2.1. Global brands and homeworker policies

As noted in Hickman (Reference Hickman2021), the Indian apparel manufacturing industry was worth $90 billion in 2021, $16.8 billion of which was exported to serve Western brands. These brands, through their supply chain management policies, have a very strong influence on the treatment of homeworkers. As described in Carpenter (Reference Carpenter2021), many brands currently have policies that there should be no homeworking present in their supply chains. In practice, given the very strong motivations for engaging homeworkers in the Indian textile industry, such policies have simply meant that no one admits to its occurrence rather than stopping such homeworking from happening. This situation clearly enhances, and reinforces, the dangers of the homeworkers being exploited. In fact, the presence of extensive, unacknowledged homeworking in a company’s supply chain represents a very clear risk of human rights abuses. Since removing the homeworking entirely is not feasible, nor socially desirable, the logical next step for the policymakers is ensure that global brands acknowledge the presence of homeworkers in their supply chains.

Ideally, as Traidcraft and Homeworkers Worldwide are campaigning for, companies should be publishing policies acknowledging the existence of homeworking and promising to manage it constructively. However, as we found when we surveyed those reports by global brands, only very few Modern Slavery reports even mention homeworking. Carpenter (Reference Carpenter2021) identified 113 textile retailers in the Noida-Gurgaon area from the open apparel registry, and 33 Western suppliers/brands were found to be connected to them through the supply chain but only 30 have records that can be traced in the public domain. Of the 30 companies that we were able to trace, 10 were identified to have openly adopted homeworkers’ policies, 6 made statements that would have effectively banned homeworking in their supply chains, while 14 were ambiguous on the matter. From this sample of 30 firms, it means that any factory acknowledging the existence of homeworkers in its labor force could risk surrendering between 1/5th (i.e., 6/30) and up to 2/3rds (i.e., 20/30) of its potential export market. We do not claim the evidence here extrapolates to the 8000 UK businesses whose reports we downloaded from the UK Modern Slavery registry. Nevertheless, this serves to illustrate the serious and real threat to our effort in Kapas Hera and Noida in bringing the homeworkers to the surface. With eShram’s large-scale registration of unorganized workers, the tracing of homeworkers in Indian textile supply chains becomes more feasible than before. This is one area that requires joint effort from modern slavery act legislators, trade enforcement laws, and ESG investors to also recognize and acknowledge the presence of homeworkers in the supply chains. This will encourage the brands and suppliers to also present their homeworkers policy openly. Political maneuver by the various interest groups, when properly orchestrated, might lead to a considerable effect on improving the livelihood of millions of homeworkers.

2.2. Digital readiness within the Indian textile supply chain

Before moving on to consider the consequences of digitizing the Indian Textile supply chain, we must study the degree of digital readiness of the actors within it. The factories themselves are, partially due to their need to work with Global brands, in general very capable of participating in any form of digital supply chain. While the subcontractors are much smaller scale operations, often single person, they are generally educated well enough and economically well off enough to at least have smartphones and could again participate in a digital supply chain. Our knowledge about the subcontractor is acquired through our contacts in India, namely staff members from Traidcraft India who were commissioned to conduct interviews locally. Apart from their local knowledge, we have included a subcontractor in the list of interviewees.Footnote ³

The homeworkers themselves, however, are currently much less well-equipped to participate. They are often poorly educated, and the vast majority simply do not possess any means of accessing the internet. Many women interviewed have (access to) mobile phones, and some have smartphones. The ownership of smartphone by these women has increased as the children’s education has moved online during the COVID pandemic. With their Aadhaar cards, the homeworkers understand how fingerprint and iris scan are used in identification. The women are familiar with IVR (Interactive Voice Response) technology as they use it to order LPG gas cylinders. Most women with a smartphone did not know how to access the internet or use internet-based applications such as WhatsApp or Facebook. It was mostly younger women, ranging from 24 to 30 years of age who expressed a certain degree of familiarity and comfort in using smartphone technology. However, they are all very eager to learn new technologies and new skills especially with the Indian government’s digital policy; all government services in Tamil Nadu are now available through eSevai portals.

