Skip to content

The project aimed to determine the total sum of contributions to the candidate committees as well as the number of contributions made by individuals.

Notifications You must be signed in to change notification settings

laudebugs/fec-data-analysis-2020

Repository files navigation

Campaign Finance Data Analysis for the 2020 Election Cycle 🇺🇸🐎 🐘 🏃‍♀️👮‍♀️

Background

The 2020 Election, which is one of 2020's major events in American history presents a chance for American's voices to be heard. But even before people vote, a lot of people have made up their minds on how they would vote in the election. One way voters express their opinions is through contributing money to the presidential candidates. Donations are made to committees registered by the Federal Election Commision (FEC). Each of these committees then, in turn uses the money in some way or another to support a candidate or oppose a candidate.

In the 2020 Election cycle, as in previous elections, the FEC discloses donations made to the candidates. This project aimed at looking into individual donations made to the presidential candidates from FEC Bulk Data download. The FEC defines individual donations as donations made by individuals to political committees and are included in the itemized list if the contribution’s election cycle-to-date amount is over $200 for contributions to candidate committees or the contribution’s calendar year-to-date amount is over $200 for contributions to political action committees (PACs) and party committees1.

The data in this research examines the itemized individual contributions made to the candidate committees: Trumps candidate committee: DONALD J. TRUMP FOR PRESIDENT, INC. (FEC Committee ID: C00580100) as well as Joe Biden's candidate committee: BIDEN FOR PRESIDENT (FEC Committee ID: C00703975).

Research Goals

The project aimed to determine the total sum of contributions to the candidate committees as well as the number of contributions made and obtain the following pieces of data points:

  1. The average donation of each candidate
  2. From what industries did each candidate receive a large share of individual donations
    1. in terms of numbers
    2. in terms of the sum of donations
  3. What is the share of employed v. unemployed v retired people who donated to each candidate

Methodology

The Timeline 📅

The individual donations file contains donations by individuals from the beginning of 2019 to the end of August 2020. The bulk data, at the time of the project, did not include individual donations for the month of September and October 2020.

Dataset Overview and Building

Each line of the dataset looks like the following

C00402669|N|M5|P|201905089149647697|15|IND|SHANKMAN, GARY|FAIRFAX|VA|220312721|SERCO INC.|SVP CHIEF FINANCIAL OFFICER|04262019|319||A995D13BAAB3A4913A9F|1329883|||4050820191647460445

The FEC provides a file description for the file.
I wrote a simple script to parse the dataset, and extracted, for each candidate committee, calculated the following:

  1. Total amount and number of donations made to the candidate committee over the time period beginning from 2019 to Aug 2020
  2. The average donation total_amount/number_of_donations towards each candidate committee

Extracting Employment status statistics

Each individual donation includes information on which company the individual is employed in or whether they are self-employed, unemployed or retired. The task was then to determine which industries contributed the most to each of the candidate committees. To match a company to an industry, I considered using an API such as Uplead that returns an overview of a company. However, most of the APIs required a 💰paid subscrition or an enterprise account. The free API from Crunchbase, which also has an enterprise version, returns a company's domain address. With the domain address, I then called the Google Knowledge Graph Search API that returns a JSON-LD result that contains a description and a link to the Company's Wikipedia page.


💡 One thought that came to mind is that this Knowledge Graph may be how Voice assistants look things up on the internet and perhaps return a simple description of, for instance, who Kendrick Lamar is. Worth looking into sometime.

If a request to the Knowledge Graph Search AOI returns a link to the Wikipedia page for the company, I scraped the wikipedia page for the Industry of the company contained within an element (if it exists) such as:

<tr>
    <th scope="row" style="padding-right: 0.5em;">
        Industry
    </th>
    <td class="category" style="line-height: 1.35em;">
        <a href="/wiki/Conglomerate_(company)" title="Conglomerate (company)">Conglomerate</a>
    </td>
</tr>

Storing the company, and it's industry within a file, I then moved to consolidate the reports.
For each candidate, I calculated the total sum, and number of donations from a particular company, then in turn for each industry and has a few metrics to work with and answer the question: What industries donated the most to each presidential candidate?

Results

Preliminary results indicate that there is a very large divide between the kinds of donors that both presidential candidates got. Trump had a larger share of contributors that were retired than did biden. Biden, on the other hand, had a larger share of donors who were are unemployed when compared to Trump. So much have these proportions changed that the divide from the 2016 election cycle is very evident.

I have outlined a report of the analysis here.


1: Individual Contribution definition from [FEC](https://www.fec.gov/introduction-campaign-finance/how-to-research-public-records/individual-contributions/)

About

The project aimed to determine the total sum of contributions to the candidate committees as well as the number of contributions made by individuals.

Topics

Resources

Stars

Watchers

Forks