaction cost, such as price, order placement, delivery, tariffs/ duties, billing, inspection and follow-up; 3) Post-transaction cost, such as failure cost, return and repair costs, etc. It is believed that taking a TCO perspective not only helps to disclose the “real price” or “hidden cost” in purchasing activities, and it is also a valuable method to analyze purchasing process in order to find improvement opportunities.

Secondly, competence and quality are important dimensions and they are related. Competence measures what the industrial sector can do, namely can the manufacturers in that low cost country only provide assembly service or they can master state-of-art technology which is driven by strong R&D capability. Quality measures how good the performance is. However, quality itself is difficult to measure. Normally it needs to be judged based on real product, let along measuring the product quality on a country level. Therefore, the logic here is to exam the quality management system among the manufactures. ISO certificate is a widely accepted measurement of quality management system; therefore, percentage of ISO certificate ownership of manufactures is an acceptable indicator.

Thirdly, the importance of logistics does not need to be over emphasised. Besides the direct transportation cost and lead time, time to export, such as time to pass custom sometime can cause delay of responsiveness. The facility of infrastructure is also need to be checked. For international purchasing, duty and tariff are important concern, especially the anti-dumping tariff and safeguard measures. These special tariff could suddenly take away all the saving potential.

Figure 1 Low Cost Country selection criterion

2.3 Product selection for Low Cost Country Sourcing

purchasing function with the general business strategy of a company. They maintain that companies as an open system formulates a business strategy based on external environment and internal competence. This strategy will be translated into certain Policy (P), and the Organizational design (O) and business Process (P) of each function in the company. Since purchasing has transformed from only a business function into strategically important activities, the purchasing policy, purchasing organization and purchasing process should be coincide with the general business strategy. Therefore, in order to select the item to purchase from outside parties, the first consideration will be what the core competence is, so the company should keep in-house, and what should be purchased in order to better use the specialty and other benefits from the suppliers.

On a micro level, Andersson et al. (2007) used a multiple case study method to link product characteristics and low cost country sourcing. They maintain that Volume of the sourced product and the demand pattern are the dominant forces to be studied in LCCS. Their primary finding is that products considered for LCCS are characterised by high volumes, low level of design changes, regular demand/regular shipments and high value density. Meredith Smith (1999) summarized the selection criterion into six categories, namely product specification and its change, product technology and its change, impact of quality failure, impact of logistics, product criticality and volatility and product/delivery cost.

inevitable, it seems reasonable that a simple, functional product with an efficient supply chain is appropriate.

Beside the above factors, the order frequency and bath size should also be considered. Product needed on a daily or weekly basis is better sourced locally to prevent delivery disruption and inventory build-up. On the other hand, items that are only purchased on a once a year scale is not proper neither, since a stable relationship with suppliers is hard to maintain and supplier search is very time-consuming. Large batch size is necessary in order to reduce unit transportation cost. Furthermore, quality is always a big concern for purchasers, especially under low cost country sourcing, because there are certain stereotypes that product from low cost countries are not satisfied. Therefore, it is better to choose those products which have less probability of quality failure, high tolerance towards quality failure and less application in the whole production process.

Figure 2 Low Cost Country Sourcing product selection criterion

2.4 Sustainability of the low cost

Why the production costs are low in certain countries? Why the low cost is relatively stable in certain period? And why after that period the production costs increases rapidly? These phenomenon are witnessed in the Far East world after World War II. The “demographic bonus” theory in demography provides some systematic explanations. “Demographic bonus” is in the form of a large group of working-age people supporting relatively fewer older and younger dependents that creates a one-time opportunity for growth (window of economic opportunity), may have accounted for as much as a third of the East Asian economic miracle (The Economist, March 15th 2003). It explains the relationship between Age Structure Transition (AST) of the population in a country with its economic development.

There are two factors which influence age structure transition, namely fertility rate (the average number of births per woman per lifetime) and mortality rate (the total number of deaths per year per 1000 people). Furthermore, the population is classified into 3 age groups: children (0~14 years), working age adult (15~64 years) and elderly (65+ years). One key indictor which monitors age structure transition is the dependency ratio. The “total dependency ratio” is calculated by using the sum of children and the elderly divided by amount of working age adult.1 According to research (Sedano, 2008; Wong, 2005; Navaneetham, 2002; Mari Bhat, 2001; Hussain, 2002, Chu & Lee, 2000), the process of age structure transition goes through several stages: I. Mortality begins to decline, causing an increase of children in the population, which increases the total dependency ratio; II. Fertility then begins to decline, which initiates a period of declining child-dependency ratios and declining total dependency ratios. Working Age Adult accounts for the largest proportion while life expectancy keeps increasing. III. Finally, the elderly population begins to increase, and then elderly dependency ratios and total dependency ratios rise again, so the society becomes aging. Fertility rate drops to around 2, which is a natural replacement for the population to stay stable. If the fertility rate is lower than 2, the total population of that country will decrease.

1

As an example, Figure 3 shows the Age Structure Transition in Brazil in the form of demographic pyramid.

