The Effects of Agricultural Financing on Agricultural Productivity in Nigeria


 The study “The Effects of Agricultural Finance on Agricultural Productivity in Nigeria” investigated the effect of agricultural financing, both public and private on the outputs of two main sectors of agriculture: crop production and livestock production. The objectives of the study are to examine the long and short run relationship of agricultural financing on crop production and livestock production, and to examine the causal relationship between agricultural finance and agricultural productivity. To achieve these objectives, the study employed two models, each using ARDL Test, Bounds Test, and Granger causality test using time series data from 1981 to 2019.Data were obtained from CBN and World Bank data bases. Dependent variables were Crop Production and Livestock Production respectively and independent variables were Public Finance, Commercial Bank Credit to Agriculture, Inflation Rate and Interest Rate. The model was tested using descriptive statistics to analyse the significance of the relationship between the dependent and independent variables. The results show that both public and private finance were positive but insignificant in the short run. In the long run, public finance remained insignificant whereas private finance was positive and significant. Thus, private financing is more effective at improving agricultural productivity than public finance. The study also revealed a negative long run relationship between interest rate and the outputs of crop and livestock production during the period. It is therefore recommended that the government encourages private investment in agricultural activity, and puts measures in place to curb corruption and embezzlement. Government should also ensure that credit facilities are provided to farmers at low interest rate to reduce it detrimental influences.


Introduction
Agricultural activity constitutes the mainstay of the rural population which account for two-thirds of Nigeria's population. It is the single most important activity and employs over seventy percent of the working population while contributing a range of thirty to forty percent of the Gross Domestic Product.
The agricultural sector is the second largest sector after oil. It was once the primary sector but investment in the sector as more people rushed to the lucrative oil sector dropped signi cantly.
Agricultural productivity in Nigeria has largely been accounted for by resilient growth performances in four constituent sub-sectors: crops, livestock, sheries and forestry (Eboh et al., 2012). The agricultural sector have in recent times signi cantly contributed to sustained growth performance in Nigeria, its major contribution seems to be much short of overall potential.
Agricultural subsectors in Nigeria have capacities that give the sector an enablement for growth.
According to CBN (2012), between 1960 and 2011, an average of 83.5% of agriculture GDP was contributed by the crops production subsector making it the key source of agriculture sector growth.
Agricultural output has uctuated widely and productivity has also declined. In terms of contribution to GDP, available statistics from the CBN shows that the agricultural sector's share of GDP increased from 28% in 1985 to 32% in 1988, dropped to 31% in 1989, rose to 37% in 1990 but fell signi cantly to 24% in 1992, it increased again to 37% in1994. It was 32% in 1996 and rose to 40% in 1998, dropped again to In today's world, the agriculture sector acts as a catalyst, accelerating the pace of restructuring and diversifying the economy that depends less on the supply of foreign agricultural products or raw materials to economic growth and sustainable development. The agricultural sector contributes mainly to the nation's development in the aspect of enhancing government revenue, expenditure and aspect of enhancing government revenue; infrastructural, living standards and also contribution to Gross Domestic Product.
The sheer size of agriculture suggests that strategies designed to promote the early stages of economic growth cannot ignore agriculture. The improvement of the economy of the rural areas in a sustainable way has the ability of increasing employment opportunities, reducing stemming pre-mature rural-urban migration, regional income disparities and reducing poverty at its root. (Anriquez and Stamoulis, 2007).

Agriculture
Agriculture is a way of life that involves the production of food, livestock rearing through shing, animal husbandry, systematic growing and harvesting of crops for production of commodities such as feed and bre among others. Agricultural activity is integral to the survival of the human race. It is the key to the development and the rise of primitive human civilisations and nurtured the growth and subsequent prosperity of these civilisations. Agriculture can be dated back thousands of years and marks man's transition from rudimentary hunter and gatherer communities to the societies that will birth the greatest civilisations.

Public Finance
Government nancing are strong instruments introduced to remove market failure (tendency for an economy to produce too much of some goods and an insu cient number of others) and can launch an economy into achieving economic growth. Government nance or expenditures are the expenses which a government incurs for (i) its own maintenance (ii) society and the economy (iii) helping other countries (Bhatia 2002).
Government nance represents the total government spending to attain the predetermined macroeconomic objectives. In the Nigerian economy public nance can broadly be categorized into capital and recurrent nancing. The recurrent nancing are government expenses on administration such as wages, salaries, interest on loans, maintenance etc., while expenses on capital projects like airports, roads, education, health, telecommunication, electricity generation etc., are referred to as capital nancing (Obinna, 2003).

