An actual storm — not the “No” campaign barnstorming of former Colombian president Álvaro Uribe — may have decided the fate of peace in Colombia. If Hurricane Matthew-induced rains had not soaked the Caribbean coast, the math says that the “Yes” vote would have carried the day by almost 90,000 votes.

Voter turnout was surprisingly low across the country. Sunday’s plebiscite to ratify a peace accord with FARC was arguably the most important vote in Colombia in a half-century. But only 13.1 million of the 35 million Colombians who were eligible to cast their ballot (37.4%) went to the polls, according to the National Civil Registry. This was a huge drop from the 15.8 million who voted in the last presidential election.

hurricane-matthew-colombia-vote-3

So with the “No” campaign winning by a mere 53,894 votes — in a 50.21% to 49.78% victory over “Yes” — it’s hard to discount the impact that the Category 5 Hurricane Matthew had on the plebiscite and thus the future of the nation.

While the vicious storm didn’t make landfall in the country (almost no hurricane ever does), its rains “triggered heavy flooding,” washed out at least two roads, killed one, and forced evacuations on Colombia’s Caribbean coast, according to the Associated Press.

At least 62 neighborhoods in the city of Cartagena alone were affected by flooding and other damage, with 5,000 families who “lost everything,” said Carlos Ivan Marquez, the nation’s disaster risk unit director. Almost 9 inches (224 mm) of rain fell on the city within 24 hours. The department of Magdalena declared a state of emergency and officials in Bolivar have had to hand out some 4,000 food kits and 1,200 blankets, hammocks, and mosquito nets.

If Mother Nature’s wrath hadn’t chosen this weekend to strike, the voter turnout in the region would almost certainly have been higher.

There are two large cities on Colombia’s Caribbean coast. Blue-collar Barranquilla is the country’s fourth-most-populated municipality, while the sun-kissed tourist dream of Cartagena comes in fifth. Just 24.9% of eligible voters turned up at the ballot in Barranquilla, and a staggeringly low 20.1% voted in Cartagena.

It wasn’t just the big cities. Across the whole coast, voters stayed home. Four of the six Caribbean departments (not including Antioquia) had a turnout rate below 25%. The other two, Sucre and Cordoba, still came in well below the national average at 34.4% and 31.7%, respectively.

Yet those who did go to the booth in these six departments overwhelmingly voted “Yes.” More than 60% chose to ratify the accord in all six jurisdictions, and some simple math shows that a better turnout would have changed the outcome.

hurricane-matthew-colombia-vote-2

To push the national results in the other direction, the Caribbean departments wouldn’t have even needed a strong turnout. Just by having the same low turnout seen across the nation — 37.4% — those six locations would have finished with 1,478,499 “Yes” votes compared to 971,914 on the “No” side. (This presumes the same “Yes”/”No” ratios seen in the actual plebiscite.)

That theoretical “Yes” margin of victory on the coast (506,586 votes), compared to the actual margin on Sunday (362,887), would have tipped nationwide results in the other direction by 143,699 votes. That is more than enough to change an outcome decided by just 53,984 votes.

If Hurricane Matthew didn’t strike, “Yes” would have won by 89,805 votes.

It wouldn’t have even taken a higher turnout everywhere. Just by getting a 37.4% turnout rate in two departments — Atlantico and Bolivar, the homes of Barranquilla and Cartagena — “Yes” would have won by 35,217 votes.

While its impossible to attribute the low turnout to the hurricane alone, it certainly looks like the biggest culprit. So as it turns out, Colombians didn’t really want to vote against the peace deal. They just stayed home due to rain.