Understanding the Offshore Mixed Sourcing Strategy: A Case Study of a Japanese Affiliated Apparel Factory in China

Abstract

The strategic decision to choose in-house production and outsourcing (make and buy) is one of the enormous research questions of international business studies. However, the dynamics of offshore mixed sourcing involving suppliers with varying capabilities remain under-explored. This study intends to elucidate how a firm optimizes the division of labour between an affiliated offshore factory and heterogeneous contract manufacture. We adopt a single-case study design to analyse a Japanese apparel firm operating in China. The empirical analysis using the plant-level data of both in-house and outsourcing Chinese factories reveals a clear strategic distinction: the affiliated factory specializes in High-Mix Low Volume (HMLV) production to manage market volatility, whereas outsourcing partners are utilized for volume production, segmented by their quality capabilities. This study contributes to the literature by demonstrating that mixed sourcing is not merely a cost-saving tactic but a mechanism to manage supply chain heterogeneity.

1. Introduction

The description of the internalisation or outsourcing of production processes abroad is still one of the enormous questions of international business studies (Buckley, 2016; Buckley & Tian, 2017; Casson et al., 2016; Choi et al., 2018). As reviewed by Buckley (2016), deepening globalisation forms a complex web of global value chains (Gereffi et al., 2005). Globalisation in the 21st century provides opportunities for emerging economies, especially China, and has brought new challenges to the global sourcing strategies of MNCs (Ju et al., 2019).

However, the landscape of global sourcing is shifting. In the post-pandemic era, strategies are evolving from pure cost efficiency to resilience and strategic autonomy. Witt et al. (2023) argue that the accelerating decoupling of global value chains from efficiency-driven logics has significant implications for firms’ international business strategies. Similarly, Cassidey et al. (2022) and Puranam et al. (2013) analytically demonstrate that a “mixed strategy”—combining outsourcing and in-house production—acts as a critical hedging mechanism against disruption risks, superior to single sourcing modes.

This strategic shift is particularly salient in transition economies like China. Mukherjee et al. (2023) highlight that institutional reforms in such economies significantly influence a firm’s vertical scope. Ju et al. (2019) provide empirical evidence that multinational firms in China are more likely to adopt concurrent sourcing strategies to navigate high institutional regulations and market uncertainty.

Despite these advances, a significant gap remains in understanding the operational mechanism of this strategy. Previous studies, such as Song (2021) and Ju et al. (2019), have predominantly focused on firm-level performance or macro-level volume flexibility, treating outsourcing partners as a homogeneous group. Few studies have empirically examined how firms manage “heterogeneity” among suppliers at the plant level. Specifically, the following question remains: how do firms allocate production volumes and varying quality requirements between their own offshore subsidiaries and a diverse set of local subcontractors?

In line with the previous studies, this study analyses the managerial decision of firms on their heterogenous sourcing strategy that combines in-house and outsourced manufacturers. In this study, we evaluate the performance of the sourcing strategy primarily through ‘Quality Efficiency,’ measured by defect and repair rates, and ‘Operational Flexibility,’ measured by the diversity of item types (SKUs) relative to production volume.

This study conducts a case study of a Japanese apparel firm that operates its factories in Japan and China and outsources to Chinese manufacturers. The Japanese apparel industry is one of the leading industries that actively promotes offshore outsourcing manufacturing in China, Southeast Asia, and South Asia because it suffers from the appreciation of the Japanese yen, rising labour costs and labour shortages (Zhao, 2011; Inoguchi et al., 2012; Iwasaki & Ueki, 2022; Iwasaki et al., 2025). This observation is consistent with that of Cohen et al. (2018), which suggests a continued decline in manufacturing activities in Japan and limited opportunities for Japanese firms to repatriate the production capacities. The authors collected plant-level data from seven factories (one affiliated factory and six subcontractors) in China that operate the textile sewing processes for a Japanese apparel firm. By utilising the data and interview-based analysis with these factories, this study intends to understand the firm’s strategic decision regarding why and how firms manage the combination of in-house production and outsourcing manufacturers.

Because of the labour-intensive processes, most of the apparel industry in Japan tends to outsource the production processes to Asian developing countries as an original equipment manufacturer (OEM) style. Therefore, it is also invaluable to select the case of the apparel industry that only maintains in-house productions. These critical motivations of this study are based on such research interest.

This paper is divided into five sections. Section 2 has a literature review related to firms’ internationalisation and sourcing strategies, including Dunning’s eclectic approach (OLI framework), concurrent sourcing, and parallel production. Section 3 provides an explanation of the methodology. Section 3 also explains the characteristics of each subsidiary factory based on interviews with the managers of the focal Japanese firm. Section 4 presents empirical findings from the combination of factory-level data and interviews. Section 5 provides the discussion and conclusion of this study on the firm’s decision to outsource production, as well as the limitations and remaining challenges of this study.

