Gender Dynamics and Upgrading in Global Value Chains: The Case of Medical Devices Penny Bamber and Danny Hamrick1 January 28, 2019 1 Penny Bamber and Danny Hamrick work for the Duke Global Value Chain Center. This paper has been commissioned by the World Bank Group as background paper for the WBG-WTO Global Report on Trade and Gender. 1 Abstract Global value chain (GVC) integration and upgrading have become key goals of trade and investment policy in developing countries. GVC trade is seen as a channel for export revenue generation, knowledge and technology transfer; and importantly, boosting employment, particularly for women. Early sectors to globalize brought with them strong opportunities for female workers – creating a large number of jobs. For many of these women, this was their first waged job out of the home and their participation in the workforce helped drive export competitiveness. As policymakers around the world set their sights on moving into the production of more sophisticated products or adopting new technologies into their operations, what is the impact on female employment? Will women be pushed out of the workforce as countries upgrade or do these new industries offer opportunities for both men and women? This study provide insights on how these changes in technology intensity may impact female opportunities in global value chains. It analyzes how increased technology intensity of products made in a country may affect gender distribution of employment by looking at the entry of two developing countries in the medical devices manufacturing GVC. The study reveals that as a rapidly growing global industry, the medical devices GVC provides strong opportunities for women in developing countries to work in a high-tech sector – compelling evidence for policymakers seeking to meet both economic and social development goals. Women accounted for over half of employees in the sector in both countries analyzed (Costa Rica and the Dominican Republic). Furthermore, these jobs are not just on the production line, but encompass professional and technical positions as well. Female technical workers are actually the fastest growing employment category; with female intensity rising as more women enroll in technical programs. The data shows that these jobs are comparatively well-paid, and permanent. This runs contrary to with predominant thinking that women in GVCs are primarily concentrated in low- paying, temporary work. 2 1. Introduction GVC-related trade in sectors such as apparel, agriculture and electronics, has generated significant female employment, offering many women their first chance for waged employment (Bamber & Staritz, 2016; Lopez-Acevedo & Robertson, 2016; Staritz & Reis, 2013). Many emerging economies are now seeking to upgrade into increasingly higher value and/or more technologically sophisticated sectors.2 While the economic imperative is clear, the gender implications of this strategic shift are more ambiguous (Kucera & Tejani, 2014; Saraçoğlu et al., 2018). In particular, important questions arise regarding the relative opportunities for men and women in these higher value GVCs, and what, if any, gender sensitive policies are required to enhance a country’s competitiveness in these sectors. While there is a growing literature on the nature of work across these different types of value chains (Fernandez-Stark & Bamber, 2018; Gereffi et al., 2011; ILO, 2016), research on gender and GVCs has primarily focused on labor-intensive and generally low-technology industries characterized by high levels of feminization. Much less research has been carried out on ‘high tech’ GVCs, or in value chain segments with higher technology intensity (Saraçoğlu et al., 2018). This study seeks to contribute to this gap in the gender and trade debate by analyzing the gender implications of, and impacts on trade and competitiveness in, a single high value manufacturing sector, medical devices. The medical devices industry offers an interesting example of a high-tech and high-value manufacturing sector. It is characterized by strong and growing global demand as populations age and healthcare expenditure expands. High technological and regulatory barriers to entry reduce footloose investments and require a skilled and capable workforce. A diverse range of products, from relatively simple, low value items such as syringes and sutures to high-risk devices such as pacemakers, provides ample scope for product upgrading in the sector. Once dominated only by developed nations, it is rapidly globalizing, providing opportunities for a growing number of developing countries to join the industry, especially in production stages of the chain. At the same time, global employment is expanding, particularly in these new locations; a large share of these workers appears to be female. 2 Made in China 2025 is emblematic of this shift in industrial policy amongst emerging economies. 3 This study utilizes a gendered GVC framework to analyze the dynamics of female participation in two emerging countries in the industry: Costa Rica and the Dominican Republic. Three key questions are addressed: (1) what is female intensity in the industry and how does it compare to the manufacturing sector as a whole; (2) does female intensity change over time as the sector grows and more technologically sophisticated products are manufactured in a particular location, and; (3) do changes in wages as a result of upgrading affect female intensity in different roles within the industry. Notably, the study uses firm-level data to isolate the employment effects of GVC-only activity from manufacturing oriented to the domestic economy; the research is also supported by stakeholder interviews. Key findings from the study have relevant implications for policymakers seeking to use high tech GVC engagement to drive economic and social goals. First, the sector expanded quickly in both countries, contributing significantly to national export baskets, making it an attractive sector for GVC entry. Second, women account for over half of employment, considerably higher than their participation in the manufacturing sector as a whole, as well as export processing zones (EPZ). Third, this female participation holds steady as the sector grows, upgrades into new, higher value products and as wages rise; and forth, these predominantly female jobs are good quality jobs despite being primarily in production stages; that is, they are sustainable and characterized by permanent contracts and higher than average wages. Gender sensitive policies in both locations in terms of access to education and supportive maternity leave approaches contributed to ongoing female participation, while industry characteristics for offering well-paying, long term, permanent employment were highly attractive for recruiting a growing number of single mothers in these countries. As such the sector offers both economic and social upgrading opportunities for its host locations, provided specific policy actions are taken. The study is structured as follows. First, we present a brief overview of the existing literature on gender and GVC-trade, focusing on the benefits of a gendered GVC approach, followed by a discussion of the methodology used. Second, an overview of the medical device GVC is provided to highlight key global industry characteristics, particularly with respect to the nature of work in the industry. Next, we examine the experiences of the two countries cases, focusing on each nation’s upgrading trajectory and gender dynamics within the industry. This is followed by a discussion of key factors raised in the literature which may explain the distribution of women/men 4 seen in the two case studies. Finally, we conclude by considering the policy implications of our findings. 2. Literature Review and Conceptual Framework Over the past two decades, a growing body of literature has emerged on gender issues in developing country participation in GVCs. This research has focused primarily on the first industries to offshore production, that is, light manufacturing sectors, such as apparel, increased international trade in agriculture, and consumer electronics (1990s-mid-2000s). Global employment in these GVCs is characterized by high levels of female participation, which has led to theories of feminization of trade. Much less GVC research has been carried out on the gendered impacts of industries that have globalized since then. However, numerous export-oriented industrialization (EOI) studies have indicated that upgrading into these sectors results in declining female intensity, contributing to the emergence of a feminization-defeminization debate around trade (Kucera & Tejani, 2014; Saraçoğlu et al., 2018; Tejani & Milberg, 2016). This section briefly reviews these analyses before outlining the conceptual framework to be used in this study. The early sectors to globalize were labor-intensive and cyclical in nature (Barrientos et al., 2011). Scholars have highlighted that countries entered into these light manufacturing and agricultural sectors in labor-intensive operations by tapping into large pools of previously unemployed female workers (Barrientos, 2014a; Kucera & Tejani, 2014; Pickles & Godfrey, 2012; Seguino, 2000). This resulted in high levels of female participation, particularly in sectors concentrated in export processing zones (Tejani, 2011); on average, 60-80% of apparel production jobs in the leading exporting countries are women (Barrientos, 2014a; Lopez-Acevedo & Robertson, 2016). Similarly, horticulture and floriculture GVCs have very large female shares (Christian et al., 2013). For many of these women, these industries provided them with their first opportunity to enter waged employment (Dolan & Sorby, 2003; Maertens & Swinnen, 2009; Pickles & Godfrey, 2012). The high levels of female employment in these sectors prompted considerable research as to why (Barrientos, 2001; Barrientos & Kabeer, 2004; Bussolo & De Hoyos, 2009; Seguino, 1997). Findings from this work resulted in a persistent gender and GVC narrative arguing that women are 5 overly represented in lower paying, labor intensive roles which rely on precarious conditions for female work to respond to fluctuations in global demand (Barrientos, 2014a; Fontana, 2006; UNCTAD, 2015). Key factors identified as contributing to this phenomenon are the gender wage gap, perceived flexibility and gender norms and stereotypes. Lower female wages meant women are more cost effective for low value stages of the value chain (Deere, 2005; Fussell, 2000; Seguino, 1997, 2000), while as secondary breadwinners, women were seen as being better positioned as flexible workers (Barrientos, 2014a; Barrientos et al., 1996; Seguino et al., 2010). Where women have failed to move into higher level positions in these industries, this has largely been attributed to lack of access to productive resources such as training and education, networks and access to finance (Fontana, 2011; Staritz & Reis, 2013). Considerably less gender research has focused on the GVCs of industries that have offshored since the mid-2000s. These industries as a group are characterized by higher technology- , knowledge- or capital-intensity, such as aerospace, chemicals and information technology. With lower employment impacts than their labor-intensive predecessors, these industries have had a less pronounced impact on female employment in host countries. They have thus attracted less attention from scholars and policy makers alike. With the exception of knowledge-intensive sectors such as IT services (Powell & Chang, 2016; Sarkar & Mehta, 2018), few industry-specific gender analyses have been undertaken mapping the globalization of these industries. Gendered analysis of participation in these sectors, rather, has been dominated by the export-led industrialization (EOI) literature. Collectively, the EOI studies have found that as countries upgrade their economies into high tech and capital-intensive industry segments/sectors for the export market, female intensity in employment declines; that is, the ‘defeminization’ of trade. Prominent scholars contributing to this theory include Caraway (2007) for East Asia; Kucera and Tejani (2014) for a sample of 36 countries,3 Jomo (2009) in East Asia, Fussell (2000) in Mexico and Berik (2000) in Taiwan. Scholars attribute both inter and intra-industry declining female employment in high technology industries/segments to wage changes and gender differentiated access to productive resources (Saraçoğlu et al., 2018; Tejani, 2011). They argue as 3 Kucera and Tejani (2014) find that technological upgrading within industries has a particularly strong effect on defeminization, which is intensified if countries move out of the labor-intensive apparel sector and into those characterized by metal fabricated products. 