Content uploaded by Peter Muriu
Author content
All content in this area was uploaded by Peter Muriu on Jul 06, 2017
Content may be subject to copyright.
Public Policy and Administration Research www.iiste.org
ISSN 2224-5731(Paper) ISSN 2225-0972(Online)
Vol.7, No.6, 2017
17
Credit Access and Savings Mobilization: Evidence from Kenya
and Tanzania
Dunstone Ulwodi
*
Peter Muriu
School of Economics, University of Nairobi, Kenya
Abstract
This paper examines access to credit and savings mobilization in Kenya and Tanzania. The economies of these
countries are closely interlinked and any shocks to one economy could have a contagion effect on the other.
While savings and access to credit play a significant role in securing investment, available evidence shows low
savings rates in Kenya and Tanzania. The paper utilizes cross-sectional data obtained from the FinAccess and
FinScope surveys of 2013 in Kenya and Tanzania respectively. The Seemingly unrelated regression equations
(SURE) approach is used to develop a policy framework for intervention. Estimation results reveal that
socioeconomic and demographic factors are key determinants of access to savings and credit in the two countries.
Specifically, education, income, age, social capital, location and gender were found to be the most significant
determinants of access to and use of credit and savings products in both countries.
Keywords: Access to credit, Savings mobilization, Seemingly unrelated regression
1. Introduction
Access to and use of quality financial services by both households and firms is of increasing concern to policy
makers across Africa. Although most African countries have undergone extensive financial reforms in the last
two decades, their financial sectors remain under-developed relative to the standards of industrial economies or
even other developing countries (Alter and Yontcheva, 2015). A recent global database shows that financial
exclusion is more pronounced in Sub-Saharan Africa where less than one in five households has access to formal
financial services (Demirguc-Kunt, et al., 2015). This paper investigates access to credit and savings
mobilization in Kenya and Tanzania. Significant improvements have been made in Kenya and Tanzania where
FinAccess data 2016 and FinScope data 2013 shows that access to formal financial system stands at 75 percent
and 57 percent respectively. Despite the significant increase in access to financial services, access to credit and
savings mobilization remain low. A vast majority of small scale savers continue to be deprived of affordable and
secure credit in a stable financial system. In addition, selective credit policies prevailing in most developing
countries have not only hindered savings mobilization from domestic sources but also increased inefficient
allocation of scarce capital (Donkor and Duah 2013).
The level of domestic savings has remained low despite concerted efforts aimed at addressing this
situation. Figure 1 shows the trend in gross savings as a percentage of GDP between 2000 and 2014. Tanzania's
savings strategy has yielded better results than Kenya. On the other hand, access to credit derived from adults
with credit from formal institutions is at 7 percent for Tanzania and 10 percent for Kenya. These low levels of
credit are not desirable if meaningful growth targets envisaged in the Vision 2030 for Kenya and Vision 2025 for
Tanzania are to be achieved
Figure 1: Gross Savings as percentage of Gross Domestic Product in Kenya and Tanzania
Source: World Bank Statistics, various years
Public Policy and Administration Research www.iiste.org
ISSN 2224-5731(Paper) ISSN 2225-0972(Online)
Vol.7, No.6, 2017
18
It is therefore imperative to examine how access to credit and savings mobilization between Kenya and
Tanzania could be harnessed in order to create a stable environment where the two economies can be bolstered.
Kenya and Tanzania form natural partners in regional economic co-operation and therefore assessing their areas
of strength would help promote development agenda in a more consistent manner. There is scant empirical
literature that exists comparing the levels of financial inclusion in the two countries. This paper makes use of the
demand side data comparing access to credit and savings mobilization in Kenya and Tanzania in more detail.
These facts raise the following questions: What factors explain access to and use of credit and savings
mobilization in Kenya and Tanzania; what are the marginal effects of socioeconomic and demographic factors
on access to savings and credit mobilization among households in Kenya and Tanzania? This paper makes
several contributions to literature and policy. First, it compares and contrasts socioeconomic and demographic
characteristics of access to savings and credit in a two country context. The findings add to the growing work in
this area (Ngendakuriyo 2014; Ellis et al., 2010). Additionally, the paper demonstrates how two separate
datasets could be enjoined to derive consistent indicators of financial inclusion. Most of the studies in this area
have succeeded in linking financial sector measured using macro-level variables to growth (Calderon and Liu
2003). Others including (Honohan 2005) have sought to establish the relationship between financial sector and
welfare improvement. Comparative studies on access to financial services have failed to show why different
countries adopt different strategies despite the fact that their economies have more or less similar socioeconomic
characteristics. In addition, rigorous analysis is often impeded by the lack of appropriate data that is amenable to
cross country comparisons; the data compilations used in this paper offer a unique opportunity to explore
financial inclusion from a multifaceted standpoint. Despite similarities existing between the two countries,
interventional strategies to address the problem of access to credit and savings mobilization are dissimilar. This
paper therefore makes unique contribution by showing the optimal relativity of access to credit and savings
mobilization required to improve household welfare.
2. Previous evidence
Factors influencing access to credit have been compared using two rounds of survey at two different points in
Kenya. Using age, gender, household size, education, marital status, distance and income, Mwangi and Sichei
(2012) provide evidence in support of the quadratic relationship between age and access to credit. Their results
showed that access to credit increased among middle aged individuals but began to decline as individuals
approached retirement age. This study however, did not make use of other variables considered critical such as
social capital. However, the age-access nexus has been authenticated by Johnson and Morduch (2007), Le Blanc
(2014), Bhuiya et al., (2001) and Diagne (1999). In Le Blanc (2014), age is found to limit access to credit while
in Ngendakuriyo (2014), an expanding age raises the likelihood of holding savings product. Bovenberg and
Evans (1990) identify aging population as lowering the likelihood of saving because such population is in its
second period whereas Foley and Pyle (2005) found the young and elderly population to have a higher likelihood
to save than the middle aged population. Foley and Pyle (2005) used panel data collected over 5 rounds starting
from 1995 to 2001. They examined household savings rate in Russia during the transition period. The
transitioning period provided a limit to the applicability of the results and therefore deviates fundamentally from
the situation in normal period.
In a related study, Mwalughali (2013) used institutional data to assess the impact of membership on
household income in Malawi. The key per capita variables considered included land per capita, income per
capita and credit per capita. The findings showed that age of the household head, sex, and distance from the
location of the institution were key determinants of household savings. However, this study did not apply the
standard sampling approaches since it used the institution as a case study. The findings are consistent with
Akudugu (2013) who used the Global Findex dataset of 2011 to assess the determinants of financial inclusion in
Western Africa. Akudugu (2013) applied the logit model and found out that only two in five adults were
included in the formal financial services in Ghana. Age of household head, financial literacy, wealth class, lack
of documentation, distance to the nearest financial service provider, lack of trust in financial institutions, poverty
of money and social networks were key determinants of financial inclusion in Ghana. The study assumed non-
hierarchical modelling unlike Pailwar et al., (2010) who use the multilevel model to assess the impact of
membership of financial institutions on rural savings. Pailwar found out that such membership explained a large
proportion of the variation in rural household saving.
