Personal Democracy Forum 2008
Note by Thirsy Mind:
Breakout session: Data geeks unite! Building the new tools for getting out the vote, getting out corruption, and going hyperlocal
Catalist
Mission: to build permanent infrastructure for progressive campaigns and candidates. State of the art tools. Using lifestyle data, niche marketing. Works for clients such as congressional campaigns. Ran an application that allowed people to check if they are duly registered to vote. Taking silos of data that exist across organizations (and internal departments) to create a virtual data warehouse. Can run data sets to see overlaps in different variables: e.g. Republicans and indepdents in NH who value choice.
Sunlight
Digitizing more data, making data more accessible, interoperable and fun. Good interaction between developers and other staff. Biggest non-profit funder of database sets? Have an open API. It’s not about access to our data, but a layer to create a directory of datasets, so they don’t need the overhead costs. Creates a widget which can be embedded into any site to show information on congressional reps. Really into RESTful web services. Using REST to make data pages accessible, interoperable.
Outside.in
Indexing the internet by location. Claims to be “post-editorial” but only allows people to post content within the US. Cares about neighbourhoods, not geodata.
Voter Genome Project
Dems have won on open connectivity, but have lost on proprietary tools. The collapse of ACT systems lost the election for Kerry. Dems paid $12/vote, while GOP paid $3/vote. V from Vendetta quote. Dems do not have a system adequate for a national micro-targeting effort in 2008. Systems need to be truly open, affordable, open to quality control, encouraging innovation. Us + Catalist is 50 people, but that’s not nearly enough people for this kind of data. Shows slides from Scent of a Woman and the Devil’s Advocate. Private vs public voter file: what kind of leaders will they produce? How will they use the information? Control vs capability. Whose party is this? It’s one thing to not sell your data to republicans, but what about competing Democrats in primaries? Wont you be tempted (like Keanu in the Devil’s Advocate). This will determine who leads our party.
Q: Where do you draw the line?
A: Personally, I would put all data out there.
Q: Who pays for it then?
A: The DNC has already paid for it. So publish it. Datasets are not patentable/copyrightable – there’s a reason for that.
Q: Isn’t this data subject to commercial licenses?
A: ...
Sunlight: Biggest challenge is reported vs generated data. Generated data comes automatically from transactions, while reported data is self-reported, needs a lot of uptake, less accurate. Instead of creating a new reported dataset, we should work on cleaning up generated data, because it’s falling from the sky. Once that’s done, we can talk about what reported data needs to be addressed.
Catalist
Mission: to build permanent infrastructure for progressive campaigns and candidates. State of the art tools. Using lifestyle data, niche marketing. Works for clients such as congressional campaigns. Ran an application that allowed people to check if they are duly registered to vote. Taking silos of data that exist across organizations (and internal departments) to create a virtual data warehouse. Can run data sets to see overlaps in different variables: e.g. Republicans and indepdents in NH who value choice.
Sunlight
Digitizing more data, making data more accessible, interoperable and fun. Good interaction between developers and other staff. Biggest non-profit funder of database sets? Have an open API. It’s not about access to our data, but a layer to create a directory of datasets, so they don’t need the overhead costs. Creates a widget which can be embedded into any site to show information on congressional reps. Really into RESTful web services. Using REST to make data pages accessible, interoperable.
Outside.in
Indexing the internet by location. Claims to be “post-editorial” but only allows people to post content within the US. Cares about neighbourhoods, not geodata.
Voter Genome Project
Dems have won on open connectivity, but have lost on proprietary tools. The collapse of ACT systems lost the election for Kerry. Dems paid $12/vote, while GOP paid $3/vote. V from Vendetta quote. Dems do not have a system adequate for a national micro-targeting effort in 2008. Systems need to be truly open, affordable, open to quality control, encouraging innovation. Us + Catalist is 50 people, but that’s not nearly enough people for this kind of data. Shows slides from Scent of a Woman and the Devil’s Advocate. Private vs public voter file: what kind of leaders will they produce? How will they use the information? Control vs capability. Whose party is this? It’s one thing to not sell your data to republicans, but what about competing Democrats in primaries? Wont you be tempted (like Keanu in the Devil’s Advocate). This will determine who leads our party.
Q: Where do you draw the line?
A: Personally, I would put all data out there.
Q: Who pays for it then?
A: The DNC has already paid for it. So publish it. Datasets are not patentable/copyrightable – there’s a reason for that.
Q: Isn’t this data subject to commercial licenses?
A: ...
Sunlight: Biggest challenge is reported vs generated data. Generated data comes automatically from transactions, while reported data is self-reported, needs a lot of uptake, less accurate. Instead of creating a new reported dataset, we should work on cleaning up generated data, because it’s falling from the sky. Once that’s done, we can talk about what reported data needs to be addressed.
