Abstract: In this study, we examine the diversity of risks that affect farming in the Northern Guinea Savanna of Nigeria. We also investigate the perspectives of these risks in relation to their economic implications on the farming enterprises. We also show that through reorganization of these risks, some derived factors have the ability to present themselves whether as corresponding to existing categorization of the variables or not and also to enable us know which of the factors is more important than the other. Gross margin and factor analytical methods were used in computing the estimated results on a cross sectional sample of 348 farming households. Results show that farmers who were grouped under natural risk incurred the least mean production cost of 11, 115.61 while the highest mean production cost of 15,998.18 was incurred by farmers grouped under production risks. The highest mean revenue of 18, 998.16 was recorded by farmers under production risk which translated into a mean gross margin of 65, 999.85. Verifying whether some derived factors would correspond to the existing categorization of 14 risk types (from five sources) which the farmers faced, results from the factor analysis and the consequent F-test from ANOVA show no marked or significant differences among the identified factors and the existing risk sources. Consequently, the individual effect or importance of the original 14 risk types that the sampled farmers considered important can be dully represented and effectively regrouped into five sources (factors) as natural, technical, social, ecosocial and biochemical.

INTRODUCTION

Risk is uncertainty that affects an individual’s welfare and is often associated with adversity and loss (Bodie and Merton, 1998). Risk is also uncertainty that matters and may involve the probability of losing money, possible harm to human health, repercussions that affect resources (irrigation, credit) and other types of events that affect a person’s welfare.

Farming is a financially risky occupation. On a daily basis, farmers are confronted with an ever-changing landscape of possible price, yield and other outcomes that affect their financial returns and overall welfare (Harwood et al., 1999). The consequences of decisions or events are often not known with certainty until long after those decisions or events occur so, outcomes may be better or worse than expected. Oftentimes, surveys have asked farmers about the most important types of risk that they confront in their farming operations (Harwood et al., 1999). These types of questions are typically part of a larger survey that inquires about producers’ risk management strategies and offers respondents a list of concerns that they can score in terms of importance. Scores generally are not ranked relative to one another, meaning that producers independently analyzed each concern on the list. Sources of risk in farming include among others production or yield risk, price or market risk, institutional risk, human or personal risks and financial risks.

Production or yield risk occurs because agriculture is affected by many uncontrollable events that are often related to weather including excessive or insufficient rainfall, extreme temperatures, hail, insects and diseases. Technology plays a key role in production risk in farming. Price or market risk reflects risks associated with changes in the price of output that may occur after the commitment to production has begun. In agriculture, production generally is a lengthy process. Livestock production for example, typically requires ongoing investments in feed and equipment that may not produce returns for several months or years. Institutional risk results from changes in policies and regulations that affect agriculture. This type of risk is generally manifested as unanticipated production constraints or price changes for input or for output. Farmers are also subject to human and personal risks that are common to all business operators. Disruptive changes may result from such events as death, divorce, injury or the poor health of a principal in the farm firm.

In addition, the changing objectives of individuals involved in the farming enterprise may have significant effects on the longrun performance of the operation. Financial risk results from the way the firm’s capital is obtained and financed. A farmer may be subject to fluctuations in interest rates on borrowed capital or face cash flow difficulties if there are insufficient funds to repay creditors.

The above risks constitute major agricultural constraints which farmers always have to contend with. In the Northern Guinea Savanna of Nigeria, risks associated with farming can be categorized into the following: Natural risks (drought, flood, wind and storm, disease and pest); social risk (theft of produce, bush fire, invasion of farms by cows); economic risk (producer price fluctuation, insufficient and untimely supply of fertilizer, insufficient credit and insufficient supply of seeds) and technical risk (poor soil, insufficient access to chemicals, scarce labour and insufficient processing facilities).

