LawBytes Podcast – Episode 29 transcript powered by Sonix—the best audio to text transcription service

LawBytes Podcast – Episode 29 was automatically transcribed by Sonix with the latest audio-to-text algorithms. This transcript may contain errors. Sonix is the best way to convert your audio to text in 2019.

Michael Geist:
This is Law Bytes, a podcast with Michael Geist.

CBC:
Distortion is a certainty and there’s no point declaring it can’t happen here. It’s already here. So what are we going to do about it? About misinformation and those who would mess with our minds and elections.

CPAC:
Are you concerned with fake news and disinformation in Canada?

CPAC:
Not really. It really comes down to are you willing to look into the information you’re trying to feed yourself in, the information that you’re trying to project out into the world? Right. If you’re constantly being told what to believe in, what to think and what to say, that’s more of a you thing than a Canada thing. So I don’t blame Canada. I blame the person or the people.

CPAC:
Absolutely. I think it’s a very important issue. It’s important to have reflective and accurate information and such. I think fake news is definitely misleading and can often construe mis interpretations of reality.

CPAC:
Well, I am concerned about fake news because there’s so much info out there already. With Wikipedia leaks and everything going online. You never know what kind of source is true. Like I’m in school and just source checking is a big thing. Like you have to check every every source that you have. So it’s fake news makes it harder now to know what’s true and what’s not.

Michael Geist:
Coming into the 2019 federal election, there were widespread concerns regarding disinformation campaigns, the prospect of foreign interference, social media advertising and manipulation and fake news. In fact, the federal government enacted legislation designed to foster greater transparency on political advertising. But on the heels of elections elsewhere, the prospect of online harms during the election appeared very real. Taylor Owen, the Beaverbrook chair in media ethics and communications in the Maxwell School of Public Policy at McGill University, set out to find out what was actually taking place online. He led the Digital Democracy Project, which studied the media ecosystem in the run up to and during Canada’s Oct. 2019 federal election by monitoring digital and social media and by conducting both regular national surveys and a study of metered samples of online consumption. The project released reports throughout the campaign on how social media was being used political advertising trends, the role of fact checking and the presence of misinformation and fake news. He joined me on the podcast shortly after the election to discuss.

Michael Geist:
Taylor, thanks so much for joining me on the podcast.

Taylor Owen:
Hey, my pleasure. I’m a fan.

Michael Geist:
Ok, well, that’s great to have you on it. You know, there’s been a few people that have been incredibly active during this election campaign beyond, of course, the leaders. And I think you’re one of them because your digital democracy project has put an enormous number of new reports and ideas out into the public sphere. So what do we start there if can explain a little bit with the project is about?

Taylor Owen:
Yeah. I mean, so the genesis of it was a sort of observation that in many of the other countries that have had big elections since the 2016 U.S. election, where there was substantial foreign interference, there was a community of scholars and sometimes non civil society and even for profit actors who were monitoring it, monitoring the information space during the election. And we didn’t see that community existing in a robust way in Canada in the lead up to this election. So we saw the that’s the genesis of the project. Was this in our perception that we could add something by doing some pretty wide scale monitoring of the media ecosystem during the election. So the way it kind of came together is we have a team at McGill that’s being run by a computer science professor, Derek Ruths, and they are they led and are still doing a fairly large data collection project on of the media ecosystem. So they’re collecting as much Twitter as possible. Facebook, public posts, all of Reddit, all news published and distributed during the election. So really just looking at all. What are the ways in which we can capture the public discourse during an election. Where we think maybe we added some mythological capacity to even what’s been happened in other countries or has been implemented in other countries is we tried to pare that online data collection activity with a number of survey mechanisms. And the reason for this is that in most of these other studies and other elections, they’ve been able to perhaps spot disinformation campaigns or say something about the narratives that were emerging during the elections. But they really didn’t get out behavioral change.

