Following Election Day, Americans everywhere alternately breathed sighs of relief and disappointment. For the average American, the election was over. It would no longer dominate the news cycle, and people could move on with their lives—at least for another four years. But, while it soon became old news for most, it didn’t for a select group of people lucky (or unfortunate) enough to work in one of the gigantic, powerful and fairly messy machines that are American political campaigns. For those people, there was still work to be done to disassemble everything and reflect on which factors had made the campaign a success or a failure. Obama’s reelection was far from a landslide victory. This was an election that could have easily tipped in favor of Mitt Romney, making both campaigns’ every effort even more important.
Recently, controversy has erupted concerning the fate of one such part of these campaigns: their technological property. The Obama campaign’s success can be attributed in part to its vastly superior use of IT resources to leverage “big data.” The term “big data” has become so popular in recent years that it has become dangerously close to being a cliché. It refers to the trend in the tech industry to use massive amounts of data relatively efficiently in ways that were previously unrealistic to implement.
Companies like Google, for instance, rely on dealing with large amounts of data and delivering it at mind-blowingly fast speeds. The type of functionality Google provides requires a large amount of infrastructure, design, programming and upkeep to create and maintain. What’s so remarkable about the past election is that it’s the first time that serious computing resources have been so important in getting a candidate elected. The result of elections has always been, and will always be, primarily determined by the popularity of the candidate, but the way that campaigns are helping their candidate gain that popularity has been changing radically.
The Obama campaign compiled a “dream team” of programmers hand picked from top-tier Internet companies like Google, Facebook and Twitter and dubbed it Team Tech. The developers hired had experience in a work environment with a fast-paced tech startup feel and deploying internet-based products. Despite the team’s wealth of skills, the task they faced was daunting by any standard. They had to create a complex infrastructure at a lightning pace. According to Ars Technica, a leading tech news source, “Scott VanDenPlas, the head of the Obama technology team’s DevOps group, [explained] it in a tweet:
‘4Gb/s, 10k requests per second, 2,000 nodes, 3 datacenters, 180TB and 8.5 billion requests. Design, deploy, dismantle in 583 days to elect the President. #madops’”
If nothing else, the Obama campaign’s success has been a great testament to the versatility, skill, ingenuity and, honestly, political power of the tech community. It also points to the necessity for future campaigns to get on Silicon Valley’s good side.
But things weren’t always this way. In 2008, the, Obama campaign “lacked an internal IT team, relying on vendors and field volunteers to pull much of the weight,” according to Ars Technica. At the time, the technology plan was code-named Project Houdini. Houdini looked like a great plan, but turned out to be incapable of handling the unanticipated high traffic loads that the campaign experienced. Essentially, it crashed because it was too popular. This time around, campaign staffers like Harper Reed, the campaign’s chief technology officer, were determined to build “a bigger boat,” after “the ‘Jaws’ moment” in 2008 that was Project Houdini.
That “bigger boat” was dubbed Narwhal, and comprised of a collection of web apps and services built on Amazon’s cloud computing services that were all designed to intelligently connect volunteers with voters. One of the application’s many functions would analyze what the person who was being called had said in other interactions with the campaign to target marketing more accurately. In other words, the Obama campaign rented server power from the shopping giant Amazon and in so doing, saved on the relatively high initial purchase and ongoing maintenance costs for the servers and the space to house them appropriately, thereby freeing the team to deal only with the actual software coding. The team also lessened their load by using free, open-source software.
“[The team] aggressively stood on the shoulders of giants like Amazon, and used technology that was built by other people,” [Reed] said. “We had a pretty good culture of using not-invented-here technologies. And we weren’t scared about it.”
Therein lies the root of the controversy now surrounding the Obama campaign. In the tech world, there’s a high level of respect for free, open stuff. More than in any other industry, people will routinely make something and give it away for free as open-source software. Consider how strange it would be to find a store giving away free clothes and you’ll see how unique and wonderful the free, open-source movement really is. The understanding is that if you’ve used open-source software as part of your software (called “branching”) then you should make your variant on the original software available for free too. That’s what developers want the Obama campaign to do now, but the campaign and Democrats in general have serious reservations.
They suggest that to make it free is to endanger the advantage they now hold over Republicans. This could jeopardize the success of future democratic candidates while allowing their better-funded opponents to use their software against them. Democrats fear they would be unable to match the strength of Republicans armed with the same software because Republicans might be able to deploy it more widely given their greater funds.
The Democrats’ fears are well founded. They do currently hold a significant advantage, given that the Republican’s recent attempt at a similar technological solution, called Orca, failed miserably. According to another tech news source, The Verge,
“A source in Boston said the tool had been hacked, and an anonymous aide corroborated the reports of crashes, telling The Washington Examiner that, ‘somebody said Orca is lying on the beach with a harpoon in it.’”
Slightly suspicious hacking incident aside, Orca had major issues from the beginning. Apparently, Orca failed in multiple ways. Technical errors and bugs were rampant, and even on a non-technical level, accurate instructions on how to use the tool were nonexistent or deeply flawed. Volunteers were told that Orca was a native mobile app, like those you get on the App Store, when it was, in fact, a web app accessed through a browser. “Training materials were vague and inaccurate,” Verge claimed, and “the URL was for an HTTPS site, but the equivalent HTTP address didn’t forward users to the right page, so anyone who mistyped or simply started with “www” would find nothing.”
But while the Democrats have reason to be afraid of the Republicans stealing Narwhal to spearhead their own efforts next time, the Democrats’ rationale doesn’t really hold much water. Firstly, the Democrat advantage lies not in the software, but in having the people who created it on their side. Deciding to not “play by the rules” and make the software open-source, especially because it could help so many other people in the next four years and beyond, could alienate those developers that are one of the party’s key assets. Numerous non-profits and charities could benefit from the software, as it could allow better targeting of phone calls, maximizing donations. Secondly, the software will probably be somewhat obsolete by the time that another campaign comes around anyway and so there’s no use in holding onto something that’s already depreciating.
But despite the developer’s objections, Narwhal will likely remain under wraps. What will be interesting to see is how these types of systems will be used next time around. With a new Republican fear of being left in the dust if their campaigns fail at technology again, and a new Democratic drive to maintain their technical advantage in the face of real competition next time around, the next election could start a trend of elections driving real technological research and advancement. To be competitive next election might necessitate, according to one Twitter user, “major outreach with Silicon Valley,” so the Republicans aren’t “laughed at,” again. That would likely be difficult, especially considering the Republicans’ more prominent (debatably unfairly, given that both sides pushed it) affiliation with the publicly unpopular SOPA/PIPA legislations that attempted to regulate the internet more closely. Regardless, Republicans cannot have, as one person so eloquently tweeted, another “clusterf**k of biblical proportions,” as some say they did this past election. Whatever ends up happening, there will probably be some big changes coming in how elections are run next time around, and I’m excited to see what that will look like. •
