As Hillary Clinton continues her book tour to relatively positive reviews, it has led to a new round of cover-your-ass (CYA) and finger-pointing from every facet of the campaign world.
From Nate Silver blaming the media for not correctly interpreting what he knew all along to Democratic pollster Stan Greenberg’s claim that he saw it all coming but never said anything publicly, everyone is pointing the finger at a culprit other than themselves.
But self-centered rehashes aside, the bigger point that is being missed is not whether James Comey’s letter had an effect (it did) or if Russian interference had an effect (data is still out, but this appears to be the case). The fact is, this race should never have been close to begin with. Clearly many missteps were made and, unfortunately for the critics, whichever scapegoat is your preference was not the nail in the coffin.
Most recently the criticism has focused on the Clinton campaign’s over reliance on analytics while foregoing more traditional polling and focus groups. But anyone who’s arguing that the Clinton campaign would have won is they had put more emphasis on polling rather than analytics is completely missing the point.
First of all, it is clear that the Clinton campaign generally seemed more interested in running up the score than ensuring they were going to win the game. The money they were putting into Arizona, North Carolina, Indiana, Missouri, Nebraska’s 2nd, pale in comparison to what they were putting into Wisconsin and Michigan. At the end of the day, all they had to do was win two of the three Rust Belt states that Obama won in 2012 (PA, WI, MI) and they didn’t set themselves up to do this.
The prime example of this is Wisconsin, where the candidate barely stepped foot and the campaign barely spent any money on TV. It should not have been a surprise then that Clinton significantly underperformed Obama in 2012.
This brings us to the second point in the polling versus analytics argument — it’s not an either-or choice. The example above is not to do a victory lap, the point is that we were picking up different things at different times with different types of research. This should be the lesson of 2016, when it comes to research: There is not one perfect way to understand what is happening in the electorate.
Any campaign that is arguing digital instead of broadcast, or vice versa, is unlikely to be a successful campaign. The same is now true for research. Anyone arguing polling over analytics, or analytics over polling, is not only missing the point, but is setting up for a failed campaign.
The public’s ability to get information from different sources has increased tenfold over the last decade, as has the ways in which people can express their dissatisfaction, or support. With this in mind, it’s very likely that one mode of research, or at a minimum the mode that you pick, is not going to pick up everything and a campaign needs to understand what it truly happening.
The divergence of bullhorns voters have at their disposal requires the ability to have a wide arrange of tools to not only listen, but also understand what you are hearing. There’s going to be, or at least there should be, a place for phone polls, online research, social listening, analytics, and so on. While this isn’t the cheeriest news for campaigns, since it’s going to cost more money, anyone who is relying on one mode of research is setting themselves up for failure.
A well-rounded research plan is an important first step, but taking both polling and analytics into account is not enough. Campaigns need to take diverging opinions into account as well. While we cannot definitively speak for the Republican side (but judging from 2012 they are not immune), it’s easier on the Democratic side to be wrong with the crowd than express an idea that is outside the mainstream.
When the prevailing wisdom was Clinton was going to win Wisconsin, Florida, and Pennsylvania, the system is not set up to listen to those whose evidence deviates. This creates a situation where people are making the smartest decision for their business’ future rather than putting themselves out there with numbers that differ.
No one likes being the person to give candidates or clients bad news, but campaigns are going to be much better served if the questions are not just focused on how we win but also on how we can lose.
Exacerbating this problem on the Democratic side is the tendency to concentrate a large amount of work in a small number of firms. With the same firms advising most of the races across the county, up and down the ticket, a system is established where one mode of thinking is prevalent and no one can question that thought process or the orthodoxy, lest they bury their future chances of being one of the chosen few.
Campaigns must create a system where diverging opinions are encouraged and listened to — you don’t want an echo chamber. If someone at the table is not talking about how you are going to lose, you should rethink who is at the table.
And finally, understand that if you hire a large firm that is working on dozens of races in a given cycle, the likelihood of you getting unique advice diminishes quickly.
As our world continues to change, and technologies advance, it would be nice to think that campaigns are getting easier. Unfortunately, that isn’t the case. This does not mean you need a presidential budget to run a local campaign. Just be wary of the easy answers. They’re more likely a mirage than based in reality.
Stefan Hankin is founder and president of Lincoln Park Strategies, a Washington D.C.-based public opinion firm. Follow him on Twitter at @LPStrategies.