News news5 stories
A signed and implemented deal would remove a central obstacle to rebuilding Gaza and establishing a new Palestinian government.
Trump announced the proposal before Hamas publicly confirmed its position; Hamas then said it had accepted. The plan calls for Hamas to surrender heavy and light weapons over six to eight months, disclose tunnels and arms sites, and transfer territory to a new Palestinian administration as Israel withdraws in stages. At publication, neither Hamas’s disarmament nor Israel’s withdrawal had begun.
The case turns on a harder question than whether Gray was a bad parent: when does ignoring danger make a parent criminally responsible for a child’s intentional killings?
“Warning signs” is too vague to judge the verdict. This is the evidence prosecutors put together:
- In September 2021, the boy used a school computer to search “how to kill your dad.”
- In May 2023, an FBI tip about an online school-shooting threat was traced to the family home. Deputies could not prove who made it, but told Gray to keep guns inaccessible.
- Gray gave his son a Sig Sauer AR-style rifle for Christmas in 2023.
- The family’s guns were not kept in a safe or secured with locks; testimony placed the rifle in the boy’s bedroom or at the top of an unlocked closet.
- The boy maintained a shrine to the Parkland school shooter. Gray said he did not know what it was.
- In August 2024, the boy texted his father, “This is the end for you,” “Whenever something happens, blood is on your hands,” and “I’m restraining myself for what I want to do.”
- Gray searched for a mental-health facility but did not complete an intake. The boy’s mother said a crisis program was supposed to evaluate him; Gray said he had planned to arrange school counseling the following week.
The defense’s answer was that the boy hid his plan, police had not substantiated the 2023 threat, and several messages had less ominous explanations at the time. That dispute—what Gray actually knew before the shooting—is why this first-of-its-kind conviction creates a difficult precedent.
Why this was a good outcome
This was not a parent being punished merely because his child committed a crime. Prosecutors showed a narrow, unusually grave chain: a prior school-shooting investigation, explicit threatening messages, an unfinished mental-health intervention, unsecured guns and a rifle Gray himself gave the boy. Holding a parent liable in circumstances this extreme may push adults to secure firearms and act on credible warnings before someone is killed.
Why this may be a bad precedent
The son intentionally committed the murders; the father did not. Treating terrible supervision as murder risks judging foreseeability with hindsight and makes the boundary of criminal responsibility hard to define. A cleaner rule would punish reckless gun access or unsafe transfer directly, instead of stretching homicide law to cover another person’s deliberate act—especially when police previously could not substantiate the online threat.
Has this happened before?
In 2024, James and Jennifer Crumbley were each sentenced to 10 to 15 years for involuntary manslaughter after their son’s school shooting.
Two bars: prosecutors sought 80 years across 27 counts at Colin Gray's sentencing, and the judge imposed concurrent terms totalling 15 years.
The judge imposed 15 years of the 80 prosecutors asked for
The 80 years was the total across 27 counts; the judge ran the terms concurrently. A Georgia jury convicted Gray of second-degree murder and involuntary manslaughter after his 14-year-old son killed two students and two teachers.
The numbers
| Years | Sentence | Note |
|---|---|---|
| Sought by prosecutors | 80 | across 27 counts |
| Imposed by the judge | 15 | concurrent terms |
Why the Crumbleys were convicted
- James Crumbley bought the 9 mm handgun four days before the shooting; Ethan described it online as his “beauty.”
- On the morning of the shooting, school staff called the parents in after finding a drawing of a gun, a wounded person and phrases including “The thoughts won’t stop.”
- The parents declined to take Ethan home and did not tell the school that James had just bought the gun.
- Prosecutors said the weapon was not secured and remained accessible to Ethan.
AP’s trial account explains the evidence presented to the juries.
Ongoing1 earlier update2 days
This story has run 2 times over 2 days, starting Jul 28.
- Oregon Fires Top Million Acres
- Ptarmigan Fire crosses into Canada after burning at least 18,000 acres — this issue
Oregon and Washington leaders are calling for reinforcements ahead of what they expect will be a grueling August.
Eighteen thousand acres is about 28 square miles: a substantial fire, though still far below the 100,000-acre threshold commonly used for a “megafire.” The more important scale is regional and national. By July 21, U.S. wildfires had burned 3.96 million acres—about 38% more than at the same point in 2025. By July 27, the national total had passed 4 million acres, compared with a 10-year average of 3.4 million by that point.
Two bars: U.S. wildfires had burned more than 4 million acres nationally by July 27, 2026, against a ten-year average of 3.4 million acres by the same date.
