Committee gatekeeping

Why does a bill most legislators would vote for never reach a vote?

Here is a state of affairs that ought to be impossible. A bill exists. If it were put to the chamber, a clear majority would vote for it. It is never put to the chamber. It expires at the end of the session having never been rejected, because it was never considered.

The instinct is to call this corruption or obstruction — someone cheated the majority. Hold that off for a moment and ask a structural question instead. A chamber can only debate one thing at a time. Someone, or something, must decide what that thing is. What powers does that role acquire, purely from being the role, before we know anything about who fills it or what they want?

Begin with the constraint that makes the role necessary. Floor time is finite and rivalrous: an hour spent on one bill is an hour unavailable to every other. Thousands of bills are introduced; a few hundred can be seriously considered. So a queue must be ordered, and the ordering cannot be done on the floor, because ordering the queue would itself consume the queue.

That forces the work outward — into committees, into a rules or business committee, into whoever holds the calendar. Now watch what that arrangement quietly hands over. To pass a bill you must assemble a majority. To stop one, under this arrangement, you need only decline to move it. Ask yourself how many people that takes.

Often one. A committee chair who never schedules a markup, a business committee that never grants floor time, a leader who never brings it up. The two powers are not symmetric, and the asymmetry is not a defect anyone inserted — it falls straight out of the fact that the queue has a keeper.

This is the whole move: the function follows from the structure, not from the preferences of the people inside it. Replace every person in the gatekeeping role with someone of opposite views and the gate is still there, still cheap to close, still expensive to force open.

Now, preferences do decide how the power is used, and political scientists have a good account of it. Cox and McCubbins's cartel agenda-setting theory argues that a majority party's most reliable power is negative — not passing what it likes, but keeping off the floor anything that would split its own coalition. The prediction is testable and largely borne out: the majority party is very rarely "rolled", that is, beaten on a floor vote it allowed to happen. It is rarely beaten because it rarely allows the vote it would lose.

Which explains the case we started with. A bill that would pass with votes from the whole minority plus a third of the majority is exactly the bill leadership least wants voted on. Its passage would be a public demonstration that the governing coalition does not command its own chamber. The gate is closed not despite the fact that the bill would win, but because it would.

Is there a release valve? Usually a formal one that is deliberately made hard. In the US House a discharge petition can drag a bill out of committee to the floor — but it needs 218 signatures, a public and countable act of defiance against one's own leadership, and it very seldom reaches the number. That difficulty is not an oversight either. A gate that any transient majority could open would not perform the sorting function the chamber built it for.

Some refinements the reasoning needs.

First, gatekeeping is not one device. It is a family: the chair who never schedules, the rules committee that grants a closed rule barring amendments, the leadership norm of not advancing a measure lacking majority support within the majority, the government's control of the order paper in a Westminster chamber that leaves most private members' bills to die for want of time. Different shapes, same underlying asymmetry.

Second, do not conclude the arrangement is simply pathological. Committees also do work the floor cannot: hearings, expertise, amendment, and the filtering out of the genuinely unserious. The gate is a by-product of a division of labour that the chamber has real reasons to want. The honest formulation is that specialised pre-floor stages buy competence and pay for it in veto points.

Third, resist the tidy conclusion that gatekeeping simply thwarts the median legislator. Sometimes it protects them. A member may be relieved never to cast a recorded vote on a bill that is popular at home and disastrous in coalition. Silence on the floor is a service to some of the people it appears to frustrate — which is part of why the release valves stay unused.

• Negative versus positive agenda power, and why leaders in most chambers hold far more of the first.
• Whether committee expertise can be preserved while lowering the cost of overriding a gate.
• Comparison with chambers where the government owns the timetable outright, and where the same power shows up under a different name.

Every term the collection defines is gathered in the glossary.