Financing Agribusiness: Potential Determinants of Financial Inclusion for Smallholder Rural Farming Communities in Kenya

Abstract

The purpose of this article is to empirically examine and compare the factors or determinants of financial inclusion in commercial banking institutions, microfinance banks, mobile banking platforms and table banking groups (informal banking groups) in Kenya. The data used in the study were obtained from 631 small and micro-sized (SMS) farming enterprises in Kenya. We use probit model to empirically establish the factors that determine the probability of small and micro-sized farming enterprises in accessing agricultural credit from the four major lending institutions. Also, we apply Heckman selection model to establish the determinants of agricultural credit rationing. OLS model is used to investigate the determinants of agricultural credit delinquency rates in the lending institutions. The results show that female-owned SMS farming enterprises are likely to access less agricultural credit amount from table banking groups, own small farms with no title deed and own less valued assets compared to male-owned SMS farming enterprises which access agricultural credit from commercial banks and microfinance banks. Further results revealed that household dependency ratio was a significant determinant factor in agricultural credit access from mobile banks, commercial banks and table banking groups. Additionally, agricultural credit delinquency in mobile banking is significantly reduced by distance to the banking agent. The availability of mobile banking agencies within village centres would considerably enable agri-entrepreneurs to make credit repayment within the required period.

Keywords

Financial inclusion financial technology agricultural credits Small and micro-sized enterprises

Introduction

It is generally accepted that extensive financial inclusion among farming households has played a significant role in the economic growth of transition economies in Asia and Eastern Europe (Swinnen and Gow 1999; Yuan and Gao 2012). The provision of affordable and accessible financial services to smallholder farmers has enabled the agricultural sector of China to be competitive in the global food market. Also, agricultural credit has the potential to increase yields outputs, enabling poor farmers to use more efficient external inputs such as chemical fertilisers, pesticides and seeds. However, lack of affordable and accessible credit services has a negative impact on the adoption of agricultural technologies and limits farmer’s ability to utilise sustainable agricultural practices in sub-Saharan African (SSA) countries (Tadesse 2014). Indeed, financial inclusion for low-income smallholder farmers has been identified as a key factor in attaining sustainable food production to feed the growing population in the region (Dong, Lu, and Featherstone 2012). In addition, affordable credit would ensure better livelihoods for rural communities by creating agricultural employment.

Just like in most of the emerging economies, commercial banks in Kenya have been reluctant to offer credit services to smallholder farmers. Smallholder farmers are considered to lack the required collateral and guarantors to qualify for commercial bank credit. Also, the success of credit applications mainly depended on savings and employment income and very little on credit scoring. Prior studies point out that the transaction cost of processing credit mainly excluded the majority of smallholder farmer populations in rural areas (Asante-Addo et al. 2017). Accessing credit from commercial banks in Kenya still requires farmers to travel more than 10 km to reach the nearest bank branch (Omwansa and Waema 2014).

The provision of financial services by microfinance banks for all the rural farming communities has widely been acknowledged as the panacea for reduction in poverty and economic growth in many developed countries (Kumar, Turvey, and Kropp 2013). The microfinance banking model has been popular for helping millions of poor people attain economic growth in Asia and Latin America. However, the success of microfinance has not been sufficiently replicated in SSA (Van Rooyen, Stewart, and de Wet 2012). The region has more than 300 million people who are economically active and in need of credit; however, only 20 million (less than 10%) have access to any form of formal credit services (Helmore and CARE/International 2009). With only 25% of the adults owning bank accounts, SSA lags behind other regions in attaining sustainable financial inclusion levels (Demirgüç-Kunt and Klapper 2013). Liberalisation and privatisation of financial services have improved credit availability to smallholders in SSA (Sheahan and Barrett 2014). Nevertheless, majority of farmers still face severe credit access and credit rationing constraints that limit their agricultural enterprises’ potential.

In emerging economies, mobile money technology has revolutionised the financial sector’s landscape. Mobile money services facilitate secure and affordable remittances and also provide quick cashless payment of goods and services and linkage to banking services (Kusimba, Kunyu, and Gross 2013). In Kenya, more people are likely to bank with mobile phones than through commercial bank account (GSMA 2012). The number of loan and deposit accounts accessed from mobile phones increased from 1.67 million in 2010 to 8.51 million in 2016 (FSD 2016). Mobile banking technology is considered to contribute towards financial inclusion for the unbanked rural farming populations. However, the penetration of mobile banking technology is still considered low in rural agricultural areas. In low-income and developing economies, majority of the population, especially women rely on traditional informal banking groups like Rotation Saving and Credit Association (ROSCA) or Accumulated Savings and Credit Associations (ASCAS) commonly referred to as table banking groups for credit (Yuan and Gao 2012). Women who are excluded from formal credit depend on informal banking schemes to finance the micro-enterprises (Gichuki, Mutuku, and Kinuthia 2015). However, despite the benefits associated with informal banking groups, some studies report that informal banking is not sustainable in bringing financial gaps in emerging economies (Gugerty 2007). In addition, the schemes are prone to fraud and default rates (Ambec and Treich 2007). Furthermore, Gugerty (2007) noted that the informal banking groups cannot sufficiently provide the financial capital needed to fund expanding micro-enterprise ventures.

Doss (2014) argued that women are likely to be involved in producing 60–80% of the world’s food, despite having lower access to agri credit, land, farm inputs and their responsibilities for household work. Regardless of growing women potential, agricultural production is still a male-dominated one. Deeply embedded cultural beliefs such as discrimination from property inheritance laws, oppressive customary laws and time constrains due to household responsibilities derail women’s entrepreneurial growth. Previous literature points out that women entrepreneurs are likely to be financially excluded from formal banking services than their male counterparts (Kumari 2017).

