Factors Influencing Participation of Crop Farming Households in Non-farm Activities in Ghana

Agricultural sector is facing serious challenges (low productivity, inadequate investment, changes in climatic conditions etc) in developing countries. The challenges facing Ghanaian agriculture had led to the increasing participation of farmers in the non-farm activities. It is therefore incumbent on researchers to investigate factors that can be modified for farmers to fully participate in non-farm agriculture activities to complement their livelihoods. This study investigated factors that influenced farming households’ decision to participate in non-farm activities using data from fifth version of Ghana Living Standard Survey. The sample size used for the study is 1368. A binary logit estimation model was used for the analysis. The results obtained from maximum likelihood estimation showed that households with greater probability to participate in non-farm activities were as follows: female-headed households; households headed by the young; households whose heads were married, households whose heads had formal education, larger households, Original Research Article Sienso et al.; AJAEES, 6(3): 117-125, 2015; Article no.AJAEES.2015.069 118 households living in urban areas and households living in the forest zone. Crop farming households who are male-headed, had no formal education, had small household sizes, lived in rural areas and in savannah areas should be encouraged and supported to diversify activities by engaging in nonfarm activities.


INTRODUCTION
Agricultural sector continues to contribute significantly to overall sustainable development and poverty reduction worldwide. It is often regarded as the engine of growth in developing countries since it plays a vital role by providing abundant food for the populace, providing input to feed industries as well as employing majority of the labour force. Irrespective of these contributions, many advanced economies have recognized the important role which non-farm activities play in socioeconomic development of their countries. There is a strong correlation between non-farm activities and the agricultural activities as well as income of farmers in rural economies.
Agricultural sector is facing serious challenges (low productivity, inadequate investment, changes in climatic conditions, postharvest losses, diseases and pest infestation etc) in developing countries [1]. The contribution of agricultural sector to gross domestic product (GDP) in Ghana is decreasing year after year. By implication, agriculture cannot support livelihoods in many developing and transitional economies on a sustainable basis and therefore diversification of livelihoods is critical [2,3]. One of the ways subsistence households diversify their livelihoods is through participation in nonfarm activities [4]. [5] described non-farm activities as the engagement in any productive or lucrative work that is away from farm plot. There are many non-farm activities that rural folks in Ghana engage in namely petty trading, seam stressing and agro-processing, blacksmithing, bicycle fitting, carpentry, building and construction, driving, etc.
In Ghana, many youths are shying away from agricultural activities. Therefore, there is the need for concerted efforts by all stakeholders to promote non-farm activities in rural areas to complement their livelihood. The challenges facing Ghanaian agriculture had led to the increasing participation of farmers in the nonfarm activities. Ghana Statistical Service [6] estimates that agriculture employs about 75% of labour in the rural sectors whilst their households' annual expenditure is 1.6 times lower than that of the urban sectors although the farmers have large household sizes. Meanwhile, it is estimated that about 3.4 million households in Ghana operate farm enterprises while about 46% of all households in Ghana operate nonfarm enterprises [6]. Participation in non-farm enterprises is a food security or diversification or livelihood strategy aimed at reducing the effects of income and food variability. It is an open secret that many households in developing countries cannot depend on only on-farm activities for livelihood empowerment. The challenges facing Ghanaian agriculture had led to the increasing participation of farmers in the non-farm activities. Many farmers in developing countries have resorted to engaging in non-farm livelihood strategies to augment agricultural activities. The government, civil society organisations and non-governmental organisations are all calling for farmers' participation in non-farm activities as a shock absorber in times of agricultural failure. It is therefore incumbent on researchers to investigate factors that can be modified for farmers to fully participate in non-farm agriculture activities to complement their livelihoods. There are a number of studies that have been done to investigate factors that influence households' decision to go into non-farm activities in many developing countries [3,4,7]. However, such factors are location and time specific. Meanwhile, they did not disintegrate crop farmers from animal farmers. They failed to realise that socioeconomic factors influencing farmers' decision to engage in non-farming activities differ with respect to the type of farming activities that one is engaged in. It is important therefore, that we find out the specific socio-economic factors that influence Ghanaian crop farming households' decision to go into non-farm activities. The question that arises is-what are the factors affecting crop farming households' participation in non-farm activities?
This paper aims at identifying factors affecting crop farming households' decision to engage in non-farming activities. Also, the magnitude of the effects of each socioeconomic factor is analysed. The findings of this research will contribute to a better understanding of factors influencing the participation of households of crop farmers in non-farm activities and its welfare effect on the households.

