Since it came to office nearly 100 days ago, Britain’s coalition government — a team-up between Conservatives and Liberal Democrats that had the potential to be awkward and ineffective, but has instead (if The Economist’s current cover story is to be believed) emerged as “a radical force” on the world stage — has made 435 pledges, big and small, to its constituents.

In the past, those pledges might have gone the way of so many campaign promises: broken. But no matter — because also largely forgotten.

The Guardian, though, in keeping with its status as a data journalism pioneer, has released a tool that tries to solve the former problem by way of the latter. Its pledge-tracker, a sortable database of the coalition’s various promises, monitors the myriad pledges made according to their individual status of fulfillment: “In Progress,” “In Trouble,” “Kept,” “Not Kept,” etc. The pledges tracked are sortable by topic (civil liberties, education, transport, security, etc.) as well as by the party that initially proposed them. They’re also sortable — intriguingly, from a future-of-context perspective — according to “difficulty level,” with pledges categorized as “difficult,” “straightforward,” or “vague.”

Status is the key metric, though, and assessments of completion are marked visually as well as in text. The “In Progress” note shows up in green, for example; the “Not Kept” shows up in red. Political accountability, meet traffic-light universality.

The tool “needs to be slightly playful,” notes Simon Jeffery, The Guardian’s story producer, who oversaw the tool’s design and implementation. “You need to let the person sitting at the computer actually explore it and look at what they’re interested in — because there are over 400 things in there.”

The idea was inspired, Jeffery wrote in a blog post explaining the tool, by PolitiFact‘s Obameter, which uses a similar framework for keeping the American president accountable for individual promises made. Jeffery came up with the idea of a British-government version after May’s general election, which not only gave the U.S.’s election a run for its money in terms of political drama, but also occasioned several interactive projects from the paper’s editorial staff. They wanted to keep that multimedia trajectory going. And when the cobbled-together new government’s manifesto for action — a list of promises agreed to and offered by the coalition — was released as a single document, the journalists had, essentially, an instant data set.

“And the idea just came from there,” Jeffery told me. “It seemed almost like a purpose-made opportunity.”

Jeffery began collecting the data for the pledge-tracker at the beginning of June, cutting and pasting from the joint manifesto’s PDF documents. Yes, manually. (“That was…not much fun.”) In a tool like this — which, like PolitiFact’s work, merges subjective and objective approaches to accountability — context is crucial. Which is why the pledge-tracking tool includes with each pledge a “Context” section: “some room to explain what this all means,” Jeffery says. That allows for a bit of gray (or, since we’re talking about The Guardian, grey) to seep, productively, into the normally black-and-white constraints that define so much data journalism. One health care-related pledge, for example — “a 24/7 urgent care service with a single number for every kind of care” — offers this helpful context: “The Department of Health draft structural reform plan says preparations began in July 2010 and a new 111 number for 24/7 care will be operational in April 2012.” It also offers, for more background, a link to the reform plan.

To aggregate that contextual information, Jeffery consulted with colleagues who, by virtue of daily reporting, are experts on immigration, the economy, and the other topics covered by the manifesto’s pledges. “So I was able to work with them and just say, ‘Do you know about this?’ ‘Do you know about that?’ and follow things up.”

The tool isn’t perfect, Jeffery notes; it’s intended to be “an ongoing thing.” The idea is to provide accountability that is, in particular, dynamic: a mechanism that allows journalists and everyone else to “go back to it on a weekly or fortnightly basis and look at what has been done — and what hasn’t been done.” Metrics may change, he says, as the political situation does. In October, for example, the coalition government will conclude an external spending review that will help crystallize its upcoming budget, and thus political, priorities — a perfect occasion for tracker-based follow-up stories. But the goal for the moment is to gather feedback and work out bugs, “rather than having a perfectly finished product,” Jeffery says. “So it’s a living thing.”