Figure 3 Age Structure Transition in Brazil

Source: Wong, 2005, “Demographic bonuses and the challenges of the Age Structure Transition in Brazil”, data originally from United Nations (2003)

The process of age structure transition is a slow process in most developed countries, which took approximately two centuries to complete. But it is witnessed that this process happened faster in certain developing countries, especially after WWII, such as in East and Southeast Asia. There are some strong evidences which show that demographic bonus contributes significantly for economic development. Some economic miracles in Asia such as Japan and China are a typical Beneficiary of this bonus. Statistics show that demographic bonus contributes for 27% of the fast economic development of China in the past 20 years. Research also shows that for the whole Asia, the contribution of age structure change to economic growth from 1970 to 1990 is around 14%. (Bloom, Canning and Malaney, 1995, appendix table 4b) However, demographic bonus is only a necessary but not sufficient condition for economic take-off. If the policy makers of the country cannot create enough employment opportunities for the large labor supply, it will lead to massive unemployment, poverty, rampant crime, civil unrest and other social-economic problems.

Because the whole population will gradually move to stage III, the demographic bonus is a “window of opportunity”. It will happen once and only once. A demographic window opens as the numbers of younger children decrease, and the signal of diminishing window is the rise of total dependency ratio which is contributed by the rise of elderly dependency ratio. When the window is closed, the bonus will become onus, because of the aging of the large amount of Working Age Adults. In that case, the economy will have heavy burden due to social security cost to elderly people. Therefore, it is very important for the countries which are still in the bonus window to build a proper social security system.

demographic bonus in the urban areas, while provide opportunities for the poor rural areas to be able to harvest demographic bonus, which results in a win-win situation.

The implication to policy makers of developing countries is that, they must develop a proper strategy for the country to fully utilize this one time opportunity. It is important to create sufficient demand by open up the economy and integrate with international trade. By exporting cheap labour, the low cost country can accumulate capital for future development. For multinationals, the implication of Demographic Bonus does not restrict to international purchasing, but other activities such as outsourcing, local production and other forms of FDI.

To summarize the literature, the reason that production cost is low mainly because there is abundant supply of labour for a period of time, while total domestic demand does not increase dramatically. This low cost is relatively stable in a period due to the effect of Demographic Bonus. When the window of opportunity is closed, labour becomes a scarce production factor and therefore cost increase significantly. When the management of MNEs makes decision for LCCS, accessing the countries’ current cost level is a part of the input. What is more important is to investigate whether that country is enjoying demographic bonus and until when. When both sides of stories are heard, a final list of candidate countries could be generated.

2.5 LCCS decision process model

model, (4) analyze how sustainable the low cost in the future, (5) finalize the country list and search for suppliers, (5) based on real quotation price, start to develop suppliers in the low cost country. In the model, the process starts with a diamond and it ends with a triangle. Each box represents one major process. The incoming arrow from the left represents input, and the outgoing arrow under the box represents the output. The above arrow represents the control activities. The contents in the red dish line indicate the main research objectives of this study.

3. The Case

Given that there is limited literature to apply Total Cost of Ownership into Low Cost Country Sourcing, this study has an exploratory nature. Therefore, case study is one of the proper strategies. By using case study, the business practice in one company is systematically analyzed and the result is also influenced by that specific industry. To conduct the research, both qualitative and quantitative methods are used. Especially, a serial of interviews with structured and semi-structured questions were performed with the managers in the company. They are regarded as experts in their work fields. During the interviews, the Delphi method was applied. By using the Delphi method, common agreement of interviewees towards certain issues is used to derive principles, while different opinions are excluded.

The European plastic pipe industry is chosen to carry out the case study. The plastic pipe industry has growing potentials on the market, since plastic pipes are replacing previous generations of concrete and metal pipes. The research is carried out in a Dutch company called SMIT. SMIT Group is the leading European supplier of plastic pipe system with operating activities in 28 European countries and over 6700 employees. The company operates in several market segments in the plastic pipe industry, including tap water, surface heating and cooling, soil and waste, rain- and storm water, distribution of drinking water and gas and telecom applications. These businesses are organized in two Strategic Business Units: Building & Installation and Civil & Infrastructure. SMIT’s activities focus on Europe, but it has long history and extensive experience on global sales through agents and selling know-how through licensing.

classified as Differentiation, because it provides a package of product and service so that customers can do “one stop shopping” when they seek plastic pipe solutions. The high value-added service is supported by large investment in innovation, and being a service provider also puts SMIT the high end of the value chain. When looking at the nature of the pipe and fitting industry, the market demand depends on the housing and construction market. The change of product technology is relatively slow and the product life cycle is long. Once the system is installed, warranties are provided for up to 50 years. Besides, across Europe, each region (e.g. Southern Europe, North West Europe, Central Europe, Nordic Europe, etc.) has its own requirement for pipe system due to climate conditions or tradition of use. This requires different materials and product design issues. Given these factors, SMIT group is managed in a decentralized way and a large range of product portfolio is maintained to meet various market demands. However, this practice brings challenge for complexity management and optimization of the product portfolio.

4. Product selection for LCCS

According to the process model, several product items must be selected as carriers In order to carry out the research. Firstly, the purchasing managers from one region of the group had a brainstorm session and a list of product family was generated. Secondly, serial interviews were conducted to ask the opinions of purchasing managers from other regions. The results of the interviews act as a screening process to narrow down the product list to 5 product families. These are:

- Brass valve: normally with a brass body and nickel coating on the surface; - Brass fitting: mainly used for connecting pipes, similar material as the

valve;

- Grating: a kind of metal cover of the drainage, raw material is cast iron; - Rubber Ring: mainly used in water applications for sealing;

- Plastic end caps: insert at the end of a pipe to protect it, used for packing purpose.

