## Implications for Poverty Reduction in Rural Households in Ghana: Formal Education

Again primary school and JSS/ Middles school categories of the level of education both have positive impact on weekly expenditure and for that matter poverty reduction given their coefficients as 0.170 and 0.173 respectively. Thus respondents with primary school and JSS/Middle school education spend on the average 17% and 17.3% respectively higher on consumption per week than their counterparts who have no formal education. From Table 1, given the mean value of the various categories of education, the number of respondents within each category is computed as; no formal education are 232; primary school education are 102 and JSS/Middle school education are 66. Thus the number of respondents with primary and JSS/Middle school education combined is 168; this number is less than those with no formal education which is the reference category. This suggests that may be those with primary and JSS/Middle school education are better able to manage their agro-processing businesses very well and so make more money which enables them to spend on consumption more than those with no education. Credit card

The coefficient of the number of dependants in a respondent’s household (depend) is given as 0.0294. This shows a positive relationship with weekly consumption expenditure. Thus as the number of dependants in the household increases by one, weekly consumption expenditure increases on the average by 3%. Possible explanation to this finding is that, the higher the number of dependants in the household, the more the need for respondents (providers) to generate more resources or income for consumption expenditure. Again the more dependants there are in the household, the greater the compelling demand for the women to engage in an off-farm or participate in an MFI programme that enables them generate additional income for consumption expenditure. This finding is contrary to Imai, Arun & Annim who instead used dependency ratio and found a significant but negative relationship with poverty proxied by Index Based Ranking (IBR) score. It is also possible that households engage their dependants particularly older ones in their production activities, thus contributing to household labour and helping to generate more income for the household consumption expenditure. If this is so, then the more the number of dependants in the household the higher the consumption expenditure will be and the vice-versa.

Moreover the number of income generating activities that the respondent engages in (numacty) is positively associated with weekly consumption expenditure. Given a coefficient of 0.0479, it suggests that if numacty increases by one then weekly expenditure on basic needs will increase on the average by 5%. It is not surprising that the respondents engage in multiple income generating activities, perhaps the old adage of not putting all eggs in one basket holds here. The region has just one rainy season. Thus off-farm income generating activities may be contributing substantially to their sustenance hence the motivation to engage in multiple income generating activities.

Again oldsav is positively associated with weekly consumption expenditure with a coefficient of 0.000381. By the coefficient as the amount of initial savings increases by GH01 then weekly consumption expenditure will increase by 0.04%. Though this may be negligible, it is an indication that respondents who are able to mobilise more savings channel these money into their income generating activities therefore enabling them earn more money for their consumption expenditure. Also initial saving may be a criterion for grant of MFI loans. If this is the case those who are able to mobilise more savings get more loans for their economic activities which also helps them spend much more in the future.