It’s been a roller coaster of a week in the U.S., with an election on Tuesday that we figured would be a nail biter. A lot of things came into play — pollsters getting it wrong, social media insanity and propaganda were commonplace, and the importance of data and analytics for projections has never been more important. On this episode of the Futurum Tech Webcast, I’m joined by my colleagues and fellow analysts Olivier Blanchard and Ron Westfall as we take a look at what went down, what we saw from an analysts’ point of view as well as a look at the impact of the election on the market this past week.
A Quick Look at Election Results and Why We Wait
We started with a look at where we are, which is, as predicted, in a state of limbo, waiting for results. We discussed the fact that one of the key things that has impacted this election is the fact that different states have different voting laws. Add to that the “small detail” of the coronavirus pandemic, and people being understandably nervous about getting out to vote in person, and the tsunami that we’re currently wading through in terms of getting votes counted is totally understandable.
Journalists and news organizations have warned us for weeks that it would likely take days to figure out the winners, but of course they didn’t always follow their own advice on election night and in the days to follow. Everybody loves making predictions.
Voting laws are complicated. State officials said that even unofficial totals in states like Michigan, Wisconsin, and PA wouldn’t be available until later in the week.
We took a quick look at some of the differences in key states where we’re waiting on results, including:
Pennsylvania, which doesn’t start processing absentee ballots until Election Day morning. There was never any way all those votes would be counted until Friday (today) at the earliest. For anyone frustrated about this, note that this is the result of deadlocked negotiations between the Democratic Governor Tom Wolf and Republicans who control the Senate.
Arizona was another state that’s complicated. The state generally has high turnout, and a robust history for voting early that often leads to late results. This is not new. In the 2018 general election it took six days before Kyrsten Sinema was declared the winner of the U.S. Senate race. In 2010, it took voters 12 days to learn they had legalized medical marijuana. This year, an estimated 88% of the primary vote in August came by mail, and in Maricopa County alone, almost 2.1 million people asked for a mail-in ballot out of 2.6 million registered voters. But here’s the thing. Mail-in ballots aren’t always mailed in, and they are not always early. Some voters wait until Election Day and take their mail in ballots to the polls, to be sure they get there. They are largely set aside until Election Day votes that are made in person are taken care of.
What’s happening in Georgia? At about Noon Central Time today, Georgia reported that thousands of military, overseas, and provisional ballots that will soon be added to Georgia’s vote count, and that roughly 8,100 absentee ballots were still pending.
And last but never least, North Carolina is waiting to see if 116,000 outstanding requested absentee ballots are returned by November 12th. It’s a waiting game, in so many places. And again, that’s nothing new.
Election Security and Voter Trust in the System. Our conversation then turned to election security.
We explored the fact that voting machine and voter database security is an issue, and why that’s a reality during election seasons.
We had instances of databases being accessed and voters receiving threating letters in the mail if they voted a certain way.
We explored an example of a voter intimidation campaign that targeted voters in Alaska and Florida.
And we covered the fact that Federal authorities were somewhat optimistic that we we had managed to make it through the voting season without a major cyberattack. They pointed to huge turnout by Americans that showed confidence in the security of their votes. It’s reported that the American operatives went on the offense, “hunting forward” in an effort to surveil the work of Russia, Iranian, and other cyber-operatives, identifying targets selected in the U.S. and working to bolster cyber-defenses at home.
USPS. We touched on the challenges the USPS faced, which definitely played a role in this election. The concern about the Postal Service’s ability to handle the massive increase in mail volume also led to so many people voting absentee in advance, and also taking their ballots in person to drop offs wherever possible.
Let’s talk about data. Our conversation turned to data, where we explored
- Data’s impact on polling
- Data’s impact on messaging
- Data’s impact on targeting
- Data’s impact on vote count projections
We explored how AI models underperformed in the 2020 US election where we saw firms like KCore Analytics, Expert.AI, and Advanced Symbolics claim algorithms can capture a more expansive picture of election dynamics because they draw on signals like tweets and Facebook messages. But there are dangers associated with that and overall, it’s still very unclear whether AI proved more or less accurate than the polls.
