K Sh K Sh K Sh K Sh K Sh M K Sh aking pension savings easy and efficient for informal sector workers Learning K Sh from Kenya’s Haba Haba pilot K Sh K Sh K Sh K Sh K Sh Contents Acknowledgments . . . . . . . . . . . . . . . . . . . . . . iii 1 Background 4 Delivery Chain Assessment Stage 1: Assess . . . . . . . . . . . . . . . . . . . . . . . . . 6 Stage 2: Enroll . . . . . . . . . . . . . . . . . . . . . . . . 15 Stage 3: Contribute . . . . . . . . . . . . . . . . . . . . . 17 Stage 4: Provide . . . . . . . . . . . . . . . . . . . . . . . 23 Stage 5: Manage . . . . . . . . . . . . . . . . . . . . . . . 25 27 Recommendations Area 1: Design considerations . . . . . . . . . . . . . . . . 27 Area 2: Design and implementation of outreach and communications strategy . . . . . . . . . . . . . . . . . . 31 Area 3: Updating business processes and administration . 34 Area 4: Institutional strengthening . . . . . . . . . . . . . 36 Prioritizing recommendations for actions in the short term . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 43 Annexes Annex A . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Annex B . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 49 References ii Acknowledgments T his report was coauthored by Himanshi Jain (Senior Social Protection Specialist), Somya Bajaj (Economist), and Agastya Yeachuri (Consultant) of the World Bank’s Social Protection and Jobs team in collaboration with the National Social Security Fund (NSSF) team: Awiti Millicent, Stephen Obare, and Evans Ombui. The team benefited from inputs by Anthony Omerikwa, former NSSF Chief Executive Officer/Managing Trustee. This publication was originally written as a background note to inform the design of the savings component of the World Bank–financed National Youth Opportunities Towards Advancement Project (NYOTA) in Kenya, which aims to increase employment and earnings and promote savings for targeted youth at the national scale. The relevance of the note prompted the authors and the NYOTA team to publish it as a stand-alone report that would enable similar schemes to use its delivery chain framework methodology to identify operational gaps and make improvements. The report benefited from peer review by Melis Guven (Global Lead of Delivery Systems) and Raphaela Beatrice Karlen (Senior Social Protection Specialist) and was prepared under the overall guidance of Paolo Belli (Practice Manager at time of report preparation) and Suleiman Namara (Practice Manager, Eastern and Southern Africa at time of completion). Nita Congress was responsible for copyediting and designing the report. The findings, interpretations, and conclusions expressed in this report do not necessarily reflect the views of the Executive Directors of the World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. For questions about this report, please email Himanshi Jain (hjain1@worldbank.org). iii M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s Photo: poco_bw/Adobe Stock iv Background H aba Haba, which means “bit by bit” in Swahili, is a voluntary pen- sion scheme in Kenya for workers in informal employment and promises to be a scheme through which they can slowly but surely save for their old age.1 The scheme, administered by the National Social Security Fund (NSSF) was launched as a pilot in 2019. Haba  Haba allows for easy, anytime, anywhere savings by informal economy workers. Registration, contributions, and access can be handled via mobile phone by dialing USSD (Unstructured Supplementary Service Data) *303# or by con- tacting the NSSF via WhatsApp. The registration process only requires an individual’s first and last names, and government identification (ID) number. Contributions can be paid in person at NSSF offices or through the mobile money platform M-Pesa. The program provides flexibility as to when and how much an individ- ual contributes. The recommended saving is K Sh 25 ($0.17) a day so that a monthly minimum of K Sh 400 ($2.73) can be achieved.2 There is no pen- alty if individuals fail to meet this target. However, if they do not meet the Haba Haba allows for easy, monthly minimum targets, they will not be eligible for short-term withdraw- anytime, anywhere savings by als under Haba Haba. If they meet the annual minimum, they can withdraw informal economy workers; its up to 50 percent of the fund balance after five years of contribution. The services can be accessed via remaining balance can only be withdrawn as a lump sum upon reaching the mobile phone. retirement age of 55. Each month, an individual’s account balance is credited with the investment return declared by the NSSF for Haba Haba. Participants can access their contribution history and fund balance by logging into their account or via WhatsApp/USSD using their NSSF ID number. The scheme aims to expand social security coverage in Kenya by providing a channel for income security at retirement for informal sector workers who are not mandated to be covered under the formal scheme. It is esti- mated that there are approximately 15.05 million workers in the informal 1  As addressed and covered in this report, workers in informal employment are those employed in informal enterprises, which are not mandated to contribute to the provi- dent National Social Security Fund. Typically, workers in informal employment are not covered by social insurance benefits such as sick pay, maternity benefits, disability pension, or old-age pension (KNBS 2020). All currency conversions of Kenyan shillings to U.S. dollars reflect the rate in effect on 2  September 13, 2023 ($1 = K Sh 146.70). 1 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s sector in Kenya (Guven, Jain, and Joubert 2021); as of June 2022, only about 2.6  million individuals were contributing to the NSSF’s mandatory scheme for the formal sector, according to NSSF data. Some of those in the infor- mal sector are above the poverty line and able to cope with shocks.3 This subgroup of the informal sector can be referred to as the “nonpoor and resil- ient” and comprises 52.5 percent of households in Kenya (Guven, Jain, and Joubert 2021). This subgroup is likely to have some ability to save and could be encouraged to save consistently through monetary and nonmonetary nudges. The coverage expansion strategy adopted by Haba Haba involves creating public-private partnerships. Since its inception, Haba  Haba has targeted trade groups and associations whose members are more organized and have relatively stable incomes such as the Digital Taxi Forum, boda boda drivers, and market vendors in select communities. This approach, primarily focused on organized groups in urban and semi-urban areas during the pilot stage, was adopted because of their likely familiarity with the NSSF (which man- ages schemes for the formal sector) and the potentially higher disposable income of these groups relative to rural areas. In the Haba Haba pilot, the NSSF collaborated with Proto Energy, the National Health Insurance Fund Haba Haba has a high inactivity (NHIF), Kenya Commercial Bank, and Safaricom to design and provide an rate—close to 80 percent of attractive bundle of benefits to meet the varying and diverse needs of those registered individuals—which in the informal economy. Among workers in the informal sector, the NSSF suggests a need to carefully estimates that there are about 60,000 taxi drivers, about 1 million workers evaluate processes beyond the in the matatu sector (based on an inventory of 500,000 matatu vehicles, initial outreach. each of which requires both a driver and a conductor), 1.3 million motorcycle taxi riders (boda bodas), 8 million market vendors, and 100,000 fishers and fish vendors who can be potential savers in the Haba Haba scheme (Guven, Jain, and Joubert 2021). Transport workers (taxi, boda boda, and matatu driv- ers) were among the first groups in the informal sector targeted by the NSSF during the pilot, owing to the organized nature of these groups and their rel- atively stable income source. Since its inception in 2019, 538,373 individuals are estimated to have registered in Haba  Haba, with 102,405 (19  percent) having active accounts—that is, they have made at least one contribution.4 The high inactivity rate—close to 80  percent of registered individuals—suggests a need to carefully evaluate processes beyond initial outreach. Ensuring the persistence of contributions matters both in terms of the adequacy of 3  These individuals are identified in survey data as those who have been faced with a shock in the last few years and were able to cope without recourse to any mal- adaptive strategies such as reducing food consumption or taking their children out of school. 4  Figures are from the NSSF and cover the period November 2019–June 2023. 2 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s individual benefits and in making the scheme viable over the long run by reducing operational costs. Estimates suggest that the administrative costs of onboarding and outreach at the take-up stage are about $7 per individual, with average annual contributions estimated at K Sh 9,000 ($61.35) (Guven, Jain, and Joubert 2021). Unless more individuals contribute regularly to the The delivery chain analysis scheme, there is a risk that administrative costs as a percentage of total con- carried out for this report tributions collected will remain high.5 identified four areas for investment in Haba Haba in The NSSF would like to focus on improving the coverage and persistence order to create a scheme that of savings in the Haba Haba scheme. However, limited technical and orga- meets the needs of informal nizational capacity, along with strained resources, limit the NSSF’s ability to sector workers and builds trust provide a seamless customer experience, track progress, and pilot innova- among existing and prospective tive strategies that could work for the Kenyan context. The delivery chain members. analysis carried out for this report, using a mix of consultations and review of operating procedures, identified four areas where investments are rec- ommended: (1) refining policy design, (2)  designing effective outreach and communications strategies, (3) updating business processes and administra- tion, and (4) strengthening institutional capacity. The overall impact of these investments would be to create a scheme that meets the needs of informal sector workers and builds trust among existing and prospective members of Haba Haba. This report includes data covering the 2019–22 period. More recent data are likely available, but the report’s aim is not to provide the most recent numbers on Haba Haba, which is a moving target. Instead, the intent is to comment on trends and highlight success factors and recommendations to improve the scheme, based on the delivery chain assessment, reimagined for social insurance. The report also provides suggestions on possible sequenc- ing of business process changes at the NSSF so it can prepare to relaunch Haba Haba under the World Bank–financed National Youth Opportunities Towards Advancement Project (NYOTA) in Kenya. 5  Data for the mandatory scheme and the Haba Haba program are collected con- jointly and cannot be disaggregated; therefore, the number of contributors, average contributions, and administrative costs for each scheme cannot be tracked in real time. All data cited in this report are thus best estimates provided by the NSSF team, based on the assumption that all those who do not have an employer contribution tagged to them are part of the Haba Haba scheme. The drawback to this approach is that individuals who are self-employed or formal sector individuals who are willing to make additional payments so as to increase the adequacy of their benefits from the mandatory scheme may be incorrectly counted as participating in Haba Haba. 3 Delivery Chain Assessment T he objective of the delivery chain assessment informing this report was to analyze the different stages of the delivery chain (from out- reach to exit) of the Haba Haba scheme from the perspective of current and prospective members. Such an assessment enables policy makers, practitioners, and stakeholders to get a bird’s eye view of the operating environment of the scheme and identify implementation gaps and operational inefficiencies. The assessment used a framework devel- oped by Lindert et al. (2020) in The Sourcebook on the Foundations of Social Protection Delivery Systems but reimagined for social insurance. The recom- mendations to improve the Haba Haba scheme, presented in the last section of this report, are grounded in findings from this assessment and relevant international experience gained through World Bank support in expanding pension coverage through voluntary schemes in Africa and Asia (Guven and The delivery chain assessment Jain 2023; Guven, Jain, and Joubert 2021; Jain and Palacios 2022). analyzed the stages of the delivery chain—from outreach Delivery systems constitute the operating environment for implementing to exit—of the Haba Haba social protection benefits and services. In recent years, delivery systems have scheme from the perspective been improved by the rapid development and deployment of three types of of current and prospective tools: (1) identification systems, (2) social registries, and (3) payment systems. members to give policy makers, All three work beyond the social protection system, serving broader functions practitioners, and stakeholders a for governments. Each has proven critical in improving the quality of delivery bird’s eye view of the scheme’s within social protection programs and systems. In many countries, social pro- operating environment so as to tection programs are not only the biggest users of these systems but have identify implementation gaps driven innovation and helped ensure widespread adoption (World Bank 2022). and operational inefficiencies. The way countries deliver social protection has implications for the out- comes of a social protection program. Social protection delivery system assessments address crucial questions such as how various elements of deliv- ery systems come together to implement programs as they were intended to function, how they ensure easy access to programs, how delivery systems can be leveraged to promote better coordination and integration, and how social protection programs meet the needs of their intended populations and pro- vide a better client experience (Guven and Jain 2023). Lindert et al. (2020) have developed a delivery system framework around the delivery chain com- prised of four stages: assess, enroll, provide, and manage. This delivery chain can be reimagined and applied to evaluate the struc- ture and functionality of social insurance (health and pension) schemes. 