C I T I Z E N I N T E L L I G E N C E A R C H I V E
CIA-HOUSE-002: Did the Money Change
the Outcome?
House Series, Model No. 2 | Session 1 working document | October 2026
THE QUESTION: AIPAC-linked groups spent millions in these races. Did the spending
match who won? This file lines up the money and the results from House Model #001. It
does not prove the money caused any result.
All 435 House seats are on the ballot November 3, 2026. This model looks only at the eight
races where AIPAC-linked spending was found and the primary is already decided. The
Michigan row is a Senate race, shown for comparison.
The eight races
RACE
AIPAC-
LINKED
SPENDING
WHO THE
MONEY WAS
FOR / AGAINST
WHO WON
THE
PRIMARY
AIPAC-LINKED SIDE
MI Senate (for
comparison)
$30.6M+
For Stevens,
against El-Sayed
El-Sayed
LOST
IL-9
$5.8M+
For Laura Fine
Daniel Biss
(Fine finished
third)
LOST
IL-7
$5M+
For Conyears-Ervin
La Shawn Ford
LOST
IL-2
$4.3M+
For Donna Miller
Donna Miller
WON
IL-8
$3.9M
For Melissa Bean
Melissa Bean
WON
KY-4 (R
primary)
$3M+ against,
~$500K for
Against Massie, for
Gallrein
Ed Gallrein
(Massie lost)
WON
NJ-11 (special)
$2M+
Against Tom
Malinowski
Analilia Mejia
(Malinowski
lost)
OPPOSITION EFFORT:
target lost (not counted
as win or loss)
MD-5
$1M+
For Adrian Boafo
Adrian Boafo
WON
THE COUNT: AIPAC-linked side won 4 (IL-2, IL-8, KY-4, MD-5). It lost 3 (Michigan
Senate, IL-9, IL-7). In 1 (NJ-11), money was spent against a candidate who lost.
What this shows, and what it does not
IT SHOWS
IT DOES NOT SHOW
Very large spending, with mixed
results. The biggest single bet ($30.6M
That the money caused any win or loss. Incumbency, polling,
local issues and other candidates' money also matter, and I

in Michigan) lost.
In the Illinois races, the money came
through new groups, not AIPAC's own
name.
have not measured them.
Anything about November. These are primary results only.
Early ideas to test (hypotheses, not findings)
1. Spending against a candidate may work differently than spending for one.
2. Timing and message may matter more than total dollars.
3. Open seats may respond differently than seats with incumbents.
4. More races are needed before any pattern can be claimed.
ARCHITECT ANALYSIS (opinion): If money always won, there would be nothing to
learn. It does not always win. The $30.6M loss in Michigan is the proof. The better
question for a voter is not "did the money win?" but "who was paying, and did they tell
me?"
STATUS: SESSION 1. Amounts and outcomes from Sludge, FactCheck.org, WTTW, JNS, New Jersey
Monitor, Spectrum News. Not yet checked against FEC filings. Open items: polling before each primary;
November results; add more races.
"I don't tell you how to vote. I give you the X-ray. You decide."
— Uma | PixelMagic / MindTraveler | People Over Party