Posts Tagged ‘polling’

Engaging a million citizens

December 15th, 2008

Here is an interesting article in the American Prospect that proposes that how Obama administration should try to engage citizens in governance, co-written by my friend Joe Goldman at AmericaSpeaks. The key sentence:

As president, Obama should signal a new kind of governance by calling on the American people to take part in a series of national discussions, each engaging 1 million Americans or more, on the issues of highest public concern, such as the economy, health care, foreign policy, energy, and climate change.”

I know that my advisor Jim Fishkin would criticize AmericaSpeaks recruitment methodology, that allows anyone to participate. He usually argues for random sampling to ensure “formal equality” in the deliberations. But there is definitely something appealing and legitimate about having an open process that offer equal opportunity to participate and make special efforts to recruit people from underserved communities.

I’ll be interested to see if any this talk translates into action!

2. politics , ,

What to watch for on election day

November 3rd, 2008

Here is a snippet of what Nate Silver, author of has to say about the election tomorrow:

7pm EST Virginia, for my money, is the most important state in this election. If John McCain loses it, his path to victory is exceptionally narrow—he would need to pull out an upset in Pennsylvania, while holding on to Florida and Ohio, and avoiding a sweep out West. Barack Obama has considerably more ways to win without Virginia, but a failure to close out the state would suggest at best a more circuitous route to victory. As Obama remains about five points ahead in most polls of Virginia, what we’re really looking for is a quick call on anything before 8 PM that would indicate that the map has indeed changed from 2004, and not in McCain’s favor.

Read the whole thing. It gives an hour by hour analysis of what to look for as the results come in.

2. politics , , , ,

Polls are tied! Could Obama really lose the election?

August 15th, 2008

My neighbor asked me: “The polls show that McCain is now tied with Obama – could Obama really lose this election?”

No. Obama cannot lose the election.

Why? Two reasons. First, I believe the polls overstate McCain support because of how they estimate likely voters. They are greatly undercounting the new voters that will vote for Obama and they are overcounting the demoralized republican turnout. (For example, new voter registration greatly favors democrats.) Second, even these flawed polls will turn around once ads like this start airing:

Yeah, Republicans will not likely be very enthusiastic this fall. I think Obama has an insurmountable 10 point lead in the popular vote that will result in a Reaganesque electoral victory.


Uncategorized , , ,

A local deliberative poll

March 19th, 2008

We’ve just completed another successful deliberative poll this past weekend. This time, I did not have to travel to a far away place; the poll was in my own county, San Mateo, CA. The topic: housing. I do however see a parallel with the EU poll. In the EU, the deliberators had to consider what expansion would mean to Turkey and the Ukraine even though people from those countries weren’t invited. In the San Mateo poll, residents were asked to consider how housing policy affects commuters from the Central Valley, SF and SJ.

Here is a 2 minute news piece on the poll, uploaded by a participant of our poll. Enjoy!

Uncategorized , , ,

Yes, we can

February 4th, 2008

Inspirational video for obama…enjoy!

btw, Obama is up to 50/50 on today. We have a real race here, ladies and gentlemen!

Uncategorized , ,

Informed voting via prediction markets

January 7th, 2008

My friend Richard sent me this interesting link and email in response to my post yesterday:

[ is] moving beyond prediction markets that try to determine who will win
the election, towards markets that predict the effect of someone
winning the election. In other words, markets that attempt to predict
the price of oil, interest rates, # troops in Iraq, etc. conditional
on the person or party that wins the election.

The design of these markets is theoretically interesting as are
questions as to whether you can get enough liquidity in these more
complex markets to get good results.

But beyond that, there is the potential, mentioned in the blog post,
of using these markets to help people decide how to vote. On one
hand, this seems ridiculous – and maybe even open to abuse or
manipulation. On the other, it’s eminently reasonable. I really want
to vote for the candidate that will be the best president and have the
best outcomes for the country. To the extent that I feel that the
markets are better predictors of these things than I am alone (or in
conjunction with the spouting of pundits or even deliberation),
shouldn’t I base my vote on them?