While the homeworkers indicated that they prefer to be paid in cash, there is a very real chance of digital payment solutions being adopted in the textile supply chain. One supplier, we consulted, was a strong advocate for this. The combination of Aadhaar and eShram is likely to solidify the role of digital payments for this group of marginalized digital users. The launch of eShram has up the level of digital readiness. Significant potential transparency and traceability advantages are expected through digitization. But as discussed in Carpenter (Reference Carpenter2021), if the homeworkers lack a strong representation in the digital supply chain, their voice will be excluded, and they will be exposed to exploitation. This then leads on to another critical role for the Self-help Groups (SHGs)—they provide an effective route to support homeworkers with meaningful, active, digital representation within any digital solutions that develop.

2.3. Digital identity and the Indian government

There is one further factor potentially pushing digitization further within India. That is the Indian government, who are keen to move toward the widest possible use of digital identities and to promote digital government. Interestingly, due to India’s centralized digital identity system Aadhar, nearly all the homeworkers have a biometrically backed digital identity as an Indian citizen or a resident in India. Until recently, what they lack is an identity as a worker, with a considerable majority of them not claiming the in-work state benefits to which they are entitled. The situation might change soon when the eShram registration is completed and when the government promises are fulfilled.

The earlier digital movement has taken the form of nationwide digital frameworks. Digital India, for example, is a campaign to ensure Government’s services are made available to citizens and residents in India electronically by improved online infrastructure with better internet connectivity. The Prime Minister Narendra Modi launched on August 15, 2014 the “Jan Dhan” scheme, which is a universal banking, financial inclusion program aiming to bring banking services to large rural areas. Finally, the Aadhaar identity system means that many Indian citizens and residents possess a digital identity. Since these identities are centrally controlled, and biometrically backed, there are several legitimate concerns about their potential use, as described in Masiero (Reference Masiero2017) and Krishna (Reference Krishna2021).

In addition to the above, there were several cases where informal supply chains have been moved to work using digital systems. The most prominent of these is the Indian ration card system. This system is responsible for distributing, on a state-by-state basis, a critical state benefit in the form of physical food. Many Indian states have attempted to digitize this supply chain, partly to reduce the amount of waste due to corruption. As described in Masiero and Prakash (Reference Masiero and Prakash2015) and Masiero and Das (Reference Masiero and Das2019), these changes had several effects, some of which are far from being seen as positive by people receiving the rations. Another example is a taxi supply chain introduced in Delhi. Again here, the taxi drivers did not benefit directly from the digitization (see Krishna, Reference Krishna2021).

One further, very relevant, example is the Indian governments recent launch of the eShram portal.Footnote ⁴ Centre for Internet and Society (2021) described the practical obstacles to register on the eShram portal. While well-meaning, and logical, the eShram scheme could potentially have a huge flaw for homeworkers in the textile industry. Given the current policies adopted by Western brands in relation to homeworking, any person formally known to be a homeworker might immediately be disqualified from further employment. This extremely unfortunate impasse further shows how delicately any digital interventions must be considered, and underscores the importance of reform needed for the policies adopted by global brands. In conclusion, there is a very strong, definite momentum toward the adoption of digital technology in India, which seems likely to drive the adoption of digital supply chains within the country.

2.4. The consequences of digitization for the Indian textile homeworkers

Given all the previously mentioned momentum driving digitization and the adoption of digital supply chains within India, we must now consider the likely effects of digitizing the Indian textile supply chain. There are two distinct cases to consider here.

