1. Introduction
Smallholder farmers play a significant role in the agricultural sector in Malawi, with over 80% of the population relying on this sector for their livelihoods
| [1] | Tuni, A. (2022). "Barriers to commercialise produce for smallholder farmers in Malawi: An interpretive structural modelling approach." Journal of Rural Studies 93, 1-17. |
[1]
. However, smallholder farmers in Malawi face numerous challenges, such as limited access to credit and financial services, which limits their ability to invest in their farms and increase their productivity. As a result, smallholder farmers in Malawi face persistent poverty, which undermines the country's efforts to improve individual household income and consumption levels and their overall standards of living
| [2] | Rebecca & Kiggunduthe. (2021). The Role of Microfinance in Reducing Poverty: A Case Study of Smallholder Farmers in Malawi. The Journal of International Development. |
[2]
. It is also a critical issue, as it is closely linked to Malawi's pillar-1 of 2063 vision of reducing poverty and improving the standards of living for all Malawians through agricultural productivity and commercialisation.
Malawi's Vision 2063 is a long-term development strategy that aims at transforming Malawi into a prosperous and developed country by 2063
| [10] | GoM. (2020). Malawi Vision 2063. Malawi Government Policy Document 2020. |
[10]
. The vision focuses on economic transformation, governance, human development, and regional integration. In order to achieve this vision, Malawi must address the challenges faced by its smallholder farmers and support their access to credit and financial services, enabling them to increase their productivity and resilience to shocks
| [3] | NPC, N. P. (2020). Malawi Vision 2063. Malawi 2063 Policy Document, 38. |
[3]
.
Food and Agriculture Organisation (FAO) defines smallholder farmers as small-scale farmers, pastoralists, forest keepers, fishers who manage areas varying from less than one hectare to ten hectares of land for agricultural purposes, and are characterized by family-focused motives such as favouring the stability of the farm household system, using mainly family labour for production and using part of the produce for family consumption
| [4] | FAO. (2012). Malawi Smallholder Agricutural Credit Project. Geneva: United Nations. |
[4]
.
According to the National Statistics Office (2020), the total land area under cultivation in Malawi is about 3.8 million hectares. Smallholder farmers cultivate small and fragmented land holdings of less than one hectare (on average 0.69 ha) under customary land tenure arrangements and produce lower crop yields than those produced in the estate subsector, and mong these smallholder farmers, female-headed households cultivate relatively smaller land holdings than their male-headed counterparts (0.84 ha compared to 1.61 ha)
| [5] | IHS-5. (2019/2020). International Household Survey Report. Zomba: Malawi National Statistics Office. |
[5]
.
Access to credit by smallholder farmers in Malawi refers to the ability of these farmers to secure loans and financial support from various sources such as banks, microfinance institutions, and government agencies to invest in their agricultural activities
| [6] | Argaw. (2017). Analysis of Determinants of Access to Credit among Smallholder Farmers in Edja District, Guraghe Zone, SNNPR, Ethiopia. Journal of Economics and Sustainable Development, 8(13), 73–80. https://core.ac.uk/reader/234647920 |
[6]
. This access is critical for the financial stability and growth of smallholder farmers in the country, allowing them to purchase inputs, upgrade their production technologies, and improve their livelihoods. Agricultural credit schemes have the potential to provide smallholder farmers with the necessary financial boost to increase their yields and in turn reduce poverty, but unfortunately, smallholder farmers in Malawi face various challenges in accessing credit, including limited access to formal financial institutions, low levels of financial literacy, and lack of collateral
| [7] | Salima, L. P. (2023). "The impact of credit access on household food security in Malawi.” Journal of Agriculture and Food Research 11, 2-3. |
[7]
.
In Malawi, there have been various initiatives aimed at increasing access to credit for smallholder farmers, including government-led programs and private sector initiatives. For example, Malawi government through the Ministry of Agriculture has established strategies and policies like the financial inclusion policy, agricultural credit schemes and Malawi Growth and Development Strategy that play crucial roles in promoting farmers' access to credit in Malawi.
Financial inclusion policy is aimed at increasing access to financial services by the unbanked and underserved population, including farmers, to enable them to access credit facilities
| [8] | GoM. (2014). The Malawi National Strategy for Financial Inclusion (2010-2014). Financial Inclusion Policy. |
[8]
. Agricultural credit schemes, on the other hand, are designed specifically to support the agriculture sector by providing farmers with credit facilities that are essential for their livelihoods. The Malawi Growth and Development Strategy (MGDS I, II and III), also recognizes the importance of agriculture in the country's economy and aim at increasing the productivity of the sector by promoting the development of agriculture-based industries and providing farmers with credit facilities
| [9] | GoM. (2017). The Malawi Growth and Development Strategy (MGDS III). (MGDS III), Malawi Government Policy Document 2017. |
[9]
.