Nature of Agriculture in Ghana
According to [8], Ghana agricultural sector is characterised by rain-fed, extensive crop and livestock system, hunting and fishing from natural water bodies. Out of the total land area of 238,535 km 2 , 57.1% can be used for agricultural purposes but only 57.6% of the agricultural land is under cultivation [9]. Agricultural production in Ghana is largely for subsistence. The average land holdings are still on atomise bases with farmers using traditional and indigenous methods of farming. It is a well-known fact that most of the farmers in the rural areas engage in farming as a way of life. The principal feature of this traditional system of farming is small scale production through slashing and burning of the vegetation with the use of hoes and cutlasses. Farmers use uncertified and low yielding varieties and seeds which in effect results in low productivity. The animals are also reared on a smaller scale with extensive and semi-intensive systems. Majority of the farmers are located in the rural areas cultivating staple food crops such as cereals (maize, rice, millet and sorghum), starchy staples which include yams, cassava, cocoyam, plantain, pepper and vegetables (tomato, okra). The main export crop for Ghana is cocoa. The rural folks also engage in the production of animals such as goat, sheep, poultry, cattle, etc. Although the proportion of active labour engaged in farming has been declining over the years, the sector still employs the largest population estimated at 50% in 2012 [10]. It is important to note that the agricultural sector provides food for the nonagricultural sector found mainly in the urban areas.

Non-farm Employment
Non-farm employment has become a topical issue in the contemporary agricultural production. People have vehemently raised several arguments for and against it. The revolution on non-farm activities can be traced back to the marginal revolution that operated on the principle of optimizing marginal returns [11]. [11] argued that technological advancement can help reduce cost and time occupied by farm work, increasing the marginal productivity of labour in agriculture thereby liberating manure labour from agricultural activities. Non-farm agriculture differs from one society to another. The description of non-farm agriculture depends greatly on the individual perspective and understanding. [12] opined that non-farm employment constitutes all activities that are not directly linked with one's crop and non-crop operations but are undertaken as backward-forward linkages to various economic sectors within the rural communities. In this study, non-farm employment is defined as any employment opportunity taken by an individual in any sector of the economy rather than on the field of farm. This includes the service sector, the industrial sector as well as the commerce and trading sectors. [2] noted that non-farm employment is an integral component of the rural economy besides agriculture.
The non-farm sector plays a critical role in poverty reduction and food security especially in developing countries [13]. It helps farmers diversify their livelihood strategies so as to minimise the effects of crop failure due to unforeseen circumstances. During lean season, some of the rural farmers do not have enough food for their households. As such they resort to engaging in non-farm agricultural production activities which serve as means of savings to purchase food in difficult times to feed their households [14]. The fluctuations in the economies especially the income levels of households and prices of foodstuff have compelled many rural farmers to engage in other non-farm activities. The seasonality nature of agriculture results in seasonal income which could not be spread throughout the year, hence farmers participate in non-farm employment to have equal distribution of income throughout the year [9].