6 countries deepen their engagement in global industries, the gender wage gap declines due to labor market tightening, the cost advantages of hiring women are eroded and men tend to be favored over women (Ghosh, 2002; Kucera & Tejani, 2014).4 As with the gender-GVC literature on low- value sectors, the productive resources access argument again posits that, as higher level of capital or technical expertise or networks are required, the barriers women face to access training and education prohibit them from advancing into more sophisticated activities within an industry or into higher technology sectors (Ahmed, 2013). In both sets of literature, gender norms and stereotypes are highlighted as major contributors to this feminization-defeminization paradigm, while at the same time, women’s reproductive responsibilities are emphasized as a major constraint to their potential to respond to opportunities in expanding export-oriented economic activities. In many parts of the world, women are viewed as preferential for certain jobs that require high levels of dexterity, or jobs that are considered to be female – such as sewing, food preparation and housekeeping, while men are favored for machine operation and maintenance (Bamber & Fernandez-Stark, 2013; Barrientos, 2001). Meanwhile, their ability, or interest, in pursuing roles with more responsibility is said to be dampened by time constraints due to their primary responsibility for reproductive, care and unpaid work, such as domestic work, childcare, and caring for the sick and elderly (Ferrant et al., 2014). Consequently, women may opt out of promotions or leadership positions, and in some cases, completely exit the industry, creating a leaking talent pipeline (Cabrera, 2009). Despite the growing literature, this research to date remains somewhat inconclusive. This is, in part, the result of two important limitations: lack of specificity, on one hand, and lack of generalizability, on the other. First, much of the existing quantitative work is done on the aggregate level due to poor data availability or disaggregation. A large number of these studies utilize UNIDO INDSTAT national employment data at the ISIC 2D level.5 This limits the ability to unpack export-oriented jobs versus domestically focused employment; to understand how specific manufacturing GVCs may operate and the impact of the lead firms that drive them; or the 4 Jomo (2009) argues that this was not the case for East Asian countries, and that the trends were in fact independent of each other. 5 Including Caraway (2007); Jomo (2009); Kucera and Tejani (2014); Saraçoğlu et al. (2018); Tejani and Milberg (2016). 7 differential effects at different stages of a GVC (e.g. production versus processing; engineering versus assembly).6 The approach also assumes spillovers from export operations into domestic manufacturing industry in terms of employment practices and trends; this is often not the case in GVCs in developing countries where EPZ based firms remain fairly isolated (Farole, 2011). Second, GVC case studies have focused on gender dynamics in specific countries in globalizing industries, yet these gendered GVCs have been limited to select labor-intensive industries (Caraway, 2007): apparel, electronics, horticulture, tourism and offshore services (Barrientos, 2014a; Christian et al., 2013; Fontana, 2006; Staritz & Reis, 2013). As a result, the literature lacks a generalizability of findings. More specifically, gender dynamics in high value manufacturing GVCs remains underexplored. This study seeks to contribute to this literature by taking a high-tech sector-specific approach and analyzing the gender implications of, and impacts on, trade and competitiveness in a single high value manufacturing sector, medical devices. This study utilizes a gendered GVC framework based on firm-level data to isolate the employment effects of GVC-only activity. Three key questions are addressed: (1) what is female intensity in the industry and how does it compare to the manufacturing sector as a whole; (2) does female intensity change over time as the sector grows and more technologically sophisticated products are manufactured in a particular location, and; (3) do changes in wages as a result of upgrading affect female intensity in different roles within the industry. This gendered GVC mapping approach can be replicated across countries and sectors.7 A gendered GVC approach is particularly useful to develop a nuanced understanding of how countries participate in these industries and their impact on men and women. Specifically, the GVC framework8 maps globalized industries into segments and roles/activities and analyzes which firms, countries and workers undertake each of these activities (Gereffi & Fernandez-Stark, 2016). 6 Tejani and Milberg (2016) use 2D ISIC codes to analyze whether wage gaps and differential education outcomes account for lower female shares in more technologically advanced sectors, arguing that technological intensity matter more than export intensity based on broad manufacturing data. However, they mention that such broad categorizations of manufacturing could mask sector and country specific experiences. 7 See Frederick (2019) for further details on this GVC mapping approach. 8 GVC analysis focuses on understanding the nature of the global fragmentation of industries into order to maximize benefits from the increasing integration of the world economy. It examines geographically dispersed operations performed by workers of varying skill, located around the world, but intrinsically linked as a single system. 8 A gendered GVC framework examines how men and women are distributed across these activities, helping to identify why they may be concentrated in particular nodes or jobs and thus providing insights as to how to mitigate potential bottlenecks to trade and competitiveness (Bamber & Staritz, 2016). At the same time, by analyzing which firms undertake each activity, it allows policymakers to better understand how the organization of lead firms, whose decisions drive the industry, may impact gender outcomes. These lead firms dampen or reinforce gender norms through activities such as recruitment, promotion, skill development, knowledge transfer, procurement, working conditions and possibilities to combine work and family life (Bamber & Staritz, 2016). The framework allows for dynamic analysis over time as countries change the way they participate within a particular industry; this is broadly referred to as ‘upgrading’ (Bamber & Staritz, 2016; Gereffi & Fernandez-Stark, 2016; ILO, 2016). A number of upgrading strategies have been identified in the GVC literature to date, covering both economic and social dimensions of chain participation (Gereffi & Fernandez-Stark, 2016). Economic upgrading strategies include improving efficiencies via technology and production advancements (process upgrading); producing higher value or more sophisticated products (product upgrading); expansion or shifting into new stages of the value chain (functional upgrading) amongst others (Humphrey & Schmitz, 2002). Social upgrading strategies focus on creating more and better jobs with good working conditions and opportunities for advancement (ILO, 2016). A gendered GVC framework further unpacks these dynamic economic and social upgrading elements to identify whether these are associated with differential outcomes for men and women. Building on the variables detailed for a gendered GVC framework presented in Bamber and Staritz (2016) (see Annex 1), the discussion of the resulting gender distribution in this high-tech sector is framed around five key issues that emerge in the existing literature. Collectively, these five issues are important from a policy perspective in terms of promoting practices that encourage female empowerment while driving trade competitiveness. (1) Do differing work conditions exist between men and women within a high-tech industry and if so, do these contribute to gender roles.9 (2) Could differential access by men and women to relevant productive resources (Barrientos, 9 This has been a key trend in light manufacturing, agriculture and tourism industries (Barrientos, 2001, 2014b; Christian et al., 2013; UNCTAD, 2014). 9 2014a, 2014b; Fontana, 2011) – in this case, education -- result in poor participation of women in sectors with higher technology requirements (e.g. Tejani and Milberg (2016), Kucera and Tejani (2014)). (3) Is there a gender wage gap effect? Lower female costs helps to explain females predominance in low-value, labor-intensive segments of many GVCs, such as assembly activities (Barrientos, 2001), are these replicated in high tech sectors as well? (4) Do reproductive responsibilities impact women’s ability to work in high-tech GVCs and what policies may support their retention in the workforce. In order to empower females, policies such as maternity leave and programs to facilitate breastfeeding and childcare are often needed. (5) Are gender norms at play in the industry? Along the GVC, multiple and sometimes divergent norms operate. For example, the norms of lead firms, such as commitments to equality and diversity are seen as driving the chain. However, these operations are embedded into multiple local contexts where the chains touchdown. Gender norms, such as perceived ability in detail-oriented tasks, demands for higher level of care work and reproductive responsibilities can limit female participation and opportunities (Bamber & Staritz, 2016; Staritz & Reis, 2013; Tejani, 2011). 3. Methodology A mixed methods approach was used to develop this analysis, including both quantitative and qualitative sources. Quantitative sources include national trade statistics published for both countries in UN Comtrade database for 2004-2016 and firm-level trade and employment statistics for each country for firms in the sector gathered in the Annual Census of Export Processing Zones; in the case of Costa Rica, this data was then cross-referenced with data from Banco Central de Costa Rica, filtering for firms registered as medical supplies and equipment manufacturers (AE80).10 Firm-level trade data included exports (HS2002, 6D, value, 2000-2016 Costa Rica, 2012-2016 Dominican Republic) and imports for limited years. For Costa Rica, employment data (number of workers, total wages) was available at the firm level for 2000-2016, although gender segregated data was only available between 2004-2011, somewhat limiting the sample. For the Dominican Republic, employment and gender data (number of workers in production, technical and professional roles; wages for production and technical workers) were available for 2006-2016. The resulting dataset for analyzing female intensity included 19 firms in Costa Rica and 25 firms 10 AE80 is approximately equivalent to ISIC Rev 3 code 3250. ISIC Rev. 4 further divides medical device manufacturing into two codes, including 3250 and 2660. 10 from the Dominican Republic, although entry years differ.11 While the number of firms overall is small, limiting analytical methods, firms accounted for >85% of sector exports in the relevant years for both countries. Firms were first mapped onto the GVC by segment and product type using the value chain and product categories identified in Bamber and Gereffi (2013). Value chain segment was determined by the composition of firm exports as components manufacturers and assemblers.12 In Costa Rica, following the methodology of Frederick (2017), assembly firms were then further defined by the specific types of products exports using HS-2002 classifications based on a 50% share of exports, allowing for analysis of sophistication of output. In the Dominican Republic, the majority of exports are categorized under one HS code regardless of output making additional analysis difficult. Available employment and gender data were then analyzed. Finally, data analysis was complemented by 28 semi-structured interviews with firms, educational institutions and public sector agencies in both countries carried out at intervals between July 2012 and November 2018. Firm interviews were conducted primarily with general/plant managers and human resources managers to understand how plants undertake economic upgrading, the characteristics of the workforce and firm approach to human capital development. It should be taken into account that no workers were interviewed for this study. 