Therefore, the general objective of the study is to explore and compare the determinants of financial inclusion in Kenya’s lending institutions. The study will focus on commercial banking institutions, microfinance banks, mobile banking platforms and table banking groups. First, the study will seek to characterise and describe perceptions that are likely to influence agricultural credit access for small and micro-sized (SMS) agri-enterprises. Second, the study will examine and compare the factors that determine agri-entrepreneurs’ probability of accessing agricultural credit from lending institutions. Third, the study will explore and compare the factors that determine the extent of agricultural credit rationing in lending institutions. Fourth, the study will investigate the factors that are likely to regulate agricultural credit delinquencies across different lending institutions.

Research Methodology

The data used in the study were obtained from a survey that was conducted between June and July 2020 in Nakuru, Baringo, Embu and Nyandarua counties in Kenya. A multistage sampling method was used to select 631 SMS agri-enterprises. In the first step, we relied on sampling frame of SMS agri-enterprises registered with the agricultural department. In the second step, proportionately clustering techniques were used to select 198 agri-enterprises from Nakuru, 141 from Baringo, 141 from Embu and 151 from Nyandarua counties. The survey data were collected using structured questionnaires after interviews with agri-enterprises owners.

Empirical Specification

In order to examine perceptions likely to influence agri-entrepreneurs’ preference for agricultural credit, we employed Garrett ranking technique. The technique enabled us to rank farmer’s preference for agricultural credit by ranking scores. The ranking technique’s first assumption is that the percentage estimate for each respondent is converted into factor scores. For every factor, the scores of each individual are added and total value of scores and mean values of score is calculated. The ranking technique is helpful in recording cases where some traits are not easily measured by ordinary methods but can be arranged in the order of merit (Garrett and Woodworth 1969). A prior study by Asante-Addo et al. (2017) uses Garrett ranking to determine households’ reasons for participation and non-participation in microcredit programmes in Ghana. Anjugam and Ramasamy (2007) apply techniques to analyse factors that influence women’s participation in self-help groups (SHGs) in India. The factor having the highest mean value is considered as the most likely to influence farmers’ preference for mobile banking credit. The formula for score value for all factors can be written as

Percentage position = \frac{100 (R i j - 0.5)}{N j}

Where Rij is denoted as rank given for the ith variable by jth respondents, while N_i denotes the number of variables ranked by jth respondents.

A number of studies have examined the factors likely to influence credit access in financial banks using different econometric models; for instance, Rahji and Fakayode (2009) applied a multinomial logit to establish the influence of credit rationing by commercial banks. Kibet et al. (2009) used propensity score matching to evaluate strategies applied by microfinance in disbursing affordable credit to smallholder farmers in Kenya. Similarly, Lukytawati (2009) investigated factors that are applied by community-based SHGs like ROSCA and ASCRA to determine members’ creditworthiness. Also, recent studies have applied probit model equation to investigate the demand for credit by farmers (Dutta and Magableh 2006; Kochar 1997). In the first stage, this study will employ probit model to establish factors likely to influence SMS agri-enterprises’ agricultural credit to be approved. The framework used in the analysis is based on the assumption that agri-entrepreneurs’ application of agricultural credit is faced by two possible outcomes: 1 for successful application of agricultural credit and 0 for failure to secure credit. Prior studies by Cheng and Ahmed (2014) and Dutta and Magableh (2006) have empirically expressed the model as

L d = f (I, H C, H F)

where L_d denotes the demand function for credit, HC is considered to be the vector for agri-entrepreneurs household characteristics and I denotes the personal characteristics of the household head. HF represents agri-entrepreneurs’ household financial characteristics like assets and income. I and HC are non-financial characteristics of credit applicants. In the second stage, we apply the Heckman two-step selection model to examine the extent credit rationing parameters applied by the lender. The Heckman two-stage model enables us to reduce sample selection bias when modelling for credit demand (Dutta and Magableh 2006; Okurut, Schoombee, and van der Berg 2004). In the application of the model, the first stage probit estimates in equation 1 generate the inverse Mills ratio (IMR), which is included as an explanatory variable in the second-stage estimation to correct the sample selection bias problem. The Heckman model can be exported as

c = \propto + β^{'} Z + e

Where z = f(I, HC, HF) denotes the exogenous variables perceived to affect lenders’ decision to grant agricultural credit applied; $\propto$ and $β^{'}$ are the parameters to be examined while e is the random disturbance term. c denotes dependent variable, a dichotomous latent variable 1 indicating access of agricultural credit, while 0 indicates rejection of credit applied. Therefore, the probability of accessing agricultural credit can be estimated as

p r o b (a c e s \sin g a g r i c r e d i t a p p l i e d) = \{\begin{matrix} 1 & i f p r o b (\propto + β^{'} Z + e > 0) \\ 0 & o t h e r w i s e \end{matrix}

Similarly, the probability of not accessing credit applied for is expressed as

p r o b (n o t g a r n t e d a g r i c r e d i t a p p l i e d) = \{\begin{matrix} 1 & i f p r o b (\propto + β^{'} Z + e > 0) \\ 0 & o t h e r w i s e \end{matrix}

where $(\propto + β^{'} Z + e > 0)$ denotes the case where the applicant applies for agricultural credit.