The national season is running ahead of its ten-year average
The Ptarmigan Fire itself has burned at least 18,000 acres — about 28 square miles, well below the 100,000 acres commonly called a megafire.
The numbers
| Millions of acres burned nationally | Acres burned | Note |
|---|---|---|
| 2026, by July 27 | 4.0 | reported as having passed 4 million |
| 10-year average, same date | 3.4 |
Three Republicans joined most Democrats, but John Fetterman voted no and the measure failed 49–50.
Roll call4 members named
Broke with party to halt the war3
- Susan Collins(R) Maine
- Lisa Murkowski(R) Alaska
- Rand Paul(R) Kentucky
Broke with party to continue the war1
- John Fetterman(D) Pennsylvania
It was the Senate’s 13th Iran War Powers vote this year. Because the resolution failed, it imposes no new legal limit on the administration’s continuing strikes or deployment of U.S. forces.
A stacked bar of the Senate vote to discharge an Iran war powers resolution: 49 yes and 50 no out of 99 cast, marked with the 50 votes that would have been a majority of those voting.
The war powers resolution fell one vote short
Three Republicans joined most Democrats; John Fetterman voted no. It was the Senate's 13th Iran war powers vote this year.
The numbers
| Category | Yes | No |
|---|---|---|
| Senate roll call | 49 votes | 50 votes |
The city gave no reason for Shon Barnes’s departure after criticism of the department’s response.
Quick bites2
CDC nominee advances
PBS NEWSHOUR: The Senate HELP Committee advanced Erica Schwartz’s CDC nomination 12–10; she supports childhood vaccines, but senators questioned whether she would resist political interference from Health Secretary Robert F. Kennedy Jr.
Elections4 races
The midterms · 96 days out · November 3, 2026
D or R take the House / Senate?
U.S. House: chance Democrats take it, Polymarket 86%, Kalshi 84%. U.S. Senate: chance Democrats take it, Polymarket 44%, Kalshi 46%. Full day-by-day figures are in the table below.
No House district is priced by either book — the House appears here as chamber control only
Each bar is one contract: will Democrats control this chamber after the 2026 midterms. The Republican share is the arithmetic complement of that price, not a separately traded market. Republicans hold the House and the Senate, so the figure printed there is the chance Democrats take them. Two venues price each contract; where they differ the lighter wedge is how far apart they are, and the panel on the right is the week.
- two venues, priced separately
- shaded = how far apart they are
- 50% — control flips here
- where it stood 7 days ago
- the last 7 days, on its own scale
- week panel: one line a party
The week panel draws one line per party, crossing where control does. Only the Democratic line is a traded price: the Republicanline is its arithmetic complement, the same mirror the bar’s other half is, and not a second market agreeing with the first. The shading around each line is how far apart the two venues are.
Market prices are not forecasts with error bars; they are the last price two sets of traders agreed on. The gap between the venues is the closest thing here to a stated uncertainty. The bar runs 0–100%; the week panel beside it is zoomed to its own data and prints its top and bottom, because a two-point move is four pixels wide on a 0–100% scale and was previously drawn there anyway.
The two chambers are not the same kind of bet. The House sits far enough from the flip line that the venues’ disagreement barely matters; the Senate sits close enough that their 2.0pp gap / 2.0pp gap is a meaningful share of the distance to 50%.
Who is defending the House and the Senate is sourced, not remembered — House, 2026-07-23 issue: the NDAA passed “216-212, with six Democrats joining a unified Republican majority.”; Senate, 2026-08-10 issue: “Republicans hold 53 of the Senate's 100 seats.”
Volume is each venue's own reported figure — Polymarket in USD, Kalshi in contracts. They are not a common unit, so read the weighting as a tilt, not a conversion.
| Day | U.S. House · Polymarket | U.S. House · Kalshi | U.S. Senate · Polymarket | U.S. Senate · Kalshi |
|---|---|---|---|---|
| 7 days ago | 86% | 84% | — | 44% |
| 6 days ago | 86% | 84% | 44% | 45% |
| 5 days ago | 87% | 84% | 44% | 45% |
| 4 days ago | 86% | 84% | 44% | 44% |
| 3 days ago | 86% | 84% | 44% | 44% |
| 2 days ago | 86% | 84% | 44% | 45% |
| yesterday | 86% | 84% | 44% | 46% |
| today | 86% | 84% | 44% | 46% |
3 closest races
cSouth Carolina Senate primarySC U.S. Senate (R)
THE HILL: SC U.S. SENATE (R): Ralph Norman (R) leads Darline Graham Nordone (R), 18.6%–16.3%; 26.3% are undecided.
