IDEA: Progressive Data Bank, to democratize and foster more effective Netroots analysis

by: Syrith

Mon Jul 07, 2008 at 13:32

Thanks especially to Chris Bowers' and Paul Rosenberg's interests and expertise, on Open Left we've seen a strongly data-oriented approach to political forecasting, strategy, and activism.  

Folks like Chris and Paul have mined the internet for data, and this data shapes their political analysis, supports their arguments, and lends credibility to their political commentary.  

They and others have aggressively 'trawled the tubes' for publicly-available datasets about American political knowledge, attitudes, and behaviors; demographic and geographic information; financial data; polling breakdowns; and more.

This body of data is foundational to the good work done in the progressive blogosphere - especially here at Open Left and at fivethirtyeight.com.   And though this data is being put to great use by some, the unwashed majority of us don't give much thought to where it comes from and to whom it is available.  


As a response, I propose that we develop a "Progressive Data Bank":
a place where many types of relevant raw data, basic analysis, and guidance on how to use and apply this information, can all be shared broadly, to democratize the enterprise of political analysis and further a wide range of progressive aims.  

Why?  Here's a (highly subjective - please don't get hung up on this!) timeline that may be illustrative:

1) The Netroots and progressive blogosphere had its beginnings as a force for the democratization of political communication and networking.  I'm thinking initially of the recount blogging in 2000, then of Iraq War opposition organizing in 2003.

2) As the blogosphere grew in sophistication and evolved, the Netroots became a force for the democratization of punditry itself.  No longer were we even primarily reporting on events and creating community - we were generating, funding and implementing new strategy as well!  I'm thinking of the 2006 congressional races and ActBlue especially.  

3) Now, I believe we've entered a third stage, where the Netroots is beginning to play a role in the democratization of political analysis.  I see this as a (more profound) extension of our earlier communications and DIY-punditry functions.  Early elements of this include the Bush Dogs campaign, Kos' direct contracting of polling firms, and especially, the growing sophistication of the forecasting approaches taken by Chris Bowers and Nate Silver.  

But the bottom line is, even as our analysis becomes increasingly nuanced, massive barriers to participation remain.  This is to say that the democratization of political analysis (by the Netroots) will necessarily remain incomplete until/unless there are reliable, accessible, user-friendly sources of data for our community to access.

And therein lies my idea.  It is my sense that this vision - a 'Progressive Data Bank' of demographic information, publicly-shared polling results, financial reports, and other marvelous and manipulatable information - simply doesn't exist.

If I'm wrong, tell me!  I'm already certain that all of you wonderful folks can point me to many great resources (pollster sites, Census datasets, GIS resources, etc., etc.) that offer some of these things, in some situations.  But I don't think there's a Netroots-oriented and -friendly resource that really brings it all together.  

There should be a Progressive Data Bank website, or maybe a wiki, in which these resources are gathered, accumulated, and made accessible.  In some cases, it might just be a matter of linking to the right places (and keeping track when websites and data move); in others, datasets could be actually hosted and shared.  

Beyond the data itself, a key component of this resource would be to offer advice and support for the execution of grassroots/volunteer analysis with the data, in all the forms it could take.  This might include tips on good statistical practices, or useful functions to manipulate information.  It might mean suggestions about visualization and representation of the data, and advice about color-coding or sorting results.  It might even be technical - how to use a function in common software like Excel or ArcGIS to make the numbers sing and reinforce the key argument in Joe Diarist's upcoming post.

At fivethirtyeight.com, Nate Silver (Daily Kos' poblano) is already Crashing the Gate of the Pollsters' Ivory Tower, much the same way that Kos and Jerome (and Matt, Chris, and Paul) are battering away at Versailles and the mainstream Democratic politico-social establishment.  But even Nate doesn't let you download his underlying data and work with it yourself (I recently commented about this fact on his site, and maybe he'll get back to me).

In conclusion - A PDB would facilitate grassroots analysis, and could even help refine or direct projects themselves.  Take American Blogger, for instance.  Surely facilitating access to / use of data could support Swing State Newspaper responses, or strengthen the Progressive Investment Effectiveness (PIE!) scoring system.

And just as importantly, a PDB would help level the playing field for the performance of sophisticated analysis - avoiding a 'new elite' of folks like Chris and Nate who have gained the expertise to find and use this data - and furthering the 'democratizing' trend for which the Netroots has been responsible, while upholding progressive values generally.  

What do you think?