4 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s Using the framework developed by Lindert et al., the delivery chain for social insurance (contributory) programs has been reimagined as shown in figure 1. Depending on the nature of the scheme, the stages and substages may be parallel processes as opposed to consecutive steps. This is increasingly the case as systems become more advanced and connected—for example, reg- istration and verification occur simultaneously if the scheme’s registration database is interoperable with the national ID database. F i g u r e 1 : Reimagined delivery chain for social insurance ASSESS ENROLL CONTRIBUTE PROVIDE MANAGE RECURRING CYCLE RECURRING CYCLE Outreach Registration Verification Onboarding Contribution Benefit Eligibility Withdrawal Benefit Payout Monitoring Exit S o u r c e : Adapted from Lindert et al. 2020. In contrast to noncontributory social assistance programs, a social insurance scheme has the distinct function of contribution collection in addition to the delivery of benefits. Because of this characteristic of social insurance schemes, analyzing their delivery systems requires looking into both government-to-person (G2P) payments and person-to-government Contributory social insurance (P2G) payments. Contributory social insurance schemes must also manage schemes must manage members’ contributions efficiently over long periods. Throughout the contri- members’ contributions bution collection substage, the social insurance institution must maintain efficiently over long periods accurate records and make them easily accessible to members. In the payout substage, the benefits need to be calculated based on the scheme’s rules and provided to the rightful beneficiary when payout is due. Increasingly, social insurance schemes receive subsidies or matching con- tributions from the government to boost take-up or ensure coverage of the poorest. Due to limited fiscal resources, the incentives are often targeted and provided to individuals based on defined eligibility criteria.6 Consequently, assessing the eligibility criteria, crediting the incentive to the individual in a timely fashion, and ensuring dynamism and accuracy throughout the entire process are crucial to the customer experience and in building trust among informal sector workers, while simultaneously optimizing administrative 6  Examples of eligibility criteria are being in the poor or near-poor category (this cri- terion is used in Rwanda’s and Vietnam’s pension schemes), working in the informal sector, and meeting a minimum level of saving (this criterion is used in India’s Pradhan Mantri Shram Yogi Maan-dhan program). 5 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s costs. Gaps in the delivery chain—either through poor operational design, poor communication, or both—can hinder uptake and consistent engage- ment in voluntary schemes (Guven and Jain 2023). Gaps in the delivery chain— The delivery chain assessment was conducted through desk research, whether caused by poor interviews with operational staff—including in information technology operational design, poor (IT), communications, collections, and claims—and targeted field visits communication, or both—can where limited documentation was available. The assessment commenced hinder uptake and persistence in with a review of operational documents provided by the NSSF; these pro- voluntary schemes. vided information on business processes as they are set out (de  jure). To complement the document review, a questionnaire was circulated among those involved in the day-to-day operations of Haba  Haba to understand the de facto processes. A detailed description of the assessment methodol- ogy is provided in annex A. Based on the responses, the Haba Haba delivery chain was established; it comprises 5 stages and 10 substages, as outlined in figure 2. The remainder of this section presents the findings of the delivery F i g u r e 2 : Haba Haba social insurance delivery chain SOCIAL INSURANCE SERVICE DELIVERY SYSTEM – NSSF Haba Haba Kenya ASSESS ENROLL CONTRIBUTE (RECURRING CYCLE) PROVIDE (RECURRING CYCLE) MANAGE Channels Self or assisted by agent View information Contribution channels Based on Based on After retirement (age Support Informal groups at Registrant market Savings & not defined due to in transport validation for Log into profile Mobile money segment age informal nature of Greivences (digital taxi Branch Call NSSF Number sector Kenyan Nationals M-Pesa Boda Bodas and Matatu operators, boda Text drivers: NHIF, NSSF, shared Paid Lump Sum on Compliance boda, matatu), market vendors, Digital Platforms Via member number membership fees to Before retirement (age not Retirement until Using Can change/update and athletes through someone association by MPESA, defined due to informal else general accident cover, life nature of sector Act NO. 45 OF 2013 Data Update Radio National ID Phone number insurance, loan facility, Saving > 5 years is operational announcements In-person split transaction with Using None yet (sponsored by NSSF Number To National Hospital Claim after demise of Claim Application Four-digit pin 50% Partial withdrawals Safaricom PLC) (for existing members) Insurance Fund Primary member USSD No. *303# Set up contribution Bank Account Real time confirmation Market vendors: Last To person nominated Feedback Announcements amount Medical emergencies by primary member at meetings, Receipts expense and NSSF Using NSSF Mobile App SMS WhatsApp bot conferences and and app For online Digital Taxi Operators: With Doctor’s note and After identification Call center gatherings of View contribution and set contributions NSSF, NHIF, Data Bundles Medical Board clearance process sector leaders MPESA Mini App WhatsApp message Monitored by reminder notifications Reminders for Last expense members: Option to validate Paintings on WhatsApp Chatbot NSSF app missed contributions life insurance, loan facility, and admit member Live and accessible water tanks funding details dashboard Outreach Registration Verification Onboarding Contribution Benefit Eligibility Withdrawal Benefit Payout Monitoring Exit S o u r c e : NSSF responses to the questionnaire. chain assessment. Stage 1: Assess The first stage of the delivery chain is the assess stage, which has two sub- stages—outreach and registration. The customer’s first interaction with a scheme is at the assess stage, and it is advisable for schemes to focus on behavioral nudges that address limited attention during outreach and the complexity of information during registration. 6 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s Outreach Outreach is vital for the success of voluntary schemes because, unlike for mandatory schemes, members have to be convinced of the benefits to want to participate. Successful outreach entails raising awareness, building trust, and being available—in person, digitally, and/or through partners—to answer questions. Through such support, informal sector workers can make informed decisions about participating in a voluntary pension scheme like Haba Haba. At this substage, potential participants would like to know the design of the scheme, how it meets their needs, why they should trust it, their options for participation, and who they can call if they have questions, among similar considerations. Transparently communicating these details helps build trust in the scheme. The outreach strategy for Haba  Haba has focused on organized groups within the informal sector with a presence in peri-urban areas. In its pilot, Haba  Haba targeted groups in transport services (digital taxi operators, boda bodas, and matatu drivers and conductors) and market vendors. These groups were selected because they had robust organizations with strong peer-to-peer relations and sizable membership, a predominantly urban foot- print, and a high degree of income predictability across their membership, During the assess stage, making them more likely to save in a scheme like Haba Haba. The Haba Haba schemes should focus on team conducted outreach sessions in peri-urban (e.g., Nakuru and Naivasha) behavioral nudges that address limited attention during and urban areas using social media (Facebook), radio announcements (spon- outreach and the complexity of sored by Safaricom), talk shows, and in-person gatherings of informal sector information during registration. groups. The marketing team also made investments to generate awareness by establishing a visual presence; this included painting information on water tanks (for Wakulima market in Nakuru), affixing advertisements on NSSF vehicles (figure 3), and posting on billboards (figure 4). An unconventional target group chosen by Haba  Haba administrators was athletes. Athletes in Kenya form a minority of the workforce in informal employment, but they have short active careers, and those who are success- ful have high disposable incomes at the early stages of their careers. This prompted Haba Haba to focus on athletes, as they are likely to retire from their primary career early and will need productive and long-term invest- ments of their disposable income to have income security in old age. There has not been a gender-specific or gender-sensitive outreach strat- egy used by Haba Haba to date. This will be an important area on which to focus in the future, notably because women are more likely to be in the infor- mal sector, tend to be financially excluded, and can have different goals or communication preferences that need to be understood and better catered to (Guven, Jain, and Joubert 2021). 7 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s F i g u r e 3 : Outreach: advertisements on vehicles F i g u r e 4 : Outreach: public billboards While there has not been an evaluation to determine which outreach strategies have been most effective in increasing take-up, consultations with the NSSF revealed that in-person conversations are the most prom- ising. In-person outreach tends to occur in collaboration with the leadership of informal associations such as the boda boda secretariat or Jua Kali groups. The group leader mobilizes the associations or group members. The NSSF uses the allotted time in the meeting to provide information on Haba Haba, followed by hands-on support for registration using the USSD code. During the COVID‑19 pandemic, such engagements also involved distributing food, masks, and maize flour to encourage attendance and participation. The NSSF has, on the recommendation of group leaders, also borne the costs of wages forgone for those attending the meetings. Two key challenges to this approach were (1)  the high costs of organizing in-person events and 8 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s (2)  drops in consistent engagement—if, for example, a group leader was not well versed in scheme rules and an NSSF representative was not read- ily available to answer follow-up questions or troubleshoot. Broad-based Without broad-based efforts efforts beyond in-person outreach did not seem to translate into immediate toward visibility and awareness take-up; however, without these efforts toward visibility and awareness rais- raising, there is a risk of ing at the nascent stage of the scheme, there is a risk of Haba Haba being voluntary pension savings “out of sight, out of mind.” schemes being “out of sight, out of mind.” A  common finding across all outreach modes is that communications related to the Haba Haba scheme is not sufficiently differentiated from the formal sector scheme. As an example, in the NSSF’s mandatory scheme, workers are allowed to top up their mandated savings with individual volun- tary contributions.7 This top-up provision does not have a name, but, as with Haba  Haba, the individual alone is responsible for making those contribu- tions. In this regard, it is unclear if NSSF billboards (figure 4) are promoting Haba Haba or the scheme for formal sector workers. Issues of message clarity extend to other materials produced by the NSSF (figure 5); here, the use of the term “voluntary contributors” refers to voluntary self-employed registrants of the formal sector but could be misinterpreted as meaning participants of the voluntary Haba Haba scheme. The scheme rules for Haba Haba—notably the ability to withdraw 50 percent of savings after five years—is not available for those in the voluntary top-up for the formal F i g u r e 5 : Common NSSF materials (user guides, web pages) that lack any reference to Haba Haba S o u r c e : NSSF. 7  See NSSF webpage, “Register as a Voluntary NSSF member.” 