Using direct benefit transfers to transfer benefits to women: a perspective from India

ABSTRACT Enabling women to have control over their entitled incomes from direct benefit transfers (DBTs) can strengthen their bargaining power and transform their banking activity, mobility, and other key outcomes. However, ensuring that women are able to effectively access these benefit transfers in an enabling environment where DBTs can lead to transformative impacts requires recognising the economic, normative, and social barriers that women face in accessing and controlling their accounts. It also requires an understanding of the incentives of financial services providers and their constraints in adapting to and addressing these barriers. This article addresses these and offers recommendations for policymakers and practitioners as they seek to improve women's financial inclusion and fully harness the potential of benefit transfers in empowering women economically and socially.


Introduction
The Government of India and various state governments have transitioned to direct benefit transfers (DBTs) as a way to transfer benefits and payments digitally to recipients' bank accounts. As of January 2019, 440 schemes covering farm and non-farm subsidies, social protection payments such as pensions and public workfare programmes, scholarships, academic fellowships, conditional cash transfers, and other government payments implement DBTs across 56 ministries, with Rupees (INR) 2,16,844 crores (US$ 2.1 trillion) transferred in total in 2018-19 (Direct Benefit Transfer Mission, Government of India 2019). Some of these schemes target women specifically, such as the Pradhan Mantri Ujjwala Yojana (PMUY) and Pradhan Mantri Matratva Vandana Yojana (PMMVY), while others, such as Mahatma Gandhi National Rural Employment Guarantee Scheme (MGNREGS) or the Krishi Unnati Yojana (KUY), go directly to beneficiary bank accounts held by any gender.
Ensuring that women have control over these DBTs in particular, and their own monies more broadly, can have transformative impacts not just on their own lives but also on their families and society as a whole. Women, particularly those who may be relatively less empowered in their households, are more likely to engage in the labour force, be more financially active and experience greater physical mobility if they have been trained on using their accounts and these are linked to their DBT entitlements (Field et al. 2016a(Field et al. , 2016b. 1 Similar evidence is available from other contexts. In Niger, cash transfers to mobile accounts helped decrease costs of accessing money, while improving women's bargaining power and household consumption outcomes (Aker and Ksoll 2016). In South Africa, financial inclusion improved women's decision-making power in the household and had a consequential increase in female labour force participation (Biljon, Fintel, and Pasha 2018). The link between DBTs and women's labour force participation is crucial to highlight, given the recent Periodic Labour Force Survey data from 2017 to 2018, which shows dismally high unemployment rates in India, particularly for women (National Statistical Office 2019).
Access to bank accounts is a necessary condition for recipients to receive DBTs. India has a long history of policy and regulatory initiatives to expand financial inclusion. 2 While the successes and limitations of these initiatives are beyond the scope of this paper, under Pradhan Mantri Jan Dhan Yojana (PMJDY), a national mission to ensure access to financial services, about 341 million accounts were opened between August 2014 and January 2019, with aggregate deposits of around US$12.5 billion as of January 2019. Of these accounts, 181 million were opened by women (Department of Financial Services, Government of India, n.d.). This is noteworthy progress towards the necessary first step of having a bank account. Of the accounts opened under PMJDY, public sector banks, which are majority owned by government, opened 79% of total accounts (Chitra 2018). This shows the promise of state-driven financial inclusion. However, dismal participation by private sector banks and limited usage of accounts demonstrates that much more work needs to be done to address market failures that have led to women's financial exclusion.
Financial Inclusion Insights (FII) data from 2017 indicate that India's gender gap in account ownership is down to 4.64 percentage points. 3 Further, PMJDY mandated at least one fixed service point in a sub-service area (SSA) for every 1,000-1,500 households, thereby addressing a pressing supply side problem. However, the World Bank's Global Findex 2017 finds that almost half of all account owners in India have an inactive account: 77% of women have an account, but only 35% have used it in the last year (World Bank Group 2017). It is thus unclear how PMJDY has improved the experience on the subsequent steps of accessing and using one's money.
