66 SOCIAL SCIENTIST result will not be a complete reform. Karnataka provides one such example. In that part of the State where the agrarian system was akin to the credit relations model—as reflected in Epstein's study cited above—tenancy abolition was introduced in 1974. As less then five per cent of the households in this district, Mandya, were tenants (Census of Agricultural Holdings, 1971) the effect of tenancy abolition was only peripheral. £ven this small proportion of ten incy was spread across all classes as is reflected in the proportion of tenants in different size clisses of agricultural households. For farms less than 2 hectares in size, 2.22 per cent of the area was under tenancy and this proportion grew marginally to 2.77 per cent for farms in the two to five hectares size class and 3.29 per cent for farms in the above five hectares category. As a result those who lost land because of tenancy abolition belonged to all sections of land holding households. Nearly 10 per cent of the area that was granted to tenants by 1.980, belonged to those owning less than one hectare and as much as 33.78 per cent of the area granted to tenants belonged to widows, minors, unmarried women and physically and mentally disabled persons (Records of the Land Reforms Office, Government of Karnataka, October 1980.) In other words, tenancy abolition in this region orly meant that a small group of cultivators belonging to all sections gained sometimes at the cost of the poorest landowning households. If the measures taken depend on both the type of system and the nature of change, the economic implications are primarily determined by the nature of change. The role of the reform in removing extra economic coersion is likely to be greater in the reforms from below when compared to the reforms from above. In the case of the reforms from above in the land relations type of system and credit relations type of system the reforms Aemselves do not necessarily lead to a change in ttie economic returns to the dominant class. As we have seen the returns in these cases depend on the extent of surplus expropriated before the reforms and the wage bill after the tenants or small holders are converted into agricultural labour. Since it is the same dominant class that can determine both, the reforms per se do not guarantee an increase or decrease in returns. In fact, if the pre reform systems are in the stage of pre-formal capit 'Jist development, the reform merely formalises the changes that have already taken place. On the other hand, in the case of the reforms from below in both types of systems we find that it is very likely that the returns to the tenant or poor peasant, improve as a result of the reforms. The political implications are also influenced more by the nature of the reforms rather than the type of system. In the case of reforms from above a mass based political movement cannot be considered essential in both the land relations as well as the credit relations type of systems. It is the reforms from below that require a mass based political movement. Consequently it is the reforms from below in both types of system that lead