Let’s talk about the market. Finally, what’s up with the market? Why are the key U.S. market indexes up? While it might seem counterintuitive, US equities have performed the best when political power in Washington is split between Democrats and Republicans.
In summary, it’s been an interesting week. Check out the webcast here:
or grab the audio here:
As always, thanks for coming by and for spending time with the Futurum Team. If you have thoughts on any of these things as it relates to the US elections, specifically related to security, technology, data, and beyond, we’d love to hear them.
Disclaimer: The Futurum Tech Podcast is for information and entertainment purposes only. Over the course of this podcast, we may talk about companies that are publicly traded and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors and we do not ask that you treat us as such.
Shelly Kramer: Hello and welcome to this episode of the Futurum Tech Webcast. I’m your host Shelly Kramer and I’m joined today by my colleagues Olivier Blanchard and Ron Westfall. Hello, gentlemen, it’s great to see you. Before we get started, I want to give our disclaimer.
This show is intended for educational and informational purposes only. We are technology analysts. We have thoughts and opinions, which we share. We might, in the course of this show, mention publicly traded companies and please know, this is not intended as investment advice of any nature, it is intended for your listening and viewing enjoyment. With that, we’re going to get started on our show, which is really all about breaking down the election week in the United States, and taking a look at security and technology and data, and how all of those played a role and are playing a role in what it is that’s going on right now in these, I think maybe what could best be characterized as, tumultuous post-election days in the United States. Is that a fair assessment? Do you think?
Olivier Blanchard: Anxiety, yeah.
Shelly Kramer: Anxiety. Yeah, it’s been a roller coaster of a week. We had an election on Tuesday. We figured it would be a nail biter, and a lot of the things that were predicted actually came into play: pollsters getting it wrong, social media insanity, propaganda here and there. And the importance of data and analytics for projections, has really never been more important. I think one of the key things that I keep… I have 14 year old twins. They’re really interested in this election and they’re learning, so we’re having a lot of conversations about how things work and some of the questions that they’ve had for me is, what’s going on? Why don’t we know? With technology the way it is, why is this process so slow? Why don’t we know what the results in Pennsylvania are? What’s happening in Georgia? Why are these things changing?
I think one thing that I have even had to step back and remind myself, and also educate myself about, is that voting laws are different in every single state. What happens in Missouri, where I am, and in Texas, where you are Ron, and in South Carolina, where you are Olivier, there just all different. And part of those differences are also party driven, and Democrats and Republicans oftentimes can’t agree on things, and so we’ve seen lots of… I don’t… arm wrestling or tug of war or whatever just trying to get all these things sorted out.
One of the things that I thought was interesting to mention was that Pennsylvania, for instance, doesn’t start processing absentee ballots until election day morning. There was never any way for those votes to be counted early, but this is sort of deadlock is a result of an impasse that… deadlocked negotiations between the Democratic governor, Tom Wolf, and the Republicans who control the Senate. Wouldn’t it be nice to know what’s happening in Pennsylvania? Well, it would be, but we’re at the misery of the laws in these particular places.
Another thing that I thought was really interesting was Arizona, and that’s another state that we’re waiting on and it’s very complicated, and that state generally has a really high turnout, and a robust history of voting, but it often leads to late results. And so we’re sitting here frustrated, but this isn’t new. One of the things I thought was interesting as I was researching this, in 2018, in the general election, it took six days before Kyrsten Sinema was declared the winner of the US Senate race, and that was a really hot, contested race. I remember that. There was another thing that I noticed that took a really, really long time to happen, and I can’t remember what that one is. But anyway, what are your thoughts on what’s happening with the confusion around election laws and this delay in voting results? Any thoughts there that you guys have to share?
Olivier Blanchard: Ron can go first.
Ron Westfall: Oh, sure. In fact, it’s back to the future, we’ve been here before. 2000, as we all know, it took over a month for us to decided who would be the US President for the next four years, and we got to know a great deal about Florida election law-
Shelly Kramer: We did.
Ron Westfall: … and Florida procedures and so forth. Something that most people would never have any thought about, until that happened. What’s playing out here, is a parallel. We’re already almost one week past the election and we still don’t know who’s going to be the US President, officially.
Shelly Kramer: Is it a week? Is it a week, because it feels like a hundred years?