In the study, we also examine the determinant factors influencing agricultural credit delinquencies among agri-entrepreneurs by using OLS model. The approach has been previously applied by Raghunathan et al. (2011) and (Weber and Musshoff 2012). The approach uses frequency of late payments during the loan repayment period to measure credit risk. Prior studies by Raghunathan et al. (2011) noted that the higher the frequency of delinquency rates increases the risk for credit defaults. Weber and Musshoff (2012) study estimated the empirical OLS model as follows:

c_{i, t} = α + β \cdot a_{i} + γ \cdot C_{i, t} + ε_{t} + \in \cdot s_{i} + e_{1, t}

where c_i,t is the dependent variable, and it denotes the percentage ratio of delinquency of agricultural credit accessed. e_1,t represents time constant for year t when agricultural credit was accessed, ∈ is the parameter of vector, while I and t are identical and independent distributed error terms with mean of 0 and a variance $σ_{e}^{2}$ .

Descriptive Statistics of the Selected Variable

Table 1 presents descriptive statistics for the surveyed SMS agri-enterprises. We observe that 78% of the SMS agri-enterprises that accessed agricultural credit from table banking groups were female who have 7.8 years of education, earn an annual average income of Ksh181,596. In addition, women-owned SMS agri-enterprises were small in size, had no title deed farm and owned less valued assets compared to male agri-entrepreneurs who access agricultural credit form other lending sources. This is a typical feature of the study location where women have limited control over productive resources and often seek credit from informal banking sources (Mulu-Mutuku and Gichuki 2017). We further note that agri-entrepreneurs with preference for credit from table banking groups access the least amount of agricultural credit, while agri-entrepreneurs with preference for commercial banks and microfinance banks received agricultural credit of Ksh146,578 and Ksh60545, respectively. In comparison to table banking groups and mobile banking platforms, accessing agricultural credit from commercial banks and microfinance attracted highest interest rates of 12.94% and 11.12%, respectively. Interestingly, we observe that agricultural credit rationing is evident among agri-entrepreneurs with preference for microfinance (84%), table banking (76%) and mobile banking platforms (75%). Also, credit delinquency rates are least expected in commercial banks and more likely in mobile banking platform. Agri-entrepreneurs who accessed credit from mobile banking platforms are young and have more education years.

Table 1.

Agri-entrepreneurs’ Household Characteristics.

Characteristic	Mobile Banking Platform (n = 200)		Commercial Banks (n = 127)		Micro-finance Banks (n = 148)		Table Banking Groups (n = 161)
Characteristic	Mean	SD	Mean	SD	Mean	SD	Mean	SD
Age of the agri-entrepreneurs	36.71	12.81	44.44	14.91	48.48	15.66	48.78	16.30
Education years of HH	10.59	2.217	8.984	3.459	9.391	4.351	7.826	3.212
Gender of agri-entrepreneurs (1 = male)	0.590	0.492	0.795	0.405	0.506	0.501	0.378	0.486
HH dependency ratio	50.41	50.41	48.04	23.85	47.11	24.03	54.17	16.91
Farm size (acres)	4.349	5.316	4.122	5.112	3.930	4.507	2.981	2.539
Title deed land	0.530	0.500	0.763	0.426	0.648	0.479	0.173	0.380
Membership in social network groups (1 = yes)	0.575	0.495	0.535	0.500	0.777	0.417	0.826	0.380
Average annual income (Ksh)	191125	15463	248172	227902	214692	148706	181596.	124425
Value of liquid assets (Ksh)	1633014	22532	2411211	3961110	1578675	1768855	1105129	13071
Value of livestock (Ksh)	140687.0	10122	527010.	36027	65079	55389	66053	82546
Amount of credit borrowed (Ksh)	43598.0	43969	146578	52663	60545	28740	27158.	11386
Credit repayment interest rates (%)	10.73	2.763	12.94	1.513	11.12	2.143	6.766	2.097
Credit rationing (1 = yes)	0.750	0.434	0.447	0.500	0.840	0.367	0.766	0.424
Outstanding dept. (Ksh)	16570	22713	62181	27514	30837	22996	8411.	4895.5
Credit delinquency rate	50.25	22.31	26.12	32.073	47.11	24.03	44.54	31.07
Distance to banks (minutes)	56.40	222.2	61.53	63.19	126.7	89.91	27.88	17.69
Distance to shopping centre (minutes)	138.6	224.4	135.63	370.2	29.43	20.64	112.73	90.58

Source: Compiled by authors from survey data.

Factors Influencing Agricultural Credit Access

In order to characterise and describe factors likely to influence agricultural credit access for SMS agri-enterprises, we apply Garrett ranking technique; the results are presented in Table 2. We find that lack of collateral requirement, high interest rates, too much paperwork and complexity in using mobile banking applications were perceived as a major hindrance to agricultural credit access. The results further show that 10% of the surveyed agri-entrepreneurs perceive credit borrowing to be a risky venture, while lack of credit information is perceived to be the least likely factor affecting agricultural credit access. The literature on credit lending in emerging markets reveals that commercial banks are unfamiliar with unique operating characteristics of small-sized farming ventures (Reyes and Lensink 2011); for instance, despite the potential of livestock production in Ethiopia, credit markets are not easily accessible for smallholder farmers (Fitawek and Kalaba 2017). Generally, commercial lenders find it hard to establish credit worthiness of small-sized agribusiness ventures, especially in cases where they lack collateral and the path to profitability is considered to be tenuous and takes long to be realised (Kibet et al. 2009; Saldias and von Cramon-Taubadel 2012). Mukherjee (2013) shows that in India, micro-enterprises were likely not to apply for credit mainly because lenders would outrightly reject their credit application based on indebtedness, delinquency or credit supply constrain.

Table 2.