2. Literature Reviews

Previous international business research has analysed the costs and benefits or advantages of domestic and offshore insourcing (in-house production) or outsourcing production (subcontracting). Among various related theories, the literature review here briefly highlights (1) the OLI framework developed by John Dunning as the theoretical foundation of internationalisation and (2) mixed strategies of insourcing and outsourcing (e.g., plural or concurrent sourcing) as a framework to understand the reality of Japanese firms’ sourcing practices.

2.1. The OLI Framework

International business studies have focused on the relationship between location choice and international production to map out the sourcing strategy of firms as their activities have become multinational. Dunning proposed the so-called OLI framework in the study of FDI to explain the pattern of international business activities. In the OLI framework, the competitive advantages of three elements, “ownership”, “location”, and “internalisation”, determine the internationalisation of firms, including the choice of location and entry into foreign markets (Dunning, 1977, 1980, 1981).

The OLI framework is based on the activities of Western firms, but it is applicable to firms from emerging economies in the 21st century (Dunning, 2001). In the case of Chinese firms, Pheng and Hongbin (2006) analysed the OLI advantages of Chinese construction multinationals (MNCs) in the international market. Erdener and Shapiro (2005) studied the Chinese family business through the OLI paradigm. “OLI is best seen as a way of looking at the phenomenon of MNCs and their activities” (Eden, 2003). Dunning’s eclectic paradigm offers the perspective of analysing the internationalisation of firms at the macro (country), meso (industry) and micro (firm) levels, respectively.

Although the eclectic OLI paradigm is still a valuable framework for understanding the benefits of firms’ internationalisation, recent academic studies tend to focus on the more complex managerial strategy of choosing the balance between “make and buy” in the context of globalisation in growing emerging economies. In practice, a firm can choose between in-house production and outsourcing, including domestic and foreign sources of supply. Outsourcing allows the firm to specialise in a particular task, in which it has a comparative advantage. The smile curve shows the benefits of vertical specialisation in activities that the firm can add more value for. Outsourcing may become one of the choices for the firm when service link costs are low enough to allow production fragmentation. Transaction costs (e.g., uncertainty, suppliers’ production costs, competition between suppliers) influence make or buy decisions (Walker & Weber, 1984). The sourcing decision also depends on the firm’s strategy for controlling the sources of supply in the value chain. Some firms will prefer retaining control over the critical value-adding processes (Mudambi & Venzin, 2010).

2.2. Mixed Sourcing Strategies

Unlike traditional concurrent sourcing studies that treat outsourcing partners as a homogeneous group, this study emphasizes the role of ‘heterogeneous outsourcing.’ We define this as the strategic use of multiple suppliers with varying capabilities (e.g., varying levels of quality control, lot-size flexibility, and technical skills) to optimize the overall supply chain performance.

Historically, studies before the 2000s focused mainly on the dichotomous “make or buy” decision (i.e., choosing between in-house manufacturing or production outsourcing) based on Transaction Cost Economics (TCE) considerations of cost minimization (Williamson, 1985). Within this framework, dual or multi-sourcing emerged as an important strategy for corporate buyers to mitigate supply risks by deploying two or more suppliers (Yu et al., 2009; Zeng, 2000). However, the global landscape has evolved, and Multinational Corporations (MNCs) have shifted from this binary choice to more complex “mixed strategies” that simultaneously utilize in-house production and external sourcing to optimize operations under changing international environments (Krzeminska et al., 2013; Puranam et al., 2013).

The concept of mixing “make and buy” encompasses strategies such as concurrent sourcing (Heide et al., 2014; Mols, 2010; Parmigiani, 2007) and plural sourcing (Heide, 2003; Mols et al., 2012). Mols (2010, p. 61) defines this as “a phenomenon where firms simultaneously make and buy the same good.” Unlike the traditional view of substitution, these modern perspectives emphasize complementarity. Concurrent sourcing enables firms to reduce the uncertainty of lot sizes and technologies, minimize transaction costs, and enhance performance (Mols, 2010; Parmigiani, 2007). Nordigården et al. (2014) further elaborate on this by analysing wood product manufacturers in Scandinavia, distinguishing between “in-house-dominant” strategies (focused on complementary competencies) and “outsourcing-dominant” strategies (focused on flexibility).

