Cash or In-Kind, Electronic or Manual Transfers? Evidence from Field Experiments in West and Central Africa

Transcription

1 Cash or In-Kind, Electronic or Manual Transfers? Evidence from Field Experiments in West and Central Africa Jenny C. Aker, Tu.s University and IFPRI JPAL SEA Policy Conference on Social ProtecAon Jakarta, Indonesia January 12, 2016 Jenny C. Aker Cash, In-Kind, Electronic, Manual

2 How should society (donors, governments and NGOs) redistribute to the poor? Cash transfer programs are an increasingly important part of social protecaon programs in both developed and developing countries Yet many social protecaon programs are sall inkind Food aid, food stamps, medicines, inputs, vouchers Especially in lower-income and conflict countries

3 If program recipients would weakly prefer an equalvalued cash transfer to an in-kind transfer, why not use cash? Cash might not be the preferred modality for redistribuang wealth TargeAng Local supply PoliAcal feasibility Paternalism SubstanAal evidence of the impact of each modality, but limited evidence of the rela%ve effects of each modality o Recent studies on cash versus in-kind transfers (GenAlini 2014, Cunha 2014, Hidrobo et al 2014, Hoddino] et al 2014)

4 Regardless of the modality, social protecaon programs present logisacal, operaaonal and security challenges This can reduce the (cost) effec%veness of social protecaon programs o It can also result in substanaal direct and indirect costs to program recipients.

5 Research Goals Research What are the relaave effects of different transfer (cash versus in-kind) on household purchases, consumpaon and well-being? o What is the cost effecaveness of each modality? How does the transfer delivery mechanism (physical versus electronic) affect the costs of implemenang such transfers? o Are there any addiaonal benefits from cashless (electronic)? Two randomized control trials (DRC and Niger)

6

7 Social DRC Implemented by Concern Worldwide in a camp for internally displaced persons (IDPs) US$130 transfer provided in three installments between September 2011 and March 2012 Provided to increase asset accumulaaon (NFIs) and meet food needs Jenny C. Aker Cash, In-Kind, Electronic, Manual

8 Jenny C. Aker Cash, In-Kind, Electronic, Manual

9 Research Design 474 households were randomly assigned to one of two transfer modaliaes T1: cash transfer. Cash transfers provided in three installments T2. Voucher. Vouchers provided in three installments o First voucher could be spent on food and non-food items at a mulasectoral fair o Second and third vouchers could only be spent on food items Equivalently-valued transfers provided at same Ame and same amounts Both groups had to travel to main urban center (15 km from camp) to pick up their transfer

10 Voucher Fair

11 Why Vouchers? Encouraging consumpaon of paracular goods TargeAng Market supply PoliAcal feasibility Security

12 Is this like other voucher programs? Differs from tradiaonal voucher transfers by: Timing had to be spent in one day LocaAon could only be spent at voucher fair (rather than vendors, kiosks or markets in different locaaons)

13 Why would cash and vouchers differ? Cash can be spend on anything, whereas vouchers could (primarily) be spent on food If the value of the voucher is more than the household would have spent otherwise on food, then cash and voucher households will purchase different things o Voucher households will spend more on food to use the full value of the transfer If voucher households cannot sell the vouchers (or what they purchased), then this will also translate into differences in consumpaon Jenny C. Aker

14 Impacts Items purchased Well-being (food security, assets, coping strategies) Costs

15 Household Purchases 100% 90% 80% 70% But voucher households were more likely to buy rice, salt and fish and bought more of them! Cash households used the transfer to buy more types of food and non-food items Voucher Cash 60% 50% 40% 30% 20% 10% 0%

16 Household Purchases Voucher households bought 6 Ames more salt (10 kg more) than cash households! 8 Voucher Cash Salt (kg) Rice (kg) Fish (number) Jenny C. Aker Cash, In-Kind, Electronic, Manual

17 Why different purchases? Voucher households had to spend the voucher on food items at the voucher fair to use the enare value of the transfer o This shi.ed expenses towards foods that were easily storable and easy to resell o But choices were sall constrained If something was too heavy, I didn t buy it I wanted to buy two boxes of salt but could only carry one, so I bought one plus other things. Voucher program recipient, Masisi camp

