Dear all,
I am working in stata 15. I have election data (https://ppeg.wzb.eu/www/ppeg_parl_2024v1.zip) from different countries (cname_en) and their parties (cmp or pname_en) and want to calculate the mean vote share (v_share) for right-wing populist parties across these countries and over time. The problem is that elections in different countries are only every 4 to 5 years. To calculate a mean, I have to first add observations for each country and each party in between elections. So if Party A in country B won 20% in 2015, until the next election in 2019 each year must display the vote share (v_share) from 2015. Additionally, for a party that does not gain any shares in a subsequent election (e.g. because they dissolved), the vote share from the previous election should not be carried forward
My first code was looking somewhat like this:
The drop of years above 2023 was necessary because the code generated values past 2023.
The problem is that for each party and year I would mostly just have one value and not the forwardcarried values of the past election. I tried also the carryforward command but it fails because each year appears several times.
Maybe there is also another way to interpolate values that I am not aware of.
I hope someone can help me with this! Also, it's my first post so if something is missing, let me know.
Thanks in advance!
Ramon
I am working in stata 15. I have election data (https://ppeg.wzb.eu/www/ppeg_parl_2024v1.zip) from different countries (cname_en) and their parties (cmp or pname_en) and want to calculate the mean vote share (v_share) for right-wing populist parties across these countries and over time. The problem is that elections in different countries are only every 4 to 5 years. To calculate a mean, I have to first add observations for each country and each party in between elections. So if Party A in country B won 20% in 2015, until the next election in 2019 each year must display the vote share (v_share) from 2015. Additionally, for a party that does not gain any shares in a subsequent election (e.g. because they dissolved), the vote share from the previous election should not be carried forward
My first code was looking somewhat like this:
Code:
cap drop year gen year = year(edate) fre year drop if year <2000 keep cname_en cmp electorate total_vote valid_vote party_id pname_en v_share year * sorting dataset sort cname_en pname_en year * Create a new variable for the last year with data for each party in each country by cname_en pname_en (year): gen last_year = year[_N] * Expand the dataset to include all years for each party in each country expand last_year - year + 1 * Create a sequence of years for each party in each country bysort cname_en pname_en (year): gen seq_year = year[1] + _n - 1 * Replace the year variable with the sequence of years replace year = seq_year drop seq_year last_year drop if year > 2023
The problem is that for each party and year I would mostly just have one value and not the forwardcarried values of the past election. I tried also the carryforward command but it fails because each year appears several times.
Maybe there is also another way to interpolate values that I am not aware of.
I hope someone can help me with this! Also, it's my first post so if something is missing, let me know.
Thanks in advance!
Ramon
