Nigerian innovation survey data
Description
The pooled cross-sectional dataset contains 2643 observations including 51% manufacturing and 41% service firms. * The dataset includes data from wave 1 (2005-2007), wave 2 (2008-2010), and wave 3 (2016-2018) of the Nigerian business innovation surveys. * The year variable identifies the different survey waves. Wave 1 was completed in 2008, wave 2 in 2011, and wave 3 in 2019. * 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 three 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 (Excel document). * Item numbering on the questionnaire for the three waves is 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) are used. * Definitions of main concepts can be found in the accompanying codebook. * 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 business innovation surveys as this will help in understanding how to treat the data for analysis. The Oslo Manual (third and fourth editions), which are freely available online, are very useful resources. To have a feel of the sectoral distribution of the sample, type in Stata: tab service year
Files
Steps to reproduce
Sampling Notes * The sample was randomly selected based on the list of establishments with at least 10 employees obtained from the register of the National Bureau of Statistics (NBS), the Nigerian Stock Exchange, the Lagos Chamber of Commerce and Industry (LCCI), and the Manufacturer's Association of Nigeria (MAN). * The registers of the Stock Exchange, MAN and LCCI include only formal firms whereas the NBS list includes both formal and informal firms. These sources were cross-referenced and any firm listed in at least two sources was automatically selected into the sample. The logic is that if a firm is listed, it is still surviving. Note that the 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 sectors of activity. * The final sample was selected by proportional probability. * The survey questionnaire was delivered by hand to all the firms, and in some instances, the selected firms 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 (sample of 1000 firms) and then repeated in 2011 (sample of 1500 firms) and 2020 (sample of 2300). The final pooled dataset includes 2643 firms, an overall response rate of 55%.