12-03-23 fin landscapes summary report

8/2/2019 12-03-23 Fin Landscapes Summary Report

1/12

FSD KF St Dpg

T s s K's f

dsp: T t vd


2/12

Government o Kenya THE WORLD BANK

By

Susan Johnson

With

Graham K. Brown and Cyril Fouillet

With support rom Financial Sector Deepening Kenya

THE SEARCH FOR INCLUSION IN KENYAS

FINANCIAL LANDSCAPE : THE RIFT REVEALED

FSD KF St Dpg

The Kenya Financial Sector Deepening (FSD) programme was established in early 2005 to support the development o nancial markets

in Kenya as a means to stimulate wealth creation and reduce poverty. Working in partnership with the nancial services industry, the

programmes goal is to expand access to nancial services among lower income households and smaller enterprises. It operates as an

independent trust under the supervision o proessional trustees, KPMG Kenya, with policy guidance rom a Programme Investment

Committee (PIC). In addition to the Government o Kenya, unders include the UKs Department or International Development (DFID),

the World Bank, the Swedish International Development Agency (SIDA), Agence Franaise de Dveloppement (AFD) and the Bill and

Melinda Gates Foundation.

This report as commissioned by FSD Kenya. The ndings, interpretations and conclusions are those o

the authors and do not necessarily represent those o FSD Kenya, its Trustees and partner

development agencies.


3/12

THE SEARCH FOR INCLUSION IN KENYAS FINANCIAL LANDSCAPE : THE RIFT REVEALED 1

INTRODUCTION

Does the rapid adoption o mobile phone based money transer

services (MMT) suggest a tectonic shit in Kenyas

fnancial landscape and the greater potential

or ormal sector inclusion - especially in savings or

transactions accounts?

What are its implications or ormal sector services?

This report examines the use o MMT within the context o inormal and ormal

nancial service use. The study investigated both the supply and demand sides

o the market and uses both quantitative surveys and qualitative interviews.

It was designed to delve into nancial service provision and use below

the aggregate ndings o the national level FinAccess surveys in order

to illuminate the dynamics o nancial inclusion or low-income

groups and hence oer insights or policy and practice.

IT FINDS THAT:

MMT seamlessly acilitates a wide array o inter-

personal transactions that Kenyans undertake within

their social networks and relationships o amily, kin, riends

and colleagues.

The logic o these transactions is one o give and take in ways that circulate

resources through these social networks

MMT has captured these inormal transactions and made them vis ible.

This logic is also replicated in the popular use o inormal groups such as

merry-go-rounds and accumulating savings and credit associations where

people circulate resources among themselves.

Saving in and through these oers benets that saving in banks

and other ormal providers does not. Give can respond to takeon a negotiable basis.

To make dramatic strides towards ormal inclusion

the nature o these transactions needs to be

understood.

METHODOLOGY

The study was carried out in the environs o three small towns.

These were chosen to represent districts in each tercile o the poverty

ranking o districts in Kenya according to the report Wellbeing in Kenya

(GOK, 2005). They were Karatina in Nyeri District, Central Province (top tercile);

Nyamira in Nyanza Province (middle tercile) and Kitui in Eastern Province

(bottom tercile). Located in the higher potential agro-ecological zones, Karatina

and Nyamira oer insight into access and use where population densities and

ood security are not primary problems, while Kitui represents the semi-arid

spatial rontier o delivery and a context o chronic ood insecurity.

The supply-side study involved identiying and interviewing as many ormal

and semi-ormal providers based in each town and its environs as possible. In

all some 59 interviews were carried out with providers across the three sites with

near complete coverage o the major banks, and MFIs.

The demand-side study consisted o two components. A survey o nancial service

use carried out through random sampling in three sub-areas o each research site.

Within these sub-locations in turn villages were chosen to achieve representation

across the sub-location where there were dierences. Local administration ocialsprovided household lists at the village level rom which households were

randomly selected. This resulted in 337 respondents in 194 households.

From this sample, 148 more in-depth interviews were undertaken

with purposeully chosen respondents, in particular where they

had started using services, to explore the reasons behind

these dynamics.