NAICS Name of the

Industry Material Labour

Manufact.+ overhead General Sales + Adm. Profit before taxes 326122 Plastics Pipe and

Pipe Fitting Manufacturing 55% 9% 14% 20% 2% 339991 Gasket, Packing, and Sealing Device Manufacturing 38% 16% 18% 26% 2%

332919 Other Metal Valve and Pipe Fitting

Manufacturing _{41% 15% 15% 25% 4%}

331522 Nonferrous (except

Aluminum)

Die-Casting Foundries 31% 18% 33% 16% 2% 326199 All Other Plastics

Product

Manufacturing 40% 14% 23% 20% 3%

The industry-wide (according to North American Industry Classification System) cost models for these products are provided in table 2. As shown in the table, share of material cost for the proposed products are quite high. Their shares of labour cost are also higher than Plastics Pipe and Pipe Fitting Manufacturing. According to the product selection criterion discussed in section 2.3, it shows that it is wise to purchase those proposed products from low cost countries.

Afterwards, a survey is conducted to the purchasing managers by using a self administrated questionnaire. The questionnaire is designed based on the Smith method (1999). In his contribution, he proposed a method to evaluate the proper purchasing areas for products with different characteristics by using multiple case studies. Different domains of supply market include local sourcing (travelling distance within 50 km or 60 min.), national-wide sourcing, sourcing within trade-bloc (such as EU) and finally purchase globally. For global purchasing, purchasers could choose to buy directly or go through a local agent. A questionnaire was designed based on the six categories of considerations discussed in section 2.3. Respondents were asked to score from -10 to 10 for the questions, and the weighted average score was calculated. The results were plotted into a matrix, which is shown in figure 5. In the case of SMIT, this is a useful tool to assess whether the selected item are really suitable to purchase globally.

Figure 5 Smith matrix (Smith, 1999, page 124)

The questionnaire contains 18 questions. The purpose of the survey is threefold. The first 2 questions are designed to learn whether the product specification (e.g. drawing) and product cost model are available within the company, not for calculation. Questions 3 to 17 are used to gather the opinions towards 5 aspects: nature of product characteristics and its change, purchasing volume and frequency, impact of quality, volatility/ criticality and freight vs. product value density. The object of evaluation is the product family, not any specific item or article. Respondents are asked to give a rating from 1 (in favour of local sourcing) to 5 (in favour of low cost country sourcing), and the results will be converted to -2~2 in order to plot the coordinate into the matrix.

There are two aspects mentioned by Smith but are not included in this model, namely product technology and product availability. The reason is, according to the interview with managers, the production technology for the 5 proposed products are relatively basic in the manufacturing sector, such as forging, cast (moulding), extrusion moulding and injection moulding. The degree of complexity of these technologies does not vary significantly therefore it will have little influence on the survey result. Secondly, given that the technologies are relatively basic, they should be available in most of the industrialized and transition economies. Therefore, these two aspects are excluded from the model; instead, the issue of purchasing volume and order frequency are emphasized by the model, because for low value-added product, the way to realize saving from global sourcing is from high volume and large batch size.

Question category Question categoryQuestion category

Question category QuestionQuestionQuestionQuestion Dimension in Dimension in Dimension in Dimension in the matrix the matrix the matrix the matrix

commodity or customized X

simple or complex product X

change of specification Y

Product characteristics and its change

speed of product life cycle Y purchasing Volume per year X Purchasing Volume

and frequency order frequency Y

probability of quality failure Y degree of impact towards failure Y degree of tolerance to failure X Quality impact

easy of correction X

degree of demand fluctuation X delivery disruption impact X Volatility/criticality

Range of application of the product Y transportation volume per full container X Freight/product value

density purchasing price per weight Y

Non-random sampling is used to select the respondents of the questionnaire because they must be experts of the proposed purchasing items. The respondents include group lead buying managers who have a general view of corporate purchasing, and local strategic buyers who have more in-depth views of each region. Regarding the number of questionnaires, 5 were designed and sent out for brass valve, 6 for brass fitting, 4 for metal grating, 5 for rubber ring and 3 for plastic end caps. In total, 23 questionnaires were sent out to 11 purchasing managers and 21 respondents were received. One of them had insufficient answer so it was excluded for analysis. Therefore, a valid response rate is 87% which is sufficient.

Findings

The answers to question 18 shows the following result: regarding the top 3 consideration for LCCS, among the 11 purchasing managers, 8 of them chose Quality, 7 voted Purchasing volume and order frequency, 5 for product nature, 5 for Volatility/criticality and 3 for Freight/product value density. This gathered opinion is not product-specific but is in general for LCCS. The number of “vote” is used directly as the weight of its corresponding aspect. The results of question 3 to 16 were converted from the scale [1, 5] to [-2, 2]. Finally, the weighted average of X and Y dimensions were calculated for each product family, and the results are shown in Table 4.