Recently, theoretical perspectives have evolved further to address post-pandemic realities. Cassidey et al. (2022) analytically demonstrate that a mixed strategy acts as a critical hedging mechanism against supply chain disruptions, offering superior resilience compared to single sourcing. Similarly, Witt et al. (2023) argues that in an era of geopolitical uncertainty, firms face pressure to reconfigure their supply chain governance in response to decoupling pressures. This strategic complexity is particularly pronounced in transition economies like China. Mukherjee et al. (2023) highlight that institutional reforms in such economies significantly alter transaction costs, compelling firms to reconsider their vertical boundaries. Ju et al. (2019) provide empirical evidence using data from China, showing that MNCs are more likely to adopt concurrent sourcing strategies to navigate high institutional regulations and market volatility. In this context, maintaining in-house capacity serves as a buffer against local institutional uncertainties.

While recent studies have clarified the macro-strategic drivers, operational mechanisms remain underexplored. Song (2021) finds that MNCs adjust production volumes flexibly among subsidiaries to respond to labour cost fluctuations. However, while these studies explain “volume” flexibility, they often overlook the “quality” dimension. Qian et al. (2023) demonstrates that the quality of buyer–supplier collaboration varies significantly depending on the extent of relationship-specific investments, with distinct performance implications for buyers and suppliers. Building on these perspectives, this study moves beyond treating outsourcing partners as a homogeneous group. We emphasize the role of ‘heterogeneous outsourcing,’ which we define as the strategic use of multiple suppliers with varying capabilities (e.g., varying levels of quality control, lot-size flexibility, and technical skills) to optimize overall supply chain performance. Unlike the firm-level focus of Ju et al. (2019) or Song (2021), the authors consider the necessity of investigating how a firm allocates specific items to specific factories based on their heterogeneous capabilities to improve firm-level efficiency (Parmigiani & Mitchell, 2009).

2.3. Offshore Mixed Sourcing Strategies

Recent business studies have focused on mixed strategies (e.g., plural sourcing, concurrent sourcing, parallel productions) to overcome the dichotomy between in-house production and outsourced production (Sørensen et al., 2022). However, international business studies have not fully investigated offshore mixed sourcing strategy, where international firms take advantage of both their own factories and contract manufacturers outside their home countries. To fill this research gap, our empirical analysis focuses on the differences between offshore insourcing and outsourcing. By utilising the plant-lot-level data for a Japanese company’s affiliate and subcontractors in China, this study will analyse trade-offs, division of labour, and the complementarity between offshore insourcing and outsourcing under the mixed strategy.

The Japanese manufacturing system has a comparative advantage in its ability to produce high-mix, low-volume (HMLV) products, which enables it to respond flexibly and agilely to changes in market demand. But it can achieve HMLV production at the expense of the economies of scale that high-volume low-mix (HVLM) production can achieve. Previous studies have paid little attention to the costs of HMLV production compared to its benefits, although Anderson (1995) investigated the production costs of product variety. Randall and Ulrich (2001) introduced market intermediation costs, which are associated with the variety-related inventory holding costs, demand uncertainty, and forecasting difficulties, in addition to production costs in a supply chain. Randall and Ulrich (2001) claim that firms can outperform competitors by developing a supply chain structure that combines scale-efficient production at a distant location with market-mediation cost-saving production close to the target market.

Based on the discussion through the literature review, our case study demonstrates a Japanese apparel firm’s mixed strategy that combine in-house HMLV productions in Japan and China with the outsourcing of HVLM productions in China to satisfy Japanese customer demands for high-mix low-priced supplies. In addition, this study extends the conceptual framework of Randall and Ulrich (2001) by introducing heterogeneity in benefits from offshore in-house and outsourcing production capabilities. To understand the competitive advantages of combining in-house production and heterogeneous subcontractors located in the same foreign country, this case study highlights a Japanese firm that operates its factories in Japan and China and uses Chinese sub-contract manufacturers for apparel sewing. These factories in China will differ in the extent of intangible know-how sharing for HMLV production. Our plant-lot-level data will show different effects of product volume and variety on production costs among the factories in China.

3. Methodology

This study adopts a single-case study approach, which is appropriate for examining complex, context-dependent phenomena such as supply chain management decisions (Yin, 2003; Eisenhardt, 1989). A single case allows for a deep investigation into the “how” and “why” of the sourcing process, which quantitative surveys often overlook.

3.1. Conceptual Framework for Characterising Production Capabilities

Before presenting empirical findings from our case factories, we would like to apply the OLI framework (Dunning, 1977, 1980, 1981) to outline the characteristics of in-house production capabilities and outsourcing in Japan and China based on the previous research on apparel sewing factories in Japan (Iwasaki & Ueki, 2022; Iwasaki et al., 2025) and our field research in China (

Table 1. The OLI framework to characterise the in-house and outsourcing factories in China and Japan.