Taylor Owen:
So did exposure to those kinds of narratives or that kind of potentially problematic content, whether it be foreign influence or just fake news campaigns or ever it might be. Did that actually change the behaviour of citizens during an election. So that we worked with Peter Loewen, a political scientist in Toronto, and his team where they ran national surveys every week for nine weeks and they ran a what’s called a metered survey, which is a sample of Canadians that allowed us to collect all of their online data consumption, online consumption or web consumption. So we saw everything that they saw over the course of a period of time during the election and surveyed them on the front and back end of that. And we did a active survey sample where we sort of recruited Twitter users and Facebook users off the platform itself and brought them into a panel survey. So we had these these wide variety of mechanisms and really we were just trying as much as possible to, shed some light during the election on on what was happening in the media ecosystem.

Michael Geist:
I mean, it’s really, from a Canadian perspective, unique to see that this kind of mass data collection, at least by within an academic environment, taking a look at what people are actually doing, surveying them to a sense, get a sense of what they think they’re doing or what they say, and then being able to pull that together. And I know that over the course of the election campaign, you were putting out regular reports that touch on a pretty wide range of issues.

Taylor Owen:
We did, we did. So we we weekly tried to release some set of findings either from little experiments that we ran each week or just reflections on some of the trends we are seeing, it ends up being much more difficult than we expected. I think there’s a reason that there are not a large number of people doing these kinds of things in real time during the election. It’s it’s not easy to put out things that have a degree of academic rigour to them on that kind of timescale. And I think we’re probably and we also say a lot more after now really after period after the election or we can take this what is a really massive dataset of all this online behaviour and then take out the metered survey, all this all the data that this sample of participants, all the Web data and all the site content from the Web sites they saw and really start to make sense of that and hopefully tell some sort of relatively robust story about the election.

Michael Geist:
So the work will certainly continue. Why don’t we zero in on on some of the studies that you did put out as I think this is still fresh. You know, one of the things I know that you took a look at was, as you mentioned, social media use during election campaign. And do you have a sense of based on the data, what does it tell you about about Canadians usage and does it increase during these these kinds of activities?

Taylor Owen:
Yeah. Look, it seemed to spike quite significantly during the election, so we tried to create measures of political activity because we really we didn’t care about all activity on these platforms, but really what could be construed as political and that activity on Twitter sort of activity on the main hashtags, by partisans, by political journalists, by candidates themselves, that cumulative discourse grew by 800 hundred percent down from the pre, pre and post election writ period. And on our Facebook, it was 250 on public posts on Facebook, which importantly is what we can see, right. We can see the public post, not the private posts went up by 250 percent. So absolutely there was a spike.

Michael Geist:
That’s that’s a massive increase in news that being did. Does the data tell you that’s being fueled or do you have an idea whether or not that’s being fueled by the political parties and the politicians themselves, or is it the public that’s fueling the discussion or is it a bit of both kind of responding to what each is seeing taking place?

Taylor Owen:
Yeah, I’d say all you all of the above. I mean, all so we try on Twitter, for example, we clustered different users. So we had a general public user base, which were just people who posted in any of the election or political related hashtags. We had all political journalists in one cluster. We had all candidates in another cluster. And we had sort of kind of some thought leader type category. So on each of each of those, the usage spiked massively. But one thing that was really, really clear was that the overall discourse was highly partisan. So you could see very clear demarcations of partisan clustering across all of those groups where individuals with particular ideological or partisan affiliations were speaking largely to themselves and sharing content that confirmed broadly the narratives of those political parties. And so that that I think is is a key aspect to it. Yes, it spiked. But also we really did see on Twitter, particularly this partisan clustering.

Michael Geist:
So it’s super interesting you suggest that that we that we’re creating a large public sphere, but one in which you’ve got a series almost of echo chambers where people are just reverberating the same messaging with within wherever their partisan views happen to land. Is that is that the case or do you? Did you find some amount of crossover? Is there a sense is there the ability to persuade or are people just becoming more and more firm in their views when they venture into, say, Twitter or Facebook?