Broad lessons from the PMJDY show that simply opening accounts can be a red herring that distracts from addressing the many barriers that prevent women from exercising their agency to use their accounts. In the fast-growing landscape of DBT payments, if women are significantly less likely than men to use and access their bank accounts, they will be systematically excluded from the potential positive impacts of state-initiated transfers. Given that women are already more economically constrained than men, if unaddressed, the gender gap in financial inclusion could exacerbate women's poor outcomes. To enable women to control their own monies in a way that is convenient, affordable, and reliable, there is an urgent need for a harder look at the incentives of public and private sector financial services providers (FSPs) and their constraints in adapting to and addressing the barriers that women face.
Men and women both face barriers to accessing and using financial services. However, women face significant hurdles and optimise their lives under a different, and often larger, set of social, economic, and normative constraints. This holds true with respect to usage of financial services as well. According to FII data, the gender gap in usage of accounts in India was 11 percentage points in 2017, with 59% of men using accounts actively in the past 90 days compared to 48% of women. For advanced uses like digital payments, the gender gap increases -42% of male account owners use digital payments, while just 29% of female account owners do (Financial Inclusion Insights 2018).
Women in India have systematically been excluded from access, usage, and control over finances, and changing this so they feel financially empowered requires a policy design approach that chips away at the compound constraints that women face. We have divided these constraints into specific areas, although we recognise that there is significant interaction between them. While we explore each constraint in turn, women may be trying to optimise their lives under all of these constraints simultaneously. As such, addressing any of them without addressing others might have limited impact. For instance, Field et al. (2016aField et al. ( , 2016b find that opening bank accounts for women and teaching them how to make transactions was insufficient. Linking their bank accounts to DBT payments was also a crucial, concomitant step to financial engagement. We outline recommendations for each focus area and recognise that it will take a host of interacting changes to improve the landscape of financial services for women. This is not an insurmountable challenge, and initiatives by the central and state governments, researchers, not-for-profit organisations, and FSPs are clearly demonstrating a path forward.
Women's constraints, FSP incentives, and role of policymakers and regulators In this section, we elaborate four areas that currently limit women's financial inclusion. These are limited physical access to banking services, lack of information and financial literacy, limited use cases and poor woman-centric product design, and high social and reputational costs. Each subsection includes background on issue, exams current solutions or activities, and outlines remaining gaps and opportunities. For each barrier, we explore the extent to which banks can address it and the challenges that banks face. FSPs on their own may not necessarily have sufficient incentives to address these constraints, making a case for the government and regulators to play a greater role as facilitator. To this end, we discuss the role of government and regulators in solving misaligned incentives and resulting inefficiencies.

Area 1: limited physical access to banking services
The Reserve Bank of India (RBI) issued a notification in 2017 that required banking outlets to be set up in villages with a population of over 5,000 by December 2017 (Reserve Bank of India 2017). Census data from 2011 shows that 53% of rural India live in villages smaller than 5000 people, suggesting that the RBI notification doesn't speak to the financial needs of over half the country (Open Government Data Platform India 2015). However, FII data does demonstrate an improvement in physical access to banking facilities, as 52% of respondents report having access to a financial point of service within 1 km away from their homes, and 83% having access within 5 km (Financial Inclusion Insights 2018). 4 While this is a tremendous achievement in such a geographically diverse country, the extent to which this has addressed the mobility constraints that women face in accessing banking institutions is yet to be determined.
Research shows that the vast majority of women in India often face large constraints to physical mobility. We define "mobility constraints" as a composite concept consisting of normative constraints that prevent women from leaving the house or travelling outside the village, opportunity costs of time in visiting the banking outlet, and financial costs of paying for transportation to and from the outlet. While it may be possible for a woman to walk to a banking outlet which is a kilometre away, in practice, norms related to women's mobility constrain them from visiting it to access their money. 5 Women may also have family members accompany them to the banking outlet for numerous reasons including lack of permission to travel outside the village alone or lack of knowledge or confidence to operate banking facilities on their own. If women either choose to or get permission to visit the banking outlet alone, the next constraints they face are in terms of actually visiting the outlet. If it is up to 5 km away, it can take between 45 and 90 min to walk there. They may need to take the day off work, arrange for child care, delegate domestic tasks to others, or complete these before visiting the outlet. If the banking outlet is beyond walking distance, they may also need to arrange for money to pay for public transportation. Even if these hurdles are crossed, receiving effective service at the banking outlet is not guaranteed due to poor customer service for women, navigating banking processes, long waiting times, and issues of liquidity and connectivity that affect the service reliability at the banking outlet. Given existing time poverty (women perform numerous domestic tasks, familial care work, and farm labour), spending a few hours in a queue to update a passbook is particularly costly.