	In-House (China)	Outsourcing (China)	In-House (Japan)
O	Operational flexibility	Capacity flexibility, Low fixed cost	Operational flexibility, Timeliness
L	Single-skilled labour abundance	Single-skilled labour abundance	Multi-skilled labour abundance, Proximity to the market (Japan)
I	Quality control and reliability	Cost	Quality control and reliability

Source: by authors.

). This process helps us understand the division of labour of among the factories in China supplying to our focal Japanese apparel firm.

The previous research (Iwasaki & Ueki, 2022) shows that the in-house production of Japanese apparel firms takes advantage of the HMLV production irrespective of their locations, while offshore outsourcing tends to favour the HVLM production. In the relationship between China and Japan, Japanese firms take advantage of the characteristics of offshore production to reduce costs and obtain abundant labour. In-house factories in China allow companies to have flexible factory operation and reliable production quality compared to outsourcing. The advantages of Japanese production are summarised as the characteristics of operational flexibility, multi-skilled labour, proximity, and timeliness (Iwasaki et al., 2025). While single-skilled labour is the characteristic of Chinese HVLM production, multi-skilled labour is the characteristic of Japanese HMLV production (Iwasaki & Ueki, 2022). Due to the distance from the production side to the Japanese market, the production in Japan has the advantage of timeliness and proximity to the market additionally. The strategic focus on the Japanese market makes the in-house production in Japan beneficial.

Table 1 presents how each OLI factor creates advantages in quality, cost, and delivery (i.e., QCD) control for in-house production in China and Japan and contract manufacturing in China. In terms of the ownership advantages (O), firms have their own offshore production capacities to secure operational at the expense of capital cost. Contrary to in-house production, offshore outsourcing gives them capacity flexibility. In-house production in China is disadvantageous in terms of timely production and delivery compared to domestic production in Japan. But Japanese affiliated factories in China compensate for this weakness with the locational advantages (L) of China, in particular the abundance of single-skilled labours. In terms of the internalisation advantages (I), intra-firm knowledge transfer makes in-house production and quality control more reliable, whereas outsourcing in China allows Japanese firms to use cost-oriented Chinese suppliers and achieve a significant cost reduction.

Based on the observations in Japanese apparel factories in Japan and China and the OLI framework, we assume that the in-house factories of Japanese apparel firms in China focus on HMLV productions or Japanese apparel firms subcontract HVLM productions to Chinese manufacturers. In other words, we hypothesise that the (1) production lot size for Chinese contract manufacturers is more likely to be larger than that for Japanese affiliated factories in China and that (2) the production defect rate for Chinese contract manufacturers is more likely to be higher than that for Japanese affiliated factories in China.

3.2. Case Selection and Data Collection

To investigate the characteristics of factories in China used for offshore mixed sourcing strategies and these hypotheses, this study conducts an in-depth case study of a single apparel company with multi-plant operations for a sewing process. We selected the focal Japanese apparel firm based on the criterion of “representative intensity” (Seawright & Gerring, 2008). The firm represents a typical HMLV manufacturer facing the classic “make-or-buy” dilemma in the Japan–China context. This firm has one affiliated factory in China (Factory A), and Factory A uses six outsourcing factories (Factory B, C, D, E, F, and G). These insourcing and outsourcing productions in China enable us to understand the offshore mixed strategy that includes parallel production and concurrent sourcing (Contractor et al., 2010; Heide, 2003; Heide et al., 2014; Krzeminska et al., 2013; Mols, 2010; Parmigiani, 2007; Puranam et al., 2013; Rothaermel et al., 2006; Theyel et al., 2018). We had several interviews with its Japanese top management before launching our in-depth case study of its affiliated factory and contract manufacturers in China.

We started this study of the focal firm’s mixed sourcing practices with interviews with its factory manager in China. Then we collected data on its overseas procurement from multiple factories in China. The authors have had several interviews with the manager of Factory A and the managers of outsourcing factories. We also had site visits to some factories. The manager of Factory A has been directly involved with the outsourcing decision of six outsourcing factories for more than ten years and manages the multi-plant operations to optimise the firm’s production in China. The interviews with Factory A manager were divided into four times (August 2016, August 2017, and March and November 2019) with follow-up email-based information exchanges. The authors also had field visits to Factory A (the focal factory), Factory B and Factory E in August 2017.

The authors obtained the monthly data for the seven factories for the period January 2015 to December 2018. The data includes key performance indicators (KPIs), including the total number of inputs (denoted as Total), the number of item types (Number), the number of defects (Defect), and the number of repairs (Repairs) (

Table 2. The key variables characterising the factories.