The first case would involve some Global brands continuing to refuse to acknowledge the existence of homeworkers in the supply chains. In this case, the digital supply chain would have to omit the homeworkers entirely from consideration. In one sense this would simply formalize the lack of visibility of the homeworkers but doing so would make their exploitation both easier to enact and ultimately harder to address. The second case is where some global brands admit to the existence of homeworkers. As seen in Carpenter (Reference Carpenter2021), perhaps around a quarter of Western companies have this status, with ongoing campaigns in the area hoping to improve the situation, so it is easy to imagine such a system being set up by some interested brands. In this case, it would be possible to set up a digital supply chain containing the brands, factories, and subcontractors as active participants. The homeworkers would, however, due to their digital position, be forced to remain as purely passive participants, unless represented by digital custodians. Whether represented by digital custodians or not, it is important to give the homeworkers a say, a representation, at the design stage of such a digital system.

In theory, a digitized system would provide solid information regarding how many homeworkers have worked on a specific batch of clothes, how much work each did, and how much payment each received. Unfortunately, the only source of this information would be the subcontractors themselves. Given the substantial structural reasons motivating the subcontractors to exploit the homeworkers, they clearly represent a highly unreliable source for this information. To make such a digital supply chain operate well, we must therefore find some second source that can audit the accuracy of the information provided by the subcontractors. This would not need to verify every transaction, merely a sufficient percentage to motivate them to report honestly. The current supply chain contains no entity well-placed to perform this task. Given the strong likelihood of digital supply chains arriving in the Indian textile industry, and the need to have such an entity to make the digital supply chain work well, there is a very strong motivation to find a way to ensure data integrity. We suggested in Sections 2.6 and 3, how governments in- and outside India, might help to promote the creation of a trustworthy distributed data depository.

2.5. Digital (identity) custodian

(Identity) custodian is a well-defined concept and construct in the SSI (Self-Sovereign Identity) literature. It has an important role in a decentralized system and especially a digital one. Acting like a trustee, custodian acts in the best interests of its wards (i.e., homeworkers in this case). Here, we expect custodians (e.g., self-help groups, the local unions, a sympathetic factory owner) to help the homeworkers bypass the digital barrier, and to stand up for their interests in the consortium designing and managing the data collection and data sharing arrangement. Indeed, we visualize this in a wider context such as the modern slavery intelligence network, whereby the victims will be represented by NGOs and CSOs as custodians.

Custodians, whether digital or not, are entities (persons or organizations) who are entrusted to act on behalf of their wards. In theory, custodian could also be an algorithm or an AI (Artificial Intelligence), provided that the wards’ interest is protected and served. Here, the key is not because the worker is skilled or unskilled, but if the worker has the capacity to claim and protect its own rights via digital or nondigital means. Indian homeworkers are typically women, rely on their husbands to access mobile phones, most likely digitally illiterate and, due to the gender and family position, have little say on personal matters. We envisage this is one group of the population who will benefit from having custodians protecting their interests.

Finally, the custodian relationship need not be a permanent one, and its function is by no means universal. One may have a custodian for identity verification, and another custodian for processing grievance. The power and duration of the custodian relationship should be well-defined. The ward should be able to terminate such a custodian relationship or change custodian at any time. These should be spelt out clearly in the governance structure of the custodian relationship.

2.6. A decentralized bottom-up data depository system

To audit the data supplied regarding homeworkers, we need a trustworthy mechanism to obtain information from homeworkers regarding how much work they have completed, their pay, and any major complaints that might have arisen. The low visibility of homeworkers makes this a formidable task. It seems very unlikely that it will be possible for an overseas brand to create and impose an external solution on the homeworkers. Even Traidcraft, a benevolent NGO with formidable experience and expertise, faced major challenges in physically locating homeworkers during the 9-month project. Any form of external appointed auditors, inevitably viewed with some skepticism due to their role, would face correspondingly larger obstacles. Instead, any solution to this problem must arise organically from the bottom of the supply chain.

While the difficulties imposed by the COVID epidemic limited our ability to directly test different models, based on our experience within the project, we were able to identify two models of organizations that might prove suitable. The first was a single, highly socially engaged, factory (or supplier) that had spent considerable time and effort building up its knowledge of the homeworkers. We have encountered a few progressive suppliers to witness how this option works. It is easy to imagine such a company being trusted to act as an auditor by a Western brand. Obviously though, there is little chance of such a model scaling in the general case. Many factories in India are not interested in the homeworkers in their supply chains, and there is little reason for homeworkers to trust them.