Studying the diversity of the risks facing farmers in the Northern Guinea Savanna of Nigeria and their economic implications particularly on crop farmers is an important milestone in the numerous efforts to provide solutions to the various agricultural constraints and challenges which resource poor farmers in sub-Saharan Africa have been battling with. The objective of this study is therefore to among other things.

MATERIALS AND METHODS

Data and study area: The data used for this study are from a cross-sectional sample of smallholder crop farmers in the Northern Guinea Savanna agro ecological belt of Nigeria. The data were obtained in respect of the 2004/05 agricultural year in a survey. The sample comprised 348 households selected from the 23 Local Government Areas (LGAs) and from the four Agricultural Development Programme (ADP) zones of Kaduna state. Structured questionnaire and checklists were used to obtain the data through personal interviews and focus group discussions. Detailed information on various aspects of the farm-household under various farming risk conditions and for different crop activities was collected. This included the household’s demographic characteristics, farm size, cropping patterns, crop output in actual and value terms, labour and non-labour inputs in actual and value terms. The risk sources and types which affected the majority of the sampled farmers were identified as:

Technical risks are those that are related to production. It is worth mentioning that identification of the risk sources and types and their nomenclature is necessarily the outcome of a focus group discussion between the researchers and the village extension practitioners who live and are conversant with the farming activities and needs of the sampled households.

Methods of data analysis: The results of the focus group discussion were employed to classify the sampled farmers based mainly on the characteristic features of the risks, they faced during the agricultural year of survey. This was done for each of the four ADP zones and made to coincide with the major cropping patterns in the study area. The cropping pattern is typically a multiple crop system which consists of maize plus one or two of other cereals or grain legumes in the Northern, drier axis and maize plus one or two grain legumes and root crops in the Southern humid axis (Table 1 and 2).

Descriptive statistical tools which include scattered diagrams and histograms are used to compare the means and standard deviations of the farmers’ characteristics which included the age, years of formal schooling, cropped area, household size and farming experience. These tools were also used to depict and compare the gross margin components of the farmers’ cropping activities, e.g., cost, revenues and gross margins.

Table 1:	Lay-out of sample and survey design by risk type and by ADP zone
Survey data: n = sample size: nn = natural risk; ns = social risk; ne = economic risk; np = production risk; nt = technical risk and nz = zonal total

Table 2:	Cropping patterns of respondents by risk type and by ADP zone
Survey data: mz = maize; gc = guinea corn; cp = cowpea; rc = rice; mlt = millet; gn = groundnut; sb = soybean; cy = cocoyam; ym = yam; pt = potatoes and cs = cassava

Analysis of Variance (ANOVA) models were used to compare the means of the cost, revenues and gross margins for the four ADP zones and under the five identified risk situations. In order to achieve this, the relationship between the cost, revenues and gross margins was subjected to ANOVA based on the following null hypotheses that the risk effects (treatment) are all equal to zero and that the zone (block) effects are all equal to zero, namely that:

The alternative to the 1st null hypothesis is that the risk (treatment) effects are not all equal to zero, namely that α_i≠ 0 for at least one value of i; correspondingly, the alternative to the second hypothesis is that the block (zone) effects are not all equal to zero, namely that β_j ≠ 0 for at least one value of j. The analyses on which the tests are based require the following generalization of the following theorem:

Where:

	=	The mean of the observations for the ith risk (treatment)
	=	The mean of the observations for the jth zone (block)
	=	The mean of all the nk observations

At a given probability level (usually 0.05), ANOVA was selected for this purpose because it enables us to conclude on the differences among the risks and zonal mean values. A factor analysis was finally carried out on a list of fourteen responses (risk types from five sources) which were included in the questionnaire prepared for the sampled farmers. Factor analysis includes both component analysis and common factor analysis. Factor Analysis (FA) and Principal Component Analysis (PCA) are statistical techniques applied to a single set of variables when a researcher is interested in discovering which variables in the set form coherent subsets that are relatively independent of one another.

