A thesis submitted to the Faculty of Commerce, Law and Management, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Doctor of Philosophy. / This research is a contribution to the field of development evaluation. Much of the evaluation practice in development and public policy remains weak in scientific validity, and challenged by issues of attribution and comparability of results across different studies. After an in-depth review of the existing literature and an analysis of the current shortfalls and knowledge gaps in programme evaluation, the research proposes a methodological framework that allows for the empirical measurement and comparison of the impact of diverse types of interventions aimed at addressing a specific outcome of interest. The evaluation framework informs decision-making in social-economic development processes, by combing elements of theory-based counterfactual evaluation, multiple-treatment meta-analysis, mixed methods, and participatory approaches. The evaluation framework is tested in South Africa by utilising the proposed package of methods through two case studies presented in this thesis, to generate evidence for policy-makers, programme managers, and investors operating in the education sector. The first is an evaluation of the impact of the corporate social investments of Anglo American Platinum in Limpopo and North West provinces, that utilised geo-spatial features of mining operations to conduct a quasiexperiment. The second is a comparative analysis of major interventions implemented in South Africa to improve learning outcomes in public schools. The education meta-analysis is the first of its kind to be conducted in South Africa, and has revealed many locally-produced impact studies which had not previously been captured by international reviews on school interventions in developing countries. The empirical work conducted in this research confirms existing theories and reveals new insights into the role of the private sector, the proximity of schools to mines, psycho-social and economic factors, learner age and home language, educational material, quantity and quality of teachers, school management, and accountability systems, in affecting education outcomes. The research highlights some of the programmes and policies which have been most effective in South Africa’s schooling sector, while cautioning about the contextual factors and methodological design features which influence the effect sizes being reported in the evaluations of development interventions. The research concludes by reflecting on the experiences, data and cost analysis challenges, and the lessons learnt from the application of the proposed evaluation approaches in South Africa’s education sector. It discusses the limitations of the framework, and how this can be further refined for future use in other countries, sectors, and development policy contexts. / GR2018

Thesis (Doctoral)--University of Stockholm, 2004. / Added t.p. with thesis statement inserted. Includes bibliographical references (p. 250-278).

published_or_final_version / Politics and Public Administration / Master / Master of Public Administration

published_or_final_version / Public Administration / Master / Master of Public Administration

published_or_final_version / Public Administration / Master / Master of Public Administration

This paper uses the case of Polychlorinated Biphenyls (PCBs) to examine the dynamics of the public policy cycle in Canada. A process approach is applied to examine the principal stages of the cycle: problem identification, agenda-setting, decision-making, and implementation. In examining these stages, the factors that drive the policy cycle and those that impede its progress are identified. The regulatory history of PCBs is traced to demonstrate some of the complexities of the policy cycle. As one of the better known hazardous wastes in Canada, PCBs have been the catalyst for the introduction of new toxic chemical regulations throughout the 1970s and 1980s. Since the late 1970s, regulatory policies have been developed and implemented for PCB use and handling. After major PCB accidents occurred during the second half of the 1980s, regulations were introduced for their transport, storage and disposal. This case study provides seven major conclusions: (1) before the right policy solution is found and implemented, several stages in the cycle may need to be repeated; (2) public perception is a key determinant of the policy problem; (3) focusing events are a critical factor in setting the agenda; (4) decision-making often takes an incremental approach due to incomplete information and divided policy jurisdictions; (5) successful implementation, particularly in divided jurisdictions, requires sympathetic officials who are supportive of the enabling legislation; (6) constituency group support is a necessary condition for policy implementation; and (7) to deal with public concerns which may impede the implementation process, officials need political and managerial skills.

Agenda-setting hypotheses inform political communications studies of media influence (public agenda-setting), as well as examinations of the policymaking process (policy agenda-setting). In both cases, studies concentrate on the salience of issues on actors' agendas, and the dynamic process through which these agendas change and effect each other. The results, narrowly conceived, offer a means of observing media effects or the policy process. Broadly conceived, agenda-setting analyses speak to the nature of relationships between major actors in a political system. This study differs from most past agenda-setting research in several ways. First, this project draws together public and policy agenda-setting work to build a more comprehensive model of the expanded agenda-setting process. Secondly, the modeling makes no assumptions about the directions of causal influence - econometric methods are used to establish causality, allowing for a more nuanced and accurate model of issue dynamics. Quantitative evidence is derived from a longitudinal dataset (1985-1995) including the following: a content analysis of Canadian newspapers (media agenda), 'most important problem' results from all available commercial polls (public agenda), and measures of attention to issues in Question Period, committees, Throne Speeches, government spending, and legislative initiatives (policy agenda). Data is collected for eight issues: AIDS, crime, debt/deficit, environment, inflation, national unity, taxation, and unemployment. The present study, then, is well situated to add unique information to several ongoing debates in agenda-setting studies, and provide a bird's eye view of the media-public-policy dynamics in Canadian politics. Many hypotheses are introduced and tested. Major findings include: (1) there is a Canadian national media agenda; (2) the salience of issues tends to rise and fall simultaneously across Canada, although regional variation exists based on audience attributes and issue obtrusiveness; (3) there is no adequate single measure of the policy agenda - government attention to issues must be measured at several points, and these tend to be only loosely related; (4) the agenda-setting dynamics of individual issues are directly and systematically related to attributes such as prominence and duration; (5) Canadian media and public agendas can be affected by the US media agenda.

Thesis (Doctoral)--University of Stockholm, 2004. / Added t.p. with thesis statement inserted. Includes bibliographical references (p. 250-278).

Title from title page of PDF (University of Missouri--St. Louis, viewed April 5, 2010). Includes bibliographical references (p. 298-354).

Thesis (M(Political Policy Studies))-University of Pretoria, 2005. / Includes bibliographical references and summary.