4. The Global Medical Device Industry The medical devices industry13 is a high-tech and high-value manufacturing sector14 with strong global female employment that deviates from widespread norms of high male dominance of manufacturing. It is a global growth industry; with trade reaching US$203B in 2016 (UN Comtrade, 2018).15 Global demand is driven by populations that are growing (emerging markets) 11 A breakdown of the division of these firms is available in the Annex. 12 Services firms were not included in the data analysis. In Costa Rica, in 2014, the largest share of services in the medical devices sector was the payment for royalties, accounting for over half of services included (Frederick, 2017). 13 The medical devices industry covers a wide spectrum of products including “all instruments, appliances and materials that are designed for diagnostic and/or therapeutic purposes to monitor, treat, prevent or alleviate disease, injuries or handicap and that do not strictly achieve their action by pharmacological, immunological or metabolic means” (WHO, 2017). 14 Based on OECD (2011) Technology Intensity Definition. 15 See Bamber & Frederick (Forthcoming) for industry definition. 2016 was the latest available comprehensive data at the time of publication 11 and aging (mature markets), rising incomes and expanding access to health care (BMI Research, 2018; CFRA, 2018).16 The production of these devices is concentrated in a relatively small number of global lead firms, which account for more than half of the world’s market share. These firms historically focused their production in developed countries with solid technical capabilities and intellectual property protection (Bamber & Gereffi, 2013). Nonetheless, increasing price pressure from buyers (BD, 2018; Boston Scientific, 2018; Medtronic, 2018), has spurred the industry to globalize and outsource production, first to increase economic efficiencies, but also to harness qualified human capital abroad and access new markets (Brocca et al., 2017). Numerous countries from Latin America, Asia and the Middle East have developed industrial policies to attract and/or develop these firms (Bamber & Frederick, Forthcoming; BMI Research, 2018; Field Research, 2018; World Bank, 2011). Several characteristics of the offshoring patterns of this industry make it particularly attractive. First, investments tend to be long-term in nature as a result of long-transfer times, and relatively low competition. Barriers to entry are considerable due to the high costs of research and product development (R&D) and regulatory compliance, while installation of new product lines can take up to three years. Second, investment sustainability is facilitated by steady demand; driven by evolving healthcare needs, demand is relatively predictable and correlated with growing and aging populations and healthcare spending plans. Third, the sector covers a diverse range of products, from relatively simple, low value items such as syringes and sutures with low regulatory coverage to high-risk devices such as pacemakers, which require strict regulatory controls and technical know-how. This wide range of product offerings provides ample scope for entry and product upgrading in the sector – even in basic assembly operations. Finally, companies typically expand production operations as plant capabilities improve, allowing countries to functionally upgrade across value chain segments.17 As a result, the sector offers a more sustainable alternative to that in GVCs such as apparel or electronics, with high potential for long term product, process and functional upgrading over time. 16 From 2004-2016, global imports of medical devices grew 120% with developing countries, particularly those in Asia, being the fastest growth markets with significant long-term potential (UN Comtrade, 2018). 17 While high value segments of the GVC – namely R&D, have remained close to headquarters, various stages of production have been offshored. Assembly tasks for low-value items were the first to be relocated with lower cost countries offering skilled labor at a considerable discount, with more sophisticated products and then components manufacturing, distribution and sales following, capitalizing on long term engagement with lead firms. 12 These characteristics have important implications for sector employment. While the size of the global medical device workforce is still small, estimated at 1.5-2M in total (UNIDO, 2018), it is growing rapidly, particularly in offshore production locations. Headcounts in manufacturing hubs, including Costa Rica, Singapore and China, grew by over 50% between 2008-2015, as medical devices firms have expanded into new destinations. Workforce quality plays a critical role in the industry’s globalization. Human capital has been identified in certain cases as the single most important factor driving site selection in the sector (Field Research, 2012; Kimelberg & Nicoll, 2012). The workforce consists of skilled and semi-skilled workers who hold permanent jobs within firms as opposed to contractual labor arrangements seen in other industries (Bamber & Frederick, forthcoming). Approximately one third of roles require either a two-year degree, technical or vocational training, while the remaining two-thirds require a minimum of high school and a few months to a year of on the job training (BLS, 2018; O*Net OnLine, 2018). Globally, the benefits of employment in the medical device industry appear to cross gender lines. Analysis of available gender disaggregated employment data for the industry shows that the medical devices GVC is characterized by sector-wide gender equity, and the female share of the workforce tends to be higher as compared to the manufacturing sector as a whole. Table 1 details female employment in the sector for the 16 countries reporting to UNIDO INDSTAT, together with the United States (US). These countries collectively account for 33% of medical devices exports in 2016 (UN Comtrade, 2018). Female employment accounts for 49%, compared to 39% in manufacturing. Leading exporters of these products (US, China, and Mexico) have similar shares of male and female employment. In 2015, 44% of the US medical devices workforce was female (BLS, 2018); in China, 50% (UNIDO, 2018), while in Mexico, the share was slightly higher with 59% (INEGI, 2018). 13 Table 1. Female Employment in Medical Devices Manufacturing, by Country All Medical Devices All Manufacturing Difference Total Female Total Female in Female Reporter Employees Employees Share Employees Employees Share Intensity China 409,457 203,067 50 83,403,475 33,566,689 40 10 United States* 309,200 134,600 44 12,335,167 3,359,417 27 17 Mexico 111,796 65,429 59 3,442,108 1,197,544 35 24 Turkey 22,242 7,680 35 3,427,395 805,576 24 11 Malaysia 19,594 11,047 56 2,088,001 688,368 33 23 Viet Nam 17,721 12,141 69 6,097,051 3,695,916 61 8 Others 10,536 4,189 40 3,366,831 876,657 26 14 Total 900,546 438,153 49 114,160,028 44,190,167 39 10 Source: BLS (2018); UNIDO (2018) Notes: All Medical Devices, based on ISIC Rev. 4 4D: 2660 Irradiation/electro medical equipment & 3250 Medical and dental instruments and supplies. All Manufacturing, C. European countries account for approximately 1/3 of global employment in the sector, however, they do not report female intensity. *US Employment statistics based on NACIS 3391, Employment, Hours and Earnings, Current Employment Survey 5. Gender Dynamics in the Medical Devices GVC: Costa Rica and the Dominican Republic To explore how gender dynamics may evolve in medical devices GVC trade and the roles of different firms in driving these changes, we look at the experience of two countries, Costa Rica (CRI) and the Dominican Republic (DRI). Each is an emerging producer within the industry, although with over a decade of experience. The medical devices industry is considered a priority sector for investment agencies in both CRI and DRI as part of their efforts to diversify exports and upgrade. Further, each nation is growing and upgrading into new value chain activities and/or product categories. From 2006 to 2016, sector exports from Costa Rica and the Dominican Republic grew at 296% and 131% respectively (UN Comtrade, 2018). Both countries have maintained high levels of female participation over time, despite this considerable export growth (Figure 1). An overview of participation in the GVC is provided for each country, followed by analysis of gendered distribution of roles and how this has changed as the countries have upgraded. 14 Figure 1. Costa Rica & Dominican Republic Exports in the Medical Devices GVC, 2006-2016 3,000 100% Female Employment (% total) 90% 2,500 80% Export (US$M) 2,000 70% 60% 1,500 50% 40% 1,000 30% 500 20% 10% - 0% 2006 2008 2010 2012 2014 2016 Costa Rica Exports Dominican Republic Exports Costa Rica Share of Female Employment Dominican Republic Share of Female Employment Source: CNFZE (2017); PROCOMER (2012); UN Comtrade (2018). Costa Rica’s participation in the medical devices industry dates back to 1985, when the first device companies - Baxter Healthcare and Abbott - established operations there. Export growth began in the early 2000s, picking up as the industry globalized. By 2016, medical devices exports reached US$2.5B, 1.2% of global medical devices exports and 17% of the country’s merchandize exports. Just a year later, the sector surpassed Costa Rica’s agricultural industry to become the largest export sector (Alvarado, 2018; Lo, 2018; UN Comtrade, 2018).18 In 2016, medical devices in the country employed just 16,417 people; 53% of these were women (PROCOMER, 2016). The Costa Rican medical devices sector consists of a consolidating base of foreign firms producing increasingly sophisticated products. This includes local subsidiaries of global lead firms serving an array of healthcare markets, such as Boston Scientific and St. Jude Medical, as well as a number of smaller firms focused on single segments such as cardiovascular and neurology. There is only one domestically owned original equipment manufacturer in the country; other local firms 18 This is due both to sector growth, as well as the closure of Intel’s plant in 2014. 15 in the industry are concentrated in supporting activities. 70% of firms are concentrated in the manufacturing of components and assembly of final goods. Numerous firms run vertically integrated operations, carrying out all production related activities from sourcing of raw materials, through component manufacture and assembly to sterilization. In 2017 and 2018, a number of firms also established offshore service operations to serve global functions in human resources, and accounting, as well as carry out product related complaints management for both products manufactured in Costa Rica and abroad.19 The Dominican Republic entered the medical devices industry a little later than Costa Rica (late 1990s), however, the industry only gained significant momentum after 2006. Over the past decade, the sector has grown in importance to become one of the leading exports, worth over US$1.3B in 2016, 0.5% of the global medical device market and 12% of national export basket (UN Comtrade, 2018).20 Insertion in this GVC marks a departure from the country’s previous dependence on the comparatively volatile textile and apparel (T&A) industry (World Bank, 2017c). Today, medical devices production is one of the most advanced manufacturing industries in the country. While firms generally export devices in the surgical instruments product category (HS9018), output in fact covers a wide range of devices from sutures to ostomy bags and IV sets as well as parts for capital equipment. Sectoral employment doubled between 2006-2016, reaching approximately 20,695;21 women account for 67% of these jobs (CNFZE, 2017). As in Costa Rica, GVC participation is based primarily on investments of global lead firms. Six of the top ten global firms had established operations in the country by 2016 and 72% of accumulated investment (US$1.1B) in the sector entered the Dominican Republic since 2008 (CNFZE, 2017). These firms are located in EPZs which account for 100% of the country’s exports in the industry (Bamber & Frederick, 2018). The majority (90%) of these products are destined to the US, although Asia is the fastest growth market for the country (Hamrick & Bamber, 2019). Activities primarily focus on labor-intensive assembly and packaging, with the more established 19 These include Allergan, Boston Scientific, ICU Medical (former Hospira), Medtronic, and Hologic amongst others (Zolezzi, 2018). 20 Based on an estimated US$8.7B in exports (UN Comtrade, 2018). 21 Export, employment and investment information in this project proposal is based on data published by the Dominican Republic Consejo Nacional de Zonas Francas de Exportación (CNZFE). 