Countries such as China and Kenya demonstrate through platforms like Alipay, WeChat, and M-Pesa that digital payments can provide financial services at scale to sparsely populated geographies. However, in India, the picture of women's digital engagement is bleak. The gender gap in phone ownership overall is 33 percentage points, while the gap in smartphone ownership is worse, at approximately 65 percentage points (Barboni et al. 2018). Researchers at Evidence for Policy Design (EPoD) found that gender norms around purity for marriage, subservience, patrilocal exogamy, and caregiving present a significant barrier to women's mobile engagement. Given this, we do not expect digital financial services on phones, by themselves, to solve women's low financial inclusion in the short to medium term.

Profitability of agent networks
Given the low levels of women's digital engagement, easy access to banking outlets for cash in-cash out services remains a necessary condition to enable women to control their monies. However, there is a fundamental misalignment of incentives in serving low-income rural customers. While improved access to financial services would benefit these customers, this is often not profitable for the FSPs. Rural customers typically maintain low bank balances, are geographically spread out, and have low transaction volumes per unit area. As a result, investments by banks in banking outlets are largely under-funded, or funded through marketing or corporate social responsibility (CSR) budgets rather than as profit centres with a strong business case. 6 While rural banking agents' median earnings doubled from US$40 in 2015 to US$93 in 2017, challenges to a strong business case include high operating costs that also doubled from US$32 in 2015 to US$62 in 2017 (Mehrotra et al. 2018). Higher levels of dedicated agents, who solely conduct agency banking and do not have additional businesses such as groceries and mobile recharges in rural areas, require these high costs be recovered from a single revenue stream. Compared to many other developing countries, banking outlets in India earn relatively low revenue per assisted transaction that is shared across banks, corporate banking correspondents (CBCs) who are agent network managers, and agent banking correspondents (ABCs), who are the retail banking outlets (MicroSave n.d.). Further, many FSPs cite low transaction volumes in remote rural areas as a challenge to setting up banking outlets. Unnikrishnan et al. (2019) found that "rural oases", locations that are far from urban centres but have relatively high levels of economic activity, have a high proportion of banking outlets that report healthy profits on volumes of 30-40 transactions a day. However, "frontier locations", characterised by low levels of population density and economic activity, are less likely to develop networks of banking outlets organically. While there are 786,000 banking agents in India (Business Correspondent Federation of India 2018), including other points of service such as bank branches, the number of financial points of service in India is 98 per 100,000 adults, compared to countries like Kenya, Uganda, Bangladesh and Tanzania, where there are between 500 and 900 financial points of service per 100,000 adults (MicroSave n.d.).
The business case for providing bank accounts and banking outlets to financially enable women can be made stronger with an expanded portfolio approach that includes products like credit, insurance, merchant payments, and financial advice, including planning for income shocks (GSMA 2016). Offering such services would give FSPs an opportunity to earn greater profits per customer, provided that the products are designed keeping in mind the "suitability principles" recommended by the Committee on Comprehensive Financial Services for Small Businesses and Low Income Households (Ananth et al. 2013). A successful suitability assessment must take into account two parameters: the customer's requirements and objectives, and their financial situation.
These products can be sourced from multiple FSPs and aggregated for the customer at the point of service. Such an approach would serve low-income women for various financial needs including payments, liquidity, mitigating and responding to shocks, and financial planning (Insight 2 Impact 2018). This would also result in movement towards "complete financial inclusion" (Ananth, Chen and Rasmussen 2012). While there exist market failures and high transaction costs in financial services like credit, insurance, and investment, deeper relationships with customers driven by trained wealth managers and geographic saturation are a promising approach to address some of these challenges (Ananth 2018).