18 Did the Type of Transfer lead to Different Well-Being? Food security (household diet diversity, number of meals per day, number of months of adequate household food provisioning) Asset ownership and savings (durable and nondurable goods categories) o But cash households were able to save some of their transfer (US$ 2) Agricultural assets (land, livestock) Coping strategies Jenny C. Aker Cash, In-Kind, Electronic, Manual

19 Why did purchases differ but wellbeing didn t? Different Purchases Vouchers restricted to food, and had to purchase those items on a specific day But Similar Well-Being Voucher households resold some goods purchased Voucher households had less bargaining power visà-vis traders had to pay higher prices Voucher and cash households shared goods (and cash) Jenny C. Aker Cash, In-Kind, Electronic, Manual 19

20 Costs per Program Recipient $16.00 $14.00 $12.00 $10.00 $8.00 $6.00 Transfer Fees Account Opening Costs Voucher prinang Materials (plasac sheeang, sacks) Transport (fuel, lodging) Staff Ame Voucher program cost US$3 more per recipient PotenAally higher security costs for voucher households $4.00 $2.00 $0.00 Cash Voucher

21 Cash or Vouchers? Cash and voucher households had very different purchasing pa]erns but similar levels of well-being o Different for some other studies Since well-being was the same and cash was cheaper to implement, cash is the preferred modality here o Goods and services were readily available on the local market, so cash was possible But costly for the NGO and program recipients: Could social protecaon transfers be implemented even more cheaply? Jenny C. Aker Cash, In-Kind, Electronic, Manual

22 Source: Concern Worldwide 22

23

24 Source: Concern Worldwide 24

25 Mobile Money Services by Country

26 26

27 Cash Transfer Niger Implemented by Concern Worldwide in response to 2009/2010 food crisis Monthly (uncondiaonal) cash transfer of $USD45 for 5 months ($USD 225 total) o 2/3 of annual per capita GDP o Provided during hungry season (May through September) to prevent severe malnutriaon and asset decapitalizaaon TargeAng o Poorest households targeted o Women received the cash transfer 27

28 Research Design 96 villages were randomly assigned to one of three intervenaons: T1. Cash: Manual cash transfer each month in the recipient s village or in a nearby village T2. Mobile: Manual cash transfer plus Zap-enabled mobile phone T3. Zap: Zap-enabled mobile phone plus the cash transfer via mobile money transfer No difference in Aming of transfers Value of the phone about $USD 5 28

29 Research Design C. Control T1. Cash T2. Mobile T3. Zap T3-T2: AddiAonal effect of the mobile money transfer approach T2-T1: AddiAonal effect of the mobile phone T1-C: Impact of the cash program (unobserved) 0.0 Pre-program Post-program

30 Mobile Money in Niger New mobile money transfer program (introduced in January 2010) o Few agents outside of capital city o Li]le consumer familiarity with the program o Limited adopaon Low mobile phone penetraaon (30%) High illiteracy rates (85%) Challenges with manipulaang mobile phones and remembering PIN codes Limited operaaonal capacity of mobile phone service provider in the mobile money sphere o Needed to link Concern s program recipient lists with Zain Zap interface 30

31 31

32 Why should this ma\er? Reduced costs of obtaining transfer can increase Ame spent on producave acaviaes and change the locaaon and Aming of purchases Reduced costs associated with informal private transfers help households to be]er cope with shocks Reduced costs increase access to informaaon and lead to more opamal decision-making Increased privacy of the cash transfer leaves more income available for the household and affects investment strategies Increased intra-household bargaining: TargeAng women can improve women s control over resources and investment in public goods 32

33 Impacts Costs Well-Being (Food Security, Number of meals) Mechanisms 33

34 Total Cost of Cash Transfer Program by Month (Euros) Cost per recipient is $.90 more in zap villages If the program conanued for another 2 months, Zap would have been less expensive per program recipient than manual cash May June July Aug Sept M - transfer Manual cash 34