Table 4 findings of the questionnaire based on Smith method

Product Product Product Product nature nature nature nature Volume/ Volume/Volume/ Volume/ frequency frequencyfrequency

frequency QualityQuality QualityQuality

Volatility/ Volatility/Volatility/ Volatility/ criticality criticality criticality criticality Freight/ Freight/Freight/ Freight/ product product product product cost costcost cost ∑ ∑∑ ∑ Valve ValveValve Valve 0.5,0.5 -1,-0.3 -1.67,0.17 -0.67,-0.7 1.7,0 -0.7,0.05 Fitting FittingFitting Fitting 1,-0.2 0.2,0.6 -0.8,-0.2 -0.5,0.2 1.6,0 -0.04,0 Gratin GratinGratin Gratingggg 0.5,0.75 -1.25,0.25 -0.5,0.36 -0.75,2 1.75,-2 -0.3,0.47 Rubber Ring

Rubber RingRubber Ring

Rubber Ring -0.13,0.63 1.75,-1 -1.63,-0.38 -1.13,-2 1.25,-0.25 -0.5,-0.4 Plastic end

Plastic end Plastic end Plastic end cap

capcap

The result in the above table shows that, first, purchasing managers think only plastic end cap is a very suitable product for global sourcing. Valve, fitting and grating should be purchased from national trade bloc, in this case, the EU. Rubber ring product is best to purchase from local suppliers. Second, the findings do not show a significant direction about proper purchasing areas, because the final coordinates are all very close to the origin point. Therefore, one conclusion from the calculation is that purchasing those 5 kinds of product globally is not strongly supported by the result. This is out of the expectation of the researcher, which could either caused by the design of research or the data input.

Going back to the original data, it is interesting to find that purchasing managers’ opinions towards whether the product should be sourced globally is heterogeneous. Respondents voted for the same direction on certain consideration aspects, but in some cases their opinions are completely opposite (e.g. one votes 2 while another votes -2). Table 5 shows the distribution tendency of data by calculating standard deviation for each question and each product. According to the table, in some cases, the standard deviation is around 1.5 within a scale of [-2, 2], which indicates that purchasing mangers do not have a consensus towards that issues.

13 1313 13 1.155 1.414 0.500 1.258 0.000 14 1414 14 0.577 0.707 0.957 0.000 1.155 15 1515 15 1.528 1.789 0.000 0.000 0.000 16 1616 16 0.577 0.548 0.500 0.957 1.000 17 1717 17 1.000 1.000 0.000 0.957 1.000

Table 5 Standard deviation of data for each question

The findings that purchasing those products from low cost countries is not significantly supported and the fact that these purchasers’ opinions are heterogeneous can be explained by the following reasons. First, low cost country sourcing is still in the initial phase within SMIT. Due to organizational inertia, some purchasers are not ready to move the supplier base or they are not fully convinced that LCCS can work within this organization. Second, SMIT’s purchasing activities are managed in a de-centralized way in order to adjust to requirement of local market. Each market has different product portfolio. Therefore, for the same brass fitting, for example, the volume could be huge in one region but limited in another.

5. Low cost country selection

It is difficult to generate a list of low cost countries by using one single quantitative threshold, e.g. GDP per capita, given that the concept and standard for “low cost country” is rather vague. Due to limited time for the study, it is planed to have a final list with 5 countries. Therefore, when generating the long list by looking around the globe, 4 areas are identified and around 5 countries were picked from each area. Table 6 shows the list of candidate countries.

Asia

Latin

America Mediterranean Area

Emerging Europe

China Argentina Morocco Bulgaria Indonesia Brazil Tunisia Romania

India Chile Egypt Croatia

Philippines Mexico Turkey Czech

Viet nam Poland

Slovakia Hungary Ukraine

Table 6 List of candidate countries for LCCS

Those countries were selected because they are frequently mentioned by business practitioner as destination of LCCS. China will be in the final list as a benchmark of comparison, in order to see whether other countries are better places for SMIT to purchase than from China. “Emerging Europe” mainly refers to former European socialism countries. These countries are joining the integration process of European Union and it is witnessed that their economies were growing fast in recent years.

countries need to be improved. Furthermore, each country has its own economic census standard and there is always a lag between latest data and present, which make it less comparable. Therefore, data of wage cost were collected instead, and data on non-wage cost were collected by other means in order to measure it indirectly.

International Labour Organization (ILO), United Nations Industrial Development Organization (UNIDO) and U.S. Bureau of Labour Statistics (BLS) provide good database on wage cost statistics. The average earning per month of employee (including all industries) based on local currency were collected from the 21 countries and the Netherlands (as benchmark). Time span includes 2001 to 2008 in order to find a comparable year, but the latest year of available data from different countries vary a lot. Then the market foreign exchange rates on Dec. 31st of each year were used to convert local currency into the Euro. The conversion process did not involve Purchasing Power Parity (PPP) adjustment, given that this study does not intend to equalize the living standards across countries, but concerns how much the employer should pay. Afterwards, the average monthly earning data are compared with the Netherlands. The result shows how much lower of monthly wage of that low cost country than the Netherlands. Table 7 presents the results of comparisons.