Taylor Owen:
So it was only when we saw this clustering that was one of the questions, right? Like is this. Is this a filter bubble or an echo chamber? Right. Is that’s the way the system is filtering content to people based on pre-existing behaviour, or is this people self-selecting into these kinds of conversations creating echo chambers? And so because we had these survey experiments running, we could test that, try and test for that. And one of things we do is we exposed people in a survey experiment to a broad range of news and saw and and asked them which ones they wanted to consume. And overwhelmingly, people choose content that supports their put their pre-disposed positions in that experiment up the really clear results. But then we can test if continued exposure to that kind of political messaging changes their beliefs. And we found it actually does. It actually strengthens their perspectives and their pre-existing positions. So that could be the spillover effect of this, is that we know people are choosing and self-selecting into communities both online and in these survey experiments, choosing content that confirms their pre-existing views and that over time their beliefs on that issue become more rigid and strong.

Michael Geist:
Right. So so rather than seeing crossover, we see a firming up of people’s positions in large measure, I suppose because they are hearing the same things, are reading the same things continuously that reinforce the views that they came in with to begin right at the start.

Taylor Owen:
Yes. I mean, that’s our perception now. Yeah. But we are really that those one final thing on that is that in this in this everything, there’s another element here, which is whether we as a society more broadly are polarized from one another and how we view people in other clusters, in other ideological clusters and some of the survey work. Peter was able to get out some of that, too, and actually found fairly high degrees of what they call effective polarization. Would just like all core dislike for other political parties or other supporters base just because they are members of an opposing group. And that is something that I don’t think people really thought existed to the degree it did in the Canadian political system. But it it came through pretty, pretty strongly.

Michael Geist:
So the you know, the aftermath of the election results where we’ve had a fair amount of discussion about polarization, more on a geographic basis between provinces or perhaps urban, urban, rural. You’re saying that you see the same kind of thing taking place in social media.

Taylor Owen:
Yeah, in terms of what people are sharing, who they’re following and the content they are consuming. It is polarized.

Michael Geist:
Yes, the one of the other areas that you focused on and had the opportunity to do so in large measure because of new legislation that required disclosures of political advertising on social media was to take a closer look at that kind of advertising of advertising, particularly, I know, on Facebook. What are the kinds of things that you found taking place?

Taylor Owen:
Yeah. I mean, that preface, I think it’s important that that this the our ability to see something inside is in part because of C-76 which mandated these ad archives. I think it’s we were able to see Facebooks because they’ve done a much better job at making that data public than the other platform companies. Remember Google, right, decided that they were not going to sell political ads rather than impose or develop the archive for the Canadian market.

Michael Geist:
But if there’s anything that’s worth exploring why that might be the case. But Facebook had already implemented this in the US to their credit and and then deployed that ad archive system to Canada in a fairly effective way. And so for this, we there are a couple of projects that we’ve partnered with that we’re kind of leading this. One was that team at the NYU Engineering School, which had been done some of the best ad archive work in the United States and the Ryerson Leadership Lab are their project on this, too. And I think it tells a few things about the ad system. And we we were they and we to a certain degree were able to see which what political parties were doing, how they were using that tool of Facebook advertising differently than each other. And there are some differences. I mean, the Conservative Party used it much more as an extension of broadcast advertising, where they were blasting similar messages to large clusters of people in across different areas of the country, whereas the liberals were using micro-targeting looks like to a much greater effect using custom lists, which it doesn’t look like the Conservative Party was doing. And you could see some of that third party activity. Right. So Canada Proud emerged as by far the biggest third party ad spender. I mean, more than the next seven, I think, combined or something. So there’s no question that they were using that tool to great effect. But it seems that that’s the one piece of it. But then because we had access and visibility to the ads and this is a new thing, right, that we can see not just who spent what, but actually what the ads themselves were that they spent on. We could do some experiments with the ads themselves. So we use this survey tool again to test whether some of those ads that the campaigns were running or working, whether they’re having a positive or a negative effect on voters, that could tell us something about the nature of this kind of advertising more generally.

Michael Geist:
And did the advertising target you coming back to the partisan playground findings: Are people advertising to reinforce the views of people who already have a particular view, or is advertising seen as an opportunity to see if they can’t pull some people away from one perspective into theirs?

