← Election Series · All house races

Our verdict

LiberalMind endorses Casey Askar — as unproven, and on watch

Election Day is Tuesday, November 3, 2026, and early voting is underway in Florida. This is the whole race — the records, the claims against each candidate, and what the sources actually show. Read the research, and decide which flaw you can live with.

The setup — beginning with a correction

This seat is open, and the common shorthand for why is wrong. The new FL-22 was created by the 2026 redraw; it is not a retirement seat. Rep. Lois Frankel (D), associated with the old 22nd, did not retire — the map moved most of her district, and she is running for re-election in the new District 23, where she won her August primary (Cook Political Report race page; Ellis Insight). Rep. Debbie Wasserman Schultz was also drawn into this geography — her home is in it — and chose to run in the new District 20 instead, "a decision generating controversy within Democratic circles" (Ellis Insight, June 11, 2026). Any copy describing FL-22 as "Frankel retiring" is factually wrong, and this article does not run it.

The district itself is a Florida oddity: from western Palm Beach County and the Weston area across the peninsula to the Naples suburbs of Collier County, spanning parts of four counties. On the new lines, Trump carried it by 10.5 points in 2024 while Biden carried the same lines by 3 points in 2020 (Cook) — a district that has voted for both parties' presidents within two cycles. Cook's PVI is R+4 (its page dates the PVI to Nov 17, 2025, so its basis relative to the new lines is unstated); FTP's re-aggregation moves the district's 2024 presidential vote from D+5.6 on the old lines to R+10.3 on the new. Both nominees are first-time nominees who have never held elected office (Cook). Democrat Pia Dandiya won her primary with 61.3%; Republican Casey Askar won a seven-candidate primary with 24% — a partisan Republican outlet's claim that he "defeated his opponent by 13 percentage points" is arithmetically hard to square with a 24% plurality in a seven-way field and is printed as a conflict, not a fact.

The raters moved all fall in one direction: Cook moved the race from Lean R to Toss Up on September 25; Sabato's Crystal Ball moved from Leans Republican to Toss-up on October 8; Inside Elections sits at Tilt Republican (per an October 19 secondary report — our research pass could not retrieve the primary ratings page for this district, and that sourcing is flagged rather than hidden). Forecast models, printed as models and not ratings, split: PollingForecast gives Dandiya 56%, Call the Map gives Askar 54%, Decision Desk HQ's model gives Askar 58%. There is no October polling — the only two public polls are August surveys: an Impact Research poll for the Democratic-aligned House Majority PAC (Aug 24–31) at Dandiya 46, Askar 44, moving to 49–44 after respondents were given information about both candidates; and a Tavern Research survey (Aug 10–12, 551 LV) at 50–50, with no undecided share stated — an unusual topline, printed as reported. The compilations' averages conflict because they average different poll sets; we print the polls, not either average.

The money is the loudest fact in the race. Dandiya raised $1.83 million in the third quarter alone, from nearly 24,000 donors this cycle, 93% of contributions at $100 or less (FEC via Florida Politics/TC Sentinel). Askar has financed himself: $4.04 million in self-loans through September 30, rising to $6.69 million lent this cycle by the pre-general report through October 14 ($400,000 of it repaid) — which is how his cash on hand reads $80,000 on October 1 and $1,881,187 on October 14: the two figures sit on opposite sides of further self-loans, printed with their dates, not reconciled into one. Outside money: No Going Back PAC (MAGA Inc.-funded) reported $5.91 million against Dandiya; House Majority PAC reported $7.84 million against Askar and separately announced a $9.7 million advertising buy — different measures at different dates, never summed here.

Pia Dandiya (Democratic)

The record: Dandiya was born and raised in Palm Beach County to parents who immigrated from Jaipur, India — her father a physician, her mother in education. She began as a teacher, became at 28 the founding principal of Democracy Prep Endurance High School in Harlem — one of the youngest high school principals in the country — served as a White House Fellow in 2020–2021 (Domestic Policy Council / Department of Education), and joined Apple in 2022 leading public-sector initiatives in education, healthcare, and workforce development, leaving in June 2025 (campaign biography compilations). Her campaign biography states 86% of her Harlem students came from low-income families and every graduate earned college admission — attributed as campaign-biography claims, in her own file's terms. She initially filed to run in FL-21 against Rep. Brian Mast, then moved to the open FL-22 after the redraw (Wikipedia, Pia Dandiya). If elected she would be the first Indian American to represent Florida in Congress. She is in the DCCC's Red to Blue program.