Despite the Government efforts discussed above, poverty still remains a persistent problem for smallholder farmers in Malawi. The new World Bank Poverty Assessment Report finds that over half of the Malawian population (50.7%) are still poor, almost no different from a decade ago and our high reliance on rain-fed smallholder agriculture is one of the core drivers of these stagnant poverty levels
. Furthermore, out of 15 million smallholder farmers in Malawi, very few farmers have access to credit from formal financial institutions in Malawi, which is the only formal alternative of our limited government subsidies in securing modern farm inputs like seeds, fertilizers, technologies and machinery required to improve farm outputs as a step towards poverty reduction
.
1.1. Problem Statement
In Malawi, smallholder farmers make up nearly 80% of the population, and 50.7% of them are still living in poverty, not different from a decade ago
| [11] | Caruso, G. D. (2022). Malawi Poverty Assessment: Poverty Persistence in Malawi - Climate Shocks, Low Agricultural Productivity, and Slow Structural Transformation (English). Washington, D. C. World Bank Group, http://documents.worldbank.org/curated/en/099920006302215250/P174 |
| [1] | Tuni, A. (2022). "Barriers to commercialise produce for smallholder farmers in Malawi: An interpretive structural modelling approach." Journal of Rural Studies 93, 1-17. |
[11, 1]
. This occurred during the transition from MGDS I to MGDS III, whose goal was to align with Malawi's long-term development vision of transforming the agricultural sector and reducing poverty in Malawi. It is therefore a major mystery why the MGDS failed to achieve these goals. This makes it doubtful that poverty will be reduced by the year 2063.
The main reason why the majority of Malawians find it difficult to enhance their economic wellbeing is lack of financial boosts like credit facilities to cushion them in times of unfavourable conditions. It is the same case with smallholder farmers in Malawi
| [12] | Kaluwa & Kunyenje. (2019). "The paradox of the financial inclusion-poverty nexus in Malawi.". African Review of Economics and Finance 11, no. 2, 38-66. |
| [7] | Salima, L. P. (2023). "The impact of credit access on household food security in Malawi.” Journal of Agriculture and Food Research 11, 2-3. |
[12, 7]
.
Government and non-governmental organizations have made a lot of efforts to reduce poverty among smallholder farmers in Malawi, and such interventions include farm input subsidies, agricultural training, and civic education campaigns on modern agriculture. However, there hasn't been much done in terms of financial inclusion (credit access to smallholder farmers).
There has been limited research conducted using the current IHS5 data that analysed the direct causal relationship between credit access and poverty. The available literature only stresses the relationship that credit access have on food security, and not poverty which is a broad and dominant variable, hence the need for further research
| [13] | Chilimba, A. (2020). The Impact of Microfinance Programme Participation on Household Food Security in Malawi. International Journal of Economics and Finance Studies. SA. |
| [7] | Salima, L. P. (2023). "The impact of credit access on household food security in Malawi.” Journal of Agriculture and Food Research 11, 2-3. |
[13, 7]
.
1.2. Main Objective
The main objective of this study is to analyse the effects of smallholder farmers’ access to credit on household poverty in Malawi.
1.3. Specific Objectives
1) To analyse the factors that affect smallholder farmers access to credit in Malawi.
2) To estimate the impact of smallholder farmers access to credit on poverty.
1.4. Hypotheses Tested
Based on the above objectives, the study will test the following null hypotheses:
1) There are no factors that affect smallholder farmers access to credit in Malawi.
2) Smallholder farmer’s access to credit does not affect poverty.
2. Literature Review
2.1. The Basic McKinnon–Shaw Hypothesis
The McKinnon-Shaw Hypothesis serves as the foundation for the direct impacts of finance on poverty. The main tenet of the hypothesis is that financial markets will operate in a way that makes it possible for the underprivileged to have access to more effective and pertinent services when financial markets are well-developed and real interest rates are positive. According to the complementarity theory put forth by
| [14] | McKinnon, R. (1973). Money and Capital in Economic Development. The Brookings, Washington DC. |
[14]
, physical and financial assets complement one another. The two presumptions that support the theory. The first is that economic agents can only finance themselves, and the second is that those profitable ventures have a lot of indivisibilities. The algorithm does not differentiate between investors and savers. Due to the indivisible nature of investments, a prospective investor must accumulate funds until they are adequate to make the intended investment. As a consequence, financial and physical capital are intertemporally complementary. The availability of profitable chances for saving and consequently accumulating can help the poor by financial services. In this manner, the savings channel, or "conduit effect," of financial services can have an impact on poverty. In order for savers to be motivated to save based on the prospect of interest income, interest rates must be liberalized and permitted to be determined by the market. McKinnon's Hypothesis is occasionally referred to as an outside money paradigm because it relies on self-financing.