A potentially more durable approach, described by Choubey et al. (Reference Choubey, Khumallambam, Mani and Preece2021), would be to encourage the homeworkers to form local self-help groups (SHGs) to provide mutual self-support and assistance, and as a group be supported by local CSOs. The SHG is a social structure that is common in Ambur, where women sometimes conduct saving-loan transactions among themselves (Choubey et al., Reference Choubey, Khumallambam, Mani and Preece2021). These groups would then automatically be able to find their local homeworkers and would be well-placed to gain their trust and work together as a collective. There are many relevant CSOs in India, with SEWA (Self Employed Women’s Association), SAVE (Social Awareness and Voluntary Education), SABAH (SAARC Business Association of Home-Based Workers), and FairWear Foundation, as some very prominent examples. Given this level of acceptance, this model is well-placed for developing organizations to support homeworkers in any digital supply chains arising in the Indian Textile industry complementing the progressive factory owners and suppliers in the previous model.

Since each local group would be small in scale, with the members knowing each other, they will be able to validate each other’s identities. Here, we refer to the data collection point where both the data contributor and the data collector must verify each other’s identity to establish trustworthiness for the data collected. In the case where this relationship is between two women living in the same community, identity verification is probably not needed. If a factory is the data collector, then the factory is likely to already have the identity of the homeworkers, or in a position to efficiently verify it. In other cases, where the data collection point is a well-established third party, for example, SAVE, then the ID verification process might be one way, that is, one only need to verify the identity of the data contributor but not the well-established data collector. The CSOs are also in a better position to perform some identity verification, checks on supply chains payments to homeworkers, as well as gather sufficient information to ascertain some average hourly wage that can be cross-checked with similar information collected by factories from the subcontractors. To protect the identity and privacy of the homeworkers, one could anonymize job and payment information, when sending aggregate information over to the factories and Western brands for analysis. The largest practical obstacle to this solution would probably involve the need to train the CSOs and especially the self-help groups in using any supporting software to report homeworker data.

If we define a collective as a collection of many local groups, there are many useful supporting roles this collective may deliver including acting as regional data collection points where useful labor statistics for this “hidden” bottom layer of the supply chains can be published. A sample of 10 Phase II interviews suggests the response to the collective initiative was mixed (Choubey et al., Reference Choubey, Khumallambam, Mani and Preece2021). The three women homeworkers interviewed at Kapas Hera said they understand what a self-help group is and associated it as “women working together and supporting each other.” They are ready to join one if offered and felt their husbands would support their decision. It is then no surprise that they are “happy to share very detailed information (including personal data, full ID) with the collective.” The women are wary of scams, and generally do not trust bank, contractor, brand, and supplier to hold their personal information. The local CSO representative thought some women might not appreciate what a collective is, but the local CSO representative is confident that the women are capable of understanding the concept of a collective, and that they would embrace it if the collective can help them get work or support them through COVID. There are other stigmas, for example, concerns about being seen as troublemakers by people higher up the supply chains, women with young children, Muslim minority, the husband’s, the family’s, and even the landlord’s attitude toward a collective could deter the women from registering with a collective. A female subcontractor was suspicious that the collective might take away her livelihood. A CSO explained that a lot of rapport building is needed to build trust before introducing a new collective; these women do not easily trust a new organization or a new set of people. As Carpenter (Reference Carpenter2021) pointed out, the proposed digital solution is as much of a social intervention as a technological one. Finally, it is also important to ensure interoperability between local groups, especially if they do not belong to the same CSO or the collectives overlap with two or more supply chains. Ideally, and to minimize data risk, the PII (personal identifier information) should be anonymized, to eliminate the collective’s responsibility toward personal identity protection.

