{"id":11125,"date":"2018-11-04T13:35:22","date_gmt":"2018-11-04T18:35:22","guid":{"rendered":"http:\/\/studentwork.prattsi.org\/infovis\/?p=11125"},"modified":"2019-01-10T23:56:06","modified_gmt":"2019-01-11T04:56:06","slug":"how-much-influence-does-the-telecom-industry-have-on-congress","status":"publish","type":"post","link":"https:\/\/studentwork.prattsi.org\/infovis\/labs\/how-much-influence-does-the-telecom-industry-have-on-congress\/","title":{"rendered":"How much influence does the telecom industry have on Congress?"},"content":{"rendered":"


\nLast year, Congress voted on not-so-unusual party lines to dismantle a rule from the Federal Communications Commission (FCC) entitled \u201cProtecting the Privacy of Customers of Broadband and Other Telecommunications Services\u201d. This rule would have prohibited internet service providers (ISPs) from selling browsing history and other kinds of personal consumer data to third-party companies without the consumer\u2019s express permission. According to Pew,\u00a0<\/span>one-quarter of adults in the U.S. spend basically every waking moment online<\/span><\/a>, generating tons of data every second that many companies can sell for advertising and other purposes. As the Center for Responsive Politics puts it<\/a><\/span>: <\/span><\/p>\n

\u201c<\/span>The resolution \u2014 which was approved by the Senate last week \u2014 blocks a [FCC] rule that would bar ISPs from selling customer data, including app usage, browsing history, even Social Security numbers, to marketers and others. Widely praised by privacy and consumer advocates when it was finalized last year, the rule hadn\u2019t yet taken effect. Now \u2014 assuming \u00a0President Trump signs it, as he\u2019s expected to \u2014 it won\u2019t.\u201d <\/span><\/p><\/blockquote>\n

President Trump <\/span>signed<\/span><\/a> the resolution to repeal the online privacy rule on April 3, 2017.<\/span><\/p>\n

These same telecom companies that wanted this rule repealed also happen to pump millions of dollars toward members of Congress and their leadership PACs every year to, I assume, influence policy outcomes. Using <\/span>a dataset<\/span><\/a> from the Center for Responsive Politics\u2019 online database, OpenSecrets.org, I decided to explore the possible relationship between the monetary contributions of the three largest telecom companies in the U.S. \u2013 Verizon, Comcast, and AT&T \u2013 from 1989 to 2017 and Congress\u2019 decision to repeal this important federal rule that would help protect consumers\u2019 privacy online. <\/span><\/p>\n

Inspiration<\/h2>\n

A post on OpenSecrets.org<\/span><\/a> discusses net neutrality, a set of guidelines that require ISPs to let all data on the internet be treated equally, prior to the December 2017 vote from the FCC to get rid of it. In that post is a great network visualization on telecom contributions to Congress:<\/span><\/p>\n

\"\"

Source: OpenSecrets.org, Net Neutrality<\/a><\/p><\/div>\n

I was also inspired somewhat aesthetically by <\/span>The Internet Map<\/span><\/a>: <\/span><\/p>\n

\"\"

Source: The Internet Map<\/a><\/p><\/div>\n

Materials<\/b><\/h2>\n

The dataset used in this visualization was <\/span>Contributions from AT&T, Verizon and Comcast to 115th Members<\/span><\/a> from <\/span>OpenSecrets.org<\/span><\/a>, a database managed by the Center for Responsive Politics. The data was cleaned up using <\/span>OpenRefine<\/span><\/a> and <\/span>Google Sheets<\/span><\/a>, then uploaded to <\/span>Gephi<\/span><\/a> to create the visualization. After finishing the design process, I exported from Gephi using the <\/span>Sigma js<\/span><\/a> plugin, which produced a set of HTML, CSS, JS, and JSON files that I could use to design an interactive version<\/a> of my network visualization to publish online. From there, I used <\/span>Dreamweaver<\/span><\/a>, <\/span>TextEdit<\/span><\/a>, and a <\/span>Github<\/span><\/a> repository to do some front-end development, then <\/span>CyberDuck<\/span><\/a> as an FTP client to upload the files to my personal site as a backup.<\/span><\/p>\n