The Search For incluSion in KenyaS Financial lanDScaPe :

The riFT reVealeD

THE NUMBERS: SAMPLE SizES

intervews carredout wth ormal

and sem-ormal

provders on the

supply sde

Respondents werentervewed n 194

randomly selected

households

in-depth ntervewswere undertaken

wth purposeully

chosen respondents

59 337 148

The Great Rt Valley


4/12

2 THE SEARCH FOR INCLUSION IN KENYAS FINANCIAL LANDSCAPE : THE RIFT REVEALED

OVERVIEw OF THE FINANCIAL LANDSCAPE

Money transer is the most used service ollowing by inormal nancial groups

and then banks

Money transer use is now almost double the level o the last FinAccess survey

The higher levels o usage o all services in this research compared to FinAccess

refect the location o the study in two higher potential areas

Access strands are shown below. The access strand approach treats MMT as

a similar service to nancial intermediation services. Since they oer transer

rather than intermediation we also show access strands excluding MMT. This

raises rates o inormal inclusion and exclusion.

Financial landscape

survey 2011

FinAccess

2009

MMT 60.8 27.9

ROSCAs 38.0 31.7

Banks 35.6 21.5

ASCAs 27.3 8.0SACCOs 22.8 9.0

MFIs 4.5 3.4

PATTERNS OF ACCESS AND USE

The ndings reveal a landscape o nancial service access that is most

consistently infuenced by gender and education.

Men are more likely to have MMT than women (and especially young men

compared to young women) and are more likely to have bank accounts but

are much less likely to use inormal nancial groups.

Having no education reduces the likelihood o having MMT compared to any

level o primary education or above; whereas having a secondary education

or above results in a higher likelihood o having a bank account.

Income was more strongly associated with a higher likelihood o using MMT

and less so or banks or nancial groups.

On the other hand, living urther away rom a bank reduced the likelihood

o either having an account or using it requently. By contrast people urther

away rom an MMT agent were also more likely to have a registration andthis underlines the importance o having the ability to receive transers even

or those living ar away. But or MMT greater proximity did not appear to be

associated with more requent use.

Visual representation o multple use o nancial services

Karatina / Kitui / Nyamira

Access ormal/inormal by gender

Percentage Percentage

100%

80%

FinAccess 2009 FinAccess 2009

excluding MMT

Financial landscape

survey 2011

Financial landscape

survey 2011 excluding MMT

40%

20%

60%

0%

Source: FinAccess 2009 and own data

Formal

Formal other

InormalExcluded

Access strands

Note: There is extensive multiple use o services spanning both ormal and

inormal sectors

5%

12%

13%

23% 23% 36% 36%

28%

23%

37%15%

33%

39%

18%

27%

32%


5/12


MOBILE MONEY TRANSFER (MMT)

MMT has been marketed as a remittance mechanism to send money home

and many transactions involve transers to household members and close

amily.

We also nd that a signicant proportion o transactions are with a wider

circle o other relatives and riends beyond amily or household members.

MMT is not used or saving as such, 34% o those registered or MMT who

had their money on phone, explained this in terms o sae-keeping and

mobility or having unds to use or onesel or others when needed, rather

than the act o accumulating savings

Terminology in local languages reveals that holding a reserve o unds to all

back on in emergencies is something people expect to have but which theydo not think o as orm o saving.

MMTs extensive use reveals the vast array o inter-

personal transactions that Kenyans undertake and

which are endemic to their nancial lives. These

transactions can be understood as embedded in

social networks and relationships. Within these

social networks, systems o give and take operate

in which avours or assistance o various orms are

given and received and o which cash is one. These transactions are usually ver y

open ended. Assistance is given with no particular expectation o return and is not

necessarily related to need. This can be understood as a orm o saving with others

through such transactions. However when needs arise, requests or assistance

can also be made and resources borrowed or received as a git. When unds are

borrowed whether or not they are to be returned will be clear but there is rarely a

set timetable or interest.