Some business models and decisions within the control of FSPs can help within the current policy and regulatory environment as well. ABCs that are not dedicated to agency banking as their only source of income and are not exclusive to servicing customers of one bank are reported to be more viable, given higher business activity and costs spread across multiple profit centres. Newer models of inventory management of digital money centrally have also shown promise in increasing profitability by preventing stock-out losses and reducing the cost of capital for banks or CBCs (Balasubramanian, Drake and Fearing 2018). 7 FSPs could also target improving the customer experience with banking agents and increasing the effectiveness of agents to provide financial services. For instance, more needs to be done to ensure liquidity of both digital money and physical cash inventory such that when women do show up to transact, there is adequate inventory to dispense. Other issues include grievance mechanisms for when women have uncomfortable experiences with agents. FSPs must incorporate a broader sensitivity to women's experiences in accessing their accounts, and train agents to be more facilitative. Recently, the Business Correspondents Federation of India (BCFI), supported by the World Bank Group, DFID and SIDBI, announced initiatives that include development of a code of conduct, multilingual content for training ABCs, frameworks for ABC certification and registry, grievance redressal mechanisms, and standards for interfaces between CBCs and ABCs. These are promising initiatives, and we are eager to understand their impact on women's inclusion.

Interoperability
This raises the next question of why FSPs don't coordinate among to provide services like deposits, withdrawals and money transfer to each other's customers, also known as interoperability. Despite the technology architecture in place, many large commercial banks are reluctant to fully support interoperable agent networks, making unit economics of providing these services less attractive for smallerscale providers (Jacqueline, Ondieki and Nanjero 2017). Particularly, the interchange fees that a bank pays for its customer using the point-of-sale (PoS) infrastructure provided by another bank continue to be a point of contention (Pani 2018). Banks that have a large customer base are against any minimum interchange fees and advocate for lower pro-rata fees, whereas smaller banks argue that current interchange fees are not attractive enough for them to deploy more PoS machines in rural areas. Further research is needed on optimum interchange pricing across the supply chain of providing financial services and to uncover other barriers to interoperability.
Another issue in FSP cross-coordination relates to using the interoperable payments infrastructure layer offered by National Payments Corporation of India (NPCI) through Aadhaar Enabled Payments System (AePS). This allows online interoperable financial transactions such as cash withdrawal, cash deposits, fund transfer, and checking balances at PoS machines through the banking outlet of any bank using the Aadhaar authentication (NPCI n.d.). The recent judgment by India's Supreme Court allowed the use of Aadhaar for DBT payments, but struck down provisions of the Aadhaar Act that allowed FSPs to use Aadhaar outside of such payments (What Supreme Court's Aadhaar Verdict Means for You 2018). Many FSPs have interpreted this to also mean that the use of AePS for withdrawal of DBTs is still allowed. However, this is seen as a setback for interoperability overall, especially for uses like deposits and non-DBT related withdrawals, and there continues to be trepidation and confusion among FSPs in using AePS. It remains to be seen how the government, regulators, and FSPs make interoperability work at scale in the light of these developments.
Given the mobility barriers women face, coverage of banking outlets needs to be continuously monitored and expanded to reach areas where setting up bricks and mortar branches is unviable. However, FSPs are reluctant to make inroads in remote, rural areas where revenues do not justify the costs of setting up outlets. Ironically, these are the regions where mobility barriers are strongest and the need for such centres highest.
For the government and the regulator to effectively address this market failure, there is an urgent need to first understand the magnitude of the issue: its nature, and the number and spread of underserved populations. FII indicates that 83% of adults knew of a financial point of service within 5 km of their households, but reliable and regularly updated public data do not exist to understand the coverage and location of access points, and how many of these banking outlets are active (Financial Inclusion Insights 2018). 8 We urgently need systems to provide updated micro-data on coverage, understand the extent and nature of under-coverage, and identify where further interventions are needed.
Once the government and the regulator understand the spread of underserved populations, they can explore instruments to address the challenge. These could include re-calibrating and making more effective instruments that are already used, such as subsidising costs for ABCs and CBCs to operate in these areas, tax incentives, or providing regulatory mandates to serve these populations. More importantly, once this information is available, FSPs can themselves identify underserved locations which can be profitably targeted.

Area 2: information and financial literacy
Bank accounts can often be a black box for low income, illiterate, and innumerate women. They face a twofold constraint: FSPs give limited information on their accounts and banking processes, and women are also often not financially educated enough to access this information independently. To partly offset costs of maintaining and servicing numerous basic savings bank deposit accounts (BSBDAs), which do not require a minimum balance to be maintained and have a limit of four withdrawals per month, FSPs cut the costs of providing customers with information on transactions in convenient and accessible ways. For instance, BSBDA holders often do not receive SMS alerts on bank transactions. This is not to say that providing women with information via SMS would improve financial literacy, as women often don't have individual phones to receive these messages and may also not be able to read. The gender gap in reading and sending SMS messages is 51 percentage points in India (Barboni et al. 2018). However, the broader point is that low-income women with BSBDA accounts face higher constraints to accessing information about their accounts compared to women with regular savings accounts.