Like McKinnon, Shaw supports financial market liberalization and claims that higher interest rates will boost savers' incomes and provide more chances to diversify domestic asset portfolios. Shaw's theory does not demand that investors be self-financing, in opposition to McKinnon. According to his debt-intermediation theory, a "collection of financial marketplaces" that act as intermediaries connect savers and investors
| [15] | Shaw, E. (1973). Financial Deepening in Economic Development. Oxford University Press. |
[15]
. It is an inside money strategy as a result. These middlemen draw assets, which raises the amount of money available for loans. Shaw's debt intermediation theory thus adds a credit channel of financing to poverty in addition to the focus on saving.
Shaw further contends that by giving knowledge about returns on savers' deposits to those who, in his view, are "working under a handicap of ignorance," financial intermediaries increase the effectiveness of the financial system and decrease allocative waste. Financial markets are fragmented as a consequence of perceived information as being expensive and incomplete. Similar to this, these marketplaces show severe credit rationing. This fragmentation is minimized by financial development, which also lowers borrowing costs and improves access to credit for the poor.
According to
| [15] | Shaw, E. (1973). Financial Deepening in Economic Development. Oxford University Press. |
[15]
, money is a debt owed to the monetary system, and its main function is to be used as a method of exchange. He contends that the amount requested is closely related to the use of money as a form of payment. Along with the savings and credit channels mentioned in both McKinnon and Shaw's hypotheses, this adds the payments channel through which finance can have an impact on the poor. The financial industry can contribute to the reduction of poverty by offering dependable and affordable payment services for regular transactions and remittances. An accommodating payment method fills a crucial exclusionary gap for the impoverished, who are typically on the margins. This makes it easy and affordable for them to engage in mainstream finance. As they send and receive payments at reasonable rates, it becomes simpler for them to trade, which has an immediate effect on their day-to-day activities.
2.2. Dealing with Poverty Through Finance
Increasing Access to Credit
Credit, in addition to saving, can improve the likelihood that the poor will take advantage of opportunities to generate income. According to the literature, increased entrepreneurship is favourably correlated with poorer people having greater access to credit. Because they are likely to have very little savings, people in poverty cannot take advantage of economic chances because of their low incomes.
Similar to this, because of market frictions, the impoverished also tend to save in kind. Access to the credit market can lessen these frictions and enable the poor to engage in micro and small businesses, which may lead to a decline in poverty. Increasing microcredit, largely based on the Grameen Bank Plan, has been the strategy that has received the most widespread adoption
| [19] | Simatele, M. C. (2021). Financial Inclusion: Basic Theories and Empirical Evidence from African Countries. In M. C. Simatele, Financial Inclusion (pp. 9-34). Cape Town, South Africa: AOSIS Publishing. |
[19]
.
2.3. Empirical Review
Access to credit for smallholder farmers is a critical element for the economic development of rural communities in Malawi and all sub-Saharan African countries, where agriculture remains a key sector for livelihoods and poverty reduction. In Malawi, like many other countries in the region, smallholder farmers face significant challenges in accessing credit due to limited financial infrastructure and low levels of financial literacy like alluded to earlier on. The consequences of limited access to credit for these farmers are often dire, including reduced investment in agricultural production, lower productivity, and a cycle of poverty.
Empirical studies have explored the relationship between credit access and poverty reduction among farmers in sub-Saharan Africa, including in Malawi. These studies have examined the effects of different types of credit programs, including microfinance, agricultural credit, and government-supported programs, on farmers' income, productivity, and well-being.
In Malawi a research was conducted on how credit availability affected family food security
| [7] | Salima, L. P. (2023). "The impact of credit access on household food security in Malawi.” Journal of Agriculture and Food Research 11, 2-3. |
[7]
. This observational study looks at how access to formal or informal credit affects family food security in Malawi, a developing nation with a credit crunch. In light of possible endogeneity between credit access and food security, the research uses the fifth Integrated Household Survey (IHS5) and the Endogenous Regime Switching (ERS) method as well as the Tobit regression model. Indicators of informal credit access include access to extension services, size of landholdings, household size, and vulnerability to shocks, according to regression findings. Key determinants of access to formal credit include education level and household size. According to the research, while having access to formal credit raises household food security, having access to informal credit lowers food security overall. Different policy repercussions are inferred from these findings. This paper therefore lacked evidence on how poverty is directly affected, hence the need for further research.
Another researcher also analysed the impact of microfinance programme participation on household food security in Malawi
| [13] | Chilimba, A. (2020). The Impact of Microfinance Programme Participation on Household Food Security in Malawi. International Journal of Economics and Finance Studies. SA. |
[13]
. The study used cross-sectional data that was collected in Malawi for the Third Integrated Household Survey in 2010-2011. It employed the Heckman Selection Model, which was deemed applicable since the selection to participate in credit programmes is typically non-random. The study established that households that participated in microfinance programmes experience improvements in their status of food security. Their paper therefore lacked evidence on how poverty is directly affected, hence the need for further research that we embarked on.