Ron Westfall: Almost a week, I know, exactly. This is so-
Olivier Blanchard: Feels like three years.
Shelly Kramer: Yes.
Ron Westfall: On par for 2020, timeframes have just been thrown out the window. It’s like, “When does the school year really start?” And so forth. But the point is it could even be Florida squared this year, hypothetically. We’re looking at more than one or two states being contested, either because of judicial review or because the state rules that, for example, automate recounts or accept a challenge if the results are within a very narrow percentage range, so stay tuned. We just have to wait for all these state results to be certified, and to see if there’s enough warrant for one of the parties or both of the parties to challenge and drag this out even further.
Shelly Kramer: I think the hard thing for me is I want to give everybody a Xanax or seven, I don’t know, because it’s… Again, for instance, North Carolina, they’re waiting on over a 100,000 absentee ballots that may or may not have been returned, and they have until November 12th. Well… So it’s kind of like take a deep breath, this isn’t the first time this has happened, calm… tone down the rhetoric, tone down the accusations. I will say one thing that has really been of concern is just concerns about safety, at these polling places where votes are being counted. These people, in many instances… Probably not Nevada, because Nevada is, I think, the universal joke about slowness. These people are heroes, they are working around the clock in many instances, and it’s under so much pressure, and I think…
The other thing that I was sharing with my kids yesterday is that I didn’t really know for sure, but states have a really vested interest in getting this right. Republicans are involved and Democrats are involved and election supervisors are involved, and it’s like, no state wants to be the laughingstock or the suspect one. And so I think that really understanding that there are legitimate, respectful, well-regarded people who are invested in this on both sides of the aisle, and just calming down a little bit and trying to be patience, I think that we need to remind ourselves how important that is. I know that that’s wasted words. Those are wasted words.
Olivier Blanchard: We live in the… It’s kind of like the era of, it’s DoorDash, right? We get on our phone, we order our food, it shows up 20 minutes later, and we’ve kind of become a little bit expecting of things moving faster just because technology has taken over so much or lives, made things easier and faster. But the electoral process in the United States, for better or for worse, but I think for better mostly, is still a very human, labor intensive process, which ensures that even if there is a little bit of fraud, you can have one person who’s going in there and wreaking havoc, but just because of the mass amount of people touching ballots and recording ballots, it’s not going to have a huge effect.
The moment that you start automating that, even if you think that you have fail-safes, you create the potential for danger, and so in a way, it would be much easier for us to be able to just pick up our phone and vote. You a phone to vote, which I think we actually have adequate security to be able to do this, but state laws are state laws, they err on the side of caution, if we had this. On the one hand you have all of these ballots that tend to come in on the same say, then you also have the mail-in ballots, the provisional ballots, which have a certain amount, usually three days for most states, that allow them to come in. But then you also have the overseas vote, especially the military vote. I think it’s really important to remember that in a state like Georgia, which might be within point one or point two percent between the two candidates, the military vote, a week from now, might still be able to make a difference.
I think the expectation of things happening very quickly, because of the technological world we live in, isn’t super aligned always with this every two years or every four year process of realizing that the electoral count model is not keeping up, for good reasons, with our mobile and very app driven world. And so it’s an expectation issue more than anything.
Shelly Kramer: I think it is interesting, and I hadn’t thought about that human touch that you bring up, and I do think that for a lot of people that’s a good thing, and I’m thinking about myself. For instance, I don’t ever vote electronically, unless I have to. I had to in this election, but I always get a paper ballot. I like the old school paper ballot, feeling like… and I trust it more, and I just think that knowing that real human beings are laboring over this process and are looking… When you’re talking about ballot curing, when you’re talking about matching signatures… I know people who had their ballots rejected because the signature didn’t match and the process they… I have a couple of friends that, as a result, had to get a provisional ballot, and were very upset about that, because it’s a small little thing or I can’t remember, some name didn’t match or something. Close personal friends who had issues like that, who were upset that they had to have provisional ballots.
