The Data — 86.9M FEC Records, Wasted Donations, 2026 Races
2000 – 2026 Federal Election Data
Democratic donors gave
$666M to races
that never had a chance.
Not because donors are uninformed. Because nobody showed them the odds before they gave. That ends here.
32%
of all Dem spend
went to losing races
339
structurally
unwinnable races
"Structurally unwinnable" is a classification, not a claim that every dollar spent had zero effect: it means the seat's Cook PVI was R+8 or worse going in, an incumbent opponent was present, and the challenging party hadn't won that seat in the prior three cycles. Some of these races were closer than the partisan lean predicted, and some spending in them may have helped adjacent races or built long-term infrastructure — this figure counts dollars spent under those conditions, not dollars proven to have accomplished nothing.
Assumptions behind the estimate: commission costs are industry-norm percentages, not vendor-reported figures (see the methodology note below for the exact rates); "unwinnable" uses the classification above, not a race-by-race outcome model; and the range is 2000–2024 FEC bulk data only. See the full methodology.
🏴
$666M
Wasted on hopeless races
R+8 or worse partisan lean — structurally unwinnable going in
💸
$100M+
Est. consultant commissions
~15% of losing-race spend went to media buyers, mail vendors, digital consultants — win or lose
🔁
13
Consecutive election cycles
The same pattern of funding unwinnable races, every cycle from 2000 through 2024
Exhibit A
Same Thing. Every Cycle.
Hundreds of millions in donor money poured into structurally unwinnable races — every two years, like clockwork. The consultants running these campaigns had every incentive to keep them alive.
Red bars = donations to R+8 or worse races that had no path to victory · Source: FEC bulk data 2000–2024
Exhibit B
The 2024 Money Pits
Candidates who raised enormous sums in races they either lost by 10%+ or won by 20%+ — money that couldn't move the needle. This is where your money went if you donated to these campaigns.
| # |
Candidate |
Race |
Lean |
Margin |
Raised |
| 1 | Colin Allred | TX Senate | R+13 | Lost by 11.2% | $96M |
| 2 | Jon Tester | MT Senate | R+11 | Lost by 14.3% | $92M |
| 3 | Adam Schiff | CA Senate | D+22 | Won by 20.4% | $38M |
| 4 | Debbie Mucarsel-Powell | FL Senate | R+3 | Lost by 12.9% | $28M |
| 5 | Angela Alsobrooks | MD Senate | D+25 | Won by 20.1% | $24M |
* Lost by 10%+ = structurally unwinnable or badly run. Won by 20%+ = safe seat that didn't need the money. Source: FEC.gov, AP Elections
Exhibit C
Where the Money Burned
Wasted donor money wasn't spread evenly — it concentrated in a handful of states where candidates ran high-profile, high-spend campaigns in deeply red territory.
By State — Top Wasted
Hopeless-race spending by state
By Race Type
Senate vs. House hopeless races
The Solution
The Information Donors Were Never Given
✗ The Old Way — ActBlue
Donate first. Find out it was hopeless later.
🔴No win-probability score before you give
🔴~6% effective fee (platform + PayPal + tip pressure)
🔴No performance tracking — no way to know your impact
🔴Recurring donations auto-enrolled — hard to cancel
✓ The VoteROI Way
Every race scored before you give a dollar.
✅Win-probability score on every race — prediction markets + polling
✅Your data never shared with campaigns
✅2.75% flat fee — less than half of ActBlue
✅Donation X-Ray: see your full giving history + outcomes
✅No recurring traps. No tip prompts. No spam.
Methodology
How We Handle Edge Cases
Redistricting: When a district's lines change mid-cycle, we re-map incumbent/PVI/rating fields to the new boundaries at the next data refresh. Historical race data before the redraw is not retroactively re-scored.
Open seats: An incumbent's retirement or resignation removes the incumbency-advantage term from that race's probability model; base_prob is re-set from the district's underlying partisan lean until a credible poll or forecast is available.
Special elections: Tracked separately from the regular cycle with their own status field; a special election's result does not change the seat's regular-cycle race entry.
PVI revisions: TBD — this site does not yet have an automated process for ingesting Cook PVI's periodic re-baselining (e.g. after new census data); PVI values are updated manually when we become aware of a revision. See docs/audit/REPORT.md for the open items list.
Put Your Money Where It Can Win.
Every race on VoteROI is scored by real win probability. Your donation goes where it actually changes outcomes — not where consultants get paid regardless.
See the 2026 Races →