Variables that are correlated with one another but largely independent of other subsets of variables are combined into factors. PCA produces components while FA produces factors but it less confusing in this study to call the results of both analyses factors. Factors are thought to reflect underlying processes that have created the correlations among variables. The techniques in factor analysis are frequently utilized in multivariate data analysis and are described in details by Tabachnick and Fidell (2000) and Sheskin (2007). Within the framework of a factor analytic procedure, a factor score can be computed for each subject on each of the factors (Sheskin, 2007). A subject’s factor is a composite score based on the relative contribution of all the variables which represent that factor. Computation of the factor scores allows a researcher to determine the relative standing of each subject with respect to each of the factors derived in the analysis. On a mathematical level, each factor can be defined by a linear equation as showed in Eq. 1 (Sheskin, 2007):

(1)

Where:

Through the use of Eq. 1, a factor score can be computed for any subject (Diekhoff, 1992; Tabachnick and Fidell, 1996; Stevens, 2002; Field, 2005). Respondents were asked to indicate which of the risk types affect their farming activities. Because of scale of measurement, the responses were converted to percentages, using the local government areas as bases. The percentage scores were thereafter used as proxies for the variables in the factor analysis. The importance of factor analysis in this study is seen in its ability to present the derived factors whether as corresponding to existing categorization of the variables or not and also to enable the researchers know which of the factors is more important than the other. In this study, results from the factor analysis are employed to complement the findings on the attributes of the risk sources and types which the sampled farmers are facing.

RESULTS AND DISCUSSION Characteristics (means and standard deviation) of sampled farmers: Table 3 shows the overview of the characteristics of the sampled farmers. The data revealed that for farmers in Birmin Gwari zone, the mean age stand at about 46 years with a standard deviation of 9.18. Considering all the four zones however, farmers in Lere zone appear to be most advanced in age with mean age of 46.68. Though, the risk bearing ability of the sampled farmers was not and cannot be tied to the farmers’ age, it is reasonable to infer that the incidence of the various farming risk faced by the farmers in the study area is a burden to mainly the middle-aged farmers. In Birmin Gwari zone, an average farmer did not have formal education. This applies to farmers in all the zones. Total farm area is highest in Birnin Gwari (6.06 ha) and lowest in Maigana (4.08 ha). The average number of persons in the farming household stands at about 9 in Lere zone.

This based on the data is the highest while the lowest average number of persons per household was estimated for households in Samaru zone. Years of farming experience are about 15, 12, 15 and 15 for Birnin Gwari, Lere, Maigana and Samaru, respectively. The data suggest that an average sampled farmer did not belong to any association or group; neither did they have access to extension and credit facilities. Average distance from the farm to farmers’ homestead is estimated to be approximately between 2 and 3 kms. Most of the farmers were affected by an average of two risk sources.

Gross margin analysis: This study uses data from farmers’ plots planted with maize and with guinea corn and rice, millet and soybean, cowpea and groundnut, guinea corn and millet, millet and groundnut, guinea corn and soybean, guinea corn and cowpea, rice and soybean, guinea corn and groundnut, groundnut and yam, guinea corn and cassava, cowpea and cocoyam, guinea corn and cassava, cowpea and cocoyam, rice and potato, groundnut and cassava. All the plots of the sampled farmers were planted with maize as the major crop in the agricultural year.

In the Northern axis of the study area, farmers make use of other portions of their plots to plant other cereal crops (guinea corn and millet) and grain legumes (groundnut, cowpea and soybean). In the Southern axis, the sampled farmers, besides maize and few pockets of other cereal crops, root crops like yam, cassava, cocoyam and potato are commonly planted on other portions of farmers’ plots. The result of the components of gross margin analysis show that farmers who were grouped under natural risk incurred the least mean production cost of 11,115.61 while the highest mean production cost of 15,998.18 was incurred by farmers grouped under production risks. The highest mean revenue of 18,998.16 was recorded by farmers under production risk which translated into a mean gross margin of 65,999.85. In spite of the highest revenue recorded by farmers under production risk and the gross margin thereof, farmers under natural risk situation recorded the overall highest mean gross margin.

