Friday, January 29, 2016

Trumping the conventional wisdom on ground game?

Effective "ground game" is considered crucial for presidential candidates to find success in state caucuses. A recent Monmouth poll included a question that serves as a clean, quantitative measure of this rather nebulous term. It asked likely caucus voters if they had been personally asked to vote for a candidate (through neighborhood canvassing, being contacted by a candidate's phone bank, etc).

In Iowa, the percentages of likely voters who reported having been contacted by each candidate's campaign and that candidate's Real Clear Politics' polling support averages:


It's immediately apparent that ground game and polling performance are strongly related. Indeed, they correlate at .79 (p-value = .002). Vigorous though that relationship is, once trend-bucking Trump is removed, the correlation increases to a nearly perfect .98 (p-value = .00000001). In other words, ground game--quaint though it may seem in a country of 330 million where mass communication is free, instantaneous, and almost infinitely scalable--appears to be deserving of the attention it receives.

The caucus process is tediously time-consuming. Consequently, turnout is much smaller than in comparable primaries, where the process mirrors that of general election voting. In 2012, just over 120,000 people--a mere 5% of Iowa's adult population--participated in the Republican caucuses. Cruz has been working the state for years. Historically, that pays dividends on caucus day.

But Trump has overcome seemingly insurmountable odds from the get-go. While Cruz and Rubio have pestered more individual Iowans than Trump has, Trump's message and mannerisms resonate with middle American radicals more viscerally than anything that comes out of the mouths of the political puppets.

The prediction markets for Iowa now have Trump at 70% and Cruz at 29%, suggesting that the smart money is on Trump to defy yet another allegedly ironclad rule of presidential campaigning. Cruz needs the snow expected in Iowa on Tuesday to turn out to be a blizzard that arrives a day earlier than the meteorologists are predicting it will. Des Moines' five-day forecast:


The Z Blog said...

I think I'd like to see the data from previous events to see if it holds up. We can look back at polling and see that the late polls tended to be spot on. The most obvious analogy here is Howard Dean's experience in 2004. Mythology says he was leading in the polls and finished last because he had no ground game. Then cue story of empty Dean buses driving around Iowa.

Reality is he started to slide in the polls much earlier and he ended up third. John Kerry started trending up in December and over performed his polling by ten points. Dean and Edwards slightly under performed their polling. Dean was at 30% in January, but the last bit of polling had him in the low-20's.

A sensible way to look at that is as people got serious in the final weeks, they swung to Kerry and abandoned Dean. The DMR poll today will be the final poll before the caucus. If you see weakening in Trump's numbers, then you better be prepared for a fifth place finish. If he is steady, then this one data set linking ground game to vote share can probably be dismissed.

Entity said...

It's interesting, though predictable, that Trump is the huge outlier here. I'm guessing Trump is a huge outlier in a lot of measures. His poll numbers haven't collapsed after some gaffe or revelation, like so many candidates-of-the-month in 2012. He's been bulletproof.

Audacious Epigone said...

Dug up a poll from 2012 with similar questions here (couldn't find any from 2008). Most to least contact in month leading up to vote:

Ron Paul
Mitt Romney
Newt Gingrich
Michelle Bachmann
Rick Santorum

And Rick Santorum ended up winning, playing the potential Trump role here, while support for the others pretty closely followed the ground game. One thing that stands out is how much vastly higher these contact numbers from 2012 are than those found in the Monmouth poll I linked to. Even Santorum, who had contacted the fewest of the five candidates asked about in 2012, contacted more likely voters (37%) than the Monmouth report shows Cruz's campaign having contacted (25%). Both these polls were conducted about a week out from the actual caucus, so that's not the explanation. Any idea?

Maybe ground game is always hyped but never really that important? I've not paid close attention to it in the past. If so, we could be looking at another data point in favor of it being over-billed.

The Z Blog said...

The thing that always pops into my head when thinking about this stuff is my own experience working for a Congressman back in the 80's and working for candidates in the 1990's. Back then, you had real phone banks with volunteers making calls and reading scripts. The person riding herd on them was a paid employee.

Then you had paid canvasing. I would get a fleet of vans full of volunteers, a cargo van with yard signs, stickers, caps, t-shirts and so on. I would sweep a neighborhood, knocking on doors, getting names, phone numbers and placing signs in yards. Often, the candidate would call these people back the next day and thank them for their support and maybe ask for money.

That's not how it is done now. Telemarketing is mostly useless in a mobile phone world where we text more than talk. Instead, campaigns have social media, email and robo calling where they leave voice mails because no one answers the phone. Door-to-door canvasing also seems to have died off.