Monthly Monthly Monthly Monthly earning EUR earning EURearning EUR earning EUR % lower than % lower than % lower than % lower than NL NL NL NL Year of Year of Year of Year of 2222 reference reference reference reference Source Source Source Source of data of data of data of data Argentina Argentina Argentina Argentina 384 86 2006 ILO Mexico Mexico Mexico Mexico 310 89 2007 ILO Chile Chile Chile Chile 398 85 2005 ILO Brazil Brazil Brazil Brazil 463 83 2006 BLS Bulgaria Bulgaria Bulgaria Bulgaria 164 94 2006 ILO Romania Romania Romania Romania 280 90 2006 ILO Croatia Croatia Croatia Croatia 794 71 2006 ILO Czech Czech Czech Czech 671 75 2006 ILO 2

Hungary Hungary Hungary Hungary 683 75 2007 ILO Slovakia Slovakia Slovakia Slovakia 596 78 2007 ILO Ukraine Ukraine Ukraine Ukraine 86 97 2004 UNIDO Turkey Turkey Turkey Turkey 672 76 2006 other3 Morocco Morocco Morocco Morocco 571 79 2005 UNIDO Egypt Egypt Egypt Egypt 108 96 2006 ILO Tunisia Tunisia Tunisia Tunisia 305 88 2003 UNIDO India India India India 113 96 2004 UNIDO Vietnam Vietnam Vietnam Vietnam 39 99 2008 other4 China China China China 163 94 2007 ILO Philippines Philippines Philippines Philippines 128 95 2006 ILO Poland Poland Poland Poland 490 82 2006 BLS Indonesia Indonesia Indonesia Indonesia 81 96 2008 other5

Table 7 Result of wage data and comparison with Netherlands

The TCO model for low cost country selection has been discussed in section 2.2. When applied in SMIT’s case, the model is extended by using more detailed indicators. The complete model is shown in Table 8, together with the weight of each indicator. The weights were determined based on purchasing managers’ experience of current China sourcing practice. Statistics for the indicators are all from sound database, especially the Global Competitiveness Report 2008-2009 (World Economic Forum) which provides the rankings of very comprehensive indicators for 144 countries. Logistics data (cost and lead time) are based on real quotation from the transportation company.

There are two kinds of data: hard data are directly collected from the database, and index data are extracted from the Global Competitiveness Report 2008-2009 (country ranking from 1~144). Then these original data are converted into a score within a range of 1 ~ 10. The original score for each country is attached in the appendix 2

3 _{Turkish Statistical institute, press release 2006} 4

Table 8 Low Cost Country selection model: indicator and weight Model

dimension

Indicator Direct measure Type of

data

Source of data

Wage cost level

(70%) Wage per month,% lower than NL Hard ILO, UNIDO, BLS,etc.

Labor Cost, 35% Non-wage labour costs (30%)

Social security payment and payroll taxes as a percentage of worker’s salary

Index Global Competitiveness Report (GCR)

Manufacturing Value-added, average annual real growth rate (2000-2006) -

33%

Hard UNIDO Manufacturing Value-added per capita,

at constant 2000 US$ prices(2006) -

33% Hard UNIDO Industrial performance (30%) _{Manufacturing Value-added as} percentage of GDP at constant 2000 prices(2006)- 33% Hard UNIDO

Quality of primary education - 33% Index GCR Primary school enrolment rate - 33% Index GCR Primary

education (10%)

Government Education expenditure -

33% Index GCR Higher education and training (30%) Index GCR Competence, 20% Innovation (30%) Index GCR ISO ownership ratio (50%)

The percentage of manufactures that receive ISO certificates

Hard Enterprise survey

Quality, 15% Supplier quality rating (50%) Index GCR Logistics cost 6 (20%)

Hard Maersk or Schenker Transportation

time (30%)

Hard Maersk or Schenker Time to export

(days) (25%) Hard www.Doingbusiness.org

Cost to export

(15%) US$ per container Hard www.Doingbusiness.org

Infrastructure rating (10%) Index GCR Logistics, 20%

* Special duty (If Applicable) Hard EU trade commission

5 _{Vietnamese Statistical bureau} 6

IPR (15%) Intellectual property protection Index GCR Efficiency of legal framework - 25% Index GCR Transparency of government

policymaking - 25%

Index GCR

Business costs of crime and violence -

25%

Index GCR Institutions

(20%)

Ethical behaviour of firms - 25% Index GCR Macroeconomic

stability (15%)

Index GCR Total local tax rate - 33% Index GCR Number of procedures required to start a

business - 33%

Index GCR Goods market

efficiency (10%)

Time required to start a business - 33% Index GCR Technological

readiness (10%)

Access rate to telephone, mobile and internet, etc.

Index GCR

Local supplier quantity - 33%

Index

GCR State of cluster development - 33% Index GCR Business

sophistication

(10%) Production process sophistication - 33% Index GCR

Social

Responsibility

(10%)

Child labour use - 10% Hard UNICEF

Business Environment, 10%

Corruption

(10%)

Corruption Perceptions Index (CPI) -

10%

Index

Transparency International

Table 8 Low Cost Country selection model: indicator and weight (continued)

is reasonable to only consider labour cost in this case. The input for wage cost level is the result shown in table 7. Non-wage labour cost is measured by the percentage of social security payment and payroll taxes in worker’s salary. These two indicators receive 70% and 30% weight respectively, because the former one comes from hard data and the later on comes from index data. Given that index is an indirect way of measuring, hard data is believed to be more reliable and therefore receives more weight.

The dimension of Competence aims to measure “what can they do?” of a country on a general level. Industry performance monitors the capability of a country’s industrial sector and the key indicator is Manufacturing Value-Added (MVA). Primary education is involved aiming to check the knowledge level of front line workers. It should be a qualifying factor and therefore only receives 10% attention. Higher education and Innovation aim to examine a country’s research and development capability and therefore receive more attentions.