Support to children at school

or in the city

Funds or investments

Rosca group contributions

Funerals

Pre-weddings/weddings

Births

Harambees

Medical expenses

School ees

Business

Paying or labour, Rent

Borrowing in cash to be

returned at a later date

Approxmate (medan)

balance held on phone

Ksh300

There is a vast diversity o purposes or transers:

Pattern o money transers in rural areas last transaction

Pattern o money transers in tons last transaction

Have received unds Have sent unds78%

66%

83%

47%Have received unds Have sent unds

Ho many ever received/sent unds

64%

ever received unds

53%

ever sent unds

Reliability: MMT oers a high level o

reliability and conormity to their service

promise

Convenience: is now extremely good with

agents in even small market centres and the

ability to undertake transactions rom the

comort o your soa

Structure: The service is very simple and

does not oer a structure in itsel.

Flexibility: The key point is that it gives

access to the fexibility that transacting with

others oers.

MMT allows exibility or inter-personal

transactions

100%

80%

sent to Received rom

40%

20%

60%

0%

Business/Utilities/Other

Friend

Other Relative

Family/hh member

Pattern o mobile money transers last transaction

14%

15%

53%

18%

37%

30%

14%

19%


6/12


34%O those registered with a moneytranser service reported holding abalance on their phone

19%Gave saety as a reason or holdingmoney on the phone 18%Said they hold money in their MMTaccount in order to have unds tosend to amily, relatives, riends when they need

them

Likelihood o having anaccount: higher or men, bettereducated; those located closer toa bank

Frequency o use is higherwith higher incomes and lowerdistances.

BANKS

While there is a stronger sense now that bank accounts are or the common

man, the qualitative data revealed that these were oten opened or

processing payments - whether rom contracts, or regular payments rom

employment or or other lump sums to be received.

This pattern is a strong reason or levels o dormancy since once payments are

not being received they all into disuse.

However, there was evidence o a gendered pattern here with women tending

to use accounts more actively or saving than payments (also because they

are less likely to be receiving lump sumpayments), and - with a very high rate o group

membership among them this also suggests

that groups may have taught them a stronger

savings and money management culture.

Bank managers reported that when approaching banks, users are looking

or the opportunity to borrow, but the rate o bank borrowing in our sample

was very low at 9% o those with bank accounts. The disappointment in

seeking loans and diculties managing them when they were received,

helps explain that banks are not seen as an easy or straightorward source

rom which to borrow.

35.6%have bank account

Male

Female

Efect o distance on probability o having

a bank account by gender50%

40%

30%

20%

10%

0%

0 5 10 15 20 25 30

Distance to nearest bank Kms

P

robabilityofhavinga

ccount

Note: Gender has a large eect relative to distance on bank access. Being a

woman is equivalent to living some 10kms urther away rom the bank

THE NUMBERS

Note: Having secondary education or higher signicantly raises the

probability o having a bank account.

Low charges

No minimum deposit

Low interest rates

Minimal procedures

Easy access to loansEasy to open

No minimum balance

Proximity

Saety

Reasons given or opening an account

70%

50%

30%

10%

80%

60%

40%

20%

0%

0 5 10 15 20 25 30

Distance to nearest bank km

Probabilityofhavinga

cco

unt

Efect o distance on probability o having a bank account by

educational level

Primary

No Education

Secondary

More than Secondary


7/12


INFORMAL FINANCIAL GROUPS

Inormal nancial groups oer access to lump sums either through ROSCA

payouts or ASCA loans, which as some men particularly emphasised -

were hard to gain rom banks.

They also oer social connections with riends as well as a means o gaining

advice and guidance.

They are more requently used by women than men.

Groups are valued as very structured mechanisms

that assist in creating the discipline o saving

due to their regularity and the diculty o not

paying in ront o peers.

Access is very local, 90% o memberships

involved walking to the meeting with 90% o

these within 30 minutes walking distance

Groups can be considered as a orm o give and

take although a more balanced and structured

orm than that o the wide range o inter-personal transactions undertaken

using MMT.

