the impact of microfinance institutions in local financial markets: a case study from kenya

Journal of International Development

J. Int. Dev. 16, 501–517 (2004)

Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jid.1088

THE IMPACT OF MICROFINANCEINSTITUTIONS IN LOCAL FINANCIAL

MARKETS: A CASE STUDY FROM KENYA

SUSAN JOHNSON*

Centre for Development Studies, University of Bath, UK

Abstract: This paper looks beyond the direct impact of microcredit provision on users to

examine whether microfinance institutions (MFIs) have had wider impacts within the local

financial markets in which they are operating. It considers the potential for both competition

and demonstration effects on other financial providers. In the context of local financial

markets in and around the small town of Karatina in Central Kenya, supply side information is

used to investigate the key changes in provision between 1999 and 2003. The paper concludes

that changing macroeconomic conditions have been the main driver increasing competition

for middle and lower income clients and that few competition or demonstration effects

resulting from the MFIs are in evidence. Copyright # 2004 John Wiley & Sons, Ltd.

1 INTRODUCTION

The main emphasis in impact assessments of microfinance programmes has been to

examine the direct effects that the provision of small loans has had on the livelihoods and

wellbeing of poor clients. However, when supporting the entry of microfinance institu-

tions1 (MFIs) into financial markets, donors and academics have also considered the

potential impact on the financial market more broadly. Specifically on the behaviour of

other financial service providers, with the expectation that their entry might precipitate

innovation and hence further expand the supply of services to poor clients. However,

studies attempting to investigate such impacts are virtually non-existent. This research

under the ImpAct programme therefore set out to examine whether MFIs have had such

effects within financial markets.

This paper starts by reviewing the types of impact on financial markets that have been

discussed in the literature. The second section then sets out an approach to examining

Copyright # 2004 John Wiley & Sons, Ltd.

*Correspondence to: S. Johnson, Centre for Development Studies, University of Bath, Claverton Down, BathBA2 7AY, UK. E-mail: [email protected] refers to the provision of small scale savings and loan services. Many types of organization areoften referred to as MFIs including, for example, BRI in Indonesia and various Credit Unions. Here I use the termto refer to organizations providing microfinance services—usually microcredit—whose origins lie in NGOs, asthese represent a very specific form of intermediation where credit funds are—at least initially—externallysourced from donor grants.

these impacts. The third section reports results from a study of financial markets in and

around the small town of Karatina in Central Kenya to examine these impacts empirically.

The fourth section concludes.

2 MICROFINANCE INSTITUTIONS IN FINANCIAL MARKETS

The 1990s saw a paradigm shift in approaches to microfinance that moved the rationale for

interventions in credit provision from one of subsidized delivery to the need to build

healthy financial systems (Otero and Rhyne, 1994). This was the result of an increasingly

rich body of detailed research into informal financial arrangements converging with

insights from the new institutional economics and practical experience of lending to poor

people (McGregor, 1988). This resulted in a number of key departures from earlier

thinking. First, poor people can and will pay relatively high interest rates for loans and

their concern is for repeated and reliable access rather than cost. Second, poor people can

and do save, and practical experience suggested that ‘compulsory’ savings requirements

linked to loan access could provide funds for on-lending.2 Third, group-based methods

(regularly found in informal arrangements) could reduce transactions costs and had the

potential to ensure that large numbers of people could be reached with the services. These

developments gave rise to hopes that microfinance institutions could expand their out-

reach, reduce their costs as they grew in size, mobilize funds for on-lending independently

of donors, and hence become independently self-sustaining so providing services in the

long term (Johnson and Rogaly, 1997).

The idea of building sustainable microfinance institutions also converged with policy

debates regarding ‘financial repression’ and its implications for financial sector reform.

The financial repression thesis suggests that measures such as interest rate controls, high

reserve requirements and directed credit policies imposed distortions on financial inter-

mediation. Hence that the removal of interest rate ceilings, for example, improves the

quality and quantity of saving and investment in the economy. On the supply side higher

deposit interest rates mobilise savings since higher interest rates produce less of a bias for

present over future consumption. On the demand side, raising the interest rate improves

the efficiency of investment through deterring low-yielding investments that would

previously have qualified (and especially in the case where investments may have been

selected by administrative systems rather than price selection). With improved allocative

efficiency of the funds invested, output growth is generated that further stimulates savings

as incomes rise (Fry, 1997).

Along with the proponents of financial repression, subsidized lending both through the

state in the form of agricultural credit schemes and NGOs, had been criticized by the ‘Ohio

School’ who took the view that this was ‘undermining’ the market with cheap credit

(Adams et al., 1984). They argued that subsidized credit led to allocative inefficiency and

could erode the in-built mechanisms of indigenous lending practices (Nagarajan et al.,

1995). By contrast, the Ohio School supported building financial organizations that could

cover their costs and be financially self-sustaining thereby widening the market for

financial services in a sustainable way (Von Pischke, 1991).

2Although the capacity of poor people to save was recognized in this way, this did not lead to the availability offlexible savings services by MFIs. This is an area of current and ongoing debate, although the need for poorpeople to have access to savings services is increasingly recognized. See Rutherford (1998; 1999); Sharif andWood (2001).

502 S. Johnson

Copyright # 2004 John Wiley & Sons, Ltd. J. Int. Dev. 16, 501–517 (2004)

However, while subsidized credit to users is seen as distorting the market and no longer

good practice, cheap capital for the establishment of MFIs has been the norm. Using neo-

classically based infant industry arguments, Hulme and Mosley (1996—also Meyer et al.

(in this volume))—argue that subsidies are valid because the benefits of developing the

technology of lending to poor people involves externalities of knowledge which cannot be

internalised by the organization itself (Hulme and Mosley, 1996). However, the case made

is a general theoretical one and has not been applied to specific financial markets to

demonstrate the legitimacy of subsidies in a particular context. To do so would require the

development of criteria to be used in assessing what a ‘correct’ level of subsidy might be

which would avoid undermining already existing financial service providers (Johnson,

1998). Thus support to MFIs has been rationalised alongside thinking around liberalization

and in this sense moves towards removing the ‘distortion’ of subsidized credit to users

(Copestake, 1996). Yet their presence and promotion remains ‘distorting’.

