Female workers in textile and garment sectors in Sri Lankan Export Processing Zones ( EPZs ) : gender dimensions and working conditions

Sri Lankan women who are employed in Export Processing Zones (EPZs) work in a range of factories. A cross-section of female EPZ workers (n = 2304) representing 6 EPZs were surveyed as part of this study, with the majority in the textile and garment industries. Largely qualitative or small-scale research has demonstrated that textile and garment workers are often subject to poor social respect, derogatory comments and exposed to harsh or poor working conditions. Of particular interest was, exploring in greater detail whether there were any quantifiably significant work-related or societal differences between female EPZ workers in textile and garment factories (n = 1878), compared to those employed in ‘other’ factory types (n = 426). Measuring ‘objective’ work conditions, Mann-Whitney U tests demonstrated significantly lower earnings and savings among textile and garment workers compared to those who worked in ‘other’ factories. Pearson’s chi-square test of contingencies revealed that although reports of abuse and harassment were generally low, textile and garment workers were more likely to experience verbal abuse. Regardless of such negativity the majority of women within this study reported high rates of empowerment as a result of their experiences working in EPZs – largely irrespective of factory type. These statistical findings suggests that the different ‘objective’ workplace experiences (and to some extent, ‘subjective’ treatment) of female EPZ workers across textile and garment and ‘other’ factory types, may have little to do with their overall sense of personal achievement or inclusion, feeling empowered as a result of having been engaged in work.


IntroductIon
The movement towards economic liberalism that took place in the late 1970s in Sri Lanka has shifted work policies from import-substitution-oriented industrialisation to export-oriented industrialisation (Arunatilake, 2012).This has created more employment opportunities for younger women in the tourism and labour intensive industries, such as Export Processing Zones (EPZs), agriculture and general factory work (Arunatilake, 2012).The opportunity for rapid economic growth also resulted from an increased population of 'working age' individuals (currently 67%), which is expected to be maintained until 2017, generating a large workforce (World Bank, 2012).
Disadvantaged female populations have long experienced subordination throughout South-east Asia from internal and external socio-political and economic powers (Heyzer, 1986).EPZs are largely female dominated workplaces especially rife with gender-inequality and poor working conditions (Engman et al., 2007;Sivananthiran, 2007, Hancock et al., 2011b).Being less likely to take part in collective action or turn down repetitive jobs, women are exploited as an inexpensive and passive workforce (Sivananthiran, 2007).The majority of EPZs do not provide their female staff with adequate income or support; employees are forced to live in squalor, work long hours -without access to day care or aid during pregnancy -with many potentially dealing with sexual and (or) physical abuse (Engman et al., 2007).
The findings discussed in this paper come from a larger study that explored issues of empowerment, economic and social capital, and female EPZ workers' perceptions of employment (Hancock et al., 2011a).In this paper, the focus is on a smaller scope; that is, textile and garment workers and a statistical analysis of a crosssection of young women sampled from 6 EPZs across Sri Lanka (n = 2304).More specifically, comparisons