16 firms carrying out sustaining engineering to support productivity improvements; 22 firms carry out assembly-only operations, while 3 firms export components. Notably, firm-level data analysis reveals a high degree of correlation between exports and employment in both countries (2016: Costa Rica, r= 0.79 (r2=0.89); Dominican Republic, r=0.94, (r2=0.62) as a result of the more labor-intensive production stage in which both countries primarily operate. 5.1. Mapping the Participation of Women in the Medical Devices Industry22 The medical devices sector in both countries is characterized by strong female participation that deviates from national manufacturing norms. This deviation is consistent with the global trends noted in Table 1. Female intensity is significantly higher than that of EPZs,23 national manufacturing sectors (approximately a 20% margin) and overall labor force participation in both nations (Table 2). In Costa Rica, it is the only EPZ sector in which women account for the majority of employment (PROCOMER, 2012), while in the Dominican Republic, the only sector that employs a higher share of women is apparel (INEC, 2017).24 Table 2. Female Intensity of Medical Devices Employment, EPZ, and Manufacturing Medical Devices a EPZs a Manufacturing b Labour Force c 2006 2016 2006 2016 2006 2016 2006 2016 Costa Rica 50* 53 40** 40 32 34 36 38 Dominican Republic 65 67 52 49 35*** 31 38 41 Sources: a) PROCOMER (2015, 2016); Own calculation based on CNFZE (2017) b) World Bank (2017b), c INEC (2017) (BCDR, 2017) Notes: *2004, **2010, ***2008. Female intensity is high across the value chain; however, this falls slightly with 22 Data analysis in this segment is limited to firms with available employment statistics for both countries (Costa Rica, n=19, Dominican Republic n=25). 23 The other sectors in which these countries participate via EPZs include Costa Rica: Services, Plastics, Metalworks, Electrical & Electronics, Food and Textiles & Apparel; Dominican Republic: Services, Textile & Apparel, Footwear, Jewelry, Agri-processing and Tobacco, Electrical & Electronics. 24 The high levels of female intensity identified by using firm-level data differ significantly from those indicated through national manufacturing statistics that were used in earlier EOI studies. This suggests that using the latter as a proxy for female participation in trade could be misleading. 17 participation in upstream production stages of the chain (Table 3). During the periods analyzed, both countries primarily participated in two stages of the value chain, assembly and components manufacturing. Employment is concentrated in firms undertaking assembly activities which account for >98% in both countries. In Costa Rica, however, many assembly firms are vertically integrated upstream into the production of some components for final goods exports. Components production typically requires more technical expertise and draws on more traditional manufacturing sector competencies. This vertical integration potentially contributes to Costa Rica’s lower female intensity relative to the Dominican Republic, which has very limited component production. Table 3. Female Intensity by Value Chain Stage Total Females Employees Share (%) Employees Share (%) Costa Rica (2011) 10,272 5,472 Components Exporters 115 1 54 47 Final Goods Exporters (assembly) 10,157 99 5,418 53 Dominican Republic (2016) 20,695 Components Exporters 405 2 259 64 Final Goods Exporters (assembly) 20,290 98 13,534 67 Source: CNFZE (2017); PROCOMER (2012) Note: Firms are categorized by their final exports; final goods exporters therefore include those firms that are vertically integrated into the components segment of the chain. Table 4 below details the gender disaggregation of production, technical and professional roles in the Dominican Republic allowing for further analysis of the types of roles undertaken by men and women in both segments of the chain.25 There is a clear gendered division of labor that is consistent across both components manufacturing and assembly stages of the GVC. In both stages, female employment is concentrated in production roles. These roles usually include line jobs such as putting together components or final goods. Comparatively, women are underrepresented in technical roles, such as machine operators. While male employment is also concentrated in production, a larger share of male employees can be found in technical and professional jobs. 25 Comparable data was not available for Costa Rica. 18 Female intensity, or share of female participation in the workforce, is also highest in production jobs (74% assembly, 67% components), but is closer to gender parity in technical and professional roles, particularly in components manufacturing where women account for 47% and 52% respectively. Women feature particularly in professional roles such as industrial engineering, and transversal functions such as finance and human resources. However, there are very few women in firm leadership positions. In the Dominican Republic in 2017, only 3 firms had female in senior leadership (General Manager/Plant Manager) (Field Research, 2017). Table 4. Gender Composition of Workforce, by Job Type and Value Chain Segment in the Dominican Republic Total Employment Employment by Gender Share, Job Type (%) Female Total Share Male Female Male Female Intensity (%) Assembly Production 15,927 78 4,486 11,776 65 85 74 Technical 2,700 13 1,581 1,164 23 8 43 Professional 1,663 8 835 853 12 6 51 Total 20,290 6,902 13,793 100 100 68 Components Production 335 83 110 225 75 87 67 Technical 45 11 24 21 16 8 47 Professional 25 6 12 13 8 5 52 Total 405 146 259 100 100 64 Source: CNFZE (2017); PROCOMER (2012). Finally, female intensity is associated with firm employment size. Small firms (employment <100) in both countries have low female participation compared to larger firms (employment>500); in Costa Rica, female intensity was 21% lower, while in the Dominican Republic, this was 13% lower (see Table 6). Smaller firms are also more likely to export components than final medical devices, with a greater focus on technical production than assembly activities. 19 5.2. Effect of Upgrading in the Medical Device Industry on Female Intensity Economic upgrading in the medical devices GVC allows actors to capture more value from their participation (Humphrey & Schmitz, 2002). Numerous upgrading trajectories have been identified in the GVC literature to track how countries improve their participation in an industry. In the case of medical devices, two specific upgrading trajectories stand out that can be empirically tested using the available data: economic upgrading based on total exports and product upgrading based on shifting export composition. In each of these upgrading scenarios, the impact on female employment is analyzed. Overall, we find that upgrading in the medical devices GVC is associated with stable or increasing female intensity. This finding is consistent across the two upgrading indicators analyzed in this section. Economic upgrading by export growth In this scenario, export growth is analyzed as a simple measure of economic upgrading (Bernhardt & Pollak, 2016). This variable provides insights into participation of a country in the GVC: Is the medical devices sector growing, stagnating or declining, and how are firms performing, in general? Also, how does this performance measure vis-à-vis other countries in terms of global market share? Table 5 shows that Costa Rica and the Dominican Republic followed similar growth patterns during the periods analyzed (2004-2011 for Costa Rica, and 2006-2016 in Dominican Republic). Using UN Comtrade data to allow for cross-country comparison, it can be seen that both countries increased their sector exports by similar amounts, as well as increasing their share of world market exports. Between 2004/6 and 2011/6, exports grew by 120% and 131%, while they gained 0.11% and 0.09% global market share respectively. At an aggregate level, both countries upgraded. 20 Table 5. Economic Upgrading: Sectoral Export Growth and Female Intensity Costa Rica Growth Dominican Republic Rate 2004-2011 Growth 2004 2011 (%) 2006 2016 Rate (2006) (2016) (2006- 2006-2016 2016) (%) National Sector Exports (US$) 514M 1,128M 120% 573M a (634M) (2.509M) (296%) 1.325* 131% 0.56% 0.67% 19% 0.49% Share of Global Market (0.55%) (1.23%) (126%) 0.58% 17% Total Employment 3,755 10,272 174% 9,124 20,695 127% 50% 53% 65% 67% Average Female Intensity b (53%) (53%) Source: a UN Comtrade (2018), Based on HS-2002 product categories defined in Table A-2; all exporters; downloaded 27/08/2018. b CNFZE (2017); PROCOMER (2012). Note: National sector exports include exports of all firms in the country, including but not limited to those in the firm-level dataset. What impact has this expansion of industry exports had on employment, in general, and female intensity in particular? This sectoral growth drove an important increase in employment, with job creation increasing at a faster rate than exports in Costa Rica and a comparable rate in the Dominican Republic. Female jobs accounted for more than half of the new jobs created by the sector’s expansion during this period (55% and 68%) and at an aggregate level, this led to a small increase in female intensity of 2-3.6%. 21 Table 6. Economic Upgrading and Female Intensity at the Firm-Level Determinant No. Exports Ave. Female Change Female % All % All New % MD Firms Length Intensity Female Job New Jobs Growth (2011/6) Intensity Creation Female (Entry- Jobs Average Std Dev 2011/6) CRI: 2004-2011 (US$M) (US$M) DRI: 2006-2016 All firms 19 64.6 105.2 5.6 48% 5% 2,570 Female/new jobs: 55% 100% Labor <100 7 3.49 5.29 5 34% 12% 119 3% 5% 1% Force 100500)26 accounted for over 80% of new jobs, as well as 80% of new female jobs in both countries. Mean female intensity in these firms was over 55% across the period; in Costa Rica this decreased slightly (-2%), while in the Dominican Republic, female intensity increased (3%). Large employers notably accounted for over 70% of medical devices export growth in both countries. Firms that were already established 26 Firm size was determined by employment at the end of the period. 22 in the country at the beginning of the period (Early investors) also accounted for the largest share of total and female employment growth. Early investors drove industry export growth in Costa Rica (97%), while in the Dominican Republic, investment by new large firms during the period led to new firms accounting for 44% of job creation and 45% of export growth. Product Upgrading Next, we unpack the potential upgrading trajectories in effect for each country and how these may impact gendered outcomes; that is, is export growth the result of more of the same or of the incorporation of higher valued products and what opportunities are there for men and women? Using disaggregate trade statistics, we examine the changes, if any, of products shipped from each country to determine if product upgrading has occurred, that is, a shift into higher value products. This is generally measured using export unit value (Bernhardt & Pollak, 2016; Kaplinksy et al., 2009; Ponte & Gibbon, 2005), but in the medical devices sector, high degrees of heterogeneity of products included under respective trade classifications can result in misleading unit values. Thus, following Bamber & Gereffi (2013), we adopt a classification of products based on regulatory barriers and complexity that contribute to differentiated product margins; products are categorized into six groups from low value to high value. Product upgrading, in this case, thus refers to shifting into products with higher regulatory requirements and, therefore, higher value (Bamber & Gereffi, 2013). Table 7. Product Upgrading Exports (US$M) Growth Share of Exports 2004 2006 2016 (2006-16) 2006 2016 Costa Rica 513 634 2,509 296% Consumables 0.21 0.22 0.031 -86% 0.035% 0.001% Disposables 441 454 1,142 151% 72% 46% Surgical Instruments 38 135 726 439% 21% 29% Other 0.359 2 9 469% 0.25% 0.36% Therapeutics 32 42 506 1120% 7% 20% Capital Equipment 1 1 126 9744% 0.20% 5% Dominican Rep. 573 1,325 131% Consumables 12 7 -42% 2% 0.5% 23 Disposables 39 216 453% 7% 16% Surgical Instruments 521 1,015 95% 91% 77% Other 0.029 65 224,038% 0% 5% Therapeutics 0.168 16 9,423% 0% 1.2% Capital Equipment 1 6 500% 0% 0.5% Source: UN Comtrade, 2018. Based on HS-2002 product categories defined in Table A-2; all exporters; downloaded 27/08/2018. Table 7 illustrates that upgrading was a result of both more of the same, as well as a diversification into new products. Both countries experienced absolute growth within original product categories with the exception of low value consumables, which shrank. At the same time, the export profiles of both countries became more diversified, with higher value product segments accounting for an increasingly larger share of exports, and original product share declining. Absent in 2006, higher value product segments in the Dominican Republic accounted for 6.5% of exports by 2016. Analysis of FDA data showed significant product upgrading within these categories as well.27,28 Firms generally started assembling one product and now produce multiple product families in their operations. This changing composition of medical device exports shows that both Costa Rica and the Dominican Republic were moving toward product categories of higher technological content (product upgrading). How has this product upgrading affected female intensity? In the case of Costa Rica, it is possible to use firm-level data to analyze how female participation evolves as product upgrading takes place for three categories.29 The results are detailed in Table 8. At an aggregate level, there is an increase in female intensity as product sophistication increases, as well as in each product category (2004/2011: disposables 49/51% compared to instruments 51/52% and therapeutics (2006/2011) 57/59%). The trend persists using firm-specific analysis, in 2011 these categories 27 The number of products registered with the FDA for manufacture grew close to 20% between 2017 and 2018 alone to 726 (FDA, 2018). 