As part of their remit, statistical agencies routinely collect data at the firm level. However this data often remains unused, apart from some cursory descriptive statistics. This is often for a number of reasons, including that the statistical agency does not see value in the data, confidentiality issues or that the records have not been digitised. In light of the scarcity of firm data in Sub-Saharan Africa, and all that we have yet to learn about the hurdles to private-sector development in low-income countries, there is potentially very high value in bringing micro data collected by statistical agencies into the public domain. This project intends to do just that, starting with Ghana and Swaziland, where the researchers have well established connections with the statistical agencies. The team will standardise the data series over time and provide supporting documentation to allow for the use of this data by the broader research community.
In Ghana the researchers aim to make available the two industrial censuses carried out in 1987 and 2003. After acquiring these two datasets, they will collate the variables so that definitions are consistent and they are comparable over time, provide supporting documentation on them, and place these in the research domain. The output will be a firm level data set available to researchers and government statisticians, enabling further work to be carried out on how the industrial structure has changed over the period from 1987 to 2003 in Ghana. The team will also investigate whether linking firms between the two censuses, and the creation of a panel dataset, is possible.
In Swaziland the research team plans to explore what data is available, digitise data which only exists in physical form and establish a panel of firms in Swaziland extending back to 1985. This data will be collected from the Central Statistical Office (CSO) of Swaziland and the Swaziland Revenue Authority. Preliminary investigation suggests that, in addition to firm censuses, the CSO has, at least since 2011, also been gathering information on product-level production and quantities sold domestically, regionally and internationally. This new dataset would thus provide a unique opportunity to examine how firms in a relatively small country which is very dependent on a larger neighbour (South Africa) respond to changes in the neighbouring country.
Authors