Perceptions on or Against Credit Access.

Reason Against Credit	Mean Score	Rank
Borrowing too risky	0.1023	5
Lack of credit information	0.099	6
No assets collateral	0.667	1
Interest rate is high	0.565	2
Too much paperwork/mobile application very complicated	0.285	3
Lack of required savings	0.263	4

Source: Compiled by authors from survey data.

Empirical Results: Determinants of Agricultural Credit Access from Lending Institutions

The results in Table 3 present the probit regression estimates of the likelihood of SMS agri-enterprises in successfully accessing agricultural credit from commercial banks, microfinance banks, mobile banking platforms and table banking groups. In the model, the dependent variable is a binary variable, that is, 1 for successful application of agricultural credit and 0 for failure to secure agricultural credit. The independent variables are agri-entrepreneurs’ demographic, financial and social capital characteristics. The magnitude of the coefficient from probit model is difficult to interpret; we, therefore, use the marginal effects estimates. The marginal effects for the value of livestock are positive and statistically significant for mobile banks, microfinance and table banking groups. This indicates that the value of livestock owned by agri-entrepreneurs would significantly influence likelihood of obtaining agricultural credit from mobile banks, microfinance and table banking groups. This confirms the previous findings by Mohamed and Temu (2008) that in a risky credit environment, farmers in need of credit would use livestock capital as collateral to secure required credit amount.

Generally, in SSA, the reliability of agricultural income for credit repayment is not always guaranteed mainly because of droughts, crop failures and market instability. However, our results show that the amount of income earned by farmers would significantly determine probability of agricultural credit access from commercial banks and microfinance bank observations also shared by Sheahan and Barrett (2014). Further finding reveals that agri-entrepreneurs’ dependency ratio was a significant factor that influences the possibility of agricultural credit access from mobile banks, commercial banks and table banking groups. Agri-entrepreneurs’ membership in social networking groups was likely to influence agricultural credit access from table banking groups. Indeed, participating in social network groups like farmer’s development groups does not only provide information on credit lending services but also offers valuable information on commodity prices and agronomic recommendations on fertiliser use (Mbo’o-Tchouawou and Colverson 2014; Tadesse 2014). With regard to agri-entrepreneurs’ personal characteristics, we note that agri-entrepreneurs’ age influences the probability of agricultural credit access from mobile banks and table banking groups. Arun and Bendig (2010) study show that there is a positive co-relationship between age and the demand for credit. Also, agri- enterprise ownership is likely to have an increasing demand for credit when they become more engaged economically with the rising age. The study findings also showed that education years of agri-entrepreneurs would negatively influence the possibility of agricultural credit access from microfinance banks. Also, agri-entrepreneurs’ gender is a significant factor that influences probability of agricultural credit access from mobile banks platforms. Distance covered to the banks was a factor that influenced the likelihood of agricultural credit access in mobile banks and microfinance banks. The rapid success and adoption of mobile money services like M-Pesa in Kenya have facilitated diffusion and penetration of mobile-based savings and loans (Gichuki and Mulu-Mutuku 2018). FSD (2016) observed that financial institutions and microfinance institutions are harnessing the idea of adopting the mobile agent banking system. This has replaced the traditional banking model of physically walking to banking halls for deposits and collections.

Table 3.

Estimate on the Determinants of Agricultural Credit Access.

Characteristic	Mobile Banking Platform Marginal Effect	Commercial Banks Marginal Effect	Micro-finance Banks Marginal Effect	Table Banking Groups Marginal Effect
Age of agri-entrepreneurs	0.104**	0.762	–0.007*	0.002***
Gender of the agri-entrepreneurs (1 = male)	2.643**	–0.067	0.150	0.018
Agri-entrepreneurs’ years of education	0.131	0.017	–0.051***	–0.001
HH dependency ratio	7.270**	0.005**	0.052	0.002***
Farm size (acres)	–6.700	–0.011	–0.005	–0.048
Title deed land	–0.543	–0.427***	0.178	–0.077
Membership in social network groups (1 = yes)	–1.001	0.135	0.115	0.095**
Average annual income (Ksh)	4.090	6.020**	6.230*	2.960
Value of liquid assets (Ksh)	–3.260	1.950	6.260	–7.720
Value of livestock (Ksh)	1.220***	2.140	5.850**	2.780**
Outstanding dept. (Ksh)	1.320*	–1.150	6.560**	0.037***
Distance to banks (minutes)	2.37***	0.014	0.060*	–0.054
Distance to shopping centre (minutes)	6.22*	0.041	–0.002	0.074
Cons		–0.743	–0.534	–7.709
Wald χ²
LR χ²	36.63	34.77	25.48	37.91
Pseudo R²	0.459	0.203	0.257	0.531
Log likelihood	–28.65	–67.75	–41.44	–42.19

Source: Compiled by authors from survey data.

Notes: *, **, *** Indicate statistically significant at the 10%, 5%, and 1% levels, respectively.

Determinants of Agricultural Credit Rationing by Lending Institutions

The determinants of agricultural credit rationing by the lending institutions are presented in Table 4. As previously indicated, the Heckman two-step selection model will be used to examine the factors that determine the extent of credit rationing parameters applied by commercial banking institutions, microfinance, mobile banking platforms and table banking groups. Despite Erec Heimfarth and Musshoff (2011) observing that agricultural income in emerging economies is considered to be volatile due to weather, diseases, pest and market instability, our study findings reveal that SMS agri-enterprises’ income would determine the amount of agricultural credit accessed from commercial banks.