Determinants of Non-farm Participation
Socio-economic studies of this nature often involve the identification of factors influencing individual's decision to participate in non-farm activities. It is imperative to note that [2] revealed in their research that the decision of individuals to participate in non-farm employment is influenced by religion, tradition and social orientation. In the study, they found out that the determinants of participation in rural non-farm economic activities are sex, age, marital status, vocational training, belongingness to group and location. These factors as asserted by them influenced non-farm participation decision of individuals positively.
In a study to examine the factors influencing farm and nonfarm income of Haoreconomy in Bangladesh, [15] found out that age and level of education had positive impact on non-farm activities. Also, [16] observed that a household's propensity to engage in a non-farm enterprise is positively dependent on years of education of household head and household size but it has a negative correction with female-headed household and distance to cities. Meanwhile, [17] disaggregated the analysis and observed that socioeconomic factors affect gender participation in non-farm activities differently. For instance, their research results revealed that the proportion of female labour has direct effect on female participation in non-farm activities but the reverse is true for males.

Theoretical Concept of Binary Logit Model
Discrete choice models have become very important in the description, explanation and prediction of individual's choices between two or more alternatives. Binary logistic model is one of the discrete choice models which can be traced back to 19 th century. The decision to participate in non-farm activities can be analysed using binary logit or probit models. Logit model uses a cumulative logistic probability function while the probit uses normal distribution function. This study used binary logit model because its results can be easily interpreted and the method is simple to analyse [18]. Also, the parameter estimates of the logit models are asymptotically consistent, efficient and normal.
Where P is the vector of probabilities of a choice, e is the base of natural logarithm; β i is a vector of estimated coefficients and  is the error term.
The description of explanatory variables is indicated in Table 1.
The forest zone has been used as the reference point (category) that is why it is dummied zero (0) in both coastal and the savannah zones. The parameters were estimated using maximum likelihood estimation technique. The marginal effects of the probability of the choice of farming household were also estimated. To determine the partial effects of the factors X i on P i, the marginal effects are computed by taking the partial derivative of P i with respect to X i . In the logit model, the marginal effects of the variables are measured as the change in the probability of a household participating in off-farm activities as a result of a unit change in X i ceteris paribus i.e.

Study Area
Ghana is a West African country located on the latitudes of 4º 44 1 N and 11º 11 1 S and longitudes of 3º 11 1 W and 1º 11 1 E. The country shares borders with Ivory Coast to the west, Togo to the east, Burkina Faso to the north and the Gulf of Guinea (Atlantic Ocean) to the south. The total land area of Ghana is 238,588 km 2 . Out of this, only 7.1% of the total land size of the country is useful for agricultural purposes whilst 57.6% of the agricultural land is under cultivation [8]. The country has ten administrative regions. The total population of the country is 24,658,823 of which half the population are actively engaged in agricultural activities either directly or indirectly. The economy of Ghana can be grouped into agricultural sector, service sector and industrial sector.

Data Source and Description
The data for the study was secondary obtained from Ghana Living Standards Survey Five (GLSS5) which was collected by the Ghana Statistical Service in conjunction with the World Bank. A total of 1,368 crop farming households were used for the analysis, out of which 684 (50%) engaged in one or more non-farm employments and the other 684 (50%) did not engage in any non-farm employments.

Demographic Characteristics of Households
The demographic characteristics of household heads are shown in Table 2 under the appendix. It can be seen that 78.7% are households headed by males while the remaining 21.3% are households headed by females. This means that majority of households are headed by males as it is a norm in African countries. The age distribution of household heads indicated that respondents within the ages of 40 to 59 years were relatively in high proportion (45%), followed by those who are 20 to 39 years and 60 to79 years. Respondents within the ages of 1 to 19 years recoded the least percentage of 0.3%. From the table, it is seen that 75% of the household heads which represent a good proportion of the working force are engaged in non-farm activities. Out of 1,368 respondents used, relatively high proportion (63.7%) of them had a household size of 1 to 5. In terms of marital status, 67.7% of the household heads were married while 32.3% were single.
Most (43.9%) of the households heads had no formal education, 25.2% had basic education (primary school to junior high school). The least percentage (8.8%) of the household heads had tertiary education. It can be seen from the table that households who earn above GH¢500 from non-farm activities were 17%. A good proportion of household heads (62%) earned less than GH¢100 from non-farm activities.  Table 3 shows the estimated results of the binary logit model. The results showed that all the explanatory variables, except region and savannah zone are significant and maintain their expected signs of the effects. A positive sign of marginal effects shows that there is a direct relationship between the dependent and independent variables whilst a negative sign shows an inverse relationship.

