The Determinants of Access to Agricultural advice in the West Cameroon Region

1.1 Background and Objective of the Study In agriculture, advice is considered important for improving the performance of farms by development factors, because it facilitates access to information and training (Rebuffel et al., 2015). Although in the countries of the South some research shows that access to agricultural advice is limited (Faure et al., 2011). In Cameroon, the evolution of agricultural policies since independence aims to seek solutions to help producers in rural areas to achieve their development objectives. They are concerned with improving technical and economic performance, structuring agricultural activities in the production chain and promoting income growth and employment in rural areas.

Access to agricultural advice can be considered as the capacity for a producer to be supported by the services of a consultative body and to benefit from the support offered to him. Thus, access to agricultural advice is perceived as the very accessibility of the advice by the strategies deployed and the objectives pursued, and the fact of being able to take advantage of its services once the producer is supervised. Several social, economic, institutional, anthropological, technological factors can therefore influence access to agricultural advice, such as the producer, the service provider or even the agricultural environment.

Access to agricultural advice can then be influenced by the disparity between supply and demand, as shown by Agunga and Igodan (2007) for producers in the United States. Or even ethnic considerations of gender or social status, which define the place of the individual in society and their possibility of accessing agricultural advice, as emphasized by Hoang et al. (2006) in their study in Vietnam. It should be noted that the gender issue in terms of access to agricultural advice is particular in certain southern countries, such as Nigeria (Lahai et al., 1999). In addition, access to agricultural advice relates to the service provider who adjusts offer according to its objectives, means, and scope (Rebuffel et al., 2015). Finally, public policies can play an important role in the supply of services by directly participating in the definition of the content of agricultural advice, in particular for taking into account the environmental or social dimensions of agriculture, and access methods (Rivera and Alex, 2004).

The objective of this article is to characterize the determinants of access to agricultural advice in the West Cameroon region within the Program for the Improvement of Competitiveness of Family Agro-pastoral Farms (ACEFA). It is therefore a question of better understanding the different aspects which promote access to advice within the farming community and which contribute to improving their living and working conditions. Results which can be enlightening for the strengthening of the agricultural advisory systems deployed in the said region and the other regions and countries of the South. 1.2 Approach and method The study was conducted with 360 producers in rural areas in the Menoua, Mifi, Bamboutos and Koung-Khi divisions of the West Cameroon region. To address the issue of access to advice, we have taken into account the approach to advice for family farms deployed within the ACEFA system, which was chosen because of its presence in all the divisions of the region. In addition, we have made the assumption that the access of farmers to agricultural advice is related to the demand of producers and the methods of providing advice. As the demand for advice varies according to the type of operation, the social characteristics of the producers and the nature of the production, the advisory service depends on the governance policies deployed within it.

The analysis of quantitative data was done using descriptive statistics and the logit model; the latter is used to determine the factors that affect access to agricultural advice among the population surveyed. If we set P the probability of accessing agricultural advice or not, and X the independent variable likely to influence the occurrence of such a situation, the mathematical formulation of the logit model is expressed in equation: 𝑃 Y = logit(P) = ln( ) = α+βx (1) 1−𝑃 by deriving P from the equation (1) we obtain: 1 P = (2) 1+[𝑒𝑥𝑝−(𝛼+𝛽𝑥) with P: probability of access to agricultural advice, 1-P: probability of non-access to agricultural advice, X: independent variable which represents the factor influencing access to agricultural advice, Y: dependent variable which indicates whether or not producers have access to agricultural advice, Exp: is the exponential function with natural logarithmic base, β: is the slope coefficient, α: is the intercept term.

This paper analyzes the factors which are determinants to the farmer’saccess to agricultural advice. As there are several factors, the logit model considers them as covariates or explanatory variables. By extending equation (1) to a multivariate casein which there are 17 covariates or explanatory variables (X , X ,..., X ) like in this study, we obtain equation (3) 1 2 17 expressed as follows: 𝑃 Y = logit (P) = ln ( ) = α+β X + β X +…+ β X (3) 1 1 2 2 17 17 1−𝑃 From equation (3), we can compute the probability P by taking the exponential (Exp) in both sides of the expression. Hence, the predicted probability value P of adopting compost is expressed as 1 P = (4) 1+[𝑒𝑥𝑝−(α+β1X1+ β2X2+⋯……………..+ β17X17)

Where, Y: farmer group being dichotomous can take two values (with 1 = access to agricultural advice, 0 = no access to agricultural advice); X : Gender, this is the sex of the respondents coded 1 = man, 2 = woman; X : Age, corresponding to the 1 2 age of the respondents in coded years 1 = ˃ 40-old-, 2 = ≤ 40-young-; X : Marital status, specifies the marital status of the 3 respondents coded 1 = married, 2 = single; X : Household, this is the number of people for whom the head of the family is 4 responsible, coded 1 = ˃ 10-large-, = ≤ 10-small-; X : Education, it refers to the education of the respondents coded 1 = 5 educated, 2 = without education; X : Experience, this provides information on the duration in years of practice of the 6 agricultural activities of the respondents coded 1 = ˃ 20, 2 = ≤ 20; X : Membership, it refers to the respondents' membership 7 in a producer organization coded 1 = yes, 2 = no; X : Training, it allows us to know if the farmers surveyed have received 8 training in agriculture before or during their activity coded 1 = yes, 2 = no; X : Area, these are the dimensions in hectares of 9 the farm surveyed coded 1 = ˃ 1ha-grand-, 2 = ≤ 1ha-small-; X : Acquisition, it refers to the mode of land acquisition by 10 respondents coded 1 = owner, 2 = tenant; X : Activity, it corresponds to the different types of main agricultural activities 11 carried out by the respondents coded 1 = agriculture, 2 = livestock, X : Residence, this is the distance from the place of 12 residence of the respondents in relation to the agricultural advisory service coded 1 = close (≤100km), 2 = distant (˃ 100km); X : Awareness, it refers to the producer'sknowledge of the existence of agricultural advisory services and the services they 13 offer coded 1 = yes, 2 = no; X : Availability, information on the participation of respondents in the activities of the advisory 14 service which supervises them coded 1 = yes, 2 = no; X : Proximity, this is the quality of the advisers' interactions with 15 producers through their presence in the field coded 1 = ≤ 14 days-near-, 2 = ˃ 14 days-distant-; X : Request, it is a question 16 of taking into account the requests of the respondents by the advisor 1 = yes, 2 = no; X : Collaboration, it is the 17 collaboration between supervised producers and non-supervised producers among the respondents coded 1 = yes, 2 = no). α: Intercept term; β , β ,…..,β (termed as βk) are respectively the slope coefficients of the explanatory variables X , X ,….., 1 2 17 1 2 X (termed as Xk), to be estimated in SPSS software. 17 The Odds Ratio (OR) derived from the equation (1) where ln (P / 1-P) called log Odds Ratio (OR), representing the degree regression is used to facilitate the interpretation of the results obtained. The exponentials of the slope coefficients βk associated to the explanatory variables are interpreted as OR of adopting the compost (or of occurrence of the event) for each increase in the explanatory variable. In general, since the OR of logit model are just the exponentials of estimated coefficients βk, the positive coefficients will usually display an OR greater than one (OR>1) whereas the negative coefficients will generally indicate an OR lower than one (OR<1). Usually, the expression 1/Exp(B) designates the inverse OR which is computed in order to facilitate the interpretation of the variables with negative coefficients (Wooldridge, 2009).