The point here is the definition of "contact" has been quickly changing in the last decade. Maybe the drop from 2012 is due to a) the number of well funded campaigns and b) a decline in the use of traditional marketing tools. People are not counting an e-mail or FB ad as contact so they don't respond positively to the question.

Audacious Epigone said...


Social media has undeniably been a huge presence in the 2016 race. I almost always first hear about Trump's latest putative gaffe/outrage by his response twitter/facebook response to the media hissy fit. Getting around the gatekeepers has never been easier.

Still, I'd expect the trend away from the quaint door-to-door/phonebank stuff to be gentler than it appears to have been. A week out in 2012, 72% of likely voters said they'd been contacted by the Ron Paul campaign. A week out in 2016, and just 25% of likely voters say they've been contacted by the Ted Cruz campaign at all during the run up.

The Z Blog said...

The consultancy is a small club so trends rocket around the industry. One of the "lessons" of the Obama win in 2008 and 2012 is the use of data analytics to narrow the focus of GOTV efforts. I'm thinking that maybe this is the cycle the Stupid Party decided to hop on board. The test of that would be the same figures for the Democrats.

It's an interesting thing. The DMR poll is out and unless they are historically wrong and I mean wildy wrong, we'll have another data point suggesting ground game is not terribly important.

Audacious Epigone said...


The DMR poll is the one Trump has consistently fared the worst in, too. He's higher than he ever has been in one of these DMR polls, so he'd seem to be in a good spot.

The Z Blog said...

Sleeping on it, the change ground game versus vote share could simply be the result of the financing of campaigns. In the olden thymes, you could not afford and army of paid staff. You had to rely on volunteers and that meant getting locals organizations, the party clubs and local money-men involved.

Today, everyone brings in their hired guns who know all the local hired guns. They craft the social media campaign, the TV campaign and the commentariat campaign. Just look at how many chattering skulls either work for a candidate or used to work for one. Mark Levin, for example, has family working for Cruz.

What Trump is proving is that a great candidate with a popular message does not need all the crutches and props to keep his campaign afloat.

Anonymous said...

Audacious Epigone,

Trump's campaign operating style is outside the frame of reference of most media political reporters, all the media editors and, it seems, most of the political consultant class who talks to the American & International press.

Trump is approaching winning the Presidency with the same due diligence that billionaire Jeff Bezos is with his Blue Origin's reusable Space Vehicle or Elon Musk is with his Space-X reusable launchers.

You have to look at Trump in light of billionaire capabilities and not the political consultant organizational template.

In this, political consultants and political campaigns remind me very strongly of oil field "job shop" firms. They throw together whatever resources at hand to achieve a project with a final sprint against a set deadline. They build it and walk away until the next project.

Trump is much more into continuity of operations. Trump does a deal, builds a project, but he puts his name on it and _runs_ it for the long term afterwards.

You have to possess much better management and people selection/training skills set for the successful management of properties afterward. This is why entrepreneurial empires like Trump's are so rare.

This is showing up in Trump's campaign.

Mil-Tech Bard said...

Audacious Epigone,

Next rock -

It is really hard to find any useful tells about Trump's ground campaign in Iowa, There was one article back in October 2015 but that's all. This is one of the reasons I hang out at "The Conservative Treehouse" web site ( It amounts to a boiler house operation for Trump volunteers and wanna be volunteers.

These folks -are not- on the inside, but they are smart, observant and have pretty good bovine waste detectors.

The following three points have shown up repeatedly there about Trump's operation in Iowa:

1. Trump's people are not talking to the media or other campaign operatives and they are running a ground game using recreational vehicles without most of the traditional brick and mortar store fronts to limit derived intelligence -- AKA the other side having people watch and dumpster dive your campaign HQ.

2. The Media knows something about the extent of Trump's ground game, but they are not talking. It is not "the editor approved narrative."

3. Trump's people in Iowa are doing the same sort of "build the relationship with future voters" contact outreach Obama's people did in 2007-08 and 2011-12.

Please note several things about the above points.

There are no reports of RV/vehicle rentals in the Trump campaign financial releases. Yet TCTH folks talk about them a lot. This could be pre-paid services or simply poor media reporting. Either way, Trump is acting to limit what his opponents know about him.

A lot of the 'relationship building steps' in Iowa look a hell of a lot like typical high end hotel "Remember the important dates of repeat VIP business customers" SOP's in action. This high end hotel Standard Operating Procedure is relational electronic database driven.