Table 3:	Descriptive statistics (characteristics) of the sampled farmers

Fig. 1:	Average variable cost incurred (by ADP zone by risk type)

Fig. 2:	Average revenue (by ADP zone by risk type)

On zonal basis, the sampled farmers in Samaru zone incurred the least mean production cost. However, farmers in Lere zone recorded the overall highest mean values of the production cost, revenue and the resultant gross margin. For ease of compassion and because most sampled farmers worked on either traditional or inherited land, the opportunity cost of land was computed for the study area and the benchmark value used as a variable cost which was included in the production costs incurred by each sampled farmers during the agricultural year. Figure 1-3 show the patterns of the average cost, revenue and gross margin. The ANOVA result shows that the mean differences in the cost along the risk types and across the ADP zones were not statistically significant.

Fig. 3:	Average gross margin incurred (by ADP zone by risk type)

This indicates that during the agricultural year, the cost incurred by the sampled farmers for crop production was not related to the differences in the risks faced by the farmers neither was it related to the spatial differences in the farmers’ plots of cultivated lands. The mean differences in the revenue generated by farmers across the zones show a statistical significance (p<0.05) indicating zonal variation in the revenues generated by the sampled farmers.

A further check (LSD tests) on the mean differences on the zonal revenues shows that the existence of the differences is between Birmin Gwari and Samaru zones; Lere and Maigana zones and between Lere and Samaru zones. There was no difference in the mean revenue between each of the zones and the grand (benchmark) mean for the study area. The difference in the revenue between Birmin Gwari and Samaru could be as a result of the sharp differences in the types of crops grown in the two areas besides maize.

For example in Birmin Gwari, cereals and grain legumes are the crops grown while in Samaru, root crops dominate the cropping activities. The same reason could be adduced for the differences in mean revenue recorded by the farmers in Lere and Samaru. The extent of maize growing in Lere accounts for the mean revenue differences between it and Maigana zone. In the study area, bulk of the maize produced come from Lere zone. The ANOVA results on the gross margin show a statistical significance (p<0.05) in the mean differences among the sampled farmers along the risk situations. For example, there were mean farmers’ gross margin differences between the pairs of natural and economic risks; natural and technical risks and production and technical risks situation. No statistical difference was observed, evident in the mean gross margins between any of the risk situation and the grand (benchmark) mean for the overall risk situation.

Table 4:	Results of the factor analysis of risk sources: factor loadings

The difference in the gross margins among the identified pairs of risk situation could be as a result of the effect of the individual types or elements within each risk source or situation. These have enormous consequences and implication on the production requirements (resources and inputs), the output and ultimately the marketing components of the crops which also affect the revenue generated from the crop and the gross margin thereof.

Results of factor analysis: Table 3 shows the factor loadings obtained after a varimax rotation of responses from the 14 questions that measured the sources from 5 risk types, e.g., natural, social, economic, production and technical risks. As can be seen (Table 3), five factors (from the factor loadings) were identified as the actual risk sources that are of utmost importance to the sampled farmers. The following are the factors and the constituents of each of them; factor one retained risk types which are purely natural; these are drought, flood and poor soil. Factor two is formed mainly from the two components of technical/production risks, e.g., scarce labor, insufficient credit facilities and inadequate supply of chemicals. It also has wind/storm (from natural risk) and producer price fluctuation (from economic risk).

Other factors and their constituents are: factor three (theft of produce and invasion of farms by cows-which are mainly social risk types; factor four (bush fire and insufficient supply of maize seeds); factor five (disease/pest and inadequate spraying equipments). The results from ANOVA analysis for the factor loadings of each of the original risk sources and for those of the five identified clusters are showed in Table 4.