35 2.5 2 Cost ($USD) to Program Recipient of Obtaining the Cash Transfer Program recipients in cash/placebo villages had to travel 2 km (one way) to get the transfer, approximately ½ hour (excluding waiang Ame) The zap group had to travel less than.5 km (less than 10 minutes) The Ame savings is equivalent to USD$.92 over the five-month period (lower bound), or 2kg of millet -- enough to feed a family of five for one day Distance (km) to Cashpoint Distance (hours) to Cash Point Cash/Placebo Mobile Money 35

36 Impacts 100% 90% 80% 99% 98% 79% Zap households were more likely to buy other foodstuffs 82% (cowpeas, condiments, meat and oil) 70% 69% 69% 60% 50% 40% 37% 46% 40% 51% Manual Cash/ Placebo M-Transfer 30% 20% 18% 18% 10% 6% 5% 0% 36 Grains Cowpeas Meat Condiments Oil Seeds School fees

37 Food Security 50% 45% 40% 35% 30% Household diet diversity was 16% higher in Zap households as compared with Mobile and Cash households 36% 43% Children in Zap households ate 1/3 more meal per ay 38% 29% 29% 27% 25% 24% 20% 17% 18% 15% 10% 5% 0% Cowpea Condiments Fats Cash Zap Mobile 37

38 Why should this ma\er? Reduced costs of obtaining transfer can increase Ame spent on producave acaviaes and change the locaaon and Aming of purchases Reduced costs associated with informal private transfers help households to be]er cope with shocks Reduced costs increase access to informaaon and lead to more opamal decision-making Increased privacy of the cash transfer leaves more income available for the household and affects investment strategies Increased intra-household bargaining: TargeAng women can improve women s control over resources and investment in public goods 38

39 Intra-Household Decision-Making Manual cash villages: o All program recipients (plus some husbands) traveled to obtain cash transfer o Majority provided cash to husband or son immediately a.er receiving the transfer o Li]le Ame for discussion on purchasing prioriaes Zap villages o Women received a discrete beep via mobile phone o Informed husband of transfer arrival in privacy of home (not right away) o Discussed purchasing prioriaes at home 39

40 Intra-Household Decision-Making Women in Zap households had greater responsibility for certain household acaviaes o Were more likely to travel to markets and be responsible for selling grains Zap households spent more on clothing for religious fesavals 40

41 Summary of Findings Mobile money decreased costs for implemenang agency and program recipients, but large fixed costs o PotenAally cheaper in an environment where mobile money is more well-known and has a larger agent network o However, issues around sensiazaaon and PIN codes must be addressed (Ghana) Benefits beyond simple cost reducaon o Increased diversity of purchases, improved diet diversity o High marginal uality of income potenaally more variaaon in outcomes under other condiaons Why? Changes in intra-household decision-making o Intra-household decision-making is very context specific o PotenAal other impacts or channels in the longer term 41

42 Policy Some evidence of impact of mobile money in Kenya and Rwanda and electronic transfers in India, but sall a fracaon of what we need to know The necessary infrastructure needs to be in place Need a supply-side (agents, plauorm) and demand side approach (sensiazaaon, trust, ability to use) o Is m-money profitable for the service providers in the shortterm? Would other financial services be profitable? Mobile money can potenaally promote financial inclusion, but will not necessarily bank the unbanked Savings, credit, insurance are all part of financial inclusion Payments can potenaally push take-up -- but will people use the services a.er the payment is made? 42

43 The Bo\om Line How should wealth be redistributed? Cash, voucher, in-kind? o Body of evidence suggests that while cash is more flexible, inkind transfers can help to achieve certain (health/nutriaon) objecaves o Depends upon targeang, local supply (availability and prices) and poliacal feasibility Regardless of the modality, either physical or electronic transfers can be used o Electronic can be cheaper, but the necessary infrastructure needs to be in place (non-trivial to set up) o Even if infrastructure exists, it could disadvantage the poor (especially if they difficulty in using the technology) 43

44 Thank you 44