LItEraturE: GEndEr InEquaLIty, WorkInG condItIonS and PoLIcy
As a result of export-oriented policies, urbanisation can increase rapidly and is a phenomenon that has been occurring in Sri Lanka since the late 1970s (Arunatilake, 2012).The urbanisation process is a feature of economic development and thus, is also responsible for affecting shifts in the work environment.There has been a decline in the Sri Lankan agriculture sector, decreasing the share of the GDP from 46% in 1950 to just 13% in 2010; in contrast, the services sector has experienced an increase from 37% to 59% and industry sector has also achieved an increase from 20% to 30% during the same period (Arunatilake, 2012).According to Amunugama (as cited in Arunatilake, 2012,), the Sri Lankan business and technology sector has grown to be "the fifth largest contributor to Gross Domestic Product (GDP)".Urbanisation has also led to a sustained migration of jobseekers from rural to urban areas, a reduction in labour supply and the subsequent development of incentives to attract workers where most needed.However, despite their declining industry, workers trained in agriculture, albeit highly skilled, have traditionally found it difficult to transit to emerging 'modern' jobs because their specific skill-sets are not necessarily interchangeable (Cali & Menon, 2013).This decline from primary industries to manufacturing has resulted in its re-classification from an import substitution to a world-leading export-oriented economy (Athukorala & Rajapatirana, 2000;World Bank, 2002).
The presence of Sri Lankan Export Processing Zones (EPZs) has steadily increased since first established in 1977, with women specifically targeted for employment in such zones in order to sustain the emerging enterprise (Engman et al., 2007;Hancock & Edirisinghe, 2012;Hewamanne, 2011;Lynch, 2002;Lynch, 2007).Engman and Farole (2012) reported that EPZs normally offer an advantage to investors compared to the domestic market by creating fiscal incentives; "allowing for a combination of duty-free imports of raw materials, intermediate goods, and capital goods as well as streamlined and onsite customs services".
They further stated the goals of EPZs are to; "create new jobs, boost growth in exports and foreign exchange earnings, facilitate economic diversification and industrialisation, and provide access to foreign technology and management expertise" (Engman & Farole, 2012).
The pressure to remain competitive in emerging industries has led employers to 'cut costs' and improve productivity by restructuring workplaces and reducing 'fixed costs' in order to accommodate fluctuations in global demand.This has consequently made neweconomy jobs "uncertain, unstable, and insecure" (Arunatilake, 2012).
Although research into the poor working conditions of EPZs has been largely informal in nature, there are clear inequalities and unfair working conditions in EPZs.Ostensibly, it would appear that on a global scale, EPZs pay higher than their industry competitors (Sivananthiran, 2007).However, it has been found that these zones provide a lower wage than the national norm for machine workers (Bharathi, 2007).The International Labour Organisation (ILO) stipulated that the Sri Lankan public sphere provides more pay than private enterprise (Sivananthiran, 2007).It is speculated that EPZs provide salaries comparable to other Sri Lankan textile and garment work environments, but the exact figure remains unclear, with many employers keeping such data confidential (Sivananthiran, 2007).What is known, is that the income of EPZ employees is dependent on the level of output and financial efficacy of individual institutions.Consequently, EPZs are recognised as having greater outcome expectations and instituting longer working hours, when compared to the rest of the industry (Sivananthiran, 2007).Attanapola (2003) reported that just after the turn of the century, young Sri Lankan women comprised four-fifths of the textile and garment industry; in 2007 it was the second highest source of foreign exchange in Sri Lanka and the cause of an exponential increase in female labour force participation (Department of Census and Statistics, 2009;Institute of Policy Studies, 2007).Textile and garment factories often operate within EPZs and are arguably at the core (female EPZ workers in particular) of Sri Lanka's national development (Ferus-Comelo, 2006;Lakshman, 2004;Yatawara, 2004).
Usually, such (textile and garment) companies share a pyramidal structure favouring subcontracting and typified by poor worker conditions and (or) rights (Ferus-Comello, 2006).Outsourcing practices in foreign countries are effective in reducing costs by instituting precarious working conditions and exploiting local workers vis-a-vis long hours and low wages, thereby increasing the company's profitability whilst reducing employment and fair labour relations.'Loose' working regulations in these locations allow inequity to continue (Fernandez & Sotelo Valencia, 2013).
Women are impacted by negative social-political rhetoric, even when not the subject of discourse or policy decisions, per se, cementing their lack of power and espousing deleterious perceptions of women without actually addressing issues of inequality (Mani, 1989).Morris (2013) argued that an increased focus on women and reducing gender inequity at a global level was warranted, given that females represent half of the world population but only own 1% of land around the world and are prone to disenfranchisement, illiteracy or poverty and exposed to physical and sexual mistreatment.Moreover, females spend a disproportionate amount of time working compared to men (representing three quarters of all hours worked), whilst only accruing 10% of accumulated salaries world-wide (Morris, 2013).Other authors agreed female workers in particular, are subject to unreasonable workloads, time-frames and expectations (see Bharathi, 2007;Hancock et al., 2011b;Fernandez & Sotelo Valencia, 2013;Ferus-Comelo, 2006;Lynch, 2007).Already disadvantaged by their 'gender', they experience entrenchment as a result of exploitative employers, which is further compounded by their (traditionally) poor socio-economic standing both in Sri Lanka and globally (Arunatilake, 2012;Fernandez & Sotelo Valencia, 2013;Ferus-Comelo, 2006;Lynch, 2007).
EPZs are often over-represented by female employees, with staff exposed to gender-inequality and poor working conditions (Engman et al., 2007;Sivananthiran, 2007).Just prior to the turn of the century, over 6000 women took action against a multinational electronics organisation in Bangalore (Ferus-Comelo, 2006).The staff argued that their employer had intentionally truncated career mobility in an attempt to minimise expense, with young females (in particular) subjected to poor environmental conditions and exploited to work longer hours without remuneration (Ferus-Comelo, 2006).These negative conditions may be comparable to those experienced by employees in the textile and garment industry in Sri Lanka, however, such cases are likely to be underreported.Furthermore, Ferus-Comelo (2006) argues that female workers tend to be in insecure and non-traditional employment (part-time or contract work).Being less likely to take part in worker unions or avoid monotonous work, women are misused as a 'cheap' and submissive workforce (Sivananthiran, 2007).It has been reported that many EPZs offer female staff inadequate income and support; typified by poor living accommodations, (most female employers struggle to pay board and food expenses); they are also subject to lengthy workloadswith minimal access to child care or maternity assistance; and instances of sexual and physical abuse are common (Bharathi, 2007;Engman et al., 2007).
Women may also be subjected to poor ergonomically designed and even 'dangerous' working conditions (such as exposure to toxic chemicals), either unaware of the dangers and (or) not in a position to mitigate their exposure to risk with instances of poor safety and worker injuries having increased over the past decade (As-Saber, 2013;Bharathi, 2007;Ferus-Comelo, 2006).Women's situations are compounded if they develop health issues, often unable to seek medical support due to its expense or (temporarily) exit employment, where both options would disempower women by removing them from the labour force (Bharathi, 2007;Ferus-Comelo, 2006).Similarly, 'marriage' can hold negative connotations for employers, with female employees sometimes tested for pregnancy under the pretence of 'randomised' drug tests.Pregnant staff members face entry-level barriers or termination of employment, leading some women in developing nations to undergo unsafe abortions so as to retain (secure) employment (Rajalakshmi, 1999as cited in Ferus-Comelo, 2006).
Although decentralised work options should improve work-life balance, employers tend to exploit homebased employees by reducing pay, whilst simultaneously increasing workloads (Ferus-Comelo, 2006).Moreover, private residences are not always reviewed by employers, hence, some women may be involved in unsafe or illegal activities which may reinforce negative stereotypes about women or traditionally female-oriented 'work' cultures (Ferus-Comelo, 2006).Although home-based work minimises cost to employers, it also potentially increases risk and jeopardises employee-employer liability (Arunatilake, 2012).
Prior to the turn of the century, Abeywardene et al. (1994) found that males were predominantly placed in supervisory or technical roles in EPZs.Two decades later, Redclift & Sinclair (2013) argued that the perception that females are unsuited to more senior roles had not improved, with males still considered for 'skilled' roles over women, sometimes even in jobs involving automated factory processes.Men are also given positions of relative safety and prominence, offered greater training or career progression opportunities and earn salaries approximately 4000 Sri Lankan Rupees (LKR) higher than the minimum wage for the nation's clothing industry which is at 5100 LKR (Bharathi 2007;Perman et al., 2004).Ferus-Comelo (2006) suggested that in developing societies, a number of 'inherent' traits have been associated with women which supposedly justifies their continued employment in low-level, manual labour often associated with the textile and garment industry.These traits include greater manual dexterity, with women supposedly more precise and able to endure mundane, albeit intricate work.Moreover, the young age of most female EPZ employees also increases their submissiveness as they are less likely to engage in socio-political actions or make complaints compared to their 'older' counterparts (Ferus-Comelo, 2006).However, Whelan (2013) argued that women do not necessarily possess 'different' skills than males, with such definitions grounded in socially constructed or poorly evidenced myths.Despite this, the situation of women is often compounded by the reality (supposed) female 'talents' may not be acknowledged or deemed useful by organisations, with workplace polices of 'gender diversity' yielding dichotomous findings with regard to improving workplace efficacy (Whelan, 2013).
Women frequently receive pay cuts due to 'tardiness', taking too many lavatory breaks, missing work due to ill health or refusing to work overtime and not meeting productivity targets (Busser, 2005).It has also been argued that female employees earn less than their male counterparts due to the perception that their entry into the workforce is 'optional' and is reflected in the reality women are over-represented in non-permanent and lower skilled positions (Ferus-Comelo, 2006).In fact, female EPZ staff typically work for a couple of years before returning to their homes (villages) to set up their own small businesses, buy property or to start families, thus turning economic wealth into social capital (Theobald, 1996as cited in Ferus-Comelo, 2006;Hancock et al., 2011a).Although this perhaps further perpetuates the myth that women are not 'serious' about employment, it also suggests that many enter into formal employment with a specific goal in-mind, namely to improve their (family's) socio-economic standing and ensure future security.Similarly, 'subjective' beliefs and personal work-life histories may also influence whether 'objective', poor working conditions are considered as such by workers.Some women, may enjoy the autonomy that results from employment or may have been subjected to worse treatment in prior workplaces (Ferus-Comelo, 2006).
Developing nations like Sri Lanka are experiencing a continuous level of transformation at both the socioeconomic and political levels, with women exposed to an inordinate proportion of disparity at each level.It has been argued that textile and garment industries in particular, are subject to negative external pressure as a result of economic globalisation (Cole, 2005).The last decades of the 20 th Century saw the combined efforts of various human-rights organisations, human service agencies, religious groups and worker collectives (including the ILO) in attempting to formally document the circumstances of workers.Together, they created greater public and global awareness of several labour and social justice concerns in Sri Lanka, including discrimination against women, the use of child workers and the emergence of 'sweatshops' (Heward, 1997;Gordon & Turner, 2000).In order to combat this negative image, the Sri Lankan Government has begun working jointly with private enterprise to reform the nation's clothing and textile industry by prohibiting child labour, allowing workers greater freedom and by dismantling 'sweatshops' (Samaraweera, 2007).
Over several years, Sri Lanka has moved towards providing better working conditions and improving the rights of workers, proscribing to international labour laws issued by the ILO (Lakshman, 2004).However, many local economists view the shift towards greater government intervention and regulation as detrimental to the country's free-market ideals.Lakshman (2004) and Yatawara (2004) stipulated that greater collective bargaining power, combined with the advent of unfair dismissal policies, will diminish workplace productivity and affect the financial output of Sri Lankan industry.However, in reality, it would appear that such 'fears' are unwarranted.Despite increased opportunities for greater security under the government and international conventions, many Sri Lankans remain unprotected by unions and there appears to be no visible state intervention present in the nation's EPZs.Furthermore, these 'supposed' new policies aimed at improving the circumstances of workers, particularly young women, are not reflected in the Sri Lankan social hierarchy.The literature suggests that traditional biases still permeate many legal, industrial and educational institutions found throughout South Asia, often favouring the position of males and continuing to limit the rights of females (Gomez & Gomez, 2004;Goonesekere, 2004;Jilani & Ahmed, 2004).
Although the circumstances of staff differ between EPZs, the most disadvantaged workers continue to be women and children and the most apparent cases being gender-based workplace discrimination.Arguably, the 'worst' of these zones are the 'oldest' (Moran, 2002).Generally, smaller institutions are not owned by private juggernauts or protected by the Organisation for Economic Collaboration and Development (OECD).Therefore, employees are more susceptible to unfair treatment and often exposed to unhygienic working environments (Perman et al., 2004).For instance, like women, child employees are sanctioned for tardiness and low productivity, paid even less and are often required to work continuously without respite.In extreme cases, women suffer from dehydration, unwilling to drink water lest they be reprimanded for taking an excessive number of lavatory breaks (Bharathi, 2007).
Sri Lankan Industry has criticised ILO guidelines that would prevent women from working night shifts.The promise that EPZs can deliver continuous work (day or night) ensures foreign interest in the zones and enacting such a policy would be unprofitable (Gordon & Turner, 2000).Furthermore, female employees that work past midnight are often left without transport, unless it is paid for out of their own salary (Bharathi, 2007).In Sri Lanka, employers can ask staff to work up to 60 hours over time a month, although they are not obliged to provide reward or compensation for these workers (Busser, 2005).With financial levies applied to employees who do not comply with requests to work additional hours, or the loss of their job, most do not refuse (Busser, 2005).
Arguably, since the standardisation of workforce guidelines and evaluation tools across Sri Lanka, there has been a decline in discriminatory EPZ practices against staff (Sivananthiran, 2007).However, this is not universally apparent (Gordon & Turner, 2000).Existing policies that promote diversity may potentially empower women and lead to cultural change over the long term, which can simultaneously perpetuate myths surrounding women.For instance, policies based on affirmative action may lead to tokenism, where women are placed indiscriminately into positions of power to meet 'targets', despite lacking the skills to contribute or cope adequately (Adams, 2013).Interestingly, women tend to adopt so-called 'male traits' when in positions of power, and thus are more readily accepted in patriarchal societal systems.As such, any presumptions that female 'talent' may weaken financial output in senior roles are potentially meaningless.Furthermore, despite only requiring a small cohort of EPZ workers to create an official Union, successful collective action amongst staff remains minimal (Busser, 2005;Hancock et al., 2011b).Many employers continue to use punitive action such as relegating workers to lower pay and job status in order to discourage the establishment of Unionised worker groups (Busser, 2005).As an alternative to Unions, EPZ workers have begun forming Employee Councils.However, unlike traditional worker collectives, they do not operate as separate bodies.Councils function under the auspices of EPZ administration and members are limited in the kinds of collective action they can take (Busser, 2005).
There are some EPZs that place greater value on their workforce and maintain a standard of work environment that is higher than the industry norm (Moran, 2002).These employers offer greater financial security, more flexible working arrangements, access to leave benefits and subsidised living or transport expenses (including child care).They focus on the health and well-being of staff with better protection on-site and access to outside medical aid (Moran, 2002).With an emphasis on personal development, such employers who have a good reputation of supporting workers are 'in demand' by workers and also provide avenues for further education and training (Moran, 2002).These employers recognise, that worker productivity is positively related to higher job satisfaction and a good working environment (Perman et al., 2004).Despite a small number of 'good' employers, the literature indicates that most EPZ administrators do not comply with the international conventions adopted by the Sri Lankan Government (Gordon & Turner, 2000).Many EPZs continue to discourage collective action, do increased (decreased) feelings of inclusion or decisionmaking in and outside of their home environment.
The focus of this paper was narrowed down to explore more specifically the experiences of textile and garment workers.Considering the negative associations of such factory types in the literature and the fact that they were the most numerous group of respondents sampled in the original surveys, it was deemed prudent to further investigate the responses of these female EPZ workers.New research questions developed for this study were as follows: 1. What were the experiences of 'textile and garment workers' compared to all 'other' Export Processing Zone (EPZ) workers sampled in the original study?
2. What were the major factors that emerged from the comparison of these two EPZ worker cohorts?
Approval for conducting the research was granted by Edith Cowan University's (ECU's) Ethics Committee prior to the dissemination of surveys in Sri Lanka.Several trained female Sri Lankan research assistants (RAs) consumed local data collection.RAs were conversant in both English and Sinhalese, also having received intensive support from the Principal Investigator representing ECU, as well as the Research Director and Project Manager from Sri Lanka's Centre for Women's Research (CENWOR), regarding proper recruitment and data collection processes.Respondents were randomly selected by RAs and recruited from the many boarding houses in close proximity to Sri Lanka's EPZs, as well as through consultation with the Sri Lankan Board of Investment (BOI) and non-government organisations such as CENWOR.Access to such resources ensured that the relevant sample populations were targeted and that these female participants felt secure.Moreover, such collaboration (coupled with community forums) also meant that the research findings would directly impact individuals and be disseminated to the appropriate policymakers (stakeholders), thereby building community capacity and positively influencing the perceived place of women in Sri Lankan society.
Eligibility criteria only required that these female EPZ workers had been employed within an EPZ for at least 12 months prior to the date formal data collection.All other demographic information pertaining to their age, position and work experience was collated as part of formal data collection.In line with ethical conduct, prospective respondents were approached in safe locations (predominantly in public areas) and care was taken not to not recognise the legal rights of workers and offer poor, often unhygienic and punitive, working environments for staff.This serves to truncate worker creativity and reduce overall output.As a result, the ILO argued that such business practices prevent growth and global expansion opportunities, negating the benefits that come from producing a highly skilled and satisfied workforce (Engman et al., 2007).
Ultimately, greater female representations in the labour force does not necessarily equate with gender equality.Although, Morris (2013) stipulates that many males are not consciously aware that their action may be disempowering others, Elson (1999) and Piper (2003) argue that patriarchal systems promote the appointment of women only in order exploit them (sometimes via negative measures of control and harassment), secure in the knowledge that women are disadvantaged and potentially powerless to assert control.Regardless of the socio-political disadvantages experienced by women, discussed above in terms of their disenfranchisement compared to men coupled with poor 'objective' working conditions, such as inadequate pay (particularly when compared to men) and working lengthy hours, or 'subjective' experiences of ill-treatment or exclusion evident in EPZs and factory work, the focus of this paper is only extended to female EPZ workers and not male workers.It was deemed important to limit the scope of this study to understand how Sri Lankan women perceived their place in a patriarchal society (workforce), particularly their exposure to positive or negative, 'objective' and 'subjective' working (societal) environments since they have been long considered essential to Sri Lanka's shift to an export-oriented economy (Hancock, 2006).