9 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s sector. A clear demarcation between the schemes is important for clarity in policy and communications. Given adequacy concerns with the NSSF mandatory scheme, it makes sense to nudge individuals to save more for their future; the challenge is that the NSSF, as alluded to above, currently does not differentiate the marketing of Haba Haba with that of the options for formal sector workers. Rough estimates by the NSSF team during consultations suggest that, on average, they spend K Sh 1,000 per person (about $7) for outreach, includ- ing attracting and ultimately registering potential members. In the absence of segregated records between the mandatory scheme and Haba Haba, it is difficult to identify the expenditure on communications and marketing activi- ties of Haba Haba alone. According to NSSF team estimates, about $417,000 has been spent on Haba  Haba outreach to date across various channels, including mobilization for member education and sponsorships. These esti- mates should be interpreted with caution, and changes are needed to ensure clear and accurate recordkeeping of costs attributable by scheme. The large-scale outreach efforts for Haba Haba are the responsibility of NSSF headquarters staff, and the responsibility to follow up with identi- fied target groups is that of the NSSF branches. There are 59 branches of For the mandatory scheme, the NSSF, with one in each county in Kenya. These branches have targets, branches can follow up with performance measurement, and monitoring frameworks, including monthly registered employers; however, reviews and feedback, but their focus to date has been on the scheme for they do not yet have a method formal sector workers. The branches also have compliance departments to follow up with the relatively experienced in following up with employers who might not be paying their unorganized and geographically dues on time, are underreporting employee salaries, etc. For the mandatory mobile Haba Haba scheme scheme, branches can follow up with registered employers; however, they do members. not yet have a method to follow up with the relatively unorganized and geo- graphically mobile Haba Haba scheme members. In consultations with two NSSF branches, the branch managers said they lacked clarity on their role in promoting the Haba  Haba scheme. There is no evidence of relationship building at the branch level between the NSSF and informal sector associa- tions. The performance management framework of the branches also does not include any performance metrics or targets associated with Haba Haba. The NSSF has a call center with five members but no dedicated number or personnel for Haba  Haba. A common helpline (800-221-2744) is used for all NSSF-related queries, including the Haba  Haba scheme. More important, while the call center staff is well versed with the rules for the man- datory scheme, there is no operational manual or clarity on regulations for Haba Haba (as per an interview with the team). According to a call center staff member, when faced with a question on Haba  Haba staff are often unsure of the correct answer and are forced to keep the member on hold 10 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s while they “redirect the call to a select few in [the] NSSF who know the rules.”8 In most instances, they need to request permission to call back with accu- rate information, at great dissatisfaction to the caller. The call center team does not have a customer relationship management (CRM) system, in the absence of which there is no tracking of customer interactions, calls related to Haba  Haba, or a record of solved/unresolved queries. Furthermore, the limited IT, technical, and human resource capacity of the call center means that staff are unable to host short in-house surveys to gather feedback from callers or members (something routinely done by the Ejo Heza call center in Rwanda). Consultations with call center, branch, and headquarters com- munications staff reveal that common questions posed during the outreach substage are “Can I partially withdraw the benefits?,” “What is the turnaround time for withdrawal of benefits?,” “What is the Interest rate I will earn?,” and “Will my contribution be invested safely?” These questions indicate that flexi- bility, returns, and trust are important topics for prospective members. Considering the above, strengthening the call center to effectively sup- port customer engagement beyond outreach (when participants are encouraged to register) is critical to improve active participation in the Haba Haba scheme. The need for such improvements is reinforced by the number of inactive participants (approximately 80 percent of current mem- bership), as well as the growing complexities of scheme interactions (due to the various bundles offered in partnership with external associations as outlined below), which would increase the number of inquiries from custom- ers and require improved capacity of call center staff to effectively address them. Ultimately, the scheme must learn from its experiences and strike a balance between customization of bundles to attract some groups and efforts to streamline processes so that costs and customer experience can be standardized. The outreach strategy needs to identify stakeholders with which to partner and be clear about roles and expectations from the partnership. Outreach and communication is not a one-time effort, but an ongoing pro- cess to ensure retention. A two-pronged approach to partnering is suggested: l Partnerships at the institutional or aggregator level (with associa- tions, banks, etc.) to ensure scalability and cost-effectiveness l Grassroots mobilization to get buy-in from informal sector workers. 8  Haba  Haba’s strategy of tailoring products for target groups—such as a bundle for digital taxi drivers to meet their needs—also can create confusion if, for example, someone from the group contacts the call center and the team cannot correctly iden- tify the membership group and Haba Haba package to which the person is entitled. 11 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s The outreach strategy needs to involve the government at different levels (national, provincial, district, and county) to increase ownership and, if pos- sible, should assign targets at each level; this was successfully done in China, Rwanda, and Vietnam (Jain and Palacios 2022). The choice of partners can be determined based on the value proposition of the collaboration. For exam- ple, partnerships with telecommunication service providers and banks could be explored to allow for auto deduction of contributions; in exchange for their Partnerships with efforts and support, these partners could receive a flat fee or a percentage telecommunication service of the funds collected. In India’s Atal Pension Yojana voluntary scheme, the providers and banks could scheme administrator—the Provident Fund Regulatory and Development enable auto deduction of contributions for either a flat Authority (PFRDA) uses a commission structure for banks to target informal fee or a percentage of the funds sector workers. Banks receive a minimum commission of Rs 100 ($1.21) on each collected. application and promotional incentives that increase with the number of sub- scriptions: Rs 20 ($0.24) for the first 100,000 applications, Rs 30 ($0.36) for 100,00–300,000 applications, Rs 40 ($0.48) for 300,000–500,000 applica- tions, and Rs 50 ($0.60) for more than 500,000 applications).9 Banks can also appoint mutual fund distributors to market this scheme, expanding the reach of the distribution chain. In addition to informal sector associations, the NSSF has partnered with private sector stakeholders to expand its outreach. Key partners in the pri- vate sector include Equity Bank, which brings vast experience in the informal sector; and Vooma, a mobile money service, to increase coverage and imple- ment the auto debit function to enhance compliance levels. The NSSF has partnered with telecommunication companies such as Safaricom to expand outreach and design attractive product bundles. Such initiatives have been predominantly targeted toward informal sector groups with a more urban presence, such as taxi driver associations, boda bodas, etc., and could poten- tially be expanded to rural areas based on the nature of the informal sector target groups in such areas. It will be important for the NSSF to be cognizant of its geographical focus when setting targets and developing the outreach strategy. The current focus has been on urban and semi-urban areas based on their perceived higher ability to save for the long run, and their greater financial/digital awareness relative to rural individuals. However, the experience of the Ejo Heza Long-Term Savings Scheme in Rwanda suggests that expanding such schemes to rural areas is not only possible but also has some distinct advan- tages (Guven and Jain 2023). For example, the presence of cooperatives in rural areas and high trust in and familiarity with local leaders in these con- texts gave Ejo Heza a reliable grassroots partner—which, once convinced of benefits of the scheme, participated at scale and continued to save. An All currency conversions of Indian rupees to U.S. dollars reflect the rate in effect on 9  September 13, 2023 ($1 = Rs 82.96). 12 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s organized group such as a cooperative with a strong peer network can be hard to find in urban contexts B o x 1 : Lessons learned from outreach efforts where individuals tend to be more mobile. during the Haba Haba pilot While it is reasonable to focus on a specific location The NSSF’s target for the pilot was to onboard 1.2 million or worker profile at the pilot stage, national scale-up people in Haba Haba from 2019 to 2022. Registrations have would entail revisiting the outreach and mobilization fallen short by a significant margin, but there are lessons to be strategy based on the characteristics of individuals learned from the outreach efforts: in a particular geographic category. l The targeting strategy of identifying organized groups and conducting outreach via group leads was effective. But Box 1 summarizes lessons learned from outreach this approach can be expensive and, without sustained fol- experiences during the Haba Haba pilot. low-up, can quickly turn into a sunk cost. l Poor planning can derail good ideas. While the target- ing strategy was well designed, care was not given to Registration strengthen the IT and call center teams, which limited the NSSF’s ability to process data and gain insights from data, The information provided during outreach is the set clear goals for branches, and reward high performers. foundation for intake and registration. Easy-to-use l Sensitization of informal sector workers to build a savings cul- registration channels can simplify the process for ture is key. Such efforts should be augmented with appropriate the intended population. A mix of digital processes incentives, behavioral nudges, and tailored communications and local networks can ensure a seamless and posi- material—which was inadequate during the pilot. tive experience for a new member. l Informal sector workers are geographically mobile, so exploring digital communications channels and Registration can be assisted, self-registration, or partnerships that transcend boundaries—such as tele- a hybrid approach. Self-registration with assistance communication service providers, savings and credit could make individuals more committed to contrib- cooperatives, and banks—can be useful. uting to the scheme, as they would be empowered l Political will and government engagement are needed to to navigate the scheme rules and become familiar sustain outreach and build long-term partnerships. with the process. Common questions at the regis- tration substage include where to register, how to access and complete paperwork, how to register digitally, how to submit paperwork or complete a digital transaction, and how to set up notifications. In a voluntary model like Haba Haba, members can join the scheme at any time and contribute at their convenience. They are not required to register on any particular date; group registrations, when they occur, are by the consent of the individual and are not an auto registration process—as in mandatory schemes where contributions are deducted from an employee’s salary by the employer. When applying to a new social insurance scheme, the applicant’s infor- mation can be collected from scratch or be based on a few mandatory points requested during registration, with the remaining details retrieved later from other administrative systems. The latter approach can expe- dite the registration process while ensuring that useful administrative data on individuals are gathered. This approach requires interoperability across 13 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s data systems. Such data exchange requests are typically performed using application programming interfaces to enable real-time data exchange. The reason for collecting these data is not to compile information on a person but rather for targeting benefits. For example, in Rwanda, the Ejo Heza scheme database is interoperable with the social registry managed by the Ministry of Local Affairs. This allows the government to offer targeted fiscal incentives (with higher matching contributions to poorer households) to encourage saving among the population. The systems should not only be aligned but should also be able to communicate with each other in real time. Thus, any changes to the administrative system database—particularly those relat- ing to socioeconomic classification, address, and phone numbers—should be reflected in the social insurance system. The ease of the registration process (and subsequent interactions) and choices of how to register play a role in a scheme’s perceived attractive- ness. Clear messaging and communications on the registration process is key in this regard. Currently, eligible individuals in Kenya (those with a valid ID who are not mandated to contribute to the formal sector scheme) can register in Haba Haba by walking into an NSSF branch, by dialing *303# on their feature phone or NSSF mobile app, or by using the M-Pesa mini-app or WhatsApp chatbot. Since the records for Haba Haba and the NSSF formal sector scheme are not disaggregated, it is not possible to identify the mode of registration most used by members. Going forward, collecting this kind of information and mining the data can allow the NSSF to better understand International evidence suggests the preferences of its target customers. The information required to regis- that contributor targets, when ter is simple, with only three mandatory fields (ID number,10 full name as per coordinated between the national ID, and phone number), and two optional fields (NSSF number for scheme administrator and local government, can improve previously enrolled members; member photo/fingerprints/sex). On average, ownership and accountability, the registration process takes about three to five minutes. and motivate local agents to innovate. The Haba Haba scheme does not yet have registration targets set by the NSSF for its branches or by the government for local offices. International evidence suggests that contributor targets, when coordinated between the scheme administrator and local government, can improve ownership and accountability and motivate local agents to innovate. Targets need to be set objectively, mindful of varying administrative capacity by district and vary- ing ability to save among informal sector workers. NSSF branch officers can help individuals install and use the NSSF mobile app, but the geographically mobile nature of informal sector workers means that branch officers need to be proactive in providing registration support. This mobility on the part of the informal sector could make it difficult to accurately reward branches Passports can also be used to register, although national IDs are preferred. 