The Ohio School has been a key player in the application of New Institutional

approaches to financial markets and has also considered the role of MFIs in developing

the rural financial market (RFM) however, they suggest that ‘the route to better RFM

performance is not well marked’ (Von Pischke, 1983, p. 12) but a well-functioning rural

financial market should have the following characteristics. It should:

* mobilize rural savings as well as disburse credit

* grow to meet expanding opportunities without the need for subsidies

* expand the array of vehicles for attracting savings

* offer varied and flexible lending terms and conditions

* have institutions which are healthy and expanding

* have active competition among formal and informal borrowers and lenders

* the costs of financial services should fall as a result of innovation

* the economically active population should have expanding access

* the capability of the RFM to take part in larger financial markets should grow.

The shifting of the financial ‘frontier’ (Von Pischke, 1991) is defined as the limit of the

activities of formal financial institutions or alternatively that it is a ‘limit that is pierced by

any financial innovation’ and in particular is expanded when poor people have their first

direct and sustainable interactions with the formal financial system. Von Pischke proposes

a four-level framework for review which looks at impact on borrowers; the intermediary

itself; financial market development and macroeconomy and macrofinancial situation. The

first level is the most familiar: impact on borrowers and their enterprises has been the main

focus of impact assessment work. The second level has been heavily emphasized by the

‘intermediary’ school (Hulme, 2000) which focuses on the building of sustainable

financial intermediaries.

Here we are concerned to examine the third level of impact on financial market

development. Von Pischke suggests that this can be established by asking the following

questions (Von Pischke, 1991, pp. 264–376) both at the level of the project or intermediary

and at the level of financial market structure:

* At the level of the project or intermediary:

� are project instruments innovative?

� did project instruments promote competition; have they proved catalytic, novel or

trivial?

MFIs: A Case Study from Kenya 503


* At the level of the structure of the financial market:

� What is the relationship of the project to the structure of the financial market and

changes in its structure?

� How did the project fare relative to other institutions in terms of market share?

� Have project interventions influenced other players and their pricing?

� What evidence is there that financial instruments have competed with non-

financial forms of savings and credit?

At the level of the financial market, von Pischke concentrates on the question of whether

projects have successfully lengthened the term structure of the market, i.e. whether they

have overcome the risks of borrowing short to lend long. Since, the role of a financial

intermediary is to overcome the risk of using funds taken in on relatively short-term

deposits and turn them into loans for much longer periods of time, the ability to

successfully manage liquidity and price risks is crucial. He proposes means of analysing

the MFI’s assets and liabilities to establish whether this is being achieved but also suggests

that where such refinements in valuation techniques have occurred they may be expected

to create demonstration effects that would further stimulate competition.

Additionally, he poses two further questions:

* whether injecting liquidity into tight financial markets has a greater impact on these

markets than where liquidity is not a problem?

* whether the intervention has successfully developed banking habits in terms of long

term relationships involving continuous deposits and repeated borrowing.

Von Pischke does not indicate the exact means of answering many of these questions.

However this approach to impact assessment in financial markets concentrates on improv-

ing their functioning as efficient allocators of resources for investment consistent with the

financial repression case. In line with the operationalising of New Institutional Economics

approaches, reducing transaction costs and managing risk are seen as the main indicators of

whether or not this has been achieved despite the inherent difficulties in their measurement.

It is clear therefore that interventions in the financial market through credit projects

established with subsidy have anticipated wider impacts on the structure and functioning

of those markets that go beyond the direct impact on users. While the ‘intermediary’

school has focused on evaluating the goal of sustainable provision, few studies have

attempted to investigate the wider impacts on financial market development. The

contributions reviewed here suggest two types of impact on market development. Mosley

and Hulme suggest that there are externalities of knowledge that will be of wider benefit to

the market as a whole, and this is reinforced by Von Pischke in suggesting that there may

be demonstration effects to other players of successful innovations. However, Von Pischke

particularly emphasizes the role of competition in creating incentives for innovation and

the creation of value through new financial instruments which breach the frontier. High

returns to innovation produce incentives to invest and in turn erode those returns—

fuelling Schumpeter’s cycle of ‘creative destruction’ which carries economic development

forward (Von Pischke, 1991, p. 201).

3 CONCEPTUAL APPROACH AND METHODOLOGY

Figure 1 presents the supply side of financial markets with respect to different categories

of financial service providers and helps to conceptualise the impact when NGO-originated

504 S. Johnson


MFIs enter these markets. First, is the direct impact that the provision of services has on

the livelihoods of the users of those services which is the main focus of most impact

assessment studies. Beyond these we can suggest that there are two domains of impact.

First, is the impact that the entry of MFIs has on other financial intermediaries. MFI

services might compete away business from other suppliers on the basis of price or service

quality and hence affect the volume of business and profitability of other intermediaries in

the market—a competition effect. Hence, this may affect the array of financial services

available in the market indirectly, for example (and as is often hoped) competing local

moneylenders out of business. Alternatively or additionally, other intermediaries may

change their own products and services to copy or mimic aspects of the MFI’s services—a

demonstration effect.

At the level of users, the competition effect is in operation when users of MFI services

change their use of other financial services, and hence we can consider changes in service use

from the perspective of clients also. This competition effect at the level of users might have

indirect impacts on non-clients also. For example, members who withdraw from a ROSCA

because they now have access to an MFI, may cause the collapse of the ROSCA and hence the

reduced availability of services for those who, for whatever reason, may not want or be eligible

to join the MFI programme. Also at the level of users, it is possible that MFI clients may

provide a demonstration effect to non-clients in terms of banking habits, for example, through

learning savings behaviour and credit discipline. This may, in turn, encourage non-members to

enter into financial transactions with other financial intermediaries.

This paper draws on supply-side data collected in Kenya to investigate whether there is

evidence for competition or demonstration effects operating at the level of financial

intermediaries as a result of MFI engagement in the market. It reviews information on

market shares and changing product mixes along with evidence from interviews with

managers and officials who were interviewed in-depth about the dynamics at work in the

market. Hence it uses respondents own reports to explain attribution. A first period of

Figure 1. Conceptualising impact in local financial markets



research was carried out in Karatina between 1999 and 2001 and supplemented by a period

of three weeks of fieldwork in February 2003 (see Johnson, 2003).

