Why Republicans cannot win: Google, Facebook, and other Big Tech companies are manipulating the flow of information to shift elections.
By Robert Epstein, Gatestone Institute
If I could show you some of the data that we collected, you would be astounded.
There is a lot more that we can find. The bottom line on this is going to be that the types of monitoring systems that we have been developing since 2016, I am pretty sure at this point that they need to be permanent, large scale, and operating in all 50 states to protect our free-and-fair elections from interference by tech companies, which can flip elections any and all ways they please without anyone knowing.
Tech will always be far ahead of laws and regulations, but monitoring is also tech. If we are monitoring them, we are doing to them what they do to us 24/7.
When you monitor, you can catch them in their shenanigans, and you can get them to back down. As you’re going to see, apparently, we got Google to back down by exposing some of the manipulations that they were engaging in before the 2020 presidential election.
The Wall Street Journal in 2018 published some emails that had leaked out of Google in which one Google employee is saying to others, “How can we use ephemeral experiences to change people’s views about Trump’s travel ban?”
You may not have heard the term “ephemeral experiences” before. I had never seen it before, but these are very brief experiences that we all have online every single day. They are things like newsfeeds, search results, search suggestions, sequences of YouTube videos. None of that is recorded anywhere. It affects us, as Google executives know full well. In fact, it has a tremendous effect on us.
It just appears, we click, it disappears. It is stored nowhere, it leaves no paper trail for authorities to trace. It can be used to manipulate, and it is being used to manipulate very deliberately and strategically, especially at Google. I think some of the other companies as well have been catching on, especially Facebook.
As it happens, I have been studying ephemeral experiences since early 2013. In 2015, I published my first scientific report on our findings in the proceedings in the [Proceedings of the] National Academy of Sciences.
This report has since been accessed or downloaded at the website of the National Academy of Sciences more than 300,000 times.
Perfect manipulation device
This was a very sophisticated study with five experiments and more than 4,000 participants. We showed that when there is bias in search results, first of all, people cannot see it, which makes it perfect as a manipulation device, and leaves no paper trail. Search results, for instance, are ephemeral experiences.
We showed in these controlled experiments that we can easily shift 20 percent or more of undecided voters to one candidate or another as we please – up to 80 percent in some demographic groups – and that was after just one search. With multiple searches having bias, the numbers go up and up and up.
What happens over a period of six months, say, if there is bias in search results? We reported in that paper in PNAS that we felt that as of 2015, you could shift so many votes that way that we calculated that upwards of 25 percent of the national elections in the world, the outcomes of those elections were being determined by Google’s search engine.
Here [on this graph], we are months before an election. I’ve got 300,000 people who are undecided. 300,000 who lean to the right – that is my red curve. 300,000 who lean left — that is my blue curve.
What happens is, we are shifting week by week by week. We are shifting people out of the undecided group into the other groups. They are making up their minds, but because of the bias in search results, anything that they search on, anything political at all, is leading them to Web pages that make one candidate look better than the other.
In this case, the Democrat looks better than the Republican. We keep shifting in this way differentially so that we keep putting more people onto the blue curve and we are putting fewer people onto the red curve.
You play that forward for six months. When you get to the election, you have an enormous gap. You have created a gap of more than 100,000 people between the votes we shift into Democrats and the votes we shift to Republicans — with no one knowing and with no paper trail. That is roughly how this works.
In 2015, I received a call from Jim Hood, who at that time was the Attorney General of Mississippi. He was very concerned about this effect, which we called the Search Engine Manipulation Effect, SEME or “seem,” from a perspective of criminal justice issue. He said to me, “They have this power. How do we know that they are using this power?”
Using real people
I did not have a good answer for him, but I became somewhat obsessed with trying to figure out a solution to that problem, and in 2016, my team and I developed the first-ever system for doing to the tech companies what they do to us. That is, we are monitoring what they are showing real people on those screens that we are all addicted to every day.