The role of gender in access to credit among urban slum residents in the Philippines was investigated
by Malapit (2012) using the survey data between 2002 and 2006. The study revealed that women were likely to
face credit constraints than men. The study focused on wealth and non-formal lenders who relied on the
reputation and history of borrowers to screen potential borrowers. They concluded that women targeted credit
interventions in urban poor settlements was necessary. The major weakness of the study was in the scope of
coverage which constrained its applicability to a wide range of settings. This study however is corroborated by
Mwangi and Sichei (2012) who found that men had a higher likelihood of accessing loans from banks and
Public Policy and Administration Research www.iiste.org
ISSN 2224-5731(Paper) ISSN 2225-0972(Online)
Vol.7, No.6, 2017
19
SACCOs, whereas women had a higher likelihood of accessing loans or credit from the microfinance institutions
and local shops. Credit constraints are likely to be higher in women because of discrimination tendencies by
potential lenders. Additionally, women are involved in social and economic roles such as homemaking, which
limit their ability to establish adequate lender networks from whom they can obtain credit.
Using a Tobit model to analyze major determinants of household savings, Mirach and Hailu (2014)
found out that the key determinants of household savings included age, sex, marital status and income. In this
study women were more likely to save than men. This is inconsistent with the previous results where men had a
higher likelihood to save than women. Women and men were found to have different propensities to save
because of the differences in their perception of risks and interest rates, as well as, other gender related factors
that influence savings behaviour. In this regard, women were more likely to save than men because of their
conservative behaviour in investment decisions. They prefer to hold their assets in fixed rather than liquid assets
(Mirach and Hailu 2014).
Mwangi and Shem (2012) concluded that effective marketing strategies are required to enhance
financial inclusion in terms of access to savings accounts and credit facilities. Marketing creates awareness on
the benefits and availability of various savings and credit products, thereby facilitating inclusion. This argument
is anchored on the fact that access to credit and savings products was influenced by several socio-economic
factors that if exploited through appropriate marketing strategies could positively affect the decision to adopt
savings and credit services. The socio-economic factors considered include age, level of education, ease of
reaching a savings facility, availability of M-Pesa services, wealth, and group affiliation. Ageing increased
access to savings and credit services in formal financial markets, but reduced access in the informal markets.
Large households were found to have high dependency ratios, which negatively affected access to credit and
savings.
Family size and composition have been found to produce uncertain impacts on access to savings and
credit. Families with large size and/or many children are less likely to save (Orbeta 2006; Loayza and Shankar
2000) consequently constraining access to credit. This behaviour is partly due to strong intergenerational links
existing in most developing countries (Gersovitz 1988) which lessen the need to save. Mwangi and Sichei (2012)
found that an increase in household size enhanced access to credit from a buyer of households’ harvest, but
reduced access in the formal financial system. This is attributed to adequate labour to produce surplus
agricultural products which could be used to access credit services.
Income facilitates access to credit and savings services, especially in developing countries where
financial inclusion is still low. Individuals with reliable sources of income are likely to have at least one savings
account to accumulate capital since they are likely to have surplus disposable income (Fernando 2007). In
addition, they can use their source of income as collateral to access credit from formal financial institutions.
Indeed Ike and Umuedafe (2013) in their study of determinants of savings in Nigeria found that the volume of
savings among rural farmers was determined by the amount of income obtained from farming and non-farming
activities. Thus, low productivity and limited access to credit were the main factors that constrained
accumulation of savings among farmers.
According to Mwangi and Shem (2012) wealth was used as proxy for ownership of assets such as
cattle and mobile phones is a key determinant of access since the wealth can be turned into cash and deposited in
bank accounts. Having a mobile phone, on the other hand, is important because it facilitates the use of M-Pesa,
which provides savings, credit, and money transfer services. Ngendakuriyo (2014) showed that wealthier
households and individuals with high levels of education had a high probability of saving in banks, SACCOs,
and microfinance institutions than in informal institutions and secret places.
Gitaharie et al., (2014) found that the probability of a household to access credit for businesses from
banks, non-bank, and individuals was influenced by income-related factors such as employment status and
poverty level. Banking public education programs also improved access by creating awareness on the available
and benefits of financial services in the country. Ownership of assets such as a house, computer, and a phone
also influenced access to credit. This suggests that wealth positively affects access to credit services since it can
be used as collateral in the formal and informal sources of credit. However, this is inconsistent with Malapit
(2012) who found that lenders in the informal sector depended more on the borrowers’ creditworthiness rather
than wealth to evaluate loan applications.
Mwalughali (2013) showed that land per capita had a positive effect on access to credit. This means
that having adequate collateral is a key determinant of access to credit. This is based on the fact that financial
institutions prefer to provide secured loans to avoid default risks, especially among poor customers with no
steady sources of income. A key challenge in providing credit based on collateral is that poor households,
especially those who reside in rural areas may not have access to collateral. As a result, they will remain locked
out of the formal financial system.
Individuals often consider the costs and the benefits associated with various credit and savings
products before making a decision to be included or not. High costs are expected to reduce access, whereas
Public Policy and Administration Research www.iiste.org
ISSN 2224-5731(Paper) ISSN 2225-0972(Online)
Vol.7, No.6, 2017
20
benefits are expected to improve access. Ngendakuriyo (2014) analysis of the determinants of household savings
mobilization in East Africa Community (EAC), established that transaction costs and deposit interest income
influenced individuals’ decisions to save in financial institutions and informal mechanisms such as secret places.
Additionally, larger households were more likely to save in informal institutions and secret places due to the
convenience of accessing the funds, low costs and lack of adequate funds to maintain regular savings accounts in
banks.
Mirach and Hailu (2014) showed that household’s propensity to save is determined by income,
inflation rate, and the interest paid on bank savings accounts. Inflation discourages savings by eroding the value
of money or income over time. Additionally, low interest rates on deposits can discourage the uptake of savings
products, especially during high inflation because savers will not be able to get value for their money. This
suggests that efforts geared towards improving the uptake of savings products should also address barriers such
as inflation, which make such products unattractive to the target population.
Empirical evidence shows that the higher the interest rate the higher is the likelihood that households
will have access to informal financial services (Kochar 1997; Mwangi and Ouma 2012). In the same vein social
capital measured by group membership is positively related to use of formal financial services. Vaessen (2001)
evaluated access to credit by the families and demonstrated that the moneylender and the household attributes
were essentials impacting access to financial services.
Mbuthia (2011) found that resource constrain was the main barrier to accessing a bank account. This
study supports the argument that access to regular income promotes uptake of savings and credit products.
Perception of the interest rate paid on deposits and the availability of formal financial institutions were also
significant determinants of savings. High deposit interest rates promote mobilization of savings since it rewards
savers for their efforts to save for the future.
A properly functioning credit market has to be established to facilitate access to credit products. This
means that the market should be large enough to avoid credit rationing, which often leads to high interest rates
and exclusion of the most vulnerable households and enterprises. According to Atieno (2001), small enterprises
in rural parts of Kenya did not have access to credit due to rationing. In the informal financial sector, credit
rationing was attributed to financial resource constraints among lenders. By contrast, rationing in the formal
sector was attributed to lending terms and conditions which acted as barriers to credit access. Apart from
rationing, access to credit was limited by lack of information concerning the available credit facilities and the
cost of accessing them. This finding is consistent with Mbuthia (2011) who found that access to credit increased
with increase in availability of information concerning available credit products. Atieno (2001) also found out
that lack of the required security had a negative effect on access to credit. This supports Mwalughali (2013) who
found that availability of collateral in terms of land increased access to credit.
Location is another factor considered in literature although diverse measurement approaches have been
applied. The findings of how location impacts savings and access to credit are mixed. While some studies show
that location is a significant determinant of access to savings and credit (Kraay 2000, World Bank 1993), others
fail to attach this importance to location in the urban neighborhoods (Kulikov et al., 2007). World Bank (1993)
pointed out that savings and credit schemes with a wide network of branches lower the transaction costs for
micro-savers and offers appropriate and affordable credit facilities because of the location of the financial
institution. This is consistent with Akudugu (2013) who found out that distance had a negative effect due to the
high cost of accessing services in the absence of financial institutions within a 5-kilometre radius. Mwangi and
Ouma (2012) also found that distance had a negative effect on access to credit. This further underscores the need
to improve availability of financial institutions to enhance access to credit and other services.