Karatina is a small town in Mathira Division of Nyeri District in Kenya’s Central

Province. Its location on the slopes of Mt Kenya means that it has fertile volcanic soils and

good rainfall, with no absolute poverty in Nyeri District when the rains are good

(Government of Kenya, 1998). However, along with the rest of Kenya, economic

conditions have deteriorated. National GDP growth rates fell from 4.6 per cent in 1996

to 1.4 per cent in 1999, and turned to negative 0.2 per cent in 2000 before recovering to

1.1 per cent in 2001.

The area’s strong agricultural potential has meant that Karatina is a small but vibrant

town and this is one of the reasons that four NGO-originated MFI programmes now

operate there, with Kenya Women Finance Trust being the most established, having started

its operations there in 1991. It also possesses a wide range of intermediaries including

banks and non-bank financial intermediaries3 (NBFIs), parastatals, savings and credit co-

operatives (SACCOs), and microfinance institutions. It was not possible to capture the

informal financial sector on the supply side in the same way. Karatina abounds with group

based financial arrangements, ROSCA and ASCAs operating in various ways—with

women in particular making extensive use of them (see Johnson, 2003). However,

moneylending as a commercial operation (i.e. not including friends and relatives) operates

on a very limited scale. What follows therefore attempts to assess the key changes and

dynamics among the financial service providers more formally defined.

4 FINANCIAL SERVICE PROVISION IN KARATINA, 1999–2003

4.1 Formal Financial Sector Intermediaries

Formal sector banks and non-bank financial institutions (NBFIs): four of the five main

banks with a national presence operate in Karatina and in addition there is a building

society4—Equity. In 1999 the commercial banks offered the usual range of current and

deposit accounts, along with loans usually secured against land, shares or cash. In 1999,

these accounted for 73 per cent of deposits by value, 55 per cent of outstanding loans and

49 per cent of the number of savings accounts (see Table 1).

By 2003, the number of accounts had fallen by 24 per cent although their proportion of

the total remained almost the same. Deposits remained stable in nominal terms signalling a

fall in real terms as prices rose 17 per cent over this period. The primary cause of the

falling number of savings accounts was increasing minimum deposit levels over the period

1999–2001 when banks decided to raise their minimum deposits and move their business

towards a high net worth customer base—primarily the business market and salaried

individuals. Customers heavily resented this and it distanced the commercial banks from

the mass market since even low-level salaried government employees could no longer

afford to leave the minimum balance in their accounts untouched. In their search for

3During the 1990s the Central Bank of Kenya brought the regulatory requirements for NBFIs into line with thoseof the commercial banks after a series of banking crises.4Building Society registration originally required lending for long-term housing purposes but during the 1990sthe Central Bank has treated this registration as little different from that under the Banking Act. However, itcannot operate cheque accounts.

506 S. Johnson


Tab

le1

.S

avin

gs

and

loan

sp

erfo

rman

ceo

ffi

nan

cial

inst

itu

tio

ns

inK

arat

ina,

Ken

ya

19

99

and

20

03

(Ksh

s’000s)

Note

Dep

osi

tsM

ember

s/sa

vin

gs

acco

unts

Loan

sL

oan

/dep

osi

tra

tio

19

99

%o

f2

00

3%

of

19

99

%o

f2

00

3%

of

19

99

%o

f2

00

3%

of

tota

lto

tal

tota

lto

tal

tota

lto

tal

19

99

20

03

Fo

rmal

sect

or:

Ban

ks

11

14

85

93

73

10

93

70

36

52

45

43

24

17

37

42

34

29

99

55

55

37

09

95

53

74

9

NB

FIs

21

13

97

37

16

00

00

10

25

66

32

52

10

00

28

70

99

59

82

00

08

62

51

Par

asta

tals

30

00

00

00

29

961

42

9961

3

Su

b-t

ota

l1

26

25

66

80

12

53

70

37

55

02

06

49

38

37

45

15

30

95

16

86

49

06

06

74

25

2

MF

Ise

cto

r:

Mai

nst

ream

MF

Is4

18

629

12

4947

11958

22911

42

8411

45

0500

5153

202

Man

agem

ent

543

184

32

6628

21

0329

10

5790

88

2050

10

29

919

3190

112

serv

ice

MF

Is

Su

b-t

ota

l6

18

13

45

15

75

31

22

87

12

87

01

12

11

04

61

14

80

41

98

17

91

56

SA

CC

Os:

Cas

h-c

rop

SA

CC

Os

61

68

44

91

12

35

10

21

43

72

83

36

25

25

03

45

34

95

71

12

66

21

23

24

8

Em

plo

yee

SA

CC

Os

79

01

50

61

47

34

39

22

77

22

04

83

80

11

71

01

14

25

71

28

97

8

Tra

nsp

ort

/bu

sin

ess

SA

CC

Os

81

25

60

11

33

73

13

96

03

96

11

14

48

11

21

70

19

19

1

Su

b-t

ota

l2

71

15

91

73

95

81

82

43

99

56

39

27

69

43

71

45

06

01

82

39

08

82

55

36

0

To

tal

15

76

90

91

01

16

76

14

91

01

10

24

49

10

07

47

68

10

07

86

47

11

00

96

85

67

10

0

Notes:

1.

Dat

afo

ro

ne

ban

kin

20

03

was

no

tav

aila

ble

soit

sn

um

ber

so

fac

cou

nts

and

vo

lum

esh

ave

bee

nas

sum

edto

hav

ech

ang

edco

mp

arab

lyto

the

oth

erb

ank

s.2

.D

ata

for

thes

ein

stit

uti

on

sis

inco

mp

lete

.3

.D

ata

for

Po

stO

ffice

Sav

ing

sw

asn

ot

avai

lab

leas

bal

ance

sar

en

ot

hel

dat

ab

ran

chle

vel

.4

.M

ainst

ream

MF

I’s

dep

osi

tsar

ein

the

mai

nm

ob

lise

db

yth

eM

FI

bu

td

epo

site

din

the

ban

kso

are

excl

ud

edfr

om

tota

ld

epo

sits

.5

.T

hes

eto

tals

hav

eb

een

esti

mat

edb

ased

on

aver

ages

from

asa

mp

leo

fg

rou

ps

of

on

eo

fth

eo

rgan

izat

ion

s.6

.D

ata

for

som

eo

fth

ese

inst

itu

tio

ns

was

esti

mat

edin

19

99.

