KAKAMPINK TROLLS? I WENT BACK TO THE DATA.

I have to start with an admission.

This topic came from a request by a commuter and transportation page. They asked me if I could look into “Kakampink trolls.”

My first reaction was to push back a little. I do not like spoon-feeding people with information.  Part of what I am trying to do with MCT is help people develop the habit of reading, researching, comparing sources and putting information together for themselves.

I would rather have readers who know how to look for an answer than readers who have to wait for me to hand them one.

The page replied that political research is not really their specialty. Fair enough.

Research is a skill, and political disinformation research is an even narrower one. I am not going to judge a commuter and transportation page according to how well it can investigate political influence operations.

Parang hinusgahan ko ang isda dahil hindi siya makalipad nang malayo.

So I took the request.

The question caught my attention for another reason: I have researched this before.

I already had studies, fact checks, platform reports and notes from the 2022 election.

I just never turned them into an article because what I had seen then did not look like the same pattern of publicly documented fake-account and disinformation operations I was finding around pro-Marcos, pro-Duterte and Uniteam networks.

That did not mean I thought Kakampinks were above criticism.

I had already spent a lot of time calling out our own behavior after 2022: the condescension, dogpiling, moral superiority and tendency to treat people who voted differently as stupid or beneath us.

I was a Kakampink in 2022. I am not one now. I realized that if I wanted to be genuinely pro-Philippines, I had to shave off my political color and be willing to criticize every camp, including the one I once belonged to.

I had also been thinking about something else: coordination itself is not the problem.

People fighting disinformation can organize, respond quickly and amplify verified information.

The line is crossed when coordination turns into deception through fake identities, hidden sponsorship, manufactured support or information you know is false - exactly what a Uniteam supporter was doing then.

So I went back to the data.

BEFORE I CALL ANYONE A TROLL

Filipinos use the word “troll” so loosely that it can become almost useless.

Someone disagrees with you, troll. Someone insults your candidate, troll. A thousand people post the same hashtag, troll army. Someone spends all day defending a politician, bayaran. No profile picture, fake account.

Any of those suspicions can turn out to be correct. None proves anything on its own.

A real supporter can be rude, dishonest, gullible or obsessive without belonging to an organized operation.

A volunteer digital activist can coordinate hashtags, share campaign graphics, promote rallies and ask other supporters to amplify posts without pretending to be somebody else.

Then there is brigading. Real users can collectively report, reply to or pressure a target. The activity can be coordinated while the people involved remain genuine individuals using genuine accounts.

A paid political influencer is another category. The person may be real and may post under a real name. The question is whether the political relationship or compensation is disclosed.

A fake-account operator goes further. One person may control several identities and use them to create the appearance that many unrelated people independently hold the same opinion.

There is also a difference between misinformation and disinformation.

Misinformation can be false or inaccurate information shared by someone who believes it.

Disinformation involves deliberate deception.

Intent is harder to prove, which is why I am not going to label every false pro-Robredo post as disinformation simply because it was wrong.

Meta uses the term coordinated inauthentic behavior for operations that work together while misleading users about who they are or what they are doing. Fake accounts are often central to that kind of activity.

Then there is the biggest claim of all: campaign control.

If I find 500 fake accounts supporting Leni Robredo, I can say I found 500 fake accounts supporting Leni Robredo. I cannot jump from there to “Leni Robredo operated them” without evidence connecting those accounts to her campaign.

The same rule applies to Marcos, Duterte or anyone else. Support is one claim. Coordination is another. Payment is another. Fake identity is another. Campaign control is another.

If we throw all of those into one bucket and call them “trolls,” we stop investigating and start insulting each other.

THE KAKAMPINK VOLUNTEER MACHINE WAS REAL

The first thing the evidence shows is that the Robredo volunteer operation was highly organized.

Before the filing of certificates of candidacy in 2021, PCIJ found at least 47 Robredo-supporting Facebook pages that had paid for political ads. Thirty-three of those pages were managed by only two individuals.

