Senate Votes 89–4 to Push the Funding Cliff to December 11 (August 2026): The Seven-Day Gap That Contains the Fed's Last Meeting of the Year
The Senate advanced a stopgap 89–4 to fund the government to 11 December. The House's version stops on 4 December — and the gap between them holds the Fed's final meeting.
Senate Votes 89–4 to Push the Funding Cliff to December 11 (August 2026): The Seven-Day Gap That Contains the Fed's Last Meeting of the Year
On Monday 3 August 2026 the Senate voted 89–4 to clear the procedural hurdle on a bipartisan stopgap that would fund the federal government through 11 December, moving on a funding deadline that is still nearly two months away. The House had already passed its own version on 21 July by 220–205, running through 4 December, and then left town until 31 August. Two chambers, two dates, one week apart — and inside that week sit the Federal Reserve's final meeting of the year on 8–9 December, complete with a fresh Summary of Economic Projections, and the November Consumer Price Index on 10 December. Markets did not react to any of this: the S&P 500 and the Dow closed at records on 4 August on Strait of Hormuz headlines, the dollar index sat near 99.9, and the 10-year yield held around 4.63%.
That non-reaction is the right starting point, not an oversight. A funding lapse is one of the few political events with an almost entirely non-credit transmission path, and the channel it does run through — the flow of official statistics — is slow, indirect, and easy to underweight until a central bank has to make a decision without a number it normally has.
- The Senate advanced a stopgap 89–4 on Monday 3 August 2026, funding agencies at current levels through 11 December. The House passed a rival version through 4 December on 21 July by 220–205 and is in recess until 31 August.
- Nothing is law yet. Current funding runs out on 30 September, and identical text must clear both chambers and be signed before that date to prevent a lapse on 1 October.
- The seven-day gap between the two dates is the whole story: the FOMC meets 8–9 December with a Summary of Economic Projections, and November CPI is scheduled for 10 December. The Senate date puts both inside funded government; the House date does not.
- A shutdown is not a default — debt interest is paid under a permanent appropriation. The CBO put the 35-day 2018–19 lapse at about $11bn of output, roughly $3bn of it permanent.
- The channel that actually matters is data. The record 43-day lapse of October–November 2025 meant October's unemployment rate was never collected, and the January 2026 jobs report was not published on schedule during the four-day February lapse.
- Markets ignored it. Records on the S&P 500 and Dow on 4 August came from Hormuz de-escalation headlines, not from Congress — dollar index near 99.9, 10-year around 4.63%.
- See how the rate, growth and risk factors are scoring the US dollar against the other seven majors on the live meter.
What actually happened, and what it does not settle
Federal funding for fiscal 2026 runs out on Wednesday 30 September, which would put agencies into a lapse on 1 October. Congress has moved unusually early on it. The House passed a continuing resolution on 21 July 2026 by 220–205 extending current spending levels through 4 December, then began its summer recess. On 2 August the Senate Appropriations leadership released a bipartisan alternative running a week longer, to 11 December, and on Monday 3 August the chamber voted 89–4 to limit debate and take up the vehicle — a margin that tells you the votes for passage exist several times over.
The Senate text keeps agencies at current levels with a handful of exceptions, and carries surface transportation authority to the same 11 December date; Roll Call has the itemised contents. Neutral coverage of the deal itself is at NBC News.
What none of this does is make anything law. A continuing resolution has to pass both chambers in identical language and be signed. The House is away until 31 August, which leaves the September calendar to resolve a difference that is, on its face, seven days on an expiry date. That is an easy problem by the standards of the last two years — and it is worth being explicit that "easy" is a statement about the size of the gap, not a forecast about whether it closes.
Two dates, two calendars
Here is why a one-week difference is not a rounding error. Both proposals push the next cliff past the 3 November midterm elections, which is the acknowledged point of a short patch. But they land on opposite sides of the densest week in the fourth-quarter macro calendar.
| House version | Senate version | |
|---|---|---|
| Funding runs through | Fri 4 December 2026 | Fri 11 December 2026 |
| Status | Passed 21 July, 220–205 | Procedural vote 89–4, 3 August |
| November jobs report (Fri 4 Dec, 08:30 ET) | Last funded day | Funded, 7 days of margin |
| FOMC decision + SEP (Tue–Wed 8–9 Dec) | Beyond the cliff | Funded |
| November CPI (Thu 10 Dec, 08:30 ET) | Beyond the cliff | Funded, 1 day of margin |
| Next deadline relative to midterms (3 Nov) | +31 days | +38 days |
Release dates are from the Bureau of Labor Statistics December 2026 schedule; the meeting dates are the Fed's own published FOMC calendar, and the December meeting is one of the four that carries a Summary of Economic Projections.
Read the table as a conditional, not a prediction. Under the Senate's 11 December date, the entire sequence — jobs, the Fed's last decision of the year, the projections that go with it, and the November inflation print — happens with the statistical agencies open. Under the House's 4 December date, the November employment report publishes on the final funded morning and everything after it is exposed. What has to be true for that to matter is narrow but specific: the earlier date has to be the one that survives reconciliation, and the December deadline has to actually produce a lapse rather than another patch. Neither is a given. But if both happen, the Federal Reserve holds a projections meeting during a funding lapse, and the CPI it would normally see the following morning does not arrive.