Quality, which is an important issue, receives 15% of weight in the whole model. This is not a signal that quality is not important. Quality can only be judged based on real product and assessing the quality level of a country is difficult. ISO ownership ratio tests whether quality management system is prevalent in a country and supplier quality rating is based on survey to management executives which is performed by Global Competitiveness Report. However, these two indicators have relatively weak measurement validity. For example, it is observed that companies can “purchase” the ISO certificate in some countries and ISO certification has become a fashion and a qualifying factor to stay in business. Therefore, the quality dimension is assigned with less weight to prevent bringing bias to the result.

The last dimension concerns the environment for doing international business. These indicators are also cost drivers for purchasing, especially when the company wants to set an establishment (e.g. a local purchasing office) in the host country. This dimension receives less weight because the model tends to focus more on tangible cost, but they cannot be ignored. In order to clearly show the share of weight for each indicator, Table 9 provides their impact factor to the final result.

Impact factor

Wage per month,% lower than NL 24,5%

Non-wage labor costs 10,5%

Industrial performance 7,0%

Primary education 1,0%

Higher education and training 6,0%

Innovation 6,0%

ISO ownership ratio 7,5%

Supplier quality rating 7,5%

Logistics cost 4,0%

Transportation time 6,0%

Cost to export 5,0%

Time for export 3,0%

Infrastructure rating 2,0%

Intellectual property protection 1,5%

Institutions 2,0%

Macroeconomic stability 1,5%

Goods market efficiency 1,0%

Technological readiness 1,0%

Business sophistication 1,0%

Social Responsibility 1,0%

Corruption 1,0%

Table 9 Impact factor of each dimension to the final result

Findings

Attractiveness for LCCS Wage lower than NL, % Indonesia 6,79 Vietnam 98,8 India 6,67 Ukraine 97,0 China 6,62 Egypt 96,0 Chile 6,46 India 96,0 Tunisia 6,22 Indonesia 95,7 Egypt 6,19 Philippines 95,0 Philippines 6,07 Bulgaria 94,0 Romania 5,87 China 94,0 Czech 5,82 Romania 89,7

Viet nam 5,81 Mexico 89,0

Poland 5,64 Tunisia 88,0 Bulgaria 5,52 Argentina 86,0 Mexcio 5,24 Chile 85,0 Ukranie 5,22 Brazil 83,3 Turkey 5,17 Poland 82,3 Morocco 4,98 Morocco 79,0 Slovakia 4,98 Slovakia 78,4 Argentina 4,83 Turkey 76,0 Hungary 4,79 Czech 75,3 Brazil 4,74 Hungary 75,2 Croatia 4,56 Croatia 70,8

Table 10 Country selection result: Attractiveness for LCCS

and cannot be significantly influenced by the design of weight. Countries in the middle range could score high on certain indicators while score low on others, therefore their ranking can be influenced by the design of weight. The above sensitivity analysis shows that the design of measurement is reasonable and valid, and it is also a process of fine-turning the model design in order to find the most suitable one for decision making.

When looking at Table 9, the result of country selection model shows that, certain country is very attractive when only consider the wage (and labour) cost, however, when a TCO view is taken, that country could look less attractive, such as Vietnam and Ukraine. This shows the importance of taking a total cost view. Besides analyzing the ranking of countries, it is also necessary to “open the black box” and find out which country is good at what. This is to investigate which of the 5 aspects (labour cost, competence, quality, logistics and business environment) contribute the most to a country’s attractiveness for LCCS. The best candidate country should score high on all aspects and its performance is relatively balanced. If a country scores high on one or two aspects but low on others, it brings some concerns for management decisions. The break-down view is shown as a Radar Chart for each country which is attached in the Appendix 3.

The issue of special duty should be discussed separately. The 21 countries could be classified into 2 groups. Countries which are members of EU have the advantage of no import duty when purchasing from there. Rate of import duty for other countries varies from 2.2% to 5.7%, depends on the product to purchase. No anti-dumping duty or sage guard measures are found against the product to purchase. Regarding policy benefit from host country on international trade, Chile has signed Free Trade Agreement with EU; therefore, there is no import duty from Chile.

This model has certain weakness. First, given that certain data (minority in the whole dataset) are difficult to collect, some indicators are not directly comparable, which decrease the reliability of the calculation. Second, this model assesses countries on an aggregate level by taking all industries or sectors. It cannot make an evaluation for specific industry; therefore, this model can only give a general direction. Whether the savings from LCCS could be realized depends on the actual situation of that industry.

6. Sustainability of low cost country sourcing

Based on the theoretical guideline of demographic bonus, a model is developed to briefly analyze the sustainability of low cost in the 21 countries. The model intends to study the “window” in the following way: first, in order to analyze when the bonus window will close for each country, population data of the 21 countries are collected from international database of U.S. census bureau, and then dependency ratio (children, elderly and total) are calculate from the year 1990 to 2030 (estimation data). Population growth rate of these years is also collected as a reference. These statistics are compared and received a score from -2 to 2. Table 11 provides an overview of data and scores. For the threshold of assigning scores, please refer to Appendix 4. The original data of dependency ratio are presented as line chart and they are provided in the Appendix 5. Second, urbanization rate is used as a measurement of population migration. Urbanization rate of the 21 countries at the year of 2020 are collected, and the rate is compared with the urbanization rate of developed economies, afterwards they received a score from -2 to 2. For the threshold of assigning scores, please refer to table A3 in the appendix 4. The design of threshold is only for the purpose of differentiating those countries on that indictor. Those two scores will be used to calculate the weighted average score, but the weight of each aspect is different. Age structure transition (AST) has the fundamental influence; therefore, it receives 80% of the weight, while migration as a facilitating factor receives 20% of weight.