Taylor Owen:
That’s interesting. We we don’t have that analysis yet. I think I certainly want to be looking for the most as I’ve weeded with around negative versus positive framing. So the experiment we ran was on whether these negative framings versus positive framings work and what effect each one has on people’s beliefs of the party and on the issue. And on that, we can show that that negative ads, just as we know, work better. But they make people angry at the people who posted them as well as the people they are. directed at. So they actually kind of increase polarization, I think, but decrease it. But they do get the message across. But no, I think on the micro-targeting stuff, we will we will be able to do more going forward on that. But it is a limitation of this, right, in that we only have access to very limited targeting data. And that’s something that many of us, civil society and academic groups using these data in the US in particular have been flagging that ad targeting legislation needs to broaden the mandatory requirements of data release there because it’s just too limited in what we can see right now. It’s a very general targeting information, right.

Michael Geist:
So a more transparent approach would really provide a great deal more insight. I just want to circle back just to make sure that I understood. So negative advertising is effective in terms of just reinforcing a negative view. But you’re suggesting that people also have a negative perspective on the person doing the negative postings as well?

Taylor Owen:
Exactly. It drives up negativity on both, but it works to get the the message of the ad transmitted. So people have a higher negative perceptions of the party at which is directed. But also of that, who posted that. The inverse is actually, to a certain degree, true too. That positive ads seem to reduce that variable of effective polarization that I mentioned. Just latent dislike for the other political group. But the parties just aren’t using it very much. So I think the parties could, if they had more positive ads, decrease effect of polarization. But they’re just not doing it.

Michael Geist:
That’s really interesting. As you know, the Herle Burly was a extremely popular and incredibly insightful podcast series that ran throughout the campaign. There was a lot of talk about negative advertising with the pervasive view, I think, of those of people like Scott Reid, Jenny Byrne that you needed to move towards negative advertising. But it sounds like you’re finding says, yes, they do work, but they also have real costs, which is, I guess, the tradeoff that campaigns have to make when they decide just how negative to go or to embrace a more positive approach.

Taylor Owen:
Yeah, I think that’s right. I think there’s some blowback there that I would think they would. Political campaigners would acknowledge that that’s the case and they see that risk as as worth taking. I think there’s a broader question, though, about what that kind of negative attack advertising does to our broader political discourse. And if it leads to just everybody disliking the opposing side more, that’s probably not a great outcome in the end of the day.

Michael Geist:
No, it’s not.

Taylor Owen:
But I think that what more working on political advertising isn’t worth framing here is that we started with the mentioned C-76. And I think looking at the who is advertising and what they were doing, it is highly likely that an a type of political advertising during an election was dissuaded by the very fact that it was going to be made public. I think we saw relatively clean digital advertising campaigns in the ads we’ve seen. And I think it’s an open question over who else would have used that tool had it remained secret. Counter-factual, obviously, we don’t know, but I think it’s worth reflecting on when we look at the effectiveness of these kinds of policies.

Michael Geist:
You know, that’s a terrific insight. It’s consistent with those that have pushed, of course, for more transparency in any number of different places where if you know that what you do may be revealed to the public, there’s greater transparency there. So you’re in the spotlight a little bit. It has it can have real effects on the way people behave. Whether that’s with your in your government or corporate accountability or in this context, knowing that the ad campaigns you run would be made available in this manner, might have some thinking twice about what they chose to do.

Taylor Owen:
Yeah, in particular, when you combine it with other elements of C-76 like limits on digital spending. Right. So there might actually be not just perception downsides to being made public, but actually penalties as well.

Michael Geist:
So the legislation is imperfect, as you’ve already mentioned, in ways that it could be improved. But it has has has did have some positive effects on this campaign. You know what, if the so that I think for those that we’re actively engaged in some of those policy issues, I think that that’s that’s a great takeaway. You know, one of the other ways, of course, that people become disheartened isn’t just through the advertising, but is through the range of the information itself, sometimes false information, sometimes active disinformation. I know that you took a look both at disinformation campaigns as well as fact-checking and the impact that that has if we could take a take a discuss a little bit. Those from a fact checking perspective. We’re seeing more and more groups I know engage in fact checking. Does it have much of an impact?