Data mining 30-years of hotel guest data world wide tells you a great deal about people, if you are willing to invest the time/effort to do so. Especially if you used outside consumer data to fill in the blanks.

Mil-Tech Bard said...

Audacious Epigone,

Last Rock

And even more important than his Open Source Intelligence/Big Data capability, Trump has a very well staffed, high end, _Counter-Intelligence_ operation at work surrounding his campaign operations.

Trump made a Million dollars a BILLION dollars by the time he was he was 42 in 1988. He never settled. He in the last almost 30 years has moved that billion dollar portfolio to something like $10 billion world wide.

Trump needed the counter-intelligence side more than intelligence to get where he has gotten since 1988, if only to prevent his 5-star hotel and especially _casino_ properties from being hollowed out by local politics, corruption and crime. And gambling brings out crime and corruption like bovine waste brings fly's.

Of his billionaire capabilities, this private Counter-Intelligence service Trump built up with his casino properties is the single greatest threat to the GOPe and Clinton campaign operations.

The Media and Media talking political consultant class simply cannot get inside Trump's head and his important campaign people cannot be turned/intimidated, because of 30-years of high end C-I vetting built into Trump operations.

It costs a new political campaign a lot of money to do the kind of VIP hotel outreach Trump is doing with his new political contacts. Political candidates & consultants have to build this capability anew every time they run a campaign.

Trump does it for very little compared to other campaigns because he is using a very small percentage of existing innate capability from the Trump corporate IT "Big Data" capabilities. The use of which was very likely pre-paid in 2014 to be outside the Federal campaign finance reporting rules for the 2015-2016 period.

I suspect some of the high end American political consultant class has this figured out, but they don't talk without someone paying them a lot and those who pay a lot don't talk to the press until long after the campaign is over.

And frankly, a lot of the fat cats paying these high end political consultants won't listen to the truths about Trump until after Trump wins.

Since these high end consultants know when/how fat cats shoot the messenger -- it is how they got there in the first place -- the full truth won't be provided to the fact cats until they are ready to listen.

Mil-Tech Bard said...

Via TCTH, the Trump campaign is intentionally not having its new voters register a Republicans prior to the Iowa caucus.

This is the Trump counter-intelligence operating style at work.


This is brilliant and why there has not been a surge in registrations.

There’s no metric you can really apply though to figure that out. The Trump campaign is not telling people to go change your registration. This is where Chuck Laudner is really savvy. What they’re telling their people, they’re giving them voter registration forms, telling them to fill it out, register as Republican and take it to your caucus night because the little trick is the state party had to print its list long ago. So you run the risk of really frustrating the voter if they change their voter registration to participate and then they show up but they say ‘Hey, you’re not on the list.’ So this is where Laudner, being a former executive director of the Republican Party who knows the ins and outs of all this stuff, just says ‘Hey, we’re just going to arm our people with voter registration forms. We’re not going to rely on the Republican Party to have enough forms. We’ll just take them there that night and register them because then there’s no little problems that could upset our supporters.’

Audacious Epigone said...


Do you think now then that the presumed causality is backwards? Do calls/canvassing merely reflect support rather than influencing it?

Trump existing as an outlier here wouldn't seem to change the feasibility of that being the case, since a lot of his support comes from people who haven't been involved in these sorts of things in the past.


How to square starting construction late with building it to last? How important is that durability in Iowa? Or do you mean later down the stretch the dividends will be paid out more from Trump's organizational approach than from the typical consultant template?

Mil-Tech Bard,

That is encouraging. I do wonder why it doesn't appear to be captured in polling data, though. Allegedly, expected first-time voter rates are going to be almost identical to what they were in 2012.

Relatedly, any coaching supporters regarding how to respond if participating in a poll? Thinking exclusively about the marginal voter who may or may not participate but who if he does will vote for the candidate we're considering, does said candidate benefit more among marginal voters if polls show him at 40% with the next guy at 30%, or would he rather the poll show the other guy at 32% and him at 30%? Is that too clever by half for a campaign to try its hand at manipulating?

Mil-Tech Bard said...

Audacious Epigone,

"Anon" above was my first rock of three.

Please note that even Politico admits Trump has been building up his campaign team for years in _complete secrecy_.

But Politico still don't realize the implications of what they have reported.



That the media saturation was driven by controversies that led to widespread condemnation of Trump in the mainstream press and by pillars of the Republican Party didn’t matter to Trump. His team knew that the more Republican voters got to see of Trump the politician, the more they liked him.

They knew because they’d studied the polls.