10  The uptake of national ID is near universal in Kenya, as IDs are needed to obtain a phone line, access buildings for services, etc. 14 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s that provide good support. One solution could be to request the location of registration by the individual during the registration substage and use that variable to calculate if a branch met its monthly targets. Streamlining the registration process, supported by enhancements to the NSSF’s IT systems, strengthening interoperability with foundational systems, interfacing with external partners, and a clear plan of action for NSSF branches to register applicants will all be needed to unlock opportunities to scale Haba Haba. Stage 2: Enroll The enroll stage includes two substages—verification and onboarding. At this stage, the individual has decided to join the scheme, but experience at A smooth enrollment process in line with member expectations the verification and onboarding substages will make an impression on the will enhance the customer member. A smooth enrollment process in line with member expectations will experience. enhance the customer experience. Clearly outlined steps conveyed through effective messaging can build trust and provide answers to frequently asked questions at the enrollment stage. Ve r i f i c a t i o n Once the applicant’s identification information is collected at registra- tion, it is verified against data in other existing administrative systems. In countries with high ID coverage, the applicant’s data collected at registra- tion can be validated using an existing de-duplicated national ID database. Any additional personal information such as date of birth, sex, location, and phone number may, if stored in the national database, be requested for cross-verification or appended to the social insurance records of that individ- ual. Data sharing arrangements and privacy controls need to be implemented to enable information sharing across the different government entities (ID system and social security fund). Once registered, members in Haba  Haba can be verified in real time through the Integrated Population Registration System (IPRS). The IPRS is interoperable with the NSSF’s administrative system. This interopera- bility allows for verification using the mandatory registration data points: the registered member’s full name, date of birth, and national ID number. Once this information is verified, applicants receive an enrollment confirma- tion via SMS (short message service) delivered on USSD or the NSSF mobile app, depending on the registration channel. Kenya’s strong foundational ID system—with over 91  percent coverage of individuals above the age of 18 (Clark, Metz, and Casher 2022)—and interoperability with the NSSF makes for a concrete example of how systems integration can make it simple, secure, 15 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s and cost-effective to provide a long-term saving scheme to informal sector workers. Systems integration can make Only one NSSF number is issued to each person. If an individual had an ID it simple, secure, and cost- number with the NSSF for contributions made to the mandatory scheme effective to provide a long-term earlier in his/her work history but now is in the informal sector and wants to saving scheme to informal save using Haba Haba, the same NSSF ID is retained. The IT system is not sector workers. sufficiently agile to track an individual’s moving in and out of schemes unless the data are specifically analyzed to determine if an employer contribution was tagged to that individual (in which case, the contribution comes from the mandatory scheme). To simplify the registration process and provide a seamless experience, the NSSF is exploring partnerships with the private sector including the use of auto enrollment features. The NSSF can benefit from partnerships with telecommunication service providers, banks, and microfinance institutions that follow know-your-customer (KYC) norms during customer registration and would have the ability to implement an auto debit saving option. These partners can facilitate sustained contributions to Haba Haba once consent from a prospective member is obtained. Such partnerships will require an interface between the NSSF’s and partner IT platforms; this can be tech- nically challenging and potentially expensive, with the need to carefully vet data to address reliability and data privacy concerns. Notwithstanding these caveats, application programming interfaces make exploring such partner- ships increasingly viable. Onboarding Onboarding is the last substage of the enrollment process, when the appli- cant confirms profile details. At this point, the applicant can view his/her verified personal information and data imported from other administrative systems. Applicants may be required to provide some personal information that is not available from existing systems. They can also complete optional fields such as desired contribution amount, preferred payment method, and preferred communications channel. In voluntary schemes, there is no pen- alty if an individual does not comply with his/her self-selected contribution frequency and amount; however, behavioral nudges, such as sending remind- ers to registered members on their chosen frequency or contribution history, could serve as a self-commitment strategy. Onboarding in Haba  Haba is completed once a notification confirm- ing the registration is sent to the member via email, text, or letter. Once onboarded, members can access and view their information by logging onto their profile on the web, by dialing the NSSF number, or by text via the USSD 16 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s interface (figure 6). Members can update their phone number and four-digit pin by dialing the USSD number (*303#).11 They can set their contribution amount (which they can change at any time) and access membership infor- mation via the WhatsApp chatbot. However, the existing USSD interface is F i g u r e 6 : Existing not very user friendly. As can be seen in figure 6, for example, it is not intui- USSD interface for customers tive to understand which option should be selected if an individual wants to access his/her contributions. A redesign of the user interface with intuitive options can help participants navigate the services; greater convenience is likely to improve trust and interest in the scheme. While contribution to Haba  Haba is voluntary, members can only with- draw 50 percent of funds after five years if they have contributed a yearly minimum amount of K Sh 4,800 (about $33). The frequency and amount of contributions during the year can vary, since the scheme is voluntary. However, members need to meet the minimum daily amount of K  Sh 25 ($0.17), the monthly amount of K Sh 400 ($2.73), or the annual amount of K Sh 4,800 to be eligible for partial early withdrawals under Haba Haba. Tools like pension calculators can help users make decisions about their con- tribution levels. These aids can be made even more tangible by specifying what the expected fund balance could buy in today’s terms: clarifying, for example, that the savings accumulated from a $100 contribution each year for 30 years could pay for the average two-story apartment in Kenya or for all five years of a son’s or daughter’s engineering degree. Incorporating such tools and behavioral nudges at the point when individuals are choosing their starting contributions can have an anchoring effect, potentially affecting the level of subsequent contributions and consistent participation (Gough and Niza 2011). Stage 3: Contribute The contribute stage includes two substages—contribution and benefits eli- gibility. At this stage, the individual has made one or more contributions to the scheme. A lack of understanding of the scheme’s benefits or difficulty in joining the scheme can dissuade registered members from contributing. It is important to include incentives in product design to attract workers and ensure they are valued and understood by the target group of the informal sector. Communication at this stage should focus on establishing a presence through multiple touch points—branch offices, brochures, call centers, radio, etc.—so individuals are reminded of the scheme and its benefits, know who The pin is needed to access the account every time on the web and via the M-Pesa 11  and USSD platforms. 17 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s to contact in case of questions, and feel safe in the knowledge that their sav- ings are secure and growing. Contribution Registered members are considered active savers if they have made at least one contribution to the scheme. Contributions can typically be made through multiple channels, but members need to be aware of those chan- nels, which should be convenient and cost-effective to promote high take-up. First-time contributors may not have known about available channels of contribution and any associated costs; some hand-holding on the part of the agent network, including staff from the Haba Haba branches or a help- line, would be useful. In mandatory systems, employers typically transfer benefits (employee and employer contributions) directly to the social security fund by deducting them from the individual’s salary. For a voluntary individ- ual pension scheme, the onus is on the individual to initiate (and continue) the contribution. Some individuals are marginalized and may have limited access to formal financial services, limiting viable contribution channels. However, mobile money and new forms of online payments are increasingly being used, and voluntary schemes should align with these initiatives and offer contribution modes that are commonly used by the informal sector in the country context. Auto enrollment and auto debit—heralded as a game changer in improving Auto enrollment and auto compliance in mandatory schemes (e.g., the UK’s Nest defined contribution debit—a game changer in scheme)—are not easily replicable in voluntary schemes unless an aggre- improving compliance in gator is used to collect contributions, consent is sought, and checks and mandatory schemes—are not easily replicable in voluntary balances are put in place to address grievances. Rwanda’s Ejo Heza scheme schemes, but can be instated uses cooperatives as an aggregator. Once a cooperative collectively decides with some advance planning. to join the scheme, the cooperative leaders deduct the Ejo Heza contribution selected by each member and transfer it to their Ejo Heza on their behalf (Guven and Jain 2023). Members can contribute to the Haba Haba scheme using mobile money, such as M-Pesa and other mobile wallets, or in cash at NSSF offices. For mobile money contributions, customers must have a wallet pin set at regis- tration and the NSSF number linked to their phone. It is also possible to pay on behalf of other members by verifying their NSSF number.12 Figure 7 shows the Haba  Haba user interface as presented in the operations manual for payments made via the mobile app. Participants can also set and update their contribution amounts using the app, which allows for flexibility in their 12  If members forget their NSSF number, they can request it or a new one. 18 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s F i g u r e 7 : User interface on mobile application for payments S o u r c e : Haba Haba Operations Manual. contributions (figure 8). During the field visits, it was noticed that the user interface did not match the simplicity of the operations manual, making this an area of improvement to be considered in upgrading the mobile app. The process differs when paying in person. When paying at the branch, cus- tomers do not need a wallet pin but only their NSSF number. Paper receipts are not issued for in-person contributions, which can make it difficult for mem- bers who are not digitally savvy to track their contributions. Text reminders are sent quarterly for missing contributions. The NSSF estimates Haba Haba coverage to be at 518,290 as of the writing of this report. Without being able to analyze Haba Haba data separately from formal sector records, it is dif- ficult to comment on who contributes and how often, or to attribute the coverage increase to any specific NSSF strategy. It is likely that the simple USSD option of using *303# to join and contribute to the scheme, along with partnerships with informal sector associations, has played a significant role in the increase in take-up. The persistence of contri- butions beyond the initial take-up remains to be seen and will likely need a good value proposition to retain members to save for the long term. Informal sector workers are a diverse group; some have the intent and ability to save over the long term, but moving from intent to action is a challenge given the unpredictable nature of their income. Commitment 19 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s F i g u r e 8 : User interface for selecting payment contribution amount on mobile application S o u r c e : Haba Haba Operations Manual. nudges in addition to flexibility of debit amounts can therefore, like auto debit, be a good strategy to help informal workers in saving for their future. The NSSF has begun discussions with banks on an auto debit function, allow- ing individuals who provide consent to save in the voluntary scheme and smooth their consumption over time through consistent savings over their life cycle. Contributions must be invested prudently to keep member savings safe and to achieve competitive returns. The collected contributions are transferred to a custodian account and then to the NSSF investment unit. Contributions are largely invested in treasury bills, infrastructure bonds, term deposits, money markets, real estate, and private equity investments. Investment returns are credited to individual accounts. The NSSF team reported that Haba  Haba scheme members receive nominal annual returns of 10  percent; this is 20 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s reportedly higher than mutual fund returns. Haba Haba does not yet have a dedicated investment policy, but the NSSF intends to segregate formal and informal sector investments in its upcoming strategic plan. For schemes that allow short-term withdrawals, like Haba Haba, the investment strategy needs to be cognizant of the rate of short-term withdrawals to ensure the scheme has sufficient liquidity to facilitate such withdrawals. Benefit/incentive eligibility In voluntary pension schemes, the benefits of saving for the long term and forgoing current consumption must meet the needs of the target popula- tion. Some schemes offer matching contributions paid by the government (e.g., in India, Rwanda, Thailand, and Vietnam). This match can be a flat ben- efit on registering, a percentage of an individual’s savings (with a cap), or a top-up available at the time of retirement (as in Colombia). The match- ing provided by the government can be targeted based on the amount of savings by the individual or his/her income category. Other incentives can also be useful in targeting informal sector workers. Insurance against short- term risks—such as life and funeral insurance, and health, accident, and auto Insurance against short-term insurance—can be attractive to workers in the informal sector who typically risks—such as life and funeral do not have access to such insurance in an affordable bundle. Workers in the insurance, and health, accident, informal sector often keep savings in cash at home, which erodes their real and auto insurance—can be attractive to workers in the value over time. Therefore, if a pension scheme offers consistent returns that informal sector. exceed inflation with the flexibility to access savings in the short run, it can be valuable for the target group. Member savings would grow over time, and be available to meet short-term needs such as housing and education. For example, in Rwanda’s Ejo Heza scheme, individuals can take out a housing loan once they have saved at least RF 4 million (approximately $3,300) in their account,13 and they can choose to withdraw up to 40 percent of savings in excess of RF 4 million (Guven and Jain 2023). The Kenyan government currently does not offer fiscal incentives to any voluntary pension scheme. In the absence of fiscal incentives to Haba Haba, the NSSF has used other incentives, such as the flexibility to withdraw 50 per- cent of savings after five years, and a bundled package to attract specific groups of informal sector workers. During the pilot, the NSSF worked with informal sector leaders to design packages for each group, but there were no registration targets set. The package for boda boda and matatu drivers had their members pay K Sh 60 ($0.41) a month—K Sh 18 ($0.12) for NHIF health insurance, K Sh 25 ($0.17) toward the Haba Haba scheme, K Sh 2 ($0.01) as All currency conversions of Rwandan francs to U.S. dollars reflect the rate in effect 13  on September 13, 2023 ($1 = RF 1,205.37). 