According to King (2012), distance and time to the nearest banking infrastructure such as bank branch
or agents are significant barriers to financial inclusion in Kenya. Despite the significant expansion of bank
branch network, the negative effect of distance and time to banking infrastructure remains a challenge. This
suggests that a large number of Kenyans are still not able to access savings and credit products. However, mobile
banking and other digital financial platforms have increased access to credit and savings services by eliminating
the costs associated with distance and time to the nearest banking institution (King 2012).
Focusing on the Kenyan economy, Mbuthia (2011) found out that a household’s decision to save in a
formal, informal, or semi-formal institution is mainly informed by among other factors the information they have
about existing financial institutions, level of household income, and credit availability in financial institutions.
Savings are likely to increase if the target population has adequate information concerning the financial health of
the formal financial institution through which they intend to save. Information improves the trust and confidence
of potential savers. As a result, individuals will tend to save in institutions that are associated with minimal risks
to avoid losing their money.
According to Johnson and Zarazua (2009), financial perception has a significant effect on the decision
to adopt and use savings account and other bank services. Specifically, the perception that an individual can
survive without a bank account leads to a decline in the likelihood of accessing a general account. The decision
Public Policy and Administration Research www.iiste.org
ISSN 2224-5731(Paper) ISSN 2225-0972(Online)
Vol.7, No.6, 2017
21
to invest and access a savings account was affected more by financial behaviour than financial perception
(Johnson and Zarazua, 2009). However, financial behaviour did not have any effect on access to and use of
insurance products.
According to Mwangi and Ouma (2012), households are more likely to borrow from the informal
sector than the formal banking sector due to their perception of the interest rates charged in the two sectors.
Borrowers perceived interest rates in the formal banking system to be very high. Thus, they avoided borrowing
from banks and SACCOs. This calls for strategies that improve access to information concerning the interest
rates charged by various financial institutions to eliminate perceptions that hinder access to formal financial
services.
The business model adopted by financial institutions can significantly influence access to savings
account and credit facilities (Allen et al., 2012). An appropriate business model must allow the financial
institution to reach the unbanked segments of the population with relevant products to satisfy existing financial
needs, while maintaining acceptable levels of profits. In a study of access to banking services in Kenya, Allen et
al., (2012) found that Equity Bank had adopted a business model that focused on serving the populations that had
been locked out by the traditional banking system. This included individuals with low income, low levels of
education, without permanent houses, and without salaried jobs. As a result, the bank contributed to significant
increase in access to savings accounts and credit among households. The increase was attributed to the massive
branch network expansion program that the bank implemented to reach households in rural and remote locations.
The resulting reduction in distance to the nearest financial institution promoted uptake of savings account and
credit products.
Recently, microfinance institutions have included in their business models financial education program
to enhance uptake of credit and savings products. Participation in savings and credit institutions’ education
programs is expected to provide the knowledge and experience that individuals require to make the decision to
be included in formal financial system. Mwalughali (2013) supported this argument in his study which found out
that participation in community savings and investments programs in Malawi had a positive effect on access to
credit and household income. The study employed a probit model whose results showed that the decision to
participate in community savings and investments programs was determined by sex of the household head, land
per capita, distance to the nearest program’s office and credit per capita.
In their study of mobile phone financial services as a business model for delivering savings and credit
products, Mbiti and Weill (2011) concluded that M-Pesa significantly increased access to financial services in
Kenya by increasing competition, which in turn reduced transaction fees among competitors. They found that
frequent users of M-Pesa services were urban residents, highly educated, banked, and affluent. The use of M-
Pesa was associated with a decrease in the use of informal savings mechanisms. Additionally, it increased the
probability of being banked. This implies that M-Pesa complements bank services by increasing demand for
financial services. Although mobile money account provided opportunities for saving, it was mostly used to
transfer funds between individuals rather than to savings account (Mbiti and Weill 2011). This situation is
attributable to the higher opportunity cost of holding funds in mobile money savings account. The implication is
that M-Pesa is not likely to promote significant positive savings behaviour even if it paid interest rates on
deposits that are as high as those provided by commercial banks.
Access to credit can positively change saving behaviour among individuals and micro-enterprises.
Financial institutions prefer to lend to customers who already hold a savings account since they are likely to have
a better understanding of such customers’ financial needs and ability to repay loans. Rogg (2000) found that
increased access to credit motivated borrowers to shift their savings from less liquid assets such as livestock and
jewellery whose returns are often low or negative into deposit accounts in banks that guarantee positive returns.
One of the factors that promoted the shift in savings behaviour was increased availability of information and
improved confidence in existing financial institutions. The positive returns from bank deposits attracted savings
from the public, thereby increasing overall access to savings services (Rogg 2000). The increase in access to
savings accounts had welfare effects, which included increased income and access to credit facilities. In addition,
it promoted the development of the financial market by improving the depth and liquidity, which in turn
improved access to various financial services.
The empirical literature has shown that the main drivers of access to credit and savings mobilization
are age of the household head, gender, education, marital status, social capital, income, family size, location,
land per capital among others. However, the principal datasets used in this paper do not have variables such as
land per capita, family size among other variables. The paper therefore focuses on the variables available in the
two datasets namely the FinAccess and the FinScope surveys. In addition, the paper adopts the seemingly
unrelated regression equation model to jointly determine the drivers of access to and use of savings mobilization
in Kenya and Tanzania.
Public Policy and Administration Research www.iiste.org
ISSN 2224-5731(Paper) ISSN 2225-0972(Online)
Vol.7, No.6, 2017
22
3. Methodology
3.1 Theoretical Framework
The random utility theory is a characteristic theory of choice which has received proportionate support in
literature (Luce 1959, Marschak 1960, McFadden 1974). This study reviews the random utility model in the
context of households’ preference for financial services. It assumes that a variety of financial services are
available for consumers and therefore the choice alternative depends purely on the amount of utility derived
from their choice. The utility theory best represents the problem of the consumer as being to maximize utility
given the budget constraints. This trademark hypothesis of purchaser decision expects that people select the
choices that maximize their own utility. The econometric methods of random utility models are based on the
reviewed work of McFadden (1974). Subsequently probabilistic models have been derived from decisions
arising from alternatives available for each given scenario. Following the random utility model and assuming
rationality in decision making by economic agents, we assume that an individual i is a utility maximiser subject
to budget constraint. The utility associated with selecting an option j among the different choice options (access
to credit and/or access to savings) can be specified as follows:
MjNiXVX
jijjijijij
,......2,1;,.....,2,1,);();(
=
=
+
=
Ζ
ε
β
U
(1)
Where,
U
ij
(X
ij
;Z
ij
)
j
represents the utility enjoyed by individual i, from accessing saving or credit product of alternative j.
X
ij
represents the observed characteristics of individual i and choice alternative j.
Z
ij
represents the unobserved characteristics of individual i and choice alternative j.
V
j
(X
ij
;β) represents the utility deterministic component.
