Mobile Money as conduit for Conditional Cash Transfers in the Philippines presented at the CPR South Conference in Maropeng, South Africa 11 th of September 2014 Erwin A. Alampay & Charlie Cabotaje Center for Leadership, Citizenship and Democracy (CLCD) NCPAG -University of the Philippines
Conditional Cash Transfers (CCT) in Philippines CCT Beneficiaries Country-wide Photograph: The Guardian/Vanderlei Almeida/AFP/Getty Images 5,000,000 4,500,000 4,000,000 3,500,000 3,000,000 2,500,000 2,000,000 1,500,000 1,000,000 500,000 0 1,012,195 2,278,559 3,121,530 4,309,769 3,937,591 2010 2011 2012 2013 2014-2016 Global trend in CCT in the world Dramatic increase in CCT as a strategy for poverty alleviation in the Philippines Source: ADB
Growth of CCT in the Philippines 40,000,000,000.00 35,000,000,000.00 Amount funded annually 33,908,902,800.00 3300% increase in 5 Years! 30,000,000,000.00 30,909,950,200.00 25,000,000,000.00 20,000,000,000.00 Amount funded annually 15,000,000,000.00 10,000,000,000.00 5,000,000,000.00 6,274,887,500.00 10,035,585,300.00 13,151,920,900.00 0.00 1,009,025,500.00 2008 2009 2010 2011 2012 2013a
Challenge of distribution Overall, GRemit only serves around 12% of CCT beneficiaries, but this peaked at 29% in 2011. However, the value has increased from Php980M to Php4.08B (peaking at Php6B in 2012) passing thru GRemit This also varies across regions 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Proportion of funds released thru GCash Other conduits Released thru Gcash/Gremit 2008 2009 2010 2011 2012 2013a BUT IS IT BETTER? Cheaper? More efficient? More secure?
Distribution Before it was done quarterly (every 3 months), but now it is done bi-monthly Benefits are 500/mo; 300/child (up to 3children)== ranges from 1600 to 2800 (if done bi-monthly) If done monthly, it would range from Php800-1400
Transaction costs for DSWD Conduit (period) Cashcard Philpost GRemit (2010-11) GRemit (2012) Mlhuillier (2013) Cost/Remarks initial Php50, no transaction costs thereafter (Landbank ATMs); Php20 other ATMs (interbank fees) Php50 Php75 Php 60 (negotiated) Php42 (Gcash was disqualified/late) GRemit (2013) Php42 (Mlhuilllier did not participate, 42 became the ceiling set, LBC was diqualiflied)
Reasons for using GRemit Pragmatic- need for more direct/closer channels of delivery because of from that transportation expenses was greater than beneficiaries were receiving; Perceived Cost savings (compared with other alternatives (e.g. using helicopters; boats) Easier logistically to find partners
Data Gathering Succeeding Surveys and field observations were conducted in 4 other provinces (but survey results are not yet included here) Interview with DSWD- CCT region IV Interview w/dswd Central Office Field Observations Survey (n=192) results from CCT implementation in field
Region IVB CCT Conduits 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2008 2009 2010 2011 2012 2013a Other conduits Region IVB Released thru Gcash/Gremit Mindoro Occidental CCT Conduits San Jose, MO, CCT Conduits 100% 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% Other conduits Released thru Gcash 90% 80% 70% 60% 50% 40% 30% 20% 10,932,700.0 0 31,791,500.0 0 25,831,100.0 0 25,017,500.0 0 Other conduits Released thru Gcash 0% 2008 2009 2010 2011 2012 2013a 10% 0% 2012 2013a
Observations regarding CCT distribution m-money was not really used G Remit Merchant partners are limited by the number of people/desks they set-up
Observations regarding GRemit CCT distribution Distribution is manual, and not simultaneous Long lines occur waiting for merchant partners to set-up, even if processing may take only a minute/beneficiary
Observations Security remains a problem NOT MORE EFFICIENT NOT MORE SECURE
Other Observations Temporary markets/vendors follow where the CCT Distributions are setup.
