Scholars of agrarian structure in India have encountered severe difficulty explaining the existence and coexistence of diverse hired labour arrangements. The present paper reviews some attempts and argues that rather than seeking to validate one or other general model, future enquiry should use inductive methods to explain locally encountered diversity. The paper summarises the perfectly competitive model of wage determination and examines the limitations of neoclassical extensions of that model for the analysis of labour exchange in rural India. These are identified as a reluctance to grapple with the coexistence of several arrangement types, the neglect of power relations, the subordination of cultural to economic explanatory factors, the assumption of the homogeneity of labour and the discounting of the importance of labourer solidarity. Orthodox Marxian approaches are then criticised for class reductionism and the implicit assumption of history-as-progress. Even partial explanations of diversity are shown to require the disaggregation of labour arrangements into component parts. Further, the types of labour arrangements themselves need to be more carefully specified to avoid semantic and ideological divergencies appearing as substantive differences in the characterisation of changes in labour arrangements over time.