MEthodoLoGy: EthIcaL conSIdEratIonS, SaMPLInG and data coLLEctIon
The original study aimed to explore the extent to which Sri Lanka's female EPZ workers were empowered or disempowered as a result of their employment (for the full report, see Hancock et al., 2011a).According to Premaratne et al. (2012); "empowerment has been perceived as a shift from powerlessness to strategic social, economic and political participation".
As such, another aim of this research was to establish whether these women had successfully converted economic wealth (accumulated income and savings from formal employment) into social capital, and whether work has led to further socio-political engagement and disrupt their daily work or family responsibilities.The female EPZ workers were informed that the findings would be kept anonymous and confidential; and any published material would not reveal individuals' work location or home village.
The survey sample included 2304 women from 6 EPZs, selected in order to ensure findings represented the 11 EPZs (and 65,598 women working therein) located across Sri Lanka's rural and urban locales (Table 1).These included Katunayake, Biyagama, Koggala, Wathupitiwela, Pallakele and Seethawake; the latter originally an Industrial Park (IP) given less financial support than EPZs.All IPs have subsequently been  reclassified as EPZs and their number has since increased (Figure 1).A further of eight focus groups with 72 women were also conducted, coupled with a review of 40% of open-ended survey responses which formed part of a qualitative analysis.However, a review of this data, which would have required identifying specific quotes or open-ended responses, is beyond the scope of this paper.
The female EPZ population has been divided into two subsamples investigated in this paper: female EPZ employees from textile and garment factories (n = 1878) and respondents who worked in 'other' factories (n = 426).Although originally considered separately, cohorts representing the textile and garment industries working in formal manufacturing, their salary had allowed them to accumulate capital and empower themselves (or their family).Although not the main focus of the survey instrument, the other aspects of social functioning listed above were also indicative of empowerment and thus allowed us to make further inferences regarding this underlying concept.Of importance was to establish the level of congruency between the feelings of inclusion, ill-treatment and empowerment evident in the survey responses and the largely negative image conveyed throughout the literature.Thus, demographic and work profiles, as well as personal experiences and perceived status were investigated in greater depth vis-à-vis new quantitative analysis, using SPSS Statistics 20.0.

