Veto Overrides: The Hardest Vote in Politics

A veto override is the legislature’s constitutional last word, proof that, pushed far enough, the lawmaking branch outranks the executive. It is also, vote for vote, the hardest thing a legislature ever tries to do. The arithmetic is unforgiving, the politics are worse, and the attempts that succeed are statehouse legend precisely because most don’t.

The arithmetic

A bill passes with simple majorities. An override re-passes it over the governor’s objection, and the price goes up: two-thirds of each chamber, in most states and in Congress (a few states set it at three-fifths). In a 100-member house that means 67 votes where 51 sufficed, and both chambers must clear their bar. Fail in either and the veto stands.

Read the fine print, too. “Two-thirds of members elected” and “two-thirds of members present” are different numbers on any day with absences, and real overrides have lived and died in that gap. Override managers know exactly which rule governs their chamber, and exactly who’s in the building.

Why the politics are harder than the math

Here’s the structural problem: a bill that passed 60–40 looks five votes short of 67, but those aren’t the hard votes. The hard votes belong to the governor’s own party, members being asked to overrule their party’s leader, in public, on a roll call the governor will read with names attached.

The governor’s office works that list relentlessly between veto and override vote: appropriations for the district, appointments, fundraising, primary-season memory. The lobbying runs one direction: the governor only needs one-third-plus-one in one chamber, so the override whip count bleeds. Members who voted yes on passage discover schedule conflicts, “concerns about the drafting,” newfound respect for executive prerogative. The pattern in the data is blunt: most state overrides fail, and most fail by one or two votes. (Congress has overridden roughly 7% of regular presidential vetoes in its entire history.)

The exception proves the mechanism: in states where one party holds supermajorities, overrides become almost routine: the math pre-exists, so the governor’s veto is just a speed bump. Everywhere else, the override is a campaign for two or three specific votes.

The override campaign

Smart sponsors start counting before the veto exists: sometimes the override math shapes final passage itself, packing the margin past two-thirds precisely to warn the governor off. Once the veto lands, the campaign has three fronts:

• Timing. Rules set the window; managers pick the day. Every absence matters when the denominator might be “members present.”
• The two or three. Override fights aren’t broadcast persuasion. They’re focused, one-member-at-a-time campaigns. What does this member need, fear, owe? District pressure is the classic lever: the governor is far away; the people in the gallery are from home.
• The story. “The governor vetoed insulin caps” is an override narrative. A drafting dispute isn’t. Public awareness determines whether defecting from the governor reads as betrayal or as courage.

Win, and the bill becomes law over the executive’s signature: the system’s rarest outcome and its purest demonstration that a determined-enough legislature answers to no one. Lose by a vote, and you’ll replay every traded favor and every burned bridge of the whole session, looking for the one decision that would have changed one name. That postmortem, veterans will tell you, is where the next session’s victory gets drafted.