Anecdotal evidence further elucidates the information constraints that women face. Our team interviewed women in Chhattisgarh who had worked for MGNREGS to ask how much money they had been paid for their last work-spell. Some had been paid plausible amounts, while others said that they had been paid inexplicably low amounts, in the single-or double-digit rupees. Such low wage transfers are implausible, since the minimum daily wage under MGNREGS is more than INR 168 anywhere in the country since April 2017, suggesting that in the absence of information on payments, women were confusing their current balance with their latest MGNREGS payment. This highlights the lack of information that women have about the amounts that are transferred into their accounts, and withdrawals made against these. It also highlights a key implication for policy implementation of DBTs for schemes such as MGNREGS or PMUY, where the frequency and amounts may vary considerably over a given time period. Women find it hard enough to know when funds are transferred into their accounts, let alone keep track of payments that are calculated based on monthly gas prices or number of hours worked under MGNREGS.
In an effort to address low levels of financial literacy in India overall, the RBI mandated that bank branches are meant to conduct new customer information sessions that provide an overview of the benefits of using bank accounts, including saving for the future and receiving government benefits (Reserve Bank of India 2009). However, it is unclear how many banks have conducted this training, and what its quality has been.
We believe there are two clear ways of breaking down information barriers: improving women's financial literacy and providing women with information on their entitlements in an accessible way.

Banking literacy
Given the right use case, hands-on financial training can significantly improve banking outcomes for women who have DBTs coming into their accounts (Field et al. 2016a(Field et al. , 2016b). Yet, most of the regulator-mandated training, and financial training globally, have not achieved desired results in improving banking activity (Miller et al. 2014;Fernandes, Lynch, and Netemeyer 2014;Ogden 2019). The government could reconsider how it provides training to address information and literacy barriers. We posit that the government and the regulator need to consider two types of training for unbanked women. First, a practical group-based training for women at the local banking agent (typically, the banking kiosk) to reduce their initial hesitation in approaching them. Second, on-the-spot assistance as women start making transactions at the banking facility. Research shows that content knowledge may be better conveyed via "just-in-time" financial education tied to a particular decision, enhancing perceived relevance and minimising the potential to forget information (Fernandes, Lynch and Netemeyer 2014).
JEEViKA, the programme for improving rural livelihood opportunities run by the Government of Bihar through Bihar Rural Livelihoods Promotion Society (BRLPS), has experimented with a model of on-the-spot training and assistance at bank outlets. Bank Mitras (BMs) are self-help group (SHG) members employed by JEEViKA who sit at the banking outlet during stipulated hours and assist women who come to the bank to access their SHG funds. The BMs help with two core activities, filling out deposit and withdrawal forms. They walk women through these basic activities and provide support as needed while women interact with the banking officials. Their scope is limited to SHG transfers, but this has made them hugely popular with women in nearby villages. SHG women tend to know the BM on a first name basis and are confident about accessing funds with their help.

Information on entitlements and timings of payments
It may not suffice to have more DBTs directed to women if women are unaware of what schemes are they entitled to and how much money should they receive. If the agent who provides this information is also key in the disbursal of funds, s/he may have incentives to provide misleading information in order to siphon off funds. The government will need to carefully consider these while identifying information channels. We recommend piloting a system with a greater role for local collectives, such as those formed through the State Rural Livelihood Missions (SRLM), to provide information to women. Since collectives are formed around trust and informal associations between women, and typically operate through group interactions, they have little incentive to deliberately provide misleading information.
It is also important for women to know when DBT payments are delivered to their accounts to avoid unnecessary trips to banking outlets. EPoD researchers collaborated with a local banking partner to provide daily interactive voice response (IVR) calls in the local language to women to inform them about transactions in their account. Preliminary interactions with beneficiaries show that women value receiving this information. However, phones often tend to be transferable within households, and there is a considerable gender gap in mobile phone access. Thus often these calls were received by spouses or other family members, who can then potentially influence women on how the money should be used, lessening the control she can exercise. More research is needed into the efficacy and unintended consequences of phone-based informational interventions.