28 The U.S. Food and Drug Administration (FDA) classifies medical devices according to their potential risk to the user (Unclassified, Class I to Class III). Class I devices are considered low risk, and do not require pre-market notifications prior to sale. Class III devices are considered to high risk and are subject to high levels of pre- and post-market scrutiny. 29 In the Dominican Republic, while the number and type of products has expanded significantly since 2006, many exporters do not differentiate between products codes for the wide variety of medical devices. This makes it difficult to analyze the impact on female intensity by product code. 24 accounted for 48%, 38% and 52% respectively. For surgical instruments, the mean is considerably lower than the aggregate as a combination of very large, female intense firms with very small, male intense firms (Std dev = 24%). Product upgrading across categories was mostly driven by the entry of new firms into the country focused on these more sophisticated products (Bamber & Gereffi, 2013).30 Firms are typically concentrated in individual product categories; only 4/19 firms had less than 95% of their output in their primary product category in 2011. This is likely the result of plants being run by specific divisions of multinational firms, rather than as company-wide operations. Table 8. Product Upgrading and Female Intensity No. of Average Share All Average Total Average Female Intensity Mean Firm Firms Exports Medical Years in Female 2004 2011 Female 2011 Devices Country Employees Intensity 2011 (US$M) Exports 2011 2011 (Stdev) Disposables 7 80.1 50% 6 2,460 49% 51% 48% (19%) Surgical 6 40.4 22% 6 1,704 51% 52% 38% (24%) Therapeutics 3 99.8 27% 3 714 57%* 59% 52% (20%) Note: * based on 2006. There were no Therapeutics exporters in the country in 2004. Source: Authors calculations based on Procomer, 2012. Functional Upgrading and Female Intensity Changes Functional upgrading is the undertaking of new, higher value activities. This assesses whether new value-adding activities are undertaken, including components production and services such as research and development and procurement (upstream) and sterilization, logistics, and post-sales services (downstream). While data was not available to empirically test whether this upgrading has occurred, or its impact on female intensity, based on extensive interviews of the clusters in both countries, we find that between 2006-2016, Costa Rica has achieved greater functional upgrading than the Dominican Republic. Costa Rica has increased its presence in a range of services, including sustaining engineering (2010/11), sterilization (2011), procurement (2016), and complaints management (2017) (Bamber & Frederick, 2018; Zolezzi, 2018). The Dominican 30 Analyzing process upgrading using the available data is not possible given the small number of firms within each category. 25 Republic functionally upgraded in terms of sustaining engineering (2012), and sterilization (2017) (Bamber & Frederick, 2018; CNZFE, 2018). Services oriented jobs in EPZs in Costa Rica are characterized by higher levels of female participation than manufacturing at 61% (2015) (PROCOMER, 2016). 5.3. Upgrading, Wages and the Gender Distribution of Role Finally, we analyze the impact of upgrading on industry wages and the distribution of female employment. The EOI literature suggests that as countries upgrade, wages rise which results in men being favored over women and a less feminized workforce. We use both aggregate statistics and firm-level data for early investors to unpack what impact, if any, occurred. At an aggregate level, wages increased 78% in Costa Rica (2005/11) and 27% for production roles and 30% for technical roles in Dominican Republic (2006/16) (calculated in current US$). Mean wage increases by early investors in DRI was even more pronounced at 94% and 47% respectively. These sectoral wages were higher than in most other EPZ sectors. In Costa Rica in 2011, only services offered higher compensation than medical devices (PROCOMER, 2016). The result is similar in the Dominican Republic (Field Research, 2017).31 Table 9. Wage Change and Gender Distribution of Roles Sector-Wide (Aggregate) Early Investors (Firm-Specific) Mean Average Average Mean Wage Average Employment Mean Wage Employment Change Female Intensity Growth Female Intensity (Stdev) Change Growth (Stdev) (Stdev) (2005-2011) 2005 2011 Costa Rica 78% 136% 52% 53% (2006-2016) 2006 2016 2006 2016 Dominican 127% 65% 67% Rep. Production 30% 115% 70% 72% 94% (37%) 118% (200%) 76% (16%) 76% (15%) Technical 27% 197% 27% 42% 47% (25%) 137% (143%) 36% (24%) 45% (23%) Professional NA 172% 51% 51% NA 204% (271%) 59% (17%) 54% (8%) Source: Own calculations based on aggregate and firm-level data for medical devices sector CNFZE (2017); PROCOMER (2012) 31 Real wages have stagnated during this period in the Dominican Republic (World Bank, 2016). 26 This wage increase, however, did not result in a decline in female participation. Aggregate results in both countries indicate that female intensity increased during this period. Available data in the Dominican Republic shows that sectoral female intensity in all three job categories increased; with technical positions increasing the most by 15% over the decade, reaching 42% by 2016. While female intensity in technical positions is the lowest, female technical positions were, in fact, the fastest overall growth category of employment (243%, compared to men 170%), despite the significant wage increase at both a sectoral (27%) and firm level (47%) over the period. Within- firm effects for early investors showed a similar large average increase for technical positions (9%). Qualitative analysis reveals that this increase in female technical roles was concentrated in quality oversight (e.g. laboratory technicians, quality technicians, etc.) and supervisory roles. These roles are key for increased adoption of new systems and technologies to drive productivity, ensure operations align with global quality requirements (i.e. process upgrading) and roll out new products (i.e. product upgrading).32 Thus, while overall sectoral wage increase was driven by a slight tightening of the labor market as more employees were hired into production and by an increase in skills requirement, neither of these resulted in a decline in female employment. In summary, three key findings should be highlighted: (1) Female employees account for the majority of employment in the medical devices sector in both countries, with a higher intensity than the overall manufacturing sector; (2) Female intensity holds steady as the countries expanded their exports and upgraded into new higher value product categories. Large, early investors played an important role in female job creation and export expansion; and (3) An increase in wages as a result of this upgrading does not result in within-industry defeminization. 6. Gender Dimensions and Feminization in the Medical Devices Global Value Chain This section examines the gender distribution outcomes observed in the context of five key issues identified in the literature as contributing to feminization/defeminization of export-oriented industries: (1) Precarious work, (2) Access to Education (i.e. productive resources), (3) Gender wage gap, (4) Supportive policies, and (5) Gender norms. Collectively, these five issues are 32 The number of sites in the Dominican Republic with ISO certification 13845 for medical devices increased more than fourfold (ISO, 2017). The impacts of potential functional upgrading into more technical component manufacturing is not captured in this effect; assembly firms indicated that they do not manufacture components at this stage. 27 important for policymakers to promote practices that encourage female empowerment while enhancing firm competitiveness. We examine how these factors may be contributing to feminization in the medical devices sector. Precarious Work Female precarious work has been highlighted as a major driver of high female intensity in other sectors. Women are said to be concentrated in cyclical or temporary production jobs due to their roles as secondary breadwinners in the family unit. In the two cases analyzed, however, temporary work in the industry is generally low due to stable demand; high regulatory barriers contribute to reduced competition avoiding major fluctuations in production output. Regulatory requirements also result in significant on-the-job training needs, and thus firms prefer to offer permanent contracts to retain staff over the long term. In the case of Costa Rica and the Dominican Republic, the share of temporary labor accounted for less than 1% of total employment during the periods analyzed (CNFZE, 2017; PROCOMER, 2012). Stakeholder interviewees suggested that the stability of employment in the industry helps to recruit women, particularly single mothers who have a lower potential to withstand employment fluctuations. In both countries, women account for a growing share of primary breadwinners, up to 40% of households.33 Stable employment with permanent contracts, which include access to social security benefits covering dependents in both countries makes the medical devices industry an attractive sector for these women. While no data is available regarding gender-segregated attrition, anecdotal evidence from interviewees suggests that this is lower for women than men, making female employees more desirable recruits for firms. Alternative job opportunities for women in both Costa Rica and Dominican Republic in the areas where EPZs are located are primarily sectors characterized by high levels of precarious work (e.g. apparel, electronics, and agri-processing). Only one industry in Costa Rica offers a compelling alterative: the global services cluster, co-located with medical devices companies, which pays higher salaries with similar conditions (Fernandez-Stark et al., 2013). This sector competes directly with the medical devices industries for male and female employees alike and the strong growth of the cluster likely 33 The latest available statistics for the Dominican Republic note that female headed 40% of households in 2015, compared to 36% in Costa Rica in 2013 (Leiton, 2014; World Bank, 2017a). Both countries have experienced an important increase in this figure over the past decade. 