Vietnam 2024 1,0 2 Morocco 2022 1,0 1 Argentina 2030 + 0,9 2 Brazil 2024 1,0 2 China 2012 0,5 0 Tunisia 2016 0,8 1 Chile 2016 0,8 1 Romania 2012 -0,2 -1 Hungary 2010 -0,3 -1 Croatia 2014 -0,1 -1 Poland 2012 -0,2 -1 Slovakia 2012 0 -1 Czech 2008 -0,2 -1 Ukraine 2012 -0,7 -2 Bulgaria 2008 -0,9 -2

Table 11 size of bonus window: overview of data and score

Finally, in order to measure whether the low-cost country are making good policies to grasp the gift of demographic bonus, the country ranking from Global Competitiveness Report (2008) is used as input data because it includes comprehensive factors of a country’s development. The ranking is converted to a score of [1, 10] and then divided by 10, so that the result of calculation acts as a factor of multiplication. It is multiplied to the weighted average result of first two aspects. The full calculation model is shown in table 12 and the calculation follows the formula:

size of window A Weight B urbanization rate at 2025 C Weight D Score E Policy and institution quality F Sustainable Purchasing Score G India 2 80% 2 20% 2,0 0,7 1,40 Indonesia 2 80% 1 20% 1,8 0,7 1,26 Egypt 2 80% 2 20% 2,0 0,5 1,00 Vietnam 2 80% 2 20% 2,0 0,5 1,00 Turkey 2 80% 0 20% 1,6 0,6 0,96 Mexico 2 80% -1 20% 1,4 0,6 0,84 Brazil 2 80% -1 20% 1,4 0,6 0,84 Philippines 2 80% 0 20% 1,6 0,5 0,80 Tunisia 1 80% 0 20% 0,8 0,8 0,64 Morocco 1 80% 1 20% 1,0 0,5 0,50 Argentina 2 80% -2 20% 1,2 0,4 0,48 Chile 1 80% -2 20% 0,4 0,9 0,36 China 0 80% 1 20% 0,2 0,8 0,16 Romania -1 80% 1 20% -0,6 0,6 -0,36 Hungary -1 80% 0 20% -0,8 0,6 -0,48 Croatia -1 80% 0 20% -0,8 0,6 -0,48 Poland -1 80% 0 20% -0,8 0,7 -0,56 Slovakia -1 80% 0 20% -0,8 0,7 -0,56 Czech -1 80% 0 20% -0,8 0,8 -0,64 Ukraine -2 80% 0 20% -1,6 0,5 -0,80 Bulgaria -2 80% 0 20% -1,6 0,5 -0,80

Table 12 calculation of sustainable purchasing score

Findings

of stepping out of the window. Certain research in China estimates that the bonus effect will end at 2015 (Pool, Wong and Vilquin, 2006). This explains why it is witnessed that labour cost in China has been increasing significantly. Countries like India, Egypt, and Philippines are in the prime of the bonus. It is estimated that the window of opportunity could be open until 2025 or 2030. Given that India is actively participating in globalization and adjusting its policy to create a good environment for economic development, it is an attractive country regarding sustainable low cost purchasing. Other Asian countries, such as Vietnam and Indonesia also have the potential, but there policy environment should be improved. Countries in Latin America are in the bonus window but the size of window is different for each country. Argentina could still enjoy the window until 2030. Brazil and Mexico could step of out the window around 2025. However, the window for Chile could close soon at around 2015. It is also deemed that Latin American countries are not utilizing this one time gift very well except for Chile. The governments must try to make better policy to increase employment and improve economic environment, otherwise the gift will be wasted (Sedano, 2008).

Countries in “Emerging Europe” are not attractive regarding sustainable low cost sourcing. It is witnessed that the population growth rate has been negative in recent years, which means in near future, those countries will experience shortage of labor. The reason that labor costs in those countries are relatively low is not due to the effect of demographic bonus. Those countries are about to pass the window. Rather, it is because their economies are recovering from the negative effect of central planning regime. Given that those countries have joined the EU (or will do in near future), it is believed that their economies will closely integrated with more advanced economies in Europe, so it can be expected that the labor cost will increase fast in near future.

regional levels in a country, so countries that are geographically big have the possibility to prolong the bonus window. The effect of demographic bonus on specific industry can also be different. For example, garment and toy production business can feel more cost pressure than equipment manufacturing when the bonus window closes. The valuable benefit of demographic bonus theory is that it can be applied not only for low cost purchasing, but also for outsourcing, local production and other forms of foreign direct investment (FDI).