To your point about Georgia, I read this morning that there are 8100 absentee ballots that are still pending and they’re expecting thousands of military, overseas, and provisional ballots, and you’ve got to get through those. To talk about election security and voter trust, my opinion is that having that human element, actually makes people trust the system more. I think another thing that happened in this election that also has caused some backup, is that we had concerns about the postal service and their ability to get mail ballot delivered on time. And so I think that what happened in a lot of instances, where people who might have otherwise done mail-in ballots, got their mail-in ballots, but then held on to them and delivered them on election day or they delivered them a couple days before election day. People didn’t seem to want to trust all the normal systems this particular year. I think that that also leads to a delay.
Let’s talk a little bit about election security in general. That was a big factor in the 2016 election in a number of ways, and for the past four years that’s been a topic of conversation. Did it happen? Didn’t it happen? Whatever, I’m not going to go there. But we do know that there were some cybersecurity issues with this election, do you want to talk about some those Ron? You’ve shown up in a suit, and I just want to be sure that we maximize your exposure because you dressed up today, and Olivier and I, we look like the shlubs that we are. Anyway, Ron do you have any thoughts.
Olivier Blanchard: We look like elites.
Ron Westfall: Right on, it’s my security suit. It’s an excellent issue. I think there are a couple of aspects here. This election cycle is unique because of COVID-19 and the pandemic, and so that did usher in all of these new voting capabilities that really hadn’t been in place before, because there was genuine concern that going to public polling places would spread the virus. That opened up the options for more mail-in voting, for more absentee ballots, et cetera. Yeah, it does present a distinct challenge this election cycle. What will be interesting is, say, in the mid-term elections or the next presidential election, will these same provisions still be in place? Because some states discovered they liked this increased flexibility, or some will dial them back.
Now, going to the issue of fundamental security. Yes, there was speculation, for example, of international intervention. But to Olivier’s point, the voting system is so heavily distributed that it’s really hard, say, theoretically, if a hacker got in and manipulated one county, that would alter the outcome of something like a national presidential election, let only a state level election. The odds of that are very insignificant, although, it’s theoretically possible. I think I’m going to dust off one term that hasn’t really been used recently but could be well-suited for election cycles because of their predictability, it’s going happen, at least in the US, every four years, and every two years for various elections, and that the technology of blockchain can actually address many of these security issues. I think that needs to be revisited after this election cycle.
If you’re really serious about having assurance across the board for all these votes, a blockchain could actually be the answer for this, even though it might not be applicable for other needs. This, I think, is a takeaway, so I’m going to actually stick to my guns and talk about why blockchain needs to be reconsidered between now and the end of the year at least.
Olivier Blanchard: I think-
Shelly Kramer: That’s a short window to get that in play, between now… That’d run you.
Ron Westfall: Well, I could talk about it.
Shelly Kramer: Yeah, you can definitely talk about it. Olivier, let’s hear what you have to say.
Olivier Blanchard: I think the more manual, the more we stick to paper ballots, I think the more likely it is that you’re going to avoid, at least, external influence from hackers, from hostile agents, from foreign hackers, and that’s definitely one of the threats with electronic voting machines. On the one hand, we know that electronic voting machines are not plugged into the internet, while the vote is ongoing. We know this. It’s usually a talking point on TV when pundits talk about this, but what they don’t talk about is these are machines. They have operating systems that need to be maintained and upgraded. And so on the one hand, you have some machines that are running very old OSs and very old software that is definitely vulnerable when one of those machines becomes connected to something or becomes upgraded or there’s a software update, some kind of maintenance, and you connect them to a network, which does happen between elections.
Then that’s when the machine becomes vulnerable to some kind of hacking, some kind of manipulation, somebody could plant something is there or even just to collect data, there’s a lot of things that they could do. On the other hand, there’s another vulnerability that has nothing to do with machines and which has to do with databases of voter registration state by state, and even per county. Wherever those databases are, on the one hand, they can be accessed, collected, and then distributed somewhere. On the dark web, as people like to say, that information can get into the hands of hostile actors or consulting groups that will help group A, group B, or group C do some targeting or better understand where to focus their efforts, if they want to create interference or target very specific message for voter suppression or else in certain areas.
Also, it’s possible that those voter registration rolls could be altered just enough, by a very smart hacker, to disenfranchise a certain percentage of voters in a particular district. You could see how, if you wanted to lower the African American votes in a particular district or particular county by three percent, you could randomly select the right ratio of voters and make a slight correction about their address, their signature, some tiny little piece of data that’s going to create an inconsistency that will prevent them from voting effectively or as smoothly as they normally would. Those vulnerabilities are very small, but they do exist.