Due to their vulnerability, anonymity and privacy are important attributes for the homeworkers, without which they do not have the confidence to chase late payments, expose sexual harassment cases, and so forth. They are even reluctant for the collective to act on their behalf if there is any hint of losing their livelihood or creating tension in the community. Harassment is a very serious matter. Any intermediation should involve experienced social workers such as SAVE in Tirupur. The technological challenge is much lighter in comparison.

The considerations for registering the subcontractors are different. In some cases, the male subcontractors act in a manner closer to a factory owner, while a female subcontractor’s position will be more akin to a homeworker. The latter might choose to register with a SHG, while the previous may onboard the digital supply chains through the factories. The subcontractors will need some reassurance that their roles do not become redundant in the new scheme. Some suppliers and brands emphasized that the subcontractors are indispensable in the complex, seasonal, and real-time process of work distribution, quality control, and delivery. Once the homeworkers (and their SHGs) are onboard, the subcontractors become the next weakest link. Careful considerations are needed so as not to jeopardize this important group in the chains.

In policy terms, then, the question becomes how the formation of the local groups is to be promoted, and how they represent the homeworkers within the digital supply chains. Western Governments can probably only help with this indirectly, by encouraging companies to form positive homeworker policies. The Indian government would naturally be better placed to have some influence in this area especially after launching the eShram portal. Having identified two sources of plausible candidates for gathering and validating the homeworker productivity and payment data, we now focus on the shape that a digitized supply chain with extensive homeworkers might take. We have not dwelt much into the rich literature surrounding trust, culture, and power dynamics in society, data ethics—the use and reuse of data of vulnerable groups, digital ethics of centralized and decentralized systems controlled by governments or private agents, and so forth. Apart from the space restriction, there is no contribution here in repeating these theories that have been well deliberated elsewhere. Instead, we focus on the context where the rights and working conditions of a sample of Indian textile homeworkers might be represented in the digital supply chains.

We do not recommend the creation of any new groups or new organizational form as this is very costly and unsustainable over the long term. We are proposing networking and sharing intelligence among existing groups and organizations. We suggested self-help groups as one important type of data collection point because of their rooted presence in some Indian communities. But in some cases, some manufacturers are better collection points. During our grant period, we had meetings with a few manufacturers who are both progressive and active in organizing the homeworkers in their supply chains. One manufacturer even actively looks out for sign of child labor exploitations in the families of the homeworkers. Similarly, we have SAVE (Social Awareness and Voluntary Education) at Tirupur who are collecting data from their 30,000 members for living wage negotiations. SAVE and the said manufacturers are already serving as data collection points.

3.1. Workers registration

The launch of eShram has made a significant leap forward in the mapping of homeworkers. Any worker who is unorganized, nonincome tax payee, and aged between 16 and 59, are eligible to register on the eShram portal. The eShram portal is a centralized database of unorganized workers seeded with Aadhaar where personal ID such as name, date of birth, address, photo, fingerprints, iris scan, and bank account details are stored. A 12-digit Universal Account Number (UAN) is uniquely assigned to each unorganized worker after registration on the eShram portal. The UAN number is a permanent number which once assigned will remain unchanged for the worker’s lifetime. If a worker does not have an Aadhaar-linked mobile number, he/she can register through Biometric authentication. After registering, the worker will get an accident insurance cover, and it is expected that in future, all the social security benefits of unorganized workers will be delivered through the eShram portal. In crisis and national pandemic-like situations, this database may be utilized to provide emergency assistance to workers. Bank details are being captured to ensure hassle-free delivery of benefits under social security schemes or any benefits by the Central/State government directly to the worker’s account. The eShram registration also adds useful information such as labor skill sets and some family details to facilitate employment. The UIDAI (the Unique Identification Authority of India) provides facility for licensed third party to verify personal ID using the Aadhaar authentication API.Footnote ⁵

The respondents at our phase II interviews said they would trust the known NGOs/trade unions and the collective to hold their PII even though the government is the body they most trust. They do not trust any other organizations (namely banks, suppliers, brands, and contractors) to hold their personal data. The homeworkers may not appreciate how the aggregated data is used, but they would like to understand the potential benefits of this data collection to them. Some women do not wish to share their Aadhaar card and bank account numbers. The Data Protection Laws must protect the storage and usage of PII by all data controllers and data processors respecting the different jurisdictions along the (global) supply chains. In addition, one should apply more stringent ethical codes due to the vulnerability of homeworkers.

