Introduction
Our project looks at contributions to political campaigns by members of various institutions in the states of New York, Ohio, and Texas. Using an interactive map, our project compares the party leanings of these donations in comparison to the regions in which the institutions reside. We selected a handful of institutions in each state that were diverse in terms of size, location (urban vs. rural), and regional political affiliation. The state of New York serves as an overwhelmingly Democratic state, while Texas serves as a Republican state and Ohio serves as a swing state.
Why should anyone care about this?
There is a commonly-held belief that institutions are heavily liberal, so we hope to explore whether or not this is true by doing a deep dive into political donations. Futhermore, we thought it would be interesting to see whether the truth of this assumption depends on the political environment geographically surrounding a given institution. We were motivated by many political controversies on our own campus in recent years, such as the Amherst Uprising, the defunding of the Amherst Republicans Club, and the publication of the Office of Diversity and Inclusion’s Common Language Guide.
We quantified the political identity of institutions by exploring which campaigns and candidates they donate too. We know that since we only looked at the donation habits of the employees at the college, this is just an estimate of the true political identity of an instution.
Data Sources
2016/2017 College Scorecard Data
- Obtained from: https://collegescorecard.ed.gov/data/
FEC Campaign Financing, Budgeting, and Spending Data
- Publicly Accessible by anyone at: https://www.fec.gov/data/receipts/individual-contributions/?two_year_transaction_period=2020&min_date=01%2F01%2F2019&max_date=12%2F31%2F2020
- This dataset is readily downloaded as a csv from the ‘export’ function on the website.
Ballotpedia
- A digital encyclopedia of American politics and elections run by a nonpartisan nonprofit organization.
- Used to manually classify donations from FEC data (above) as democratic, republican, nonpartisan, or 3rd party
2016 County Level Census Data (from BigQuery)
- Google’s Big Query provides a public dataset regarding US Census information on demographics, populations, etc. in certain years on a year by year basis that can be accessed from the link below:
- https://console.cloud.google.com/bigquery?project=bigquery-public-data&p=bigquery-public-data&d=census_bureau_acs&t=county_2016_1yr&page=table
2016 State Level Census Data (from BigQuery)
- Google’s Big Query provides a public dataset regarding US Census information on demographics, populations, etc. in certain years on a year by year basis that can be accessed from the link below:
- https://console.cloud.google.com/bigquery?project=bigquery-public-data&p=bigquery-public-data&d=census_bureau_acs&t=state_2016_1yr&page=table
- https://console.cloud.google.com/bigquery?project=bigquery-public-data&p=bigquery-public-data&d=census_bureau_acs&t=state_2016_1yr&page=table
US General Election Presidential Results
County Names and GEOID’s
Variables
From the College Scorecard Data, we used:
- Institution Name
- City
- State Abbreviation
From the FEC’s data on individual contributions, we used
- Contributor Name
- Recipient (the group or political committee receiving the donation)
- State (that the contributor resides in)
- Employer (Institution the donor belongs to)
- Occupation (of the donor)
From Ballotpedia, we used:
- Organization Name
- Political Affiliation (democratic, republican, nonpartisan, and 3rd party)
From BigQuery’s ‘census_bureau_acs’ ‘county_2016_1yr’ dataset:
- GEOID Code
- Median Income of the County ($)
From the ‘county_2018_1yr’ dataset, we can collect the same type of information as above, except on a state-wide basis
- GEOID Code
- Median Income of the County ($)
From the US General Election Presidential Results we used:
- Percentage Democratic Votes
- Percentage Republican Votes
- State Abbreviation
- County Name
From the County Names and GEOID’s dataset, we used:
- County Name
- State Name
- GEOID Code
Results
Our map serves to show how colleges donated and counties voted. The interactivity allows for one to click on a county or college and view specific data. We also perused our data a little farther and found a few interesting caveats. The following analyses explore who is donating, and how many donations are coming from each of the 28 instiutions. These donations were summarized by from the number of donations so in our analysis; multiple donations from the same person are counted as separate donations. If an individual donated multiple times, each of them would count as a separate donation.
Click here to see interactivity for Texas
Click here to see interactivity for New York
Click here to see interactivity for Ohio
A look at who’s donating
Professors and Researchers were among the top donors from the institutions that we looked at which were from Ohio.
