Abstract
In order for a pluralist democracy to function well it is necessary that the means to bring forward policy proposals should be dispersed widely among a variety of political parties, pressure groups, and institutions. The goal of this thesis is to define and solve the problems of creating a computer model of tax-benefit policy suitable for a pluralist society. Computer modelling of tax-benefit policies poses two serious problems. Firstly, can such computer models be sufficiently easy to use so that non-experts can use them without the need for computer specialists? Secondly, can they be flexible enough so that truly innovative policies can be simulated, or must the user be restricted to a narrow set of policy options? The first section of the thesis defines the major dilemmas involved in creating a tax-benefit model. Chapter 1 outlines the problems inherent in solving the aforesaid problems of usability and flexibility. Chapter 2 shows how far these problems have been solved already by describing the features of eleven existing models of the British tax-benefit system. The second section of the thesis examines the store of knowledge on which an ideal tax-benefit model should be based. Chapter 3 discusses the various principles which have been advanced as yardsticks by which to measure tax-benefit policy. Chapter 4 examines four case histories of major changes to the tax- benefit system in the past. Chapter 5 makes a detailed examination of the form of behavioural response to tax-benefit policy which has been most frequently explored by policy researchers : the extent to which work incentives are affected by tax-benefit policy. The third section describes the characteristics of a new computer model and a new computer language which have been constructed for this thesis which attempt to deal with the problems of tax-benefit modelling in a pluralist society. Chapter 6 outlines the general features of the data, the new computer language, the Policy Simulation Language, and the new model the Policy Option Model which have been created. Chapter 7 gives a detailed explanation of how the features of the Policy Simulation Language and the Policy Option Model have been implemented. Chapter 8 describes the characteristics of the Policy Option Model's user interface. Chapter 9 gives examples of the output from the Policy Option Model. In order to collect the data for chapter 10 the Policy Option Model was demonstrated to a number of people in the field of tax-benefit modelling who use other models. Chapter 10 contains a selection of the comments of these people after these demonstrations. Chapter 11 outlines possible future developments of the Policy Option Model and discusses the extent to which it has succeeded in solving the problems of tax-benefit modelling in a pluralist society.