To have lump sum when its

your turn 61%

To socialise with riends 57%

To help when there is other

emergency 54%

1 kg o mae ATM Wthdrawalcharges

1kg Ksh30

Structure: Inormal nancial groups are highly structured through

weekly or monthly contributions.

Flexibility: While much literature on groups emphasises the

discipline that they oer, this overlooks their fexibility. This fexibility

is provided through the social relationships that people have in their

groups and can be reerred to as negotiability. Through these

groups users gain access to resources both directly and through the

social relationships that they make in the groups themselves. It is

this that provides the fexibility in response to need.

Reliability:

it takes time to nd groups that are trustworthy andcan deliver services and many people nd this dicult to achieve.

Financial costs are low with ew ees or charges.

Convenience is high or those who can nd trustworthy groups.

Inormal nancial groups ofer structure but

exibility through social connections

Reliability: In the case o banks, their reliability in terms o delivering

the promised service at the promised price is high, although the price or

many is signicant. At the time o the research a withdrawal ee o Kshs30

rom an ATM represented the price o a kilo o unmilled maize which

would eed three people or one meal.Convenience: Convenience in terms o proximity is poor or those out o

town, but speed o service and privacy can be good.

Flexibility and Structure: Since deposit accounts are now highly

fexible in terms o withdrawals, this element is strong but the fipside

is that structure or helping people to save regularly is non-existent.

Moreover, fexibility in relation to loans is not a eature.

Group membership

% o sample ho are members o:

Any nancial or welare group 73

Group with a nancial unction( ROSCA, ASCA, savings or investment)

51

Group with a welare unction 53

Group with only a welare unction 22

Three main ansers on hy belonging to inormal group

Banks ofer reliable and exible savings services, but ew loans, no savings discipline and are costly

inormal nancal

groups are the second

most used nancal

servce

51%

Odwolla Group Savngs and Loans assocatons (GSL) clustered meetngs n

Rachuonyo, Nyana Provnce.


8/12


SACCOS AND MFIS

SACCOs are more used by men and those with primar y education but distance

is not important in having an account.

Patterns o use are biased towards monthly produce payment receipts with

very low levels o more requent deposits. Surprisingly, distance aects

requency o use positively.

The ease o getting loans through SACCOs on the basis o both shares and

cashfows is attractive, especially or those who have built up their shares

over years.

Although SACCOs are seen as more reasonable than banks, complaints about

the transparency o charging practices were in evidence.

MFIs are being used more by women than men, by the more educated andby business owners.

The main advantage and reason or joining MFIs was access to loans including

the act that multiple loans could now be held

However, the disadvantages cited included application ees, the vetting

process when taking a loan, high interest rates, paying as a guarantor when

others deaulted, an uncaring attitude rom the MFI when payment was

dicult and conscation and auction o items.

MFIs have the highest eective interest rates among all regulated service

providers, while SACCOs have the lowest.

THE NUMBERS: PERCENTAGES

The proporton

usng MFi accounts

s only 4.5%

66% depost at least

monthly, but only

2% more requently

and 39% receve

automatc payments

SACCOs Easy

access to loans; 31%

had outstandng

loans (only 9%

or bank account

holders)

4.5% 66% 31%

SACCOs provide reliable services with goodaccess to loans

Reliability: SACCO are on the whole good at delivering the promised

product. In terms o saety, we encountered one case where a SACCO

had collapsed and respondents had lost their unds. However, cost is

seen as more reasonable than the banks in relation to the amounts o

unds that clients are managing, but complaints about charging suggest

that there is some lack o understanding and weak transparency.

Flexibility: Most SACCO FOSA accounts allow or fexible withdrawals

but some retain a limit to the requency o transactions though this

was not cited as a problem by respondents. In terms o access toloans that enable cash fows to be managed, this is certainly stronger

than or banks with the ability to take loans or school ees and other

emergencies being strong.

Structure: By being a channel or produce payments some

respondents cited using this as a means to accumulate sums and leave

them or a couple o months beore withdrawal. In addition by making

automatic deductions or share purchase rom payments they can also

assist building o shares against which loans are made however, this

is again seen as problematic or those receiving very small payments.

