We're sorry but this page doesn't work properly without JavaScript enabled. Please enable it to continue.
Feedback

Pointing Fingers at 'The Media'

00:00

Formal Metadata

Title
Pointing Fingers at 'The Media'
Subtitle
The Bundestagswahl 2017 and Rise of the AfD
Title of Series
Number of Parts
167
Author
License
CC Attribution 4.0 International:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor.
Identifiers
Publisher
Release Date
Language

Content Metadata

Subject Area
Genre
Abstract
The German election in September 2017 brought a tectonic shift to the layout of German politics. With the AfD in parliament far-right illiberalism has reached the mainstream. We investigate the communicative developments underlying this rise. Using web-scraping and automated content analysis, we collected over 10.000 articles from mainstream-news and far-right blogs, along with over 90GBs of Tweets and thousands of Facebook-Posts. This allows us a deep insight into how public discourse works in 2017 Germany.
Keywords
HypermediaSound effectMechanism designReading (process)Embedded systemNatural numberFormal languageProcess (computing)Focus (optics)Density functional theoryWordProcess (computing)Revision controlFormal languageHypermediaPresentation of a groupIdentity managementTheorySlide ruleNeuroinformatikCASE <Informatik>AreaMathematical analysisResultantSound effectCore dumpOvalAxiom of choiceMUDMilitary baseArmStudent's t-testBitGroup actionGastropod shellModemRational numberIterationPhysical systemQuicksortDialectGoodness of fitMechanism designGame theoryParameter (computer programming)Multiplication signInheritance (object-oriented programming)GenderMathematicsReading (process)DiagramDemosceneInternetworkingM-theoryNatural languageEvent horizonSelf-organizationCentralizer and normalizerPosition operatorExterior algebraSet (mathematics)Uniqueness quantificationComputer animation
AverageLink (knot theory)Letterpress printingGreen's functionInformation securityIntegrated development environmentState of matterCentralizer and normalizerLink (knot theory)Term (mathematics)CurveCASE <Informatik>TwitterGraph (mathematics)Frame problemMultiplication signQuicksortIntegrated development environmentAverageData dictionaryScripting languageHeat transferRippingRange (statistics)Electronic mailing listException handlingOcean currentNumberMusical ensembleShared memoryCountingRight angleBuffer overflowStrategy gameStack (abstract data type)DemoscenePrice indexEmailLie groupEvent horizonHypermedia2 (number)Theory of relativityBlogPlotterSound effectOrder (biology)Mathematical analysisCategory of beingGroup actionCausalityBoss CorporationWritingSlide ruleWordMoment (mathematics)ResultantCovering spaceAreaRing (mathematics)Letterpress printingInformation securityDrop (liquid)Table (information)Physical systemLevel (video gaming)Line (geometry)Mechanism designFocus (optics)Inheritance (object-oriented programming)Diagram
MathematicsHypermediaCASE <Informatik>GoogolTwitterCausalityWordSound effectHuman migrationNumberBitVarianceDirection (geometry)CASE <Informatik>MIDIParameter (computer programming)QuicksortMathematical optimizationNP-hardMathematical analysisLink (knot theory)HypermediaCurveBounded variationRange (statistics)Graph (mathematics)Representation (politics)Time seriesTwitterScaling (geometry)Point (geometry)Line (geometry)Vektor-autoregressives ModellLie groupTerm (mathematics)Software developerSearch engine (computing)GoogolMathematicsLogical constantMechanism designTheoryInformationDecision theoryAbsolute valueResultantArchaeological field surveyStatisticsProcess (computing)Theory of relativityRow (database)Frame problemMultiplication signMultiple RegressionMereologyRevision controlPole (complex analysis)Group actionDensity functional theoryDynamical systemWater vaporRegulator geneExterior algebraTemporal logicVector spaceOrder (biology)Bookmark (World Wide Web)Boss CorporationPulse (signal processing)MeasurementDiagram
TwitterMultiplication signQuicksortMathematicsHypermediaGraph (mathematics)Right angleInternetworkingAlgorithmCuboidDynamical systemSoftwareConservation lawMetreNumberIntegrated development environmentInformationAreaTwitterInsertion lossMatching (graph theory)Task (computing)Positional notationWeightMereologyService (economics)Graph coloringExpected valueDensity functional theoryGreatest elementGreen's functionExtreme programmingElectronic mailing listWeb pageWritingValidity (statistics)Group actionOrder (biology)TrailLine (geometry)Wave packetLogicSound effectComputing platformPhysical systemFrequencyExterior algebraPoint (geometry)Power (physics)Computer wormMathematical analysisHome pagePhysical lawMessage passingTerm (mathematics)Centralizer and normalizerArchaeological field surveyDot productFile Transfer ProtocolGraph (mathematics)SphereShared memoryDrawing
