Policy Research Working Paper 10631 - Data Gaps in Microdata in the Context

They do not necessarily represent the views of the World Bank and its affiliated organizations, nor do they represent the views of the Executive Directors of the World Bank and the governments they represent, or the views of UNHCR. [...] For instance, the UNHCR MDL was launched in 2019 and despite efforts to retrospectively cover and publish older micro-level datasets collected before the inception of the UNHCR MDL, the extent to which older datasets are still missing from the database is unknown.25 23 The results and codes for the analysis are all available from . [...] Figure 3 Panel A plots the relationship between the size of FDP (or the total of refugees, Venezuelan refugees and migrants, and IDPs combined) and the number of publicly available FDP datasets.27 The upward slope seen in the plot indicates that overall there is a positive correlation between the number of FDP and the number of publicly available FDP datasets. [...] Figure 5: Prevalence of Topics in Datasets A) Percent of FDP datasets covering a B) Differences in percent of datasets particular topic in the survey modules covering a particular topic in the survey modules between FDP datasets and non- FDP datasets in the WB MDLs Notes: Panel A shows the percent of non-project-specific FDP datasets collected after 2010 covering each of the 14 different topic are. [...] First, extract the information on the questionnaire modules in the metadata of the identified FDP datasets as well as all the non-FDP micro-level datasets in the WB MDL.37 Second, calculate the frequency of words mentioned in those modules for both FDP datasets and non-FDP datasets.