Taylor Owen:
Yeah. And so we wanted to look at that on whether exposure to correct information in response to potentially false information or a predisposition to believe something on a topic that wasn’t true, whether that could actually change people’s opinions on it. And we found two main things which one is relatively positive and the other shows its limits. I mean, for one, we found we did experiments on both climate change facts and immigration facts. And on both of them, we found that for people who believed false things on either of those. So for refugees, it was number of refugees coming to Canada. And on climate change, it was some basic facts around the Paris agreement and the causes of climate change. If people were presented with corrected facts, they would change their present, their their their own facts. Right. So they would update their knowledge on the issue and and agree to respond with those new facts. That’s the positive thing that the fact checking could work. Then the limit of it is that for partisans, it does not change their policy position on those issues. What I think tells us something really interesting about how people come to their policy positions. It’s a lot more than just the facts, right? So on immigration, it’s more than just how many absolute number immigrants they think we are. We we or refugees, we accept a year and on climate change. It’s not just about whether they think climate change is human induced. It’s about something more. And so, yes, we can change their knowledge on an issue, but we can’t necessarily those that fact-checking exercise doesn’t change their policy positions for partisans.

Michael Geist:
That’s pretty. It’s awfully discouraging, actually, when you think about it. If you’ve got people who who know from a factual perspective that the policy they’re advocating for, the underlying facts are inconsistent with the policy, but yet they still will advocate for the policy because it’s consistent with their broader partisan view.

Taylor Owen:
Right. And I mean I mean, in part that’s understandable because people support policies for this wide range of reasons might be values based. It might be identification with a particular ideology. Right. It might be the reason that the perception that other people have pushed of alternate policies in the past. Right. There’s also the reasons we might come to believe certain things, but it’s there supports certain policies. But it’s. But facts are just a piece of that, unfortunately. And then, look, there was another sort of really discouraging aspect that we found here, which was that discouraging. But I think offer some pause, particularly to the media and journalists, which is that it actually didn’t matter who is doing the fact checking and had the same level of effect. So we tested whether it mattered if a political candidate, a friend on social media or a news, a reputable news outlet, one of the top most trusted news outlets in the country. And it actually made no difference, which if I were a journalist or a media organization, that would probably give me pause.

Michael Geist:
It would suggest that people are people. People may be open to hearing new facts. But the the degree to which they look at what are otherwise viewed as reliable, authoritative media sources is just one of a number of different places where they can get their information and look at those sources as being roughly equivalent.

Taylor Owen:
Yeah, I think that’s right. And we did a number of tests around the perceptions of the media more broadly, and there’s some positive things in that I mean, people thought that people are pretty high levels of trust in traditional media. The country has much higher levels than we see in other United States, for example. We also found kind of surprisingly that the the more media someone consumes, the more likely they were to be misinformed on policy issues and election issues. So again, that would if I was a journalist, that would give me cause on the effect I’m having on the electorate when if I if I as I as a citizen consume more journalism, I am more likely to be misinformed on the main issues of the election.

Michael Geist:
I wonder if that just speaks to there’s so many different. We’re being bombarded with so many different issues and perspectives that would you the more you consume, sorting through all of those different issues is increasingly challenging.

Taylor Owen:
I think that’s right. And partisanship, when you layer that on, makes all these variables worse. And that’s really one of the things we sort of landed on here, is that partisanship really does have a its force in many variables. And so if you add partisanship on top of media consumption and if you add social media consumption on top of that, to the degree of social media coverage, it’s the worst. So the worst the most misinformed people we tested were partisans who consume a lot of media, mostly on social.

Michael Geist:
People who otherwise if you were to ask, I assume I tell you that they believe that that they are very well informed and here’s why they’re so well-informed.

Taylor Owen:
And that’s a nail on the head there. And that’s the variable that we tested, the difference between being uninformed and misinformed. Right. So uninformed people were actually very willing to be corrected because it actually might not be that a problem that would have a problem in terms of disinformation, which we’ll talk about because they’re OK at being me being corrected. It’s the misinformed people who think they know think they are right. The bigger problem.