The point is you have to use real people if you are going to monitor them. You cannot do what some professors at Columbia University have done or what reporters of The Economist did. You cannot use an anonymized computer because Google recognizes that easily enough, and they do not send biased anything to such computers.
You have to look over the shoulders of real people. That is what we learned how to do in 2016. We had to recruit people. We had to recruit a very diverse group, obviously, of Democrats, Republicans, Moderates. We had to develop special software that we installed on their computers.
If you go to tamingbigtech.com you can get the piece from 2018 about how we developed that first monitoring system.
In 2016, we recruited 95 field agents in 24 states. We looked over their shoulders, with their permission. We preserved 13,000 ephemeral experiences on Google, Bing, and Yahoo. That means we preserved their searches and the search results and the 100,000 Web pages to which the search results linked. That allowed us to look for bias, to see if there was any bias in those search results.
Indeed, in all the days we looked at before the election, we saw pro-Hillary Clinton bias and never any pro-Trump bias. I was a Clinton supporter, so that was fine with me, but do we really want private companies manipulating elections on a large scale with no one knowing and without a paper trail? That is the question.
This bias was also present in all ten search positions on the first page of Google Search results. These two graphs show something disturbing: the few field agents that we had who were communicating with us by Gmail, they are people who could easily be identified by Google, they saw virtually no bias at all.
In 2018, we expanded the system. We had 160 field agents preserve almost 50,000 ephemeral experiences and about 400,000 Web pages. Again, on all the days we looked at before the elections in 2018, the midterms, we found consistent liberal bias in search results – enough over time, if that bias had been present through six months, to have shifted millions of votes with no one knowing.
There appeared to be Google bias. There was no such bias in the other search engines. This was definitely a Google phenomenon.
We went wild. For the 2020 presidential election, we recruited more than 700 field agents. This time, we focused on swing states where we knew the action would be. We preserved more than 450,000 ephemeral experiences – not just search results – on Google, Bing, Yahoo, the Google Home page, YouTube, and Facebook.
Our field agents were quite diverse. I am going to skip over gender and age. We have a very diverse group and, more importantly, very diverse politically. Here you see the red, that’s conservatives. These are self‑identified, of course. The blue is liberal. The purple is moderate, and so on. That allows us to see whether bias is being directed toward one group or another.
This is five days before the November 3rd election and including November 3rd. Sure enough, once again, we are seeing in that red ellipse, we are seeing Google bias on the liberal side, and no such bias on the other search engine. This is again just summarizing the difference between what Google was showing, which is highly biased in one direction, and what these other search engines were showing.
The other [search engines] do not have much influence on election, however, because they have hardly any users. The question is: was the biased information just going to liberals last fall? The answer is no. In fact, the content going to conservatives was slightly more liberal than the content going to liberals. For all groups, we were getting liberally biased content.
‘Up-next’ algorithm
Let’s shift over to Google’s home page. Some people saw vote reminders on Google’s home page, some people saw vote reminders for months. That too can shift votes: who was seeing those vote reminders, which could determine who sends in mail ballots, and can determine who registers to vote.
What you see again is that pattern. It is so disturbing. The highest level of those vote reminders was seen by moderates, then liberals, then conservatives at the bottom — fewer vote reminders going to one particular political group. Over time, that has an incredible impact on an election.
We shift over to YouTube. Again, we collected so much data. YouTube’s “up‑next” algorithm, that is scary, let me tell you. Now 70 percent of the videos in the world that people watch are suggested by Google’s “up‑next” algorithm that basically determines what video comes next.
In one of the links from Google, we have a two‑minute video in which the CEO of Google YouTube, Susan Wojcicki, was telling her employees, explaining to them how they were altering the “up next” algorithm to push up content that they think is valid, and to suppress content that they think is not. Of course, you are all aware of that suppression.
Disturbing results. Over 93 percent of the videos that were being suggested to people, of the political videos in the days leading up to the 2020 election, were deliberately biased. Now were those biased videos suggested to liberals? No.