21 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s split transaction charges, and the remaining K Sh 15 ($0.10) as membership fees. The split transaction software by Safaricom is used to transfer the total contribution to the NSSF and the NHIF, respectively, without any additional action from the member, thus making the single contribution go toward both pen- F i g u r e 9 : Advertising of specific bundles for target groups (sticker advertisements) sion and health insurance. The Digital Taxi Forum, which has approximately 11,000 members, received a bundle that, for K Sh 100 ($0.68), provides for daily insurance from the NHIF, accident insurance covered by KCB Bank, and a complementary mobile data bundle plan from Safaricom. The NSSF printed adver- tisements and pamphlets for this group (figure  9); consultations revealed, however, that few members continued to save. Among the reasons noted were that Safaricom soon after released a discounted mobile data bundle that made the group deal with Haba Haba less attractive to members. Reportedly, those who contributed were not promptly credited for the complementary mobile data bundle, and cus- tomer service was unable to help them; this made them apprehensive about contributing again. The Haba  Haba scheme customized packages for informal sector asso- ciations to attract their members, but poor follow-through, a lack of targets and an incentive structure, and no recourse for promptly address- ing grievances led to slow growth in coverage. Partner commitment has been challenging to sustain. Even when memorandums of understanding were signed with partners, the three-year duration lacked a clear roadmap for long- term engagement. The COVID-19 pandemic also limited the NSSF’s ability to have in-person meetings and information sessions. The NSSF call center team was not always aware of the tailored deals provided to each target group, which led to confusion. Tailoring customized bundles for each group was also time-consuming to execute and costly in terms of developing partnerships with a range of organizations. Each tailored bundle needed database syncing, which entails design and integration changes, and updating all the teams with relevant information. Discussions with teams at the NSSF highlighted that they were rethinking the sustainability of this tailored approach in terms of duration (including a phaseout plan), exit options, etc. The complexity of such bundles can also increase the administrative burden on the NSSF in maintain- ing partnerships and integrating platforms. Lessons from the Haba Haba pilot suggest that customization and flex- ibility have distinctive advantages, but need to be balanced against the consideration that benefits be easy to understand, recall, and manage. The NSSF partnered with the NHIF, Safaricom, and others to ensure a full 22 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s package of support to its members; the resulting bundles, however, entailed extensive logistical and administrative costs, as outlined above. Distinct bun- dles can be expensive from a time and resource perspective, especially when institutional policies are not aligned. Furthermore, a robust monitoring and evaluation framework is needed to assess the attractiveness of benefits and their impact on registration and contribution. Stage 4: Provide The provide stage includes two substages—withdrawal and benefit payout. At the withdrawal substage, the individual interacts with the scheme to eval- uate if he/she is eligible for premature withdrawals; at the benefit payout substage, the individual interacts with the scheme to receive a benefit (retire- ment or early withdrawal) payout. At the time of withdrawal or payout, the focus of the message should be on outlining the eligibility criteria and whether the individual meets them, the documents needed if eligible, the application process, and the expected wait time. As at other stages, submis- sion of a withdrawal or payout request should accommodate the needs of a diverse group of individuals, including minorities and females. Withdrawal In pension schemes that allow short-term benefits, members, after some years of contribution, can request a withdrawal or be entitled to a benefit. Voluntary schemes like Pension schemes typically do not allow 100 percent of funds to be withdrawn Haba Haba that are designed before retirement, as the primary goal of the scheme is to support members for informal sector workers are during their old age. Voluntary schemes like Haba Haba that are designed for more likely to offer flexible informal sector workers are more likely to offer flexible withdrawals given the withdrawals given the high high liquidity needs of the target group. Members can withdraw 50 percent liquidity needs of the target of their savings in Haba Haba after having consistently contributed for five group. years. Because the scheme was launched in 2019, the first cohort of members will be eligible to withdraw their savings in 2024, subject to meeting the spec- ified contribution requirements. Emergency withdrawals from the scheme were not allowed during COVID-19. In the case of medical emergencies, a doctor’s note and clearance from the medical board counts as eligibility for withdrawal—similar to the rules for the mandatory scheme. Benefit payout Pension schemes have retirement as the trigger for a benefit payout. The payout on reaching retirement can be a lump sum, a pension for life based 23 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s on a benefit formula, or a scheduled withdrawal—that is, savings provided over an agreed time period (e.g., a period equal to average life expectancy). Under current provisions, the NSSF does not pay any pensions, even to those in the formal sector. It is a provident fund, whereby benefits are paid as a lump sum and equal the contributions plus interest accrued over the years. In Haba Haba, there is no mandatory retirement age due to the informal nature of the workers. Members can access full withdrawal benefits from the age of 50, which is the same as in the mandatory scheme for the formal sector. Payouts at retirement are processed by the NSSF claims unit; figure 10 illus- trates the Haba Haba benefit claims process. Schemes that offer insurance for short-term benefits—like accident or life or health insurance—will have some members who become eligible for claims before reaching retirement age. Claims under insurance need to be paid out after verifying eligibility (e.g., death of member in the case of F i g u r e 1 0 : User interface for benefits claim through mobile application S o u r c e : Haba Haba Operations Manual. life insurance, childbirth in the case of maternity benefits). Managing claims and disbursing payouts quickly once a claim is verified will require systems to be interoperable and administrative arrangements to be in place. Thus, a scheme that offers multiple short-term benefits can be more expensive to handle from an administrative standpoint, unless the benefits are out- sourced to an insurance firm, which can spread out the costs as part of its portfolio. Offering voluntary schemes as collateral for housing is an incentive some schemes would like to provide; before implementing such incentives, the social security institution should have in place clear governance arrange- ments on these credit facilities. 24 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s Stage 5: Manage This last stage of the delivery chain includes two substages—monitoring and exit. At the monitoring substage, the scheme collects feedback on customer experience and grievances, and monitors changing attitudes/preferences of existing and prospective members. During the exit substage, the intention is to ensure that the scheme does not include any member who is no longer eligible. The exit substage helps ensure the integrity of the data and avoids errors and opportunities for corruption (e.g., moneys frozen in suspense accounts and/or payments made to ghost pensioners)—problems that can arise if accounts with a completed payout are not closed. Monitoring Monitoring contribution behavior and member experiences is important in understanding what works and what needs improvement. Members, Inclusive monitoring aimed prospective registrants, and the intended target market can hear about a at learning and at correcting scheme or institution from a variety of sources and can have multiple inter- any loopholes is important in actions with scheme administrators/partners, making such monitoring providing a positive experience essential. Members can face challenges at multiple points—registration, to members. accessing data, benefit disbursements, withdrawals, claims, etc.—and the challenges faced can differ by group (e.g., women versus men, urban versus rural, smartphone users versus feature phone users). Inclusive monitor- ing aimed at learning and correcting any loopholes is critical to providing a positive experience to members. Since the Haba Haba pilot, the NSSF has been deliberate in seeking feedback through education forums, social media platforms, and the call center; and progress is monitored through a dash- board. Once the IT systems have been revamped to enable disaggregation of records, the dashboard will allow real-time information on take-up and sav- ings behavior to be available for policy makers. Exit The exit substage looks at processes used in the scheme to determine when a beneficiary is no longer eligible for benefits. Reasons for program exit from a pension scheme include death, emergency withdrawal, and retirement. Typically, programs follow a four-step process when removing ben- eficiaries from the system: an exit trigger, criterion validation, exit decision, and beneficiary notification. Exit triggers are updates that result in changes to a beneficiary’s basic information stored in the system. Verification of the exit criteria is rarely documented in a program’s operations manual, but can 25 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s be a useful check-and-balance procedure to avoid someone’s being wrong- fully exited from the scheme. Schemes that offer funeral benefits should have information on the named beneficiary stored in the database. Also, a clear, accountable mech- anism needs to be in place to certify through official means (such as a death certificate) that the member has died. Since the NSSF has a long history of managing the mandatory scheme where survivors are eligible for lump sum benefits, it can verify such claims. To date, there have been no survivor claims under the Haba Haba scheme. Photo: Petr Kapuscinski / World Bank 26 Recommendations T he Haba  Haba delivery chain assessment has led to a more informed understanding of business processes, gaps in implemen- tation, and lessons learned from the pilot. In particular, three key challenges have emerged regarding con- sistent collection of contributions from the informal sector: B o x 2 : Features of voluntary schemes that make them attractive l Informal workers’ incomes are undetectable l Simplicity in design and execution can make it easier for and vary, making it administratively difficult people to understand the scheme and therefore increase to mandate contributions based on income. word-of-mouth referrals and improve trust. l Ensuring compliance with a mandated con- l Balancing short-term needs with long-term goals in tribution is difficult, as these workers do not scheme design can be a plus for informal sector workers. have a registered employer through which Workers who are not mandated to contribute to pensions administrators can track contributions. are more likely to have higher susceptibility to other short- term risks against which insurance is limited. Flexibility l Informal workers’ low and irregular income in withdrawals or bundling with other insurance can be means that the administrative cost of col- attractive for this group. lecting and managing their contributions can be high. l Monetary incentives offered by the government or a sponsor in the initial stages of the scheme can, at a mini- mum, help improve scheme uptake while awareness grows. Insights from the delivery chain assessment, com- bined with emerging international experience with l Partnering with aggregators that can influence poten- voluntary schemes (box  2), suggest that there are tial members and perform an auto deduction role as employers do in formal sector pension schemes can be a four key areas Haba  Haba should focus on as it game changer in scaling up coverage. The aggregator must gears up to relaunch after the pilot; these are sum- be in close contact with its members to observe and moti- marized in table 1: vate them to join. If the aggregator can also auto deduct contributions after securing member consent, they can l Refining the policy design indeed be effective partners in growing the scheme. l Designing effective outreach and communi- cations strategies l Updating business processes and administration l Strengthening institutional capacity. Area 1: Design considerations 1 . Improve the attractiveness of the Haba Haba scheme through incen- tives, behavioral nudges, and partnerships. 