Michael Geist:
So you you mentioned disinformation or misinformation online and as a good segue, that’s that’s, of course, one of the reasons we saw some of the legislative activity to try to expose activity, expose advertising online to address that issue. And what are the real concerns about interference with the campaign fuelled by disinformation. I know that you attempted to try to take a look at some of that activity as well to the extent to which to identify whether or not it was happening in Canada. What did you find?

Taylor Owen:
Yeah. So ultimately, the election, when sort of events emerged that people were suggesting might be disinformation campaigns or inauthentic activity. Representative of inauthentic activity, we took a pretty close look at it and will more so now get a bit more time. But the high level takeaway is that we did not see a large amount of either fake news, news and false news designed to appear like traditional journalism or regular journalism or the kind of inauthentic behaviour that we’ve seen in other countries, either domestic or international. So hashtags being fueled by bot activity, for example, or or false narratives and memes sort of spreading virally in an authentic way. We did not see a lot of it. There was some of that, absolutely, but not a lot. And for what we did see, it appeared to be in this kind of a perverse effect of an earlier thing we talked about, which was the clustering of the online discourse. That clustering, it looks like, provided a bit of a buffer for the spread and distribution of that problematic content. So if you look at something like the Buffalo Chronicle stories that kind of took off, two big ones which were sort of probably the clearest cases of fake news distribution and inauthentic behaviour that emerged. Those were just largely distributed amongst conservative communities on Twitter and Facebook. And that sort of to me implies that the problematic content we’re seeing confirmed existing biases rather than changed behaviour on voting. That does mean it’s not a problem, right? It still it probably made our discourse a bit more toxic and it probably led a certain partisans to feel more entrenched in their views and more negatively about other political parties. But it probably didn’t change your votes. And that’s that’s something we can pull out of this conversation.

Michael Geist:
And that’s, you know, I guess that’s encouraging on one level, too, that the fears of disinformation campaigns, whether foreign influenced or otherwise, didn’t emerge. The same time, though, I think the identification of the impact that that fake news has yet not to necessarily deceive or bring people over into one went from one camp to another, but rather to further entrench the kind of divide that it sounds like you you’ve identified across a range of different studies really from. Yeah. Who do you trust more broadly to misinformation to even the way people are engaging in advertising. The same trend keeps reoccurring in a number of different studies you conducted.

Taylor Owen:
It doesn’t like so. So one thing that you probably saw that analysis on Twitter kind of took off mid-campaign around the potential foreign influence of Trump supporters in uncertain hashtags in Canada. So this researcher found a pretty high level of Make America Great ID on accounts, talking, driving the anti Trudeau hashtags that popped up. And so were suggesting that these bots are authentic, coordinated behavior. So one of the things we did do a little bit of a deep dive on on these anti Trudeau hashtags is about 40 that emerged during the election. So we looked at over a week, about seven hundred thousand tweets on those 40 hashtags, trying to see if like. Is that the case? Right. Is there? Is it because I think that that would be a significant concern. Right. There was a coordinated Make America Great Again community that were bombarding the Canadian election. And what we found was there were a lot of accounts that had these kinds of ID, but they didn’t look to the American look pretty clearly. They were Canadian. They weren’t necessarily tweeting in a manner that suggested they were bots or even inauthentic or even coordinated. But at what it looks like, they were were just a bunch of partisans who tweeted a lot. Right. Tweeted 100 times a day, largely things that confirmed their existing views or their hatred of whatever it might be. And we’re just sort of having this conversation, this collective conversation under these anti Trudeau hashtags. And when we looked at other hashtags that were pro Trudeau or pro NDP, we actually found very similar types of behaviour. These people who are just tweeting a lot with fairly toxic kind of memes and content, speaking to each other, can reconfirming their preexisting biases. And your point that tells us something important about our political discourse, right, that there are these toxic sub communities. They are highly polarized. They are not exposed to information that in any way challenges their their beliefs. But is that inauthentic? I don’t know.

Michael Geist:
What’s notable as well is, as you point out, that this is not a left issue or a right issue. Although each side, I think of it that terms you see the same kinds of behaviour you’re suggesting occurring, whether you’re on the left of the political spectrum or on the right.

