Des Moines Register Likely Voters Not Likely

The end does not justify the means. Des Moines Register predicts a couple dozen elections fairly well, so no one questions the details anymore. They just consider it the Gold Standard and accept whatever it says as a matter of faith.

But math isn't based on faith.

As Jeff Roe points out, despite his clearly partisan motivations, the Des Moines Register called 3019 registered voters and determined that 602 were likely to caucus which is a 19.9% turnout. Traditionally turnout is 6% turnout. A 19.9% turnout would yield 386,000 GOP caucus voters when the highest all time was 131,000. In other words, not very likely.

Why would the Des Moines Register allow 602 people to represent Iowa?

Math.

In one case the math would render the Des Moines Register poll as unimportant, in another case it might bankrupt their polling budget, assuming they had the manpower to accomplish the task.

To get a 6% turnout of the 3019 registered voters they contacted, that would mean only 181 would provide the answers for their poll. A standard margin of error for such a small sample is 7.2%.

Can you imagine the media reporting on a poll with a 7% margin of error?

That would mean that Trump's numbers are 95% . to be somewhere between 21 and 35. Cruz would be somewhere between 16 and 30. Rubio would be somewhere between 8 and 22. In that scenario, an analyst could plausibly suggest that Rubio could win the Iowa Caucus.

The alternative math solution is to greatly increase the number of registered voters reached. For a 6% turnout that means reaching 10,033 registered voters. That would cost the Des Moines Register at least 3 times more money, with at least 3 times as many staff to contact them. Do they have that much staff? And with all the poll fatigue, it might take a lot more calling to reach a given registered voter. So is 4 times the cost and 4 times the callers an unreasonable requirement?

The Des Moines Register claims they aren't trying to predict turnout, but that is a screen to keep the discussion away from a huge margin of error or a huge cost of doing correct polling.

The fundamentals of statistical sampling assume your sample represents the general population. Who believes that people that are judged by the Des Moines Register are THREE TIMES MORE LIKELY to vote than the general population have views that represent the general population?

On the Democratic side the numbers also show questionable math.

They again accepted 602 (more on that suspicious number later) registered voters as likely to vote in the democratic caucus which again produces the 386,000 number of actual democratic caucus voters even though 2008's record SMASHING attendance had only 239,000 democratic caucus voters. A number 1/3 that size (130,000) seems more reasonable which would match (after inflating for population growth) the 2004 Democratic Caucus attendance when we had the revolutionary Dean, the eventual nominee Kerry, Speaker of the House Gephardt from next door Missouri and many others to bring out voters.

But cutting that much means just 203 likely voters with a 6.9% margin of error. This means analysts trying to guess whether Clinton's 45 really is a number between 38 and 52 and Sander's 42 really is a number somewhere between 35 and 49.

In other words, the analysts and media pundits would dismiss the Des Moines Register poll as just another poll result in a field of dozens of polls.

Or the Des Moines Register could triple or quadruple its staff to try to reach 10,000+ registered voters who would actually answer their phone call.

The most likely result is the Des Moines Register would not have the resources to contact that many registered voters so they would either be forced to report a too large margin of error, or hope that their credentials keep anyone from looking under the hood at their suspect math.

The last minor point is their choice of EXACTLY 602 people to vote in the GOP caucus and the same exact number in the democratic caucus. There is 0% reason to believe the size of the two parties caucuses will be the same. They Democrats have much larger caucuses whenever both parties have no incumbents. This forced choice of tallying the voices of more GOP likely voters than is going to happen is for appearances. They don't want to confuse data illiterate media analysts and pundits with HONEST MESSY data (which is normal in the real world). If they present two different margins of error, their might be questions. People might look closer at the data and their methods.

Finally, there are much worse polling methods used. The CNN one at Christmas time 2011 with 0% independents in its sample launched the viability of Rick Santorum. That was far worse than this Des Moines Register poll.

This post is mainly to point out that math counts. Changing the math to make things look good to the media does not.

Who's on the Bubble for CNBC GOP Debate

The short answer is that likely all 10 left from the last main debate will be together again and there will only be 2 more polls that determine this. The longer explanation follows.