Methods<\/b><\/h2>\n

Below are the methods I used to create this visualization, starting from finding and cleaning the dataset to designing the interface for the interactive version of the visualization.<\/p>\n

The dataset
\n<\/strong><\/span>The dataset<\/span><\/a> I retrieved from OpenSecrets.org was already very organized, clean, and understandable. After some help from Professor Sula, I was able to wrap my mind around avoiding making this visualization a multimodal network (something that doesn\u2019t work well in Gephi) by grouping the nodes representing telecom companies in with the nodes representing members of Congress, with the telecom companies serving as the source nodes and the amount of their contributions serving as the edges and weights between nodes. This also meant that this entire visualization was directed, with the flow of the relationship only going in one direction: from Verizon, Comcast, and\/or AT&T to a member of Congress. It\u2019s important to note here, as the authors of this dataset included <\/span>in the data table<\/span><\/a>, that contributions counted in this dataset did not come directly from the companies themselves, which would be illegal, but rather \u201cfrom the organizations’ PAC, their individual members or employees or owners, and those individuals’ immediate families. Organization totals include subsidiaries and affiliates.\u201d<\/span><\/p>\n

Data cleanup
\n<\/strong>As said above, this dataset was quite clean as-is, but there were a few things I had to do to adapt the data for how I wanted to use it for this lab, including creating separate node and edge tables to upload separately into Gephi. <\/span><\/p>\n

To create the nodes table, I removed all empty columns as well as a column that shows each record\u2019s political party and amount of contribution by political party. I also closed other columns that I didn’t need for the initial visualization until I had the Congressperson\u2019s name column, source column, weight column, and some attribute columns. Not knowing if the other columns would import into Gephi correctly or not, I decided to take them out for now and put them back in later after creating the visualization. The same process was used to create the edge table, taking out all columns except for the necessary ones that would allow the edges to match between nodes. <\/span><\/p>\n

Creating the network
\n<\/strong>In order to indicate the varying amount of contributions between telecom companies to members of Congress, having the nodes representing telecom companies be positioned toward the middle of the graph repelling those congresspeople with little or no contributions outward while pulling those with larger and more frequent contributions inward seemed like the most logical way to show those relationships. In Gephi, this also seemed to be the only network that would work when I ran different types of layout templates over the data. After having the program run \u201cForceatlas 2\u201d as a layout, I then used \u201cExpansion\u201d to spread the nodes out even further. <\/span><\/h4>\n

After I was satisfied with the initial layout, I started editing the appearance of nodes and edges. Because I wanted to show the relationship between contributions from telecom companies and the votes from members of Congress on the 2017 internet privacy bill, I decided to color the nodes representing congresspeople by their votes (blue for \u201cyea\u201d, red for \u201cnay\u201d, and green for those who didn\u2019t vote) and the three nodes representing telecom companies as white, with a black outline around every node in the graph to help them stand out. For the network edges, I used the weight (or \u201cthickness\u201d of the edges) to indicate the amount of contributions from each company toward each congressperson as well as hue (white for thinner lines, with a gradient to red as the lines become thicker\/the contribution amounts go up). <\/span><\/p>\n

For still images of my visualization, I exported a PNG as well as a PDF with labels included and one without. I knew that the bulk of presenting this visualization would happen by creating a version that was online and interactive<\/a>, so I installed the Sigma js plugin for Gephi, exported the files needed to build the site, and opened up Dreamweaver. <\/span><\/p>\n

