^2 SOCIAL SCIENTIST analytically rich studies) he does not consider explicitly the question of the unequal access of landlords and tenants to input and product markets, which also colours the choice of contract. Caballero presents what is essentially the same argument as that of Martinez-Alier, but algebraically rather than diagrammatically. He concludes that the allocative efficiency of wage labour may be greater than that of sharecropping, but the inefficiency of the former resulting from supervision costs, makes it less preferable. Several of the empirical studies highlight some of the theoretical insights about choice of contract and implications for agrarian growth or stagnation. Thus, Keegan's excellent analysis of sharecropping in the arable highveld of South Africa shows concretely how sharecropping in this context was far from being a pre-capitalist relic, but was rather a functional relationship created and structured by the needs of white settler-based agricultural capitalism. Keyder's discussion of Turkish agriculture brings out the forces which caused sharecropping to appear—indebtedness resulting from the small peasantry's forced involvement with the world market; and disappear—technical change in the form of tractorization which enabled the substantial spatial extension of cultivation. Adrienne Cooper, in her description of sharecropping in Bengal, focusses on some aspects of these contracts which have been ignored or bypassed in much of the theoretical literature—perhaps because of the sheer difficulty of dealing with them? Most importantly, she stresses the substantial variations in details of contracts, rent payment and patterns of input cost sharing. Also, the enforcement by landlords of other supplementary forms of payment, such as forced labour and extra cesses, creates difficulties for the easy and common theoretical assumption of a given, universal rent rate (usually half-share). The most significant aspect of the empirical articles—and one which is marked by its absence in much of the theoretical work in this volume (barring the work of Pearce)—is the importance of politics and socio-political power, not only in determining the form of the sharecropping relationship but also its dynamics and evolution through time. Mandle shows how in the post-bellum American south, the sharecropper-based plantation economy sprung up ultimately from the political inability of Hack ex-slaves to ensure land distribution in their favour. Desmond OilFs account of the mezzadria movement among Tuscan sharecroppers in United Italy illustrates that in this case, class political conflict rather than economic struggle for the betterment of conditions, dominated the scene. Stolcke and Hall, discussing the historical evolution of contracts in plantations in Sao Paulo, Brazil, emphasize a political approach. They ^rgue that the transformation of labour contracts (e.g., from slavery to ^ sharecropping to wage labour) is determined by the interrelation between the system of labour exploitation and the pattern of worker resistance. This brief account does not do justice to the various articles, or to the advantage of having a number of studies of this nature brought