7.F

or

20

03

dat

ao

nsm

all

SA

CC

Os

for

wh

ich

up

dat

edin

form

atio

nw

asn

ot

coll

ecte

d,

dep

osi

tg

row

thh

asb

een

assu

med

atap

pro

x1

0p

erce

nt

per

yea

r;w

ith

mem

ber

ship

un

chan

ged

.8

.F

or

20

03

dat

ao

ntr

ansp

ort

/bu

sin

ess

SA

CC

Os

was

no

tco

llec

ted

—d

epo

sit

gro

wth

at5

per

cen

tp

ery

ear

and

stat

icm

emb

ersh

iph

asb

een

assu

med

.



affordable places to save, clients moved to the Equity Building Society and to front office

services (FOSAs) opened by the SACCOs (see below).

By early 2003, this had started to change. All the main banks had lowered their

minimum deposits significantly, precipitated by the fact that all of them had been

chasing high net-worth individual customers who were increasingly scarce given recession

and retrenchment in both public and private sectors. Their revised products have been

accompanied by aggressive marketing campaigns with the previously rare event of bank

officials visiting businesses and individuals around the town on a door-to-door basis.

On the lending side, in 1999, collateral based lending—especially against land—was

very limited in Karatina (Johnson, 2004). Land registration and titling was initiated in

Kenya in the 1950s and aimed at creating a land market that would support the financial

market. This finding echoed those of other research that suggested there was little evidence

to show that formal land titling has increased the use of credit (Pinckney and Kimuyu,

1994; Place and Migot-Adholla, 1998). The reasons for this can be explained by the

political nature of the land question in Kenya alongside social and cultural constraints.5

Social constraints include the safety-net role of land and the fact that the agreement of

wives and other family members is required via Land Boards before men can formally use

land as collateral which brings a complex myriad of family and lineage relationships into

the transaction. These relationships also complicate bankers’ ability to repossess land in

the case of default (see also Shipton, 1992) to the extent that by 1999 one banker explained

that it was a ‘silent norm’ to avoid taking agricultural properties as collateral because

foreclosure was a ‘nightmare of a process’.

By 2003, one notable change was the emergence of unsecured personal loans offered by

the banks based on an assessment of cash flow through the newly revised current or

savings accounts, with many of the banks deliberately marketing these loans linked to

savings accounts. This has been assisted by the development of credit-scoring techniques

that enable decisions to be made quickly. However, these decisions are not usually made

locally but sent to Nairobi despite the fact that most of these banks are now online. This is

probably to protect the integrity of the credit-scoring process so that local staff cannot bias

the system. Earlier attempts to introduce such systems on-line at branch level failed for

this reason.

Overall these developments have raised the nominal value of lending by approx-

imately 22 per cent, at a time when there has been a 17 per cent increase in the price index

suggesting an increase of 5 per cent in real terms. This appears to have raised the loan to

deposit ratio from 37 per cent in 1999 to 49 per cent in 2003, which, while not substantial,

is perhaps significant given the state of the economy. The interest of the main commercial

banks in lending to the middle- to lower-income market has been largely precipitated by

the changing macroeconomic situation. With falling Treasury Bill rates they have no

longer been able to undertake riskless lending to the government and at the same time, this

has precipitated a new move to cash-flow rather than collateral based lending.

Among the commercial banks, it is the Co-operative Bank that has particularly focused

on developing microfinance products, as a result of liberalisation of the sector and the need

for product diversification (Bell et al., 2002). It developed a small business individual loan

product using stock and household assets as collateral but based on an assessment of cash

flows. It was aimed at clients at the upper end of the microfinance product range at the time

(1999). However, in the Karatina branch outreach of this product has remained quite low

5See Johnson (2004) for a more detailed explanation.

508 S. Johnson


although performance in other branches has been stronger. In addition, the Co-operative

Bank introduced a savings account for small savers with an (at the time) low minimum

balance, unlimited withdrawals and no fees. The account was initially marketed in a very

limited way since its profitability was uncertain. Moreover, its heavy promotion risked

undermining the normal savings account product whose minimum balances were being

raised during 1999 and 2000 alongside those in other banks. Subsequently, it has benefited

from the presence of the product in gaining customers from other banks during the ‘race to

the top’.

Compared to the commercial banks, however, Equity Building Society is aiming more

directly at the middle and lower end of the market and making unsecured loans up to a

ceiling of Kshs50 000 (approx. US$666) without the assistance of credit-scoring methods.

Collateral needed for loans above this amount is negotiated pragmatically and may, for

example, include the deposit of a title deed without it being formally charged. In this range

they have introduced a range of products including medical, education, salary advances,

farm input and business loans. This wide range of products means that they are also the

only formal sector organization that is really competing at the bottom end of the market for

both poor personal customers and those with small businesses who are MFI clients.

Lenders with parastatal and government origins have been in decline in the late 1990s.

In particular the Agriculture Finance Corporation (AFC) has a huge bad debt portfolio

which it has been unable to restructure and closed its office in Karatina in 1999 and no new

lending had taken place over the previous two years. Other parastatals such as The

Industrial and Commercial Development Corporation (ICDC) and Kenya Industrial

Estates (KIE) have a very small presence in the area. This sector was estimated to account

for 4 per cent of the value of outstanding loans. Apart from the Post Office for which

deposit data was not available,6 none of these organizations mobilized deposits.

4.2 NGO-originated Microfinance Institutions

The NGO based microfinance sector can be split into two sub-categories. First are what I

call the ‘mainstream’ MFIs which had donor or other external support. Kenya Women

Finance Trust was the biggest of these programmes and has its Mt Kenya Region Branch

in Karatina. FAULU, and the Small and Micro Enterprise Programme (SMEP—whose

origins lie in the National Council of Churches of Kenya) are the next largest programmes.