ε
j
is the utility random component
3.2 Empirical Model
The basic model assumes that utility is derived from accessing savings and/or credit products subject to several
factors. Among the factors that affect access to savings mobilization and credit include age, income, education,
marital status, social capital, location and gender. If we let U
ij
be individual i's utility of alternative j, then j takes
either 0 or 1. Subsequently, we assume that
ijijij
XvU
ε
β
+
=
),(
(2)
The terms are as described in equation 1. Assuming a utility maximization framework and linearity in parameters
of the cumulative density function, equation 2 can be expressed as a probit model as shown in equation 3;
iii
xy
εβ
+=
(3)
y
i
is the likelihood that financial mode i (where i is each of the result variable) will be selected. The compressed
form of the replica to be approximated thus takes the form:
),(
iiii
xyy
ε
=
(4)
Where, y
i
is the relevant dummy variable signifying access strands: formal, formal other, informal, and excluded.
xi represents independent household attributes in the FinScope and FinAccess datasets. ε
i
is the error term. In the
empirical model in equation 5, non-income factors mainly socioeconomic and demographic have been included.
s
IncometalSocialCapiEducation
tusMaritalStaLocationGenderLogAgeSqrtAgeFAccount
υβββ
β
β
β
β
β
β
++++
+
+
+
+
+
=
876
543210
(5)
FAccount is a binary measure of access to savings and credit account.
3.3 Seemingly Unrelated Regression Equations (SURE) Models
Equation 5 describes the behaviour of account ownership that facilitates access to savings and credit in response
to a set of explanatory variables. This is a comparative study of Kenya and Tanzania which involves multiple
equations with standardized variables. In each country, at least four equations involving four dependent variables
have been estimated. However, one approach to deal with multiple equations that explain different economic
phenomena may be to jointly estimate them. This is because such equations may have correlated error terms as
expected. When these conditions are met, the resultant equation takes the form of a seemingly unrelated
regression equations (SURE) model where separate linear equations are related. The system of equations
generated individually are assumed to be related and therefore the appropriate method was to estimate them
jointly. The paper considered the model comprising of four equations of the form:
∑
===+=
itiijtijti
kjMiTtxy ,....2,1;,......2,1;.....,2,1,
εβ
(6)
where y
ti
is the t
th
observation on the i
th
response variable which is to be explained by the i
th
regression
Public Policy and Administration Research www.iiste.org
ISSN 2224-5731(Paper) ISSN 2225-0972(Online)
Vol.7, No.6, 2017
23
equation, x
tij
is the t
th
observation on j
th
predictor variable appearing in the i
th
equation, β
ij
is the coefficient
associated with x
tij
at each observation and έ
ti
is the t
th
value of the random error component associated with i
th
equation.
3.4 Definition and measurement of variables
3.4.1 Dependent Variable
The response variable is defined by the four access strands namely: Formal, formal other, informal and excluded.
These strands encompass the access to credit and savings mobilization i.e. formal credit/savings, formal other
(credit/savings), informal (credit/savings) and excluded (credit/savings). Individuals choose savings and credit
products independently from the options given according to their maximization objectives.
3.4.2 Independent Variables
The independent variables considered in this paper have also been used in previous studies (Mirach and Hailu
2014; Malapit 2012; Johnson and Morduch 2007). However, the permutation of these variables is exclusively
defined in this paper. The variables under investigation include: age, income, education, gender, marital status,
location and social capital.
3.5 Sources of Data and Sample
The paper utilized the Financial Access (FinAccess) and FinScope survey datasets for 2013 for Tanzania and
Kenya. The Kenya sample clusters were selected to guarantee representation at all levels. At individual level, the
determination of the sample was done using KISH technique that randomly selected adult individuals. An
aggregate of 6,449 interviews were recorded. In Tanzania, an aggregate of 800 specification ranges were secured
both on the territory and Zanzibar Island. The aggregate meetings directed were 8,000 producing an aggregate
sample of 7,987 interviews.
4. Empirical results and discussions
4.1 Descriptive Statistics
Table 1 presents the summary statistics obtained from the cross-sectional data for Kenya and Tanzania. The
means and the standard deviations are shown. The difference in means was not large enough depicting similarity
in data.
Table 1: Summary Statistics for the single equation model
Variable Obs Median Mean Std. Dev. Min Max
Formal Savings 14436 0.00 0.096 0.295 0 1
Formal Other Savings 14436 0.00 0.076 0.264 0 1
Informal Savings 14436 0.00 0.212 0.409 0 1
Excluded Savings 14436 1.00 0.549 0.498 0 1
Formal Credit 14436 0.00 0.021 0.143 0 1
Formal Other Credit 14436 0.00 0.084 0.277 0 1
Informal Credit 14436 0.00 0.079 0.270 0 1
Excluded Credit 14436 0.00 0.217 0.412 0 1
Log Age squared 7127 7.37 7.393 0.779 6 9
Education 3860 2.00 1.926 1.299 0 6
Marital Status 7127 3.00 2.204 1.108 0 3
Social Capital 3639 0.00 0.135 0.458 0 3
Income 5395 1.00 1.253 1.292 0 6
Location 7127 0.00 0.450 0.498 0 1
Gender 7127 0.00 0.466 0.499 0 1
4.2 Determinants of savings mobilization
The seemingly unrelated regression equation (SURE) model was used preferably because it offered the best fit
for the available data. In this paper, a joint estimation of the models for Kenya and Tanzania was preferred
largely because of the similarities in the design of the data and the key drivers of access to savings mobilization
in the two countries. The two datasets used in this paper had different assumptions underlying data collection
methods but with many similarities that warranted a joint estimation. This was further made easy through
standardization of variables across datasets.
The findings are presented in Table 2. The log age squared is significant across all access strands
except formal savings strand. However, it is observed to be positive for formal other strand while negative for
informal access strand and excluded strands. Therefore, as individuals grow old, they are more likely to use
formal other access strand than formal access strand. This is seen in the relatively large coefficient for formal
other strand compared to other access strands. On the other hand, age appeared to reduce the use of informal
Public Policy and Administration Research www.iiste.org
ISSN 2224-5731(Paper) ISSN 2225-0972(Online)
Vol.7, No.6, 2017
24
access strands for savings and excluded strands. This indicates that older individuals would not find it prudent to
use more risky avenues to access savings. This view is consistent with the precautionary principle which states
that older individuals tend to consume what they saved over the years. However, Efobi et al., (2014), using the
Global Findex data set observed that age had a negative bivariate inclination to use of banking services. Camara
et al., (2014) investigated barriers to financial inclusion in Peru and found out that age appeared to be a
significant factor albeit with little importance. They further noted that the turning point existed at age 53 when
the relevance of age is suppressed. The findings further show that the risk involved in informal strand is
unwarranted for older members of the society and especially those living on pension and old age security
resources. In both countries age is strongly associated with access to savings mobilization from formal other
strand. However, age is not found to be significantly associated with formal access to savings.
Education appears to be a significant factor. Our findings show that education is positively associated
with formal savings access strand. More educated individuals will positively influence adoption of formal access
to savings. This is depicted by the large coefficients ascending from the lower levels of education to university
degree. Individuals with university education are shown to be more likely than those without education to access
formal channel of saving. Similarly, education was also found to positively influence formal other strand of
savings access. The coefficients increase in size as the level of education increases indicating that higher levels
of education has greater influence on use of formal other access strand of saving. Although education was found
to be associated with exclusion, the magnitude of influence diminishes as you move from lower levels of
education to higher levels. It should however be noted that university degree was not found to be significant with
exclusion. This implies that as individuals acquire more education, they are more likely to use formal savings
access strand than they would be excluded. The findings in this study however deviate from the findings in
Kulikov et al., (2007) that indicate that higher levels of education were more likely to lead to lower savings
because higher education reflects higher expectations in terms of income and less uncertainty in future incomes.