27 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s Ta b l e 1 : Summary of key recommendations for the Haba Haba voluntary pension scheme Area Recommendation 1. Improve Haba Haba attractiveness through incentives, behavioral nudges, and partnerships Area 1 2. Strike a balance between customization and scalability Design considerations 3. Improve system readiness to process short-term benefits and pension payout 4. Use targeted surveys to inform scheme improvements 1. Implement a customer relationship management system 2. Relaunch Haba Haba and create a dedicated team to work on Haba Haba within the NSSF 3. Develop a detailed FAQ manual and educate NSSF employees on the scheme Area 2 4. Adopt an inclusive communications strategy Outreach and communications 5. Update the operations manual to reflect the de facto and not the de jure 6. Develop pamphlets or digital calculators to show individuals what they can expect to receive 7. Balance outreach effectiveness with associated costs 1. Undertake a well-planned and coordinated IT modernization project 2. Digitalize the verification process for Haba Haba membership and improve interoperability across Area 3 stakeholder databases Updating business processes 3. Improve data analytics to support planning and decision making 4. Create a dedicated web page/app option for Haba Haba 1. Institute a clear role for NSSF branches in promoting Haba Haba Area 4 2. Strengthening monitoring, learning, and evaluation Institutional strengthening 3. Build human resource capacity and leverage external expertise 4. Develop strategic and sustainable partnerships l Provide monetary incentives for (at least) those most vulnerable. Contributors to the Haba Haba scheme are recommended to pay at least K Sh 25 ($0.17) per day so that a minimum of K Sh 400 ($2.73) per month can be achieved.14 The monthly target is reasonable, and anything less than the minimum will likely be inadequate savings for retirement. On the other hand, meeting the recommended minimum can be difficult in some months for informal sector workers, espe- cially the most vulnerable. Fiscal matching for vulnerable groups can be considered to incentivize their participation. Unless these groups Kenya’s NSSF law mandates that minimum savings in accounts managed by 14  the NSSF must be K Sh 400 per month; since there is currently no separate law for Haba Haba, its accounts technically fall under this purview. 28 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s have savings for their old age, they will likely depend on the state for old age allowances. The growing number of elderly in the country over time would result in higher future spending for the government. Thus, incentivizing workers to save to meet their short- and long-term needs is a sound fiscal policy. Almost all countries with voluntary pen- sion schemes offer some degree of matching (Jain and Palacios 2022) notably when the scheme is in its nascent stages. Before national rollout is considered, a pilot with monetary incentives to vulnerable groups can help policy makers evaluate the impact on take-up and savings levels. l Behavioral nudges include highlighting the safe and stable returns of the scheme. Informal sector workers who save in cash see their benefits eroded over time due to inflation. If Haba  Haba’s secure and relatively high returns can be conveyed clearly to informal sector workers through communications materials and in-person visits, it could help make an attractive case for the scheme. Stable returns on savings are the most sustainable incentive the NSSF can offer its participants—including through Haba  Haba—so it is important to highlight this aspect. An evidence-based approach, l Explore the potential for auto deduction in partnership with aggre- grounded in targeted surveys gators. The aggregator model adopted by the NSSF is a promising and focus group discussions, strategy, and partnering with organizations that can auto enroll appli- should be used to understand cants and/or auto deduct contributions can be a game changer. The the perceptions of the target NSSF’s ongoing conversations with mobile money operators and group and inform scheme banks that have strong networks in urban and peri-urban areas is a design. step in the right direction in this regard. l Nudge existing members to register spouses, children, and other family members. Existing members are likely convinced of the ben- efits of saving with the NSSF and are familiar with the processes. A marketing campaign to encourage them to spread the word or save for the retirement security of other family members can be launched. In India, the Gift a Pension initiative was launched to allow relatively better-off individuals to pay contributions on behalf of their domestic workers (Borate 2022). l Incorporate gender-sensitive approaches and gender-responsive solutions. The operational team should adopt a gender lens at each stage of the delivery chain. The question scheme architects should ask of grassroots mobilizers is whether their strategy to enroll and attract individuals to save is sensitive to the needs of women. To this end, the agents recruited for mobilization should include women, who might be able to better connect with and convince women in the informal sector to join. Roadshows or enrollment drives could be focused in market areas where women work during the day, as many of these 29 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s women will be unlikely to be able to take time off during the day to register and have childcare or other family responsibilities after work. Where feasible, incentives and nudges should aim to respond to the needs of women where and as these can be ascertained. The World Bank–financed NYOTA project does this by including a maternity insurance benefit as an added incentive for women who save in Haba Haba. The design being contemplated is one where women who have saved a minimum amount in the 12 months prior to giving birth to a child would be eligible for a lump-sum benefit at the time of the child’s birth. During this period, women would also be enrolled to receive free prenatal checkups under the government’s Linda Mama program. The objective was to provide women who save in Haba  Haba with a short-term tangible benefit they value—and, in doing so, improve the health and well-being outcomes for themselves and their chil- dren. The NYOTA project will evaluate the impact of this benefit and make recommendations for fine-tuning or for including it as a perma- nent benefit for all Haba Haba beneficiaries. Such gender-responsive solutions can ensure that women, who make up the majority of infor- mal sector workers, see a clear value in joining the scheme and are attracted to contribute. Research suggests that women savers tend Research suggests that women to be more regular, so focusing on them could also help the scheme savers tend to be more regular, reach scale much sooner. so focusing on them could also help the scheme reach scale 2 . Strike a balance between customization and scalability. Haba Haba’s much sooner. experience to date shows the importance of, and challenges in, tailor- ing incentives for each aggregator/informal sector group. Going forward, the NSSF should adopt an approach that balances customization efforts with the administrative need to achieve economies of scale and sim- plify scheme administration. Haba Haba needs to be scalable for it to be viable. Buy-in from multiple stakeholders will ensure that the scheme is not beholden to any one partner. Next steps include the following: l Create a consortium of aggregators that includes government agen- cies, the private sector, and civil society. Explore partnerships based on shared values, clarify expectations, and hold annual meetings to mon- itor progress. l Use targeted surveys and impact evaluations to assess which cus- tomized incentives are cost-effective and can lead to both take-up and persistence of savings. l The NSSF needs to be in the driver’s seat and work alongside aggre- gators and local government officials to achieve targets at scale. 3 . Improve system readiness to process short-term benefits and pen- sion payouts. The Haba  Haba scheme was launched in 2019, so early 30 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s withdrawals and retirement payouts are unlikely to arise immediately. But systems and processes need to be in place to ensure that anyone who claims a benefit has a smooth experience. Confusion or delays in pay- ment due to unpreparedness can lead to negative word of mouth, which will hurt the scheme and the NSSF’s image. 4 . Use targeted surveys to inform scheme improvements. The flexibility built into Haba Haba (e.g., allowing partial withdrawals after five years, bundling to offer protection against other short-term risks) reflects its understanding of the liquidity needs of informal sector workers. An evi- dence-based approach, grounded in targeted surveys and focus group discussions, should be used to understand the perceptions of the target group and inform scheme design. Area 2: Design and implementation of outreach and communications strategy An effective communications and mobilization strategy at all stages of the Currently, communications delivery chain is key to the scheme’s continued success. The following are is an afterthought and not suggested improvements in Haba  Haba’s marketing and communications mainstreamed into NSSF approach. operations, either for the mandatory scheme or 1 . Implement a customer relationship management system. Currently, Haba Haba. communications is an afterthought and not mainstreamed into NSSF operations, either for the mandatory scheme or Haba  Haba. A CRM system will allow the communications team to record and monitor com- munications between customers and the NSSF. Given the voluntary nature of participation in Haba Haba, it is extremely important to focus on member communications. Currently, Haba  Haba members can reg- ister their grievances by visiting the branch in person, on social media (Facebook and X [formerly known as Twitter]), through written corre- spondence, or via the call center. However, inquiries and complaints are not logged, responses on social media are not monitored, no feedback is sought as to whether a customer’s complaint was resolved to his/her sat- isfaction, and the call center does not submit any reports to management for its review. It is imperative that investments be made in procuring and deploying a CRM system and training relevant headquarters and branch staff to use it. 2 . Relaunch Haba Haba and designate a dedicated team to work on its implementation. Haba  Haba is yet to be known by its name. It does not have its own website (unlike the Ejo Heza scheme in Rwanda); and 31 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s in its marketing material, it is labeled “NSSF ni Haba Haba.” The NSSF, for its part, is well known but unfortunately evokes perceptions of mis- trust stemming from past issues of delayed payments and mismanaged assets. While this is no longer the case, there might be value in disasso- ciating Haba Haba from the NSSF and relaunching it as a stand-alone product on its own merits. There is no dedicated team for Haba Haba within the NSSF, including for communications. A voluntary scheme targeting the informal sector needs an agile and innovative communications approach that is well coordi- nated with the strategy team. Working in silos will be ineffective when the pressing need is to create a new playbook to expand coverage to the informal sector. Consultations revealed that communications teams were absent from some of the outreach to informal sector leaders. When they were present, they were not always equipped to answer technical ques- tions about the scheme. The lack of internal coordination on Haba Haba can signal to informal sector workers that their money might not be safe in this scheme. It is thus important to have a strong and dedicated team for Haba Haba, at least in the initial stages. Over time, NSSF policy makers can decide if the core team on Haba Haba should have a perma- nent presence or be absorbed into the standard NSSF business lines once the scheme kicks off. A common and shared 3 . Develop a detailed FAQ manual and educate NSSF employees about understanding of the scheme’s the scheme. Haba Haba needs a detailed FAQ (frequently asked ques- rules and operational steps will tions) manual like the one available for the NSSF mandatory scheme. help to minimize variances in During consultations, it was found that staff did not know, for example, customer experiences and avoid what the partial withdrawal rules are for someone who contributes to confusion. Haba Haba for a few years but thereafter gets a job in the formal sector and can no longer contribute. Knowledge of the distinction between a voluntary scheme for mandatory workers and the Haba  Haba scheme was also unclear. Interpretations of scheme rules also differed across dif- ferent personnel to whom the World Bank team posed questions. There is no single source of comprehensive information on policies and processes regarding Haba  Haba. As the NSSF moves out of the pilot phase, a common and shared understanding of the scheme’s rules and operational steps should be established. This will also minimize variation in customer experiences and avoid confusion. 4 . Adopt an inclusive communications strategy. The diversity within the large group of informal sector workers in Kenya means that user experi- ence of business processes and the context of communications messages needs to be looked at from multiple angles—including urban versus rural, youth versus middle-aged, feature phone users versus smartphone users, 32 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s women versus men, and individuals with and without phones. Communication is also needed at each stage of the deliv- F i g u r e 1 1 : Ejo Heza sliding ery chain, not only during outreach. For example, once people brochure, showing estimated have registered, they should be contacted to notify them returns about the scheme’s rules and to congratulate them for invest- ing in their future; after their first contribution, they should receive a communication to build trust that their money was safely received; they should be reminded when they miss a contribution or when one is due; and they should be con- tacted periodically regarding investment returns accruing on their savings. Such interventions throughout the delivery chain will improve trust in the scheme and make members more comfortable in saving larger amounts. 