rESuLtS
A total of 2304 women responded to the survey questionnaires.EPZ workers employed in textile and garment factories (n = 1878) were grouped together and compared to respondents who worked for 'other' types of EPZ factories (n = 426).Although, the levels of education between the two groups were similar (Table 3), a further Pearson's chi-square test of contingencies indicated that there was an association between education level and the type of factory.Represented statistically as χ 2(1, N = 2302) = 6.99, p = 0.008, it would appear that the textile and garment workers sampled were more likely to report lower levels of education (such as O-levels and incomplete O-levels) than employees from 'other' factories, who appeared to report higher educational achievement (in terms of A-levels and tertiary level qualifications).
Demographic details of the subsamples are displayed in Table 3.The average age of this cross-section of women was approximately 24 years with the majority aged between 20 to 24 years.Most reported that they had never been married, with around one-fifth of women from both groups married at the time of the survey.Almost two-thirds of the subsample in the textile and garment industry (n = 61.4%)and over half of those representing the 'other' factory types (n = 55.2%), had completed their G.C.E O-levels (the second highest level of secondary education in Sri Lanka); with a further 36.2% of textile and garment workers and 41.1% of 'other' respondents having completed their G.C.E A-levels (the highest level of secondary education in Sri Lanka).Discouragingly, only a minority from each subsample had obtained tertiary level qualifications, with less than 2% of textile and garment workers and less than 1% of 'other' factory workers, never completing their G.C.E.O-levels.The implications of these and other findings will be discussed in the following sections; however, it is clear from this included some of the highest proportions of EPZ workers sampled (Table 2).This, coupled with the focus on the exploitation of textile and garment workers and general scrutiny placed on EPZs by stakeholders and academics in the literature, led to the combination of the women into one subsample.The remaining factory types were then amalgamated into a new 'other' category which included female EPZ workers in shoes, machinery, computers and electronics, jewellery, cigars (cigarettes), haberdashery, fishing, toys, food, and miscellaneous industries.
These two sub-samples formed the basis for comparison in this paper.It was explored whether any differences were apparent between the two sets of women in terms of their experiences within EPZs and Sri Lankan society.Data collection instruments were designed to provide insights into the way (young) women may be empowered or indeed disempowered as a result of their employment.In line with the poor work environments and treatment faced by EPZ workers described in the literature above, several 'objective' work conditions were measured.Variables included, hours (years) worked or spent in overtime, income, opportunities for promotion, the amount of earnings saved (and whether this was through informal or formal financial institutions) and if any remittance was paid to family per month.Survey items also explored 'subjective' work (societal) experiences, such as feelings of inclusion, decision making power and whether women felt empowered.Cases of verbal, physical or sexual harassment at work and experiences of public humiliation (for being an EPZ worker) were also documented.
Empowerment 1 was measured using a single selfreported item that asked women whether, as a result of substantive analysis that the female EPZ workers were generally young but moderately well educated.
Further comparisons were made across a range of 'objective' work conditions.Analysis involved the use of Pearson's chi-square test of contingencies for comparing nominal data, while the Mann-Whitney U test was used for comparing ordinal, interval and ratio data which was also implemtned in order to compensate for issues of non-normality and size disparity between subsamples.Table 4 displays variables collated, such as averge working hours, opportunites for promotion and financial information, highlighting any similarities or differences between the two groups.
As presented in Table 4, women across both groups worked a similar number of years.However, women employed in textile and garment factories worked longer total weekly hours than women who worked in 'other' factories.Textile and garment workers were also more likely to work overtime than those who worked in 'other' factory types.The odds ratio demonstrate that textile and garment workers were 2.23 times more likely to have worked overtime than those women who worked in other factories.However, this has not translated to earnings, since, there was no difference in overtime monthly salaries.
Textile and garment workers also reported a significantly lower total monthly salary, which included overtime salary in comparison to the total salary earned by women who worked in 'other factories'.Encouraginly, the amount saved as part of bank plans did not differ between the two groups of women.However, the amount saved through informal savings schemes was significantly higher among women who did not work in textile and garment factories.Total savings were also significantly lower among textile and garment workers.Finally, respondents from the textile and garment industry reported that on average a significantly greater proportion of their earnings went to their family income when compared to their counterparts in 'other' factories.
These results indicated that respondents working in texile and garment EPZs were often disadvantaged across several work-related variables, compared to those employed in 'other' factory types.Although, highlighting such disparity was important, focusing purely on 'objective' work conditions did not reveal information about their treatment, experiences of inclusion or autonomy (identifed as problematic in the literature).Therefore, it was also necessary to compare more 'subjective' variables, such as, cases of abuse, harassment and humiliation, feelings of