Agricultural Financing
Agricultural nancing is often seen as any of several credit vehicles used to nance agricultural transactions, including loans, notes, bills of exchange and bankers acceptances. Credit promotes productivity and enhance standard of living by breaking vicious cycle of poverty of small-scale rural farmers. Credit to agriculture has also been viewed as the major constraint on the expansion of both large and small-scale farming. Agricultural nancing refers to all efforts to provide easier access to such credit facilities.
Agricultural nancing is crucial to agricultural and rural development in Nigeria. Approximately 70% of the population is concentrated in rural areas with agriculture not just as their source of income but also a lifestyle. Keynesian (Modern Concept) view: this opines that Government must play a positive role in order to regulate the economy by government spending and taxes in the most desirable manner. This school of thought discredits the belief of classical that supply creates its own demand and the automaticity of the economic system to generate full employment and growth by itself without interference. Keynes believes that the propensity to consume reduces as income increase and the propensity to save increase as income increase. This will bring about disequilibrium in the economy as consumptions (aggregate demands) do not grow proportionally with savings when income is rising. Thus, to maintain income, employment and growth it is necessary to off-set the effects of reducing demand for outputs by a corresponding increase in public expenditure. Hence, if undesirable economic conditions are to be avoided the gap between the income and expenditure must be lled either by increasing propensity to consume in the economy or by increasing government expenditure.

Review of Empirical Studies
Yusuf and Mohammed (2017) assessed the effect of public expenditure on crop production in Yobe State using primary data. The study adopted survey research design and collected cross-sectional data. The study used Chi-square to test the hypothesis and found out that public expenditure does not have effect on the output of crops and therefore recommended that a policy framework needs to be instituted in the diversi cation of public expenditure. However, the study was carried out in Yobe State and include only capital expenditure and there is need for similar study to be replicated in Kogi State using both capital and recurrent expenditure.
Uremadu, Ariwa and Uremadu (2018) examined the effect of government agricultural expenditure on agricultural output in Nigeria using time series data from Central Bank of Nigeria from 1981 to 2014. Vector Error Correction model estimation technique was used and the study revealed that total government agricultural expenditure (TGEX) had a positive and signi cant effect on agricultural output (AGO) in Nigeria during the period studied. The major weakness of the work was that the independent variable was not disintegrated; the study used total government expenditure as the independent variable. Therefore, the current study will disaggregate the expenditure into capital and recurrent.
Rufus and Oyewole (2018) empirically evaluated the nexus between public spending on agriculture and Nigerian output growth. The ndings show that agricultural sector output has positively impact on the economic growth in Nigeria over the period under study. The main weakness of the study was failure to test for long-run relationship between the variables using Johansen cointegration analysis.
Mgbanya, Onwumere, Eze, Nwokenekwu& Igwe (2018) assessed the impact of the national recurrent expenditure on Nigeria's agricultural growth from 1990 to 2017. The Wald test result showed that the Fstatistics (23.126) was greater than the F-tabulated (4.32) at p-value < 5%, this implies that the recurrent expenditure on agriculture has a signi cant impact on the agricultural share of GDP from 1990 to 2017.
However, this study overlooked capital expenditure on agricultural growth and also relied on only secondary data for data collection. Brown and Ajayi (2015) in their study examined the effects of government spending on the agricultural sector in Nigeria. The study revealed that total government expenditure had positive and signi cant impact on agricultural output in Nigeria during the period under study. Based on the above ndings, the study recommended for an increase funding of the agricultural sector in Nigeria. However, the study fails to disaggregate the components of public expenditure.

Theoretical Framework
The model is anchored by the Keynesian theory of public expenditure which makes up a key component of public nance. It is a macroeconomic theory of the total expenditure Keynes believed in the use of public expenditure as a tool of scal policy to stimulate the economy and encourage increased productivity as well as economic growth. Public expenditure can be used to nance the agricultural sector and improve agricultural productivity.