Bob Jordan (@bobjordanapp) wrote on Sep 29:

Chuck Todd, NBC’s political director and host of Meet the Press, has expressed a willingness to shorten the stage after 11 candidates plodded through a three-hour marathon debate on CNN two weeks ago.

Todd has a lead role in establishing debate rules for CNBC, an NBC affiliate.

“Let’s just say the goal is to create a threshold that candidates have to meet to qualify for the stage rather than committing to putting 10 candidates on the stage. And I don’t think we should commit to more than 10-candidate debates,’’ Todd said in an interview on ESPN radio. “You have to be viable. So now we’re in debate three it’s time to show viability and only the viable ones survive.”

A day later, the CNBC published its debate qualifier rules: 

To qualify for the primetime debate, candidates must average 2.5 percent on the national polls from NBC, ABC, CBS, Fox, CNN, and Bloomberg — released between September 17 and October 21.

Because of basic math six candidates (Trump, Carson, Fiorina, Bush, Rubio and Cruz) already have a high enough average that they could get 0’s on the next 5 polls and still qualify.

But that brings up another question. How many more polls will there be? 

By looking up the frequency of national polls by these organizations, it is virtually certain that neither Bloomberg, ABC nor CBS will do another national poll before October 21. ABC finished their 2nd poll this year on Sep 10 (their previous poll ended Jul 19), CBS finished their 2nd poll on Sep 13 (their previous poll ended Aug 2) and Bloomberg finished their 2nd poll on Sep 21 (their previous poll ended Aug 2). NBC, as parent company of CNBC “might” be tempted to rush out a poll just 3 weeks after their last poll on Sep 24 (their polls always end on Thursdays and they average a poll once every 7 weeks), but doing one so quickly might seem to be less objective-minded, and raise credibility issues to make such a radical departure from their previous pattern of polling every 6 to 8 weeks.

That leaves just CNN and Fox, who have been averaging polls every 5 and 4 weeks respectively. Fox’s last poll ended Sep 22, and is the chief sponsor of the GOP, so it seems very likely they will release another poll on or just before the deadline. 

CNN likes to end polls on Sundays and Oct 18 is the last Sunday before the deadline. It means slightly rushing a new poll since their last one ending on Sep 19, but it seems likely they will do one.

So most likely situation is that there are only TWO MORE polls.

What does this mean for the remaining candidates? Of the candidates at the last "kiddie table" debate, the highest polling averages (using the CNBC criteria), at just 0.5%, belong to Santorum and Jindal. They would have to average a whopping 6.5% to get to the debate. Neither has ever polled that high on ANY national poll this year.

This leaves 4 candidates currently averaging between 2.75 and 4%. Two of them (Kasich and Christie) could get goose eggs on the next two polls and still qualify for the debate. Huckabee just needs to get a 1% on just one of the next two polls to qualify.

Seem like a lock. But is it?

While it is a highly likely Huckabee will bottom out before reaching 0% on Oct 21, but since both the CNN and Fox polls are likely to come out just before the Oct 21 deadline and they only report 0 or 1% (not 0.5%), it isn’t unreasonable to see zeros. 

But there is one sign of hope for Huckabee. CNN’s last poll had Huckabee at 6% (vs. 3.2% on their trendline at the time). Perhaps something about their polling method gives Huckabee a bump. So maybe that means CNN will be down to 3% near the Oct 21 deadline meaning Huckabee would make the cut even if NBC also came out and gave him a 0.

That leaves one candidate, Rand Paul. 

If there are one or two polls before the deadline, as long as the new results average 2%, he is in. Looking at the forecasted trends above based on all national polls, he should get by with 3%. 

What about house affects? The last time Fox and CNN polled about Rand Paul he got 2 and 4% respectively, about 10% higher than the 2.7% average he was trending on for all polls. So he seems like he will just get in. 

It gets a little more dicey if NBC also does a poll because with three polls he needs to get a total of 7 points so there is not much room to spare on the forecast. 

If something important and negative gets a lot of coverage in the next month (the trial involving his father, sister and campaign strategist comes to mind), that might be enough to drop .5 % in one poll, which rounds to 1% and he just misses the cut.

Does CNN's polling methods save Mike Huckabee? Will Rand Paul avoid negative, poll depressing news? We'll know in a few weeks.

