This project examines a key assumption which underlies one of the main approaches to poverty reduction currently advocated and practised by many international development agencies. Enormous energies and resources are devoted to institutional reform in order to improve the investment climate and thus promote economic growth. The assumption is that institutional reform comes first and investment follows. The project investigates whether this widely assumed sequence applies in the real world or whether, in fact, investment and growth provide the impetus for institutional reform. The project suggests a new way of examining this issue by drawing on comparative intra-country evidence and by combining quantitative and qualitative methods. Vietnam has data on growth, investment and the quality of governance, disaggregated by province and component of governance reform, for five consecutive years. The availability of such panel data makes it possible to examine the sequencing and dynamics of reform. Complementary qualitative research methods will be used to check the quantitative results and explore political dynamics at work. The project was designed and will be executed jointly by a team of IDS Fellows and Vietnamese collaborators.

Our source of firm data is the Vietnam Enterprise Census. This is an annual census of all firms with more than 10 employees with an additional random sample of smaller firms. The data includes a wide range of information on firm characteristics including: sector, employees, assets, legal type, performance, source of capital, and investment. We have data from 2000 to 2010. The number of enterprises increased rapidly over this period from 42,123 in 2000, to over 250,000 in the later years, reflecting the strong growth in private sector activity over the decade. Unfortunately, matching firms across years for the early years is extremely difficult. However, from 2006, a standardised firm identifier was used allowing us to construct an (unbalanced) panel of firms from 2006-2010. Our final panel dataset contains 391,631 firms. Fifteen percent of the firms are measured for all five years, but others are measured less frequently either because of firm entry and exit during the time period under investigation, or because they fell below the GSO threshold of ten employees for inclusion in the census in a particular year and therefore were only subject to random selection, leading to gaps in the data. To measure the quality of local economic governance we draw on the Provincial Competitiveness Index (PCI). The PCI is a composite index of provincial economic governance which has been calculated each year since 2006 by the Vietnam Chamber of Commerce and Industry (VCCI). It is based upon a mail-out survey to a random sample of firms in each province. The survey asks a range of questions about firms’ perceptions of local economic governance, as well as concrete measures of their experience of local governance. A particular strength of the PCI is that it focuses on aspects of local governance which are under the control of the provincial administration. It therefore excludes factors such as the quality or availability of national roads, airports and ports which would bias the index in favour of larger cities or provinces. Firm responses to the questions are combined into a set of nine sub-indices reflecting provincial performance on: • Entry costs • Land access and tenure security • Transparency • Time costs of regulatory compliance • Informal charges • Proactivity of the provincial government • Business support services • Labour training • Legal institutions Provincial scores on each sub-index represent the province’s performance on that topic relative to the performance of other provinces in Vietnam. The overall PCI index is a combination of the sub-indices, yielding an overall score for the quality of economic governance in each province. The published PCI scores use a weighted sum of the sub-indices, with weights determined by the influence of each sub-index in predicting different aspects of firm performance. We use this published PCI, since it is observable to decision makers in firms and in government. In addition to the Enterprise Census and PCI data, we also draw on a range of provincial statistics from the GSO Statistical Yearbook.