Total	The number of inputs from raw materials.
Shipment	The number of shipments.
Number	The number of item types each factory produces in one month.
Defect	The number of defects at each production process.
Repairs	The number of repairs. After repairing, the variable becomes the “Shipment” volume.
Sewing_Defect	The number of defects occurred in the sewing process. This variable is a sub-category of the variable for Defect.
Month	The number of months in which each factory has a shipment.

Source: by authors.

). We use these indicators to derive production lot size, defect rate, and related variables for our empirical analysis.

The authors also have an online meeting in December 2020 to update the information during the COVID-19 pandemic. These interviews help increase the validity of data collection (Yin, 2003). The authors discussed with the manager by showing the preliminary calculation of the production data and our interpretation of what the data will reflect of the daily business to improve the reliability of our study.

3.3. Data Collection and Triangulation

To ensure scientific validity, we employed data triangulation (Denzin, 1978) by combining two data sources:

• Quantitative Data: Monthly Key Performance Indicators (KPIs) from 2015 to 2018 (N = 48 months’ observations for each factory).
• Qualitative Data: Semi-structured interviews with factory managers for each factory.

4. Empirical Findings

4.1. Overviews of the Case Factories

The authors conducted in-depth interviews with the factory manager of Factory A. The interviews are complementary to help understand the characteristics of each factory. We also had factory visits to its subcontractors to collect their data. However, we could not collect the detailed information necessary for a comparison between the factories due to the limited data disclosure. The following is a brief summary of the characteristics of each factory.

• Factory A is established in the late 1990s by the case company. Factory A accounts for almost half of the production lots in the company’s Chinese production. The number of workers is around 150–200, and Factory A has several facilities, including wash-and-wear treatments, embroidery, and printing to textile. Factory A played the role of a mother factory in the establishment of Factory B and Factory E.
• Factory B is a small factory (the number of workers is less than ten), but the quality is unstable due to poor management. Factory B focuses on small lot productions, and the turnover rate of workers is minimal. This factory was established at the end of the 2000s with technical and capital assistance from Factory A.
• Factory C is also a relatively small factory with around 30 workers. The manager of Factory C is not able to manage the workers, so the turnover rate is relatively high. Factory C has other business partners and receives orders from Factory A based on labour costs.
• Factory D is not the main factory for Factory A and mainly receives orders during the low season. The number of workers is the largest of the six subcontractors, and Factory D has other main business partners.
• Factory E is a relatively larger factory, and the number of workers is around 40. Factory E is 100% dependent on orders from Factory A because it was set up in the late 2000s with technical and capital assistance from Factory A.
• Factory F is an irregular subcontractor and provided no specific information for this study.
• Factory G is also an irregular subcontractor. The number of workers is small. Factory G is called in when the delivery date does not seem to be on time.

4.2. Characteristics of the Case Factories

Table 3. Factory overviews.

Factory Name	Number of Workers	Characteristics of Firms	The Turnover Rate of Workers	Quality of the Products	Notes
Factory A	Around 150–200	The affiliated factory of the focal firm	Low	High	The mother factory for Factory B and Factory E
Factory B	Less than 10	A regular subcontractor	Low	Not Low	Established with the assistance of Factory A
Factory C	Around 30	A sub-regular subcontractor	High	Not Low
Factory D	N/A (large-sized among the 6 subcontractors)	A sub-regular subcontractor	High	Not Low
Factory E	Around 30–40	A regular subcontractor	Low	High	Established with the assistance of Factory A
Factory F	N/A	An irregular subcontractor	N/A	N/A
Factory G	N/A (small-sized)	An irregular subcontractor	N/A	N/A

Note: Due to the limited information, the author could not get clear details of Factory F and Factory G. Source: By authors.

shows the average of the KPIs listed in Table 2 for each factory. To ensure the confidentiality of the raw data, these KPIs for each contract manufacturers (i.e., Factory B–G) are compared to Factory A, which is associated with the Japanese company.

To identify the characteristics of each factory, we introduce the variable for the total monthly number of repaired items (denoted as Repairs). The reason why Repairs can be a proxy for production costs is that the payment to workers and subcontracted manufacturers is based on the number of products that meet the quality requirements needed to pass the quality inspection by the affiliated factory (Factory A) of the Japanese firm. The repair process is only a time-consuming cost for the factories, because the time for repair work does not generate additional fees but could be used to obtain processing fees.

We apply the OLI framework to findings from the interviews and the data in Table 3 and

Table 4. The ratio of the characteristics of factories.