In fact, more were being suggested to conservatives than liberals. Georgia: We went all out, more than a thousand field agents throughout Georgia. We preserved more than a million ephemeral experiences, and we found something completely different.
Notice I have for the first time a little green smiley face in the upper left. That is because we — apparently by going public with our findings in October and November and December of 2020 — we got Google to stay out of Georgia. This shows you bias on the Google search engine, that line, it is practically at zero.
Google turned off the bias like flipping a light switch, and yes, they have that ability. We know that from Google whistleblowers and leaked documents. Again, comparing the search engines, there was no bias in Google in Georgia. Comparing the Google’s bias in the presidential election to Georgia: again, very high bias in presidential and virtually no bias in Georgia.
Not able to stop them
Now, how about those go‑vote reminders, which we saw a lot of in the presidential election. Google sent out no go‑vote reminders in Georgia. This tells you, you see, that you can get these companies to back down.
We have graph that shows you the growth of our monitoring systems from 13,000 ephemeral experiences preserved in 2016 to 1.5 million in 2020 and 2021.
Facebook and Instagram, yes, we were monitoring them, but we were not able to stop them. They were definitely sending out lots of vote reminders in Georgia. We are working now to establish a permanent large‑scale modern system in all 50 states.
That is the only way in my opinion to protect democracy, our children, and human autonomy from manipulation by a big tech, and not just the Google of today, but the Google that comes next.
If you can imagine a control room where we have lots of people connected to maybe tens of thousands of field agents around the country, and where we are looking for shenanigans, bias, and manipulations, and which we are exposing as we find them, these companies will back down, and we will preserve the free and fair election.
More information, there is my email address, mygoogleresearch.com, you can get links to all content, and you can also support my work at that link. I mentioned taming big tech. Then aibrt.org, that’s the institute where I do the research.
* * *
Question: In the swing states you studied on November 3, 2020, what was the differential of votes between Trump and Biden, and what would they indicate? Was that a turnable overall number, or what?
Dr. Epstein: If we extrapolated the level of bias that we saw in those four swing states nationwide and go back a few months, assuming that level of bias had been present for a while, that would have shifted in the presidential election a minimum of six million votes to Joe Biden. These are not small effects. These are huge effects.
This is not like all the nonsense you see about – and I’m sorry if I’m offending anyone – but the nonsense about voter fraud. Those are tiny little incidents compared to what is happening here on a massive scale. That is because Google search results are seen each day in the United States 500 million times.
This is not like stuffing a ballot box with a few hundred ballots. This is a massive manipulation. That is why this monitoring has to be set up. No laws, no regulations will ever stop this. Monitoring can stop it cold.
As far as I know, I’m the only one in the world doing this research and this monitoring. That is because Google is paying off everyone else.
Question: What major changes in tech manipulations have you seen since 2016?
Dr. Epstein: Tech manipulations, the only change I have ever seen was what apparently occurred on the night of October 29th, 2020. That is the day that we went public with some of our findings. Google became aware of the massive amount of monitoring that we were doing this time around.
That night, we saw them literally shut off manipulations that they were engaging in. Unfortunately, we also saw them pressure the New York Post into withdrawing a story about our monitoring system.
They fired the reporter. When we were analyzing data, this was a few days before the election, November 3rd. We decided this time around, we are not going to wait until months after the election. We are going to see what we have now. We did some preliminary analyses. We found very high levels of bias and so on.
I had a personal referral to a reporter at the New York Post, Ebony Bowden. She loved the story. Her editor loved the story. She wrote the piece. She read some of it to me on October 29th.
Then, that night, Google changed what they were showing people. Around 11:57 PM that night, the piece got killed. They asked Google to comment before they went to press. They were supposed to go to press the next morning. It was a big story. It was a big story about tech interfering with the election. It was, I thought, powerfully written.
