Dear all,

I am conducting an analysis where I want to see the effects of weather on income in Bangladesh. The weather data i have is panel-form from various weather stations in the country over te time span 2004-2016. As for income, the data I want to use from the Household Income and Expenditure Surveys (HIES) conducted in three waves, 2005, 2010 and 2016. Link to data description:

http://catalog.ihsn.org/index.php/ca...dy-description

The problem I am facing is how to best match entities from the HIES-data sets and make a pseudo-panel in STATA without losing variation in my variable of interest, i.e. total income. For every household we have specific data on different charactersitics, such as age, main income source, district, number of people in household, rural/urban etc.