June 2015
Sri Lanka Journal of Social Sciences 38 (1) empowerment and levels of community involvement, decision making or social inclusion reported between the two sub samples.Another Pearson's chi-square test of contingencies was used to evaluate whether factory type was related to respondents' experiences of illtreatment in the workplace and society (Table 5).
As presented in Table 5, the majority of women in each subsample did not experience abuse, harassment or humiliation.However, verbal abuse was the most common negative experience reported, having represented almost 30% of textile and garment workers and nearly 12% of 'other' workers.Further analyses indicated that there was a clear association between the type of factory worked in and reports of verbal abuse.The odds ratio indicated that the chance of experiencing verbal abuse were 3.13 times greater for textile and garment workers than for women who worked in 'other' factories.Experiences of public harassment were almost equal, representing 19.2% of textile and garment workers and 19.7% of employees from 'other' factories.Interestingly, there were no other major statistical associations found between experiences of abuse, harassment and humiliation and the type of EPZ factory worked in.
Encouragingly, empowerment of women was high across both subsamples (representing 95.3% of each group).A minority of textile and garment workers (22.3%) and 'other' workers (19.2%) participated more in community and political activities, whereas increased participation in decision making within the home was reported by 54.4% of textile and garment workers and by 61.7% of 'other' respondents.This difference was statistically significant, and the odds ratio indicated that respondents who worked in factories 'other' than textile and garment EPZs were 1.35 times more likely to report greater decision making power in the family home.
However, the opposite was found regarding participation in decision making outside the home.There was a significant association found between factory type and increased decision making outside of the home, with the odds ratio revealing that the probability of textile and garment workers participating in decision making outside of the home was nearly three times (2.97) as more likely to be than those employed in 'other' factories.Experiences of increased social inclusion were almost evenly divided between each group and figures demonstrated that over half of women surveyed felt greater inclusion within their community.These more 'subjective' findings, linked to treatment and inclusion, perhaps convey a more positive image for textile and garment workers than the 'objective' working conditions described above relating to hours worked and wage (savings particularly for personal use, rather than their family's).

