Neighborhood and agricultural clusters across states of India

In this study we trace how number and members of income clusters have changed in Indian agriculture over the last four and a half decades. Two features which stand out in our results are that not all geographical neighbors belong to the same cluster and clusters include both geographical neighbors and non-neighbors. To identify the factors driving a pair of states to common cluster we then use a logit model and find that smaller is the relative difference between them in terms of mechanization, infrastructural support, deviations from normal rainfall and price differences, higher are the chances that they will be in the same income cluster. Between contiguous and non-contiguous state pairs we find that apart from the common factors, smaller relative differences in irrigation support, rainfall and price differences additionally brings non-contiguous states together.