Convenience: In terms o convenience, we have interestingly ound

that those who are urther away use them more oten, suggesting that

these provide an important service in payment process despite the

problem o proximity. Speed o service was not an issue, and privacy

is again good.

Members o the Krnyaga Dstr ct Farmers SACCO Socety Ltd, beng served at the Kagumo branch.


9/12


TRANSACTIONS SIzES

A comparison o the amounts users transacted in these services shows that the

median amounts that are withdrawn rom banks, MMT and groups are all under

Kshs 4000. Comparing amounts deposited indicates that medians or banks and

SACCOs are in the region o Kshs5000 while those or groups are Kshs100, and or

MMT are Kshs1500. By comparison median loans rom banks, SACCOs and MFIs

are in the region o Kshs30,000-40,000 compared to ASCA loans o Kshs2000.

While such large amounts may be needed rom time to time, the more requentneed is to manage cash fow or these smaller lump sums which the data suggests

are Kshs5000 or less. By contrast to the inormal sector, the ormal sector does not

eectively cater to loans o this level.

THE NUMBERS: MEDiAN AMOUNTS

Ksh 4,000Medan amounts wthdrawn rom banks and SACCOs

are under Kshs 4,000

Ksh 2,000Medan se o ASCA loans and ROSCA payouts

Ksh 30,000 - 40,000Medan loan se or banks, SACCOs and MFis

Efective interest rates by sector ith and ithout equity deposits

Loan sie

Ksh


10/12


Financial inclusion operates at an individual rather than ahousehold level

Finally, investigation o the intra-household income, expenditure and

nancial management show that men and women tend to manage income

streams independently to meet particular expenditure responsibilities.

Income pooling and joint management most oten occur when income is

jointly earned.

Financial service use tends to be similarly independent and consultation

and co-operation is most likely to occur when household assets are used

as collateral or repayments aect the ability o a partner to ull their

expenditure responsibilities.Along with the evidence above on the gendered use o ormal and inormal

nancial groups, this suggests the importance o approaching nancial

inclusion through a gendered analysis, since transactions costs will need to

be much lower or ormal inclusion to be aordable or women.

CONCLUSION THE RIFT REVEALEDThis evidence suggests that it is groups and inter-personal transactions that

oer the greatest fexibility in response to the cash fow management needs

o poor people.

Money transer now acilitates transactions which enable cash fow

management more eciently within existing social networks. So should

someone have assisted with Kshs300 last week and they are in need o

Kshs500 this week then it is likely that such ass istance could be orthcoming.

Hence by circulating unds through these inormal networks savings are

used to secure both nancial support and social connections. Given the

diculty o getting loans rom banks, especially o the relatively smallamounts o a ew thousand shillings that are needed to address the kinds o

cash fow needs people have on a requent basis, they are not an attractive

place to put unds when they are available. Putting a ew hundred shillings

in the bank results at best in getting less back than was put there in the

rst place due to charges, and secures neither the potential or additional

nancial support nor social connections in the process.

Hence the rit that MMTs popularity has revealed is that between the way

the ormal and inormal sectors operate. The extensive array o inter-personal

transactions that it has brought to light operate in circuits o give and take

which eectively allow or saving and borrowing in inormal mechanisms

- that is, with other people. Inormal groups operate with some similarcharacteristics and underline the extent to which transactions are embedded

in social networks and relationships.

The task or ormal nancial service providers is to nd ways to learn rom

the eatures o inormal services that are highlighted here in terms o 'give

and take' and negotiation in order to provide services that will better respond

to their needs.

Median desposits, ithdraals and balances held, diferent types o services

Last Deposit Last withdraal Balance held

Bank 5000 4000 4000

SACCO 6000 3000 1750

ROSCA* 100 2000 n/a

ASCA* 100 2000 n/a

Secret 250 300 300

Money transer 1500 1100 300

NB: * - median across all memberships in a group with an ASCA or ROSCA unction rather than by individual.

The Gthura branch o Famly Bank, Narob.