The study results further reveal that SMS agri-enterprises’ income would determine the amount of agricultural credit accessed from the mobile banking platform. This can be explained by the fact that in mobile credit platforms, borrowers take very short time to know the decision on whether the loan was granted after application. In addition, the process is swift and requires less paperwork, and eligibility and amount credit granted requires developing good credit rating scores and less of income status. Mobile money technology has enabled easy transfer, deposit, savings and withdrawal of money; these services were previously unavailable to the majority of rural populations who rely on farming (Omwansa and Waema 2014).

The value of livestock owned by SMS agri-enterprises determines the amount of agricultural credit offered by microfinance banks and table banking groups. The findings reveal that titled deed farm would significantly guarantee SMS agri-enterprises to receive agricultural credit applied for from commercial banks. While explaining the casual relationship between land title and credit access, Boucher, Barham, and Carter (2005) noted that formal lenders with limited information on the borrowers require significant collateral to mitigate the risk of defaults. Also, lack of land titles prevents full collateralisation of the main asset owned by poor farming households in developing countries. The study estimates also reveal that shorter distance to the bank positively influenced the agricultural credit amount received from mobile banks. Generally, mobile banking has credit processing efficiency that is cost-effective and not offered by other lending platforms. Using their phones, agri-entrepreneurs easily repay credit and make cash withdrawals from mobile agents in rural centres. Unlike in conventional banks, it would take agri-entrepreneurs longer time to process credit and access cash from the banking branch located far from the farms. The study results reveal that membership in social network organisations would significantly determine the agricultural credit amount accessed from informal banking groups. Asante-Addo et al. (2017) and Cheng and Ahmed (2014) also found evidence that membership association by agri-entrepreneurs significantly increased probability of SMS agri-enterprises in accessing credit amount applied for 23%.

Table 4.

Estimates on the Determinants of Being Credit Rationed.

Coef.	Mobile Banking		Commercial Banks		Micro-finance Banks		Table Banking Groups
Coef.	Coef.	Std. Err.	Coef.	Std. Err.	Coef.	Std. Err.	Coef.	Std. Err.
Agri-entrepreneurs’ age	0.042**	0.019	0.002	0.009	–0.012	0.011	0.037**	0.012
Gender of the agri-entrepreneurs (1 = male)	1.103**	0.460	–0.178	0.308	0.199	0.393	0.289	0.372
Agri-entrepreneurs’ years of education	0.054	0.087	0.041	0.039	–0.147**	0.051	–0.030	0.062
HH dependency ratio	0.026*	0.014	0.016**	0.006	0.004	0.008	0.032***	0.009
Farm size (acres)	–0.004	0.036	–0.029	0.029	–0.019	0.036	–0.075	0.062
Title deed land	–0.163	0.424	1.153**	0.320	0.435	0.410	–0.119	0.456
Membership in social network groups (1 = yes)	–0.378	0.422	0.373	0.268	0.453	0.384	1.517***	0.441
Average annual income (Ksh)	1.860*	1.970	1.560**	6.910	1.640	1.110	4.770	1.260
Value of liquid assets (Ksh)	–13.00*	7.010	5.570*	3.240	1.450	1.220	0.037	1.120
Value of livestock (Ksh)	4.910	2.910	6.320	4.250	16.05**	6.340	4.410**	1.960
Outstanding dept. (Ksh)	5.330**	2.120	–3.000	4.800	19.05**	7.920	0.006**	0.013
Distance to banks (minutes)	0.022	0.016	0.883	2.161	–0.005	0.256	–0.087	0.099
Distance to shopping centre (minutes)	0.009**	0.004	0.401	0.267	0.013	0.010	0.011	0.020
Cons	–4.752	1.717	–0.743	0.766	–0.618	1.194	–7.697	1.466
Inverse Mils ratio lambda	2041.	9351	–15869		–6588.2		8657
Rho	0.109		–0.351		–0.407		0.919
Sigma	18725		45176		16`182		9416.5
Number of observations	120		126				159
Wald χ² (13)	46.41		45.07		85.22
Uncensored observations	110		75		27		120
Censured observations			51		79		39
Prob. > χ²	0.000		0.2840				75.99

Source: Compiled by authors from survey data.

Notes: *, **, *** Indicate statistically significant at the 10%, 5%, and 1% levels, respectively.

Table 5.

Estimates on Determinants of Agricultural Credit Delinquency.

	Mobile Banking		Commercial Banks		Micro-finance Banks		Table Banking Groups
	Coef.	Std. Err.	Coef.	Std. Err.	Coef.	Std. Err.	Coef.	Std. Err.
Agri-entrepreneurs’ age	–0.007	0.019	–0.021	0.018	0.096^***	0.027	–0.010	0.012
Gender of agri-entrepreneurs (1 = male)	–0.212	0.481	0.1915	0.610	–0.072	0.706	–0.298	0.374
Agri-entrepreneurs’ years of education	0.296^***	0.090	–0.057	0.071	0.044	0.070	–0.056	0.051
HH dependency ratio	–0.057	0.036	0.024^*	0.013	0.031^**	0.014	0.032^**	0.015
Farm size (acres)	0.030	0.047	–0.037	0.054	–0.051	0.054	0.009	0.066
Title deed land	–0.224	0.438	–0.314	0.568	0.330	0.663	–0.979^*	0.540
Membership in social network groups (1 = yes)	0.457	0.396	–0.427	0.552	0.887	0.718	2.529^**	1.140
Average annual income (Ksh)	–1.480	1.790	1.710^*	1.040	–1.080	2.280	3.020	1.510
Value of liquid assets (Ksh)	–2.220^**	1.020	1.250^**	6.1208	1.290	1.980	–9.430	1.487
Value of livestock (Ksh)	–2.108	4.860	–6.110	6.400	–3.510	3.480	–2.140	2.210
Outstanding dept. (Ksh)	4.920	1050	5.950	1.105	0.029	0.002	0.048	0.418
Distance to banks (minutes)	–0.001^**	0.005	0.002	0.004	–0.008	0.003	–0.002	0.011
Distance to shopping centre (minutes)	0.078^**	0.003	0.028	0.046	0.017	0.016	0.004^*	0.002
Credit amount accessed	–2.770	4.730	6.480^*	5.770	–0.021	0.018	0.267	0.020
Interstates	0.174^*	0.096	–0.189	0.191	0.041	0.148	–0.065	0.089
Number of observations	110		75		52		120
Log pseudolikelihood	–123.6		0.089		0.1519		0.075
Log likelihood	51.16		–90.59		–58.22		–158.8
Wald χ² (15)			17.89		20.86		26.11
Pseudo R²	0.1048		0.089		0.151		0.075
Prob > χ²	0.0000		0.268		0.1412		0.036