After the filing period, PCIJ found another 64 support pages buying ads for Robredo, more support pages than it had identified for any other presidential candidate at that stage. Most openly disclosed that they supported Robredo.

Robredo spokesperson Barry Gutierrez also told PCIJ that volunteers coordinated with the campaign team either to boost official posts or produce their own materials.

During the official 90-day campaign, PCIJ’s review of Facebook Ad Library data found ₱58.75 million in advertising supporting Robredo. Around ₱11 million came from her official pages, while much of the rest came from supporters.

That is a large and organized political operation. It is not, by itself, evidence of a troll farm.

And I have no problem saying that because I believe people pushing for accountable government should organize too.

If thousands of real people openly coordinate around factual information, I do not suddenly call them trolls because they support a politician.

The harder questions begin when we look at how some of that organization was used.

THE PART KAKAMPINKS SHOULD NOT DENY

Some Kakampink groups engaged in brigading.

A Cambridge University Press study on KpopStans for Leni documented a March 9, 2022 “Mass Report Hour.”

Participants circulated lists of Twitter and Facebook accounts they believed were spreading false information, shared reporting instructions, posted screenshots after reporting targets and celebrated when some accounts were suspended.

The Parallel Public Spheres research on the 2022 election also discussed this kind of collective targeting as brigading.

I understand the instinct behind it. If you believe an account is spreading lies during an election, reporting it can feel like public service. But once a group organizes people around selected targets, we are no longer talking about unrelated users independently deciding to report something.

The participants may have been genuine volunteers using genuine accounts. The activity itself was still coordinated pressure against selected targets.

And yes, some Kakampinks were simply assholes online. There was condescension, dogpiling and the habit of treating people who voted differently as if they were stupid, uneducated or beneath the people correcting them.

Being on the side of good governance does not give anyone permission to mock someone’s English, parade their mistakes for applause or turn disagreement into public humiliation.

You can be correct about a politician and still behave badly toward another Filipino.

YES, FALSE INFORMATION CAME FROM SOME PRO-ROBREDO ACCOUNTS TOO

There is no reason to pretend otherwise.

AFP debunked posts claiming Robredo’s February 2022 Naga rally was the biggest campaign rally in Philippine political history. The posts attached supposed crowd estimates to the Philippine National Police and the Associated Press. Representatives of both organizations denied issuing those figures.

VERA Files also documented posts falsely presenting a 2019 photograph of journalist Atom Araullo holding a Gabriela Women’s Party placard as proof that he supported Robredo in 2022. Araullo himself confirmed that the photo came from a Gabriela event he was covering and had nothing to do with Robredo’s presidential campaign.

So if the claim is simply that false information also circulated from pro-Robredo accounts and pages, the answer is yes.

The harder question is how much of it there was and how it compared with false information moving in the other direction.

Tsek.ph’s second election analysis covered 685 fact checks published through the end of April 2022. Among false or misleading claims involving Ferdinand Marcos Jr., 92 percent were favorable to him. Among disinformation involving Robredo, 96 percent was negative toward her.

That dataset is not a census of every false post produced during the election. It reflects claims checked by Tsek.ph’s partner organizations, so it has limits.

Still, it does not show an evenly divided disinformation environment. False pro-Robredo content existed, while the available fact-checking data also showed Robredo absorbing overwhelmingly negative disinformation and Marcos benefiting overwhelmingly from favorable false or misleading claims.

Both can be true at the same time.

THE FINDING THAT CHANGED MY EARLIER VIEW

This is where my old impression needed correction.

A 2023 Internews study looked at covert political influence during the 2022 election across Facebook, Twitter, YouTube and TikTok.

Researchers identified 1,425 influencer accounts they assessed as likely involved in covert political campaigning, representing at least 584 unique influencers across platforms.

The researchers also modeled the possible financial value of commissioned influencer work.

In their model, influencer activity supporting Marcos was valued at around ₱351 million, Robredo at around ₱135 million, Isko Moreno at ₱42 million and Manny Pacquiao at ₱1 million.

