Abstract: Rural women play a significant role in domestic and socio-economic life of the society such that national development is not possible without developing this important and substantial segment of our society. Considering therefore, the importance of women in agriculture, this study was conducted to examine the role of women in farm management decision-making process. Data were collected in 10 areas of decision making, keeping in mind their importance for agricultural development. The Ordered Probit Model was employed in addition to the descriptive statististics to analyze the data. It was discovered that the socio-economic characteristics of the women farmers significantly affect their decision making in agriculture.
INTRODUCTION
The role of women in agricultural production in Nigeria can never be overemphasized. They perform crucial roles in the domestic and economic life of the society. Rural and national developments can hardly be achieved with the neglect of this important and substantial segment of the society (Kijshor et al., 1999). In recognition of the important role of women in nation building, the Nigerian Government more than ever before is keen upon rural poverty alleviation as a way of improving the economy. As such, focus is on planned and desirable change in the rural societies in the form of agricultural development. The success of these planned change programmes is however hinged largely on the rational decision making process of the women. A lot of literatures (Amali, 1989; Okonjo, 1991; Auta et al., 2000; Rahman and Alamu, 2003) have shown the various contributions of women to agricultural production in Nigeria. The role of women in decision-making process in agriculture has however not widely been explored. Male dominance in decision making in the household and economy has continued even in areas where women are the key providers of labour because the influence of women has not been recognized. The women have more or less been relegated to play second fiddle position in homes and the economy. There is therefore the need to correct for this anomaly.
Considering therefore the importance of rural women in agricultural production, this study was conducted in Zaria area of Kaduna State to examine the level of participation of rural women in the decision-making in different areas of agriculture and also to study the factors influencing their participation in the decision-making process in farm management.
MATERIALS AND METHODS
The study was conducted in Chikum and Igabi Local Government Areas of Kaduna State. A double stage random sampling was employed in obtaining the sample for the study. Four villages were randomly selected out of which 50 women were selected randomly from each village to form the sample size. A total of 200 women were therefore randomly selected for the study. The data were collected for the 2005/ 2006 cropping season with the aid of structured questionnaire. The information gathered includes age, level of education, religion, wealth status and size of land holding. Furthermore, 10 areas of decision-making were also examined in this study to determine the role of rural women in decision making process in agriculture; bearing in mind their importance for agricultural production and development. The level of participation of women in agriculture were analysed by studying the responses on a three point scale:
| • | Whether the women farmers were only consulted |
| • | Whether their opinions were also considered and |
| • | Their involvement in taking final decision |
Simple descriptive statistics and the ordered probit model were used in the analysis.
Model Specification
Since the dependent variables of main interest, factors that determine women
participation in farm management decision making had an ordinal categorical
nature, the ordered probit model was employed for the analysis of the polychotomous
response data. Considering the ordinal probit model, let
Y* = β’Xi+εi |
Where Y* is the underlying latent variable that indexes the level of participation of the woman in a given farm operation, Xi is a vector of parameters to be estimated and εi is the stochastic error term. The latent variable exhibits itself in ordinal categories, which could be coded as 0, 1, 2, 3,.....…, k. The response of category k is thus observed when the underlying continuous response falls in the k-th interval as
Where δi (i = 0, 1, 2, 3) are the unobservable threshold parameters that will be estimated together with other parameters in the model. When an intercept coefficient is included in the model, is normalized to a zero value (Green, 2000) and hence only k-1 additional parameters are estimated with. The probabilities for each of the observed ordinal response which in this study had 10 responses (0, 1, 2, 3) will be given as:
where 0 <δ1<δ2<...<δk-1.ø
is the cumulative normal distribution function such that the sum total of the
above probabilities is equal to one.
| Table 1: | Description of explanatory variables used in the participation model |
The marginal effects of the regressors X on the probabilities are not equal to the coefficients. The marginal probabilities could therefore be calculated from the probit model as:
where φ(.) is the normal density function, the threshold parameter and Xk the k-the explanatory variable. The marginal effect was computed directly from the SHAZAM software employed in the analysis of this study. The list of variables used in the main empirical model is given on Table 1. The dependent variable indexes if the level of participation of the woman is low, medium or high. The value is 1 if low, 2 if medium, 3 if high and 0 otherwise.
The explanatory variables employed in the probit regression model are defined in Table 1. The first three are dummy variables that account for the level of education, religion and the wealth status of the women. Size of land holding was measured in hectares. The age of the women is split into various groups to examine the ole of age groups in women participation in farm decision making.
RESULTS AND DISCUSSION
The data regarding the extent of participation of women in the various areas of farm decision making process is presented in Table 2. The participation of the women in decision making was quite minimal. Table 2 reveals that in each of the farm operations, less than 20% of the women were consulted, except in the sourcing of farm credit where about 28% were consulted; about 13% or less of the women had their opinion considered in each of the farm operations except in storage and marketing where about 46% had their opinion considered. However, only between 1 and 2.5% major took the final decision in all of the farm operations.
The extent of women participation in farm management decision making process depends on a number of factors. Table 3 shows the maximum likelihood estimates of the parameters of Probit regression model characterising the role of women participation in farm management decision making. Ninty percent of the women farmers were correctly classified by the model. The log-likelihood ratio test showed that the estimated model with the set of explanatory variables for the data was better compared with the null model. There was therefore a significant relationship between the log of odds {In [P/(1-P)]} and the probability of the level women involvement in farm management decision-making and the explanatory variables included in the model; suggesting that these variables contribute significantly as a group to the explanation of the role of women in decision-making in agriculture. The Madalla and Mcfadden prove the goodness of the model used in the analysis.
Based on the Probit model analysis, the age group, education, wealth status
and tenancy variables were found to have significant influence on the level
of women participation in farm management decision making. The negative signs
on age groups with ages 35 years and below and the positive signs on the age
groups variables with ages 36 years and above imply that the level of participation
of the women in farm management decision making increases with age: This therefore
suggests that the age of the woman farmer is an important variable influencing
the involvement of the woman in farm management decision making.
| Table 2: | Extent of the participation of women in decision making in agriculture |
| *Figures in parenthesis are the percentages | |
| Table 3: | The maximum likelihood estimates of the probit model |
| *Sig at p<0.05; *Sig at p<0.001; LR test: 49.81; Madalla R2: 0.68 Mcfadden R2: 0.7; Percentage of correct predictions: 98 | |
However, only the 36-50 and >50 age groups were significant at the 5% level. Hence older women participated more in decision making process in the different areas of agriculture than their younger age group counterparts. The significantly positive sign on the EDU variable might be attributed to the high level of knowledge and experience about improved farm practices acquired by the educated woman farmer. This helps her to influence major decisions being taken in the home, farm management inclusive.
The wealth status of the woman is also a major determinant of the role of the woman in farm management decision making. The more financially strong the woman is, the more her involvement decision making process.
In regards to tenancy, the significant positive sign implies that landless women do not make significant participation in farm management decision making. The landless woman undertook farming on leased land and could not as such take a major decision that has to do with the land without the owner’s consent.
Lastly religion was found not to have any significant influence on the level of the woman’s participation in farm management decision making. This might be due to the second fiddle position the woman is forced to play (Sa’adu, 2002) in the country irrespective of her religion.
CONCLUSIONS
It can therefore be concluded that though the woman farmer is heavily involved in agriculture in Nigeria, the level of her participation in farm management decision making is quite low. This can be attributed to the age, education, land tenancy and the wealth status of the woman. Majority of the women interviewed were however found not to be formally educated and are of the low income group.