Determinants of Households' Decision to engage in Non-farm Activities
From the regression results, age is significant at 1% and also meets the a prior expectation. The negative sign shows that households headed by the young (the youth) have greater probability of participating in non-farm activities than households headed by relatively elderly people. This shows that younger household heads have greater ability to influence the household members to go into off-farm activities. This could be attributed to the fact that young people are adventurous and more susceptible to change and try new things (are more daring) than the older generations indicating that non-farm activities have bright future. This result is consistent with the findings of [19] who observed that young people have high propensity to participate in non-farm economic activities. Meanwhile, this revelation is contradictory to the observation made by [20] that age decrease offfarm diversification in Nigeria.
The variable sex was negative and significant. This indicates that female headed households had greater probability of participation in nonfarm activities than male headed households.
These findings met the a priori expectation following that women are noted for their engagement in non-farm activities such as petty trading, seam stressing and agro-processing. These findings are consistent with the results of similar studies conducted recently by [2,19].
It can also be seen from Table 3 that households whose heads have relatively high education have greater probability of participating in non-farm activities. This met the a priori expectation. It is an open secret that highly educated household heads have less time for other jobs since they are usually engaged in full time jobs. Considering the fact that farming is time consuming, most of them simply shy away from farming activities. Marital status of household heads significantly and positively affects the decision of one to engage in non-farm activities. This shows that married people have a greater probability of participating in non-farm activities premise to the fact that it is a diversification strategy for family support.
The study also revealed that the probability of a household engaging in non-farm activities increases with household size. These findings conformed to the a priori expectation. Households with larger household sizes have more family income which can be used for nonfarm activities to earn extra income for family up keep. More so, there is a positive relationship between locality of the household head and the non-farm participating decision. This study shows that non-farm activities are more prevalent in the urban centres particularly, the national capital. Households in the forest zone have greater probability of participating in non-farm activities than those who stay far away from forest zone.

CONCLUSION AND RECOMMENDA-TIONS
This study focused on the factors affecting crop farming households' decision to participate in non-farm activities in Ghana. The maximum likelihood estimation results of the binary logit model revealed that; households headed by young people, females, people with more years of formal education and married people had higher probability to participate in non-farm activities Also, farmers with larger household sizes, farmers living in urban areas and those living in the forest zones had higher probability to participate in non-farm activities. It can be concluded that education, marital status, household size, sex, age and location of farmers significantly influence crop farmers' decision to engage in non-farm activities. Those factors should be targeted when one wants to encourage crop farmers to diversify into nonfarming activities.
Considering the fact that non-farm activities are very important in providing alternative livelihoods for crop farming households, it is therefore recommended that stakeholders in the farming sector (extension agents, opinion leaders, NGOs, policy makers) should roll out programmes which aim at helping them diversify farming into nonfarm activities. Since households headed by married people, younger people and females have higher probability of engaging in non-farm activities, governmental and non-governmental agencies in crop farming subsector should implement projects, programmes and policies such as farmers in craftsmanship models as well as skill training models that will encourage unmarried men and women, elderly people as well as males to enter into non-farm activities.
The youth in ICT models currently implemented by government should be extended and repackaged to include the elderly. With this, they will be able to earn some income from the nonfarming activities to help them improve upon their livelihoods. Also, people with less number of years of formal education, families with smaller household sizes and families living in rural areas should be encouraged to diversify their farming activities into non-farming activities through public education and entrepreneurial training. This will help them consolidate their livelihoods.