Finally, K-REP whose origins are an NGO based MFI became registered as a bank in

1999, but is treated here as part of the MFI sector. Since it suffered difficulties in Karatina

when it first went to work there it has expanded only cautiously.

A further microfinance model is the managed ASCA model operated by three

organizations in 1999, but many more by 2003. In this model, women form groups and

on-lend their own savings to each other in the form of a revolving loan fund (RLF) which

is essentially an Accumulating Savings and Credit Association (ASCA). They meet

monthly and the role of the NGO is to provide management services to the group. These

organizations are entirely independent of donor funding. This model is characterized by

high interest rates on short-term advances (called gubasho)—at 10 per cent per month.

However, longer-term loans are also offered at much lower rates for a period of a year or

6Data for savings held by the Post Office were not available because accounts are held centrally and not on abranch basis.



more. A key feature of this model is that the high interest on these short-term loans

accumulates in the women’s fund and is used to pay out dividends at the end of the year

which can be quite high—20–60 per cent of savings, but this of course depends on the

overall performance of the fund. If the fund does perform well, then these dividends also

substantially reduce the net interest rate being paid on loans.

Table 1 indicates that the mainstream MFIs accounted for approximately 1 per cent of

deposits by value, 4 per cent of outstanding loans and 2 per cent of savings accounts in

1999. By contrast, estimates for the managed ASCA organizations suggest that they

accounted for 3 per cent of deposits by value, 10 per cent of loans outstanding and

10 per cent of the number of savings accounts. This is somewhat surprising, as the managed

ASCA model is little known (see (Johnson et al., 2002)). It is fully sustainable—indeed it

grew out of the need for a local NGO operating with women’s groups to survive after

donor funding was withdrawn in 1994. The excess of loans over deposits (see loan to

deposit ratio in table one) reflects the accumulated margin in the fund between interest

earned and dividends on loans. According to these figures its outreach is greater than the

mainstream MFIs—who have developed with considerable donor funding although their

operations in the Mt Kenya Region are in general now financially self-sustaining. However

by early 2003, mainstream MFIs had increased their membership by 49 per cent over 1999

representing some reasonably aggressive expansion during the period.

The main development in the products of the mainstream MFIs has been the introduc-

tion of school fees, emergency and medical loans usually at similar interest rates and very

similar terms and conditions and similar repayment frequencies to the main working

capital loan products. In addition, most of the MFIs are experimenting with or gradually

introducing individual lending products. This changing product mix reflects a new interest

in listening to clients and responding to their needs, in part in response to competition

among the MFIs. This has been accompanied by a concern to improve service quality and

customer care. They all reported initiatives or intentions to cut their application to

disbursement times, some have paid attention to the quality of records reaching the group

on a weekly basis, along with emphasis on the honesty and integrity of staff. KREP is the

only MFI not to have introduced medical, school fee or emergency products. Rather it has

used its newly acquired banking status to experiment with going up market with individual

lending against collateral. Unfortunately, this strategy resulted in poorer repayment

performance than in their familiar microfinance market and they have now decided to

concentrate on the market they know best.

While membership increased by 49 per cent between 1999 and 2003, the outstanding

loan value of the MFIs increased by 77 per cent. However outstanding loan balance is

Kshs23 000 compared with Kshs20 000 in 1999, a nominal increase of 15 per cent in a

period when prices increased by at least 17 per cent.7 If this is used as a proxy for depth of

outreach, then this data suggests that the expansion has not tended to bring in poorer clients.

The managed ASCAs have been facing significant problems of portfolio management in

particular with the poor macroeconomic performance of the Kenyan economy. Poor

repayment has resulted in the stagnation of the revolving funds. It has been difficult to

fully establish their financial performance. The ability of the ASCA managers to enforce

repayment can be limited, costly and face objections from the members themselves.

Despite this, the model demonstrates that there is a large market that the MFIs with their

7The New Kenya Overall Price index indicates that prices rose by 17.3 per cent between December 1999 and June2002, (see CBK Statistical Bulletin June, 2002).

510 S. Johnson


concentration on small business lending have not reached. Table one suggests that

managed ASCA membership fell by 44 per cent. However, the 2003 figure reported here

is probably a significant underestimate of membership since a number of new organiza-

tions offering the service had been set up and in a number of cases these took groups from

the bigger organizations with them. A best guess would suggest that membership has

remained static or fallen slightly. However, these new organizations have also also taken

the model into new areas beyond Karatina.

The proliferation of this model to new organizations in and beyond Karatina is evidence

of the low barriers to entry involved in its replication and its attractiveness to users. To

offer the service only requires knowledge of how the model operates and the printing of

some stationery along with sufficient funds to enable free service to be offered to groups

for a period of three months when they are first recruited. For the providers, this is usually

facilitated by paying a small commission to a field officer when she recruits a group

and only paying her a salary once the group is actually paying management service

fees. While the figures above reflect the outreach of the model in Mathira alone, estimates

for mid-2003 suggested that the model was reaching some 36 000 across Central Province,

reaching as far as Nyahururu and Nanyuki moving north, and Njoro moving south.

On the demand side, its extensive take up is because of the small size of savings and

loans that it can deal with, the ease with which loans are accessed at meetings, the

periodicity of meetings and the ability to negotiate repayment. This makes it useful to a

wider range of clients, in particular those in rural areas such as farmers, and those with less

steady income streams. The negotiability point is key. When people face livelihood shocks

they need to be able to use their social networks to either find funds to repay, delay

repayment or even negotiate a new loan. User-owned groups offer this possibility and

directly contrast to the mainstream MFI model (Johnson, 2004). In the user-owned model

the group can allow delays in repayment or a new loan because it is allowing its own funds

to be held for longer. In the mainstream MFI model the MFI uses the group to enforce

repayment of its own funds and there is little scope for negotiation as it is the other group

members who are forced to pay for the member who cannot.