Kibet et al., (2009) corroborate the findings in this paper by demonstrating that the education level of the
household head positively affects the savings behaviour.
This paper also investigated the effect of marital status on access to savings. Marital status was found
to be positively associated with the formal other strand of savings access but negatively associated with
exclusion. The results further showed that marital status was not found to be significantly associated with formal
savings strand. Individuals with divorced status were 7 percent more likely to use formal other strand of access
to savings than those in single marital status. On the other hand the married individuals were found to be more
likely to access savings from the formal other strand than the single status individuals. According to Sinha
(2014), divorced women were more likely to obtain support of the children from the responsible man yet
develops her own mechanisms of fending for herself. This behaviour exposes her to a variety of financial
services including savings access that would offer collateral for the future. The paper shows that widows were
more likely to access informal savings by 6 percent. Singles or married individuals were less likely to be
excluded from accessing savings by a factor of 0.0662.
This paper also investigated social capital which has been defined as group membership. However, the
groups under investigation may not necessarily assume a formal organization. Observation on group
membership depict a positive association with access to savings at all levels except exclusion strand. Individuals
with membership in more than one group are more likely to have access to savings than those with no
membership. The findings concur with literature from both developing and developed countries. The strength of
the group coefficient is influenced by the amount of assets the group owns. The greater the coefficient and
positive it is, the more the social capital influences savings mobilization (Kulikov et al., 2007; Camara, et al.,
2014). However, social capital was found to be insignificant in relation to savings access exclusion.
Income in general was found to increase access to savings across all strands except the exclusion
strand. However, we observe that with increased income earnings, the coefficients for income tend to reduce in
the exclusion model indicating that increased income has a tendency to discourage financial exclusion. The
paper also examined the effect of location on access to savings. Location defined as rural or urban was found to
be significant for formal other and informal strands as well as exclusion strands. We observe that being urban
compared to rural increases access to savings from formal other and informal channels by a factor 0.035 and
0.0528 respectively but reduces exclusion by a factor 0.0409. This indicates that more urban dwellers are likely
to access savings than rural dwellers. This has led attributed to skewed distribution of financial services across
the country in favour of urban areas (Llanto 2015). According to Llanto (2015), many financial services favour
urban areas due to easier penetration arising from the high level of concentration.
The distribution of financial services almost always affects gender. In most developing countries rural
dwellers constitute the majority with less resource endowments (Addae-Korankye 2014). The analysis shows
that men are less likely to access informal savings compared to women. Similarly, the men have less likelihood
of being excluded from accessing savings compared to women. The findings are consistent with several other
studies which view women as greatly disadvantaged (Demirguc-Kunt 2013; Aterido et al., 2011, Asiedu et al.,
Public Policy and Administration Research www.iiste.org
ISSN 2224-5731(Paper) ISSN 2225-0972(Online)
Vol.7, No.6, 2017
25
2013).
Table 2: Results for the Seemingly Unrelated Regression Model for Savings mobilization
Variables Formal
Savings
Formal Other
Savings
Informal
Savings
Excluded
Savings
Log Age squared 0.00836 0.0380*** -0.0198** -0.0207**
(0.00656) (0.00731) (0.00825) (0.00981)
Some primary 0.0180 0.0554*** 0.104*** 0.111***
(0.0170) (0.0190) (0.0214) (0.0255)
Primary completed 0.0291* 0.0768*** 0.0884*** 0.114***
(0.0173) (0.0193) (0.0218) (0.0259)
Some secondary 0.0932*** 0.109*** 0.157*** 0.118***
(0.0232) (0.0258) (0.0291) (0.0346)
Secondary completed 0.152*** 0.171*** 0.168*** 0.0901***
(0.0202) (0.0224) (0.0254) (0.0301)
Technical after Secondary 0.244*** 0.190*** 0.111*** 0.0526
(0.0271) (0.0302) (0.0341) (0.0406)
University degree 0.296*** 0.332*** 0.0129 -0.0131
(0.0386) (0.0430) (0.0486) (0.0578)
Divorced 0.0133 0.0726** 0.0458 0.0447
(0.0254) (0.0283) (0.0319) (0.0380)
Widowed -0.00435 0.0344 0.0617** -0.0438
(0.0193) (0.0215) (0.0242) (0.0288)
Married/living together 0.0114 0.0487*** 0.0272 -0.0662***
(0.0149) (0.0165) (0.0187) (0.0222)
Belongs to one group 0.0423** 0.0129 0.392*** 0.135***
(0.0210) (0.0234) (0.0264) (0.0314)
Belongs to two groups -0.0277 0.110** 0.479*** 0.211***
(0.0397) (0.0442) (0.0499) (0.0594)
Belongs to three or more groups 0.0362 0.197*** 0.505*** 0.271***
(0.0523) (0.0582) (0.0658) (0.0782)
USD 1-15 0.0131 0.00377 0.0157 0.143
(0.0135) (0.0150) (0.0169) (0.0201)
USD 16-30 0.0887*** 0.0660*** 0.0505** 0.163
(0.0171) (0.0190) (0.0215) (0.0256)
USD 31-75 0.129*** 0.137*** 0.0724*** 0.156
(0.0199) (0.0222) (0.0251) (0.0298)
USD 76-150 0.115*** 0.220*** 0.120*** 0.180
(0.0299) (0.0332) (0.0376) (0.0446)
USD 151-300 0.174*** 0.259*** 0.173*** 0.105
(0.0352) (0.0393) (0.0443) (0.0527)
USD 301-1000 0.355 0.336 0.568** -0.411
(0.217) (0.242) (0.273) (0.325)
USD 1001-2000 -0.0558 -0.0718 0.799** 0.686
(0.306) (0.341) (0.385) (0.458)
USD greater than 2000 0.849*** 0.780** -0.223 0.801
(0.307) (0.342) (0.386) (0.459)
Urban -0.0124 0.0350*** 0.0528*** -0.0409**
(0.0116) (0.0130) (0.0146) (0.0174)
Male -0.00940 -0.0115 -0.122*** -0.0485***
(0.0118) (0.0131) (0.0148) (0.0176)
Constant -0.0421 -0.314*** 0.221*** 0.408***
(0.0516) (0.0575) (0.0650) (0.0772)
Observation 3,468 3,468 3,468 3,468
R-squared 0.126 0.110 0.158 0.077
Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
4.3 Determinants of Access to Credit
Table 3 shows that log age squared is significant for formal other access strand and exclusion strand. An increase
Public Policy and Administration Research www.iiste.org
ISSN 2224-5731(Paper) ISSN 2225-0972(Online)
Vol.7, No.6, 2017
26
in age marginally increases the likelihood of access to formal other strand of credit. The results further show that
individuals with advanced age are less likely to obtain credit although exclusion diminishes.
Education is found to be positively associated with access to formal credit. Specifically, having
secondary education compared to having no education was found to increase access to formal credit by a factor
0.0678. Similarly, having technical training after secondary and university degree were more likely to increase
access to formal credit by a factor of 0.0752 and 0.183 respectively. The coefficients of education increase from
the lower level to the highest level. These findings are corroborated by a survey on access to Credit by
Smallholder Farmers in Kenya in which Kiplimo et al., (2015) established that the marginal effects of education
showed significant and positive effects on access to credit. The findings have also been supported by the findings
in Johnson and Morduch (2007) who concluded that at higher levels of education individuals are able to engage
more in activities that generate higher incomes and therefore their ability to access credit is improved.