Another research in Nigeria examined how small-scale farmers' access to credit could reduce their poverty in Kwara State
| [16] | Agba, D. Z. (2018). "Reducing Small Scale Farmers Poverty through Credit access in Kwara State Nigeria.". International Journal of Development and Sustainability 8, no. 2. |
[16]
. Under the guidance of the Landmark University economics department, a study was performed in 2018 on the use of credit by small-scale farmers and its effects on the reduction of poverty in the state of Kwara. The Cobb-Douglas production function, which was modified for use in the study, was used to calculate the productivity of small-scale farmers using the conventional least square technique. The research assessed both borrowers and non-borrower’s poverty levels as well as profitability and net farm income. The difference between credit users and non-users was found to be insignificant, despite the fact that credit users had greater productivity, profitability, and Net farm income. Additionally, it was discovered that farmers who had recourse to credit experienced poverty at a lower rate than farmers who did not. Hence, the study concludes that credit can ensure poverty reduction and also assist to include small scale farmers in the development process if it is made available in sufficient quantities. Because credit can have a positive effect on reducing poverty, the study suggests that formal financial institutions and the government work together to increase the amount of credit accessible to small-scale farmers.
Recently in 2022, another paper examined the effect of credit availability on welfare disparity in Malawi
| [17] | Dunga, M. F. (2022). "The Impact of Access to Credit on Welfare Inequality in Malawi.". Studia Universitatis Babes-Bolyai Oeconomica 67, no. 2. |
[17]
. In order to fill the gap left by earlier studies on credit, this research assesses the effects that access to credit has on welfare inequality in Malawi. The research used data from Malawi's 2017 Integrated Household Survey and propensity score analysis to look at how access to credit may affect household welfare in Malawi by using consumption per capita as a stand-in. The research went on to investigate the welfare disparities between households that access credit and those that do not using the generalized Lorenz curve, Theil indices, and the Gini coefficient. The findings indicated that households with access to credit experience lower levels of inequality than those without, which has a beneficial effect on welfare. However, a deeper look at the Theil's indexes revealed that variables unrelated to credit access had a greater impact on inter-household disparities than credit access itself. The findings suggest that access to credit has a beneficial influence on welfare inequality, though this impact is relatively small. Consequently, it is implied that policies aimed at improving credit distribution should persist. To get the desired outcome, a more comprehensive strategy for reducing inequality should be implemented simultaneously at the household and national levels.
Another recent paper also added to the body of knowledge by analysing how the central area of Ghana's access to credit and financial services affected the reduction of poverty in that year
| [18] | Ampah, S. N. (2017). "Effect of access to credit and financial services on poverty reduction in central region of Ghana.". International Journal of Business and Social Science 8, no. 8. |
[18]
. The goal of the research was to ascertain, from the viewpoints of Micro, Small and Medium Enterprises (MSME's), how access to credit and financial services affected the reduction of poverty in Ghana's Central Region. In most nations, micro, small, and medium-sized businesses play a major role in efforts to combat poverty and promote economic growth.
Many nations around the world have long acknowledged the significance of MSMEs. However, numerous studies have pointed to a lack of access to credit and financial services as the primary barrier to the expansion of MSMEs and the eradication of poverty in Ghana. This study's specific goal was to determine how financial services and credit availability impacted efforts to reduce poverty. Four sub-hypotheses were operationalized from one main hypothesis that covered the specific objective and the different indicators of poverty (growth in income, increase in consumption expenditure, acquisition of business assets and ability to educate children). This cross-sectional research sampled 370 owners of Micro Small and Medium Enterprises using cluster sampling methods. In an experiment that was conducted in November and December of 2016, a questionnaire was used as the data gathering tool. Cross tabulations and multiple regression analysis were carried out using SPSS. According to the research, having access to credit and financial services had a negligibly positive impact on income growth, consumer spending growth, and the purchase of corporate assets. However, the research discovered that the ability to educate children as a measure of poverty had a significant impact on access to credit and financial services. As a result, the research disproved the relevant null hypotheses. Along with pointing out limitations, the study draws conclusions and suggests ideas for additional research in light of the results.
Now adopting Grameen Bank Model, which is a microfinance lending model developed by Nobel Peace Prize winner Muhammad Yunus in the 1970s in Bangladesh. This research has separated access to formal credit through banks from the microfinance institutions credits which are of difference in nature to analyse their impact on smallholder farmer’s poverty. This model provides small loans, or microcredit, to impoverished individuals who lack access to traditional banking services. The loans are often used to start or expand small businesses, such as selling goods in a market or starting a small farm. The model places a strong emphasis on empowering poor people, who often face greater barriers to accessing credit and other financial services. As a result, over 90% of Grameen Bank's borrowers are women and the poor
| [19] | Simatele, M. C. (2021). Financial Inclusion: Basic Theories and Empirical Evidence from African Countries. In M. C. Simatele, Financial Inclusion (pp. 9-34). Cape Town, South Africa: AOSIS Publishing. |
[19]
.
The Grameen Bank Model has been widely praised for its success in alleviating poverty, promoting entrepreneurship, and empowering women. It has inspired the development of microfinance institutions around the world and has helped millions of people in poverty to improve their lives through access to credit and financial services.