28 contributes to the lower levels of female intensity in Costa Rica compared to the Dominican Republic. Gender Wage Gap Central to the gender wage gap hypothesis is that low-value, labor-intensive segments of value chains, such as assembly activities, may be highly feminized because of comparatively lower female wages. While gender wage gaps persist in both countries at the national and sectoral level, it is relatively low and has declined in recent years (Enamorado et al., 2009). Costa Rica reports that the national gender wage gap declined from 5.6% in 2011 to just 1.8% in 2016 (OECD, 2018).34 The difference is slightly larger in the Dominican Republic, although this reaches 6% in 35 2016 (Instituto Nacional de Estadisticas, 2016a). Under the hypothesis highlighted in the literature, relative lack of a gender wage gap should lead us to expect a similar share of men and women in the industry that declines over time. First, neither country have exhibited declining female shares. Second, this tendency cannot explain the larger female intensity in the Dominican Republic (67%). Of course, national gender wage gaps may mask geographic, public/private sector, industry-specific and degree-specific wage differences. In the Dominican Republic, for example, women with a technical degree earn 66% of the salaries of their male peers (INTEC, 2016). Further research mapping wages gaps in key geographical areas for EPZ areas in which medical devices companies are found would be necessary to uncover if this a driving factor. Access to Education and Training Human capital requirements for the industry are relatively high compared to other sectors; therefore, access to education and training is a key determinant for female participation in the industry. At a minimum, production workers require high school education, technical workers should have a one to a two-year technical degree and professional staff, a Bachelor’s degree in a relevant field (O*Net OnLine, 2018). Additional specialized training programs and degrees focused on medical devices specifically can increase employability and advancement potential in the industry. 34 This makes it one of the lowest amongst OECD members. 35 Average hourly income, employed persons, by occupational sector and gender 2008-2016. 29 Overall, both countries demonstrate high female participation in all levels of general education (Table 10). Policies supporting equal access to education have been present in both countries for a considerable time; Costa Rica’s programs to foster gender parity in education were launched in the 1990s (Guzman Stein & Letendre Morales, 2003).36 The rate of women completing high school, technical and tertiary education programs is similar to that of men in both countries, except for technical education in the Dominican Republic where lower rates (42%) potentially contribute to lower female intensity in technical jobs in the industry. At a more granular level, however, detailed statistics in Costa Rica show there is still a shortage of female graduates of technical programs for the industrial sector (<18%), with the exception of quality control (50%) (ECCTI, 2018). Comparatively high shares of females in tertiary education (57%) correspond to gender equity in professional levels in the medical devices sector. Table 10. Access to Education and Training, Select Courses, Share of Women, 2016 Education Costa Rica Dominican Republic a Share of the population, % Women 50% 50% High School Enrollment, % Women a 50% 52% Technical Education, % Women 49%b 42%c Tertiary Education, % Women 55%b Post Graduate (Doctoral), % Women 51%b 57%c Industry Specific Training Programs Medical Devices Operator Course – Production d, e n.a. 62% (2017) Quality Inspector g 50% -- f 48% (Materials) 63% (2017) Engineering Specializations d, e 41% (Industrial) Masters in Medical Devices Engineering (TEC) f 37% -- Source: a World Bank (2018); b Instituto Nacional de Estadisticas (Costa Rica) (2016); c Instituto Nacional de Estadisticas (2016b); d ECCTI (2018); e Rodriguez (2018); f Cubero (2018). Specialized programs, developed in concert with the industry, have more mixed gender statistics. The new programs, launched over the past three to five years, could have an impact on the gendered division of labor in the future. Female participation in both operator and professional courses in the Dominican Republic reflect the high level of women in the industry, at 62% and 36 Ley General de Educación Número 66-97 promotes equal access to education in the Dominican Republic. 30 63% respectively. In Costa Rica, gender equity in operator levels begins to decline in post-graduate studies; the Masters in Medical Devices Engineering Program offered by Tecnológico de Costa Rica (TEC) – the most advanced industry-specific program in the country - has seen female participation fall from 37% to just 14% over the past three years (Cubero, 2018). Support for Reproductive Responsibilities The presence or lack of gender-sensitive policies is often considered to have an impact on female participation in an industry and potential for advancement. In both countries, there are several initiatives in place at the national, industrial park and firm-level that help to reduce reproductive burdens on women and thus to maintain them in the workforce. These efforts have helped to improve their participation in the workforce, in general, and in this sector in particular in both countries. Specifically, both countries have established laws to support maternity leave and post- natal breastfeeding allowances (Table 11). Overall attrition rates in the Dominican Republic, that has high levels of female participation in the production segment of the chain, are low (~1-2%). While attrition is higher in Costa Rica, this is attributed to a tight labor market rather than the exit of women from the workforce. Table 11. Select Gender Policies, Costa Rica and Dominican Republic Policy Costa Rica Dominican Republic Maternity Leave a Required by law (Labor Code Art. 94-100; Required by law; 100% pay for 98 days 1996); (pre & post-natal); extended from 84 to 100% pay for 120 days divided pre & post- 98 days in 2017 to align with ILO natal norms. Breastfeeding Required by law (1993) 1 hour/day up to 1 Required by law up to 1 year old Facilities/Times b year old (Politica Publica de Lactancia Materna) Employers with +30 female employees must provide a feeding room (regulation in process) Childcare facilities Not required by law Not required by law (park or elsewhere) Source: a WEF (2017); b León and Thomas (2017); MEPyD (2012) 31 Gender Norms and Relations Specific GVCs may have the potential to change gender employment patterns in the countries in which they operate. However, where gender norms of the host countries coincide with those of lead firms in the industry, these are likely to be reinforced in local operations (Bamber & Staritz, 2016). The majority of medical devices firms located in Costa Rica and the Dominican Republic, and particularly early investors, are of US or European origin; by 2011, in Costa Rica, US firms accounted for over 60%, while the share was even higher in the Dominican Republic where even the plants of European firms such as B. Braun (Germany) are managed by US company divisions (Bamber & Frederick, 2018). Legal requirements (i.e. Equal Employment Opportunity Act), combined with a strong commitment of leadership to diversity and inclusion goals and a substantial focus on gender equity amongst its client base,37 have helped to contribute to greater gender parity in the industry. Female participation is higher than in most other manufacturing sectors (Liftstream and MassBio, 2017), some 10% higher than overall manufacturing (see Table 1). However, female representation in mid-level management (29%) and senior leadership (24%) roles are still limited, and the female intensity decreases with each level (Liftstream and MassBio, 2017). In Costa Rica and the Dominican Republic, these firms have shifted manufacturing employment trends to match their own global operations. In both countries, the sector employs considerably higher share of women than manufacturing in general (see Table 2). Less equitable practices in senior leadership and management, on the other hand, have reinforced local norms. There are very few women in senior leadership roles; in the Dominican Republic, there are only three female general managers in the sector. This reflects similar results in the economy at large. Only 15% of firms in Costa Rica and 11% of firms in the Dominican Republic have women in senior leadership positions, and this share falls as firm size increases (ILO, 2017). 7. Conclusions and Policy Implications As technology-, knowledge- and capital-intensive industries, like medical devices, adopt 37 All major lead firms have either an established policy towards increasing female participation in the industry, or include female employment under the umbrella policy towards increased diversity and inclusion. The healthcare sector is one of the best performing sectors overall in the US in terms of gender equity. Of the top 100 Forbes list for best place for women to work, 25 are healthcare providers. 32 globalized production platforms, there are numerous opportunities for developing countries to upgrade in GVCs. While employment scale in these sectors is far below that of predecessors, such as apparel, the quality of the work is considerably improved. The two cases examined in this study illustrate that with the appropriate set of policies in place, these industries can contribute to more equitable employment outcomes for both men and women in trade. Key takeaways for policymakers include: These industries offer good quality jobs for women and female inclusion in the workplace can enhance trade competitiveness. Globally, women tend to account for more than half of medical devices employment; on average, the female intensity is 10% higher than overall manufacturing (Table 1). The country cases illustrate that employment tends to be characterized by relatively high wages and permanent contracts. These characteristics make the industry attractive for the rising share of female breadwinners in many developing countries, in Costa Rica, the Dominican Republic and beyond. These women are increasingly looking for the same degree of stability in their jobs as men, particularly those with permanent contracts with benefits that extend to their dependents. As a result, women will be just as likely (if not more) to seek out stable job opportunities as men, and their low attrition rates can enhance overall competitiveness. This changing dynamic not only boosts available human capital for trade and upgrading, but also puts pressure on governments to attract GVC investments with higher quality work. The type of FDI attraction is important for gender goals; GVC investors in these industries may differ in their gender policies and practices. Gender parity differs by segment, size, and product focus. Large scale MNC assemblers accounted for the largest share of export growth and female job creation, smaller investors and component manufacturers tended to have a comparatively lower female intensity. Specific products categories within these industries vary in their female intensity with more sophisticated therapeutics offering more opportunities for women than disposables; upgrading between product categories, however, generally requires new investors. Individual plants are typically run by specific MNC divisions and thus are unlikely to diversify into the export of products from other divisions. Overall, investors are likely to import their gender approaches; where these coincide with host country practices, gender norms will be reinforced rather than altered. For example, while the medical devices sector is strong in gender 33 equity in production, there is minimal participation of women in senior leadership at the global level, and in both cases of offshoring. This deterrent to long-term professional development can undermine long-term attraction and retention goals in the industry. Investment agencies seeking to enhance gender equity in GVC trade should include gender analysis of investors’ practices in other locations. Entry into and upgrading within this GVC and ensuring female employment gains requires specific policies. Participation in more knowledge and technology intensive sectors requires a series of key investment policies, from incentives to infrastructure. However, the most important factor is generally the availability of human capital with the relevant qualifications. This can be achieved by fostering increased cooperation between the sector and educational institutes as occurred in both countries. Ensuring gender equity in access to those specific qualifications is then the critical step to support gender goals. At the production and professional levels, equal access to and participation in high school education and industry-specific courses and engineering degrees provided the basis for high levels of female participation in both country case studies. In countries where females have low technical participation, drawing attention to niche job profiles can provide a strong entry point. The increase in female participation in technical training programs – particularly in quality control, had a strong impact on female technical hiring in both countries. Quality oversight is critical to competitiveness in medical devices and other highly regulated industries. Supportive family/maternity policies that help to reduce the burden of reproductive responsibilities foster female participation in EPZs. In highly regulated sectors, where competitiveness depends on a well-qualified workforce with strong knowledge of industry requirements, retention is critical. Both countries have government and firm-level policies, such as maternity leave, and breastfeeding facilities. These policies that support women’s efforts balance their work-life roles ensure ongoing female participation and retention. These policies have been widely adopted in many developed countries and firms. This exploratory study has made important advances in using a firm-level data approach to mapping gender distribution in the medical devices GVC. Nonetheless, it has its limitations. 