I would be dishonest if I left Robredo’s number out because it complicates the image of a purely volunteer Pink campaign. I would also be dishonest if I turned that into, “Leni Robredo spent ₱135 million on trolls.”

The researchers warned against reading the estimates that way. They did not have data confirming that every influencer who openly supported a candidate had actually been commissioned by that candidate or campaign.

What the study gives us is narrower. There is serious evidence of a covert political influencer industry during the 2022 election. Accounts supporting Robredo were part of the researchers’ model, and the estimated value attached to pro-Robredo influence activity was substantial.

That was the part I had not given enough attention before. The Pink online operation had a huge volunteer base, but volunteer activity was not the whole picture.

PAYMENT STILL DOES NOT TELL US WHO CONTROLLED THE OPERATION

A real influencer can be paid to promote a candidate while using a real identity. That raises questions about disclosure and campaign spending, especially when audiences are not told that political content was commissioned.

A fake-account network is a different operation.

Meta’s reporting before the 2022 election shows how commercial this business can become. The company said it removed around a dozen clusters focused on fake engagement. In one case, a social-media management agency used more than 700 accounts to post and share political and entertainment content.

Meta also found people running inauthentic engagement activity for several candidates during the same election. Some operators are simply selling a service.

So an account promoting a politician does not automatically tell us who owns it, who pays it or who controls it.

WHERE THE EVIDENCE STOPS

After going through the platform reports and research I could find, I did not find a Meta takedown or comparable platform finding that tied a coordinated fake-account network to Robredo’s official campaign.

That does not prove that no such operation ever existed. It means I cannot claim that one did based on the evidence I found.

Political impersonation also makes attribution difficult. A fake account can pretend to support anyone. A paid operator can work for several clients. A volunteer can look like a professional operator, while a professional operator can pose as an enthusiastic volunteer.

A pink profile photo proves very little. So does a hashtag.

WHY I DID NOT PUT THE KAKAMPINK AND DDS OR UNITEAM OPERATIONS IN THE SAME BOX

In 2019, Meta removed 200 Philippine Facebook and Instagram pages, groups and accounts for coordinated inauthentic behavior.

Meta said the people behind the activity used both authentic and fake accounts, concealed who they were and were linked to a network organized by Nic Gabunada, who had led social-media work for Rodrigo Duterte’s 2016 presidential campaign.

During the 2022 presidential race, Twitter suspended more than 300 accounts from what it described as the Marcos supporter base for violating its platform-manipulation and spam policy.

The Marcos camp questioned whether those accounts actually belonged to its supporters, and Twitter did not say the official Marcos campaign controlled them.

That qualification stays. I cannot demand strict proof when I look at Kakampink activity, then throw that standard away when I look at Marcos supporters.

But when I put those platform actions beside Tsek.ph’s fact-checking data and the research on fake-account operations in Philippine politics, I still do not see evidence showing identical operations on the two sides.

That was my reading before, and most of it survived this second look. What changed was my understanding of the Pink side because I had treated it too heavily as a volunteer operation.

There were real volunteers. There was brigading. There was false information. There was covert influencer activity.

What I still did not find was public evidence proving that the official Robredo campaign ran a coordinated fake-account troll farm.

SO, WERE THERE KAKAMPINK TROLLS?

It depends on what you mean.

If you mean real Kakampinks who harassed people, joined pile-ons, mass-reported accounts, mocked voters or spread false claims, yes. Those people existed.

If you mean organized Kakampink digital activists, yes. That is documented.

If you mean coordinated brigading, yes. Researchers documented it.

If you mean covert influencers supporting Robredo, there is serious research supporting that category, with a very large warning attached to the spending estimates.

If you mean fake-account operators working for the official Robredo campaign, I did not find enough evidence to state that as fact.

And if somebody has better evidence, I want to see it.

My standard has not changed. Organize. Respond quickly. Amplify verified information. Work together against disinformation. But do not manufacture people who do not exist, hide political sponsorship, fabricate evidence or knowingly spread false information.

If a Kakampink crosses that line, call it what it is.

Apply the same standard to everyone else.