Groundbreaking New Political Data Site Launches in Philly

It's called Crowdpac.

Screenshot of the Crowdpac homepage.

Screenshot of the Crowdpac homepage.

There’s a brand-new player in the mayor’s race. It could be a gamechanger — not for any candidate in particular, but for all the political nerds who read this blog, and potentially, for voters. It’s a new website called Crowdpac, which bills itself as the go-to source for the “best objective data on US political candidates.” All in one place, you’ll be able to track political donations, make donations, fill out a sample ballot and compare candidates’ right/left leanings on different issues, from the mayoral candidates down to the 42 candidates for Court of Common Pleas.

Above all, Crowdpac aims to bring a level of sophisticated data-analysis that, to our knowledge, has never been attempted before at the local level — in Philly or elsewhere.

When Crowdpac launched last September, it was focused only on national politics. Its central feature, so far, is the Crowdpac score, which CNN described as a “novel political ranking system.” The Washington Post explained some the Crowdpac methodology:

The Crowdpac score is calculated from a mixture of how candidates have voted, who they’ve donated to, who has donated to them and what they’ve said. The site has compiled over a million words, 15,000 contributions and 3,100 votes for Sen. Mark Udall (D-Colo.), for example. The patent-pending algorithm then sifts through the data and generates a rating.

Crowdpac’s founders are a former global development manager at Google, a senior advisor to UK Prime Minister David Cameron and a Stanford political-science professor who developed the algorithms. The company is receiving start-up seed money from Bay Area venture capital companies and Creative Artists Agency. It’s non-partisan, but it’s very much for-profit.

Philadelphia is Crowdpac’s first foray into municipal politics. Why Philly? Partly because the timing was good for the company. But Crowdpac was also drawn to the high-profile role Super PACs are playing the campaign, and to the byzantine nature of Philadelphia politics generally.

Crowdpac figures if its algorithm can help make sense of the complicated, relationship-driven nature of politics in Philadelphia, then it can untangle anything.

“The relationships are playing so much more of a game in Philadelphia, then they would potentially in Congress or Senate or president,” says Liz Jaff, a consultant who worked on the Philly launch. “It’s an interesting test case to see how much we can do.”

Crowdpac’s primary algorithm spits out a political rating, ranging from ultra conservative 10C to ultra liberal 10L. The local Philadelphia ratings are based on two inputs: voting records and who’ve they’ve taken campaign contributions from. For example, if a candidate’s main donor is known for primarily giving to far-right, Ted Cruz-style Republican candidates, then that candidate would have his score tilted significantly toward the right.

As you can see above, Milton Street is on his own island (like he has been throughout the campaign), the only candidate to receive a conservative Crowpac score. The rest are lumped pretty closely together as a moderate-Democratic cohort. But, as you might expect, the ratings grow stratified on individual issues like education.

These comparisons seem accurate, but not particularly revealing. By now, anyone remotely following the race should know that Anthony Williams is the pro-charter/vouchers guy; Jim Kenney is a staunch public-school defender.

But a separate interactive feature showcases Crowdpac’s true data-crunching potential. Using public expenditure reports, Crowdpac created a “donor flow” diagram to show the scope and magnitude of each contribution brought in by the candidates. Take a look at this City Council diagram.

Screenshot from

Screenshot from

This is pretty fantastic stuff. Or, more accurately, it has the clear potential to be fantastic. Crowdpac’s Philly operation is very much a work in progress. The site isn’t yet tracking independent expenditure committees (the Super PACs). The chart above also disaggregates PACs that really ought to be lumped together to provide a clearer picture of who’s funding who (Local 98 controlled PACs, for instance, should probably appear visually as one PAC, not a bunch of individual ones as they do now).

The donor flow diagram is only available for City Council candidates and two of the mayoral candidates so far, due to the availability of expenditure reports. Indeed, the lack of timely campaign finance data — which has everything to do with state and local disclosure laws, not Crowdpac — hinders the usefulness of the site. Give it a full election cycle’s worth of data, and it’ll be a lot more impressive.

Maybe sooner than that. Jaff says there are plans to start tracking Super PAC independent expenditures soon, and she was receptive to the suggestion that affiliated PACs ought to be lumped together.

Down the road, Crowdpac’s ambitions are even larger. The site intends to provide a framework for grassroots recruiting of political candidates, giving would-be donors the ability to pledge a campaign contribution to a candidate not yet in the race. It’s too late for this primary election, but that sort of tool could, theoretically, get talented but reluctant candidates off the sidelines by quantifying for them the level of financial support they could expect to receive if they were to run.

Crowdpac’s ultimate goal? “To help end the stranglehold of big money donors and special interests on the political system and to help create a more representative democracy.”