But to counter Ron’s point a little bit. While it’s true that on the whole a small impact in a county or a district during a national election seems insignificant, we also know that national presidential elections, in the US, are often decided by about a dozen counties in the United States. If you have access to that data and over the course of several elections you test and you probe and you get the right formula, you could get luck and be able to perform that kind of mission that would result in disenfranchising just a few thousand votes in very strategic areas that will affect the rest of the election. We have to be cautious because sometimes very small and very hard to detect can be extremely, extremely effective. All right…
Shelly Kramer: Extremely effective. I read when… Speaking about election security, I have some friends on the internet, no, I have some friends who are actually very, very immersed in efforts to stop voter suppression and I’ve learned so much from them this year and some of their efforts. Beyond that, I know that I watched… I’m also in some Facebook groups that are specific to certain interests and things like that, and vary the course of the last month, I’ve seen some really interesting things shared by people. Things that they received in the mail, emails that they have received. When we talk about things like voter suppression, I have personally seen examples of what it is people receive. One of the things, one of the campaigns that we saw was voters in Florida and Alaska received emails and they… These particular actors, and they believe that these were Iranians, Russians, and some other cyber operatives, but I think that this was identified as an Iranian effort.
These voters in Florida and Alaska, they received emails warning, “Vote for Trump or else.” And the US officials determined that this was Iranian influence specialists. Oh, way to be an influencer. But their emails said, “You are currently registered as a Democratic, and we know this because we have gained access into the entire voting infrastructure.” And this is what one of the emails obtained by the Alaska public media said, “You will vote for Trump on election day, or we will come after you. Change your party affiliation to Republican to let us know you received our message and will comply.”
This is not a slam on Republicans or Donald Trump in any way, but what it is the look at one set of copy for a message that somebody got. And you know what? If my mom got that email, and by the way, my mom’s dead, so she didn’t get on but… My parents would be, “What? I’m nervous.” And this is a little bit like a ransomware campaign, right? “Do this so we know you got our message.” When you talk about the ability to hack into voter registration databases, I have to believe that’s incredibly easy to do, because what we see from a security standpoint at the government level, is very, very outdated, lackadaisical, cringe worthy. This is the kind of information that’s available to hackers and they can do a lot with that.
One of the things that I also saw, as it relates to security is that what happened in this election is that the Department of Defense went on the offense. Instead of playing the defensive game, they went on the offense and they really wanted to make it through the voting season without a major cyber attack. I think we accomplished that. There were somethings here and there, but I think as of today, we don’t have knowledge, or I certainly don’t have knowledge, of a major cyber attack that happened. One of the things that… American operatives call this hunting forward, so they surveilled the work of Russia, the work of Iranian and other cyber operatives, and they really tried to get out in front of it. I know that I have seen reports, over the last couple of months, identifying something here, identifying something here to be on the look out for.
I always actually like it when they don’t really share many detail, because what’s the… I always wonder, when you share exhaustive details about campaigns, where we’ve thwarted cyber attacks, how much good does that really do us? I don’t want to tell people what we did to catch them. I just want to catch them.
Olivier Blanchard: It’s not too bad to let people know, “Hey, look, be on the look out for this.” Right?
Shelly Kramer: Right.
Olivier Blanchard: And especially this kind of ransomware adjacent campaign, which to your point, would be very effective with my parents, because they would completely believe it, and-
Shelly Kramer: But it’s scary too, when somebody says to you, “I know where you live. I have your phone number.” That’s scary to me.
Anyway, I think that campaigns like this can work.
Olivier Blanchard: They can.
Shelly Kramer: And to your point, sometimes all it takes is they just have to work a little.
Olivier Blanchard: Exactly. It doesn’t take much. Now, can we talk about polling?
Shelly Kramer: Absolutely. Let’s talk about polling.
Olivier Blanchard: Let’s talk about polling, because as tech analysts, I think we can maybe shed a little bit of light on what happens and why in 2020, especially, after the disaster in the polling world of 2016 predictions, we still don’t seem to be getting really good polling in the United States. In the age of data, in the age of big data, what’s going on? What’s happening? On the one hand, just kind of as a meta… And please, feel free to step in and tell me where I’m wrong or if I’m missing something.