MereologyLinker (computing)NumberDifferent (Kate Ryan album)Context awarenessTwitterComputer virusQuicksortInternetworkingHypermediaRight angleStrategy gameField (computer science)Web pageSystem callBitInformationBoundary value problemBlogDemosceneDegree (graph theory)CASE <Informatik>RobotPrice indexTerm (mathematics)Group actionLevel (video gaming)Automatic differentiationElement (mathematics)ExpressionMultiplication signVideo gameDigital mediaMappingLie groupDimensional analysisCategory of beingWebsiteData storage deviceExtension (kinesiology)Point (geometry)Information privacyLink (knot theory)Row (database)Density functional theoryLinear regressionMessage passingLikelihood functionSound effectGenderJSONXML
WindowPresentation of a groupInternetworkingHypermediaPairwise comparisonSurjective functionContent (media)Multiplication signProcess (computing)BitWordLogicState of matterGraph (mathematics)Annihilator (ring theory)Virtual machineMeasurementCASE <Informatik>Negative numberLink (knot theory)ResultantAlgorithmLatent heat2 (number)Medical imagingTwitterQuicksortConjugacy classEmailWage labourCodeObservational studyScripting languageBlogNumberFrame problemFormal languageMathematical analysisConsistencyCovering spaceOracleBlock (periodic table)Point (geometry)Sound effectArithmetic meanSelectivity (electronic)Software testingDegree (graph theory)FacebookAtomic numberFocus (optics)Identity managementRight angleMatching (graph theory)RobotLinker (computing)Different (Kate Ryan album)Line (geometry)Proof theoryJSONXML
MedianSystem programmingHypermedia
Transcript: English(auto-generated)
The next speaker is born and raised in Germany. He lives and works as a PhD student in Canada, as a member of a research group on extremist
politics and democratic systems. And he'll give us an insight into the public discourse in Germany focused on the so-called alternative for Deutschland. Please welcome Alexander Bayer.
Thank you very much. Thank you, people, for showing up in the Saalborg. Thank you, the internet, for watching. Very big thank you for the organizers for giving me the opportunity to give this little
talk. Yeah, my name is Alexander Bayer. And everywhere I went this winter, I didn't have to wear a winter jacket because the temperatures were very mild. And I will tell you in a minute why that matters. As already said, I'm a member of a research group in Vancouver, where we look at what
happens, how extremist parties and politics fare in democratic systems. And we decided to focus this research project on the fascinating, for researchers, fascinating case of Germany and asking the questions if we can point fingers and is it a valid judgment
to say the media is to blame for the rise of the AfD. For anyone who decided at the end of 2017 that they would spend most of 2016 in hibernation, which seemed like a pretty good idea at the time, I will give a quick rundown what happened.
So we had an election in September, and the domino piece that was Germany fell, domino piece in a sense that all around in Europe, far-right parties had considerable success in the past, in the recent past, and Germany was the sort of last stalwart in central
Europe where a far-right party did not get into government. This happened in September, and it did not only get into parliament, it also, the way
it looks like now, it might become the official leader of the opposition. So when these results came in, pundits were really quick to call the shots. The dominating sentiment was that it was the media's fault. They took the positions of the AfD and gave disproportionate amounts of coverage to this
far-right extremist party. And this sentiment had a lot of truthiness to it, so it had a lot of, yeah, sure, I can see why that, everyone that opened a newspaper or opened a news website, stories
about the AfD seemed, or about something that's related to the AfD, seemed to dominate coverage. This went along with a little bit of a felt truth, a truth that was perceived by people about how the campaigning season was a lot of season and not a lot of campaigning,
despite Martin Schulz's best efforts, a whole lot of sunshine, but not a lot of conflict, and this was something that then was perceived to be very, well, I don't want to say very skillfully, but somehow filled by the AfD and the topics that are of concern to this party.
So what are we doing today here? First off, I'm a political scientist by trade, and political scientists like theory. I know that this is an event where theory might not at the forefront of everyone's minds,
but it is for me because talking and arguing to political scientists about theory is kind of like mud wrestling with a pig. You do that for two or three hours, and then you realize, oh, this pig actually enjoys this. So I'll be sort of, I have one slide on what we, what previous theories would suggest
have happened and how it could have happened. Then I'll show you what kind of data we have collected to systematically answer this question and talk about public discourse in Germany, then to the meat and potatoes of the talk
about how the campaign unfolded in the media, and I will then, to end, I will show some more data that is a bit different that paints a picture on why this election was a special election and why it was sort of a perfect storm
of an election for a far-right party, and why this actually makes us claim that the media could be said to have behaved pretty reasonable as a little teaser. Okay, theory, one slide. Two possible mechanisms of media effects.