3. Methodology and Results
The study adopted the Heckman probit selection model as recommended by Wooldridge and Heckman to correct and account for sample selection bias
| [20] | Wooldridge, J. (2015). Introductory econometrics: A modern approach. In J. Wooldridge, Introductory econometrics (pp. 557-564). Cengage learning. |
| [21] | Heckman. (1988). The Microeconomic Evaluation of Social Programs and Economic Institutions. Chung Hua Series of Lectures by Invited Eminent Economists (Vol. 14). The Institute of Economics, Academia Sinica, Taipei. |
[20, 21]
.
According to Heckman, the Heckman Probit selection model is a statistical model that is used to estimate the relationship between a binary outcome variable and a set of explanatory variables in the presence of sample selection bias
| [21] | Heckman. (1988). The Microeconomic Evaluation of Social Programs and Economic Institutions. Chung Hua Series of Lectures by Invited Eminent Economists (Vol. 14). The Institute of Economics, Academia Sinica, Taipei. |
[21]
. The model was developed by James Heckman, and it assumes that the sample selection bias arises from an unobserved variable, known as the selection variable.
In the Heckman Probit selection model, there are two equations: the selection equation and the outcome equation. The selection equation is a probit regression model that estimates the probability of selection into the sample. The outcome can be estimated using lin-log regression model to estimate the relationship between the binary outcome variable and the explanatory variables, after controlling for the selection bias.
The Heckman Probit selection model eliminates sample selection bias by accounting for the correlation between the selection variable and the outcome variable. The model estimates the impact of the selection variable on the outcome variable and corrects for this bias by incorporating the inverse Mills ratio, which is a function of the probit estimate from the selection equation, into the outcome equation.
The inverse Mills ratio is used as a correction factor in the outcome equation to adjust the coefficients of the explanatory variables for the selection bias. This correction factor ensures that the estimated coefficients reflect the relationship between the variables in the population, rather than the biased relationship that results from the sample selection.
3.1. Model Specification
Now borrowing a leaf from Masanjala, the probit model on smallholder farmer’s access to credit is specified as below
| [22] | Masanjala, W. (2006). Cash Crop liberalization and poverty alleviation in Africa: Evidence from Malawi. Journal of Agricultural Economics, 233. |
[22]
:
(1)
The welfare model adopted from
| [30] | Mukherjee & Benson. (1998). The Determinants of Poverty in Malawi, (pp. 340-341). |
[30]
is then modelled as follows:
(2)
An empirical problem that is common in this model is finding appropriate identification variables
| [22] | Masanjala, W. (2006). Cash Crop liberalization and poverty alleviation in Africa: Evidence from Malawi. Journal of Agricultural Economics, 233. |
[22]
. Estimation of participation is only possible when variables in the selection model and in the second welfare function are not common. In this case, the variables in the two models are different; thus, the model can be estimated.
: representing the natural logarithm of total daily consumption to measure welfare/poverty. This is in a continuous variable format. The more the households spends, the better their welfare.
represents both formal and informal credit access, these are all the households that actually received loans both from the Banks, loan institutions and money lenders.
: This is a community variable that describes the availability of any financial institution in the community. It is expected that households who live in communities that have these financial institutions will have an upper hand in accessing credit than those that literally don’t have them in their communities.
Residence (rural=1 and 0 otherwise): According to Kadale, Malawi's rural regions receive poor bank branch service, which limits their access to bank credits and other financial services
| [23] | Kadale, C. a. (2009). Supply Side Study of Financial Inclusion in Malawi. Final Report. Finscope Study. |
[23]
. It is anticipated that location and credit access will have a negative association. It is also anticipated that there will be a negative relationship between location and poverty reduction in rural areas because of such limitations on access to credit.
IMR: Inverse mills ratio is a variable that has been incorporated in the welfare function to capture the selected/estimated credit access. This study will use this main variable to see if it has impact on poverty or not as used by
| [13] | Chilimba, A. (2020). The Impact of Microfinance Programme Participation on Household Food Security in Malawi. International Journal of Economics and Finance Studies. SA. |
[13]
.
: This variable measures all land that the households own in the community. It is expected that it will have a positive relationship with credit access.
: This variable captures all livestock in total that the household own. It has a positive relationship with credit access.
Household Size: This variable captures the total number of people in a household. The larger the size of the household, the more they will spend and the higher will be their daily consumption, so as long as daily consumption is concerned, their relationship will be positive. Households of more people are likely to be well-off.
Head_Age: This variable captures the age of the household head. As the household age increases, there are more chances of being accepted and being able to access credit and so the relationship will be positive.
: Captures the potential non-linearity in the connection between age and credit availability
| [24] | Fungáčová & Weill. (2015). Understanding financial inclusion in China. China Economic Review, 34, pp. 196-206. |
[24]
. The likelihood that someone will have access to credit increases as age does. However, there comes a point where an increase in age no longer increases the likelihood of reducing poverty or increasing access to credit, resulting in a negative relationship.
