The models show numerous factors adding to the likelihood that a household will make use of the arable land to which it has access. Although several variables are statistically insignificant, the robustness of the estimates and the magnitude of the probability changes make the models a useful starting point for discussion.
As discussed earlier, access to non-mechanized agricultural equipment has a large impact on the decision to farm. Households with those tools and equipment are approximately 9 percent more likely to engage in agriculture. The probability change varies little by sex of the household head. This relationship between ownership of non-mechanized equipment and land use does not imply a direction of causality. Whether households that want to farm acquire implements or those that have implements engage in farming is not the primary issue. The important conclusion is that households without these assets are much less likely to farm. If LRP beneficiaries are unable to purchase or acquire them from other sources, and cannot use grant monies for anything other than land and fixed improvements, the probability of engaging in agricultural production sharply declines.
The amount of cropland also affects the likelihood a household engages in farming. Additional land should increase the total return to agriculture and make it a more attractive livelihood creation option. However, the LRP as currently configured will provide relatively small plots of land for its beneficiaries. For a beneficiary household receiving two additional hectares of arable land it increases the probability of farming by 5 to 8 percent depending on the type of household. Female-headed households respond more strongly to additional land, suggesting, as discussed in earlier sections that these households are more land constrained than their male-headed counterparts. .
Traditional economic analysis of land tenure held that private ownership of land provides the strongest incentives for land use, preservation, and fixed investment.