Two markets, one question
South Carolina Senate primary
Jul 31, 12:00 PM CT
widest venue gap 0.8pp- Polymarket, solid
- Kalshi, dashed
- shaded = the gap between them
South Carolina Senate primary. Darline Graham Nordone (R): Polymarket 54.35%, Kalshi 53.50%; Ralph Norman (R): Polymarket 42.0%, Kalshi 41.5%. Widest disagreement between the two venues: 0.8pp. Full history is in the table below.
7 days agoMarket-implied chance of winning, out of 100 — prices, not polls. Source: Polymarket · Kalshi.
Prices are what each venue’s traders were paying, right-aligned to the snapshot: the last sample is today, the one before it yesterday. Series run 6–8 samples because the venues listed these markets on different days; a short line starts late rather than being stretched across a week it was never priced over.
Money-weighted: Darline Graham Nordone 54% · Ralph Norman 42%. Volume is each venue's own reported figure — Polymarket in USD, Kalshi in contracts. They are not a common unit, so read the weighting as a tilt, not a conversion.
Darline Graham Nordone (R)7 days agoPolymarket$303K—Kalshi1.3M contracts · 53%/54%68%venue gap—7-day move-16ppRalph Norman (R)7 days agoPolymarket$77.2K—Kalshi502K contracts · 41%/42%18%venue gap—7-day move+22ppEvery sample behind the chart. South Carolina Senate primary, snapshot Jul 31, 12:00 PM CT. Day Darline Graham Nordone · Polymarket Darline Graham Nordone · Kalshi Ralph Norman · Polymarket Ralph Norman · Kalshi 7 days ago — 68.0% — 18.0% 6 days ago — 73.0% 20.5% 17.0% 5 days ago 70.05% 73.0% 19.5% 19.0% 4 days ago 72.55% 73.0% 17.5% 19.0% 3 days ago 73.75% 73.0% 21.0% 17.0% 2 days ago 74.15% 72.0% 20.5% 24.0% yesterday 59.00% 56.0% 37.5% 38.0% today 54.35% 53.5% 42.0% 41.5% fIowa SenateIA U.S. Senate (D)Turek won’t commit to Schumerone book
THE HILL: IA U.S. SENATE (D): Josh Turek (D) would not say whether he would back Chuck Schumer (D) as party leader.
Two markets, one question
Iowa Senate
Jul 31, 12:00 PM CT
one venue only- Kalshi, dashed
Iowa Senate. Josh Turek (D): Kalshi 43.5%. Full history is in the table below.
5 days agoMarket-implied chance of winning, out of 100 — prices, not polls. Source: Kalshi.
Prices are what each venue’s traders were paying, right-aligned to the snapshot: the last sample is today, the one before it yesterday. Every series here is 6 samples long.
Every sample behind the chart. Iowa Senate, snapshot Jul 31, 12:00 PM CT. Day Josh Turek · Kalshi 5 days ago 39.0% 4 days ago 40.0% 3 days ago 46.0% 2 days ago 46.0% yesterday 47.0% today 43.5% eMichigan Senate primaryMI U.S. Senate (D)
THE HILL: MI U.S. SENATE (D): Abdul El-Sayed (D) leads Haley Stevens (D) 54%–39%; 6% are undecided.
Two markets, one question
Michigan Senate primary
Jul 31, 12:00 PM CT
widest venue gap 0.4pp- Polymarket, solid
- Kalshi, dashed
- shaded = the gap between them
Michigan Senate primary. Abdul El-Sayed (D): Polymarket 96.45%, Kalshi 96.25%; Haley Stevens (D): Polymarket 3.20%, Kalshi 3.65%. Widest disagreement between the two venues: 0.4pp. Full history is in the table below.
7 days agoMarket-implied chance of winning, out of 100 — prices, not polls. Source: Polymarket · Kalshi.
Prices are what each venue’s traders were paying, right-aligned to the snapshot: the last sample is today, the one before it yesterday. Series run 7–8 samples because the venues listed these markets on different days; a short line starts late rather than being stretched across a week it was never priced over.
Money-weighted: Abdul El-Sayed 96% · Haley Stevens 3.6%. Volume is each venue's own reported figure — Polymarket in USD, Kalshi in contracts. They are not a common unit, so read the weighting as a tilt, not a conversion.