SRLMs can also help address this information gap. JEEViKA implemented an initiative to transfer INR 12,000 to members who built toilets in their homes. Beneficiaries learnt that the funds had been transferred to their accounts from their JEEViKA didis at group meetings. Receiving this information in-person in their villages saved a trip to the bank to find out if the money had been deposited. They knew exactly when they could visit the bank to withdraw their money, and how much money they could expect in their accounts.

Area 3: use cases and product design
Because women have long been excluded from earning an income and subsequent access to financial services, many women we interviewed across Bihar, Uttar Pradesh, and Chhattisgarh had internalised the norm that it is best to leave money matters to male household members, and that they didn't need independent control over their income. According to FII data from 2017, 45.39% of women who had a bank account or were able to use someone else's account thought they didn't need the account or never thought of using it. Bernhardt et al. (2017) observe low average returns of female-run enterprises because women's capital is invested in their husbands' enterprises rather than their own. Households may be optimising whose skills can provide higher returns to investments, however, such optimisation creates a norm where women may feel discouraged from financial engagement overall.
Women's banking outcomes improve if funds go into their individual accounts. This points to the need for considering more opportunities for funds to be targeted towards women, which in turn means improving women's participation in common DBT schemes, such as by creating enabling conditions for women to work in MGNREGS, as well as having more household-level DBT schemes targeted towards women with funds directed to their accounts. Field et al. (2016aField et al. ( , 2016b show that opening accounts and training women on how to use it is necessary but insufficient to improve their financial engagement. Linking their accounts to their MGNREGS payment ensured resources were transferred to these accounts, thus providing opportunities to women to transact with their accounts. Governments also need to ensure that funds meant for women are directed to their accounts rather than their spouses' or relatives' account. The Supreme Court has upheld the constitutional validity of Aadhaar, and it remains mandatory for availing social welfare schemes. At the same time, the Court ruled that bank accounts do not necessarily have to be linked to Aadhaar. If bank accounts are not linked to Aadhaar, it loses potency in solving the targeting issue. 9 The government is now considering other ways to encourage people to have their accounts linked to Aadhaar (HT Correspondent 2018). Whether through Aadhaar or more direct monitoring, there is a strong case for the government to ensure that DBTs targeted to women flow into their own privately held accounts.
The use case barrier is exacerbated by the dearth of financial services that account for social and economic constraints that women face in particular. This further reduces the value proposition for women to use these accounts. Since most DBTs are not gender-based, women may opt for funds to be directed to their husband's, father's or son's bank accounts. Well-designed financial services that speak to women may incentivise them to direct and control these funds in their own accounts. By and large, FSPs have not dedicated resources to understanding financial services that are useful and engaging for women. Formal financial services remain deeply underpenetrated in India, visible in the modest bank credit to GDP ratio, households continuing to save substantial amounts in gold and real estate, and inadequate take-up of insurance (FE Bureau 2017). Women may find informal financial services such as chit funds (informal rotating savings and credit associations) and SHG savings more appealing, as women may personally know and trust the service provider, and these services permit savings in smaller quantities that might be considered more suitable.
To address this, the recommendations of RBI's Household Finance Committee (Reserve Bank of India 2017) are useful, particularly with respect to supply side interventions. One recommendation is an essential minimum financial kit for households with financial products such as a simple nofrills account, simple term life insurance, automatically triggered catastrophe insurance, a basic NPS account for pensions, and access to institutionally provided unsecured debt (Halan 2017). The report also points to the opportunity for FSPs to offer better-designed products, such as microsavings products with reminders, where the deposit size per transaction could be very small, and savings instruments linked to merchant or remote payments that allow for rounding up small transactions and channelling them to savings.
To provide women with financial services that are suitable, affordable, reliable, and convenient, FSPs need to understand their behaviour, needs, social, and economic constraints, and embed these lessons across the product development cycle that speak to women's experiencesfrom product features, marketing and communications, distribution and customer onboarding, customer experience at points of service, and grievance redressal. We posit that there are a few reasons that FSPs continue to offer a product experience that doesn't work for women.