11/12


Implications or nancial inclusion - a credit-led savings strategy

To date the discourse on nancia l inclusion has been savings-led. That is, it has

proposed that fexible transactions accounts which are low cost and convenient

will oer sucient cash fow management benets to drive inclusion. The

evidence presented in this report suggests that even i accounts were very low cost

and convenient this would not produce the revolution in inclusion desired. Rather

it suggests that there is an under lying logic o give and take and social connections

in peoples transactions which this approach ails to recognise. Instead then, a

strategy o credit-led savings is proposed: that is that depositors will be attracted

to deposit unds in ormal institutions i an acceptable reciprocating borrowing

proposition is clearly in place.

The logic o this is two-old. First, is the give and take nature o transactions.

Bankers were very clear that people expect a acility and that the lack o this deters

depositors. For users, access to borrowing embodies a relationship with a nancialinstitution without which there is no reciprocal element. The ormal sector -

especially the banks still ace a big hurdle o perceived appropriateness and trust

or poorer clients. Overcoming this requires that banks reverse the paradigm and

put their trust in clients rather than vice versa.

Second, is the evidence o amounts needed. These data suggest that useul lump

sums or poorer people involve access to loans o a ew thousand shillings whereas

ormal providers are rarely interested in such small loans below Kshs10,000 and

yet it is these that really enable cash fow to be managed in relation to requent

demands or small lump sums. Access to such loans needs to be quick, easy and

at low cost. It might operate in the orm o a revolving credit acility, that oncepaid o gives access to a similar or incrementally larger amount to give incentives

over time or good perormance. While there are obviously signicant issues in

operationalizing such a strategy viably, the point being made here is that this

evidence suggests that providers are currently operating ar rom where demand

in the low income market really is.

Third, the gendered rit in use o the ormal and inormal sectors was pointed

out above. This was in terms o men being more likely to need ormal serv ices to

process payments or lump sums o income while women oten handle smaller

income streams and their income streams tend to be independently managed.In addition it was indicated that women may have stronger savings cultures

supported by nancial groups. Given these constraints, inclusion o women into

ormal sector services aces additional hurdles since transactions costs will need

to be much much lower and access to services must be very convenient since they

usually have lower mobility beyond the homestead compared to men.

In the current environment o concern about over indebtedness credit-led

savings might be seen as a problematic strategy. However, the scale o debt

being proposed here is in act relatively small scale and lower than most MFIs

have oered even or rst loans. Second, there is evidence rom other contexts

eg. rom SaeSave in Bangladesh, that loans given at account opening , can - overa period o years - turn into a net savings position with the customer rather than

vice versa.

This strategy is proposed as a means to stimulate thinking about what the

relational dimension o nancial services highlighted in this report might mean

or developing a dierent way o thinking about routes to inclusion. In terms

o the specic approach to products suggested here, this has the potential

benet o initially providing a route to small-scale loan access or consumption

smoothing, and lumpy expenditures such as school ees, agricultural inputs, stock

purchases and so on. It may be possible to incorporate methods o incentivising

saving into product design and to support the discipline o this. The point about

this is that there must be reliability o access in the relationship and clarity over

the understanding o the benets oered at dierent stages o the relationship.

Through this there is also the potential to build the inormation base on clients

that will enable them to access larger amounts in the uture i they have the need

and capability to do so.

Members o new Odowa group durng a Group Savngs and Loans assocatons (GSL) cluster meetng n Rachuonyo, Nyana provnce. FSD has been partnerng wth CARE

internatonal (Kenya) snce 2008 n a GSL development project n Kenya whose ams are to rene the basc GSL model and test the use o nnovatve channels or delverng

tranng and support to groups to mprove cost-efectveness.


12/12

[email protected] www.sdkenya.orgFSD Kenya is an independent Trust established to suppor t the development o inclusive nancial markets in Kenya

4th Floor Kenya Re Towers Of Ragati Road, Upper Hill P.O. Box 11353, 00100 Nairobi, Kenya

T +254 (20) 2718809, 2718814 M +254 (724) 319706, (735) 319706

FSD KF St Dpg

12-03-23 fin landscapes summary report

Documents