Table 5:	Results of ANOVA from the risk sources and factors

The F-test from the ANOVA analysis indicate that there are no significant inter-source differences for every risk source. The F-test also show no significant inter-factor differences for every identified factor. In the following discussion, the only variables showed in Table 3 are the very important ones based on their factor loadings. As a Rule of Thumb (Tabachnick and Fidell, 2000), only variables with loadings of ≥0.32 are interpreted. The greater the loading, the more the variable is a pure measure of the factor. Suggestions based on Comrey and Lee (1992) are that loadings in excess of 0.71 950% (overlapping variance) are considered excellent, 0.63 (40% overlapping variance) very good, 0.55 (30% overlapping variance) good, 0.45(20% overlapping variance) fair and 0.32 (10% overlapping variance) poor.

On the strength of the foregoing, the individual effect or importance of the original 14 risk types that the sampled farmers considered important can be dully represented as shown on Table 5. These risk types are effectively regrouped into five sources (factors) as natural, technical, social, ecosocial and biochemical (Table 6). Assigning a nomenclature to each of the otherwise identified risk sources is evidenced from the elements of the identified new risk sources (factors) which have clearly been re-allocated.

However since, the ANOVA results show no marked or significant inter-source and inter-factor differences, the overall impact of the risk elements whether in their original sources or in the identified factors on maize production cannot be attributed to a particular risk source or factor. Instead, the enormity of the effects of the risks faced by the farmers is only described by the loadings in the five identified factors.

This study examined the diversity of risks that affect farming in the Northern Guinea Savanna of Nigeria. It also investigated the perspectives of these risks in relation to their economic implications on the farming enterprises. The study also showed that through reorganization of these risks, some derived factors have the ability to present themselves whether as corresponding to existing categorization of the variables or not and also to enable us know which of the factors is more important than the other.

Table 6:	Risk types reassigned to the identified new (derived) risks (factors)

Results indicated that farmers who were grouped under natural risk incurred the least mean production cost of 11,115.61 while the highest mean production cost of 15,998.18 was incurred by farmers grouped under production risks.

The highest mean revenue of 18,998.16 was recorded by farmers under production risk which translated into a mean gross margin of 65,999.85. In spite of the highest revenue recorded by farmers, under production risk and the gross margin there of farmers under natural risk situation recorded the overall highest mean gross margin. The ANOVA result indicates that during the agricultural year, the cost incurred by the sampled farmers for crop production was not related to the differences in the risks faced by the farmers neither was it related to the spatial differences in the farmers’ plots of cultivated lands.

The mean differences in the revenue generated by farmers across the zones show a statistical significance indicating zonal variation in the revenues generated by the sampled farmers. Furthermore, ANOVA results showed a statistical significance in the mean differences of farmers’ gross margin between the pairs of natural and economic risks; natural and technical risks and production and technical risks situation.

As demonstrated by the factor analysis, five factors (from the factor loadings) were identified as the actual risk sources that are of utmost importance to the sampled farmers. Consequently, the individual effect or importance of the original fourteen risk types that the sampled farmers considered important can be dully represented and effectively regrouped into five sources (factors) as natural, technical, social, ecosocial and biochemical.

CONCLUSION

From the findings, the following observations can be drawn: the difference in the gross margins among the identified pairs of risks could be as a result of the effect of the individual types or elements within each risk source or situation. These have enormous consequences and implication on the production requirements (resources and inputs), the output and ultimately the marketing components of the crops which also affect the revenue generated from the crop and the gross margin there of the overall impact of the risk elements whether in their original sources or in the identified factors on crop production cannot be attributed to a particular risk source or factor. Instead, the enormity of the effects of the risks faced by the farmers is only described by the loadings in the five identified factors. Generally, the aggregate effect of both the initially identified risks and the derived risks is one that influences the profitability of the resource poor farmers. This has impact on the measures of farm business performance such as the net cash flow generated by the farming activities or the net farm income earned.