Skip to content

Two Million Ballots Named No President

Two million ballots named no president. Nobody talks about them.

Everyone remembers who won in 2022, and by how much. I went looking at the numbers around the edges instead, and found one I had never seen anyone talk about.

Two Million People Who Voted For Nobody

Turnout was 82.98 out of every 100 registered voters. That is very high. People showed up.

But showing up and picking a president are not the same thing.

2,213,371 ballots came back with no valid vote for president. Blank, or marked in a way the machine could not read.

2,213,371
Ballots that named no president

That is more people than voted for the man who came fourth.

Then I looked at the line right below it on the same sheet. The one for vice president.

3,682,855 blank or spoiled there. That is 1,469,484 more than for president.

Same people. Same piece of paper. Same day. Nearly one and a half million more of them chose to leave the second line alone.

I cannot tell you why from this data. Maybe they liked nobody. Maybe they only cared about the top job. But it was a choice, and it is bigger than most candidates' whole vote.

The Margin

The winner got 31,629,783 votes. Second place got 15,035,773.

That is 2.1 votes for every one. A gap of 16,594,010 people.

It was not close anywhere except one place. Out of 18 areas counted, the winner led in 16.

His best was Region II, up north, with 86.2 out of every 100 votes. His weakest was Region V, Bicol, with 17.7 — the runner-up's home region, and the only place the whole national picture flips over.

Where I Got These, And Why You Should Squint

This is the shakiest source on my whole site. I want to say that plainly instead of hiding it at the bottom.

The real record is a document from Congress. COMELEC has one too. I tried to fetch both with a script. Both said no — they answer people with browsers and turn away programs.

So I used Wikipedia's copy of that record. Wikipedia copied it carefully and says where it got it. But it is still a copy of a copy.

Three things make that less bad than it sounds.

First, Wikipedia pages change. So every number I saved carries the exact version number of the page it came from. Anyone can go back and look at that same version.

Second, I printed the links to the real documents on the page. My script cannot open them. Your browser can.

Third, and this is the useful one: I made the numbers check each other.

The Checking Caught Three Real Mistakes

Each region's votes have to add up to that region's total. Simple rule. It caught three bugs I would never have spotted by reading.

The first: the winner's number in each region is wrapped in special formatting to make it white. My cleaner threw away the formatting and the number inside it. So the next number slid over into its place. The chart showed Marcos losing his own home region. That one at least looked wrong.

The second was sneakier. Some cells carry a colour code like #fe18a3. My reader pulled the digits out of the colour and glued them onto the vote count. One region showed 1.8 billion votes in a country of 110 million people.

The third was the quiet one. Percentages like 84.69 lost their dot and became 8469. Nothing looked broken. Things just stopped matching.

None of those three would have shown up by staring at the page. The adding-up rule caught all of them.

And The Source Itself Has Mistakes

Once the reading was right, the rule caught something else. It was not my bug this time.

In three regions, the "everyone else" column does not match the percentage printed next to it. Region IV-B says 7.44 out of 100, but its own numbers work out to 12.94.

I did not quietly fix those. I worked that column out from the total instead, and wrote the disagreement into a file so it stays on the record.

There is one more. The regional table only accounts for 97.8 out of every 100 votes. About 1.18 million votes sit in no region at all.

So I never add the regions up and call it the country. They do not reach it.

What I Left Out

The old page had senators, party lists, province maps, turnout per region, and charts of voting machines.

I could not reach real numbers for any of it. So it is gone, and I wrote down where you would have to go to get it.

The page is much shorter now. Everything left on it can be checked.