dIScuSSIon
This study sought to investigate whether any differences were quantifiably apparent between  (Attanapola, 2003;Department of Census and Statistics, 2009;Hancock, 2006;Hancock et al., 2011a;Institute of Policy Studies, 2007).There was little difference in the demographics between the two groups of women in terms of age and marital status.However, textile and garment workers were more likely to have a lower education compared to respondents representing 'other' factories.
A number of 'objective' work-related differences were found between the two subsamples.Those from textile and garment EPZs worked longer weekly hours and they were also more likely to report working overtime.This however did not necessarily translate to greater earnings.Textile and garment workers' monthly incomes were significantly lower than their counterparts in 'other' factory types (a finding perhaps expected, Ferus-Comelo, 2006).Savings were also significantly lower among textile and garment workers.It is surmised that this may be a result of their lower income, as well as a greater proportion of salary being contributed to family income through remittance.Given the disproportionate number of negative environmental and financial outcomes of having worked in the textile and garment industry, it could also be inferred that perhaps women with lower educational achievements had less opportunities to work in 'other' factory types with (arguably) better working conditions.
Although there were a number of differences reported between women in terms of their 'subjective' experiences, the 'treatment' of textile and garment workers by others and their feelings of inclusion (when compared to 'other' factory types) were not as discouraging.Experiences of verbal abuse, physical abuse, sexual harassment and public humiliation were generally low across both groups.However, findings demonstrated that whilst the majority of workers did not experience any form of mistreatment, there was a greater (repair) clothing might be considered more relevant than computer literacy at the village level.Given that, a proportionately 'small' number of respondents felt more autonomous outside the home or engaged in political and community activities, somewhat confusing (albeit encouraging) was the finding that regardless of their place of work, almost the same percentage of each sub sample (over half) reported feeling socially included.
Furthermore, a near consensus reported feeling 'selfempowered' as a result of being employed; this was despite clear differences in 'objective 'work conditions between textile and garment workers and respondents from 'other' factories.Moreover, it could also be inferred that given the proportionately small amount of verbal, physical or sexual abuse and public humiliation experienced by both cohorts, that such positive'treatment' may have a larger bearing on feelings of empowerment and inclusion (a sentiment reiterated in Hancock et al., 2011a).A possible explanation may be that despite the disparity in 'objective' work conditions, the fact each sample was (largely) treated positively, these 'subjective' experiences shaped respondents' interpretation of their situation -for example, if managers (co-workers) treated women with respect, then being asked to work overtime may not have been perceived as exploitative.Also potentially significant was that textile and garment workers showed higher remittance rates than their contemporaries in 'other' factory types.Given the cultural importance of providing for 'family' (also see Hancock et al., 2011a), this may indicate why women belonging to textile and garment factories had higher feelings of empowerment, but not why they had less 'familial' decision-making authority.
The literature suggested that poor working environments were typical across factory work and EPZs (Ferus-Comelo, 2006;As-Saber, 2013;Bharathi, 2007).However, findings indicated that although the length of hours textile and garment respondents worked (particularly in overtime) was higher (a trend supported by Sivananthiran, 2007); the total income earned was lower and amount saved was less than 'other' factory workers.Such environmental and financial issues may be of less significance than the actual 'act of working' in respect to increasing women's self-efficacy and perceived inclusion.Women sampled were relatively homogeneous regarding educational and socio-economic background and also in terms of length of employment, suggesting that employment in EPZs is one of only a few limited employment opportunities available to most women in the nation.Paradoxes between the 'objective' and 'subjective' variables indicated that there may have been other (unrecorded) personal beliefs or events that likelihood of being subjected to negativity if working in textile and garment factories.This result is consistent with previous research that has found that textile and garment workers experience poor social status and are often degraded by members of their community because of their work (Lynch, 2007;Hancock, 2006;Hancock et al., 2011b).
It is important to note that over half of each sub sample reported increased decision making in their private residence, thereby indicating a substantial number of respondents experienced positive, familial-level change as a result of working (a trend reflected in prior research, Hancock, 2006;Hancock et al., 2011a).Despite this, respondents representing textile and garment EPZs were less likely to experience greater autonomy and control inside the home compared to employees from 'other' factories, perhaps reflecting negative societal attitudes towards the textile and garment industry and EPZs in general (Hewamanne, 2003).The relationship between hours worked and time spent with family is inversely related; where employees working fewer hours will have greater work-life balance, engaging with family and meeting cultural expectations for women and thus affording them greater respect at the family level (Hanock et al., 2011a).
Overall, a smaller proportion of each sample believed that they held more power and influence in their community since having engaged in formal employment; potentially reflecting the reality depicted in the literature that EPZ workers are perceived negatively in Sri Lankan society.This is in line with findings that a small proportion of respondents from each sub sample were engaged socio-politically.Statistically, this represented slightly less than one-quarter of textile and garment workers and approximately one-fifth of respondents from the 'other' subsample.This lack of community and political engagement compared to within the home, were also trends in prior research (Hancock, 2006;Hancock et al., 2011a).However, it should be noted that textile and garment workers were almost twice as likely to report greater decision making power outside of the home, compared to 'other' respondents.
From these statistical findings, it is not clear why there was disparity in terms of decision making in and out of the home, or the possible relationship with factory type.Greater control may have been afforded to textile and garment employees in society, with community leadership linked to their status in organisations (slightly more likely to have been promoted) and because their skills may have been more transferable outside of work contexts, for instance, being able to manufacture may have shaped overall experiences.For example, it could be assumed that formal employment fosters greater self-confidence, increased social power as well as greater respect from others inside or outside of the family home, enabling greater decision making.Conversely, it is possible that some respondents always felt as though they had sufficient decision making power inside or outside of the home, therefore, their situation may have not changed since engaging in formal employment.This paper is limited given that greater insight into these differences cannot be gained from the purely quantitative data analysed here.It is therefore proposed further quantitative and qualitative investigations be conducted into the working conditions and treatment experienced by textile and garment workers and employees representing 'other' types of factories.What is evident is that despite exposure to poorer 'objective' work conditions and some 'subjective' negative experiences than 'other' factory employees, textile and garment worker responses indicated comparable levels of 'empowerment' and 'social inclusion'.This indicated that the absence of negative 'subjective' experiences in the workplace was potentially more important than the presence of 'objective' factors, such as lower pay or hours spent at work.The majority of both cohorts had not experienced any abuse and a near equal minority (less than one-fifth) of each sub sample had been exposed to public humiliation, an encouraging finding which was at odds with the literature regarding the perceived treatment of EPZ workers (Engman et al., 2007).