Her platform, and its price tag: "I am running for Congress to make that American Dream more within reach for others – to lower our costs and to stand up to the corruption in DC that has big corporations and billionaires calling the shots while Floridians pay the price" (campaign website). Her education plank, verbatim from her site: "We'll invest in universal pre-K so kids arrive at kindergarten ready to learn, fully fund strong public schools and the teachers who power them, and build the job training and apprenticeship pathways…" (via The Center Square). That plank is a candidate's-own-promise, and we price it the way we price every endorsee-class promise — with math: the National Institute for Early Education Research estimates high-quality universal preschool at $12,500 per child; at 70% of U.S. 3- and 4-year-olds, that is roughly $70 billion annually (NIEER estimate via The Center Square). Her campaign prices the promise at nothing. A $70-billion-a-year federal program offered without a funding line is an unclosed promise, and it sits at the center of her platform rather than at its edge.

The attacks on her — single-partisan-source, unverified, printed as such: The Floridian Press (Republican-leaning) calls her "a millionaire lobbyist" and reports she "came out against middle-class tax cuts and the child tax credit"; an outside-group attack summarized in the same piece alleges she opposes data centers while invested in AI data-center assets and stock in military contractors (Palantir, Honeywell) and in an ETF invested in Chinese-owned/operated companies, and that her husband worked for a company that bought a major data-center company in which she still holds stock. Every one of those claims rests on her financial disclosure, which our research pass did not read. They are UNVERIFIED, and none is printed here as fact. One checkable kernel exists on data centers — it concerns her opponent's shifting position, documented from his own audio, below.

Pros: A builder's résumé — founding principal, White House Fellow, Apple public-sector lead — with a genuine small-donor fundraising base (24,000 donors, 93% small-dollar) that is the opposite of a bought candidacy; home-county roots in the district's population core; a two-to-one primary win.

Cons: She has never held elected office; her signature promise is an uncosted $70-billion-class federal program; her broader frame — corporations and billionaires "calling the shots" — is her party's standard populist frame, with no documented break from that party on any of it. And she filed for a different district first: the seat she is asking for is the one the map opened, in the place she is from — both things are true, and voters can weigh the sequence.

The flaw, plainly: Dandiya's flaw is an unclosed promise at platform scale, offered by an unproven candidate: the math failure is not an opponent's characterization, it is priced from her own plank by the early-education field's own institute. Unproven is a posture; unclosed is a finding.

Casey Askar (Republican)

The record: Askar (legal name Kousay Askar) was born in Iraq and fled persecution for his faith and ethnicity; his family settled in the U.S., and he has long been based in Naples. He joined the U.S. Marine Corps at 18 and built a franchise empire — founder of the Askar Family Office, franchisor of four pizza brands including Blackjack Pizza, and one of the largest franchisees of Church's Chicken and Dunkin' Donuts (47 Dunkin' locations in South Florida), employing "thousands" (EIN Presswire biography; EquityNet). He is a father of seven. Two corrections to the easy labels: he is a first-time nominee, not a first-time candidate — he ran for Congress in FL-19 in 2020, finishing third in the Republican primary behind Byron Donalds (opposition-research compilation; Paradise Progressive). And the carpetbagging criticism aimed at his Naples base is printed as criticism: Naples is in Collier County, which the new district does reach; his residence outside the district's Palm Beach/Broward population core is a fact, the label is an argument.

The 2020 aftermath — documented litigation: On 2020 primary day, a text message falsely claimed Donalds had dropped out; Donalds accused Askar's side, Askar denied it, and Askar sued Donalds for defamation in November 2020. A judge ruled against Askar, finding he failed to make his defamation case by clear and convincing evidence (Paradise Progressive, Sept 2021 — a partisan local blog reporting a court ruling; the ruling is a checkable court record, and the blog is the source located; we print it at that strength).

The business record — court-documented items (Florida Independent News, Oct 2026, from court records): (1) In 2008, Mr. Pita founder Frank Lombardo sued Askar for wrongful termination and breach of contract; the court ruled Askar and Pita Franchise Corp. owed Lombardo about $900,000. (2) In 2010, Pita Franchise Corp. filed for bankruptcy protection (E.D. Mich.); Askar agreed to pay Lombardo $425,000 in a settlement and the bankruptcy was approved. (3) His Southeast Enterprise Holdings was sued in 2023 in Palm Beach County alleging workplace discrimination/sexual-harassment retaliation — settled after mediation. (4) The same holding company faces a current negligence suit over a January 2025 slip-and-fall in a Palm Beach Gardens Dunkin' and settled two other 2023 customer negligence suits out of court. Settlements are not admissions, and we do not print them as any. The $900,000 judgment and the 2010 bankruptcy are adjudicated and filed facts, reported from court records, and they are the heaviest documented items in his file.