When Trump first surged to the top of the Republican pack just weeks after his announcement, pundits pointed to Trump’s terrible favorability numbers to argue that his support had a hard ceiling and that other candidates would overtake him when the sprawling 17-member Republican field thinned. Indeed, Trump walked into the contest with a poll showing only 16 percent of Republicans viewed him favorably while 65 percent viewed him unfavorably.

But when Trump touched off controversy, his favorability skyrocketed, and by mid-July that same ABC/Washington Post poll found that 57 percent of Republicans viewed him favorably.

If you don’t have the same risk tolerance, he’s got you beat before you’re even in the game.”

Trump doesn't take a Sh*t without a plan.

When is a perceived political risk not a risk?

When you know more about the voters than other politicians.

The question Politico left unasked is how did Trump know more.

See Link below

Read more:

Mil-Tech Bard said...

>>That is encouraging. I do wonder why it doesn't appear to be captured
>>in polling data, though. Allegedly, expected first-time voter rates
>>are going to be almost identical to what they were in 2012.

It is called the "Bradley Effect."

It takes an act of courage for a cucked white to publicly support Trump.

Part of what Trump is doing is a tribal in-group versus the world campaign strategy.

This Neo-Jacksonian uprising was planned.

It's success was based on Trump's non-uniparty perceptual filters seeing what was in the polling data all along.

I'll give you useful polling example from 2010 and combine it with a pair of current events.

The San Berdoo ISIS terrorist attack toggled over the Jacksonian's this election cycle, an overwhelming majority of whom are Tea Party or Tea Party supporting.

See this John Zogby 2010 polling on the Jacksonian nature of the Tea Party, particularly the parts on how terrorism moved Sen. Scott Brown’s voters in 2010.

Boiling Tea
John Zogby

"Contrary to the often repeated claim that Tea Partiers ‘lack agreed upon set of views,’ our data shows that terrorism and perceived unwillingness to talk about it in a straightforward manner might be another issue around which opposition to Washington will rally. Ironically, shifting their attention from health care might make Tea Partiers angrier. “

AKA Tea Party = Walter Russel Mead described "Jacksonian", and San Berdoo is what crystalized the Tea Party around Trump in 2015.

Trump's adding Sarah Palin's endorsement the week before the Iowa caucus must be seen in that light of that knowledge.

Next Rock

Mil-Tech Bard said...

The following was written shortly after the Zogby polling in 2010 and captures the essence of why Trump's statements after Paris and San Berdoo cemented the Tea Party to Trump come hell and high water.

May 06, 2010
Obama Stands with Muslims as He Promised
By Kyle-Anne Shiver

Carefully protecting feelings can be a priority of a mommy tending her babies. Protecting a group's feelings at the expense of a nation's real safety is a huge no-no for a Commander in Chief. And if Barack Obama would prefer to be the nurturing mommy to all Muslims, he ought to resign from the presidency.

Trying to be both a mommy to Muslim sensibilities and to perform adequately as our president is a losing formula for every real peace-loving American citizen.

It's a terrible thing when Americans have to face the day not only needing to know where their own children are, but also asking the haunting question: Do you know where your Muslim neighbor is, and what he's up to?

We have to ask that question now because our president and his people will not.

So in the end, Barack Obama has brought about the exact scenario he has worked so hard to avoid. If Americans were assured that our president and our national security employees are doing the necessary watching and profiling, then we would also know we don't have to do it. We could be nice, hospitable, and open to our American Muslim neighbors and coworkers.

But when we know that the people paid to fight terrorism refuse to see the obvious, then we are necessarily put on high alert. We take on their jobs. We watch. We stare. We shy from the company of those we know might become our worst nightmare.

Trump's Calling out American Muslims in Patterson NJ for supporting 9/11/2001 attacks after the Paris Terrorist Attack and his calling for a Muslim Immigration Pause after San Berdoo attack were not impromptu risks.

There were carefully planned, weighed, polled and staffed by Trump between the Charles Hebdo attack and the Paris Attack.

He could do those "risky statements" because he paid attention to what the polls actually said.

Not what Political consultants thought the Uniparty political elites and the donor class billionaire "whales" paying the freight were ready to hear.

Audacious Epigone said...

Mil-Tech Bard,

Ah, I see. Pardon the oversight.

Fascinating, thanks.

pithom said...

Well, Cruz won, with Rubio a close second. I guess ground game really is important, and Trump is not immune to ignoring it.

Audacious Epigone said...

Trump was second, Rubio was third. Yes, it looks like it is important at least in states with small populations and caucuses. Fortunately, the next two are primaries, which should play to Trump's strengths--ie actual popularity among the electorate--more than holy roller Iowa did.