There's this normative, very endearing and wonderful idea that if you read something that someone carefully crafts and he or she constructs an argument that is well-written,
well-made, you read this, you take it in, you're persuaded by that, regardless of what this argument is. 60 years of media research suggests that this doesn't happen. Pre-existing opinions are extremely difficult to change in each and every single one of us,
even though we're likely to admit that, no, no, no, sure, I'm a rational thinker, I take standpoints if they're convincing to me, and I internalise them, but it's not how this works. The second possible effect, and the one that will be of concern to us today
at the core of the presentation, is something that's called priming. So, the media can't tell people what to think, it can't persuade people independently of the previous opinions that people have, but it's really, really successful in telling people what to think about.
It's super good, the media is super good, reading something is very effective in bringing something to the front of your mind. And here I can tell you why I told you about my choice of attire in winter. The vast majority of you probably thought when I said this,
oh, I didn't have to wear a winter jacket. Wow, who's this guy? But maybe a few of you thought, yeah, sure, it was pretty mild, that's climate change. So, without naming the issue, there's a chance that I primed a few of you to consider climate change and pull that in your frontal lobe at the front of your mind.
And this is important, this is the central thing that we have to consider if we ask if the media wrote up a party like the AfD.
Also important to consider here is that priming is easier, or there's an indirect effect of priming as well, where a topic that is owned by a specific party, that's the thing that then favors the party subsequently. So, if the media writes a lot about refugees, a xenophobic far-right party
that has this problem of refugees at the core of their agenda will reap in benefits in our minds, in that its agenda will fall on fertile ground.
So far, the theory, that's all. So, what do we do based on this theory? We collected data, lots of data. We understand this text that we collected to be data, and we use natural language processing to analyze it. Natural language processing basically means that we're giving language to a computer
that wasn't written specifically to be understood by a computer and try to extract meaningful analysis based on what the computer is doing with this. So, we used some sifting methods to collect about 8,500 articles
from four central German news websites, Focus, Bildt, Welt and Spiegel, and we have that result in a unique data set that, to our knowledge, no one else has. If so, please reach out to us. And this data set is so unique that it deserves at least six fire emojis.
It is also pretty exciting because that was pretty cheap. We were two people that were mainly concerned with collecting this data, and I don't want to calculate my hourly wage, but it was almost done with no financial expense.
And this is cool because social scientists were faced with this problem with this very interesting case of Germany falling in line, very delayed, with lots of other countries around it in terms of the far-right party getting their scenes in parliament.
And we can use methods that are available to us if we're sitting down and reading on Stack Overflow and teaching those methods to us to systematically try to answer this question. Let's dive right in. The share of party mentions in online news. So what we did for each day,
we calculated what the total number of mentioned political actors is. We did that based on work lists that we carefully crafted that included candidates' names and party abbreviations and party names and things like Kanzlerin and Kanzlerkandidat for the CDU, CSU and the SPD respectively.
And we let that thing rip through our little R script that we have. So the average of mentions of each party over the course of the campaign looks something like this. Between July 1st and September 24th,
that's the time frame that we concentrated on, we see a clear incumbency bonus, the Kanzler bonus, the Kanzlerin bonus for the CDU, CSU, Social Democrats, High Twenties, and the AFD at 10.7%. Here we might say, at smaller parties,
a little note to the green and the left. So with this dictionary method, it's kind of tricky because we can't say, oh yeah, well, we're just going to count every occurrence of Grüne and Linke for the green party and the left party because then we get stuff like the green banana and the left hand that is counted for them.
So that's why here we are only using candidates' names, that's why they probably, they sort of underperform, but for our purpose of talking about why if the AFD got favoured by the media, we are sort of letting that drop out of the table. So, the story here is, over the course of the campaign, 10.7% of mentions were happening that mentioned the AFD.
Basically, case closed, right, AFD got 12.7% in the election. That doesn't really sound like it was favoured by the media. And a few of you might know this, this analysis from a blog post that me and Constanze Kurz wrote for Netspolitik
sort of like 45 seconds after the election when we worked on truncated data, and we also focused on print media. And this is sort of what this graph looked like that we based our conclusion on. AFD didn't really get any disproportionate amount of coverage. It actually is, in the last week of the campaign,
the last weeks of the campaign actually is outperformed by the FDP. Science is the current state of erring, so now that we have better data in terms of online news data, this whole story looks a bit different if we take the average over the whole course of the campaign
and actually have it shown to us day by day. This is what I want to focus on now. So, just looking at the sort of tail end of this all the way to the right when we get close to the election date.
