Creating a panel dataset using STATA

[Joseph Lubuto]

Hello colleagues

I am analysing the impact of COVID 19 on firms. I want to use two waves of the World Bank Enterprise Surveys (One collected in 2019 and a follow up collected in 2020 after the Covid19 out break. How can I create a panel dataset for from these two waves so that I can analyse the data using propensity score matching. The variable names are different. I have attached the pictures of the datasets.

Joseph

[Carlo Lazzaro]

Joseph:
you should first -rename- the variables that you want to use as -panelid- and -timeid- in your would-be panel dataset, so that they share the same name..
Then you can use -append- to create the panel.

[Joseph Lubuto]

Hi Carlo

Thank you for the feedback. Would you by any chance have the syntax for doing that. I am still learning the STATA language.

Kind regards
James

[Carlo Lazzaro]

Joseph/James:
the name of the -panelid- (-idstd-) seems already OK.
What you should do is -rename- ing the -timevar- in the second wave (ie, in the second dataset that you're going top -append- to the first one):

rename a14d COVa20d
rename a14m COVa20m
rename a14y COVa20y

Then you can go -append-.

[Joseph Lubuto]

Thank you very much Carlo. I have actually renamed all the variables of interest so that they match in both files.

With warm regards
Joseph James Lubuto

[Joseph Lubuto]

After appending the second wave of the enterprise survey I have noticed that there are missing values/observations on the dataset from the second wave (follow-up survey). I have a categorical variable (a6a) and a continuous variable a6c having missing values/observations. Since they have the same id, I was expecting to have labels or values on the dots (.). For example

idstd	a6a	a6c
680991	small	5
680991	.	.
680992	large	456
680992	.	.
680993	small	5
680993	.	.
680994	medium	56
680994	.	.

Considering that it was a follow up and the id (idstd) variable is the same, I want the dataset to look like this:

idstd	a6a	a6c
680991	small	5
680991	small	5
680992	large	456
680992	large	456
680993	small	5
680993	small	5
680994	medium	56
680994	medium	56

DETAILED TABLES

Table From First Wave

Table From Second Wave