It got killed that night. Of course, I was dealing with them via email very deliberately knowing that all the New York Post‘s emails are shared with Google, which is also true for the New York Times, by the way, and the Guardian, and hundreds of other major news organizations, and schools, and I could go on and on.
I knew I was sharing all this information with Google; that was deliberate on my part. The piece got killed, and then very shortly after that the reporter was fired.
I tried to contact her recently to find out because I would like to obviously keep in touch, but I do not know where she is right now. I know for sure she was very upset when that piece got killed.
They have tremendous power and they exercise it any and all ways they choose.
Other than that, we got Google to back down with this extensive monitoring, these people just proceed with incredible arrogance. They do anything they want because they are completely unregulated.
What I am studying is occurring at a whole different level from voting machines. It is something people cannot see but that has a tremendous impact on thinking, behavior, and voting, but has nothing to do with voting machines. It has to do with the information people are getting when they are online.
They are getting information when they conduct a search on anything at all. We used thousands of neutral search terms this time around to see what information these companies were supplying to people when they were doing searches. People get vote reminders coming on Google’s Home page or Facebook’s Home page.
In other words, people are subjected to influence all day long on their mobile devices, their phones, their computers. That is what we study. It is a very large‑scale manipulation occurring not just in the US, but these companies are impacting more than three billion people this way every single day, 24 hours a day.
I have been in regular touch with members of Congress now for a number of years. I did testify before Congress. There are some members of Congress, Josh Hawley, Ted Cruz, and a few others, who are very interested in these issues, state attorneys general who are interested. There are 52 of those state attorneys general involved in an antitrust action against Google.
There is government interest, but without any doubt at all, I can tell you that they are not going to solve these problems. The few public officials we have who are concerned about these kinds of things that I have been studying, they are not going to solve the problem.
If the government passes any kind of legislation at all, if they change any regulations at all, it is all going to be completely toothless. A lot of these people are in Google’s pocket. Even conservative organizations. A lot of conservative organizations now are accepting money, large donations, from Google.
There are conservative organizations that have made a deal with the devil because they are accepting money from Google.
It is not in our government’s interest to change this. They must be rather pleased with what’s happening, especially among Democrats. Of course, Republicans generally do not like regulation. They want the companies to have free rein. Generally, that’s true, but you cannot let the water company have free rein. There are certain companies, say supplying natural gas, which you just have to regulate.
When it comes to control over the gateway to all information around the world, I’m sorry but they have to be strictly regulated. It’s not going to happen, however, because Google owns the Democrats and the Republicans do not like regulation.
You cannot separate the politics here from the real issues.
Our children are being brainwashed 24 hours a day by these tech companies. I have five children. I’m very concerned about this. I do not want my kids and grandkids growing up in a world that is literally being run by private tech companies.
It is necessary to establish permanent large‑scale monitoring systems. That is what I am focused on right now.
I am still in touch with members of Congress, with the Attorneys General. Anyone who wants to listen to me, I’m happy to help them. I’m still doing that, but I do not have any hopes at all that it is going to do any good.
They will levy some big fines. France fined Google $270 million. A judge, I think in Ohio, made a very strong judgment against Google. Those things will continue to happen. The EU has levied €10 billion in fines against Google since 2017; it hasn’t affected them at all.
That stuff will continue, but it is toothless. We are talking about a company whose revenues keep growing by double digits. They will bring in this year $160 billion with an extremely high profit margin. They are growing by leaps and bounds, both Google and Facebook, around the world. Laws and regulations are not going to stop them.
Question: Could you remind us please what the number is that you might need to take the big step. The amount of money, it is tax‑deductible, right?
Dr. Epstein: Oh, yes. The nonprofit side is $50 million. We do not need it all at once. To set up the system in all 50 states, that is a $50 million project. Roughly a million dollars per state. That would allow us to keep the thing going for a long time. In other words, part of that money would be going towards development efforts.
Then for people who want to make investments on the for‑profit side, we are willing to work with people because there is room here to establish a for‑profit company, a Hewlett Packard if you want to think of it that way, that provides commercial services for organizations, campaigns, law firms, companies of all sorts that want to monitor these tech companies.