The unverified items — printed as allegations or not at all: An FEC complaint (MUR 7745) alleged his 2020 campaign took a $3 million loan at 0% interest from Northern Bank & Trust, below the rate FEC rules require; a 2020 complaint alleged he misrepresented Harvard's Owner/President Management program — a 12-week executive course conferring no degree — as a Harvard MBA. Both rest on an opposition-research site; neither complaint's disposition was verified, no FEC finding is cited, and we do not print either as fact. A 2020-era profile did independently note that his campaign biography at the time claimed a Harvard Business School MBA — that narrower fact (the claim appeared) is documented; what it proves beyond a padded bio line is not.

His platform and its documented shift: Cost of living first — "Tackling the rising cost of eggs, milk, and gas is [his] top priority… slash overburdening taxes and regulations"; on health care, "end Obamacare once and for all, and replace it with a vehicle that works for all Americans… preserve Medicaid and Medicare" (campaign website via Ballotpedia News). On education: "We must educate our children and not allow them to be indoctrinated by un-American, divisive, and socialist education reforms" (via The Center Square). On data centers, his position moved in public: in the primary he said "I don't want them in the state of Florida"; in October 13 meet-and-greet audio obtained by Florida Politics he said "I want them in rural areas. I don't want them near our homes. And I don't want them near our schools." Both quotes, with dates — the position narrowed from a ban to a zoning rule, and voters are entitled to watch which version he governs on. The DCCC's label for him is "shady businessman" — a partisan characterization; its documented substrate is the litigation and bankruptcy record printed above, which is heavier than the label and does not need it.

Pros: A self-made record of exactly the kind our mission honors — an immigrant who fled persecution for his faith, wore the Marine uniform, and built businesses employing thousands; a campaign financed from his own pocket ($6.69 million lent), answerable to no donor; a conservative platform on taxes, health care, and education stated without trimming.

Cons: A documented business-litigation file with an adjudicated ~$900,000 judgment and a corporate bankruptcy in it; a defamation suit he brought and lost; a data-center position that moved between the primary and the fall; and a 24% plurality nomination. His record of proven conduct — the species we price heaviest — is the worst in this race, and it is not close.

The flaw, plainly: Askar's flaw is proven personal conduct in business: a court judgment, a bankruptcy, a failed defamation suit — all documented, all his. It is not an unproven accusation and we do not price it as one. It is the record of how he has handled other people's money and other people's claims when he held power in the only arena he has held it. We print it at full weight, because the call below is made with it on the scale, not swept off it.

OUR CALL

LiberalMind endorses Casey Askar — as unproven, and on watch.

This is the hardest call in the Florida tranche, and we make it the way the doctrine makes hard calls: by species. Against Askar stands proven conduct — a judgment, a bankruptcy, a lost suit — in business, a decade and more old, in a record that also includes building real enterprises and employing thousands. Against Dandiya stands her own signature promise: a universal pre-K program priced by the field's own institute at roughly $70 billion a year, offered with no funding line, as the centerpiece of a party-standard platform from a candidate who has never held office and documents no break from that party. Proven-versus-unproven cuts for her on biography and against her on the only promise she has actually priced for us: her program's math does not close, and a representative's first job is voting on other people's money. Askar's arithmetic, whatever else is in his file, is his own money, lent to his own campaign, in the open.

The extremist rule is not doing this work — Dandiya is a party-standard Democrat, not an extremist-tier program author, and we do not stretch the word. The flawed-beats-proven-inability rule is not doing it either; neither candidate has led in office. What decides it is narrower: in a race between an unproven Republican carrying a documented conduct file and an unproven Democrat carrying an unclosed $70-billion promise and her caucus's program, we take the flawed man over the unclosed math — and we say, in print, that his first term will be audited against the file above by the publication that endorsed him.

If the conduct file is the weight you cannot carry, the honest alternative in this race is no endorsement, and we understand the voter who lands there. We land on endorsement, with the file printed whole, because the alternative's flaw is not smaller — it is merely newer, and denominated in the voters' money rather than his own.

This race is one of hundreds in LiberalMind's Election 2026 coverage. For the complete breakdown — every race, claim by claim — go to liberalmindhq.com.

Vote early. Research your candidates. Choose the best moderates, whatever their party — no extremism from either side. And encourage your friends and family to do the same.

Sources

Election result — November 4, 2026

Winner: Pia Dandiya (Democratic) — called by the Associated Press. This result was appended on November 4, 2026 (results wave 1); the original article above is unchanged, exactly as published before Election Day.

AI-assisted article — but the sources are human, and the facts are very real.

← Back to the Election Series