Taylor Owen:
I should say we see the similar types of behaviour, but very different magnitudes. At least in the Canadian context, where the scale of the activity on the anti-Trudeau hashtags was radically disproportionate from the ones on the existing on the left. And and I don’t have a good explanation for why that’s the case.

Michael Geist:
Okay. So that so that while the types of behaviour are similar in terms of just the volume of activity, there are differences depending on which obstacle spectrum you look at.

Taylor Owen:
In the Canadian context in this election. Absolutely.

Michael Geist:
Right now, they’re thinking ahead before we close. You’ve talked about the impact of Bill C-76, talked about the prospect of some further research and the likelihood of the kinds of things you could do now that you’ve got that large dataset from a pure policy perspective. Do you have a takeaway or two about what governments ought to be thinking about as we look ahead? Next, elections obviously a number of years away. But these kinds of issues are going away.

Taylor Owen:
Might be another number of years away. A few months, I guess, to watch. Greatly hoping it’s not. But look, I mean, I think what we talked about in C-76 is that the far from perfect. Canada had the advantage of coming after a number of big elections where this problem was being analyzed. So just by coming 8th or whatever it was in a series of democratic elections since 2016. We had the benefit of hindsight, right, to look back and see what happened. The elections and the government, to their credit, I think did some things to try and alleviate the low hanging fruit that were the easiest mechanisms to interfere in election that we saw in other campaigns. The ban on foreign advertising, important money, I think is an easy. It was an easy one. The ad archives making some of the stuff visible also played a role. Some of the limits on third party advertising, though controversial, I think probably dissuaded a certain type of activity. The Foreign Interference Committee. I think that probably maybe, I don’t know, dissuaded certain foreign actors. Those are all counterfactuals and we can’t know. But we do know that we did more than what a lot of other countries had done in response to this perceived threat. And we had much less. I think we had a cleaner election in terms of the information space. So that I think is a broad reflection. The other piece of it, though, would be. Are we simply regulating and researching, I would say from our perspective, too, based on previous campaigns and the methods that were revealed in previous campaigns, and so are we studying, regulating for and studying the right things? And I’m not totally convinced we are. I think there’s there’s highly likely dark spaces that aren’t revealed by any of these mechanisms. What we are allowed to see from the research perspective or what is being limited on the regulatory side. And I think that there’s just a broader question about the nature of our political discourse in the online space. And is it the kind of discourse that we think ultimately benefits a democracy and that that’s obviously not a problem that will be addressed with necessarily more regulation. But I think its the question we need to ask ourselves, is this the kind of discourse we think is most beneficial for an election? I don’t have a great answer to that, but it’s a concern I have.

Michael Geist:
Well, I think it’s certainly a question worth asking and to sure that we get informed answers or at least an informed discussion around it. I think we’re reliant on precisely the kind of data and reports that you’ve been generating. So it’s, I think, a really important service that you’ve pulled together.

Taylor Owen:
Thanks so much.

Michael Geist:
Thank you. Thank you for joining me on the podcast today.

Taylor Owen:
My pleasure.

Michael Geist:
That’s the Law Bytes podcast for this week. If you have comments suggestions or other feedback, write to lawbytes.com. That’s lawbytes at pobox.com. Follow the podcast on Twitter at @lawbytespod or Michael Geist at @mgeist. You can download the latest episodes from my Web site at Michaelgeist.ca or subscribe via RSS, at Apple podcast, Google, or Spotify. The LawBytes Podcast is produced by Gerardo LeBron Laboy. Music by the Laboy brothers: Gerardo and Jose LeBron Laboy. Credit information for the clips featured in this podcast can be found in the show notes for this episode at Michaelgeist.ca. I’m Michael Geist. Thanks for listening and see you next time.

Quickly and accurately convert audio to text with Sonix.

Sonix uses cutting-edge artificial intelligence to convert your mp3 files to text.

Thousands of researchers and podcasters use Sonix to automatically transcribe their audio files (*.mp3). Easily convert your mp3 file to text or docx to make your media content more accessible to listeners.

Sonix is the best online audio transcription software in 2019—it’s fast, easy, and affordable.

If you are looking for a great way to convert your mp3 to text, try Sonix today.