Source: Compiled by authors from survey data.

Notes: *, **, *** Indicate statistically significant at the 10%, 5%, and 1% levels, respectively.

Determinants of Agricultural Credit Delinquency

The estimate results for agricultural credit delinquencies are presented in Table 5, as indicated previously. OLS model is used to conduct the empirical analysis where the dependent variable is the percentage ratio of agricultural credit delinquency rate, while the independent variable includes agri-entrepreneurs’ personal characteristics, agri-entrepreneurs’ household financial characteristics and social capital characteristics. Interestingly, our findings indicate that agricultural credit delinquency in mobile banking would be significantly reduced by distance to the banking agent. This implies that the availability of mobile banking agencies within village centres would considerably enable farmers to make credit repayment within required period. Whilst mobile banking platform is considered to provide small and low amount of credit, we note that interest rates significantly influence credit delinquency. The results further reveal that the value of assets owned would significantly reduce agricultural credit delinquency in mobile banking.

The study investigation also found that dependence ratio would influence the likelihood of agricultural credit delinquency from commercial banks, microfinance and table banking groups. Previous studies show that mainly, the unforeseen household expenses and dependence from relatives would endanger loan repayment (Maurer 2014). Interestingly, the amount of credit granted would positively and significantly influence agricultural credit delinquency in commercial banks. This finding is consistent with Weber and Musshoff (2012) who also find a significant relationship between larger disbursed loan amounts and higher loan delinquency rates. Similarly, Maurer (2014) attributes higher credit risks to larger amount of loans. Considering that more women are likely to source credit from table banking groups banks than men, we note that ownership of land that has a title deed would significantly reduce agricultural credit delinquency in table banking groups. However, the distance covered to the shopping centres where table banking group meetings take place has a significant influence on agricultural credit delinquency rates. Membership in social organisations significantly influenced the likelihood of farmer’s delinquency rates for table banking groups. This can be explained by the fact that farmers are likely to use fellow group members as credit guarantors.

Conclusion

This article explored and compared the determinants of credit agricultural credit access in commercial banking institutions, microfinance banks, mobile banking platforms and table banking groups in Kenya. In the study, we apply Garrett ranking technique to describe factors likely to influence agricultural credit access for SMS agri-enterprises. To empirically examine and compare factors that are likely to influence SMS agri-enterprises receiving agricultural credit from four lending platforms, the study used probit model. We use Heckman model to empirically explore and compare factors that determine the extent of agricultural credit rationing across different lending institutions, while in the third stage, the study used OLS model to establish the factors that would influence agricultural credit delinquency in lending institutions.

The main observations from the study revealed that lack of collateral required by banks was considered as the main impediment to agricultural credit access by agri-entrepreneurs. Also, high-interest rates were a concern when considering agriculture credit access from the available credit sources in Kenya. Nevertheless, we note that the penetration of mobile banking credit that requires less collateral and depends on credit scoring to provide credit was the preferred agriculture credit source for 31% of agri-entrepreneurs, while table banking groups is an important source of agriculture credit for 25% of surveyed agri-entrepreneurs.

The study results revealed that the household dependency ratio increased the probability of accessing agricultural credit from mobile banks, commercial banks and table banking groups. Considering the fact that some microfinance banks require clients to have required amount of savings that determine the credit amount accessed, we observe that SMS agri-enterprises income significantly determines the probability of obtaining agricultural credit from microfinance banks. Also, the proximity to mobile banking agents increases the probability of agri-entrepreneurs accessing agricultural credit from mobile banks. Similarly, banking proximity to the banks increases the probability of agri-entrepreneurs accessing agricultural credit from microfinance banks. This is partly explained by the microfinance model that facilitates banking agents to serve clients at their locations rather than making long trips to the banking halls.

With regard to factors that influence agricultural credit rationing, the study results indicate that title deed farms, the value of assets and income significantly determine whether agri-entrepreneurs received the desired agricultural credit amount from commercial banks. Conventionally, formal banking systems heavily rely on collateral such as farm machinery, titled land and household assets information to verify and determine the amount of credit to disburse to agri-entrepreneurs. The other important observation of the study was that proximity to banking agents determined the amount of agricultural credit accessed from mobile banks. Most of the agri-entrepreneurs who are financially excluded from formal banking systems are likely to seek credit from mobile banking networks. Mobile banks’ credit relies on credit scoring for loan disbursement unlike the traditional banking system, which relies mainly on collateral, profile documentation and enterprise financial cash flow to guarantee for credit. Also, credit application process takes less time, and mobile agents offer deposits and withdraw services that are easily convenient and accessible within the rural settings.