Abdul El-Sayed (D)7 days agoPolymarket$683K—Kalshi2.8M contracts · 96%/96%65%venue gap—7-day move+32ppHaley Stevens (D)7 days agoPolymarket$457K36.95%Kalshi2.8M contracts · 3.6%/3.7%38.00%venue gap1.1pp7-day move-34ppEvery sample behind the chart. Michigan Senate primary, snapshot Jul 31, 12:00 PM CT. Day Abdul El-Sayed · Polymarket Abdul El-Sayed · Kalshi Haley Stevens · Polymarket Haley Stevens · Kalshi 7 days ago — 65.00% 36.95% 38.00% 6 days ago 64.50% 66.00% 35.60% 37.00% 5 days ago 64.50% 62.00% 35.35% 38.00% 4 days ago 64.50% 69.00% 30.45% 33.00% 3 days ago 70.00% 75.00% 23.95% 27.00% 2 days ago 82.50% 83.00% 17.40% 17.00% yesterday 94.50% 95.90% 5.40% 4.40% today 96.45% 96.25% 3.20% 3.65% A stacked bar of one poll: 54% for Abdul El-Sayed (D), 39% for Haley Stevens (D), and 6% undecided.
ChartMI U.S. SENATE (D): Abdul El-Sayed (D) leads Haley Stevens (D) 54%–39%; 6% are undecided.
15-point lead, with 6% undecided.
Abdul El-Sayed (D)Haley Stevens (D)UndecidedThe numbers
MI U.S. SENATE (D): Abdul El-Sayed (D) leads Haley Stevens (D) 54%–39%; 6% are undecided. Percent Abdul El-Sayed (D) Haley Stevens (D) Undecided Poll 54% 39% 6%
The other 1 race
Governors1
dWisconsin governor primaryWI governor (D)Barnes exits as Hong leads
AP: WI GOVERNOR (D): Mandela Barnes (D) quit after the state party reviewed past misconduct allegations; a Marquette poll had Francesca Hong (D) ahead 38%–16%.
A judge ruled returned absentee ballots cannot be replaced, so votes already cast for Barnes—or another candidate who quit—stay locked in.
Two markets, one question
Wisconsin governor primary
Jul 31, 12:00 PM CT
widest venue gap 1.5pp- Polymarket, solid
- Kalshi, dashed
- shaded = the gap between them
Wisconsin governor primary. Francesca Hong (D): Polymarket 96.30%, Kalshi 94.85%. Widest disagreement between the two venues: 1.5pp. Full history is in the table below.
7 days agoMarket-implied chance of winning, out of 100 — prices, not polls. Source: Polymarket · Kalshi.
Prices are what each venue’s traders were paying, right-aligned to the snapshot: the last sample is today, the one before it yesterday. Series run 7–8 samples because the venues listed these markets on different days; a short line starts late rather than being stretched across a week it was never priced over.
Money-weighted: Francesca Hong 95%. Volume is each venue's own reported figure — Polymarket in USD, Kalshi in contracts. They are not a common unit, so read the weighting as a tilt, not a conversion.
Francesca Hong (D)7 days agoPolymarket$82.6K—Kalshi803K contracts · 95%/95%80%venue gap—7-day move+15ppEvery sample behind the chart. Wisconsin governor primary, snapshot Jul 31, 12:00 PM CT. Day Francesca Hong · Polymarket Francesca Hong · Kalshi 7 days ago — 80.00% 6 days ago 80.90% 84.00% 5 days ago 85.50% 84.00% 4 days ago 83.25% 81.00% 3 days ago 81.25% 90.60% 2 days ago 95.80% 94.40% yesterday 95.65% 94.00% today 96.30% 94.85%
“Still in doubt” means no leader above 80% on either book. Prices captured Jul 31, 12:00 PM CT. These are trader-implied estimates, not polls. Separate contracts trade independently, so opposing prices may not add to exactly 100%.
Not news news1 story
The wage gap is real; longitudinal data suggest much of it appears before the wedding.
Replot and analysis: Crémieux; data: NLSY79; method follows Dougherty (2005)
The BLS figures compare different groups at one point in time; they do not show that marriage caused the higher pay. The longitudinal analysis found that men’s earnings began rising roughly six years before marriage, and a younger NLSY97 cohort showed the same broad pattern. Older first-time grooms showed no premium, men in “shotgun” marriages had the same premarital gains as other grooms, and wage declines often began before divorce. The likelier explanation is selection and maturation: traits and life changes that make marriage more likely may also raise earnings.
Grouped bars of 2023 median usual weekly earnings for full-time workers: married men $1,421 against never-married men $945, and married women $1,143 against never-married women $876.
Married men out-earned never-married men by half
A longitudinal sample found men's earnings began rising roughly six years before marriage, so much of this gap appears before the wedding.
The numbers
| Median weekly earnings, full-time workers, 2023 | Married | Never married | Note |
|---|---|---|---|
| Men | $1,421 | $945 | a 50% gap |
| Women | $1,143 | $876 | a 30% gap |