Education: This variable capture years of schooling to represent levels of education for the household head. For the purposes of this study, education levels have been reclassified into four categories: no education, primary education, secondary school education, and tertiary education. The probability of obtaining credit is predicted to rise with educational attainment. Additionally, it is anticipated that, all other things being equal, higher levels of education will improve a household head's human capital, which will have a positive impact on welfare. So, for all households who fall in the four categories of education, they will assume the value of 1 and 0 if otherwise.
Dependency ratio: The household dependency ratio captures the number of dependent individuals in a household, such as children or elderly persons, compared to the number of working-age adults who are responsible for supporting them. A negative relationship of the dependent variable to welfare of the household is expected.
Livestock_Ownership: This variable was added to the adopted model to capture if the household has livestock or not. This variable is expected to have a positive relationship to credit access in a household since it will boost their wealth for collateral.
Employment: This variable explains whether the household head is employed or not. Employment as per the definition of NSO, is any pierce of work that earn the household some earnings, e.g. ganyu. It is expected that this variable will have a positive impact on the wealth of the household to access credit easily
| [5] | IHS-5. (2019/2020). International Household Survey Report. Zomba: Malawi National Statistics Office. |
[5]
.
Non-agricultural enterprise: This variable captures a number of enterprises that a household runs apart from agriculture, these are in a form of profit-making side businesses which include shops, or any individual business per household. This variable is expected to have a positive relationship with credit access as it will boost the householder’s wealth for collateral.
Social cash transfer: This variable captures all householders who had the privilege of receiving any money from any social programme to sustain them in any way during the period that the data was collected. This variable is expected to have a negative relationship as those receiving them will not have the willingness to reach out to either formal or informal credit.
Urban Distance: This variable captures the distance from the communities to the nearest urban or town. It is expected that those areas who are closer to the urban areas will be privileged with so many benefits in both their knowledge about credits and welfare. This variable is expected to have a positive effect to credit access (the shorter the distance, the better positive impact it has on the dependent variable).
3.2. Multicollinearity
Multicollinearity test was conducted because high correlations among the independent variables can inflate the standard errors of the coefficient estimates, leading to unreliable and unstable results. By identifying and addressing multicollinearity, we enhance the precision and interpretability of our model, ensuring that the estimated effects of credit access on household poverty are not distorted by redundant information.
Multicollinearity arises when all or some of the explanatory variables in a regression model have a perfect or nearly perfect linear relationship, resulting in infinite standard errors and indeterminate regression coefficients. One has two options for handling this issue: either do nothing or adhere to some general guidelines
| [25] | Gujarati, D. (2004). Basic Econometrics, 4th Edition. McGraw-Hill. |
[25]
. If multicollinearity is discovered in our data, the study will not do anything as done by
| [12] | Kaluwa & Kunyenje. (2019). "The paradox of the financial inclusion-poverty nexus in Malawi.". African Review of Economics and Finance 11, no. 2, 38-66. |
[12]
. Multicollinearity is basically a data deficiency issue, and sometimes there is no way around the data that are available for empirical analysis, claimed
| [12] | Kaluwa & Kunyenje. (2019). "The paradox of the financial inclusion-poverty nexus in Malawi.". African Review of Economics and Finance 11, no. 2, 38-66. |
[12]
. Furthermore, a linear combination of the regression coefficients can be estimated reasonably effectively even if one or more of them cannot be estimated with higher accuracy.
Table 1. Multicollinearity Results.
Variable | VIF | 1/VIF |
Household Head Age SQ | 6.75 | 0.15 |
Household Head Age | 6.74 | 0.15 |
Household Size | 1.01 | 0.98 |
Residence | 1.01 | 0.99 |
Education | 1.00 | 0.99 |
Employment | 1.00 | 0.99 |
Credit Access | 1.00 | 0.99 |
Multicollinearity test shows that there is no any correlation between the variables. Adding age squared to age allows modelling more accurately the effect of different ages, rather than assuming the effect is linear for all ages. The effect of nonlinearity of age squared accounts for the effect of older individuals’ decision in all the models.
3.3. Correct Model Specification
A ‘link’ test was conducted to determine whether the model has been correctly specified. Any instruction for single-equation estimation can be followed by a link test. In this test, the squared independent variable is added, and the non-squared model is contrasted for relevance. In comparison to the un-squared form, a model without a link error will have a non-significant t-test result.
Table 2. Correct Model Specification Results.