Building the user interface
\n<\/strong>While creating a Github repository and uploading the files that I exported from Gephi, I discovered that the Sigma.js plugin only includes edits made in the appearance tab of Gephi, not the preview section. I also noticed that some settings don\u2019t export at all, including node outlines, edge weights, and edge colors. After modifying some of the HTML and CSS to change fonts, text, and background color, I also opened the included JSON file to modify edge weights as well as the visualization description on the left side of the interface. When a user clicks on a congressperson\u2019s profile, I noticed that a lot of information they might not need is in the slide out panel, including modularity, in\/out degree, and other information, but I ultimately decided to leave it in. I also re-exported the files from Gephi again after changing the telecom company nodes from white to slightly grey to help differentiate them more in the interface, then replaced the data.json file to update the appearance in Github. <\/span><\/p>\n

After I was satisfied with the interactive version of this visualization, I used Cyberduck to log into the backend of my own personal website where I transferred the files via FTP to make it go live there. Throughout this post, I will link to my Github repository. <\/span><\/p>\n

Results and Interpretations<\/b><\/h2>\n
\"\"<\/a>

A still image of the networked data, with edges weighted and colored by contribution amount (white and thinner edges note lower contributions, red and thick edges note higher). Blue nodes are “yea” votes, red nodes are “nay” votes, and green nodes are those congresspeople who didn’t vote on the resolution. I also created PDF versions with labels<\/a> and without<\/a>, as well as an interactive version<\/a> where you can dive deeper into the data (best viewed on a desktop device).<\/p><\/div>\n

Some of the contribution amounts that congresspeople had received that struck me the most weren\u2019t necessarily the huge amounts that some members got, but rather the much smaller amounts received by a handful of them. For example, Representative <\/span>Pramila Jayapal received $250 from Verizon and $750 from AT&T, putting the node representing her on the outer reaches of the visualization. Even more curious was Carol Shea-Porter who, according to OpenSecrets.org, received $82 from Comcast, $5 from Verizon, and $12 from AT&T. Of course, as noted above, these smaller contributions didn\u2019t come straight from the telecom companies themselves, and likely didn\u2019t even come from their PACs. Rather, they could have come from an employee, a subsidiary, or another person or thing affiliated with the company but not the actual company itself. Three other interesting outliers were Representatives Warren Davidson, John Sarbanes, and Greg Gianforte, who didn\u2019t receive any contributions at all from 1989-2017, and were therefore furthest out from the center of the graph. Gianforte, famous (or infamous) for <\/span>body-slamming a reporter the night before he was elected to Congress<\/span><\/a>, was not officially sworn in until June of 2017 and couldn\u2019t have voted on the repeal of the internet privacy rule. Davidson and Sarbanes were both \u201cnay\u201d votes on the repeal.<\/span><\/p>\n

It\u2019s perhaps not surprising that the overall vote totals almost completely fell on party lines, with only a few Democrats voting \u201cyea\u201d and a few Republicans voting \u201cnay\u201d; that\u2019s how pretty much every vote distribution looks in Congress these days. After looking through the data in the visualization, I\u2019ve ultimately concluded that party affiliation may have been a more powerful factor than just contributions from telecom companies. For example, the c<\/span>ontributions over time to John McCain \u2013\u00a0the late Senator from Arizona who had received a total $794,881 from all three companies \u2013\u00a0may have assisted him in his \u201cyea\u201d vote, but Democrat Steny Hoyer has received a total $659,490 in contributions from all three companies in his political career, and he casted a \u201cnay\u201d vote. In total, the three companies donated $20,101,359 and $25,665,737 from 1989-2017 to Democrats and Republicans, respectively, meaning republicans received $5,564,378 more in contributions over time. When we look at the graph overall, the \u201cyea\u201d and \u201cnay\u201d votes seem to be pretty evenly dispersed among members who received the most donations, represented by the clusters in the middle of the graph that are closest to the telecom company nodes.<\/span><\/p>\n

Lastly, an interesting point <\/span>discussed by OpenSecrets.org<\/span><\/a> is the Republican votes that came from the House: Republicans who voted to repeal the federal rule had an average contribution amount of $138,000 from telecom companies, while the 15 Republicans who voted to keep the rule had received a relatively small average contribution of $77,000 each, showing a possible correlation between contribution and votes among House Republicans.\u00a0<\/span><\/p>\n

