Innovation is widely accepted as an important ingredient for growth and development, particularly coming from small and medium-sized enterprises. However, the conditions under which such enterprises in developing countries innovate and constraints they face are scarcely understood due to poor data collection. In sub-Saharan Africa, this dearth of data is especially pronounced, as local data collection agencies routinely gather data but fail to make it public or standardized. The aim of this project is to create a harmonized, standard data set based on two industrial innovation surveys carried out in 2007 and 2011 in Nigeria. Read more about "Creating micro level dataset innovation Nigeria"

The final output can be downloaded below or at Creating a Micro Level Innovation Dataset in Nigeria and doi.org. This may be updated periodically with fresh data or bug fixes. So far, the data has found wide application in completed and ongoing empirical research both in Nigeria and abroad. Some examples:

  • Egbetokun, A. A., Mendi, P. and Mudida, R. ‘Complementarity in firm-level innovation strategies: a comparative study of Kenya and Nigeria’ (forthcoming in Innovation and Development).
  • Egbetokun, A. A., Oluwatope, O., Adeyeye, A. D. and Sanni, M. ‘The role of industry and economic context in open innovation: Evidence from Nigeria’ UNU-MERIT Working Paper wp2014-073 under review with Technovation (downloadable at Merit Publications)
  • Agbude, E. ‘Source of knowledge as a determinant of innovation performance. Evidence from manufacturing firms in Nigeria’ Master’s Thesis, Maastricht Graduate School of Governance, the Netherlands

The Data

Sampling Notes

  • The sample was randomly selected based on the list of establishments with at least 10 employees obtained from the National Bureau of Statistics (NBS) and the Nigerian Stock Exchange.
  • The Stock Exchange list includes only formal firms whereas the NBS list includes both formal and informal firms. These two sources were cross-referenced and any firm listed in both sources was automatically selected into the sample. The logic is that if a listed firm is still surviving. Note that firm exit rate is particularly high in Nigeria.
  • Subsequently, all other firms were stratified into six geographical zones (North-East, North-West, North-Central, South-West, South-East, South-South) and sector of activity.
  • The final sample was selected by proportional probability.
  • The survey questionnaire was delivered by hand to all the firms, and in many instances, some of the selected firms did no longer exist. In every possible case, the missing firm was substituted with another one in the same sector and geographical location.
  • The survey was first carried out in 2008 (initial sample of 1000 firms) and then repeated in 2011 (initial sample of 1500 firms). The final pooled sample includes 1359 firms, an overall response rate of 54%.

Features

  • The dataset includes data from wave 1 (2005-2007) and wave 2 (2008-2010) of the Nigerian innovation surveys.
  • The year variable identifies the different survey waves. Wave 1 was completed in 2008 and wave 2, in 2011.
  • The service variable sorts the observations broadly into manufacturing and services.
  • The id variable identifies each unique firm. Repeatedness was ignored because repeated cases are only about 2.5%.
  • As much as possible, variables have been matched across the two waves.
  • Due to coding changes and some inconsistencies in the survey instrument, a few variables could not be matched.
  • Any variable that could not be matched is retained in its original form.
  • Some of the variables have notes attached to them. The notes are consistent with what is in the accompanying codebook.xls
  • Item numbering on the questionnaire for the two waves are not consistent. Thus, rather than use question numbers for variable names – as is commonly done – intuitive variable names and labels (defined in detail in the accompanying codebook.xls) are used.
  • Definitions of main concepts can be found in the accompanying codebook.xls.
  • It is strongly recommended that users thoroughly familiarize themselves with the accompanying codebook as well as the questionnaires for each of the waves before applying the dataset. This is crucial especially because of the skip patterns. While everything was done to ensure that the skip patterns were all correctly established, there can be no guarantee of perfection.
  • It is also strongly recommended that users be familiar with the nature of innovation surveys as this will help in understanding how to treat the data for analysis. The Oslo Manual, which is freely available online (Guideline for Collecting and Interpreting Innovation Data), is a very useful resource.

Recommended acknowledgement

  • The National Centre for Technology Management (NACETEM), Ile-Ife, Nigeria carried out the original surveys with funding from the Nigerian government and the New Partnership for African Development (NEPAD).
  • Abiodun Egbetokun of NACETEM prepared this pooled dataset with funding from the Private Enterprise Development in Low-income Programme (PEDL) of the UK's Department for International Development.
  • For more on the programme visit pedl.cepr.org and on this specific project, see Creating Micro level dataset innovation Nigeria.