	Factory A	Factory B	Factory C	Factory D	Factory E	Factory F	Factory G	All Factories (Average)
Number	1	0.146144	0.138686	0.100104	0.313869	0.054745	0.110532	0.317434
Total	1	0.129254	0.180593	0.16575	0.605124	0.060072	0.132422	0.386
Shipment	1	0.130163	0.182284	0.166488	0.616763	0.0616	0.133645	0.388899
Repairs	1	0.20221	0.30702	0.236038	0.418556	0.046958	0.234168	0.401558
Sewing_Defect	1	0.148129	0.224449	0.202279	0.503917	0.033055	0.121677	0.379711
Defect	1	0.143815	0.204222	0.192289	0.55909	0.031937	0.147067	0.386486
Lot_size	1	0.922052	1.268616	1.887357	1.983115	1.212166	1.551715	1.391434
Sewing_Defect_rate	1	1.078564	1.304637	1.239539	0.837455	0.706283	1.094991	1.050426
Repairs_rate	1	1.551682	1.789528	1.417693	0.685025	0.713172	1.733287	1.260387
Defect_rate	1	1.045035	1.212271	1.188876	0.916897	0.629407	1.39295	1.055573
Number of month (max = 48)	48	46	42	35	48	20	21	260

Notes: Lot_size = Total/Number; Sewing_Defect_rate = Sewing_Defect/Total; Repair_rate = Repairs/Total; Defect_rate = Defect/Total. Source: By authors.

to summarise the characteristics of each factory as follows.

• Factory A has the highest number of item types and total production among the seven factories and is close to the HMLV production style.
• Factory B is the quasi-main supplier and receives orders almost every month. The total production is the second lowest, but the number of items is relatively high. Compared to Factory C, Factory B concentrates on HMLV production.
• Factory C is the quasi-main supplier and receives orders almost every month. It ranks third in total production and fourth in number of items.
• Factory D is also a quasi-main supplier and has received orders in 35 months. The total production is the fourth, and the number of items is the second lowest. Compared to Factory C, Factory D is relatively focused on HMLV production.
• Factory E is the main outsourcing factory and receives orders every month. The total production and the number of item types are the second highest, but the lot size is relatively large (HVLM production).
• Factory F is an irregular subcontractor (20 months in 4 years). The total production and the number of items are the lowest among the subcontractors.
• Factory G is an irregular subcontractor (21 months in 4 years). The total production is the fifth (more than factory B’s) and the number of items is the fifth. The production style is relatively HMLV.

We also conducted statistical tests to examine whether outsourcing factories exhibit systematically higher defect and repair rates than the affiliated factory (Factory A). Specifically, our null hypothesis was that the mean defect rate (Defect/Total) and repair rate (Repairs/Total) of the outsourcing factories are equal to those of Factory A, against the alternative that outsourcing factories perform worse. We applied both a t-test and a Wilcoxon signed-rank test to the factory-level ratio data in Table 4 to ensure robustness against potential non-normality. The results showed no statistically significant difference (p > 0.05) in defect rates or repair rates between Factory A and the outsourcing group (see

Table 5. t-test and Wilcoxson Test Results.

Defect Rate	Repairs Rate
External factory average: 1.064 (approximately 6% higher than Factory A)	External factory average: 1.315 (approximately 31% higher than Factory A)
t-test p-value: 0.581 (no significant difference)	t-test p-value: 0.180 (no significant difference)
Wilcoxon test p-value: 0.563 (no significant difference)	Wilcoxon test p-value: 0.156 (no significant difference)

Notes: based on the data from Table 4.

).

Rather than treating this as a simple null result, these findings warrant a deeper qualitative interpretation. The absence of a significant quality gap between Factory A and its outsourcing partners is itself a theoretically meaningful outcome. Our interview evidence suggests that this quality convergence is not accidental: Factory A has deliberately invested in the technical development of its principal subcontractors (particularly Factory B and Factory E), including sharing production know-how, training workers, and providing capital assistance during their establishment. This mechanism of intra-network knowledge transfer effectively reduces the quality gap. In other words, quality parity is an achieved outcome of the firm’s governance strategy—not merely a statistical artifact—and underscores the “mother factory” function of Factory A as a quality anchor for the supply chain. The quantitative data thus corroborates the qualitative finding that the firm maintains consistent quality standards across its heterogeneous supply chain through active relational governance.

Figure 1 maps out the positioning of the seven factories according to the volume size of inputs in the vertical axis and the number of item types (i.e., product mix) of the productions in the horizontal axis.

