search

openoakland / opendisclosure

THIS PROJECT IS UNMAINTAINED - SEE: https://github.com/caciviclab/odca-jekyll AND https://github.com/caciviclab/disclosure-backend-static
http://opendisclosure.io/
49 stars 40 forks source link

Maps: Total Contributions by ZIP code; Total Contributions per Capita by ZIP code #274

Open evanwolf opened 9 years ago

evanwolf commented 9 years ago

I'd love to see a map showing TOTAL money raised by ZIP code for the whole election. Perhaps one color with gradients from light to dark?

And another showing TOTAL money raised per voter by ZIP code.

tbarreca commented 9 years ago

$ raised per capita shows how hard campaigners have to work

This isn't obvious to me, Phil. Could you please explain a bit about the thought underlying the assertion?

Thanks,

On Fri, Sep 5, 2014 at 5:11 PM, Phil Wolff notifications@github.com wrote:

I'd love to see a map showing TOTAL money raised by ZIP code for the whole election. Perhaps one color with gradients from light to dark?

  • Was my neighborhood a target?

And another showing TOTAL money raised per voter by ZIP code.

— Reply to this email directly or view it on GitHub https://github.com/openoakland/opendisclosure/issues/274.

Tony Barreca LinkedIn: http://www.linkedin.com/in/tonybarreca Skype: tonybarreca Twitter: tbarreca Mobile: (510) 710-5864

evanwolf commented 9 years ago

@tbarreca, I'm assuming effort in fundraising is roughly proportional to the number of people you're raising money from. This may be bogus in the world of big donor fundraising (bundling?) but it's consistent with my experience in NGO and political fundraising. If true, then...

Or not. If there's another interpretation of $/person, I'm up for it. Maybe the willingness+ability of a community to fund campaigns?

But when we show this number by candidate (so 94612 gave $4/person in this election, and Candidate A raised $2 of it and Candidate B raised $1 and the rest split the final $1) it provides additional perspective on the relative ability of each candidate to extract money from these neighborhoods.

tbarreca commented 9 years ago

Thanks, Phil. Let's discuss this at next Tuesday's hack night.

I think you misconstrue the meaning of "dollars per person." I am hard pressed to imagine that a sophisticated fund raiser would construe the meaning of "person" geographically, as in "by the number if people in a particular Zip code."

I think that fund-raising efficiency, which seems to me to be at least part of what you're getting at, is more naturally and simply stated as $/contributor. That seems to me to be the interesting sense of "person" in this context.

I shall look forward to a conversation.

On Sat, Sep 6, 2014 at 3:27 PM, Phil Wolff notifications@github.com wrote:

@tbarreca https://github.com/tbarreca, I'm assuming effort in fundraising is roughly proportional to the number of people you're raising money from. This may be bogus in the world of big donor fundraising (bundling?) but it's consistent with my experience in NGO and political fundraising. If true, then...

  • Lots of money vs Little money: Returns.
  • Lots of money per person vs. Little money per person: Returns On Effort.

Or not. If there's another interpretation of $/person, I'm up for it. Maybe the willingness+ability of a community to fund campaigns?

But when we show this number by candidate (so 94612 gave $4/person in this election, and Candidate A raised $2 of it and Candidate B raised $1 and the rest split the final $1) it provides additional perspective on the relative ability of each candidate to extract money from these neighborhoods.

— Reply to this email directly or view it on GitHub https://github.com/openoakland/opendisclosure/issues/274#issuecomment-54730572 .

Tony Barreca LinkedIn: http://www.linkedin.com/in/tonybarreca Skype: tonybarreca Twitter: tbarreca Mobile: (510) 710-5864

lla2105 commented 9 years ago

Hey everybody, I agree with Tony - let's talk about this on Tuesday. Make sure to grab me for this convo Tuesday night.

evanwolf commented 9 years ago

For me this is first about correcting for a map's perceptual distortion.

For example, two zips appear the same size but one may have twice as many people or ten times the financial power of the other. We frequently correct in dataviz by showing data adjusted for a common factor that's different the zones. If it's population, you would use show $ per capita. If it's a proxy for the financial health of a district, you might show $ per $million in property value.

The other goal is to expose patterns of influence.

We know Oakland is not a geographically level playing field. We have rich and poor zip codes, gentrification, ethnic enclaves, and zips with more or less commercial, industrial, and vacant property. Did the people of a poorer zip code punch outside its weight class for for some candidates? Did a zip code with more small businesses support the pro-business council candidates? How much did a population's buying power affect how much they spent? Did you raise money in Montclair or Eastmont in absolute terms? Compared to the other candidates?