Politics on Rails

Social analysis and programming

NOMINATE

How do your legislators stack up? Compare their votes on a simple graph:

Ted Cruz ACA poll results:

Ted Cruz took a poll on facebook:

It didn't seem to go as planned. One article says it “blew up in his face”. I wrote a program to gather the results for him.

Yes: 1621 - 39%
No: 2515 - 61%

The earlier comments, which are easily visible on facebook, are heavily biased in favor of "yes". Thus the media response. But overall the "noes" get the majority, as you'd expect on a conservative facebook page.

Many people didn't follow instructions, so the numbers aren't perfect. (For example, one person wrote “No doubt at all I'm better off,” which was counted as a “no”.) 230 comments weren't included due to facebook weirdness.

Here's the ruby code, if you're curious:

=begin
A program for estimating the results of Ted Cruz's facebook poll.
=end

# login email and password
email = '[email here]'
pass = '[password here]'

require 'mechanize'

url = 'https://m.facebook.com/story.php?story_fbid=517779935000978&id=315496455229328&p=0'

agent = Mechanize.new

# hiding my Mechanize identity
agent.user_agent_alias = 'Linux Mozilla'

# logging in to facebook
login = agent.get('https://www.facebook.com/')
form = login.forms[0]
form.fields[1].value = email
form.fields[2].value = pass
form.submit

cruz = agent.get(url)

yes = 0
no = 0
failed = 0

# main loop
x = 0
begin
  if x != 0
    cruz = cruz.links.last.click
  end

  begin
    cruz.search("div[class='msg']").last.children.each do |post|
      if post.text[0..2].downcase == 'yes'
        yes += 1
      elsif post.text[0..1].downcase == 'no'
        no += 1
      end
    end
  rescue
    failed += 1
    puts 'page failed'
  end

  if x%10 == 0
    puts ''
    puts "#{x.to_s} pages done."
    puts "Failed: #{failed.to_s}"
    puts "Yes: #{yes.to_s}"
    puts "No: #{no.to_s}"
  end
  x += 1
end until cruz.links.last.text != "View next comments"

puts ''
puts ''
puts 'All done.'
puts ''
puts "Yes: #{yes.to_s}"
puts "No: #{no.to_s}"
puts "Failed: #{failed.to_s}"

Statistics advice from Alex Reinhart

Statistics Done Wrong, a free online book.

"Statistics Done Wrong is a guide to the most popular statistical errors and slip-ups committed by scientists every day, in the lab and in peer-reviewed journals."

HT: Andrew Gelman.

Statistics advice from Bill Burr

Chris Cantwell demonstrates the inherent contradiction of the FSP

In a rare spectacle, Chris Cantwell got himself kicked out of the FSP. Now he's a 'porcupine non grata'.

It's not hard to see why. In an incendiary blog post, he wrote, “the answer, at some point, is to kill government agents,” among other things, while the FSP guidelines clearly say “Participants may be removed for promoting violence, racial hatred, or bigotry.” And this, right after the Concord PD implied the FSP was a 'terrorist' group in a federal grant application. Promoting violence, therefore kicked out, case closed.

But the reaction to Cantwell's removal highlights a quandary for the FSP. Posts on facebook with the tag #standwithCantwell received dozens of likes. A facebook event promoting a hypothetical debate with Chris Cantwell at PorcFest (an FSP event that Chris Cantwell is now banned from) has hundreds of people RSVPed to go.

Apparently a group of rule-hating libertarians just can't take it when someone lays down the law, even in a clear case like this. And that's a major problem.

Any political effort, to be successful, needs to appeal to a lot of people. And if the effort is going to be associated with a brand, such as the Free State Project, then that brand needs to be cared for so that it can be appealing.

Chris Cantwell shows why that can't be done. The FSP's primary goal is to get 20,000 libertarians to move to New Hampshire. Libertarians hate rules. If the FSP became known for kicking out wayward members, that could seriously impact recruitment. So it almost never does that. But because it doesn't, the FSP brand becomes associated with every PR nightmare created by a Free Stater.

FSP president Carla Gericke likes to say the FSP is just the bus [and not involved with movers in New Hampshire]. And the FSP's policy is definitely designed as a bus. But in real life, it's more than a bus. Once you get to New Hampshire, you're a 'Free Stater', forever associated with the Free State Project. You quickly meet a lot of other Free Staters, make friends with Free Staters, and get involved in the nebulous 'Free State movement'. Somehow, you're still on the bus.

The term 'Free Stater' dates back, as far as I know, to the first Free State Project movers. It was probably a convenient label in the past, giving new movers a ready-made community to join when they get to New Hampshire and a comforting group identity.

But there are so many libertarians in New Hampshire now, split into factions and subfactions, all with different strategies that require different, often contradictory, public faces. How much good does the term really do, compared to the amount of damage it does by burdening movers with a controversial label?

The FSP can't be both the bus and the public face of libertarianism in New Hampshire. At one or the other, it will fail. To actually achieve 'liberty in our lifetime', the FSP needs to remove its failing brand from New Hampshire. Other organizations should step up to represent different factions of the liberty movement, like Free Keene, the New Hampshire Liberty Alliance, or even a new third party, and not struggle with being falsely associated with each other, or with Chris Cantwell or any other 'Free Stater'.

The FSP needs to be a bus – and only a bus.

Free Staters vote monolithically (on liberty issues)

William Tucker uses the NH Liberty Ratings graph to argue that 'Free State Project lawmakers form [a] monolithic voting bloc'. But the data isn't on his side.

Here's his modified graph, with Free Staters colored green:

On the y-axis, which measures the Liberty Ratings from the NHLA, some people may see a monolith. Free Staters are all clustered in the top quarter of the graph-- not surprising, since Free Staters claim to vote for liberty.

On the x-axis, however, the argument falls apart. Free Staters range halfway across the spectrum. Michael Sylvia, at .98, is farther right than Bill O'Brien. Joel Winters, on the other hand, is about on par with rogue, moderate Republican Susan Emerson. Any bloc that contains these two rival ends of the Republican Party isn't a bloc at all, much less a monolithic one. And since this axis measures the left/right spectrum, it's the axis most partisans care about.

Tucker seems to overcome this discrepancy by excluding Joel Winters from the Free Stater 'bloc'. But on what grounds? Joel Winters is as much a Free Stater as any of the others, with a Liberty Rating to prove it. In fact, he was the first elected Free Stater in New Hampshire.

And although it isn't apparent on the graph, he isn't an outlier. The two legislators closest to him on this graph are Steve Vaillancourt and Neal Kurk, two high-profile libertarians. Together, these three demonstrate that this is a space libertarians (and Free Staters) actually inhabit.

Libertarians are usually dismissive of the traditional left/right spectrum. To a large extent, this graph reflects that. There are enough conflicts with the left's economic policies that libertarians don't reside on that side of the graph, but otherwise they're all over the place, from the middle to the far right.

So as far as liberals and conservatives are concerned, there really isn't a Free Stater bloc, just a diverse group of people trying to push their own quirky agenda, who at times align with the right, but often don't align with either side.

Next »

Design and content by Will May. Powered by LocomotiveCMS.