5. Discussion and Conclusions

5.1. Discussion

As previous studies focus on the mixed strategy, this study conducts a case study to understand how firms combine in-house production and heterogeneous suppliers in the same foreign country. Although this study uses the single case of a Japanese company with offshore in-house and outsourced production facilities, the authors obtained in-depth information through data collection and interviews. This research method has a limitation to generalise the findings but will help understand the managerial decisions of the firms’ sourcing strategy. Some insights can be gained by combining quantitative analysis and interviews.

According to the literature, firms take a mixed sourcing strategy to cope with multiple problems associated with demand uncertainty, volume uncertainty, technological uncertainty, difficulty in evaluating subcontractors’ performance, and the need for a shorter lead time (Puranam et al., 2013; Mols, 2010). This analysis identified Factory B and Factory E as principal subcontractors for Factory A. When considering sourcing the outside supplier or in-house production, the manager of Factory A relies on its in-house production and Factory B for complicated items and Factory E for mass production. The interviews support this observation. As the previous section and Table 3 show, historically, Factory A has had long relationships with Factory B and E since their establishment with Factory A’s technical assistance.

While the quantitative indicators (e.g., lot size ratios in Table 4) for Factory A appear similar to some outsourcing factories, the qualitative interviews clarify a key distinction: Factory A handles technically complex items that require higher skill levels, whereas outsourcing factories typically handle simpler and high-volume items. Thus, similar statistical performance metrics actually imply Factory A’s superior capability in managing complexity without compromising quality.

These indicators may imply that the manager of Factory A orders relatively simple items to irregular subcontractors (e.g., Factory F and G). The reason is quite challenging to estimate precisely. Still, one high possibility is that Factory A would like to keep the business relationship with irregular subcontractors to deal with the fluctuation of demands and management of the short delivery lead time. This may also explain the relation with Factory C and D. While Factory C and D have relatively high defect rates and repairs rates (Table 4), Factory A constantly provides orders to Factory C and D. Due to their comparatively large production capacities, keeping the business relationship may be significant as managerial decisions.

From a managerial point of view, factory managers need to focus on the performance of the sourcing strategy to optimise the total production costs and benefits. Factory managers need to manage both in-house and outside contractors under the high uncertainty of the demands and technological boundaries. One of the ways to deal with these difficulties is a mixed sourcing strategy that uses the HMLV production of complex items at the affiliated factory and outsourcing the HLMV and simple products to subcontractors concurrently. Although the fluctuation and uncertainty of the demand bring enormous difficulties in the production management of apparel sewing factories, these demand-side challenges also encourage firms to strengthen their sourcing strategies and enhance technical cooperation and business relationships between the manufacturers. Our interview-based analysis provides evidence helpful for a better understanding of the relationship between offshore subcontractors and in-house plants of the international firms. Therefore, our findings are consistent with the previous studies of international business research.

Additionally, it should be noted that the quantitative data covers the period from 2015 to 2018, representing the supply chain dynamics under normal operating conditions (pre-pandemic baseline). Although the COVID-19 pandemic disrupted global supply chains starting in 2020, our follow-up interviews revealed that the trust and operational protocols established during the pre-pandemic period were crucial for the firm’s resilience. Therefore, the analysis of this dataset remains valid for understanding the structural foundations of the mixed sourcing strategy.

Trade policy considerations, particularly tariffs and rules of origin, represent an important dimension that falls outside the scope of this study. During the study period, Japan–China trade was governed by MFN tariff rates under the WTO framework, which created a baseline cost structure affecting sourcing decisions. More recently, the entry into force of the Regional Comprehensive Economic Partnership (RCEP) in 2022 has introduced preferential tariff schedules alongside rules of origin requirements for textiles and apparel. Under such arrangements, the degree of in-house control over production processes can determine whether finished goods qualify for preferential treatment, thereby linking governance mode choices directly to trade policy outcomes. How firms strategically configure offshore mixed sourcing in response to evolving tariff regimes and rules of origin requirements remains an important question for future research.

It is also important to situate the observed sourcing patterns in the context of China’s evolving labour market. During the study period, China experienced a structural tightening of the supply of low-cost labour in coastal manufacturing regions, accompanied by sustained wage increases (Cui & Lu, 2018). These dynamics have asymmetric implications for in-house and outsourced operations. For affiliated Factory A, for example, rising labour costs could accelerate investment in semi-automated technologies or skill upgrading, which could widen the capability gap relative to lower-tier subcontractors. Conversely, cost pressures may threaten the viability of irregular subcontractors such as Factories F and G as buffer suppliers, thereby reducing the ‘insurance value’ of the mixed sourcing portfolio (Song, 2021). As China’s cost advantage narrows, the possibility of partial reshoring to Japan becomes more plausible, particularly for high-complexity HMLV items where Japan’s multi-skilled labour force has a comparative advantage (Iwasaki et al., 2025). The way in which the focal firm has adapted its sourcing configuration in response to these dynamics is an important question for future research.