Married individuals or those living with their partner were shown to have a higher likelihood of
accessing formal other credit channel than being excluded. Similarly, divorce status increase the probability of
accessing informal credit than the other strands of access. However, a sizeable chance exists for divorced
individuals to be excluded from accessing credit. The widowed individuals had minimal chance of accessing
formal other credit. The findings obtained from this study are corroborated by Ololade et al., (2013) whose
findings lay credence to the role of marital status in determining access to credit. The study found out that not
being married reduced immensely the likelihood of accessing credit. Social capital was found to be a significant
determinant of access to credit. Membership to several groups increased the likelihood of accessing informal
credit than those with no group membership by a factor of 0.252. The findings indicate that the type of group
could have greater influence on the choices individual's make. Informal groups would increase access to
informal financial services. The findings are corroborated by Heikkilä et al., (2016) who investigated the role of
social capital in promoting financial access in Uganda. The study established that individual's social capital is
significantly associated with access to credit from formal financial institutions.
Income is one of the most important factors influencing access to credit in Kenya and Tanzania.
Across all credit access strands, the results showed that increased income would have a higher probability of
accessing formal credit. However, the size of the coefficients of income on informal and excluded access strands
are declining as income increases. This shows that increased income is more likely to reduce access to credit
from informal strand. Similarly, more income has been shown to reduce the level of exclusion from credit access.
Zins et al., (2016) established that increased income was associated with higher levels of financial inclusion.
Similarly, our results are corroborated by Allen et al., (2012) and Fungacova and Weill (2014) who used the
worldwide data and established that wealthy individuals were more likely to be financially included than the
poor. In addition, the findings in this paper are consistent with (Al-Hussainy et al., 2008; Johnston and Morduch
2007) which attach greater importance to income as a determinant of access to credit.
Although location measured as urban and rural was not significant for both formal access channels
(Formal and formal other credit), it was found to be significant in relation to informal and excluded channels.
Being urban relative to rural, an individual was less likely to access informal credit. This could be attributed to
the widespread mainstream financial services in the urban than in the rural areas (Mlachila et al., 2013; Doan et
al., 2010). Similarly, being urban was less likely to exclude individuals from accessing credit. Similar findings
were reported in Mlachila et al., (2013) showing the importance of variety in promoting financial inclusion.
The results further shows lesser likelihood of men to access informal credit compared to women and
the coefficients taper towards formal credit access. This shows that women are more constrained to access credit
across different strands than men. According to Mayada (1994), women are considered more constrained in the
formal financial markets which limits their capacity to access credit. In another study, Navajas and Tejerina
(2006) showed that the main challenge for women to access credit was the cost associated with it.
4.4 Marginal Effects on Savings Mobilization
Marginal effects of the variables on savings mobilization are showed in Table 4. The marginal effects of income
and education on formal savings are more pronounced. For instance, a one unit increase in education level
increases the likelihood of accessing formal savings by 4.6 percent compared to income which is likely to
increase access to formal savings by 4 percent. Age, marital status and social capital increase access to formal
savings marginally. However, location and gender have the effect of reducing access to formal savings.
Tesfamariam (2012) carried out a study to investigate the determinants of savings mobilization among rural co-
operators in Tigrai region in Ethiopia. The empirical results showed that education was statistically significant
but negatively influencing savings. Results from Abdelkhalek et al., (2010) however showed that education
determines formal savings mobilization more than income. These findings corroborate the results in the current
study which show that education determines access to formal savings more than income.
Public Policy and Administration Research www.iiste.org
ISSN 2224-5731(Paper) ISSN 2225-0972(Online)
Vol.7, No.6, 2017
27
Table 3: Results for the Seemingly Unrelated Regression for Credit access
VARIABLES Formal
Credit
Formal
Other
Credit
Informal
Credit
Excluded
Credit
Log Age squared 0.00245 0.00861* -0.00968 -0.0145**
(0.00393) (0.00521) (0.00683) (0.00571)
Some primary -0.00432 0.00681 0.0343* 0.0371**
(0.0102) (0.0135) (0.0177) (0.0148)
Primary completed 0.0107 0.0452*** 0.0617*** 0.0471
(0.0104) (0.0137) (0.0180) (0.0151)
Some secondary 0.0145 0.0775*** 0.0317 0.0308
(0.0139) (0.0184) (0.0241) (0.0202)
Secondary completed 0.0678*** 0.107*** 0.0615*** 0.0193
(0.0121) (0.0160) (0.0210) (0.0175)
Technical after Secondary 0.0752*** 0.102*** 0.0703** 0.0630
(0.0162) (0.0215) (0.0282) (0.0236)
University degree 0.183*** 0.243*** 0.0200 0.0412
(0.0231) (0.0307) (0.0402) (0.0336)
Divorced 0.00520 0.0827*** 0.135*** 0.127***
(0.0152) (0.0201) (0.0264) (0.0221)
Widowed 0.00199 0.0470*** 0.0151 0.0702***
(0.0115) (0.0153) (0.0201) (0.0168)
Married/living together 0.0157* 0.0581*** 0.0212 0.0540***
(0.00890) (0.0118) (0.0155) (0.0129)
Belongs to one group 0.0297** 0.0311* 0.183*** 0.0370**
(0.0126) (0.0166) (0.0218) (0.0183)
Belongs to two groups -0.00144 0.0502 0.0971** 0.0282
(0.0238) (0.0315) (0.0413) (0.0345)
Belongs to three or more groups
0.00939 0.0317 0.252*** 0.0540
(0.0313) (0.0415) (0.0544) (0.0455)
USD 1-15 -0.000574 0.00162 0.0473*** 0.0470***
(0.00807) (0.0107) (0.0140) (0.0117)
USD 16-30 0.00347 0.0276** 0.0318* 0.0354**
(0.0102) (0.0136) (0.0178) (0.0149)
USD 31-75 0.0580*** 0.0580*** 0.0301 0.00134
(0.0119) (0.0158) (0.0207) (0.0173)
USD 76-150 0.133*** 0.120*** -0.00341 -0.0194
(0.0179) (0.0237) (0.0311) (0.0260)
USD 151-300 0.175*** 0.108*** 0.0130 -0.0343
(0.0211) (0.0280) (0.0367) (0.0307)
USD 301-1000 -0.0583 0.373** 0.201 0.350*
(0.130) (0.172) (0.226) (0.189)
USD 1001-2000 -0.0184 -0.0626 -0.157 -0.119
(0.184) (0.243) (0.319) (0.267)
USD greater than 2000 -0.0562 0.898*** -0.0596 -0.0353
(0.184) (0.243) (0.319) (0.267)
Urban 0.00713 0.00537 -0.0336*** -0.0465***
(0.00697) (0.00924) (0.0121) (0.0101)
Male -0.0156** -0.0270*** -0.0462*** -0.00843
(0.00705) (0.00934) (0.0122) (0.0102)
Constant -0.0249 -0.0998** 0.141*** 0.117***
(0.0309) (0.0410) (0.0537) (0.0449)
Observation 3,468 3,468 3,468 3,468
R-squared 0.115 0.085 0.063 0.038
Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Men and women have been shown to have varying propensities to access savings mobilization with
men carrying the highest probability. However, Bajtelsmit and Bernasek (1996) find that women had a higher
Public Policy and Administration Research www.iiste.org
ISSN 2224-5731(Paper) ISSN 2225-0972(Online)
Vol.7, No.6, 2017
28
propensity to access savings mobilization than men. They showed more women with higher proportions of fixed
assets than men. Even more recent studies have also not established conclusive findings on the sign of the
coefficient of gender on savings mobilization (Anang et al., 2015, Mirach et al., 2014). The fundamental reason
cited in these studies is the low tolerance to risk for women negatively affects them in the short run rather than in
the long run. Similarly, differences exist between savings behaviour of rural and urban dwellers. A change from
urban to rural areas decreases the probability of accessing savings by 0.011. The rural-urban differences have
been investigated and varied conclusions reached. For instance, rural dwellers are more likely to engage in
informal savings as well as saving in secret places. This is mainly driven by both the supply side constraints as
well as the individual level factors.