Lack of user-centred design
Myths among FSPs continue to exist. The worst are around homogeneity of men and women as consumer segments (Dalberg Advisors and Design 2019), and lack of business case to serve women beyond severely limited cases such as withdrawal of DBT payments and remittances. At best, it is the belief that feminising existing products with heteronormative features such as the colour pink or adding shopping benefits or higher withdrawal limits is a useful differentiator (Zee Business n.d.). However, these initiatives do not speak directly to the entirety of women's experiences with financial products, including features, distribution channels, and communications, messaging, and customer service. Designing for women in varied social and economic contexts, and not considering women as a homogenous group, requires FSPs to have a more nuanced view of consumer segments and understanding them across demographics, behaviour, and psychometrics. Instead of focusing on using their existing products as the point of departure, FSPs could also try turning their attention to investing in first-hand research on the gaps in products' experience to suit customers.
Some microfinance institutions (MFIs) that are concerned about serving their female client base better, including prominent players in Maharashtra and Tamil Nadu, have mentioned that their clients want to keep their bank balances and transactions private from the rest of their family. One MFI designed information services with specifically this in mind, with women informed of their repayment status when they visit the branch to repay their loan amount. These women prefer not to receive messages or calls on the household mobile phone. However, these examples do not reflect the norm, and are much less found in FSPs that operate on scale.

Lack of incentive to invest in research and inaccessible research outputs
FSPs are generally not incentivised to establish robust institutionalised processes to independently understand the product needs of their low-income customers. Nor do they invest resources into ensuring that feedback from low-income customers reaches their product development teams to design more customer-centric products. This is driven by perception of lack of business case, and lower returns on investment for products for low-income consumers. Some of the research and resources to advance customer-centric design exist already, such as the Customer-Centric Guide by CGAP, the Human Account by Dalberg Design, and USAID'S Gender ICT Survey toolkit. Wide research exists on designing products and services that work for women. This includes reducing product complexity and the cognitive load to understand product processes, reducing informal transaction costs such as paperwork and bureaucratic impediments that can exacerbate the feeling of embarrassment and shame, recognising preference for illiquid savings that are less likely to be appropriated by family and friends, using more visuals and less text for illiterate customers, and addressing mobility constraints. Further, people have relational reasons (perceived trust and a sense of belonging) for product choice (Houggard 2018). This is most visible in the use of informal financial services. However, many FSPs are either not aware of this research, find it difficult or abstract to apply to their businesses, or find that independent research is sometimes at odds with their own mental models. It is rare that research is translated into languages that make passive cascading of the research possible. Researchers could do more to disseminate their work among FSPs, help them walk through the findings, and involve FSPs closely during the research itself to better understand the barriers to design and deliver more women-centric products. Additionally, researchers and FSPs need to collaborate and gain insights from administrative data to understand women's experiences with financial services, and where they fall off on the user journey from registration towards active advanced usage of financial services.

Area 4: social and reputational costs
Women also face a variety of social costs in financial access. An MFI partner in Uttar Pradesh voiced concerns about women's perceptions of amounts that were "worthy" of depositing. Depositing amounts lower than INR 500 (approximately US$7) or INR 1000 (US$14) was embarrassing, and as such women may store cash at home until they had a large enough amount of money to deposit.
The State of the Agent Network, India 2017 report highlights that women form 51% of the customer base of female banking agents, but only 42% of customer base of male banking agents (Mehrotra et al. 2018). It also provides anecdotal evidence that women are more comfortable transacting at kiosks managed by women agents. This provides a case for regulators to consider setting minimum targets for women-led banking outlets. SRLMs in several states in India have also invested in having women set up and operate banking outlets. This model holds immense promise and needs to be rigorously evaluated to motivate scale up.
Women face numerous issues related to trusting their FSP. While the government may move towards increasing DBT payments for women, it could consider simultaneously establishing a body responsible for regulating and incentivising the various agencies involved in the delivery chain, each of who may have their own priorities which may otherwise not be necessarily aligned with women's access and usage of DBTs. If this isn't feasible, they could also consider entrusting this function to its current organisation. In either case, the body responsible could also commission further research to address existing evidence gaps, and work with various organisations on policy recommendations based on evidence generated. One particularly important task that this body could take on would be establishing grievance redressal mechanisms designed with women in mind. It may be a single toll-free helpline number where women could call to state their grievances. The calls could also be initiated by the government itself. The cell could triage the issue and inform beneficiaries of bottlenecks; for instance, it could be that the respective ministry hasn't processed the payment, that women are unclear about the amount of payment received, or that issues exist at the financial service provider's end. This body could use information from various ministerial online MIS systems to address the beneficiary's query. This helpline number could be publicly displayed at each banking outlet and customer service provision point.