Overall, I really enjoyed working with and interpreting this data, and I was actually surprised to find that the telecom industry had less of a clear-cut effect on these votes than I had initially thought they would, with the division between Republicans and Democrats seeming to play a larger role. While there is absolutely influence from contributions (the companies probably wouldn\u2019t be donating millions of dollars a year if there wasn\u2019t), it seems to be less of a factor than party divisions in our current political climate.<\/span><\/p>\n

Reflections for the future<\/b><\/h2>\n

For future network visualizations, I\u2019d like to learn more about the ins and outs of R, as I couldn\u2019t seem to get it to work for what I wanted in this dataset (I ended up using OpenRefine to clean the data). <\/span>I\u2019d also like to learn more about Gephi, Sigma.js, and how these files can be customized more overall. While I\u2019m able to make\u00a0edits to HTML, CSS, JS, and JSON files, there were many areas in the interactive visualization that I wanted to edit more but with no apparent place in the exported files where I could make those edits. <\/span><\/p>\n

One big takeaway from this assignment was that finding a good dataset makes a ton of difference in the entire visualization process overall. I also found that my own predictions about what I\u2019d end up interpreting from this data were actually not as obvious as I thought they\u2019d be,\u00a0with the data having a deeper story to tell.<\/span><\/p>\n","protected":false},"excerpt":{"rendered":"

Last year, Congress voted on not-so-unusual party lines to dismantle a rule from the Federal Communications Commission (FCC) entitled \u201cProtecting the Privacy of Customers of Broadband and Other Telecommunications Services\u201d. This rule would have prohibited internet service providers (ISPs) from selling browsing history and other kinds of personal consumer data to third-party companies without the…<\/p>\n","protected":false},"author":549,"featured_media":11126,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"jetpack_post_was_ever_published":false,"_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_jetpack_memberships_contains_paid_content":false,"footnotes":""},"categories":[149,342],"tags":[274,272,263,275,233,39,243,106,155,271,273],"coauthors":[327],"class_list":["post-11125","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-labs","category-networks","tag-att","tag-comcast","tag-congress","tag-dreamweaver","tag-drew-stanley","tag-gephi","tag-lab-reports","tag-network-visualization","tag-openrefine","tag-telecom-industry","tag-verizon"],"jetpack_featured_media_url":"https:\/\/i0.wp.com\/studentwork.prattsi.org\/infovis\/wp-content\/uploads\/sites\/3\/2018\/11\/featured_image.png?fit=1340%2C682&ssl=1","jetpack_sharing_enabled":true,"jetpack_shortlink":"https:\/\/wp.me\/paBdcV-2Tr","_links":{"self":[{"href":"https:\/\/studentwork.prattsi.org\/infovis\/wp-json\/wp\/v2\/posts\/11125","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/studentwork.prattsi.org\/infovis\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/studentwork.prattsi.org\/infovis\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/studentwork.prattsi.org\/infovis\/wp-json\/wp\/v2\/users\/549"}],"replies":[{"embeddable":true,"href":"https:\/\/studentwork.prattsi.org\/infovis\/wp-json\/wp\/v2\/comments?post=11125"}],"version-history":[{"count":10,"href":"https:\/\/studentwork.prattsi.org\/infovis\/wp-json\/wp\/v2\/posts\/11125\/revisions"}],"predecessor-version":[{"id":11213,"href":"https:\/\/studentwork.prattsi.org\/infovis\/wp-json\/wp\/v2\/posts\/11125\/revisions\/11213"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/studentwork.prattsi.org\/infovis\/wp-json\/wp\/v2\/media\/11126"}],"wp:attachment":[{"href":"https:\/\/studentwork.prattsi.org\/infovis\/wp-json\/wp\/v2\/media?parent=11125"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/studentwork.prattsi.org\/infovis\/wp-json\/wp\/v2\/categories?post=11125"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/studentwork.prattsi.org\/infovis\/wp-json\/wp\/v2\/tags?post=11125"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/studentwork.prattsi.org\/infovis\/wp-json\/wp\/v2\/coauthors?post=11125"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}