There is room for a big commercial operation as well. Then the commercial operation, the for‑profit, could and over the long term be helping to support the nonprofit, which is where we would do this very sensitive monitoring such as the election monitoring.
Question: Please what is your opinion of the Supreme Court in all this? Clarence Thomas and Justice Alito wanted to see evidence presented. The other judges apparently shot that down. Any idea what they could have been thinking?
Dr. Epstein: It’s hard to understand what’s happening there. Especially, it’s hard to understand their attitudes towards tech. I’m in a waiting mode when it comes to the Supreme Court. Remember, again, Republicans do not like regulation. I’m holding my breath here.
Eventually, when some of the big cases against big tech go to the Supreme Court which is inevitable, I have no idea at the moment what the Supreme Court will do because we are, especially conservatives, are very anti‑regulation. People do not understand that we have to make an exception for these tech companies.
We have to understand that they are as essential now as air and water. They must be regulated. Access to all information, the gateway to all information, that has to be regulated. Our brains are shaped according to the information we receive. We have to protect that process.
Question: Our politicians take an oath of office to uphold the laws of the constitution. Clearly, many are not doing that. Is there anything you say that can be done to hold them to their office?
Dr. Epstein: There are members of Congress who get it, who understand. Then there are people who go with the donors. I can give you an example. Elizabeth Warren was one of the early members of Congress to publish a plan for regulating big tech or “breaking them up.” That is complete nonsense.
You cannot break up the Google search engine, for example. It will not work. You cannot break up Facebook’s social media platform. You would be splitting families in half around the world. By the way, Google is one of Elizabeth Warren’s major donors. I bet you that breakup plan that she proposed was actually developed at Google.
If you “break up big tech,” all you are doing is forcing them to sell off a few of the companies they bought. Google buys on average another company every week. That just enriches them. It enriches the major shareholders.
It does not take away the three big threats: the threat to democracy, the threat to free speech, and of course the manipulation and the surveillance that’s occurring. “Breaking them up” will have no impact on the big threats that these companies pose.
Question: Do you see anything that could be done to ensure that when all these people take the oath of office, it is done seriously?
Dr. Epstein: I am a very optimistic, idealistic person, always have been. When it comes to politics, we have to face the sad truth. People are generally out for themselves. They often put their personal priorities and their desire to get reelected ahead of their ideals, the ideals of this country, or the ideals that are embodied in the constitution. We’ve all seen this over and over again.
I can tell you, by the way, here that a lot of the emails I send to people end up in their spam boxes because most people use Gmail. Google blocks my emails.
At one point, they blocked my access to the Google search engine. I have screenshots showing people what they have never seen before, which is timeouts when I’m trying to connect to the Google search engine. I no longer use any Google products at all.
For those of you who might be interested in privacy issues, you should go to an article of mine, which is at myprivacytips.com. That’s where you’ll see my essay on how to increase your privacy, protect the privacy of your kids. That begins with the sentence “I have not received a targeted ad on my phone or computer since 2014.” I’ve gotten much smarter in how I use technology.
It is a very useful link. Then mygoogleresearch.com, that is where there are links to of my work. Those are the main places for people to get information.
I published an article, I think it was in the Epoch Times, explaining why in general, Republicans cannot win. It is because there are so many different methods available to these tech companies for manipulating opinions and votes without people’s awareness. We have discovered about a dozen of them since 2013.
Right now, we are studying one called the YouTube Manipulation Effect, or YME. We are quantifying the impact of that “up‑next” algorithm. The impact of supplying people with video after video after video, which might, in fact, be very biased, as we found in our election monitoring where highly biased videos were being suggested.
If you think of all of these different techniques and imagine them all being used at the same time, and imagine the major tech companies all having the same politics, we’re talking about an unstoppable force, especially when you throw in with that the fact that 96% of donations from Google go to members of one party.