7. Discussion of results

7. 1 Discussion of LCCS country selection

Morocco 4,98 0,50 5,48 Mexico Argentina 4,83 0,48 5,31 Ukraine Czech 5,82 -0,64 5,18 Turkey Poland 5,64 -0,56 5,08 Morocco Bulgaria 5,52 -0,80 4,72 Slovakia Ukraine 5,22 -0,80 4,42 Argentina Slovakia 4,98 -0,56 4,42 Hungary Hungary 4,79 -0,48 4,31 Brazil Croatia 4,56 -0,48 4,08 Croatia

Table 13 Final calculation of LCCS country selection

By comparing the two rankings, some interesting findings emerge. First, India, Indonesia and Tunisia remain as top 5 countries in both rankings, which imply that for those three countries, not only they are attractive destination for LCCS, but the low cost is sustainable in the medium term (3~5 years). The positions of Vietnam, Philippines and Egypt are promoted after considering sustainable purchasing issue, which means they are also worth being considered by management for investment in the long term, but with more caution. It is important to note that the ranking of China and Chile are lowered down when consider the future. This shows that in the near future, labour cost in these countries could increase significantly. They are getting less competitive for low cost sourcing given that their economies are experiencing transition from cheap production factor driven to innovation driven, which means they are getting more attractive for other types of foreign investments. Other Latin American countries rank in the middle range of final calculation. Those countries are still enjoying the demographic bonus, but a lot of work should be done to improve the economic and business environment, and they are more suitable for Northern American companies to do LCCS. Countries in “emerging Europe” score at the bottom mainly because the low cost is not sustainable in the near future.

conservative to go to this country, mainly due to the problem of corruption based on their practical experience. Although corruption has only 1% impact power in the TCO country selection model, when comes to detailed business matters, it could act as a major barrier. Therefore, management can use the decision making process model as a useful tool to get general direction, but final decision making needs to consider other risks which cannot be directly assessed by the model.

In order to test the effectiveness of the whole country selection model, the company began to search for suppliers in India, Tunisia and Turkey. Within limited time, several quotations of different products were received from suppliers. Although judging from the initial price offer no significant saving can be obtained from those suppliers, this is not regarded as a signal that the whole selection model is invalid. Instead, it is attributed to the fact that those suppliers are from unknown business network. Receiving an opportunistic offer from a “cold contact” is common in purchasing practice. Therefore, further supplier searching and development are required.

Based on the above result, when it comes to the management decision of which country to go, it is recommended that big country is relatively good choice. By “big country” it refers to size of territory and also the scale of economy. The reason is that, firstly, countries with large territory normally possess various and abundant natural resources, which ensures the country to be independent on resources consumption. Secondly, Countries with large territory normally have unbalanced situation among different areas regarding economic development and age structure transition, such as China, India and Brazil. For example, the eastern costal regions of China are more developed than inland areas and attract most of foreign investment. The population of some costal provinces have finished age structure transition and are becoming aging, while inland provinces can still enjoy the demographic bonus.

together for the benefit of firm-level scale economies and reduced transportation cost and congestion cost, so that they can serve large local markets from a few plants. When firms are drawn into one cluster, the local land becomes scarce resource so land price intends to rise. Given that living cost in the agglomeration area (normally big cities) is higher than peripheral areas, therefore, firms need to offer higher labour compensation to attract workers. The effect of industrial agglomeration also contributes to the inequality of development among different areas. Statistics of wage development in China shows that the 2008 average wages in coastal provinces7 is 35.2% higher than interior provinces (China Statistical Year Book, 2008). Therefore, due to industrial agglomeration, cost of production factor, such as land rent and labour cost will gradually rise. Then this industrial cluster will become less cost competitive, therefore, these firms tend to relocate to less developed areas, and the developed area will experience industry upgrading, normally from manufacturing to service business. For big countries with unbalanced development situation, there is much more room to relocate the low cost driven industries. Since this process is relative slow, the increase of cost for the whole country is also slower than small country. Besides, countries with less scale on economies, such as Vietnam and Thailand, can be relatively easier be influenced by global economic crisis. Examples include the Asian financial crisis happened in 1997 and recent economic crisis of Vietnam in 2008.

7.2 Discussion of organization issue

In section 4, the finding of item selection for low cost country sourcing shows that, purchasing managers’ opinions on whether a product should be purchased from a low cost country are much diversified. The original answers from respondents show that some purchasers provide strong support while some others are rather conservative. This is a signal that Low Cost Country Sourcing, as new purchasing practice, is not widely accepted by purchasers in

the company. Therefore, it is important to find out what is hindering the progress and provide a solution.

Kamann and Bakker (2004,2007) did a long term research to study how does a new purchasing practice spread in a company, and why does this process differs between different companies. A framework called contagion process is proposed and there are many factors influencing this contagion process. Firstly, purchaser’s personal trajectory together with various networks that they have been through contributes to the way that the purchaser believes how purchasing practice should be organized. Secondly, due to corporate culture or tradition, a company has its own way of “getting things done”, which is a process of social negotiation, and everybody comply with this social order. This way of doing business is defined as Socially Negotiated Order (SNO). Therefore, even if the purchaser has his/her own opinion on purchasing practice in some cases, they need to follow the SNO. However, there are exceptions. The purchasing function has different social status in companies. Normally companies in fashion industry of trading business pay more attention to the voice of purchasing, but in other industries, purchaser still does not have high weight in management decisions. The social status of different functions in companies is defined as Negotiated Social Order (NSO). If a purchaser could gain higher status in the NSO in a company, he/she can pervade the management in order to put his opinion towards organizing purchasing into business practice, in order to influence the SNO.