Model Speci cation
The model incorporates public nance given to the agricultural sector and commercial bank credit to agriculture as major independent variables. Agricultural output broken down into crop production and livestock production are respective dependent variables. In ation rate and interest rates are further included as control variables within the model. Following the above, the implicit forms of the models are: Where CP is crop production LP is livestock production PF is public nance to the agricultural sector.
CCA is commercial bank credit to the agricultural sector.
INFR is the annual in ation rate.

IR is the lending interest rate
The explicit forms of the models are speci ed as: a 0 is the intercept,a 1 , a 2 , a 3 , and a 4 are coe cients of their respective variables. While µis the error term.

A priori Expectation
From the models speci ed, the a priori expectation is written as α 1 > 0, α 2 > 0, α 3 > 0, α 4 < 0. Public nance and commercial bank credit to agriculture are expected to have a positive relationship with crop production and livestock production respectively. Interest rates and in ation rates are expected to have a negative relationship with the crop and livestock production outputs.

Estimation Techniques
Unit root test, Autoregressive distributed lag, Bound's test and Granger causality were employed to estimate the model and to achieve the objectives.

Augmented Dicky Fuller Test
Unit root test was conducted at level and at rst difference, using 5% signi cance level as the criteria. Considering the results of the table below, it shows that all variables exempting in ation are stationary at rst difference where as in ation is stationary at level.

4.3: ARDL Co-integrating and long Run Form Result
In light of the results displayed below, the R-squared has a value of 0.99. This shows that the independent variables explain changes in the dependent variable of crop production to the tune of 99%.
The co-e cient of the adjusted R-squared equals 0.99, which con rms the extent to which variation in the chosen independent variables explains variation in the chosen dependent variable.
The Durbin Watson statistic evaluates serial correlation within a model. The Durbin Watson has a value of 1.77, which indicates that there is positive serial correlation within the model.
The probability value of the F-statistic equals 0.0000 showing that the model is signi cant at the 5% level of signi cance.
ARDL result is to determine the variables that are responsible for short run shocks within the model. The errors are corrected by the magnitude of the R-squared that is after another regression have been carried out.
The dependent variable of this study is the output of crop production. The coe cient of public nance in the current year is 0.026146. The sign is positive, which indicates that public nance is positively related to the output of crop production. Public nance might have been expended in the provision of credit facilities to crop producers which had positive effect on the output. However, probability value of 0.4251 suggests that public nance of the current period is insigni cant.
In the rst lag, the coe cient of public nance is -0.006248. Public nance is found to be inversely related to the output of crop production, in the rst lag. Public nance of the previous year has negative effects on current levels of crop production. This can be attributed to the embezzlement of public funds, rather than investment in crop production. Although government budgets show increasing levels of funding, this money does not meet its intended recipients. Therefore, a unit increase in public nance will decrease crop production output. The probability value of 0.8380 also indicates that public nance is insigni cant to crop production output.
The coe cient of public nance in the second lag is -0.154746 also denoting a negative relationship between crop production and public nance in the second lag. This negative relationship is caused by the high level of corruption and fraud that occurs at high levels of government. Although, the probability value of 0.0000 suggests that public nance in the rst lag is signi cant at 5% level of signi cance.
The coe cient of commercial bank credit to agriculture which is denoted by CCA is 0.042627 is positive. Thus, there is a positive relationship between commercial bank credit to agriculture and the output level of crop production in the current period. A unit increase in commercial bank credit to agriculture will result in a 0.04262 increase in the output level of crop production. The probability value of 0.3849 indicates that this is insigni cant at the 5% level of signi cance.
The interest rate of the current period has a coe cient of -0.002114 which is negative. Therefore, interest rate has an inverse effect on crop production. This conforms with apriori expectations. The probability value of 0.7447 also indicates that interest rate is insigni cant.
The ARDL result overall is statistically signi cant and negative. The coe cient is -0.266230, which is negative and complies with apriori expectations. The probability value is 0.0009, which is statistically signi cant at the 5% level of signi cance.
The ARDL also speci es the nature of the long run relationship within the model. From the probability value of the long run coe cients, one can conclude if there is a signi cant relationship between the respective dependent variable and the independent variable in the long run.
The coe cient of in ation equals 0.449455 with a probability value of 0.0125. Thus, it can be concluded that in the long run, there is no co-integration between in ation and crop production output. In ation does not affect crop production output in the long term.
Public nance has a coe cient of 0.682227 as well as a probability value of 0.0000. Considering the probability value, we conclude that there is no signi cant co-integrated relationship between public nance and the output of crop production. Plainly put, government nancing geared towards the agricultural sector has no signi cant effect on the long term outputs of crop production.
The analysis also presents the long run coe cients of commercial bank credit to agriculture which has a value of 0.160114 and a probability value of 0.3301. There exists a positive long run relationship between commercial bank credit and crop production output. From the probability value, we can conclude that there is existence of co-integration between commercial bank credit to agriculture and crop production output. That is to say that commercial bank credit to agriculture affects crop production in the long run.
Interest rate has a coe cient of -0.007940 with a probability value of 0.7324. The negative coe cient indicates that there is an inverse relationship between interest rate and crop production in the long run. This denotes the existence of co-integration and asserts that there is a signi cant long run relationship between interest rates and crop production.