In my opinion, the most important questions facing voters are: what is the consequence of your voting decision? How will the world be different if one candidate wins rather than the other? Using prediction markets to understand these consequences seems entirely reasonable to me.

1. technology, 2. politics , , , , ,

Reading the tea leaves

January 6th, 2008

Obama’s win in Iowa was quite impressive for two reasons: it is important and it was indeed unexpected. On’s prediction market, Obama implicit probability of capturing the nomination jumped from below 20 percent on January 1 to 64 percent after the Iowa caucus. See the huge spike in this graph of Obama’s prospects over the last month:
In comparison, Huckabee’s prospects went down on Intrade with his victory. Apparently, he did not do as well as expected. See the chart here:
According to Intrade, Obama has about a 90 percent chance of winning New Hampshire and South Carolina (separately) and about an 80 percent chance of winning them both. So, the market shouldn’t move much if Obama actually wins those states, because his chance of capturing the nomination already includes a high expectation that he’ll win in those states.

It is also interesting how little “bandwagon effects” that are baked into these probabilities. The market is giving Clinton and Edwards a pretty good shot at dethroning Obama on Super Tuesday even in the high likelihood of Obama running the board on the early primaries.

I guess it really has been one of those elections that is too close for anyone to call, and several more opportunities for candidates to shake things up again.

2. politics ,

Considered opinions about the European Union

October 18th, 2007

I am finally back in the US. After three back-to-back sleepless nights as a global data cruncher, I have the final results of the first-ever European-wide deliberative poll. I’ll have more to say about deliberative polls and this project, but for now, here is a short summary of what we did and what the results are.

What we did:

For the first time ever, a scientific microcosm of Europe was gathered to a single place, the European Parliament building in Brussels, to deliberate in 22 languages about key issues facing the future of the EU and its member states. The participants became dramatically more informed about key issues and changed their views. Participants from the 12 new member countries had different starting points in their opinions but generally changed their opinions more, growing closer in their views to those from the older member states. Over a long weekend, the participants deliberated about economic, social and foreign policy, reflecting on “Europe in the World.”

The results:
* Participants were more likely to support for sacrifices for pensions after deliberation than before
* They were less likely to support for enlargement, mostly coming from new member states learning the reasons against enlargement old member states
* In general, participants from new member states changed more and changed towards old member states
* Very significant knowledge gains
* Participants were more educated than non-participants and in general had small but statistically significant difference in attitude and other measures

Here’s a great press release that has a lot of juicy details:

On that site, you can also find video and other details.

There has also been a fair amount of press too. Here are two examples:,,2193515,00.html#article_continue

2. politics , ,

“SLOPs” and the unintended irony of “freeps”

September 7th, 2007

SLOPS were originally “self-selected listener opinion polls” and now often refer to today’s ubiquitous “self-selected online polls.” As I found out in today’s roundup on, Ron Paul’s supporters are defrauding every SLOP that they can find. They’ve been excluded from one straw poll for their antics, and succeeded in winning a different poll about the latest republican debate based on text-message voting. At, we’ve had similar problems with (“freeps!”) and other libertarians making sure that they were more than adequately represented in our polls.

Now, my advisor at Stanford Jim Fishkin loves to lampoon this old chestnut. He vigorously believes that researchers should control the public opinion process, through techniques like random sampling and moderated discussion. I certainly think that there is a place for his deliberative polling enterprise, but I don’t find it in opposition to SLOPs.

I think SLOPs have an important but different role to play. When I presented at the SXSW conference, I asked for a show of hands for who was republican. Not one of the 100 or so people in the audience raised their hands. SLOPs can give a powerful sense of “who’s in the room,” but only when we have adequate understanding of how the counting is done and what room we are talking about.

Which brings me back to the irony of Paul’s supporters and other free marketers that abuse the rules and intentions of these SLOPs. Aren’t they just proving why rules and regulations are needed to avoid anarchy?

1. technology, 2. politics , ,

Olark Livehelp