The order of this is surprisingly close to the actual election results. The parties actually do get in in the order that they came out of the election. But we do see a little curve that gets closer to a curve that should be bigger,
and this is where the, well, I don't want to say magic, but this is where the interesting stuff lies. So let's look at the curves one after the other. The CDU-CSU, as you would expect as the incumbent, anything that is remotely political in domestic and international politics
will score mentions for the chancellor and the CDU-CSU. That's why this curve is considerably higher than the others, but we do see a downward tendency the closer we get to the campaign when campaign coverage shifted from the incumbent to the competitors, especially the underdog competitors,
which is kind of a bad transfer to the SPD now. But if we look at the curve of the Social Democratic Party, there's a slight bump around August, and Martin Schulz really tried to drive home this issue of justice as the central campaign promised, and there's another little slight hump around September 1st,
beginning of September when the televised debate happened. But the overall trend is pretty linear. It doesn't seem to be, if we would just smooth this plot out to be a straight line, it probably would be pretty much horizontal. Not so for the AFD.
So, remember, over the course of the campaign, they got 10.7% on average of mentions. And that's true if we calculate an average of that. Of course, this looks like it scores considerably lower than the two major parties. But something happens in late August, and all of a sudden, this party gets actually close to the Social Democrats.
Starting in late August, the tendency becomes one that is pretty considerably upwards. And if we take the average of only the two last weeks before the election, we get to a number of 19.6 of all mentions are talking about the AFD there.
Which is something, if we think about the mechanisms of priming, those are short term effects. We're looking for things that happen over a short term or have an effect in a pretty short term. So this is something that is extremely, extremely important. At the beginning of this timeframe,
where the plot becomes something that has a trend that shows upwards, around August 28th, where the first little summit occurs, two things happened. One, a refugee boat capsized in the Mediterranean, an event that we sadly have to see terrifyingly often,
and one of our people died. And the second thing that happened was that Alexander Gaulland, in an interview, claimed that a German politician should be dumped in Anatolia. And it's interesting if you talk about, if you extract the topics that are covered in relation to the AFD,
before and after this moment, before this August 28th, it's a lot about Alice Weidel writing emails where it turns out she's not the public persona that she claims she is, and it's a lot about internal rifts of this far-right party,
the internal tensions between the super far-right wing and the far-right wing. And afterwards, there's a surprising amount of citations of this, oh, we should dump this person in another country.
So let's not be an indicator that this strategy of sort of provoking a scandal paid off. But let's, before we get into that, let's look into the topics that were covered over the course of the campaign. We did the same thing, we developed topic dictionaries
with keywords for each category, and we let our script read through all the data and count occurrences. So, looking at this, we see a sort of band there in 10% range where it's all a colourful rainbow,
where the topics don't really diverge from each other, except for that topic of domestic security, which is there at the low end of the range. But we do have one topic that stands out quite considerably in the early months of the Wacom Summit, which is European Union, general European Union topic.
This is because on July 1st, Helmut Kohl, the internal chancellor, got the first European act of state, and a lot of things were written about his legacy in terms of the European Union, and lots of people showed up from Strasbourg and Brussels and paid their respects. This is why this topic seems like,
or this is why this topic comes in as strong as it does here. Another topic that has a sort of unusual curve here on our graph is the topic of the environment. Our dictionaries that we developed were topical, and so what causes this steep, steep summit there in early August
is the Dieselgibfe, the diesel summit, where German car manufacturers try to sort of get out of the fact that they basically ripped off customers with selling cars that emitted toxic amounts of poisonous gas and dust.
This is why this is extremely important in the low 40% range in early August. Afterwards, the trend line points steeply down. A topic that was pretty consistent over the course of the campaign in its overall dynamic,
but the role that it played is the topic of immigration, and immigration means migration and refugees in our case here. And now thinking about what that means in relation to our theory on priming, we would think that, sure, that's a topic that is owned by the AFD.
It's super tightly connected to that party's rise. So this is something that does favour a far-right party like those are, like it is. But we can do a sort of more systematic investigation into this.
So this graph shows you the polls. Each dot represents polling results for the AFD, and the line is the average out of those polls, again, over the course of the timeframe that we surveyed. Pretty much constant until mid-August,
and all of a sudden we have increasing variance and we have a tendency, a trend line, that points upwards. And now this is where the heart of the story lies. Is this dependent on the mentions that the AFD got in the media? This is the orange line.
Now we have a different scale of our graph, that's why it looks way more nervous than in the bigger one that we had. Difficult to say. If you have data like this, time-serious data, you actually want to get rid of trends in terms of what the analysis should be like. So one way to do this in a graphic representation
is by not showing the absolute values and how they develop, but only showing the change from day to day and plotting that. This is what this graph does. So here, these two lines dance around the zero mark because, especially the blue one, where it's the polling results,
there wasn't a lot of variation from day to day. It's in incremental steps that the curve points up and down. It gets a bit higher in variance after the mid of August. Whereas the AFD mentions in the media, they stay rich in variance.
Hard to tell if anything systematic is there. You would think that after the first third of August, those lines are connected. We ran an analysis, a vector autoregression model, time-serious statistics.