The study findings illustrate that the amount of agricultural credit borrowed would determine agricultural credit delinquency rates in commercial banks. Generally, financial banks find it more costly to lend and collect money from small enterprises than a big firm. Also, the findings showed that accessing high mount of agricultural credit would influence delinquency in mobile banks, formal banks, microfinance and informal banks.

In conclusion, we note that most of the studies report that women farmers are likely to produce less because they have lower quality land, less access to fertiliser and other inputs and receive less credit and extension support. FAO’s State of Food and Agriculture reports that if women had access to productive resources as men, they are likely to increase their farm yield by 20–30%. From the study, we observed that despite the rapid rate of financial inclusion, women agri-entrepreneurs mainly rely on informal banks for agricultural credit, which tends to be unreliable. Also, women-owned SMS agri-enterprises are likely not to control productive resources like land that has a title deed. This limits the women agri-entrepreneurs’ ability to access agricultural credit from established lending institutions. The study recommends development of agricultural credit policy accessibility in rural areas. In addition, the agricultural credit products that are flexible and have longer repayment periods to cover the periods between capital investment (planting) and revenue generation (harvesting) should be established.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

ORCID iD

Castro N. Gichuki

References

Ambec

, and Treich

2007. ‘Roscas as Financial Agreements to Cope with Self-control Problems’. Journal of Development Economics 82 (1): 120–37.

Anjugam

, and Ramasamy

2007. ‘Determinants of Women’s Participation in Self-help Group (shg)-led Microfinance Programme in Tamil Nadu’. Agricultural Economics Research Review 20 (347-2016-16640): 283–98.

Arun

T. G.

, and Bendig

2010. ‘Risk Management Among the Poor: The Case of Microfinancial Services’. IZA Discussion Papers, No. 5174. Bonn: IZA.

Asante-Addo

, Mockshell

Jonathan

, Zeller

Manfred

, Siddig

Khalid

, and Egyir

Irene S.

2017. ‘Agricultural Credit Provision: What Really Determines Farmers’ Participation and Credit Rationing?’ Agricultural Finance Review 77(2): 239–56.

Boucher

S. R.

, Barham

Bradford L.

, and Carter

Michael R.

2005. ‘The Impact of “Market-friendly” Reforms on Credit and Land Markets in Honduras and Nicaragua’. World Development 33 (1): 107–28.

Cheng

, and Ahmed

A. D.

2014. ‘The Demand for Credit, Credit Rationing and the Role of Microfinance: Evidence from Poor Rural Counties of China’. China Agricultural Economic Review 6 (2): 295–315.

Demirgüç-Kunt

, and Klapper

2013. ‘Measuring Financial Inclusion: Explaining Variation in Use of Financial Services Across and Within Countries’. Brookings Papers on Economic Activity 2013 (1): 279–340.

Dong

, Lu

Jing

, and Featherstone

Allen M.

2012. ‘Effects of Credit Constraints on Household Productivity in Rural China’. Agricultural Finance Review 72 (3): 402–15.

Doss

2014. ‘If Women Hold Up Half the Sky, How Much of the World’s Food Do They Produce?’ In Gender in Agriculture, edited by Agnes

, Quisumbing Ruth

Meinzen-Dick

, Raney

Terri L.

, André

Croppenstedt

, Behrman

Julia A.

, and Amber

Peterman

, 69–88. Dordrecht: Springer.

10.

Dutta

, and Magableh

2006. ‘A Socio-economic Study of the Borrowing Process: The Case of Microentrepreneurs in Jordan’. Applied Economics 38 (14): 1627–40.

11.

Erec Heimfarth

, and Musshoff

O. J. A. F. R.

2011. ‘Weather Index-based Insurances for Farmers in the North China Plain: An Analysis of Risk Reduction Potential and Basis Risk’. Agricultural Finance Review 71 (2): 218–39.

12.

Fitawek

, and Kalaba

2017. ‘Rural Financial Services and Effects on Livestock Production in Ethiopia’. International Conference on Livestock Finance and Value Chain. Royal Swazi Spa, Ezulwini, Swaziland.

13.

FSD. 2016. FinAccess Household Survey . Nairobi, Kenya: Finacial Sector Deepening

14.

Garrett

, and Woodworth

1969. Statistics in Psychology and Education , 5th Indian ed.Bombay: Vakils, Feffer and Simons Private Ltd.

15.

Gichuki

C. N.

, and Mulu-Mutuku

2018. ‘Determinants of Awareness and Adoption of Mobile Money Technologies: Evidence from Women Micro Entrepreneurs in Kenya’. Women’s Studies International Forum 67: 18–22.

16.

Gichuki

C. N.

, Mutuku

Milcah Mulu

, and Lydia

Nkatha Kinuthia.

2015. ‘Influence of Participation in “Table banking” on the Size of Women-owned Micro and Small Enterprises in Kenya’. Journal of Enterprising Communities: People and Places in the Global Economy 9 (4): 315–26.

17.

GSMA. 2012. ‘Mobile Infrastructure Sharing. S. Report’. https://www.gsma.com/publicpolicy/wp-content/uploads/2012/09/Mobile-Infrastructure-sharing.pdf

18.

Gugerty

M. K.

2007. ‘You Can’t Save Alone: Commitment in Rotating Savings and Credit Associations in Kenya’. Economic Development and Cultural Change 55 (2): 251–82.

19.