5 . Update the operations manual to highlight de facto busi- ness processes. The delivery chain assessment revealed that, at multiple stages, the de facto process or interface differs from that documented in the operations manual (as can be seen in the annex). A business process review to identify those areas where the de facto is different from the de jure is needed, and decisions should be made on how to close the gaps. The operations manual should also be widely circu- lated among those in the NSSF who work on the Haba Haba scheme. 6 . Develop pamphlets or digital calculators to show individ- uals how much they can expect to receive. Contributors to the Haba Haba scheme would be interested in knowing how much they can receive if they save in the scheme. Estimating this amount will depend on various factors, including how much they contribute, how long they contribute, when they withdraw, and what the average investment return is during this period. While these are all assumptions and require cave- ats to be put in place, there are benefits in showing a range of outcomes and how they will vary if an individual saves more or longer. Rwanda’s Ejo Heza scheme, a defined contribution plan like Haba Haba, created an easy-to-understand, interac- tive brochure for this purpose (figure 11). Similar approaches, either print publications or digital on the app, can be devel- oped. Short (less than one minute) videos that show how to register, contribute, check balances, and get enrollment infor- S o u r c e : Ejo Heza brochure. mation can be shared on WhatsApp and YouTube. 7 . Balance outreach effectiveness with associated costs. The NSSF’s current outreach approach works out to a cost of about $7 per 33 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s individual contributing approximately $61.35 per year. This cost will likely increase, given the challenges of outreach involving targeted messag- ing for diverse and geographically dispersed target groups, the growing complexity of Haba Haba as it introduces various bundles, and the effec- tiveness of in-person channels. The NSSF should consider the sustainability of its outreach approach to ensure that the program is sustainable in the long term. Area 3: Updating business processes and administration Good-quality data are fundamental to the efficient administration of any pension fund. With a potential membership of millions of individuals, maintaining the accuracy and security of information and being able to trou- bleshoot in-house are critical. The Haba  Haba scheme can take multiple steps to securely collect and use data to improve its effectiveness. 1 . Undertake a well-planned and coordinated IT modernization effort. The core IT system of the NSSF, the Social Security and Pension Administration An IT modernization initiative System (SSPAS), was implemented in 2013 by a third party. The system for the NSSF is long overdue, has not been updated since then, and interfacing the old system with given the legacy challenges it that of new partners is time-consuming and expensive. These challenges faces with fragmented records notwithstanding, the NSSF has developed interfaces with NHIF systems and the inability to disaggregate and the ID database for validations. An IT modernization initiative for data. the NSSF is long overdue, given the legacy challenges it faces with frag- mented records and the inability to disaggregate data. The cost of such an effort needs to be carefully considered and a plan put in place so that data are not lost during migration. The pros and cons of available sys- tems in the marketplace should be considered, and a competitive bid for the project sought. The NSSF IT team should work closely with the firm deploying the new systems so that they have the capacity to manage the systems independently and are not dependent on third-party contractors. l Segregate Haba  Haba data from that of the mandatory scheme and clean records. There is no separate database for Haba  Haba members, and it is not possible to distinguish them (in real time) from formal sector NSSF members. Further, according to a study com- missioned by the World Bank, the NSSF database has instances of multiple members (as many as 20) with the same member number. Determining whether a member is a contributor to the mandatory scheme or to Haba  Haba requires an arduous back-end analysis of membership records (to identify whether an employer ID is tagged to the member). Segregation of records and cleanup/migration of past 34 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s records are warranted under the IT modernization effort. The lack of good-quality and real-time data for Haba Haba limits the ability to evaluate how many women contribute to the scheme, how often, and at what contribution level. Going forward, having gender-disaggre- gated data on all of these indicators in the Haba Haba scheme will be important in understanding and tracking progress in attracting women savers to the scheme. l Develop a beneficiary management system. In the absence of a beneficiary management system, delays or mistakes in processing withdrawals/payouts can occur, requiring significant administrative effort from the branch offices and local administrations to identify beneficiaries. This can be particularly challenging in the case of infor- mal sector workers who tend to be geographically mobile. l Hire the right talent to manage the IT modernization. In-house expertise is needed to allow for smooth data migration once the IT procurement and deployment are complete. A data verification exer- cise for existing members needs to be launched to correct any issues with legacy records. Any suspension of updates to the database needs to be promptly communicated to the NSSF Board and staff, including the call center. Branch staff should be involved in these efforts, and steps should be taken not to replicate the legacy issues. 2 . Segregate and report administrative costs to better manage expenses. Interviews with the NSSF revealed that the administrative costs for engagement with the informal sector under Haba Haba are high. The NSSF’s current recordkeeping practices make it difficult to specifically identify Haba Haba’s administrative costs, but the tailored commu- nication approach combined with lower contribution levels and limited scalability suggest that these costs are high as a percentage of total con- tributions. This is understandable for a new scheme, especially one that The NSSF should quantify targets informal sector workers. In India and Rwanda, the government Haba Haba’s administrative costs, as it cannot manage what opted to subsidize administrative costs for a few years until the scheme it cannot measure. achieved scale. In the medium run, however, optimizing administrative costs is important. Accordingly, the NSSF should quantify these costs, as it cannot manage what it cannot measure. As coverage expands, admin- istrative costs as a percentage of contributions will decrease, so it is important to look at multiple indicators, if possible—for example, the cost of attracting a new member, the cost of retaining a member, staff costs, the costs of mobilization, and so on. Finally, as the scheme moves from piloting to national implementation, its expenses, including revenues and payouts, should be reported in annual reports, as is done for the manda- tory scheme. This will improve accountability and promote transparency. 35 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s 3 . Digitalize the verification process and increase interoperability. In-person verification in branches for first-time contributors to Haba Haba is quite inconvenient. The NSSF is rightly working toward digitalizing the verification. Priority should also be given to establishing interoperability with government databases (e.g., ID, social registry, youth employment) and aggregators with which there is a trusted long-term relationship. Data privacy is extremely important, and data governance arrangements should be put in place before data are shared across institutions. Most recently, the NSSF made its systems interoperable with those of Kenya Commercial Bank. Lessons learned from this process can be useful when exploring integration with other banks, savings and credit cooperatives, and mobile network operators. 4 . Improve data analytics to support planning and decision making. The NSSF, including Haba Haba, is amassing huge amounts of data—but if these data are not organized, they cannot inform policy making. Data should be used to generate standard reports with targets and trends for planning and policy making. Monitoring dashboards should guide dis- cussions in regular meetings with stakeholders, aggregators, and branch offices. 5 . Create a dedicated web page/app option for Haba Haba. Higher opera- tional efficiency and improved member experience can be achieved if the web page/app is intuitive and customer friendly (like M-Pesa). To this end, the Haba Haba web page/app must be easy to find, and shareable as a link with friends and family to help publicize the scheme by text, email, and word of mouth. Area 4: Institutional strengthening 1 . Institute a clear role for NSSF branches in promoting Haba  Haba. A change in mindset and new Branch officials understand the NSSF’s organization and processes as training are needed for NSSF well as the importance of increasing coverage to the informal sector. branch officials. Currently, though, they are an untapped resource. The role of branch offi- cials in supporting Haba  Haba outreach efforts is unclear at best and absent at worst. They do not receive regular information, education, and communications materials from headquarters and have limited guidance on the features of Haba Haba that should be highlighted to attract infor- mal sector workers. NSSF branch officials have experience with mandatory schemes where the task is not convincing people to join but ensuring compliance. A change in mindset and new training are needed. One or two officials from 36 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s each office can be trained to work on Haba Haba. Concrete next steps that the NSSF can take include the following: l Share the Haba Haba FAQ manual with all branches and invite them to information sessions to sensitize branch staff. l Identify one or two branch staff members who will regularly coordi- nate with the Haba Haba team at the head office. l Agree on a methodology to tag all new Haba  Haba members to a branch. This will enable the NSSF to monitor branch performance with regard to outreach, registration, contributions, redress of grievances, and other relevant functions. The methodology can be developed based on consultation with the branches. A communications budget for Haba Haba can be allocated based on the targets met. l Encourage the branches to work with local officials and partners to meet targets. l Institute an awards system for branches that meet targets and fea- ture them in annual reports. 2 . Strengthen monitoring, learning, and evaluation (MLE). The Haba Haba scheme lacks an MLE framework. The pilot disclosed many lessons, but these were not documented systematically. Without an MLE framework, there is a risk of not being able to identify what works well and what does not. Policy makers are unable to make informed decisions An M&E framework should be unless they have objective evaluations that show the cost and impacts developed with clearly identified of the decisions being made. Voluntary pensions are an emerging area performance metrics, data in social security with promising developments, but it remains a diffi- sources, review processes, and cult endeavor with no magic bullet to expand coverage. It is important quality assurance protocols. to learn from international experience and adapt this knowledge to the Kenyan context. A monitoring and evaluation (M&E) framework for Haba  Haba should be developed with clearly identified performance metrics, data sources, review processes, and quality assurance protocols. Learning should be encouraged internationally (through study tours of countries with relevant experience) and internally (headquarters, branches, among stakeholders, etc.). Impact evaluations and surveys should be designed to evaluate spe- cific interventions. 