We couldn't find any systematic relation in a time frame that made sense for our theory on priming, which is a few days that we're looking for. So if you talk about time-serious, we talk about lag and lead. So you try to connect a data point that is further down the line with a data point that is not as far down the line,
and nothing of statistical significance showed up here. This kind of stumped us. We thought, when we looked at this, there was something. We took a step back, and we considered another possibility as to why the media reported as they did.
Did the media just give the people what the people wanted? And here is why I want to talk to you about why this was a special election.
I adapted this graph from the Berliner-Morgenpost, and they based it on surveys conducted by Infotesteemup onto the data I didn't have any access to. But this impressively shows why there was a special election. In five out of the six preceding elections,
employment was the topic that was on top of people's minds when they made the decision in terms of which party to vote for. Employment means unemployment. In 2017, with unemployment being at record lows, and after 2015, having a Syrian civil war still going on,
having refugees come into Europe, immigration jumps on out as the topic that was the most important for people. And here, if we look at also the topics that are further down,
the important scale for voters, those are all topics where one could conceivably think that those can be spun in a way that they are connected to this refugee situation. Social injustice, economic injustice, that's something that a party like the AfD can very effectively turn into
an idea on group-based conflict, it's us versus them. The same with pensions. Oh, those people come here to take our jobs and our money, and especially from the old people, from our L&E. So 2017, the Bundestagswaal 2017,
is a special case if we consider it compared to other parties. So now having this situation where we find that it's something that basically never happened in recent history in Germany before in terms of what made people decide at the polls,
we wondered, OK, well, is there a way to more accurately measure this demand side of things, this need for information of voters? And what better way there is to measure the salience in the population than to look at Google queries?
So we collected Google trends data, more specifically, the Google searches on refugees, general term, and again, here's this way to even out a trend line, this is the daily change in how this topic developed.
And if we put our daily change of AfD mentions over that, we do see that there's something there. There's some sort of systematic relationship. And then crunching these numbers and putting them again
through a vector autoregression model, we come to the conclusion that with a lag of only one day, Google searches for refugees actually lead AfD mentions in the media. So if on Tuesday a higher number of people in Germany googled refugees,
on Wednesday the AfD was mentioned more often than the day before. The end effect wasn't big, but it was there and it was significant. We also, of course, consider the alternative. The magic word is here, it's Granger causality. So you can actually calculate and reliably calculate
that the temporal succession means that one follows the other. And so all of a sudden, it becomes a bit difficult to point the fingers at the media, because if the media just reacts to an interest,
it operates like a business, if we like it or not. There's the normative idea of the media, especially in a country that is rich in high-quality publications, as is Germany, that the media is a public good that educates people and brings out the best in them, in challenging them
and persuading them of the best side of the argument. But at the end of the day, in the online world, it is a business with a measurable outcome. You have clicks, you have trackers, you have ad durations that you can measure, and so you can see which articles are favoured and which articles people last the longest on.
And we're not saying this important distinction to make here, we're not saying that there's a direct causal link between people googling refugees and the media directly reacts to that prompt because there's some search engine optimisation guy or girl in every media publishing house
that monitors what people are interested in. We're saying that there's an intermediate step there. It's not a direct cause, it's just a sort of delay that is in there that allows for other mechanisms to get in. So we're wondering, what about the consumer focusing on the demand side?
And in 2017, there's a few things that you could actually look at to gauge what the demand side demands, and we decided to focus on Twitter, because without actually knowing this when we first started out
with collecting all this data, we decided to set up a Twitter scraper, and that way, between September 1st and September 24th, we collected 4.5 million tweets that contained keywords, that contained any one of a list of keywords that had political connotation.
So looking at this body of data, we can extract things like the top 200 most used hashtags, and if we do that, and we count the tweets that contains one of the top 200 hashtags,
and we pay special attention to which one of these hashtags are decidedly pro-AFD, we get to a number that 30.9% of the tweets that contained any of those top 200 hashtags actually contain one that is in favour of the AFD, whereas if we count the decidedly no AFD, the anti-AFD,
no AFD in all ways of spelling and capitalisation and so forth, that's only 1.2%. And here it becomes a bit ticklish. So in order to give a better idea of what role Twitter might have played
in our little relationship here between the demand side and the supply side, the supply side supplying the news, we have a beautiful network graph. So this is a retweeting network.
We extract all the mentions of an actor, each dot is a Twitter user, each line is five or more retweets. Retweets, we are aware of that. Retweets don't automatically mean endorsement. You might retweet something that is outlandish and crazy, but for the sake of visualising what the weights are on Twitter,
we're treating them as the same. Anyone who has ever worked with network graphs of that size, they take a long time to generate, and it's kind of tough to label them, so I'm very proud that I was able to do so. If we look at this island down there,
that blob, that blue blob, those are accounts that cluster around AFD accounts. The colouring here was done by a walk trap algorithm. I just adjusted the colours that that algorithm used to actually match the colours in the German party landscape.
