Network Effects and Land Redistribution : A Quasi-Natural Experiment in Zimbabwe

The paper investigates whether positive network effects may have existed between large-scale commercial farmers and small-scale communal farmers prior to the recent land redistribution in Zimbabwe. A difference-in-difference approach is used where measurement is carried out using several data sources including farm level, geographic and survey information for cotton farmers in Mashonaland Central. It tests whether the removal of large-scale farmers has resulted in a decline in productivity for those small-scale farmers close to redistributed land when compared to those with more distant farms. A significant negative productivity effect is found in addition to a country-wide negative redistribution effect in general (due mainly to wider economic and political instability over the last 10 years)