concLuSIon
This study explored various gender dimensions and the extent work in formal manufacturing industries was experienced by a cross-section of young, female EPZ workers.An aim of the research was to establish whether these women had turned economic wealth into social capital; measuring their levels of socio-political engagement and feelings of inclusion or autonomy within their families and communities.Also of salience were cases of harassment and humiliation experienced as a result of their employment.Given the high proportion of textile and garment industry employees and the negative working conditions (treatment) of such workers (and EPZ staff in general) depicted in the literature, it was deemed prudent to investigate the impact factory type had on women's perceived experiences of employment and resultant self-efficacy.
This paper investigated potential differences in 'objective' work-related conditions and 'subjective' social-related experiences among women (n = 2304) who worked in Sri Lankan textile and garment factories (n = 1878) and women employed in 'other' factory types (n = 426).This paper highlighted several similarities and differences between the subsamples.Results demonstrated that textile and garment workers earned and saved significantly less money, even though they worked significantly longer weekly hours than their counterparts in other 'factories' and were more likely to have been promoted.Textile and garment workers were also more likely to experience verbal abuse and less likely to be engaged in-home decision making.However a comparable proportion of each subsample experienced relatively low public humiliation as a result of their employment in EPZs, suggesting factory type may have little bearing on public perceptions of workers (positive or negative).
Reported feelings of empowerment were high and feelings of social inclusion were equally rated among all women surveyed regardless of factory type.Overall, the findings suggest that the act of working was perhaps more important than 'objective' work conditions or the kind of factory employed in.Moreover, the 'subjective' experiences of women are potentially more influential than extrinsic variables such as levels of income, overtime and savings when compared to other female workers.The paper provided a snapshot of women employed in EPZs across Sri Lanka and given the large sample size could be viewed as generalisable to other settings.Despite the limitations of quantitative studies, which meant that the intrinsic links between variables could not be fully contextualised, this paper is presented for enhancing understanding about working conditions, gender issues and labour in Sri Lanka's rapidly evolving, export-oriented economy.