It happens to be the party that I like, but still how do you stop a force that has access to all of those different means of manipulation and is using them? This is what all the whistleblowers have told us in the past couple of years.
That is what all the leaked documents have shown us, the leaked videos, the leaked emails. It is all very consistent. Yet these people go in front of Congress and they just lie, lie, lie, lie, lie. Congress does not even call them on it.
Question: What are some of the lies, are there examples?
Dr. Epstein: For goodness’ sake, right before I testified, which was in summer of 2019. Google was asked, under oath, I believe it was by Senator Josh Hawley, he was asked, “Does Google have any blacklist?” The Google person said, under oath, on camera, “No, Senator. We do not.”
It was not many months later that Zach Vorhies, who had been a senior software engineer at Google for eight and a half years, walks out of Google with 950 pages of documents and a video, which he sent immediately to the Attorney General of the United States, in which there were blacklists among other things. Actual Google blacklists that are labeled “blacklists.”
Again, think of the arrogance. Shouldn’t you call them something else? I would. They were actually called blacklists. The blacklists are real. I published a big investigative piece in U.S. News & World Report, I think that was in 2016 or so, which is called “The New Censorship.” It was about nine of Google’s blacklists.
I had never seen one, but I know as a programmer that they must exist. It’s one of the simplest ways that they can manipulate people, just by adding names or websites to blacklists, and then making sure that content is suppressed. Very, very simple programming technique.
Again, I wrote about nine blacklists while not having seen them. It was much later that Vorhies walks with actual Google blacklists. Meanwhile, Google’s representatives, Congress says under oath, “No, Senator. We have no backlist.”
Question: What, in your view, can or should be done to ensure a free and fair election in 2022 and beyond? Do you see any of that is possible and able to happen?
Dr. Epstein: I published an article in Bloomberg Business Week the day before I testified before Congress. There is a simple way to end Google’s monopoly on search, their index, their big database they use to generate search results, just make it public.
That is a very simple, very low‑key regulation. There’s ample precedent for it. That would be great, but that would require people in DC getting their act together. I do not think it is going to happen.
The only other method that we have to assure fair elections in 2022 is monitoring. We know how to do it, and on a very large scale. It is just money. It is not even, in the grand scheme of things, much money.
When you consider that we are spending billions now in campaigns, $50 million to get this system going is not all that much. That is how you do it. We not only can collect the data very, very rapidly and analyze it rapidly, we know now that we can get these companies to back off.
Of course, if we have massive amount of data incriminating them, they would be risking not only massive fines, but they would be risking criminal charges. They would be risking going to prison if they did not back off. If they knew these systems were running 24/7, they wouldn’t even try these manipulations. They wouldn’t even dare to do these things.
Question: These big tech companies, is it ideological with them, or they looking for cheap labor? What are they thinking?
Dr. Epstein: It depends on the company. With Google, it is definitely ideology. It was founded by people who are utopians, and utopians can be very dangerous people. They hire people who share their views, their visions, their politics.
One of the leaks from Google, not long ago, was an eight‑minute video called “The Selfish Ledger,” about the ability Google has to re‑engineer humanity according to company values. I can send you a link to a transcript I made of the film, and I can send you a link to the film itself, which leaked from the company.
Ideology is a big part of what Google is doing, no question about it. All the whistleblowers have repeated this. Facebook is harder to figure out because it is run with an iron hand by one man, Mark Zuckerberg. I think he is more interested in money than anything else.
Google is the biggest threat. They have the most power to manipulate, and they do so aggressively and strategically unapologetically. They are extremely arrogant people who think they have the power of gods. You know what? They do, and they exercise it.
Dr. Robert Epstein is Senior Research Psychologist at the American Institute for Behavioral Research and Technology (AIBRT). A Ph.D. of Harvard University, he has been a pioneer in discovering and studying the power that Google and other tech companies have to manipulate our elections, indoctrinate our children, and undermine our autonomy and freedom.