5.2. Theoretical and Managerial Implications

This study contributes to the literature on global sourcing and supply chain management in three ways. First, it extends our understanding of mixed sourcing strategies beyond the traditional volume-flexibility perspective. While recent studies such as that by Song (2021) emphasize that MNCs adjust production volumes among subsidiaries to cope with cost fluctuations, our findings highlight the critical role of “quality heterogeneity.” We demonstrate that the decision to “make or buy” is not binary but a strategic allocation of complexity—where the in-house factory handles HMLV items to mitigate technical uncertainty, while outsourcing partners manage standardized volume production. This nuances the “concurrent sourcing” theory (Parmigiani, 2007) by adding the dimension of product complexity (SKUs) as a determinant of governance mode.

Second, this research answers the call to integrate resilience into sourcing theory. Cassidey et al. (2022), Puranam et al. (2013) and Witt et al. (2023) theoretically argue that decoupling and geopolitical pressures compel firms to reconfigure supply chain strategies beyond pure efficiency logics. Our case study provides the empirical mechanism for this argument: the focal firm utilized its in-house facility not merely for cost efficiency, but as a “control tower” to maintain supply chain continuity during volatile periods. This supports the shift in International Business literature from a purely efficiency-based Transaction Cost Economics view to a resilience-based perspective.

Third, we contribute to the institutional view of sourcing in transition economies. Consistent with Mukherjee et al. (2023) and Ju et al. (2019), who argue that institutional voids in China drive firms toward concurrent sourcing, our study reveals that the in-house factory functions as an instrument of “governance.” By providing technical guidance to subcontractors, the focal firm reduces information asymmetry and behavioural uncertainty, effectively managing the “acquiescence” of suppliers as described by Teng et al. (2023).

For practitioners, particularly managers of MNCs operating in labour-intensive industries, this study offers actionable insights. First, managers should abandon the “one-size-fits-all” approach to outsourcing. Instead, they should adopt a portfolio approach based on item complexity. Complex, high-risk, or urgent items should be retained in-house (even at higher costs) to preserve quality and flexibility, while standardized items are allocated to subcontractors to leverage scale economies.

Second, the in-house offshore factory serves a strategic function as a “Mother Factory” that accumulates technical know-how and diffuses it to outsourcing (sub-contracting) partners. Managers should invest in the technical capabilities of their subsidiaries, as this investment is the prerequisite for effective control over a heterogeneous supplier network.

Third, in an era of supply chain disruptions, maintaining a “mixed” footprint is not an inefficiency but an insurance policy. As shown in our case, the ability to shift production loads between in-house and external lines allows firms to absorb shocks that would otherwise paralyze a pure-outsourcing or pure-in-house model.

5.3. Limitation and Further Research

There are plenty of limitations of this research that authors need to investigate. Although our study sheds light on how a firm combines its in-house and heterogeneous outsourcing productions in the same countries by utilising the empirical analysis of plant-level data, our findings are still based on the single firm-level analysis in the Japan–China relation. Thus, further research is necessary to generalise our findings.

Second, due to limits on the number of plants in this study, our study may have a limitations in understanding the strategic positioning of the overall Chinese productions in the firm.

Finally, we needed field research to update the data and information, but we could not materialise it due to various constraints during and after the COVID-19 pandemic. The mixed sourcing strategy is dynamic and will change according to business environments. More specifically, the pandemic fundamentally reshaped global supply chains, challenging several assumptions on which this study was based. The pandemic exposed vulnerabilities in geographically concentrated supply networks, accelerating the shift towards resilience-first sourcing structure in favour of efficiency (Cassidey et al., 2022; Witt et al., 2023). Our follow-up interviews in December 2020 suggest that relational trust built with subcontractors during the pre-pandemic period supported the firm’s initial response to disruption. However, it remains an empirical question whether the mixed sourcing configuration observed in 2015–2018 still applies. Future research may collect post-2020 data to determine whether the division of labour between Factory A and its subcontractors has changed, and whether maintaining irregular subcontractors has become more or less valuable in the context of ongoing supply chain disruption.

This study has limitations inherent to a single-case study design. The findings are derived from a specific Japanese apparel firm and its supply chain in China, which may limit its generalizability to other industries or regions. Future research should examine whether these patterns of heterogeneous mixed sourcing hold true in other labour-intensive sectors or different geopolitical contexts.