Please reply to or comment on this post with any questions or thoughts about the data. Powered by this dataset. The FEC data is messy and self-reported; I can't guarantee the accuracy of the underlying data.
Each dot is a company. X = number of contributions (or distinct donors), Y = total $ (both log scale). Per employee divides by a rough US headcount — distinct donors per 1,000 employees is a workforce participation rate. Color = % Democratic (of DEM+REP). Click a dot to filter charts 2 & 3.
Left = number of distinct donors, right = total $. Top 20 job titles by distinct donors. Bars colored by Democratic share.
Distinct donors, number of contributions, or total dollars per period, stacked by recipient party.
One line per recipient category over time. % of donating employees = distinct people who gave to that category ÷ all distinct donors that period — e.g. pick Big tech to read "what share of these employees' donors gave to a Republican committee?" Shares need not sum to 100%: anyone who gives to more than one category is counted in each. % of total dollars instead splits each period's dollars across the shown categories (those do sum to 100%).
Left = absolute stacked total (chosen metric); right = 100% party composition of the same metric. Colored by recipient affiliation.
Each dot is a job title. X = # distinct Republican donors, Y = # distinct Democratic donors (both log scale), dot size = total $, color = % Democratic of DEM+REP donors. The dashed diagonal is parity — dots below it lean Republican, above lean Democratic. Hover for the 3 most common employers.
Each dot is a company. X = rough US employee headcount (log scale), Y = distinct donors (log scale), dot size = total $, color = % Democratic (of DEM+REP). Dots well above the trend punch above their headcount in political giving. Click a dot to filter charts 2 & 3.
Each row is a company, sorted by participation rate. Left = distinct donors per 1,000 employees (donors-per-employee, scaled for readability); right = absolute total $ contributed. Bars colored by Democratic share of DEM+REP.
Each dot is a company. X = # distinct Republican donors, Y = # distinct Democratic donors, dot size = total $, color = % Democratic of DEM+REP donors (distinct people). The dashed diagonal is parity — dots below it lean Republican, above lean Democratic. Click a dot to filter charts 2 & 3.
Contributions, dollars, and distinct contributors per period, summed across the checked companies and stacked by recipient party; the 100% charts below each show party share. Distinct contributors are counted once per period, but anyone who gave to more than one party is counted in each. Use the checkboxes to add or remove companies from the aggregation.
Each dot is a first name (≥10 distinct people). X = political leaning, % Democratic of DEM+REP donors — Democratic on the left; Y = number of distinct DEM+REP donors with that name (log scale). Dot size = total $, color = leaning. The dashed line is 50/50 parity. Labels mark the "frontier": the most common name at each leaning, plus the far-left/right extremes.
Each dot is a company. X = political leaning, % Democratic of DEM+REP donors — Democratic on the left; Y = number of distinct DEM+REP donors (log scale). Dot size = total $, color = leaning. The dashed line is 50/50 parity.
Each dot is a job title. X = political leaning, % Democratic of DEM+REP donors — Democratic on the left; Y = number of distinct DEM+REP donors (log scale). Dot size = total $, color = leaning. The dashed line is 50/50 parity. Every view keeps a ≥20-donor floor (titles with fewer than 20 distinct DEM+REP donors are hidden).