Shelly Kramer: We always do Olivier.
Olivier Blanchard: I know you’re there, you won’t hesitate-
Ron Westfall: Yeah, actually.
Olivier Blanchard: … but I’m giving you permission just to be magnanimous about it. I think that we’re so used to Facebook, Google, YouTube to have all these platforms that are extremely good at collecting data, at crunching data, at coming back with very specifically target messaging, in terms of time, place, et cetera. The data game of these platforms is so good that when we look at pollsters, there’s a disconnect. How is Facebook’s data so good, how is Google’s data so good, But pollsters, who do this for a living, can’t get this right? I just wanted to make people realize that Facebook is a closed ecosystem where you’re pushing in data all day long, all of your behaviors, all of your clicks, all of your comments are analyzed, and you’re on Facebook.
But a lot of people are not on Facebook, and many voters are not on Facebook, they’re not really using Google for certain things, in a way that would help pollsters. There are different types of data, and there’s a lot of commercial data about your purchasing behaviors, things that you like and don’t like, your tastes, whatever, that don’t really translate into national polls at all. If pollsters were to stick to the data being collected by Google and Facebook and did nothing else, they would have a very skewed view of the United States.
For instance, pollsters in 1936, polls were off by an average of 12% compared to where the actual voter sentiment was. Now, it’s two percent. We’ve gone from 12% in the ’30s to two percent, but-
Shelly Kramer: That’s pretty good.
Olivier Blanchard: That’s very good. And we are actually really within two percent, even though it seems like we’re not. Response rates, however, have gone from 36% in the mid-’90s to six percent today, so response rates are going down, and obviously, phone surveys were kind of the big way of doing polls in the past, now, that’s not the case.
Shelly Kramer: We quit answering our phones.
Olivier Blanchard: We quit answering our phones, so the internet is the next best thing. The problem with the internet is on the one hand, about 10% of Americans are not on the internet, and they’re a very specific subgroup, so you can’t really target them this way. There’s also no central register of email addresses the way that there’s a central register of phone numbers and addresses. You can’t canvas the same way on the internet. That’s another issue. So what you end up with on the internet is really having to shift to an opt-in survey model. Opt-in surveys tend to be six to 10% off the actual benchmark data that we get from pollsters that do national benchmarking surveys. Now, you’re six to 10% off, which coincides with the error of the pollsters, when they’re looking at the national average and sentiment between Biden and Trump or any candidate for that matter. They also tend to have a really hard time surveying under 30 voters, so you were saying [crosstalk]-
Shelly Kramer: And there were a lot of them.
Olivier Blanchard: Yeah, were lots of them this time, and also Hispanic adults are very difficult to reach with opt-in online surveys. On the one hand, we have this expectation of big data, huge data crunching and specificity, because of the apps that we use and the technology platforms that we use. But at the same time polling for political purposes is a completely different scape. Secondly, we don’t really have an effective mechanism for pollster to be able to get a clear view of what everybody feels. They’ll get very good data in some areas and very poor data in others, and that’s what you’re seeing. I think that Ron might have some insight in to how AI is being used, perhaps not as well as it could be, in trying to bridge that gap between the six to 10% off and where we actually need to be.
Shelly Kramer: Let’s hear it, Ron.
Ron Westfall: Wonderful segue, you took the words out of my mouth, Olivier. Yes, there was some expectation that AI engines would out perform traditional polling institutions, and we’ve seen, for example, aggregators, like RealClearPolitics and FiveThirtyEight, be incredibly off in terms of the outcome of the 2020 election, not just at the presidential level, but also down ticket at the states levels, in terms of Senate and House races. And it reinforces your point, there’s just have to come up with better ways of sampling. We know one takeaway already is that sampling rural areas is more challenging, so there just has to be ways to figure out how to either work around that or get better information from rural areas.