4.3 The SACCO Sector

The third main sector is the co-operative sector. SACCO regulation was gradually

weakened over the 1990s, first with the reduced interference of government officers and

then these changes were eventually institutionalised in the 1997 Co-operative Act. This

liberalization has given rise to new SACCOs based on common bonds such as transport

operators or business people. Here we categorize the SACCOs with reference to the

common bond and highlight three types—those based in cash crops, employment and a

new and growing set of SACCOs mainly based in the transport sector among matatu

(public taxi) owners. As an example, Kenyan employee SACCOs operate with members

buying shares on a regular basis—usually deducted monthly direct from salary. These

shares can only be withdrawn on leaving the SACCO, and maximum loan amounts are

usually a fixed multiple of shares. In rural SACCOs, a single share is usually bought in

order to join and non-withdrawable deposits made as deductions from regular income e.g.

tea or coffee payments.

Two cash crop SACCOs based on the common bond of tea farming were started in the

1990s and are competing for the membership of the division’s approximately 6000 tea

farmers. The coffee co-operative system has been in a state of transition in the late 1990s



with the splitting up of the divisional level marketing co-operative covering some 25 000

farmers into 13 small co-operative societies. Further, as a result of government directives,

the Union Banking Section of the District level marketing co-operative has been

transformed into an independent Nyeri Farmer’s Society SACCO, which has a branch

in Karatina. The combination of low international coffee prices and this re-organization

has been the cause of relatively low lending volumes in the past three years, since many

farmers have had insufficient income from coffee to repay past loans, and the re-

organization has provided opportunities to avoid doing so as the links between the

factories and the SACCO weakened.

There are four small employee-based SACCOs in and around Karatina with between 50

and 300 members each, which are based in medium-sized institutional employers such as

the Town Council, hospitals and schools. Alongside these is the much larger Nyeri District

Teacher’s SACCO which caters for primary school teachers in the district. While there are

also members of national level SACCOs for teachers and government officials living in

Karatina, these were not captured in the supply side data.

The SACCO sector as a whole accounted for 17 per cent of the value of deposits,

18 per cent of the value of outstanding loans and 39 per cent of the number of savings

accounts in 1999. This sector therefore accounted for a relatively high proportion of the

number of savings accounts but a low proportion of deposits by value, suggesting much

lower average deposits than the banking sector and the fact that SACCOs—especially

those based in cash crops—tend to deal with poorer clients. By 2003, the SACCO sector

as a whole had suffered a fall in the total number of accounts of 21 per cent. However, this

decline was almost entirely attributable to the transformation of the coffee co-operative

Union Banking Section to SACCO as this now required farmers to buy shares in the

SACCO, whereas previously they were automatically given accounts through which to

receive their payments. However deposits in the SACCOs have increased by 46 per cent

over 1999 since they have been developing front office services (called FOSAs) with

deposit services for both members and non-members in open competition with the banks.

In Karatina there are now three SACCOs with branches offering these services and two are

in the same street as the banks where there were none in 1999. They offer computerised

services, can give their customers chequebooks (which even though not part of the clearing

system allows payees to come and draw cash and hence gives added convenience), and

offer services such as school fees cheques. It is difficult to gauge the numbers of accounts

that might have moved from the banks to the FOSAs. This is because many SACCOs have

front office accounts for members through which dividends, loans etc are paid but which

they may not actually use in the way they would a bank or NBFI savings account e.g.

having their salaries paid through them. However, approximate figures suggest this is

probably at least 5000 (approximately 15 per cent) since 1999 and this is probably a

conservative estimate since SACCOs such as Nyeri Teachers have been aggressively

marketing their FOSA services to their members. This development of FOSA services has

been unregulated to date, although supported by the Co-operative Bank which has also

provided technical assistance in many instances.

The banks and SACCOs are actively competing over salaried employees, such as

teachers and government workers, and tea farmers since these are the people who have

regular incomes in the area. As a result, the SACCOs are also changing their by-laws to

extend their common bonds to include these people and offering them unsecured advances

against income streams. There has been aggressive competition between the FOSAs and

Equity over this segment, but it is a segment that the banks are also interested in.

512 S. Johnson


4.4 Interest Rates

Providers quote lending interest rates in declining or flat rate terms and table two details

these rates for particular institutions in 1999. The main change in rates since 1999 has been

in the banking sector where base rates have fallen significantly. All other lending rates in

the market have remained virtually unchanged. In 1999, commercial bank base rates were

in the range 20 to 25 per cent but by the end of 2002 these had fallen to 14–18 per cent.

Treasury bill rates—which have been the main underlying determinant of bank base

rates—underwent further falls in the early part of 2003 to 7–8 per cent by February and

dropping further to 3 per cent by mid-2003.

The MFIs quote annual flat interest rates and in the late 1990s these were therefore—in

headline number terms—lower than bank base rates quoted on a declining balance basis.

Since few customers are aware of the difference, MFIs were perceived to be cheaper than

the banks. This is now changing. Since bank base rates are falling, some clients are now

aware that banks are quoting lending rates over than the MFIs. With this development,

MFIs are likely to start coming under much heavier pressure from clients regarding

interest rates.

5 CONCLUSIONS

The evidence presented here indicates that the mainstream MFIs have relatively limited

market outreach by comparison with other models both in terms of clients and volumes.

They have expanded outreach in a relatively well-off part of the country by almost

50 per cent8 in the last four years and at a time when the economy was performing poorly

but this seems to underline the outreach limitations of the model commented on by others

Table 2. Interest rates quoted and effective 1999

Financial intermediary Interest rate quoted Equivalent annual decliningbalance interest rate*

KWFT 22% flat per annum 82%

KREP 18% flat per annum 69%

Faulu 22% flat per annum 102%

SMEP 22% flat per annum 66%

Co-operative

Bank—Biashara loan 2.5% flat per month 61%

Equity 1.5% flat per month N/a

2% per month reducing N/a

Employee SACCO 12% declining per annum

(for a 3 year loan) 49%

Managed ASCA 10% per month on > 240%

outstanding balance

Bank (model case) 25% declining per annum 41%

*Effective annual rates have been calculated using an IRR method for a loan of Kshs20 000 over a year andinclude the effect of fees levied and compulsory savings where these are part of the product. They will inevitablychange when calculated for different amounts over different periods.

8The data in table one does not take into account multiple membership of MFIs (or similarly of multiple use ofother services). There was some evidence to suggest that over the period 2001–02 clients had taken multiple loansand over-exposed themselves with a consequent rise in default. MFIs in the area started to more systematicallyshare information on defaulters as a result.



due to the relatively inflexible design of the loan product on offer (Hulme, 1999; Wright,

2000). By contrast the little-known managed ASCA model has much wider coverage

although it has suffered in recent economic conditions.