3 . Build technical capacity within the NSSF. Haba Haba was launched as a pilot using NSSF resources. To expand, Haba Haba needs changes in terms of strategy, IT, communications, the call center, and project man- agement. It will need a steady and dedicated team with the proper skill set and a precise terms of reference that is answerable to management and can work with the branches. Innovation will be needed for Haba Haba to 37 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s thrive in a world where social media use is increasing, attention spans are shortening, misinformation is common, and multiple options are avail- able for individuals to save. The research arm of the NSSF needs to be strengthened with in-house data/statistical expertise to analyze data Innovation will be needed for and conduct targeted surveys. Having the right talent available to it is Haba Haba to thrive in a world critical for the NSSF to push the frontiers of the social security agenda. where social media use is increasing, attention spans are 4 . Develop strategic and sustainable partnerships. Based on experiences shortening, misinformation is from the Haba Haba pilot, the following aspects are crucial when build- common, and multiple options ing partnerships: are available for individuals to save. l Engage in partnerships with shared values to foster and sustain long- term relationships. l Put in place a budget for integrating with systems of long-term partners with data privacy and security protocols in place. l Periodically touch base with partners to understand the level of com- mitment and address any grievances. l Develop a memorandum of understanding and, in cases of bundles with negotiated premiums, have a legal contract. Prioritizing recommendations for actions in the short term This section presents a sequencing of recommended actions to be taken that will enable the NSSF to relaunch Haba Haba for the 190,000 NYOTA bene- ficiaries within a short time frame. Learnings from NYOTA and investments made under it can then be built on for a national scale-up of Haba Haba. Because the NSSF is a large organization with experienced staff, it will be possible for work to begin on all four components simultaneously under a team leader for each component. Component 1: Upgrading NSSF IT systems and streamlining business pro- cesses for Haba  Haba. IT forms the backbone of any scheme that needs to register and accurately trace individual savings. The current situation in which Haba Haba records are aggregated with formal sector records hinders the achievement of multiple objectives. The first priority area thus should be to address this challenge. l Undertake an IT modernization effort focused on onboarding any new Haba Haba beneficiaries into the new system and to clean up and migrate the old data over time. 38 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s l For all new beneficiaries, digitalize the verification process for Haba  Haba membership and improve interoperability across data- bases to reduce the time needed for enrollment. l Create a dedicated and easy-to-navigate webpage, or an app option or WhatsApp account, for Haba Haba that will allow individuals to pose questions, seek prompt assistance, and check balances. l Develop a strategic, dependable, and long-term partnership with banks and mobile operators to ensure smooth collection of contribu- tions and prompt notification to members. l Improve system readiness to process benefits promptly and reduce the wait time once a person has submitted a claim. l Improve data analytics so policy makers can monitor outcomes in real time and take corrective action. l Update the operations manual to reflect de facto and not de jure processes. Component 2: Institutional strengthening and clarity around Haba Haba’s goals and targets and how to achieve them. Kenya is a trail blazer when it comes to digital adoption and the prevalence of mobile money in day-to-day transactions. However, a scheme like Haba Haba requires some on-ground presence to be complemented with the use of digital payments because it requires building trust and explaining how the scheme can benefit individu- als. The NSSF, with its partnerships with banks and mobile money operators and a brick-and-mortar presence nationwide through its field offices in all 47 counties, is best suited to meet this need. To leverage its unique capability, the NSSF needs to strengthen itself as an institution by providing clarity on staff roles and responsibilities, improving staff technical capacity, and rec- ognizing outstanding performers. It also needs to coordinate better with external partners and internally with branch officers so that everyone under- stands the vision and works toward a shared, defined goal. Specifically, the NSSF should undertake the following: l Institute a clear role for NSSF branches in promoting Haba Haba, set objectives, and develop a mechanism to recognize and reward branches that meet their objectives and targets. l Build human resource capacity among staff and leverage external expertise to complement gaps in staff capacity. l Strengthen monitoring, learning, and evaluation so NYOTA learnings are accurately captured. l Develop strategic and sustainable partnerships. Component 3: Outreach and communications. The NSSF has a commu- nications team, but it currently manages communications for both the 39 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s mandatory scheme and Haba Haba—without much clarity on the latter. For the mandatory scheme, the team depends on the law and years of train- ing to answer member questions, but for Haba Haba, there is no law or FAQ manual on which to rely. This leads to confusion within the communication team and call center staff about Haba Haba benefits and rules, to the det- riment of the scheme. l Develop a detailed FAQ manual and educate NSSF employees on the scheme, notably those who are tasked with managing it. l Implement a customer relationship management system so calls and complaints can be recorded and tracked. l Adopt an inclusive communications strategy that considers each stage of the delivery chain. l Develop pamphlets or digital calculators to indicate to individuals the funds they can expect to accumulate if they save a certain amount consistently. l Monitor the costs of mobilization and assess the cost-effectiveness of individual mobilization strategies. Component 4: Design considerations. Global experience around voluntary pension schemes is still evolving (see table 2), which means that there are relatively few solutions that can be recommended for leading to scale-up. The differing ability to save, social norms around saving, financial awareness, and trust in institutions/governments are all likely to influence take-up and persistence rates. Therefore, in developing a design that will work for Kenya, the NSSF can draw on international experience but will need to evaluate different approaches in its own context to identify what works best for its population. The following steps can be prioritized: l Draw on international experience around matching contributions, auto debit, and access in the short term, which are common features across all successful schemes. l Strike a balance between customization and scalability to avoid tai- loring for each worker profile (e.g., transport workers, craft association members) by instead choosing design elements that would attract a larger group (e.g., women or youth). l Use targeted surveys to inform scheme improvements. 40 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s Ta b l e 2 : International experience around incentives and nudges for informal sector workers Design of scheme Government financial incentives Flexibility or nudges Institutional arrangements India: National Pension System–Lite (NPS-Lite) (2000), Atal Pension Yojana (APY) (2015), Pradhan Mantri Shram Yogi Maan-dhan (PM-SYM) (2020) NPS-Lite: direct Matching incentives given by NPS and APY do not allow early access NPS is administered by NPS contribution (DC) government for all schemes, but for before retirement; PM-SYM allows Trust, an autonomous body a limited period lump sum withdrawal under select set up by the government of APY and PM-SYM: conditions India; APY is administered contributory defined As of 2023, no further fiscal by PFRDA (pension fund benefits incentives for NPS-Lite and APY; Banks are incentivized (monetarily) in regulator); PM-SYM is 50% of contributions to PM-SYM APY when they enroll a new member or administered by the Ministry still paid by government when someone saves consistently; APY of Labor and Employment contributions are auto-debited from bank PM-SYM is on the one-stop government portal (eSHRAM), which has over 200 million registered individuals Rwanda: Ejo Heza Long-Term Savings Scheme (2018) DC scheme 100% matching for poorest Up to 40% of savings in excess of Rwanda Social Security Board households, 50% matching for $4,000 can be withdrawn administers the scheme; a vulnerable households if they dedicated team from the meet a minimum savings goal; life Ministry of Finance launched insurance and funeral insurance for the pilot initially all households that save a certain amount All monetary incentives are paid for by the government of Rwanda but are expected to be phased out once scheme reaches scale Thailand: Social Security Fund (SSF) Article 40 (1999), National Savings Fund (NSF) (2011), Retirement Mutual Fund SSF Art 40 and Government matching for both No early withdrawal allowed SSF manages the mandatory NSF-DC with select SSF and NSF, with the latter scheme and is the contribution options  scheme allowing for much smaller administrator for one of the contributions other schemes Pakistan: Hybrid Social Protection Scheme Short-term savings Government matching for those Flexibility to withdraw own savings but Benazir Income Support targeted for poor who meet a minimum saving not the government matching before Programme (BISP) provides and vulnerable amount 2 years nationwide social safety net women transfer to vulnerable families 41 Photo: Petr Kapuscinski / World Bank Annex A Methodology The methodology encompassed a combination of (1) primary data collection through interviews with various stakeholders, co-led with NSSF representa- tives and Haba Haba branch officers, and including users such as informal sector group and community representatives, as well as technical experts at the World Bank; and (2) desk research to augment our understanding and identify best practices within the industry. To complement the document review, a questionnaire was circulated among those involved in the day-to- day operations of Haba Haba to understand the de facto processes. Based on the responses, the Haba Haba delivery chain was established. This approach allowed us to capture both the qualitative nuances of stakeholder perspec- tives and the quantitative metrics necessary for performance evaluation. The primary aim was to gain insights into the efficiency and effectiveness of the delivery chain processes within the NSSF. Specifically, we sought to understand the interplay between key stakeholders, identify bottlenecks, and propose recommendations for improvement. The team onboarded the NSSF leadership team to be core members in con- ducting the analysis. The delivery chain requires minute details and nuances to be addressed which required accessing the NSSF’s institutional memory. Close engagement highlighted undocumented aspects of the pilot that the NSSF had conducted in rolling out Haba Haba. It also highlighted silos of information across departments and teams. Several stakeholders at the NSSF headquarters were consulted as part of the assessment process including the chief executive officer (CEO), the head of strategy, the head of ICT, the manager of registration and collections, the claims team, the communications team, the back-end call center team, and the grievance redressal team. Beyond headquarters, the team met with branch officers, end users, and other relevant parties during targeted field visits where limited documentation was available. These interviews served as the primary source of qualitative data, providing invaluable insights into the intricacies of the delivery process. 43 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s Additionally, the team engaged in desk research to supplement our under- standing, identify industry benchmarks, and gather secondary data to validate our findings. Key component identification Through continued engagement, a technical mission, collection and review of NSSF documents, and real-time iteration and application of the delivery chain, the team completed its analysis. Regular engagement facilitated in gap filling and a nuanced approach to capture details of how Haba Haba was conceptualized, advertised, and achieved. These components were care- fully mapped to delineate the flow of operations and identify potential gaps or inefficiencies. This fed into the recommendations provided in this report, which were presented and agreed upon at a meeting chaired by the NSSF CEO. Data analysis Given that the data were collected from interviews and desk research, iter- ative rounds of analysis were conducted, presented, and then incorporated into the report, allowing the team to refine and validate findings through tri- angulation of multiple data sources. l The data were collected and analyzed over a one-year period in 2023. l To ensure the validity and reliability of the findings, the team triangu- lated data across primary interviews, NSSF documents, and secondary research. Any discrepancies identified were addressed through fol- low-up meetings and peer reviews. l Informed consent was obtained from all participants, and, where requested, the confidentiality and anonymity of sensitive information shared during interviews was respected. 44 Annex B This annex contains screenshots from Haba Haba’s Operations Manual. F i g u r e B . 1 : Steps for registration of new member 45 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s F i g u r e B . 2 : Steps for activation of member post registration F i g u r e B . 3 : Member main menu 46 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s F i g u r e B . 4 : Opting into benefits from partner organizations to access bundle of benefits 47 M a k i n g P e n s i o n S a v i n g s E a s y a n d Ef f i c i e n t f o r I n f o r m a l S e c t o r Wo r ke r s F i g u r e B . 5 : Business split transaction with partners so the member pays and receives for the bundle 48 References Borate, Neil. 2022. “Gift Your Domestic Help a Pension through WhatsApp.” Mint February 16. Clark, J., A. Metz, and C. Casher. 2022. “ID4D Global Dataset 2021: Global ID Coverage Estimates.” Washington, DC: World Bank. Gough, O., and C. Niza. 2011. “Retirement Saving Choices: Review of the Literature and Policy Implications.” Journal of Population Ageing 4 (1): 97–117. Guven, Melis U., and Himanshi Jain. 2023. “The Promise of Ejo Heza: A Brighter Future for All Rwandans. Executive Summary.” World Bank, Washington, DC. Guven, Melis U., Himanshi Jain, and Clement Joubert. 2021. Social Protection for the Informal Economy: Operational Lessons for Developing Countries in Africa and Beyond. Washington, DC: World Bank. Jain, Himanshi, and Robert Palacios. 2022. “Assessment of VSS’s Voluntary Pension Scheme and Policy Recommendations.” World Bank, Washington, DC. KNBS (Kenya National Bureau of Statistics). 2020. “Informal Sector Skills and Occupations Survey (ISSOS): Basic Report.” Lindert, Kathy, Tina George Karippacheril, Inés Rodríguez Caillava, and Kenichi Nishikawa Chávez, eds. 2020. Sourcebook on the Foundations of Social Protection Delivery Systems. Washington, DC: World Bank. World Bank. 2022. Charting a Course Towards Universal Social Protection: Resilience, Equity, and Opportunity for All. Washington, DC: World Bank. 49 © 2023 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington, DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the govern- ments they represent. The World Bank does not guarantee the accuracy, completeness, or currency of the data included in this work and does not assume responsibility for any errors, omissions, or discrepancies in the information, or liability with respect to the use of or failure to use the information, methods, processes, or conclusions set forth. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be construed or considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specif- ically reserved. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Suggested citation: World Bank, “Making Pension Savings Easy and Efficient for Informal Sector Workers: Learning from Kenya’s Haba  Haba Pilot, World Bank, Washington, DC. Cover: Retirement icon by Lia Radiah/Noun Project.