- JW

Hardball Most Competitive since Obama '08 Election

On August 26, 2013 MSNBC made an adjustment to its evening lineup, moving Ed Schultz into Chris Matthews' live showing of Hardball at 5 PM and turning his repeat broadcast at 7 PM into his live showing.

MSNBC President Phil Griffen said:

"This move will concentrate the 'Hardball' audience to one key time period and enhance the flow of our evening programming"

The Huffington Post's media critic Jack Mirkinson said

"The move could also be something of an admission about the continuing struggles of Schultz's 8 PM replacement, Chris Hayes. His wonky show has failed to match the ratings the more blustery Schultz put up in the time slot."

Politico's media critic Dylan Byers assessment was:

"MSNBC's evening lineup will now return to a mishmash of two different kinds of liberals who, though united behind a progressive agenda, do not always see eye to eye on aesthetics."

A common problem with critics and pundits is that they have no training in analyzing data. They will tell you the ratings (number of viewers) were up (or down) vs. yesterday, or last month, or last quarter, or last year and give you no context or explanation as to why that particular comparison is relevant (or not).

The following graph shows that Phil Griffen made a GREAT move when he shifted the lineup. The 7 PM hour is now more competitive than it has been in 5 years (just after Obama was elected and also when Obama beat Clinton in the primary). The 8 PM hour while not doing nearly so well as in Olbermann's heydays is as well now as it has in the last 4+ years not counting the runup to the 2012 election. (ALL MSNBC shows did MUCH better than normal in the month or two before and after the '08 and '12 elections)

Now, if you are an accountant, you will want to know how the ratings change on a show because that will affect your revenue, but media critics/pundits aren't (usually) accountants. They want to tell you who is hot or not.

The world of cable news is mostly driven by what is happening today. Ratings go up or down for a show vs. the prior day because of whatever events are happening that day, whether there is a substitute host, which day of the week it is, etc. In a larger scale, the "events happening that day" are more interesting to mostly politics-driven cable news shows when you are nearing election day in a presidential year, or a major primary, or a major speech and also major news events (e.g government shutdown, a hurricane, a bombing, etc). So to compare the ratings of one show does on a day where something major happened to another day where nothing significant happened is generally pointless.

What is more relevant is how did a show do against its major competition.

That is why the graph above compares how close MSNBC was to Fox (i.e. 120% means MSNBC did better than Fox, 80% means Fox won). Because day to day news events will make the trends hard to follow, I averaged entire weeks. In this case I chose an average of the previous 4 weeks to make the trends evident. (A single major event can cause a ratings spike triple of a typical day, sometimes even 7 times greater). I also excluded special shows (e.g. live coverage of the convention, or live election results, or the president's State of the Union address, etc). I did not try to take into account when the normal host was not broadcasting that day only because I didn't have that data readily available. I am guessing the the vacations hosts take are roughly the same and often at the same time of year so that this is likely to have insignificant impact in a 4 week average.

I also chose to use the ratings in the key Demo (age 25-54) not only because advertisers (revenue) care mostly about this view age range, but because it is pointless for any cable news network to try to compete with the millions of retirees who are dedicated to Fox. The overwhelming majority of Fox viewers are over 65. That has always been true and the addition of Megan Kelly to their prime time lineup has not changed that.

Finally, ratings focused media critics should maybe give Chris Hayes a break as they look at these comparative numbers. Chris Hayes is now doing much better than Lawrence O'Donnell (who hosted at 8 PM between Jan 24, 2011 through Oct 21, 2011) and Ed Schultz (who hosted from Oct 24, 2011 to March 25, 2013) with the exception of the big ratings bump that every MSNBC show got 2 months before and after the 2012 presidential election and even better than Keith Olbermann did in his final year and a half (July 2009 to Jan 2011). While my data doesn't go before Dec 2005, I would bet (excluding the time around the 2004 election) that Chris Hayes current ratings vs. Fox are better than Keith Olbermann's were during his first 3.5 years of Countdown (Mar 2003- Oct 2006) or his prior 2 years at MSNBC ('97 to '98).

So give Chris Hayes some more time. As Fox's viewers continue to die off, there may yet again be a point where MSNBC is regularly competitive with Fox's O'Reilly.