3.2. Mapping of global supply chains

After the labor registration, supply chain mapping is the next most urgent task together with the need to persuade the brands to buy into the open homeworkers’ policy. For brands that have a homeworking policy, they are only concerned with those homeworkers in their supply chains, and not just any homeworkers. This traceability is especially important when dealing with harassment cases. Due to the multi-layers and potentially many-to-many dynamic relationships, supply chain information is notoriously difficult to map. At the time of writing, the Open Apparel Registry (OAR https://openapparel.org/) is among the best solutions available for supply chains mapping.Footnote ⁶ OAR is a free and interactive web portal for maps of garment facilities across the globe; it shows which organizations are affiliated with each facility. The OAR compiles and makes accessible data and information disclosed from multiple sources, including brands, retailers, civil society, multi-stakeholder initiatives, manufacturers, and government databases. The information is combined into a central, open-source map and database that provides transparency in clothing supply chains. Each garment facility is assigned a unique ID. The collated database of facility names, addresses, and affiliated parties is powered by an advanced name and address-matching algorithm.Footnote ⁷ The Business & Human Rights Resource Centre explained how this open-source map of garment facilities across the globe has helped human rights campaigners to respond quickly to abuse in supply chains.Footnote ⁸

3.3. Mapping of modern slavery risks

The Traffik Analysis Hub is a collaboration between Stop-the-Traffik and IBM (Watson), and others aiming to share information about human trafficking across all industries and sectors. The authenticated partners upload data from a large variety of sources. In addition, unstructured open-source data is ingested at scale—including thousands of publicly available news feeds. Using IBM Watson—AI, machine learning, and natural language recognition—an intelligent “golden tagging” schema is applied to the data based on terms and incidents related to human trafficking. The output is actionable data visualization and analytic with supporting interpretive narratives.Footnote ⁹ At the time of writing, the Traffik Analysis Hub has an added feature that can overlay the OAR supply chain map onto their human traffic map to facilitate supply chain risk assessment.

3.4. Information on minimum and living wage

From Sections 3.1 to 3.3, a natural next step is to combine, or overlay the textile supply chain labor information from eShram onto facilities such as Traffik Analysis Hub risk map and the OAR supply chains map, or the Indian equivalent. One useful information to overlay could be the hourly wage and pay rate of the homeworkers collated via the network of local groups in Section 2.5. Homeworker Worldwide, Cividep, and Petland presented a collaborative project in Delhi that is captured in the Hidden Homeworkers Toolkit published by Homeworker Worldwide.Footnote ¹⁰ The objective is to record work done by the homeworkers with agreed price/wage by capturing information on a job card. The data fields include Agent/Subcontractor, (Homeworker) name, ID number, Month, Date, Article, Pairs, Rate, Value, Grand Total, Advance, Net pay, and the signatures of the worker, agent, and internal auditor. The agents/subcontractors use the job card to record the number of pairs of uppers allocated to each homeworker, the agreed piece rate, and the subsequent wages paid. Both agent and homeworker were asked to sign the card providing a paper trail. In the digitized process that we envisage here, the agent could keep a digitized record via an app, and the local support group could use the app to upload a photo image of the job card, which will then be scanned or manually input the data into the system. A sampling audit process could help to ensure the integrity of the data before porting it into a regional, national, or international collection point such as the OAR for deriving the minimum/living wage information at a physical location where the local support unit is based. Once the transaction and cost information are aggregated, there will not be personal data, or the danger of infringing GDPR or data privacy laws.