The more dynamic players in the market have been the banks (particularly Equity

Building Society) and SACCOs. However, the key factor driving these developments has

been competition amongst them in the context of changing macroeconomic conditions and

SACCO deregulation. For the banks, a rush to target the high net worth individual retail

market led to its saturation, again in the context of macroeconomic decline in which public

and private sector employees were being retrenched. High treasury bill rates in the late

1990s and the problems of collateral-based landing—particularly against land—led them

to neglect lending to rural people. But this is no longer possible with the precipitous fall in

these rates over the last three years and they are again having to learn how to lend with

credit scoring and cash flow lending beginning to take hold.

Although evidence is limited, the view that mainstream MFIs are serving a different

market segment to the banks and SACCOs appears increasingly less defensible. While it

may be argued that neither the banks nor SACCOs are targeting poor people or business

people, it is not clear that MFI services are suited to or used by poor clients. A borrower

profile of five Kenyan MFIs including KREP, KWFT and Faulu (McGregor et al., 1999)

compared education and income profiles with those reported by the GEMINI survey of a

nationally representative sample of small and micro-enterprises. This showed that MFI

clients tended to be better educated and a significant proportion had annual incomes above

Kshs200 0009 compared to the national average—see Table 3.

Results of the CGAP poverty assessment tool for KWFT clients further supports the

view that MFI coverage is skewed towards the better off. The proportion of their clients in

the poorest tercile (i.e. 33 per cent by national standards) is approximately 16 per cent. In

the middle less-poor tercile is approximately 33 per cent with some 50.5 per cent of their

clientele in the better-off tercile (Microfinance Gateway, 2003). With this evidence and the

fact that MFIs are developing individual loan products in order to keep their older clients,

and given that MFI clients often have multiple identities, for example, as business people,

farmers and sometimes employees also, we can suggest that banks and SACCOs are now

more likely to be targeting the same clients.

In terms of demonstration effects, it is hard to identify any such effects of group-based

approaches to other financial intermediaries. While Equity experimented a little with

group lending in Karatina in 1999 this did not yield good results. Equity’s identity as a

microfinance provider is therefore not associated with a narrow understanding of group-

based lending or demonstration effects from the mainstream MFIs. Rather, it has been

engaged with donors and a wider discourse about how to move beyond this rather narrow

view of microfinance and has been improving its savings products as well as developing

loan products and mechanisms for unsecured lending. Being able to term this ‘micro-

finance’ seems to have in itself provided a rationale, focus and further motivation to serve

this market (see also (Coetzee et al., 2002). Similarly the Co-operative Bank’s move into

microfinance has been stimulated at this level. Indeed, it may seem redundant to consider

demonstration effects at this local level when the products and services of most

organizations—especially banks—are managed from headquarters. Recent debates

over the commercialization of microfinance suggest that moves by commercial banks to

‘downscale’ into microfinance markets result from demonstration effects at national rather9The overall poverty line in 1998 per adult equivalent was approximately Kshs23 808. The annual averageexchange rate for 1998 was US$1¼Kshs 60.4.

514 S. Johnson


than local levels (see Drake and Rhyne, 2002; Woller, 2002)). However, this study has

exposed providers and dynamics—especially the SACCOs and ASCAs—whose presence

and operations would tend to be missed in a national level analysis.

There is limited evidence of a demonstration effect regarding ASCAs. In an area where

group-based finance has long been practiced, local people realize that in ASCAs and

SACCOs rather than in MFIs the interest they pay comes back to them in the form of

dividends. This is one reason for the popularity of the managed ASCA model and, while

difficult to establish in quantitative terms, our research continually came across newly set

up independent ASCA groups. User-ownership offers a range of features that are attractive

to users. The negotiability of loans in these systems in the event of livelihood shocks was

pointed out above. When circumstances arise which result in repayment difficulties,

members know that they will have a range of options in dealing with the situation. In

SACCOs, loan products include school fees and emergency loans. In informal groups,

members have the opportunity to explain themselves (‘voice’) which can result in

negotiating an alternative payment schedule; in some cases the member may be given

an additional loan to overcome the circumstances; or that the member can use the social

networks that arise in the group to gain support—whether that support is financial or

social and whether from the group or individuals. In addition, members set the interest rate

on loans and know that it will not change without them knowing or voting for it—a point

that was particularly important during the late 1990s when both economic conditions and

bank rates were volatile and uncertain. Finally, when borrowing, the member is usually

risking only her existing savings, future income (from salaries or cash crops) or possibly

household items—land is not at risk (see Johnson, 2004).

The main conclusion is therefore that MFIs have had only very limited demonstration or

competition effects on other players in the financial market. Rather, it is becoming

increasingly clear that there are a number of ways in which the wider dynamics of the

financial market are having, and are likely to continue to have, significant impact on the

MFIs. The changing macroeconomic environment has put pressure on the formal sector

financial intermediaries to change their strategy and consider how to tailor their products

to the middle-income market. In doing this they are finding new ways to lend primarily

based on assessment of income flows. Christen and Drake have, moreover, pointed out that

the ‘ultimate irony’ might be that it is the mainstream financial intermediaries that are

ultimately more effective at this provision than microfinance NGOs (Christen and Drake,

2002, p. 14).

Table 3. Comparative indicators for clients of 5 MFIs and a nationalsample of small and micro-entrepreneurs

Five MFIs GEMINI

Education (%):

None 4.2 20.4

Primary 37.9 55.3

Secondary 49.3 23.2

Post-secondary 8.6 1.2

% with annual income less than:

Kshs10 000 7.9 62.0

Kshs200 000 24.6 1.5

Note: The Gemini study used slightly different income ranges so the percentagesreported are for amounts less than Kshs12 840 and above Kshs240 000.Source: (McGregor, Alila et al., 1999).



Further, the familiarity and abundance of group-based financial systems in this area,

means that local people are also prepared to set up groups to respond to their financial

needs and recognise the advantages of doing this. In addition, as MFIs move to individual

loans—as they are under pressure from their clients to do—they are increasingly moving

to meet the banks on their own ground.