So we do have a hefty continent at the bottom right that connects all kinds of people to the AFD. There, if you look at the little appendix below here that is coloured in brown,
that is mainly organised around a movement called Reconquista Europe, which is an even further right-wing movement that is sort of directly tacked to this island of the AFD, and the connecting node is Björn Höcke,
which is quite interesting. So we have the AFD down there. We have the other parties up there, the rainbow that is the pluralistic political landscape. We have those two extreme points. They are at the super top right, and they are at the bottom left.
Those are very extremely extreme Twitter user parties. It's the UDP and the Freienwelle. So they don't seem to engage with the nodes that are in the centre here. But what is also valuable to note is that
for the other parties, for the established parties, starting from the left in orange, the pirate party in then red, the Social Democrats in purple, Sti Linke in green, yellow FTP in black, the Conservative party, CDU, CSU. All of these parties have a central node,
a central account, around which a lot of other users are fanned out. So for each party there is a smaller number, or a relatively small number of accounts that are highly favoured in how often they are retweeted. AFD doesn't have that.
So this is of course a projection of something that's three-dimensional in a two-dimensional place, so there might be some skewing going on here in terms of how it shows on our screen, but even turning it and trying to identify which party is at the centre wasn't really possible. So the internal rifts and the internal power struggles,
they do show in how members of the party are retweeted. Also interesting to note is which nodes, which users are connecting these two continents, so to speak. One is that blue dot is a polling aggregator.
Of course everyone is interested in getting their polling numbers out. And that's tough to see here, but there's a beige user in the middle there, which is Welt.de, so one of the media publications that we actually collected data on and surveyed. Another thing that I'm just going to mention here briefly
is that light pink coloured insert between the greens and the central grey beige dot. Those are Jan Böhmermann, Die heute schon extra 3.
The dynamics are clear that we have this party that is pretty well organised on social media and thus is able to dominate a media agenda that is based on algorithms, basically.
If you think about how the logic of information dissemination works on Twitter with trending hashtags, if you have a party that is as tightly clustered within itself as the AFD shows up here,
there's a good chance that that will influence what all of us get to see when we check out the Twitter homepage. Now I know that probably a good chunk of you have burning questions in their mind, and are going to want to know, so how many of these bright blue blobs are bots,
are Twitter bots? We tried to find that out using a tool called the Bot-O-Meter, which is something that has an API available online where you can submit, it's a project from a research team in Indiana, where you can submit the name of a Twitter user and then it runs lots of analyses and analyses,
lots of things about this user, the frequency of tweets, the time at which it tweets, who is it following, who is it followed by, who is it talking to, that kind of stuff. But when I tried to submit that, I broke their API.
So if they happen to watch, I apologise, that was me. So I wasn't able to do so in time, but there's a bunch of talks tomorrow that talk exactly about that thing, so I'm happy to have this as a lead-in for the day tomorrow. So what can we take from this?
A Bundestagswa 2017 was a perfect storm for a far-right party like the AFT. You had a high-issue salience of the topic that is at the centre of its agenda, and you have a sort of unregulated Wild West of social media.
We will see how that changes with recent law changes come into effect, where all of a sudden the platform itself has some liability to which kind of messages are spread, but if that's effective for Twitter, it's a whole other bag of worms. So, in that sense, that's what I was sort of hinting at,
in this issue environment, we have people be interested in the topic that is central for the party like the AFT. The media behaved pretty surprisingly predictable,
at least for the topic, for the publications that we covered, it did so. And for the context that we're arguing here, that the AFT only gets like 20% of the share towards the end of the campaign, is something that is a little bit surprising.
And that also leads into a different question of what does this, oh, it's the journalist's fault, actually mean? What does it really mean? This sort of is based on this normative expectation of the media being an impartial deliverer of information.
And if you think about what else is going on on the internet with alternative media and an alternative news sphere establishing itself with news blogs like, well, I don't want to call any names, because there's a sort of scene of far-right fringe blogs in Germany
that we also collected. And so we're further down the line, we're going to look at what the topics were that were covered in that and how that connected to influencing public opinion in Germany. But having said this, with these alternative ways of getting your news information being available,
if you have the press, if you have the mainstream press not covering a party like the AFT to a certain extent, you only give the fodder to those cries of Ludenpresse, a mendacious press,
in members of the population that are sort of at the risk of being lost as audience members. So it's kind of difficult to call the shots here and actually point the fingers at the media, because they delivered on informing on an interest
that existed in the population before they reported on something like the AFT. And with this, I want to leave it at that. I thank you very much for your attention, and I'm highly, highly eager to hear questions and prompts and ideas how we could pursue this further.
Feel free to attend the microphones. Even the microphone I don't see behind the cameras.
Let's start with number two. It should make some sound, thank you. Thank you very much for your amazing work. I've got only one question. Do you plan on releasing those collected data and on what license? That's a question that we ask ourselves, too.