Helmore

and CARE/International. 2009. Bringing Financial Services to Africa’s Poor, Microfinance in Africa: State of the Sector Report . Geneva: CARE International.

20.

Kibet

L. K.

, Benjamin

, Mutai

Desterio E. Ouma

, Ouma

Shem A.

, and George

Owuor.

2009. Determinants of Household Saving: Case Study of Smallholder Farmers, Entrepreneurs and Teachers in Rural Areas of Kenya’. Journal of Development Agricultural Economics 1 (7): 137–43.

21.

Kochar

1997. ‘An Empirical Investigation of Rationing Constraints in Rural Credit Markets in India’. Journal of Development Economics 53 (2): 339–71.

22.

Kumar

C. S.

, Turvey

Calum G.

, and Kropp

Jaclyn D.

2013. ‘The Impact of Credit Constraints on Farm Households: Survey Results from India and China’. Applied Economic Perspectives and Policy 35 (3): 508–27.

23.

Kumari

(2017). Microfinance through Women Self-Help Groups (SHGs) for Grass-root Level Empowerment: An Empirical Study of Varanasi, Uttar Pradesh, India . Kassel: Kassel University Press GmbH.

24.

Kusimba

, Kunyu

Gabriel

, and Elizabeth

Gross.

2013. ‘Social Networks of Mobile Money in Kenya’. Working Paper 1. Irvine: Institute for Money, Technology and Financial Inclusion.

25.

Lukytawati

2009. ‘Factors Influencing Participation and Credit Cconstraints of a Financial Self-help Group in a Remote Rural Area: The Case of ROSCA and ASCRA in Kemang Village West Java’. Journal of Applied Sciences 9 (11): 2067–77.

26.

Maurer

2014. ‘Where is the Risk? Is Agricultural Banking Really More Difficult than Other Sectors?’ In Finance for Food, edited by Kohn

, 139–65. Berlin, Heidelberg: Springer.

27.

Mbo’o-Tchouawou

, and Colverson

K. E.

2014. Increasing Access to Agricultural Extension and Advisory Services: How Effective are New Approaches in Reaching Women Farmers in Rural Areas? ILRI (aka ILCA and ILRAD).

28.

Mohamed

K. S.

, and Temu

A. E.

2008. ‘Access to Credit and Its Effect on the Adoption of Agricultural Technologies: The Case of Zanzibar’. African Review of Money Finance and Banking 2008: 45–89.

29.

Mukherjee

2013. ‘Coexistence of Formal and Informal Credit Markets in India-A Study of Entrepreneurial Choices for House-based Industries’. Research in Applied Economics 5 (3): 37.

30.

Mulu-Mutuku

, and Gichuki

C. N.

2017. ‘Influence of Mobile Money on Control of Productive Resources Among Women Micro Entrepreneurs Participating in Table Banking in Nakuru, Kenya’. IMTFI Final Report. https://www.imtfi.uci.edu/research/2015/mulu-mutuku_gichuki_2015.php

31.

Okurut

, Schoombee

Andrie

, and Servaas van der

Berg.

2004. ‘Credit Demand Rationing in the Informal Financial Sector in Uganda’. Paper on the DPRU/Tips. Cornell Conference on African Development and Poverty Reduction.

32.

Omwansa

, and Waema

2014. ‘The Impact of Pure Mobile Microfinancing on the Poor: Kenya’s Musoni Experience’. Working Paper 2. California, Irvine, IMFTI.

33.

Raghunathan

U. K.

, Escalante

Cesar

, Dorfman

Jeffrey

, Ames

Glenn C. W.

, and Houston

Jack E.

2011. ‘The Effect of Agriculture on Repayment Efficiency: A Look at MFI Borrowing Groups’. Agricultural Economics 42 (4): 465–74.

34.

Rahji

, and Fakayode

2009. ‘A Multinomial Logit Analysis of Agricultural Credit Rationing by Commercial Banks in Nigeria’. International Research Journal of Finance Economics 24 (91): 97–103.

35.

Reyes

, and Lensink

2011. ‘The Credit Constraints of Market-oriented Farmers in Chile’. Journal of Development Studies 47 (12): 1851–68.

36.

Saldias

, and von Cramon-Taubadel

2012 ‘Access to Credit and the Determinants of Technical Inefficiency Among Specialized Small Farmers in Chile, Diskussionsbeitrag’. Discussion Papers 1211. Göttingen: Department of Agricultural Economics and Rural Development (DARE), Georg-August University of Göttingen.

37.

Sheahan

, and Barrett

C. B.

2014. Understanding the Agricultural Input Landscape in sub-Saharan Africa: Recent Plot, Household, and Community-level Evidence . Washington: The World Bank.

38.

Swinnen

J. F.

, and Gow

H. R.

1999. ‘Agricultural Credit Problems and Policies During the Transition to a Market Economy in Central and Eastern Europe’. Food Policy 24 (1): 21–47.

39.

Tadesse

2014. ‘Fertilizer Adoption, Credit Access, and Safety Nets in Rural Ethiopia’. Agricultural Finance Review 74 (3): 290–310.

40.

Van Rooyen

, Stewart

, and de Wet

2012. ‘The Impact of Microfinance in sub-Saharan Africa: A Systematic Review of the Evidence’. World Development 40 (11): 2249–62.

41.

Weber

, and Musshoff

2012. ‘Is Agricultural Microcredit Really More Risky? Evidence from Tanzania’. Agricultural Finance Review 72 (3): 416–35.

42.

Yuan

, and Gao

2012. ‘Farmers’ Financial Choices and Informal Credit Markets in China’. China Agricultural Economic Review 4 (2): 216–32.