In Karatina’s financial markets, MFIs appear to be operating amongst competitors

increasingly interested in the same customers for their loan products who can also offer an

array of savings products. While there is no reason to think that clients will necessarily

stop borrowing from more than one provider, the loan terms and conditions (including

interest rates) are increasingly attractive elsewhere and lack the requirements for group

meetings and guarantees. The microfinance mantra that it is access rather than price that

matters, may no longer be true for the clients of these MFIs. In addition, the Microfinance

Bill may now become law within a year and provide a basis for the formal regulation of

MFIs by the Central Bank and hence enable them to intermediate deposits. The process of

conversion for those that decide to do so will present them with significant costs and work

to meet governance and financial requirements. In the meantime, competition for deposit

accounts is strong and players such as Equity Building Society and the SACCOs are

learning how to serve this market—and how to link loan products to their savings

accounts. After conversion, the MFIs may find that they are competing in a well-served

market with high service standards and must be ready for this. Might they become the

casualties of ‘creative destruction’?

ACKNOWLEDGEMENT

This paper draws on data gathered under two research projects. The first round of data was

collected between 1999 and 2001 with funding from the Finance and Development

Research Programme funded by the UK’s Department for International Development (see

www.devinit.org/findev). The second round of research was carried out in 2003 with

funding provided by the ImpAct programme (see www.ImpAct.org). I am grateful to

James Copestake and Gary Woller for comments on earlier versions of this paper. I am

particularly grateful to George Muruka and Edward Kinyungu who assisted with the field

research in 2003. All errors and omissions remain my own.

REFERENCES

Adams DW, Graham DH, von Pischke JD. 1984. Undermining Rural Development with Cheap

Credit. Westview Press: Boulder, CO.

Bell R, Harper A, Mandirenga D. 2002. Can commercial banks do microfinance? Lessons from the

Commercial Bank of Zimbabwe and the Co-operative Bank of Kenya. Small Enterprise

Development 13(4): 35–46.

Christen RP, Drake D. 2002. Commercialization: the new reality in microfinance. In The

Commercialization of Microfinance: Balancing Business and Development, Drake D, Rhyne E

(eds). Kumarian Press: Bloomfield, CT.

Coetzee G, Kabbucho K, Mnjama A. 2002. Understanding the Re-birth of Equity Building Society in

Kenya. MicroSave-Africa: Nairobi.

Copestake JG. 1996. NGO-Donor collaboration and the new policy agenda-the case of subsidised

credit. Public Administration and Development 16: 855–866.

516 S. Johnson


Drake D, Rhyne E (eds). 2002. The Commercialization of Microfinance: Balancing Business and

Development. Kumarian Press: Bloomfield, CT.

Fry MJ. 1997. In Favour of Financial Liberalisation. Economic Journal 107: 754–770.

Government of Kenya. 1998. First Report on Poverty in Kenya. Ministry of Planning and National

Development: Nairobi.

Hulme D. 1999. Client drop-outs from East African Microfinance Institutions. MicroSave-Africa.

Hulme D. 2000. Impact assessment methodologies for microfinance: theory, experience and better

practice. World Development 28(1): 79–98.

Hulme D, Mosley P. 1996. Finance Against Poverty: Effective Institutions for Lending to Small

Farmers and Micro-enterprises in Developing Countries. Routledge: London.

Johnson S. 1998. Programme impact assessment in microfinance: the need for analysis of real

markets. IDS Bulletin 29(4): 21–30.

Johnson S. 2003. Analyzing the role of gender norms in financial markets: evidence from Kenya.

Seminar Paper given at ISS, The Hague, 17th November 2003, mimeo.

Johnson S. 2004. ‘Milking the Elephant’: financial markets as real markets in Kenya. Development

and Change 35(2): 249–275.

Johnson S, Mule N, Hickson R, Mwangi W. 2002. The managed ASCA Model: innovation in

Kenya’s Microfinance Industry. Small Enterprise Development 13(2): 56–66.

Johnson S, Rogaly B. 1997. Microfinance and Poverty Reduction. Oxfam: Oxford.

McGregor JA, Alila P, Copestake J, Johnson S, McCormick D, Njoka J. 1999. The impact of

interventions in the small and micro-enterprise sector in Kenya. University of Bath: Centre for

Development Studies.

McGregor JA. 1988. Credit and the rural poor: the changing policy environment in Bangladesh.

Public Administration and Development 8: 467–82.

Microfinance Gateway. 2003. CGAP poverty assessment tool: assessment of client poverty levels—

KWFT.

Nagarajan G, Meyer R, Graham D. 1995. Effects of NGO financial intermediation on indigenous self

help village groups in the Gambia. Development Policy Review 13(3): 307–316.

Otero M, Rhyne E (eds). 1994. The New World of Micro-enterprise Finance. IT Publications:

London.

Pinckney TC, Kimuyu PK. 1994. Land tenure reform in East Africa: good, bad or unimportant?

Journal of African Economies 3(1): 1–28.

Place F, Migot-Adholla SE. 1998. The economic effects of land registration on smallholder farms in

Kenya: evidence from Nyeri and Kakamega districts. Land Economics 74(3): 360–373.

Rutherford S. 1998. The savings of the poor: improving financial services for the poor in

Bangladesh. Journal of International Development 10(1): 1–15.

Rutherford S. 1999. The Poor and their Money. University of Manchester: Manchester, UK.

Sharif I, Wood G (eds). 2001. Challenges for Second Generation Microfinance: Regulation,

Supervision and Resource Mobilization. University Press: Dhaka.

Shipton P. 1992. Debts and trespasses: land, mortgages, and the ancestors in Western Kenya. Africa

62(3): 357–388.

Von Pischke JD. 1991. Finance at the Frontier: Debt Capacity and the Role of Credit in the Private

Economy. World Bank: Washington DC.

Woller G. 2002. The promise and perils of microfinance commercialization. Small Enterprise

Development 13(4): 12–21.

Wright GAN. 2000. Microfinance Systems: Designing Quality Financial Services for the Poor.

University Press: Dhaka.



the impact of microfinance institutions in local financial markets: a case study from kenya

Documents