A shutdown is not a default
Strip out credit risk and two channels remain. The first is growth, and it is modest and largely reversible. The Congressional Budget Office put the cost of the 35-day 2018–19 shutdown at roughly $11bn of output — about $3bn in Q4 2018 and $8bn in Q1 2019 — of which approximately $3bn was permanently lost rather than recovered when workers returned (CNBC's summary of the CBO analysis). On the growth factor, that is a rounding error that shows up as a quarter of noise and then unwinds.
The second is risk appetite, and the historical record is genuinely counterintuitive: the S&P 500 returned more than 10% during that same 35-day shutdown. The driver was a dovish turn from the Federal Reserve, not the shutdown, which is exactly the point — shutdowns have repeatedly been swamped by whatever the monetary story was at the time. Treating one as a risk-off catalyst has been a losing read for a decade.
What 2025 and 2026 already proved about the data blackout
The information channel is the one with a track record of doing real damage, and there is no need to hypothesise about it, because it has now happened three times in ten months.
The lapse that began on 1 October 2025 ran a record 43 days to 12 November. The Bureau of Labor Statistics was effectively closed, with staff furloughed and no releases published. The consequence was not a delay. The household survey — the Current Population Survey, which produces the unemployment rate, the participation rate and every ratio derived from them — cannot be collected retroactively, and it was not conducted for October. That month's unemployment rate will never exist. October payrolls were eventually folded into the following release, but for the first time in more than 900 months, the survey simply did not happen (Axios).
Then 2026 delivered two more. A partial lapse from 31 January to 3 February affected roughly half of federal departments and delayed the January employment report. A second, confined to the Department of Homeland Security, ran from 14 February to 30 April — 76 days, a record for a single-department closure, though it left the statistical agencies alone.
The dollar read: which factor is actually moving
The meter scores the US dollar against the seven other majors on five factors — the framework is set out on the about page — and a funding fight touches three of them unevenly.
The growth factor takes a small, temporary hit that reverses — on the CBO's own numbers, a tenth or two of a percentage point in one quarter. The risk factor has historically not responded at all, and the 2018–19 precedent argues against reading a shutdown as a safe-haven event. That leaves the rate factor, which is where the whole thing lives, and it moves in a specific and easily misread way.
The Federal Reserve held its target range at 3.50–3.75% on 29 July 2026, and futures have been pricing a path drifting up toward roughly 4% by year-end rather than down. A data blackout into a projections meeting does not obviously shift the centre of that distribution — it widens it. A committee that cannot see the November inflation print has less basis for changing anything than for leaving policy where it is, and a rate expectation that becomes harder to update is one that trades on positioning and prior guidance rather than on incoming evidence. That is a different market from the one traders have been running since the July payrolls sequence, and it is the reason the December calendar deserves attention now rather than in late November.
Why the market shrugged, and what would change that
On Tuesday 4 August the S&P 500 and the Dow Jones Industrial Average both closed at records, the S&P up about 1.8%, with the Nasdaq Composite outperforming. None of that came from Capitol Hill. The session was driven by optimism over a possible interim US–Iran arrangement to reopen the Strait of Hormuz, which knocked Brent sharply lower and lifted risk assets. The dollar index barely moved, holding near 99.9, and the 10-year Treasury yield sat around 4.63%.
This is what a correctly-priced political event looks like: a risk with a low probability, a small expected loss, and a two-month fuse gets almost no premium. The same logic applied to the tariff-deadline sequence earlier in the summer, where the market repeatedly declined to pay up for a scheduled binary until the mechanism connecting it to cash flows was clear.
Three things would change the picture, and it is worth naming them precisely rather than watching the headlines generically:
- The reconciled date. If the House's 4 December language prevails over the Senate's 11 December, the December FOMC and the November CPI move from inside funded government to outside it. That single edit is the difference between a scheduling footnote and a live macro event.
- Whether September actually closes the gap. The House returns on 31 August with roughly four session weeks before the 30 September expiry. A clean resolution in early September retires the October risk entirely; a drift into the final week reintroduces it, and the recent base rate for that is not encouraging.
- Whether a December lapse would touch the statistical agencies at all. The February 2026 DHS-only shutdown lasted 76 days without interrupting a single BLS release. Scope matters more than duration — a targeted lapse and a government-wide one have almost nothing in common as macro events.
The reader takeaway is deliberately narrow. Do not trade the shutdown headline; watch which of two dates ends up in the statute, and then check what that date does to the release calendar. A funding fight reaches the dollar through the Federal Reserve's information set, and this particular one has been scheduled — by accident or otherwise — to land within a week of the Fed's most consequential meeting of the year. For the wider policy-to-markets chain, the July FOMC decision remains the reference point for how this committee behaves when the data is ambiguous.
Educational macro context only — not investment advice.