Comparison of M-Money costs for sender SMART Money GCASH Remittance Charges Total (Max) (for Charges (Php20 Range sender per 1000) Ph1-Php500 Php5 Php5 Php20 Php 501-1000 Php5 Php5 Php20 1001-1500 Php 5 + 2.5 Php7.50 Php40 1501-2000 Php 5 + 5 Php10 Php40 2001-2500 Php 5 + 7.50 Php12.50 Php60 2501-3000 Php 5 +10 Php15 Php60 above 3000 Php 5 + 2.50/Php500 above 1000 Php 15 ++ Php80++ NOTE: There s also a one-time initial Card Application fee of Php100 for SMART
Comparison of m-money charges for recipient Remittance Range Ph1-Php500 Php 501-1000 SMART thru ATMS Php5 (BDO); Php 15 other networks Php5 (BDO); Php 15 other networks 1001-1500 Php5 (BDO); Php 15 other networks 1501-2000 Php5 (BDO); Php 15 other networks 2001-2500 Php5 (BDO); Php 15 other networks SMART merchant ( 1% per tnx) Php 1-5 Php-5-10 Php10-15 Php15-20 Php20-25 Gcash merchant None None None None None
Comparison of m-money Costs for DSWD/beneficiary (total) Remittance Range Current GRemit model SMART Money Gcash Direct model Ph1-Php500 Php42 Php10 Php20 Php 501-1000 Php 42 Php 15 Php 20 1001-1500 Php 42 Php17.50 to Php 22.50 Php 40 1501-2000 Php 42 Php25-30 Php 40 2001-2500 Php 42 Php32.50- Php37.50 Php 60 2800 Php 42 Php 43 Php 60 Cost Threshold
Initial Survey Results Beneficiaries farther away were more resistant to using mobile money, and those nearer were also more willing (considering that ATM option would also be there) Most of the respondent beneficiaries were SMART subscribers (also validated in 4 other CCT areas) Many were more familiar with SMART Padala centers Those who have had previous experience using SMART money were more likely to report interest in using this option In 4 other areas: SMART and Gcash Centers were present BUT resistance was absence of signals; electricity
Stages of Access* Interest: No lines No ID needed Less transport cost Good in Emergency USAGE ACCESS: 33% (63/192) have used SMARTmoney 1.5% (3/192) have used GCash SKILLS ACCESS: 34% (65/192) know how to receive money using a cellphone MATERIAL ACCESS: 49% (95/192) have a cellphone; 86 (or 90%) are subscribed to SMART No cell: Can t afford Not a priority No signals MENTAL ACCESS: 136/192 were interested or willing to receive CCT directly to the cellphone (71%) * Adopted from Van Dijk (2009), as conceptualized in Alampay & Bala 2010
RESULTS OWNERSHIP of a cellphone was not statistically significant as to whether they were interested in m- money for CCT (Chi=2.89, df=1, a=.089) Pre-existing Knowledge to use m-money was statistically significant to willingness to use this as a conduit. Those who know were more willing Chi=7.729, df=1, a=.005) Distance to known claim/redemption was also statistically significant to willingness (Chi 12.96, df=1, a=.000) Those closer were more willing.
Findings In actual GRemit implementation: Benefits were not apprent: Not more efficient, not more secure Implementation was no different, and technology model was not developed Actual overhead cost (for distribution of transfers) not yet known, but rather based on bidding; unknown/variable costs for beneficiaries Feasibility of using a mobile money model: Projected cost using SMART Money or Gcash can be potentially lower, especially for smaller and more frequent cash transfers. SMART would be a more technically feasible alternative should direct transfers to cellphones are considered (more centers, subscribed to by more beneficiaries)
Field Evidence: Other Emerging Models UNDP- Cash for work program (post-haiyan) Globe Banko- interest earning mobile based bank account (applied by Mercy Corps) USAID-SIMM Project ---building up m-money use in the local economy (salaries; loans, utility payment) Other country experience (see Aker et. al 2011; Nigeria e-wallets for farmers)
RECAPITULATION M-Money for conditional cash transfers is Technologically feasible in some areas Based on Typhoon Haiayan experience Consider availability of complementary systems/services M-Money if applied in CCT can be financially more cost efficient Need to strenghten local m-money ecosystems
What needs to be done Developing local m-money ecosystems To reduce need to cash out Education On use Alternative systems (SMART, Gcash, Globe Banko) About variance in fees Posisble fees in ATMs
Thank you. eaalampay@up.edu.ph or cabotaj@yahoo.com