34 Specifically, it only analyzes two countries in the industry and uses primarily descriptive statistics due to the number of firms present. Future research can expand this analysis to a broader range of countries entering and upgrading in the industry to help further refine the reasons behind the unexpectedly high female participation in the sector. Indeed, further sector-specific research using a gendered GVC framework into similar high tech or knowledge intensive industries would help to strengthen these findings and identify additional opportunities where trade can help close the gender gap. These sectors differ from early offshoring industries in terms of cyclicality in demand and supply, governance approaches by lead firms, and investment patterns. 35 Bibliography Ahmed, Ghada. (2013). Global Value Chains, Economic Upgrading and Gender in the Call Center Industry. In C. Staritz & J. G. Reis (Eds.), Global Value Chains, Economic Upgrading and Gender: Case Studies of the Horticulture, Tourism, and Call Center Industries. Washington, D.C.: World Bank. Alvarado, Laura. (2018). Medical Device Exports From Costa Rica Surpass Agriculture Sector For First Time in History. The Costa Rica Star. January 31 from https://news.co.cr/medical-devices-main-product-costa-rica-exports-surpassing- agriculture-sector-first-time-history/70267/. Bamber, Penny and Karina Fernandez-Stark. (2013). Global Value Chains, Economic Upgrading and Gender: The Horticulture Industry. In C. Staritz & J. G. Reis (Eds.), Global Value Chains, Economic Upgrading and Gender: Case Studies of the Horticulture, Tourism, and Call Center Industries (pp. 11-42). Washington, D.C.: World Bank. Bamber, Penny and Stacey Frederick. (2018). Central America in Manufacturing Global Value Chains. Durham, N.C.: Duke Global Value Chains Center. ---. (Forthcoming). The Medical Device Global Value Chain. Durham, NC: Duke Global Value Chain Center, Duke University. Bamber, Penny and Gary Gereffi. (2013). Costa Rica in the Medical Devices Global Value Chain: Opportunities for Upgrading. Durham, N.C.: Duke Global Value Chains Center. Commissioned by the Costa Rican Ministry of Foreign Trade. Bamber, Penny and Cornelia Staritz. (2016). The Gender Dimensions of Global Value Chains. Geneva: International Centre for Trade and Sustainable Development. Barrientos, Stephanie. (2001). "Gender, Flexibility and Global Value Chains". IDS Bulletin, 32(3): 83-93. ---. (2014a). Gender and Global Value Chains: Challenges of Economic and Social Upgrading in Agri-Food: European University Institute. Robert Schuman Centre for Advanced Studies. ---. (2014b). "Gendered Global Production Networks: Analysis of Cocoa–Chocolate Sourcing". Regional Studies, 48(5): 791-803. Barrientos, Stephanie, Richard M. Auty and John Toye. (1996). Flexible Work and Female Labour: The Global Integration of Chilean Fruit Production. In Challenging the 36 orthodoxies (pp. 125-141): New York:St. Martin's Press; London:Macmillan Press. Barrientos, Stephanie, Gary Gereffi and Arianna Rossi. (2011). "Economic and Social Upgrading in Global Production Networks: A New Paradigm for a Changing World". International Labour Review, 150(3-4): 319-340. Barrientos, Stephanie and Naila Kabeer. (2004). "Enhancing Female Employment in Global Production: Policy Implications". Global Social Policy, 4(2): 153-169. BCDR. (2017). Encuesta Nacional de la Fuerza Laboral, 2017. Santo Domingo, Dominican Republic: Banco Central de la República Dominicana (BCDR). BD. (2018). Annual Report 2017. Franklin Lakes: Becton Dickinson. Berik, Günseli. (2000). "Mature Export-Led Growth and Gender Wage Inequality in Taiwan". Feminist Economics, 6(3): 1-26. Bernhardt, Thomas and Ruth Pollak. (2016). "Economic and Social Upgrading Dynamics in Global Manufacturing Value Chains: A Comparative Analysis". Environment and Planning A, 48(7): 1220-1243. BLS. (2018). Occupational Employment Statistics (OES) Survey. Washington, D.C.: Bureau of Labor Statistics (BLS), United States Department of Labor. BMI Research. (2018). Global Medical Devices Report: BMI Research, Fitch Group. Boston Scientific. (2018). Annual Report 2017. Boston: Boston Scientific. Brocca, Michele, Kevin D'Souza, Ted Sisko, and Samantha Baron. (2017). Medtech Manufacturing's Inflection Point: BCG. Bussolo, Maurizio and Rafael E. De Hoyos. (2009). Gender Aspects of the Trade and Poverty Nexus: A Macro-Micro Approach: Washington, D.C.: World Bank; Houndmills, U.K. and New York: Palgrave Macmillan. Cabrera, Elizabeth. (2009). "Fixing the leaky pipeline: Five ways to retain female talent". People & Strategy, 32: 40-45. Caraway, Teri. (2007). Assembling Women: The Feminization of Global Manufacturing. Ithaca: Cornell University Press. CFRA. (2018). Health Care Equipment and Supplies: S&P Global. Christian, Michelle, Barbara Evers and Stephanie Barrientos. (2013). "Women in Value Chains: Making a Difference". Capturing the Gains Revised Summit Briefing 6.3. CNFZE. (2017). Census de Zona Francas, 2006-2016. Santo Domingo, Dominican Republic: 37 Consejo Nacional de Zonas Francas de Exportación (CNFZE). CNZFE. (2018). First Medical Device Sterilization Company Comes to Dominican Republic. Santo Domingo: Consejo Nacional Zonaas Francas de Exportacion. Cubero, Jorge Mauricio. (2018). Coordinator Master of Medical Devices Engineering (TEC) Personal communication with P. Bamber. November 14. Deere, Carmen Diana. (2005). The Feminization of Agriculture? Economic Restructuring in Rural Latin America. Geneva: United Nationals Research Institute for Social Development. . Dolan, Catherine and Kristina Sorby. (2003). Gender and Employment in High-value Agriculture Industries. Washington, DC: World Bank, pp. 1-90. ECCTI. (2018). Talento Humano Técnico en Costa Rica. 2014-2017. Retrieved November 23, 2018, from http://eccti.or.cr/talento-humano-tecnico.html. Enamorado, Ted, Ana Carolina Izaguirre and Hugo Nopo. (2009). Gender Wage Gaps in Central American Countries: Evidence from a Non-Parametric Approach. Washington, D.C. : Inter-American Development Bank. Farole, Thomas. (2011). Special Economic Zones in Africa: Comparing Performance and Learning from Global Experience. Washington, D.C.: World Bank. FDA. (2018). Establishment Registration & Devices: Dominican Republic. In U. S. F. a. D. A. (FDA) (Ed.). Washington, D.C. Fernandez-Stark, Karina and Penny Bamber. (2018). Promoting Skills, Jobs and Competitiveness in Global Value Chains: Policies and Practices for Developing Countries. In A. Sakamoto (Ed.), Skills and the Future of Work: Strategies for Future Growth in Asia Pacific. Bangkok: International Labour Organisation. Fernandez-Stark, Karina, Penny Bamber and Gary Gereffi. (2013). Costa Rica in the Offshore Services Global Value Chain: Opportunities for Upgrading. Durham N.C.: Center on Globalization, Governance and Competitiveness. Commissioned by the Costa Rican Ministry of Foreign Trade. Ferrant, Gaelle, Luca Maria Pesando and Keiko Nowacka. (2014). Unpaid Care Work: The Missing Link in the Analysis of Gender Gaps in Labour Outcomes. Paris: OECD Development Center. Field Research. (2012). Costa Rica: Summer 2012 Series. Personal communication with P. 38 Bamber, G. Gereffi & S. Frederick. July 2-6. ---. (2017). Dominican Republic Field Research. Personal communication with P. Bamber. December ---. (2018). Singapore Field Research. Personal communication with P. Bamber. July. Fontana, Marzia. (2006). Survey of Research on Gender and Trade: Insights, Gaps and Coverage. Paper presented at the Gender in Global and Regional Trade Policies: Contrasting Views and New Research ---. (2011). Integrating Gender Dimensions into Trade Projects. from http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/TRADE/0,,contentMDK:2288 2654~menuPK:2644066~pagePK:64020865~piPK:51164185~theSitePK:239071,00.html . Frederick, Stacey. (2017). GVC Modules and Applications. Paper presented at the Seminar on Accounting for Global Value Chains. from https://unstats.un.org/unsd/trade/events/2017/luxembourg/presentations/CH10_Frederick %20GVC%20Modules.pdf. ---. (2019). Global Value Chain Mapping. In G. Gereffi, S. Ponte & G. Raj-Reichert (Eds.), Global Value Chain Handbook. Cheltenham, United Kingdom: Edward Elgar Publishing. Fussell, Elizabeth. (2000). "Making Labor Flexible: The Recomposition of Tijuana's Maquiladora Female Labor Force". Feminist Economics, 6(3): 59-79. Gereffi, Gary and Karina Fernandez-Stark. (2016). Global Value Chain Analysis: A Primer. Second Edition. Durham, North Carolina, USA:: Duke University Center on Globalization, Governance & Competitiveness (Duke CGGC). Gereffi, Gary, Karina Fernandez-Stark and Phil Psilos. (2011). Skills for Upgrading: Workforce Development and Global Value Chains in Developing Countries. Durham: Duke University Center on Globalization Governance & Competitiveness. Ghosh, J. (2002). "Globalization, Export-Oriented Employment for Women and Social Policy: A Case Study of India". Social Scientist 30(11/12): 17-60. Guzman Stein, Laura and Anne Letendre Morales. (2003). Género y educacion en Costa Rica. Geneva: UNESCO. Hamrick, Danny and Penny Bamber. (2019). Pakistan in the Medical Devices Global Value Chain. Durham, NC: Duke Global Value Chains Center. 39 Humphrey, John and Hubert Schmitz. (2002). "How Does Insertion in Global Value Chains Affect Upgrading in Industrial Clusters". Regional Studies, 36(9): 1017-1027. ILO. (2016). Decent Work in Global Supply Chains. Geneva: International Labour Organization. ---. (2017). Women in Business and Management: Gaining Momentum in Latin America and the Caribbean. Geneva: Bureau for Employers' Activities, International Labour Organisation. INEC. (2017). Encuesta Nacional de Empresas, 2017. San Jose, Costa Rica: INSTITUTO NACIONAL DE ESTADÍSTICA Y CENSOS Costa Rica (INEC). INEGI. (2018). Encuesta Anual de la Industria Manufacturera 2013-2016. In INEGI (Ed.). Mexico City: Instituto Nacional de Estadistica y Geografia (Mexico). Instituto Nacional de Estadisticas. (2016a). Encuesta Nacional de Fuerza de Trabajo: Ingreso promedio por hora de los ocupados por sector de ocupacion y sexo 2008-2016. Cuado 3.40-4. Santo Domingo, Dominican Republic ---. (2016b). Encuesta Nacional de Fuerza de Trabajo: Población Económicamente Activa (PEA) por sexo y año, según nivel educativo, 2008-2016. Santo Domingo Instituto Nacional de Estadisticas (Costa Rica). (2016). Encuesta Anual de Hogares: Población de 15 anos o más por nivel de instruccion según zona, regi'pon de planificación y sexo. San Jose: INEC-CR. INTEC. (2016). Brechas de género: entre el dicho y el hecho de la autonomía de las mujeres en la República Dominicana. Santo Domingo, Dominican Republic: Instituto Tecnológico de Santo Domingo (INTEC); Centro de Estudios del Género. ISO. (2017). ISO Survey 2017. Geneva, Switzerland: International Organization for Standardization (ISO). Jomo, Kwame Sundaram. (2009). "Export-Oriented Industrialisation, Female Employment and Gender Wage Equity in East Asia". Economic and Political Weekly, 44(1): 41-49. Kaplinksy, Raphael, Jeff Readman and Olga Memedovic. (2009). Upgrading Strategies in Global Furniture Value Chains. Geneva: UNIDO. Kimelberg, Shelley McDonough and Lauren A. Nicoll. (2012). "Business Location Decisions in the Medical Device Industry: Evidence From Massachusetts". [Article]. Economic Development Quarterly, 26(1): 34-49. Kucera, David and Sheba Tejani. (2014). "Feminizations, Defeminization, and Structural Change in Manufacturing". World Development, 64: 569-582. 