Table 4: Marginal effects on savings mobilization
Variable dy/dx
Std. Err.
Z
P>z
[ 95%
C.I. ]
X
Log Age squared 0.012
0.006
1.850
0.065
-0.001
0.024
7.402
Education 0.046
0.004
12.060
0.000
0.039
0.054
2.093
Marital Status 0.002
0.005
0.510
0.611
-0.007
0.012
2.134
Social Capital 0.008
0.012
0.660
0.508
-0.015
0.030
0.134
Income 0.040
0.005
8.660
0.000
0.031
0.049
1.205
Location -0.011
0.012
-0.940
0.349
-0.034
0.012
0.612
Gender -0.009
0.011
-0.780
0.435
-0.031
0.013
0.595
4.5 Marginal Effects on Credit Access
Similar to access to savings mobilization, education and income influence access to credit in significant ways.
Table 5 shows that an increase in education by one level has the potential of increasing access to credit by a
factor of 0.021 while increasing income by one unit would increase access to credit by a factor of 0.027.
Although gender appears to decrease access to credit, being female would cause a decline in the probability of
accessing credit than being male. Table 5 shows the marginal effects on access to credit.
Table 5: Marginal Effects on Credit Access
Variable dy/dx
Std. Err.
z
P>z
[ 95%
C.I. ]
X
Log Age squared 0.006
0.004
1.520
0.129
-0.002
0.013
7.402
Education 0.021
0.002
8.990
0.000
0.016
0.025
2.093
Marital Status 0.006
0.003
2.050
0.040
0.000
0.012
2.134
Social Capital 0.004
0.007
0.590
0.553
-0.010
0.018
0.134
Income 0.027
0.003
9.720
0.000
0.022
0.033
1.205
Location 0.011
0.007
1.620
0.104
-0.002
0.025
0.612
Gender -0.017
0.007
-2.540
0.011
-0.031
-0.004
0.595
5. Conclusion and Policy Implications
The paper aimed at examining the socioeconomic and demographic characteristics of access to and use of credit
and savings mobilization in Kenya and Tanzania. Both socioeconomic and demographic factors are key
determinants of access to savings and credit in Kenya and Tanzania. The findings have important policy
implications as follows; when more people are financially included, the risks associated with informal and/or
exclusion are avoided. In addition, promotion of education and employment policies may provide the much
needed impetus to increasing access to financial services. Access to credit ensures that households are not
deprived of income to smoothen their consumption. Majority of the people in Kenya and Tanzania are relatively
poor and therefore require stable source of income for their livelihoods. It is in this regard that access to credit
and savings becomes a critical factor to improving the livelihoods. Education plays the most critical role in
ensuring that choices preferred by individuals are not only rational but also beneficial. More education has been
shown to lead to better earnings; enabled people to live in urban areas and facilitated creation of networks to
fulfill their desires.
References
Abdelkhalek, T., Arestoff, F., de Freitas, M., and Mage, S., (2010), A Micro-econometric analysis of household
savings determinants in Morocco. African Review of Money Finance and Banking, pp. 7-27.
Addae-Korankye A., (2014), Causes and Control of Loan Default/Delinquency in Microfinance Institutions in
Ghana. American International Journal of Contemporary Research, (12), pp. 1740-1755.
Akudugu, M., (2013), The determinants of financial inclusion in Western Africa: Insights from Ghana. Research
Journal of Finance and Accounting, (10) pp. 1-10.
Al-Hussainy, E., Thorsten B., Asli Demirguc-Kunt, and Bilal Z., (2008), “Household Use of Financial Services”.
World Bank Working Paper, pp.502-516.
Allen, F., Carletti, E., Cull, R., Qian, J., Senbet, L., and Valenzuela, P., (2012), Improving access to banking:
Public Policy and Administration Research www.iiste.org
ISSN 2224-5731(Paper) ISSN 2225-0972(Online)
Vol.7, No.6, 2017
29
Evidence from Kenya. University of Pennsylvania, pp. 1005-1017.
Alter, A and Yontcheva, B (2015), Financial Inclusion and Development in the CEMAC, IMF Working Paper
WP/15/235
Anang, B. T., Dawuda, I., Imoro, L., (2015), Determinants of savings habit among clients of Bonzali Rural Bank
in the Tolon-Kumbungu District of Ghana. UDS International Journal of Development, Volume 2 No.
2.
Asiedu E., Isaac N., and Akwasi N., (2013), “Access to Credit by Firms in Sub-Saharan Africa: How Relevant
is Gender?” PERI Working Paper, (550).
Aterido R., Beck T., and Leonardo I., (2011), “Gender and Finance in Sub- Saharan Africa: Are Women
Disadvantaged?” The World Bank, Policy Research Working Paper (5571).
Atieno, R., (2001), Formal and informal institutions' lending policies and acccess to credit by small-scale
enterprises in Kenya: An empirical assessment. Nairobi: University of Nairobi.
Bajtelsmit, L., and Bernasek, A., (1996), Why do women invest differently than men?. Financial Counseling and
Planning.
Bhuiya A., Chowdhury M., and Syed A., (2001), Micro-credit and emotional wellbeing: Experience of poor rural
women from Matlab, Bangladesh. World Development, pp. 1957-1966.
Bovenberg, L., and Evans, O., (1990), National and Personal Saving in the United States: Measurement and
Analysis of Recent Trends, IMF Staff Papers, (3).
Calderon, C., Liu, L., (2003), The direction of causality between financial development and economic growth.
Journal of Development Economics, pp. 321–34.
Camara, N., Peña, X., and Tuesta, D., (2014), Factors that Matter for Financial Inclusion: Evidence from Peru
(1409).
Demirgüç-Kunt, A., Klapper, F., and Singer, D., (2013), Financial inclusion and legal discrimination against
women: evidence from developing countries. World Bank Policy Research Working Paper, (6416).
Demirgüç-Kunt, A., Klapper, F., Singer, D., and Van Oudheusden, P., (2015), The Global Findex Database 2014:
measuring financial inclusion around the world. World Bank Policy Research Working Paper, (7255).
Diagne, A., (1999), Determinants of access to and participation in formal and informal credit markets in Malawi.
International Food Policy Research Institute, Discussion Paper (67).
Doan, T., Gibson, J., and Holmes, M., (2010), What determines credit participation and credit constraints of the
poor in peri-urban areas, Vietnam?.Does Education Matter? Evidence from Rural Bangladesh’,
Econometric Reviews. (3), pp. 1-144.
Donkor, J., and Duah, A., (2013), Relationship between Savings and Credit in Rural Banks with Specific
Reference to Ghana. International Journal of Business and Social Science, pp 8.
Efobi, U., Beecroft, I. and Osabuohien, E., (2014), Access to and use of bank services in Nigeria: Micro-
econometric evidence. Review of development finance, 4(2), pp.104-114.
Ellis, K., Lemma, A., and Rud, J., (2010), Investigating the Impact of Access to Financial Services on Household
Investment. Overseas Development Institute, Working Paper.
Fernando N., (2007), Low Income Households’ Access to Financial Services: International experience, Measures
for Improvement and for Future.