We would love to collect the data, and ultimately it will happen, but we have to make sure that we actually have the right to do so with the way we collected it, but we're definitely looking into that. Okay, number five. Yeah, you. Okay, hello.
Is this working? It's tempting. I'm from the Netherlands to compare these experiences with the AFT with the experience in the Netherlands. You know, we had Wilders, we had Verdunck, we had Fritan, and now we have Baudet, and it seems that there is a major difference with the AFT, because, frankly, I don't know the name of the leader of the AFT.
It used to be Frank Petry, and now I don't know. But in the Netherlands, the leaders of those populist right-wing parties, they were very good in manipulating the media. They were sending out messages, sending Köder in Germany,
what's the word, like, in sending out provocations, and that attracted attention of the media, so there were people saying, you shouldn't react on all provocations. But anyway, they were geared to draw attention, and I wonder whether AFT has been, to the same extent,
active in the field of drawing attention purposely, using even agencies that are specialized in advertising. Great question. There is this idea that the AFT was very skilful at sort of insinuating scandal
and purposely doing things on a public stage that would draw attention to them. For example, I say it again, this expression by Alexander Gaulland to dispose of a German politician, or the other leading candidate, Alice Weidel,
leaving a talk show while it was being broadcast. So there definitely is this element of actually taking a scandal and using it for pushing your own agenda. Whereas if they used ad agencies for the media campaign,
they did, their campaigning was highly professionalized in terms of what their posters were and how their campaign ads were worked, and they did work with a company that also was involved with Donald Trump's campaign.
But in terms of new media, or online media, it's not that new anymore, in terms of what they did on online media, I just only have an anecdotal sense
if they use something like bots, which is also a way of buying attention. I can sort of tell you about one specific case where we investigated which Twitter users were the most active in tweeting on the AFD,
on German Twitter. Tomorrow there's a talk about a Twitter user called Ballerina, which is a name that has been out there, which there's great indication that that is definitely a bot that has been planted and has been controlled by someone else, or by any group of actors that is not actually a ballerina.
What we found was a Twitter user called teletubby007 that tweeted in those three weeks that we surveyed 6,500 times, and mostly just retweeted calls
to go and cast your ballot that were all put out by the central AFD accounts. It didn't have a lot of followers, something like 500 or so, but it just kept retweeting over and over and over. When we actually wanted to check out the page of that bot,
it was deleted, the user was deleted. To answer your question, this higher degree of personalization that the Pateifel de Freiert has in the Netherlands is not as extreme for the AFD in Germany
because there's more leading candidates and there's internal rifts, like Geert Wilders is basically his own party. That's not the same, but the strategy to use scandal and to use something that is outrageous and push the boundaries a little bit more and then jump back and say, oh no, we did not mean that at all in this way,
that is the exact same spot on strategy that he used. Perhaps I should add that Wilders made it like... Excuse me, many people queuing. Okay, then I'll stop. Okay, thank you. We have questions from the internet then. Yes, QGTG is asking, why do you come to the conclusion that this was a special election
while the last election in Austria has exactly the same issues? Don't you see this as some sort of a global effect? That's true. A Syrian civil war that pushes people to flee from war and save their livelihood is something that is not only felt in Germany,
but for the context of Germany it's a special election. This sort of situation has never occurred in this way before. But absolutely, each election in Europe basically since 2015 was a special election in that sense,
but not in terms of the outcomes in a way, because far-right parties in other European countries already had their foot in the door, and especially in Austria with the F.P.E. were pretty well established with previously having been part of a government and now being part of a government again. But for Germany,
in what the issues were that were on top of people's minds, that's the special case that I'm in. Okay, microphone number three, please. Thank you. First, I really appreciate the sincerity and transparency of your talk. Thank you very much. We need more of this in such circumstances
and maybe less polemics sometimes. There's just a little trifle in your method where I was wondering how did you filter the linker and klühne stuff. How exactly did you do it? Did you maybe count all the mentions of klühne with a capital and non-capital G
and linker with a capital L and non-capital and then filter it out further? Or did you do it the other way around? I know that you focused specifically on the AFD stuff and maybe you were focused on representing all the parties that might be relevant,
but I would still be interested in that part. Thanks. That's a great question. The thing is that we used, after we collected the text, before we put it through the analytic methods, we put it all into lower case
just so we could have a consistent way of analyzing. And with capitalization, sometimes it just trips up the way to treat this. And that's why we ran into these issues with linker and klühne where we had to resort to only taking basically the candidates' names
and then also linkerpartai and klühnepartai and a few conjugations. So, dear linkerpartai, Dean wrote dramatically the cases. We conjugated them through. But yeah, since our focus was on the AFD,
we weren't especially concerned with that, which is unfortunate, I admit that. But for the purpose of this talk, we decided to just use this workaround. Okay, thanks. Okay, microphone 6, please. Hello, thanks for your interesting presentation. I'm wondering if you and your team,
so you looked at mentions of the different parties, but I'm wondering if you looked at the content of the articles and how they talked about it, if they were talked about positively or negatively. Thank you very much. It's a great question that we actually did consider. I'll answer this question with a counter-question, as social scientists like to do.