Granger Causality
According to the results of the analysis below, granger causality runs uni-directionally from in ation to crop production. It is evident that In ation granger causes Crop Production output whereas, crop production does not granger cause in ation. Hence, this relationship is one-sided in nature.
It can also be inferred that public nance does not granger causes crop production output and neither does crop production output granger cause public nance. Thus, there is no causal relationship existing between the two variables.
Similar to public nance, commercial bank credit to agriculture does not granger cause crop production output and neither does crop production output granger cause commercial bank credit to agriculture. There is also no causal relationship existing between the variables.
In a similar vein, interest rate does not granger caused crop production outputs and neither does crop production output granger cause interest rates. There is also no causal relationship existing between the variables.  According to the below results, the existence of a co-integrating relationship can be inferred. The Fstatistic has a value of 7.775595. This is greater than the upper bound value at 5% level of signi cance which is, 4.01. Therefore, it can be concluded from the results of the Bound's test, that the variables are indeed co-integrated which implies that there is existence of long-run equilibrium relationship existing between the variables in this model.

Autoregressive Distributed Lag Result
In light of the results displayed below, the R-squared has a value of 0.99. This shows that the independent variables explain changes in the dependent variable of crop production to the tune of 99%.
The co-e cient of the adjusted R-squared equals 0.99, which con rms the extent to which variation in the chosen independent variables explains variation in the chosen dependent variable.
The Durbin Watson statistic evaluates serial correlation within a model. The Durbin Watson has a value of 1.87, which indicates that there is positive serial correlation within the model.
The probability value of the F-statistic equals 0.0000 showing that the model is signi cant at the 5% level of signi cance.
ARDL result evaluates the short run relationships within the model. The errors are corrected by the magnitude of the R-squared that is after another regression have been carried out.
The dependent variable of this model is the output of livestock rearing. The coe cient of public nance in the current period is 0.021635, which is positive. This indicates that a unit increase in public nance will cause 0.021635 in livestock output, which is in accordance with apriori expectations. The probability value of 0.2307 however suggests that the results are insigni cant at 5% level of signi cance.
Public nance in the rst lag has a coe cient of -0.050524. This negative value suggests an inverse relationship between the previous public nance and current livestock production. A unit increase in public nance will cause a -0.050524 increase in livestock output in the rst lag. The probability value of 0.0062 indicates that this value is statistically signi cant at the 5% level of signi cance.
Commercial bank credit to agriculture is also a major independent variable within the model with a coe cient of 0.024618. This denotes a positive relationship between commercial bank credit to agriculture and livestock production. A unit increase in commercial bank credit to agriculture will yield a 0.024618 increase in livestock output. The probability value of 0.4934 indicates that this result is statistically insigni cant.
Interest rate in the current period has a coe cient of 0.001747, which denotes a positive relationship between interest rate and livestock production. Although the probability value of 0.6785 denotes that this result is statistically insigni cant. Whereas interest rate in the rst lag has a coe cient of -0.010611, which indicates that interest rate in the rst lag is negatively related to livestock production. This value is also signi cant at the 5% level of signi cance with a probability value of 0.0050.
Current in ation rate is also positively related to livestock production with a coe cient of 0.076532. Thus, a unit rise in in ation level will cause a 0.076532 increase in the output of livestock production. The probability value of 0.0003 denotes the signi cance at 5% level of signi cance.
The overall result of the ARDL is signi cant at the 5% level of signi cance with a probability value of 0.0301. The coe cient is negative which conforms to the apriori expectations.
The analysis also presents the nature of the long run relationships present within the model and test for the existence of co-integration.
The long run coe cient of public nance is a value of 0.572163 with a probability value of 0.0004. It can be inferred from the probability value that there is no co-integration between livestock rearing and public nance.
Whereas commercial bank credit to agriculture has a coe cient of 0.188583 with a probability value of 0.3763. The probability value suggests that there is co-integration between livestock rearing and commercial bank credit to agriculture. It can be inferred from the coe cient that the long run relationship is positive in nature. Therefore, a unit increase in commercial bank credit to agriculture will result in a 0.188583 rise in livestock production in the long run.
In ation rate has a coe cient of 0.586275 and a probability value of 0.0833. It can be concluded from the probability value that there is a long run relationship existing between in ation rate and livestock production.
Interest rate also has a coe cient of -0.005768 and a probability value of 0.8666. This denotes the existence of a co-integrated relationship between interest rate and livestock rearing. The coe cient indicates that the long run relationship is negative and that and increase in the interest rate will result in the decrease in livestock output in the long run.

