Boundary Revisions: the Deconstruction of Moral Personality in Rawls, Nozick, Sandel and Parfit

The boundaries between selves and within selves are notions that have been subject to constant readjustment in the debates between liberals and communitarians. Through the archetype theories of Rawls, Nozick, Sandel and Parfit this essay re-enacts the deconstruction of one of the intuitive certainties of earlier liberal thought. Counterintuitive alliances are struck across the liberal-communitarian divide in demonstrating that the impersonality required by any theory of justice puts pressure on every notion of personality within it. Parfit radicalizes the problem of identity for everyone, permitting us to view old problems from new perspectives.     

Swings and gerrymanders

It is usual for electoral analysts to calculate the swing necessary at the next election to unseat a government, assuming that the swing is a uniform one. However, swings are not uniform. The non-uniform swing which would, on average, unseat a government is shown to be different from the uniform swing. For example, at the next Australian election for the House of Representatives, a uniform swing of 1.5 per cent against the current government would cause it to lose five seats, leaving it in a minority of 73 out of 148 seats, but a non-uniform swing of only 0.8 per cent would, on average, have the same effect. Implications of this variability for the size of the majority, the possibilities of a Parliament with an even number of seats being equally-divided, and the use of this analysis to detect gerrymandering and to correct for it, are considered.     

Representative voter theorems

The representative voter is an individual whose strict preference for any alternative x over any alternative y implies 1) x strictly defeats y by majority rule, if there are an odd number of voters, and 2) x weakly defeats y otherwise. This result holds for the median voter if x is his ideal point or if preferences satisfy a generalized symmetry property, but not in general. We examine a formal condition that guarantees the existence of a representative voter and an economic model in which this condition holds. We also indicate a method for estimating representative voting behavior that is justified when a representative voter exists, and compare this with a method for deriving median voting behavior from an estimated demand curve.     

Problems of co-ordination in upazila administration in Bangladesh

The present government of Bangladesh has extensively reformed the administrative system and upgraded the previous thana to upazila (sub-district) to serve as the focal point of development administration. A new tier of local government called Upazila Parishad (UZP) has been created and a large number of functions have been devolved to this institution along with the placement of central government employees at its disposal. The reformers argue that the present system will remove the inadequacies of the previous administrative system and will bring about an improvement in the co-ordination of field services at the upazila level. However, the actual situation represents a totally different picture. The desired integration— both vertical and horizontal—and co-ordination in local level planning is non-existent due to the absence of any clear direction from the central government. Proper co-ordination among the field services in the upazila administration does not exist due to the anomalous relation between the upazila and the national government, including excessive central control; the involvement of multiple agencies of the central government in guiding and supervising; ambiguity in government thinking; the absence of clarification from the ministries concerned and agencies on the timing of different development programmes; and conflict over the allocation of resources. There is also conflict between the different tiers of local government in respect of authority and resources. Relations between public representatives and officials at the upazila level are strained because a clear allocation of authority to political executives is lacking; areas for co-operation among different agencies have not been clarified; and public representatives and government officials lack mutual respect.     

Decentralizing water supply services in developing countries: Factors affecting the success of community management

This article examines the problem of extending access to potable water through conventionally controlled government services and describes the roles of non-governmental organizations and community management associations in improving service delivery and maintaining local water systems. Six sets of factors that are crucial to the success of community management are identified: adequate incentives, sufficient skills and resources, appropriate processes for water systems operations and maintenance, effective interorganizational relationships, appropriate technology, and effective systems of monitoring, evaluation and feedback. The components that must be taken into consideration in designing and implementing programmes for decentralizing water supply systems through community management are set out.