General methodology is developed here to deal with the association between a a binary variable and network connections with or without confounding covariates. Also the case when the network is observed at several time periods is treated. As an application we consider the diffusion of organic farming in the province of North Karelia in Finland. It turns out that organic farms are more clustered than would be expected under pure random allocation. The neighborhood effect remains when adjusting for the production lines of the farms. The spatio-temporal analysis shows that new adopters are more often found within the neighborhoods of each others and of earlier adopters.

No abstract

In 1960 a commission headed by Professor Richard M. Titmuss submitted a blueprint of social policy for the consideration of the government o f Mauritius -then a self-governing British crown colony. The proposals which covered the whole field of social security, health and welfare were dominated by the explicit purpose of slowing down the excessive rate of population growth in the island (see Titmuss & Abel-Smith 1968). Therefore, as was reported, the members of the commission hesitated whether or not to include a family allowance scheme among their suggestions. Their enquiries had shown that there were family needs which obviously could not be met without family assistance. On the other hand, however, they felt that the introduction of such measures could violate the objectives of their program by contributing to the persistence of high level fertility. The report made the following comment (ibid., p.

No abstract

No abstract