Granger Causality
The results of the Granger Causality test below shows that Public nance does granger cause the output of livestock rearing, whereas livestock rearing does not granger cause public nance. This denotes that the relationship between these two variables is unidirectional in nature.
Another unidirectional relationship exists between in ation and livestock rearing. Where in ation granger cause livestock rearing, but the same cannot be said for livestock rearing granger causing in ation.
This also applies to interest rate, which shares a unidirectional relationship with livestock rearing. Where interest rate granger causes livestock rearing but livestock rearing does not granger cause interest rate?
However, commercial bank credit to Agriculture does not granger cause livestock rearing and neither does livestock rearing granger cause commercial bank credit to agriculture. This indicates that there is no causal relationship existing among the variables.

Conclusion
The study concentrated on the effects of agricultural nancing on agricultural productivity in Nigeria. From examining the ndings, it can be inferred that agricultural nancing has an overall positive effect on agricultural productivity.
In the short run, the effects of public and private nance were largely insigni cant due to the long-term nature of agricultural ventures. It often takes years to see the results of agricultural investment.
In the long run however, only private nancing proved to have any relationship with crop production and livestock rearing. Public nance did not have any signi cant long run relationships with either crop production or livestock rearing. It can be concluded that public nance has no signi cant effect on crop production or livestock rearing.
It can also be inferred that neither public nor private nance share a causal relationship with the outputs of crop production and livestock rearing.
In light of the ndings of this study, applications of suitable policies are necessary to ensure increase in agricultural productivity and all the bene ts it affords. Such policies will also be bene cial to the nation's economy and achievement of economic growth through agricultural development.

Policy Recommendations
As indicated by the above ndings, the following policies are recommended Firstly, the futility of public nance in improving agricultural productivity has been proven, due to the large amount of corruption and embezzlement. Discrepancies between the actual amount allocated to the sector and recorded funds exist. Government should put in extra measures to curb corruption and ensure that these funds reach their intended recipients. Establishment of necessary checks and balances as well as creation of devoted anti-corruption agencies are recommended, in order to curb fraud.
Secondly, the Nigerian government should focus more on encouragement of private investment through special directives to commercial banks among others. Governments should encourage nancial institutions to make certain percentage of their credit facilities available to the agricultural sector. This will increase private investment and therefore increase the long run outputs of crop production and livestock production.
Thirdly, interest rate has been proven to have an inverse long run relationship with the outputs of crop production and livestock production. High interest rates make it di cult for farmers to procure credit facilities and as a result, cause declines in productivity. Government should therefore regulate interest rates to ensure that credit facilities are provided to farmers at lower interest rates. This will encourage farmers to make use of credit facilities to improve agricultural yield.
Finally, in order to improve agricultural productivity, it is important to provide the necessary public goods that aid in conducting business and other economic activities. Infrastructural decay is a major issue faced by the agricultural sector, eliminating it will greatly help agriculture in Nigeria. Public funding should be geared towards providing adequate infrastructural facilities to create a more conducive business atmosphere. Facilities such as good roads, electricity, water etc should be concentrated in rural areas where the bulk of agricultural activity takes place. FUNDING: Authors' are responsible for the funding of all the research processes, while Afe Babalola University , Ado-Ekiti (ABUAD) will be responsible for the payment of the publication fee.