1.
No specific instrument was used to identify empowerment or disempowerment.Only one survey question required women to self-report whether they felt empowered or not, by answering 'Yes' or 'No' and this has been clarified in-text.Furthermore, there was no question, scale or instrument querying experiences of disempowerment per se, however inferences of (dis) empowerment have been made.No Cronbach's can be given regarding empowerment because it was a single question.Cronbach's is only used when there are a number of items (questions) assessing particular constructs.For example, Rosenberg's self-esteem scale has several items and Cronbach's can be used to assess how reliable those items are in assessing the overall construct of self-esteem.

Figure 1 -
Figure 1-The Board of Investment in Sri Lanka's interactive map showing the location of Export Processing Zones in Sri Lanka in 2014* Source: BOI (2014) *Reproduced with the Board of Investment (BOI's) permission in February 2014.The status of Industrial Parks (IPs) have been redefined as EPZs, although IPs are referred to on the BOI website) Investment (BOI, 2010).

Table 1 :
Number of participants by EPZ compared to all women in EPZs (BOI, 2010)7 -2009Questionnaires & the Sri Lankan Board of Investment(BOI, 2010)

Table 2 :
Frequency and percentage of respondents working in textile and garment EPZs and 'other' types of factories.
Source: Adapted from 2007 -2009 questionnaires (*In the original study, the miscellaneous category was classified as 'other' -to avoid confusion, it has been renamed for use in this paper)

Table 3 :
Demographic details of textile and garment workers (n = 1878) and workers in other factories(n = 426) Sri Lanka Journal of Social Sciences 38 (1)June 2015

Table 4 :
Weekly hours, monthly salary and savings amount between women who worked in textile and garment factories and women who worked in other factories

Table 5 :
Social experiences between workers of textile and garment factories and workers of 'other' factories