Source | SS | df | MS | | Obs | = 4,732 |
Model | 1,040.5 | 2.0 | 520.2 | | F (2,4729) | = 704.31 |
Residual | 3,492.9 | 4,729.0 | 0.7 | | Prob>F | = 0.0000 |
Total | 4,533.4 | 4,731.0 | 1.0 | | R- Squared | = 0.2295 |
| | | | | Adj R-Squar.. | = 0.2292 |
| | | | | Root MSE | = 0.85944 |
ln Daily Consumption | Coefficient | Std Errors | t | P>t | | |
_hat | 3.8*** | 0.8 | 4.9 | 0.000 | | |
_hat Squared | -0.2*** | 0.05 | -3.58 | 0.000 | | |
_constant | -10.4*** | 2.9 | -3.58 | 0.000 | | |
The results show that all the models are properly specified as the results suggest that the model is statistically significant, with the predictors and the constant term having significant effects on the dependent variable. The R-squared value indicates that the model explains a moderate amount of the variance in the dependent variable, and the coefficients provide information about the direction and magnitude of the relationships between the predictors and the dependent variable.
3.4. Regression Results
Due to the nature of the model, the coefficients cannot be interpreted directly to represent probabilities
| [25] | Gujarati, D. (2004). Basic Econometrics, 4th Edition. McGraw-Hill. |
[25]
. In order to interpret them, they have to be at the margin, that is to say, the derivatives are used, and these can be directly interpreted.
Table 3 below presents the marginal effects results.
Table 3. Twostep Heckman Model Marginal Results.
Variable | dy/dx | z | P>z |
Total Land per Household | 0.01 | 0.450 | 0.656 |
Household Size | 0.08 | 7.840 | 0.000 |
Household Head Education | 0.27 | 4.660 | 0.000 |
Household Head Age | 0.01 | 0.650 | 0.516 |
Household Head Sex | 0.07 | 2.200 | 0.027 |
Household Head Age Squared | 0.00 | -1.190 | 0.234 |
Dependency Ratio | 0.00 | -0.370 | 0.710 |
Community Financial Institution | 0.98 | 21.650 | 0.000 |
Residence (urban/rural) | -0.25 | -4.190 | 0.000 |
Household Head Employ (Yes/No) | 0.51 | 9.560 | 0.000 |
Non-Agriculture Enterprises | 0.10 | 1.280 | 0.199 |
Distance to nearest Urban Centre | 0.02 | 21.270 | 0.000 |
Cash transfer received | -0.14 | -2.220 | 0.026 |
Livestock ownership | -0.02 | -0.390 | 0.695 |
Total Livestock Owned | 0.00 | -0.430 | 0.665 |
Farm inputs subsidies received | 0.03 | 0.560 | 0.576 |
mills | | | |
lambda | 0.41 | 9.070 | 0.000 |
rho | 0.45 | | |
sigma | 0.91 | | |
Wald chi2 (7) | 181.76 | | |
Prob >chi2 | 0.00 | | |
Joint Significance of the Heckman Selection Equation:
In terms of model fit, the sigma (σ) value is 0.9092036, which represents the standard deviation of the error term in the outcome equation and the Wald chi-square test statistic result of 181.8 with 7 degrees of freedom, resulting in a significant p-value of 0.000 suggests that the Heckman Twostep Selection model as a whole is statistically significant.
The rho (ρ) value of 0.4502300 represents the estimated correlation between the error terms in the credit selection equation and the outcome equation of poverty. It means that there is 0.4502300 degree of association between the unobserved factors affecting the selection process and those influencing the outcome variable.
The positive rho value suggests a positive correlation between the error terms in the credit selection model, indicating that unobserved factors that affect the credit selection process also tend to influence the poverty outcome variable in the same direction.
The lambda estimated value of 0.4093481, with a z-score of 9.070 and a significant p-value of 0.000 indicates that the lambda parameter is statistically different from zero, providing evidence of selection bias. The lambda parameter represents the correlation between the selection equation and the outcome equation. A significant lambda suggests that the selection variable included in the model plays a significant role in determining the outcome variable, indicating the presence of selection bias.
3.5. Credit Access Regression Results
Table 3 above presents the results of the both the credit access selection model and the second model that is measuring outcome of welfare. The following are the detailed explanation of the selection variables’ results:
Community Financial Institution
This variable came out to be very significant with a p value of 0.000 and a positive relationship with credit access, which means that, holding other factors constant, introducing a financial institution in a community will bring almost 97% chance of smallholder farmers to access credit. This is what is practically expected, and also in line with many empirical findings that have been conducted before
| [18] | Ampah, S. N. (2017). "Effect of access to credit and financial services on poverty reduction in central region of Ghana.". International Journal of Business and Social Science 8, no. 8. |
[18]
.
Residence
The second variable is residence or location, which was found to be significant at 1% (p-value 0.000). The study shows that there is a negative relationship between the residence of the household and credit access. Holding other regressors constant, credit access status for a household located in the rural area was lower by 24.8% compared to urban based households. This is also in line with theoretical and logical expectation that most people living in rural areas are always disadvantaged when it comes to credit access. These findings are also similar with what Chilimba found in his study in Malawi
| [13] | Chilimba, A. (2020). The Impact of Microfinance Programme Participation on Household Food Security in Malawi. International Journal of Economics and Finance Studies. SA. |
[13]
.