Foley, M., and Pyle, W., (2005), Household Savings in Russia during the Transition, Middlebury College
Economics, Discussion Paper, (05-22).
Fungáčová, Z., and Weill, L., (2014), Understanding financial inclusion in China. China Economic Review.
Gersovitz, M. (1988), Saving and Development, Handbook of Development Economics, Chenery, H.B. and
Srinivasan, T.N. (ed.), (1), Elsevier Science Publishers: pp. 382-424.
Gitaharie, B., Soelistianingsih, L., and Djutaharta, T., (2014), Financial inclusion: Household access to credit in
Indonesia . Working Paper in Business and Economics, 4(1), pp. 1-14.
Heikkilä, A., Kalmi, P., and Ruuskanen, P., (2016), Social capital and access to credit: Evidence from Uganda.
The Journal of Development Studies, pp. 1-16.
Honohan, P., (2005), "Measuring Microfinance Access: Building on Existing Cross-Country Data." World Bank
Policy Research Working Paper No. (3606).
Ike, P., and Umuedafe, D. (2013), Determinants of savings and capital formation among rural farmers in Isoko
north local government area of Delta State, Nigeria. Asian Economic and Financial Review, 3(10),
1289-1297.
Johnson, S., and Zarazua, M., (2009), Financial access and exclusion in Kenya and Uganda. Bath: University of
Bath.
Johnston Jr., Don, and Jonathan Morduch (2007), The unbanked: Evidence from Indonesia. New York: Financial
Access Initiative.
Kibet, L. K., Mutai, B. K., Ouma, D. E., Ouma, S. A., and Owuor, G. (2009). Determinants of household saving:
Case of smallholder farmers, entrepreneurs and teachers in rural areas of Kenya. Journal of
Public Policy and Administration Research www.iiste.org
ISSN 2224-5731(Paper) ISSN 2225-0972(Online)
Vol.7, No.6, 2017
30
Development and Agricultural Economics, 1(7), pp. 137-143.
Kiplimo, C., Ngenoh, E., Koech, W., and Bett, K., (2015), Determinants of Access to Credit Financial Services
by Smallholder Farmers in Kenya. Journal of Development and Agricultural Economics, 7(9), pp.
303-313.
King, M. (2012), Is mobile banking the tyranny of distance to bank infrastructure? Evidence from Kenya. IIIS
Discussion Paper 421(1), pp. 2-55.
Kochar, A., (1997), “An Empirical Investigation of Rationing Constraints in Rural Credit Markets in India”
Journal of Development Economics, (53), pp. 339 – 371.
Kraay, A., (2000), Household Saving in China, The World Bank Economic Review, 14(3), pp. 545-570.
Kulikov, D., Paabut, A., and Staehr, K., (2007), A Microeconometric Analysis of Household Saving in Estonia:
Income, Wealth and Financial Exposure, Research Department Eesti Pank.
Le B., Porpiglia, T., and Zhu, J., and Ziegelmeyer, M., (2014), Household Saving Behaviour and Credit
Constraints in the Euro Area. De Nederlandsche Bank Working Paper No. (428).
Llanto M., (2015), Financial Inclusion, Education, and Regulation in the Philippines. ADBI Working Paper (541).
Loayza, N., and Shankar, R. (2000), Private Saving in India, The World Bank Economic Review, 14(3), pp.571-
594.
Luce, D., (1959), Individual Choice Behaviour, John Wiley, New York, NY.
Malapit, H.,J.,L., (2012) Are women more likely to be credit constrained? Evidence from low-income urban
households in the Philippines. Feminist Economics, 18(3), pp.81-108.
Mayada M., Baydas F., Meyer L., (1994), Credit rationing in small-scale enterprises: Special programmes in
Ecuador. Journal of Development Studies. (2), pp. 279- 309.
Marschak J., (1960), Binary Choices Constraints on Random Utility Indicators. In Arrow K. (ed), Stanford
Symposium on Mathematical Methods in the Social Sciences, Stanford University Press, Stanford,
California.
Mbiti, I., and Weil, D., (2011), Mobile Banking: The impact of M-Pesa in Kenya. Cambridge, MA: National
Bureau of Economic Research.
Mbuthia, A., (2011), Households' savings decisions in Kenya . Nairobi: Kenyatta University .
McFadden, D. (1974), Conditional logit analysis of qualitative choice behavior. In: P. Zarembka, Editor,
Frontiers in Econometrics, Academic Press, New York.
Mirach, T., and Hailu, Y. (2014), Determinants of household savings in Ethiopia: A case of North Gondar Zone,
Amhara regional state. International Journal of Development and Economic Sustainability, 2(4),
pp.37-49.
Mlachila M., Park G., and Yabara M., (2013), “Banking in sub-Saharan Africa: The macroeconomic context”
International Monetary Fund.
Mwalughali, O., (2013), The impact of community savings and investment promotion program on household
income and credit maket participation in Kasungu District, Central Malawi. Lilongwe: University of
Malawi.
Mwangi W., and Shem O., (2012), Social capital and access to credit in Kenya. American Journal of Social and
Management Sciences, (3) 1.
Mwangi, W., and Sichei, M., (2012), “Determinants of Access to Credit by Individuals in Kenya: A Comparative
Analysis of the Kenya National FinAccess Surveys of 2006 and 2009”, European Journal of Business
and Management, (3), 3.
Mwangi W., and Ouma, A., (2012), Social capital and access to credit in Kenya. American Journal of Social and
Management Sciences, (1) 1, pp. 1-9.
Navajas S., Tejerina L., (2006), “Microfinance in Latin America and the Caribbean: Connecting Supply and
Demand”, Bank of international development, Washington DC
Ngendakuriyo F., (2014), "Determinants of household savings mobilization across EAC countries: An
Exploratory Analysis". Financial Sector Development and Regionalisation Project (FSDRP) I: EAC
Secretariat.
Ololade A., and Olagunju I., (2013), Determinants of Access to Credit among Rural Farmers in Oyo State,
Nigeria.Global Journal of Science Frontier Research, (2), pp. 17-22.
Orbeta, C., (2006), Children and Household Saving in the Philippines. Philippine Institute for Development
Studies, Discussion Paper, Series No.(2006- 14).
Pailwar, V., Kaur, J., Saxena, K., and Nijhara, M. (2010), Impact of membership of financial institutions on rural
savings: A micro-level . International Business and Economics Research Journal, pp. 1-10.
Rogg, C., (2000), The impact of access to credit on the saving behavior of microentrepreneurs. Oxford:
University of Oxford.
Sinha A., (2014), Carbon Emissions and Mortality Rates: A Causal Analysis for India (1971-2010).
International Journal of Economic Practices and Theories, (4), pp. 486-492.
Public Policy and Administration Research www.iiste.org
ISSN 2224-5731(Paper) ISSN 2225-0972(Online)
Vol.7, No.6, 2017
31
Tesfamariam, K., (2012), Saving behaviour and determinants of saving mobilization by rural financial co-
operators in Tigrai Region, Ethiopia. Journal of Agribusiness and Rural Development, pp. 26.
Vaessen, J., (2001), “Accessibility of Rural Credit in Northern Nicaragua: The Importance of Networks of
Information and Recommendation” Savings and Development, pp. 5 – 31.
World Bank (1993), The East Asian Miracle: Economic Growth and Public Policy. A World Bank Policy
Research Report, New York: Oxford University Press.
Zins, A. and Weill, L., (2016), The determinants of financial inclusion in Africa. Review of Development
Finance, (1) 6, pp.46-57.