Anyone in this room use Amazon Mechanical Turk and works on hits to earn a few cents here and there? No? Okay, so I can speak freely. There's a method that uses cheap labor on Amazon Mechanical Turk and presents each worker with two sentences,
out of which they have to change the one that is more positive. And so we wanted to use this to train a machine learning algorithm to actually get a way to gauge the sentiment of positive and negative in the text that we had collected. We started that in early December, and we had a workbook with 4,000 so-called hits,
4,000 little jobs, 4,000 comparisons, and when this job was done, five or six days later, we sort of put that through a test and compared it with our own hand coding that we had done,
and it turned out that one worker on Amazon Mechanical Turk spent over seven hours and worked, of those 4,000 little jobs that we had, he worked 3,980, and over 1,400 of which he did in less than two seconds.
Which is unfortunate because this person, question mark, probably used a script, probably used a bot or just randomly clicked. The coding didn't match up at all with what we did hand-wise ourselves,
and that really screwed up our approach there. If any of you plan on doing some hits in the New Year for Amazon Mechanical Turk, and you're asked to compare two sentences that mention a political actor in Germany,
you can send me an email and maybe a screenshot and tell me how much you appreciate that we're paying six cents for each comparison. But that's a story where we don't have any sentiment in this analysis here. In this context, I'm very much a ghost of Christmas future.
In your Twitter data, where you take retweets as well, do you determine what are quotes and what are direct retweets? Because in my experience, and I work with this in Denmark and in the UK,
a lot of people like to distance themselves from what the AFD and similar are saying by quoting everything they're saying and giving them the press. That's a very good point to make. We did not make any distinction between quotes and retweets,
but we did filter based on five retweets by thinking, okay, if you occasionally feel like you have to point something out that is outrageous and ridiculous that a member of a party says on Twitter, you would be inclined to do so less than a certain amount of time.
We also tried it with other cut-offs. The graph basically always looked the same. But if we think about what this means for how the demand side is influenced, it doesn't matter, basically, if you're retweeting out of endorsement or out of spite. That's the logic why we decided to use mentions and retweets.
Another question from the internet. Yes, noob23 is asking, do you think that the window of commonly acceptable ideas in the so-called overtone window was shifted to the right by the ideas of the AFD echoed in the media?
That's a good question. Something that comes to mind here is that media use is epiphenomenal. The question is, does something happen in you because you use a certain media outlet or do you use a certain media outlet because something happened in you?
From the sense that I got, I would say that the degree to what is acceptable definitely was shifted over the course of this campaign, that all of a sudden we're questioning if remembering the Holocaust should be something that is at the heart or very close to German identity.
That's something that a political scientist would have never expected, that this cleavage can be opened up again in a way that is so potent as it did now. So it definitely did something to the overall discourse in Germany.
Whereas that is an effect of media reporting on the AFD would require us to use something like the sentiment analysis to actually determine how the media talked about which aspect of the AFD agendas.
I can see some movement behind microphone number 8. I'm sorry. Thank you very much. Thank you for your work. I still do have a critical question. Basically, the things you showed is something like we all know. We could see this happening last year, I mean this year, in the last election.
So I'm wondering now whether the method you used, which was basically focusing on quantity, is sort of mirroring what was happening. And I'm wondering if you would keep working on it or like you used buzzwords and you used the media instead of like narrowing it down
or using more specific questions. And I was wondering if you have these results now and you have proof for them, what are your next questions and how can you continue to use the data you have
to make it more specific so we can really have some outcome and some conclusions coming from this? It's an absolutely wonderful question. Of course, we thought about using this data further down the line. Our initial plan was to connect this not just with salience data that we derive from Google searches,
we also have Facebook data that we collected that we wanted to look into, but it's a bit challenging to actually analyse comments in depth onto language because language tends to be way more fluid and you have certain problems with selection and self-selection.
So you really, really have to be careful to cross-connect which person that comments on Facebook is the same person and thus, if you only do quantitative stuff, would appear disproportionately.
As I mentioned, we have also collected data from far-right blogs, from news blogs, that very actively endorse the AFD and their topics. And so we're planning to pull this into the analysis along with data from the German longitudinal election study
where in this timeframe that we surveyed in the data, each day, 100 people in Germany were called up and asked about their feelings towards specific parties and actors. So we actually have day-by-day data once it comes out on what people thought about those actors.
So we're planning to pull that in as a more reliable measure for salience. Thank you very much. I'm very sorry about time. So there will be no more questions right now in front of the audience. Alexander Bayer, thank you very much. A warm applause please.