State governments are already launching such initiatives. In Telangana, live calls were pushed to 46,000 farmers to ask about receipt of cheques related to the Rythu Bandhu scheme. Phone-based monitoring led to a 3.9% increase in farmers receiving transfers on time and a 1.5% increase overall (Muralidharan et al. 2019). While further research is needed to determine the impact of such a system on women, these initial results are promising. Another example of government intervention for monitoring payments is an intervention in MGNREGS, called PayDash, developed and currently being studied by EPoD researchers, which monitors the payment process and generates realtime, actionable information for officials on delayed payments and the administrative steps where the payments are delayed (Dodge et al. 2017). Both initiatives are examples of governance reforms that address the challenge of information asymmetry that limit the ability of policymakers to monitor the quality of public service delivery. Since these are also supply-side interventions, they partly circumvent the challenges of power imbalance and information asymmetry that citizens face in holding frontline workers accountable for services.

Conclusion
DBTs have significant potential to transform women's lives and lead to societal progress. However, women face several barriers in accessing their entitlements from DBT programmes. In this article, we elaborated the key barriers that women face, including low access to information, financial literacy, technology, mobility constraints, lack of compelling use cases, and social and reputational costs. We provided examples of initiatives taken by some FSPs to adapt to these barriers, while digging deeper into the constraints that FSPs typically work with, particularly around the economics of serving remote rural areas, and the challenges to investing in understanding women as consumers and offering products that meet their needs. We made recommendations that governments, regulators, and researchers can consider to address these critical barriers to foster an ecosystem that ensures women are able to exercise their agency in using these benefits and payments.
India is one of the few countries in the world that is in a position to achieve this goal. It has a strong interoperable payments infrastructure layer, a mature system of DBT payments across an array of government schemes that go directly to recipients' accounts, and a biometric identity verification system that has reached almost all adults. India's innovations in financial inclusion and digital payments are being studied by many countries, and we strongly believe India has an immense opportunity to lead the way in harnessing DBTs to financially empower women. 3. Account ownership gap in adults was 6% according to Findex 2017. We believe the difference with FII data from 2017 findings may be attributable to the way the questions are phrased and the different periods of data collection. 4. In 2018, 43% of respondents report having access to a financial point of service within 1 km and 81% report having access within 5 km. It's not yet clear if this is a statistically significant decline. 5. We use 'banking outlet' to encompass a broad range, including business correspondent, banking agents, bank branch, financial point of service, customer service point, and kiosk. 6. The Companies Act in 2013 specifies that corporate entities above a certain size must dedicate 2% of their pre-tax average profit from the last three years to CSR. 7. When people give agents physical cash for deposits or money transfers, agents have to transfer the money they hold digitally into someone else's account (depositor or someone to whom money is being remitted). Both digital money and physical cash are thus inventories of the agency banking business. 8. The authors do not endorse 5 km as a reasonable target for women's access, and further research in this area would be useful. 9. If bank accounts are not seeded with Aadhaar, Aadhaar can't be used as a payment address, therefore leaving account numbers as the most used payment address to transfer funds. It is likely that the account numbers given by recipients of DBT payments may belong to someone else in the family.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
This work was supported by Bill and Melinda Gates Foundation.

Notes on contributors
Rashi Sabherwal is a Senior Research Manager at EPoD India at IFMR. She works on the gender portfolio of projects spanning research questions on women's constrained mobile engagement and declining female labour force participation in India. Previously, she worked at the Jameel Poverty Action Lab (J-PAL) on the Energy and Environment portfolio.
Devesh Sharma works as Program Officer, Country Finance at the Bill and Melinda Gates Foundation on digital public financial management reforms. Previously, he worked on strengthening cash transfer programmes, measurement of the state of financial services, women's economic empowerment, taxation advice, and securities markets compliance. All views are personal.
Neeraj Trivedi is the Associate Director at Evidence for Policy Design (EPoD) India at IFMR. He leads the capacity building portfolio that focuses on building capacities of policymakers and government agencies to use evidence and economic theory for decision-making. In addition, Neeraj also leads EPoD India at IFMR's policy research engagements in the financial inclusion portfolio. Prior to joining EPoD India at IFMR, Neeraj was with Pratham, a leading educational nonprofit in India, for seven years.