Household Head Employment
A third variable in credit access selection model is the employment status of the household head. This variable is significant, with a positive relationship with credit access, which means that holding other variables constant, there is 51% possibility that the household head who is employed will access credit than those who are not employed. Another paper also found the same results in 1998 when he was studying the Income and employment effects of microcredit programmes
| [26] | Khan & Khandker. (1998). Income and employment effects of microcredit programmes: Village‐level evidence from Bangladesh. The Journal of Development Studies, 35(2), pp. 96-124. |
[26]
. At the beginning of our study, this result was also what we expected to find, as people with employment tend to have more knowledge about credits, and for the sake of smallholder farmers, employment gives them collateral for creditworthiness as well.
Distance to nearest Urban Centre
Another variable is the distance of the household communities to any nearest big urban centre of Blantyre, Lilongwe, Mzuzu or Zomba. The reasoning behind this variable is that people who live near these areas, even though they might be living in rural areas, they tend to enjoy and benefit from the civilisation effects of these areas, so their knowledge and exposure gives them an advantage in being creditworthy than those staying very far from these urban areas. This variable is significant with a p-value of 0.000, the variable’s relationship with credit access is positive, this is contrary to our expectation of negative relationship with credit access in the study.
Further, the study found out that what the analysis above find is actually reasonable according to
| [27] | Hodson, R., Dwyer, RE. and Neilson, L. A., 2014. Credit Card Blues: The Middle Class and the Hidden Costs of Easy Credit. The Sociological Quarterly, 55(2), pp. 315-340. |
[27]
. In his analysis, Hodson found out that these households that live near urban areas do not mostly own the land, houses or properties they have because they are middle class, but rather they borrow them, which makes it difficult to use these as collateral in order to access such formal loans. The study continues to say that these middle-class people living near urban areas fail to bear hidden costs of unsecured loans hence the longer the distance to the nearest urban areas, the easy the smallholder farmers find it to access credit due to ownership of their property which can be used as collateral and also for them not to have problems with hidden costs of unsecured loans and credit.
Cash transfer received
Social cash transfer that the households in our sample received from donors and governments played another bigger role in influencing these farmers to access credit or not. The variable is significant with a p-value of 0.02 representing 2% and their relationship is negative. This means that almost 14% of people who were privileged enough to be receiving these social cash found no reason to seek credit, hence, they did not participate in both formal or informal credits. These results are in line with the study expectations.
All variables which are insignificant in this study were omitted for interpretation.
3.6. Poverty Regression Results
Table 3 above also presented the results of the second part of the regression model. In this regression, impact of credit access on poverty was estimated by calculating the Inverse Mills Ratio (IMR) and
Table 3 above shows the results in the form of lambda under mills results.
Inverse Mills Ratio (IMR) – Credit Access
This variable accounts for participation in the credit programme and corrects credit access for selection bias. Holding other regressors constant, IMR shows that there is a significant P-value of 0.000 on smallholder farmers welfare. As interpreted by Chilimba, a positive IMR shows that those who participated in the program (credit) experienced higher levels of consumption than those who did not take credit
| [13] | Chilimba, A. (2020). The Impact of Microfinance Programme Participation on Household Food Security in Malawi. International Journal of Economics and Finance Studies. SA. |
[13]
. This implies that the mean total real daily consumption per household is likely to be higher for households that participated in the credit programmes than those that did not participate in them. This is in line with the findings of
| [13] | Chilimba, A. (2020). The Impact of Microfinance Programme Participation on Household Food Security in Malawi. International Journal of Economics and Finance Studies. SA. |
[13]
. who conducted a similar study in Malawi and found that households that participated in microfinance programmes high levels of consumption (high food security), as measured by real daily consumption than those who did not participate, which is mirroring our current findings.
Household Size
The second significant variable in the outcome welfare model is the size of the household with a p-value of 0.000. The relationship with daily household consumption is positive, meaning to say that holding all other factors constant, any increase in a member of the household will have a positive increase of 7.5% in their daily consumption which is this study’s measure of welfare. This is according to theory and also what this study was expecting to find out of this variable.
Household Head Education
This variable is significant in this study model of welfare of the household with a p-value of 0.000 which is approximately 1% and its relationship with daily consumption is positive. Holding all factors constant, if a household head is educated, the study found out that there is almost a 26% chance that their daily consumption growth. This is in line with the study expectations and other findings from other researchers
| [31] | Mussa. (2017). Contextual effects of education on poverty in Malawi. 2017. |
[31]
.
Household Head Sex
Last significant variable on the welfare function is the sex of the household head. This variable has a significance of 2% with a positive relationship to daily consumption which is the me measure of poverty in this study. This means that holding other factors constant, household headed by men have a 7% chance of increasing their daily consumption than households headed by women. This is usually because of gender issues in such communities that hinder women to participate in many innovations and activities to improve the welfare of their households.
Share This Article
Twitter
Linked In
Facebook
Copy
Download