A second one is that there does need to be more refinement of ethnic breakdowns. For example, grouping Cuban Americans with Mexican Americans is not a good idea. It is going to throw off your data information in places like Florida, for example. These are some lessons learned, but then also showing like in AI the information is just as good as what the information that’s being accessed by the pollsters, so even the AI engines, can be off. In fact, if you looked at Advanced Symbolics’ Polly engine, it was off even more than the pollsters. This I think is indicative of the fact that the AI community also needs to learn the same lessons as the pollsters that we’re pinpointing here.
This is something that while a analytical outfit like KCORE was pretty good at predicting elections in places like Taiwan and UK, they can’t really use the same methodology for the US elections, quite simply. I think this is something that’s a bit of another surprise and a clear takeaway that’s the entire universe of the polling community is going to have a black eye from this and also, quite frankly, the AI is community. Nobody really nailed this right, except for some outliers, and this is just fascinating. Here we are in the era of big data and we still have to get the right data. What we need is an era of correct, precise data, not big data.
Shelly Kramer: I think that the pollsters, the whole data and pollsters and all that, is so interesting, but I think that a lot of the information that I relied upon, and we all take in and process information differently. But what I expected to happen in this election happened. User behavior is user behavior, and we know that in certain markets, voters tend to do this, and we know that in other markets, tend to do this. It’s like the Arizona, we know what they’re going to do in Arizona. We know that a lot of their voting is going to be done by mail, and we also know that, generally speaking, Republican voters go to the polls. Now, that’s in a non-COVID year, so you have to factor that in, and when you also factor in the concerns about a global pandemic and people being cautious and everything else…
But I think that what I expected to happen, and we’re certainly not through this yet, was that there would be certain things that would happen on election day, and then following election day, other things would happen, and we would see those results beginning to change, because of the great influx of the mail-in votes and this is a huge year for that. I didn’t really ever look at it as pollsters were incredibly off, and again, I didn’t sit in front… I knew and still am trying to cling to sanity, so I just didn’t part myself in front of the TV all day, every day, because I knew that I didn’t want that. Again, trying to cling to some sense of sanity. So to me, I didn’t look at this as the pollsters were so, so off, but what I did find was that… And analysts, I think the same knock can be fairly made against us. We all want to predict. We all want to make predictions and we all want to be right.
And I think that in spite of the fact that we went in to election day 2020 really knowing that we were not going to go to bed at night and know what the outcome was, we knew that, everybody told us that, almost everybody, but, yet, we expected that to happen, and the news media across every outlet, projecting and guessing and everything else, and I think that that’s all contributed to our sense of frustration and our sense of not knowing. But, again, we went into this, [inaudible] we went into this knowing, really knowing, that this probably wasn’t going to be decided on election day or the next morning.
Olivier Blanchard: I think that’s the difference between analysts and pollsters though. The analyst will get the trend lines right. They’ll say, “We think that so and so is going to win. We think so and so is going to happen.” And so you kind of paint this bigger picture of very long-term, big macro lines. The pollsters are there to give you the data and be as specific as possible and give you that insight that allows the analyst to get to that point. The issue is, I look at, for instance, in my own state, South Carolina, we had a very interesting Senate race between Lindsey Graham, who has been a senator for a while and he’s head of the Senate Judiciary Committee, so it’s a pretty important job, and Jaime Harrison, who’s a relative newcomer, and for the entire election, pollsters were saying, “It’s very close. They were neck to neck. It’s 49% to 49%. It’s going to be really, really close.” And it was nowhere close.
Shelly Kramer: Yeah, it was nowhere close.
Olivier Blanchard: … double-digits difference, and so on the hand, analysts can be generally right and get away with it, pollsters who are off by that much, have some explaining to do.
Shelly Kramer: What was that… Take that campaign, in particular, what has been the feedback in terms of how were those pollsters so far off?
Olivier Blanchard: I think that pollsters, in this case, and I don’t want to point any fingers, I have no evidence, this is just my opinion, I think sometimes pollsters, like analysts, like anybody else, can get into a mode of telling the story or giving the insight that they are paid to give. I think that there might have been, on the side of some pollsters for what every reason, whatever the motivation was, a tendency to want to make this race closer than it really was.
Shelly Kramer: Interesting. Well, it was one of the… Was it 90 million? I know that I have his campaign, Jaime Harrison’s campaign and Amy McGrath’s mixed up in some instances, but I know that those two campaigns were incredibly, incredibly well-
Olivier Blanchard: Well-funded.
