How the Veto and Veto Override Work

Article I, Section 7 of the Constitution lays out the process. Once a bill passes both the House and Senate in identical form, it goes to the President, who has ten days, not counting Sundays, to act on it. This is the same final step this site's How a Bill Becomes Law page covers as the last stage a bill has to clear.

The President has three options within that window: sign the bill into law, veto it, or simply take no action. If Congress remains in session and the ten days pass without the President signing or vetoing, the bill becomes law anyway without the President's signature, an outcome that has happened but is uncommon since presidents generally prefer to take a clear position.

A veto is not just silence. The Constitution requires the President to return the bill to the chamber where it originated along with a written statement of objections, formally called a veto message, explaining specifically why the bill was rejected. That written objection becomes part of the public record and is what Congress responds to if it attempts an override.

Once a vetoed bill is returned, the originating chamber has to decide whether to attempt an override, covered in detail in the next tab, or let the bill die where it stands. Either way, the veto itself does not permanently end the underlying policy debate, Congress remains free to revise the bill and try again in a future session.

A pocket veto gets its name from the idea that the President simply pockets the bill, taking no action at all, rather than issuing a formal veto. It applies specifically when Congress adjourns during the ten-day window in a way that prevents the bill from being returned to it, which under the Constitution means the bill does not become law.

The key practical difference from a regular veto is that a pocket veto requires no written veto message and, more importantly, gives Congress no bill to vote on for an override. Because the bill never comes back to the chamber that passed it, there is nothing left for Congress to hold an override vote on, the pocket veto is final in a way an ordinary veto is not.

Exactly what counts as an adjournment severe enough to trigger a pocket veto, as opposed to a routine short recess during which the bill could still technically be returned, has been disputed and litigated across multiple administrations. Modern Congresses rarely take the kind of extended, session-ending recess that made pocket vetoes common in earlier eras, which has made the tool less central to modern practice than it once was.

Because of that legal ambiguity, recent presidents have generally been cautious about relying on a pocket veto when a regular veto is available, preferring the clearer, harder-to-dispute path of returning the bill with formal objections. Disputes over specific pocket vetoes have occasionally ended up in federal court, with results that have not fully settled every edge case of when the tool legitimately applies.

A line-item veto would let the President strike out specific provisions, often individual spending items, from a bill while signing the rest into law. That is fundamentally different from the regular veto's all-or-nothing choice, where the President can only accept or reject an entire bill exactly as Congress passed it.

Congress gave the President this power for the first time with the Line Item Veto Act of 1996, and President Clinton used it 82 times in 1997, the one full year it was in force, mostly to strike individual spending provisions and a small number of tax breaks from larger bills before signing the remainder into law.

The Supreme Court struck the law down in Clinton v. City of New York (1998), ruling 6 to 3 that it was unconstitutional. The Court held that letting the President unilaterally cancel individual provisions of a bill Congress had already passed amounted to unilaterally amending or repealing a duly enacted statute, a power the Constitution's Presentment Clause in Article I, Section 7 does not give the President.

Since that ruling, every presidential veto has applied to an entire bill, never to a single line or provision within one. Proposals to revive some version of a line-item veto, generally through a constitutional amendment rather than an ordinary statute this time, have been introduced in Congress periodically since 1998, but none has come close to the supermajority support a constitutional amendment requires.

To override a veto, the chamber where the bill originated votes on the exact bill again, and this time needs two-thirds of the members present and voting, rather than the simple majority that passed the bill the first time, to succeed. That is a significantly higher bar than ordinary legislation requires, and it is deliberately set that high under Article I, Section 7.

If the originating chamber clears that two-thirds threshold, the bill moves to the second chamber, which must separately reach the same two-thirds bar on its own vote. Only if both the House and Senate independently clear that supermajority does the bill become law over the President's objection, a single chamber succeeding is not enough on its own.

Because the threshold is so high, a successful override generally requires support well beyond the party that controls Congress, typically including a meaningful number of the President's own party crossing over to vote against the veto. That cross-party requirement is precisely why overrides remain uncommon even in Congresses that pass a large number of bills the President later vetoes.

If either chamber falls short of two-thirds, the override attempt fails and the veto stands, ending that specific bill unless Congress chooses to revise it and start the legislative process over from the beginning. There is no partial override and no compromise vote, and in practice a failed override vote ends the bill.

Presidential vetoes themselves are common, presidents have issued thousands of them across American history, but successful overrides are comparatively rare precisely because assembling a two-thirds supermajority in both chambers is difficult under ordinary political conditions, especially when a president's own party controls a substantial share of seats.

The War Powers Resolution of 1973 is one of the most frequently cited modern examples. Congress passed it specifically to limit a president's ability to commit U.S. forces to extended combat without congressional authorization, and the override succeeded over President Nixon's veto only because it drew significant support from members of Nixon's own Republican Party in addition to the opposition.

The Taft-Hartley Act of 1947, which restricted certain labor union practices, is an earlier example from the other party's presidency. Congress overrode President Truman's veto with substantial crossover support from congressional Democrats, showing the same pattern, an override succeeding only with backing that crossed the sitting president's own party line.

Franklin D. Roosevelt holds the record for the most vetoes issued by any president, 635 across his twelve years in office, and Congress overrode only a small handful of them, commonly cited as nine. That an unusually high-volume user of the veto still saw the overwhelming majority of his vetoes stand illustrates just how hard the two-thirds bar is to clear, even across a presidency long enough to generate hundreds of opportunities to try.