40 Leiton, Patricia. (2014). Uno de cada tres hogares en Costa Rica es dirigido por una mujer. La Nacion. May 12 from https://www.nacion.com/economia/indicadores/uno-de-cada-tres- hogares-en-costa-rica-es-dirigido-por-una- mujer/ENPCUPDCSNG6TKKWO7LLAVLZMY/story/. León, Viviano de and Juan Eduardo Thomas. (2017). A partir de hoy la licencia por maternidad se amplía de 12 a 14 semanas. Listin Diario. February 9 from https://listindiario.com/la- republica/2017/02/09/453436/a-partir-de-hoy-la-licencia-por-maternidad-se-amplia-de- 12-a-14-semanas. Liftstream and MassBio. (2017). Opening the Path to a Diverse Future: Creating Gender Balance in Massachusetts LIfe Sceines Sector. Cambridge: Massachusetts Biotechnology Council. Lo, Chris. (2018). Inside Costa Rica’s Super-Sized Medical Device Sector. Verdict Medical Devices. August 30 from https://www.medicaldevice-network.com/features/inside-costa- ricas-super-sized-medical-device-sector/. Lopez-Acevedo, Gladys and Raymond Robertson (Eds.). (2016). Stitches to Riches? Apparel Employment, Trade, and Economic Development in South Asia. Washington D.C.: The World Bank. Maertens, Miet and Johan Swinnen. (2009). Are African High-Value Horticulture Supply Chains Bearers of Gender Inequality? Paper presented at the FAO-IFAD-ILO Workshop on Gaps, Trends and Current Research in Gender Dimensions of Agricultural and Rural Employment: Differentiated Pathways Out of Poverty., Medtronic. (2018). Annual Report 2017. Minneapolis: Medtronic. MEPyD. (2012). Ley de Estrategia Nacional de Desarrollo 2030. Santo Domingo, Dominican Republic: Ministerio de Economía, Planificación y Desarrollo. O*Net OnLine. (2018). Home. Washington, D.C.: National Center for O*Net Development; sponsored by U.S. Department of Labor OECD. (2011). ISIC REV. 3 Technology Intensity Definition. Paris, France: OECD Directorate for Science, Technology and Industry. ---. (2018). OECD Data: Gender Wage Gap. from https://data.oecd.org/earnwage/gender-wage- gap.htm#indicator-chart. . Pickles, John and Shane Godfrey. (2012). "Economic and Social Upgrading in Global Apparel Production Networks". Capturing the Gains Briefing Notes(No 2). 41 Ponte, Stefano and Peter Gibbon. (2005). "Quality Standards, Conventions and the Governance of Global Value Chains". Economy and Society, 34(1): 1-31. Powell, Catherine and Ann Mei Chang. (2016). Women in Tech as a Driver for Emerging Economies. New York: Council on Foreign Relations. PROCOMER. (2012). Census de Zona Francas. San Jose, Costa Rica: Promotora de Comercio Exterior de Costa Rica (PROCOMER). ---. (2015). Balance de Zonas Francas: Beneficio Neto del Régimen para Costa Rica 2010-2014. San Jose, Costa Rica: Promotora de Comercio Exterior de Costa Rica (PROCOMER). ---. (2016). Balance de Zonas Francas: Beneficio Neto del Régimen para Costa Rica 2011- 2015. San Jose, Costa Rica: Promotora de Comercio Exterior de Costa Rica (PROCOMER). Rodriguez, Raquel. (2018). Adozona: Medical Devices Course Enrollment in Dominican Republic. Personal communication with P. Bamber. 13 November. Saraçoğlu, Dürdane Şirin, Emel Memiş, Ebru Voyvoda, and Burça Kızılırmak. (2018). "Changes in Global Trade Patterns and Women's Employment in Manufacturing, 1995–2011". Feminist Economics, 24(3): 1-28. Sarkar, Sandip and Balwant Mehta. (2018). What do Workers Gain from Being in a GVC? ICT in India. In D. Nathan, M. Tewari & S. Sarkar (Eds.), Labour in Global Value Chains in Asia. Cambridge: Cambridge University Press. Seguino, Stephanie. (1997). Export-Led Growth and the Persistance of Gender Inequality in the Newly Industrialized Countries. In J. Rives & M. Yousefi (Eds.), Economic Dimensions of Gender Inequality: A Global Perspective (pp. 11-33). London, Wesport, CT: Praeger. ---. (2000). "The Effects of Structural Change and Economic Liberalization on Gender Wage Differentials in South Korea and Taiwan". Cambridge Journal of Economics, 24(4): 437- 459. Seguino, Stephanie, Günseli Berik and Yana Van Der Meulen Rodgers. (2010). An Investment that Pays Off: Promoting Gender Equality as a Means to Finance Development. Berlin: Friedrich Ebert Stiftung International. . Staritz, Cornelia and Jose Guilherme Reis. (2013). Global Value Chains, Economic Upgrading and Gender: Case Studies of the Horticulture, Tourism, and Call Center Industries. Washington, D.C.: World Bank. Tejani, Sheba. (2011). The Gender Dimension of Special Economic Zones. In T. Farole & G. 42 Akinci (Eds.), Special Economic Zones in Africa: Comparing Performance and Learning from Global Experience (pp. 247-282). Washington D.C.: World Bank. Tejani, Sheba and William Milberg. (2016). "Global Defeminization? Industrial Upgrading, Occupational Segregation and Manufacturing Employment in Middle Income Countries". Feminist Economics, 22(2): 24-54. UN Comtrade. (2018). UN Comtrade Database: Medical Devices Exports to World (HS02-9018- 9022). Retrieved August 10, 2018, 2018, from http://comtrade.un.org/. UNCTAD. (2014). Trade, Gender and the Post-2015 Development Agenda Post-2015. Geneva: United Nations Conference on Trade and Development (UNCTAD). ---. (2015). Ensuring Gender-Sensitive Implementation of the Post-2015 Development Framework. Geneva: United Nations Conference on Trade and Development (UNCTAD). UNIDO. (2018). INDSTAT Industrial Statistics Database (ISIC Rev. 4; 4 digit). Retrieved July 8, 2018, from https://stat.unido.org. WEF. (2017). The Global Gender Gap Report. Geneva: World Economic Forum. WHO. (2017). Global Atlas of Medical Devices. Geneva: World Health Organization. World Bank. (2011). Moving up the Value Chain: A Study of Malaysia’s Solar and Medical Device Industries. Washingon DC: The World Bank. ---. (2016). Special Economic Zones in the Dominican Republic: Policy Considerations for a More Competitive and Inclusive Sector. Washington, D.C.: World Bank Group. Trade and Competitivenes Global Practice. ---. (2017a). Hogares encabezados por mujeres (% de hogares con jefa de hogar). Washington, D.C. ---. (2017b). Labor force, female (% of total labor force). Washington, D.C. ---. (2017c). Zonas Francas En República Dominicana: Consideraciones de Política para un sector más competitivo e incluyente Santo Domingo, Dominican Republic: World Bank. ---. (2018). World Development Indicators. World Development Indicators Retrieved November 15, 2018, from http://data.worldbank.org. Zolezzi, Sandro. (2018). Costa Rican Investment Promotion Agency (CINDE). Personal communication with P. Bamber. January 18. 43 Annex. Table A. 1. Gender Dimension Indicators of a Gendered GVC Framework Dimension Indicator Participation of women by sector, GVC stage, occupation, or job profile Patterns of segregation Women’s share of supervisory, managerial, professional, technical, and clerical work Vulnerable Women’s share of informal/contract/seasonal employment employment Women’s share in permanent jobs with contracts Women’s participation rate in TVET and other external training programs Access to training Women’s participation in “on the job” training programs Average wages by gender by GVC stage Wage gap Ratio of women’s wages to men’s wages for similar work Average number of working hours for women vs. men Number of occupational health and safety incidences for women vs. men Share of women with access to social security Working conditions Share of women with access to paid sick leave Share of women with access to maternity leave Provision of on-site child care and nursing facilities Share of female workers with discrimination/sexual harassment claims Source: Bamber and Staritz, 2016 Table A. 2. Medical Devices Product Categories, Based on Trade Data Classifications Product Product HS Code HS96 Codes 6-Digit (HS02-07 changes) Category Examples Aggregation 901831: Syringes, with or without needles 901832: Tubular metal needles and needles for sutures Needles, 901839: Needles, catheters, cannulae etc. (medical) (changes to Catheters, syringes, cannulae & the like in HS02) Disposables catheters, 90183, 3006.9 9018391010-90: Infusion equipment tubing, IV sets, 9018399010-20: Infusion and transfusion of serum bandages 9018399090: Other needles and catheters, cannulae and the like 3006.91: Appliances identified for ostomy use Medical & Dental 90184 901841: Dental drill engines (expands to dental drill engines, whether/not Surgical Instruments, 90185 combined on a single base with other dental equipment in HS02) Instruments Forceps, 90189 901842: Instruments and appliances, used in dentistry 44 Medical 3006.1 901850: Ophthalmic instruments and appliances (expands to “” nes 90.18 in Scissors, HS02) Dialysis 901890: Instruments, appliances for medical, etc. science, nes (expands to Devices, Instruments & appliances used in medical/ surgical/veterinary sciences, incl. Defibrillators other electro-medical apparatus & sight-testing instrument, nes in 90.18 3006.1 Suture materials, sterile catgut, etc. 902111: Artificial joints (changes to 902131: Artificial joints HS02) 902119: Orthopedic/fracture appliances, nes (changes to 902110: Orthopedic/fracture appliances in HS02) Artificial body 902121: Artificial teeth parts, hearing 902129: Dental fittings, nes aids, Therapeutic 902130: Artificial body parts, aids, and appliances, etc. (changes to 902139: pacemakers, 9021 Devices Artificial parts of the body other than teeth, dental fittings & joints in HS02) crutches, 902140: Hearing aids, except parts and accessories implants, 902150: Pacemakers prosthetics 902190: Orthopedic Appliances, nes (expands to appliances which are worn/carried/implanted in the body, to compensate for a defect/disability (excl. of 9021.10-9021.50) in HS02) 841920: Medical, Surgical or Laboratory Sterilizers 901811: Electro-cardiographs 901812: Ultrasonic scanning apparatus 901813: Magnetic resonance imaging apparatus MRI, 901814: Scintigraphic apparatus Ultrasound 901819: Electro-diagnostic apparatus, nes (expands to “” used in medical/ machine, X- surgical/dental/ veterinary sciences (incl. apparatus for functional exploratory 841920 rays, Patient examination/for checking physiological parameters), nes in 90.18) in HS02) Capital 90181 Monitoring 901820: Ultra-violet or infra-red ray apparatus (expands to “” used in Equipment 90182 Systems, medical/surgical/dental/veterinary sciences in HS02) 9022 Blood Pressure 90221: Apparatus based on the use of X-rays, whether or not for medical, Monitor surgical, dental or veterinary uses, including radiography or radiotherapy 90222: Apparatus based on the use of alpha, beta or gamma radiations, whether or not for medical, surgical, dental or veterinary uses, including radiography or radiotherapy apparatus 902230: X-ray tubes 902290: Other, including parts and accessories 300510: Dressings, adhesive; and other articles having an adhesive layer, Consumable Bandages and packed for retail sale for medical surgical, dental and veterinary use. s/Medical 3005 dressings 300590: Wadding, gauze, bandages ad similar (excluding adhesive dressings) Supplies impregnated or coated with pharmaceutical substances, packaged for retail. Other Breathing 9019 9019: Mechano-therapy appliances; massage apparatus; psychological aptitude- Appliances devices and 9020 testing apparatus; ozone therapy, oxygen therapy, aerosol therapy, artificial 45 other mechano- respiration or other therapeutic respiration apparatus; therapy devices 9020: Other breathing appliances and gas masks (excluding protective masks having neither mechanical parts nor replaceable filters) Source: Bamber & Frederick (2018) ; see also Bamber & Gereffi (2013) for another application. Table A. 3. Medical Devices Entrants Per Year, By Country Entry Year 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Costa Rica 7 2 2 0 4 1 2 1 - - - - - Dominican Republic - - 12 1 1 2 1 2 1 2 0 2 1 Note: Entry year based on first year employment data was reported. Firms present in the first year of analysis for each country (2004 and 2006 respectively) are noted as entrants for that year. 46 Table A. 4. Descriptive Statistics of Datasets for Firm-Level Analysis Total Exports Total Female Female Medical Devices (US$,M) Employment Employment Intensity Exports (US$,M) CRI Longevity 2004 2011 2004 2011 2004 2011 2004 2011 Entry 2004 2011 Mean 5.6 74.8 64.6 536 541 266 288 0.47 0.48 0.43 26.8 60.8 Median 6 16.7 14.9 94 199.08 50 110 0.50 0.52 0.50 0 11.6 Std Dev 2.5 110.0 105.2 790 757 391 407 0.18 0.20 0.23 71.5 99.9 Range 7 278.0 381.7 1,977 2,668 991 1,394 0.56 0.65 0.75 280 348.9 Min. 1 1.5 0.15 19 10 2 1 0.10 0.1 0 0 0 Max. 8 280.0 381.8 1,997 2678 993 1,394 0.65 0.75 0.75 280 349.01 Sum 106 747.8 1,227.7 3,755 10,272 1,854 5,472 3.3 9.0 8.3 482.6 1,154.8 Count 19 9 19 7 19 7 19 7 19 19 19 19 Total Exports Total Female Female Medical Devices (US$,M) Employment Employment Intensity Exports (US$,M) DRI Longevity 2012 2016 2006 2016 2006 2016 2006 2016 Entry 2012 2016 Mean 8.6 59.6 57.0 760 828 494 552 0.72 0.61 0.59 39.8 49.8 Median 10 13.4 16.6 742 355 500 225 0.74 0.68 0.67 1.3 6.4 Std Dev 3.8 68.0 70.0 598 948 374 682 0.15 0.18 0.19 61.9 71.9 Range 11 175.7 244.6 1,559 3,419 1,003 2,550 0.64 0.76 0.81 169.5 243.7 Min. 1 2,884 0.016 47 30 45 2 0.32 0.067 0.1 0 0 Max. 12 175.7 244.6 1,606 3,449 1,048 2,552 0.96 0.83 0.91 169.5 243.7 Sum 215 1,193 1,425 9,124 20,695 5,933 13,793 8.6 15.4 14.9 995.5 1, 246 Count 25 20 25 12 25 12 25 12 25 25 25 25 47