Shelly Kramer: … so much money. Yeah.
Olivier Blanchard: When you look… Mitch McConnell and Lindsey Graham are very important opponents for Democrats, so I can see how their perceived vulnerability would have brought in a lot of dollars from out of state. And I know that a lot of people were donating money from… and I’m not talking about corporate interest, I’m talking about just voter donors-
Shelly Kramer: Absolutely.
Olivier Blanchard: … were just donating to Jaime Harrison’s campaign and Amy McGrath’s campaign, from every state.
Shelly Kramer: I think I heard that the bulk of that money actually came from out of state, so I think that’s interesting as well. Not necessarily a bad thing. Let’s take a shift here and let’s talk about the market. Let’s talk about what happened in the market. I know you have some thoughts on this Ron, in terms of why the key US market indexes are up. What do you think?
Ron Westfall: Well, perhaps, it is a consolation for everybody regardless of their preferences politically, and that is, quite simply, the markets prefer gridlock, and it’s looking like that will be case, at least at the national level and for that matter at the state level. Even though the presidential election outcome is undecided right now, it looks like there is a solid likelihood that the Republicans will retain the Senate, that’s not official, but the indicators are pointing in that direction. In the House, there is a smaller majority for the Democrats, what that slender majority will be, is still yet to be determined. But the upshot is is that historically, the major US indexes have performed better in this scenario, and it’s for a couple of fundamental reasons.
One, there’s less likelihood of major policy shifts that can upset markets in that regard. And secondly, Jerome Powell, not exactly a household name, but he is the US Federal Chairman, and he’s a steady hand. He is ensuring that as we work through both the global COVID-19 pandemic, as well as the undecided aspect of the US presidential election, that there won’t be any surprises in that regard, when it comes to fiscal and monetary policy. So this is a bit of a shining light and I think something that is helpful as we steer through this next era of unknowns during the year 2020.
Shelly Kramer: The year 2020. The year that will never end. That’s what it is.
Olivier Blanchard: 2021 might be even better.
Shelly Kramer: Well, there’s a lot ahead, that’s for sure. Do you have anything you want to contribute about the market or in closing here, Olivier?
Olivier Blanchard: Yeah, everything that Ron just said is spot on. What I’ve heard, the feedback that I’ve heard from my friends in New York, who deal with this stuff a little bit more than even we do, is that politics aside and however they may have voted, whatever, in terms of the markets, they like that there’s going to be stalemate and gridlock in Washington. Which is kind of counterintuitive, because which ever side you are politically, you don’t like gridlock. We complain about gridlock, but the market actually see this as stability right now, because, obviously, if Biden wins, there’s going to be a radical shift in the White House and in terms of foreign policy and all sorts of other aspects of governance. But as long as the Senate doesn’t turn Democrat, there is going to be this kind of, okay, we’re not [crosstalk]-
Shelly Kramer: Checks and balances.
Olivier Blanchard: … anything major done. There’s not going to be a lot of real big shuffling that could upset markets, so that creates, ironically, expectation of stability that’s actually health for the markets right now, for investors. Yeah, markets should be up for a while. Had we had a complete shift in one direction of the other, I think that would have been very destabilizing for the markets, so that’s not the case. What’s bad for fans of politics and fans of policy in general, is actually good for investors, in this case, in the short term, anyway.
Shelly Kramer: Yeah. Well, and that’s not a bad thing. It’s not a bad thing [crosstalk]-
Olivier Blanchard: No, we can use a little stability in the middle of a transition like this, because everything else is still shifting sands.
Shelly Kramer: I think we could definitely use some stability. Well, gentlemen, thank you very much for hanging out today. It’s always a pleasure. And for our viewers and our listeners, thank you for stopping by, and we’ll see you again soon.
- Successful Digital Transformation Requires Unleashing Your IT Team’s Potential — Here’s How to Do That - April 19, 2021
- SAP SuccessFactors Users Reap Big Benefits from Axonify for Improved Efficacy of Frontline Employee Training Programs - January 8, 2021
- Automation Anywhere Launches AARI, its Digital Assistant for Work - October 7, 2020