Professors and Legal Assisants were among the top donors from the institutions that we looked at which were from Texas.
IT Support Associates, professors, and technical writers were among the top donors from the institutions that we looked at which were from New York.
Institutions and Donations
| Name | Donations |
|---|---|
| Bowling Green State University | 898 |
| Kenyon College | 1139 |
| Marietta College | 12 |
| Oberlin College | 3312 |
| Ohio State University | 24609 |
| Ohio Wesleyan University | 540 |
| University of Mount Union | 95 |
| University of Toledo | 1247 |
Ohio made over 1,000 individual donations to ACTBLUE (19560), IT STARTS TODAY (6780), FRIENDS OF SHERROD BROWN (1165), SWING LEFT (1045), and DCCC (1035).
| Name | Donations |
|---|---|
| Baylor University | 2341 |
| Dallas Baptist University | 168 |
| Lamar University | 373 |
| Southern Methodist University | 3231 |
| Texas AM University | 49 |
| Texas Tech University | 3624 |
| Texas Woman’s University | 222 |
| Trinity University | 926 |
| University of Texas at El Paso | 1043 |
| University of Texas at Rio Grande Valley | 42 |
Texas made over 500 individual donations ACTBLUE (9538), BETO FOR TEXAS (794), and SWING LEFT (779).
| Name | Donations |
|---|---|
| Adelphi University | 1139 |
| Cornell University | 28210 |
| Elmira College | 108 |
| Hamilton College | 2027 |
| Molloy College | 291 |
| Rensselaer Polytechnic Institute | 672 |
| Roberts Wesleyan College | 10 |
| SUNY Canton | 63 |
| SUNY Geneseo | 1047 |
| United States Military Academy | 52 |
New York made over 500 individual donations ACTBLUE (15467), IT STARTS TODAY (14089), SWING LEFT (710), and DCCC (500).
Analysis
Our analyses involved a few subjective decisions. To classify the political leaning each of the 386 FEC committees, we used the nonpartisan website ballotpedia. To classify a counties’ leaning, we looked at the 2016 presidential election results. We computed the difference in % of Democratic Votes and % of Republican Votes and this was the basis of how we implemented the red/blue color scale seen in our maps. We decided that a 50/50 donation cutoff would not be a durable estimate, so our threshold was greater than 60% donations to one party to classify a region as Democratic-leaning or Republican-leaning.
To classify the institutions as a Democratic or Repubican donating majority, we took looked at the % of donations going to Democratic committees, Repubican committees, and Nonpartisan committees from the individuals belonging to the respective institution.
Future Work
Although our work provides interesting and important insight to the political leanings of institutions of higher education, it only scratches the surface of a very complex investigation. Our study was limited in that only 8-11 institutions per state were included, and only 3 states were explored. Furthermore, we classify the “political affiliation” of an institution by the percentage of donations given to Democratic and/or Republican campaigns. We recognize that this may not necessary holistically represent the political affiliation of the institution as a whole, but we believe that it is a good way to classify institutions (especially since voting records are not public information). In the future, this project could be extended to more states and more institutions so that a more complete investigation could be done. With more time (as we had to manually look up each committee donation), more institutions could be classified. With a larger amount of data, statistical analysis of the percentage of donations going to republican or democratic in comparison to an institution’s region could be possible, leading to a statistically valid conclusion (rather than anecdotal). We may also be interested in extending the capabilities of our interactive choropleth by adding an implementation involving Shiny in order to view the distributions of variables like the county median income, population demographic, etc.
Conclusion
Our primary goal was to explore whether or not the political affiliation of the region surrounding an institution had an effect on the political leanings of that institution. We also wanted to explore the general claim that institutions of higher education are usually liberal leaning. Although there are limitations to our study (to be discussed further), we found that all of the colleges we investigated were Democratic. Furthermore, we found the political affiliation of the region in which an institution resides to have no effect on the political affiliation of the majority of the members of the institution. Since all the institutions’ donations were majority Democratic, we explored whether the political affiliation of the region affected the percentage of Democratic donations –– we found the region’s political affiliation to have similarly no effect on the percentage of donations that were Democratic.