Creating The Dataset
To create this dataset, our group started by looking through the list of gun vendors put together by Chris Faraone, founder of the Boston Institute for Nonprofit Journalism, and tried to cross reference the list with lobbying data found on the Secretary of State’s website. We were able to confirm the existence of several, but not all vendors in the system, as well as see which lobbyists they had hired.
Faraone sent us a list of gun shops in Massachusetts to look up along with the list of vendors he had already supplied. We decided to try a different searching approach, and instead searched for companies in the corporation database on the Secretary of State’s website to get a list of the CEO or other management positions. We then searched their names on the OCPF database, and also put the names of the company into the “employer” field of the OCPF database to find donations from more employees. We used Table Capture to extract the data from OCPF and put it into a Google Sheet so we could all collaborate on it. We used the same header names as OCPF but eliminated the address and just kept city and state.
To delve deeper into lobbying data, we went back to the lobbying database and searched for entities from another list Faraone put together in his research on the topic. These included state and local police groups as well as gun interest groups. We manually entered data about the client, the lobbyist name, the amount the lobbyist was hired for, the dates of their employment, and the efforts they were paid for.
Currently, this dataset is missing data about lobbyist donations made on behalf of gun companies. This limits the impact of the data and its potential to draw connections between money and legislation. Additionally, there were only four of us inputting data, so we don’t have as large of a dataset as possible. We also could not find data for some vendors and shops. Something else to note about this data is that while some of the money appears to be coming from Connecticut, it is because some employees, and often CEOs, of gun companies in Massachusetts live in Connecticut. However, their interests are inarguably tied to Massachusetts politics.
In order to use Graph Commons, we had to reformat our datasets to fit a Google Sheet template provided to us by Graph Commons.
For the ‘Nodes,’ we decided on three categories based on the data we collected: Recipients (referring to politicians) and Contributor (referring to the person—typically a gun manufacturing executives, employees, or gun retailers—making the donation). These categories are based on the Direct Donations sheet of our dataset.
The categories based on the Lobbying sheet in our dataset are Lobbyist (referring to the individual lobbyist or lobbying entity) and Clients (referring to the group—like The Massachusetts State Police Commissioned Officers Association—who hired the lobbyist). Taser International, Inc., now known as Axon, is a recurring client. We decided to keep Taser International, Inc. for the purposes of visualizing our despite the company’s name change in 2017.
For the ‘Edges,’ which connect the nodes in the visualization, we had to define the different kinds of possible relationship between nodes. We chose DONATED TO, AFFILIATED WITH, and CONTRACTED BY to describe the relationships we discovered in our datasets. We decided to keep just one node and edge per transaction, despite multiple donations. We plan on including the total amount of donations in the description of the node.
For repeated donations, lobbying contracts, etc. we only included them once.
In our research, we noticed that donations also come from spouses and other relatives. We added the marriage connections to the visualizations for the three pairs of spouses we found in our dataset. We added a new edge, or connection, to the visualization to show the marriages among the contributors.
To see our finalized project, click here.