Data Stories · Congress

The Attention Budget That Wasn't

A tempting theory says Congress has a fixed amount of attention, so every crisis it takes up must crowd something else out. Twenty-five years of hearings say otherwise — and the failure is the finding.

Who this data represents Published, transcribed U.S. House and Senate committee hearings (via CoCoHD/GovInfo), 1998–2022, counted by committee and month. This is a pre-registered negative result: we tested a specific model of congressional attention and report that it failed, along with what actually holds.

Here is a theory with the ring of truth: legislative attention is scarce and roughly fixed. There are only so many hearing days, so many staff, so many members. When a crisis forces its way onto the schedule, it has to displace something — the crowded-out committee loses its slot. Attention is a budget, and shocks reallocate it. It is a clean idea, and we pre-registered four predictions from it. Three of the four were wrong, and the ways they were wrong are more interesting than the theory.

0.70

month-to-month variation (CV) in hearing volume — a budget would be flat

2 of 7

datable shocks that visibly moved the responsible committees

46%

of 2019's hearings held in pandemic-year 2020

expands

what attention does under a real crisis — it does not reallocate

The calendar rules the calendar

A fixed budget would show up as a roughly flat monthly total. Instead the total is a sawtooth. Hearing volume swings by a coefficient of variation of 0.70 within every single Congress — collapsing to near zero every August and December recess, spiking when the chambers are in session, and varying nearly two-fold from year to year. The dominant force in the congressional schedule is not the news. It is the congressional schedule.

Figure 1 · Hearings per month, House and Senate

Monthly count of committee hearings, 1998–2022

This is not a fixed budget. The recess sawtooth dominates everything, and the one unmistakable break is 2020, when the pandemic cut the schedule roughly in half. If attention were a fixed pie being resliced, this line would be flat. It is anything but.

Most crises never register

For each of seven datable shocks we named the responsible committees in advance — the ones whose jurisdiction the crisis fell under — and asked a simple question: in the six months after the shock, did those committees take a bigger share of their chamber's hearings than in the year before? For five of the seven, the answer is no. The Deepwater Horizon spill, the Flint water crisis, Cambridge Analytica, the Enron collapse, even January 6th — none produced a detectable rise in the responsible committees' share of the calendar in the months that followed.

Figure 2 · Did the responsible committees' share of hearings rise after each shock?

Responsive committees' share of chamber hearings, 6 months after vs. 12 months before (ratio)

Only two shocks clear the bar. Hurricane Katrina and the 2008 financial crisis raised the responsible committees' share by more than 40%. The rest sit at or below no change — because a modern scandal produces one marquee hearing, not a wave of them, and because the response often lags past the window or lands in a special committee that didn't yet exist.

This is less surprising than it first seems. When Facebook's chief executive testified about Cambridge Analytica, that was one hearing, extensively covered — not twenty. The January 6th investigation that consumed 2022 ran through a select committee created for the purpose, months after the attack, not through the standing committees whose jurisdiction it nominally touched. Hearing counts measure the machinery of sustained oversight, and most shocks, however loud, don't spin that machinery up. They get a hearing, not a season of them.

When attention does move, it expands

The two shocks that did register — Katrina and the 2008 financial crisis — reveal the second broken prediction. Under the budget theory, the responsible committee's gain should come at some other committee's expense. It didn't. After Katrina, homeland-security hearings didn't just take a larger slice; the whole chamber's hearing volume rose — the responsible committees expanded and the others held roughly steady. A genuine, sustained crisis doesn't make Congress trade one subject for another. It makes Congress hold more hearings. Attention is not a fixed pie. It is elastic, and a real emergency stretches it.

The third prediction — that crowd-out would fall on low-salience committees like Agriculture or Veterans' Affairs — has nothing to crowd out once the budget premise fails, and we found no such displacement.

The one shock that emptied the calendar

Only one event in twenty-five years did what a shock is supposed to do to a budget: overwhelm it. In 2020 the pandemic cut congressional hearings to 46 percent of the prior year's — not by reallocating attention but by making the ordinary in-person hearing impossible. It is the exception that proves the rule in the strict sense: the schedule is not defended by a fixed budget that redistributes pressure. It simply runs on its calendar until something makes the calendar impossible, and then it empties.

None of this says Congress ignores crises — it says hearings, counted this way, are a coarse instrument for measuring attention, and that the tidy economics of a fixed budget is the wrong model for how a legislature schedules itself. The honest picture is messier and more institutional: a calendar-bound body that mostly runs on routine, expands for the rare sustained emergency, marks most shocks with a single hearing, and can only be stopped by a pandemic. We went looking for a budget. We found a calendar.