2024 U.S. Presidential Election
County-Level Analysis & Methodological Review
ANALYSIS AND FORECASTS · JANUARY 16–30, 2025
Key Findings
6.25M
Fewer Votes for Harris
than Biden nationally (−7.7%)
−9.1%
Black-Majority Counties
Harris vs. Biden across 76 counties
−17.6%
Latino-Majority Counties
Harris vs. Biden across 86 counties
5 Months
Ahead of Pew
Findings published before Pew Research’s validated-voter study
The Genesis J2T Approach at a Glance
Genesis J2T analyzes complete, state-certified vote counts for every county in all 50 states, not survey samples or margin shifts.
Raw-vote baselines. Each candidate is compared with their own past results and their party’s earlier nominees, pointing to where voters stayed home and where they switched, a pattern later confirmed by Pew’s validated-voter data.
Cross-sector factors. Vote changes are overlaid on income, prices, crime, immigration and demographic data to show where those factors coincided with voter shifts.
Multi-modal reasoning. Temporal, comparative, causal and analogical analysis across states, updated as official certifications are released.
Speed. Findings published in January 2025, nearly five months before Pew’s validated-voter study.
Executive Summary
Between January 16 and January 30, 2025, the Jumptuit Group broadcast a four-part Datacast series, “Why Vice President Kamala Harris Lost the Popular Vote,” within the Genesis J2T platform interface. Drawing on state-certified raw vote counts for every county in all 50 states, and integrating regional economic, demographic, crime and immigration datasets, the series identified the core structural drivers of the 2024 presidential election outcome nearly five months before the Pew Research Center released its validated-voter study on June 26, 2025. The analysis establishes three primary findings:
Asymmetric Democratic turnout collapse. Nationally, Vice President Kamala Harris received approximately 75 million votes, 6.25 million (about 7.7–8%) fewer than President Joe Biden’s 81.2 million in 2020. In major Democratic strongholds the shortfall was about double the national rate: roughly 16% in California and 15% in New Jersey.
Divergent drivers in minority-majority counties. Across 76 Black-majority counties, Harris’s raw vote fell by 183,281 (−9.1% overall; −14.3% in the median county), a pattern that pointed primarily to core Democratic voters staying home in non-competitive states. Across 86 Latino-majority counties, her vote fell by 406,439 (−17.6% overall; −16.7% in the median county), a pattern that indicated both turnout loss and active switching toward Donald Trump. Both patterns were later confirmed by Pew’s validated-voter data.
Cross-sector factors. By overlaying raw vote changes onto Bureau of Economic Analysis (BEA) income and regional price indices, U.S. Census demographic metrics, local crime statistics and Pew unauthorized-immigrant population estimates, the Genesis J2T model indicated that the same voter shifts recurred wherever rising living costs, declining real incomes, property crime and sharp demographic or immigration pressures converged.
THE DATACASTS · JANUARY 16–30, 2025
Series Overview
The four Datacast episodes examine state-certified 2024 raw vote totals against 2020 and 2016 baselines across key bellwether and populous states.
Episode 1 · California (Jan 16, 2025 · 6:23). Harris trailed Biden in 57 of 58 counties (−1.8M votes; −16.2%), with steep drops in urban, Black and Latino areas; San Bernardino, Riverside and Imperial flipped.
Episode 2 · New Jersey & New York (Jan 22, 2025 · 6:16). New Jersey’s margin fell from 16 to 6 points (Harris −388K); New York’s fell to 12 points (Harris −625K, Trump +327K), led by the Bronx, Queens and Nassau.
Episode 3 · Texas, Arizona & Nevada (Jan 27, 2025 · 6:03). Trump won 242 of 254 Texas counties and 14 of 18 border-area counties (flipping Starr); won Arizona by 187K (Maricopa +71K) and flipped Nevada as Clark narrowed.
Episode 4 · Florida (Jan 30, 2025 · 5:54). Trump topped 6M votes; Harris fell 12% and won 6 of 67 counties. Miami-Dade flipped Republican for the first time since 1988, along with Duval, Hillsborough, Pinellas and Osceola.

Democratic Raw Vote, 2016–2024 (Clinton, Biden, Harris): Alameda, San Francisco and Los Angeles Counties, California
METHOD
Genesis J2T Methodology & Competitive Advantage
Core Analytical Pillars
Unlike traditional election coverage that relies on two-party margin shifts or post-election survey samples, the Genesis J2T platform evaluates complete, state-certified vote counts for every county in all 50 states through four mechanisms:
Raw quantitative baselines. Compares official vote totals for the same candidate across cycles (Trump 2016 vs. 2020 vs. 2024) and for the same party across nominees (Obama 2012, Clinton 2016 and Biden 2020 vs. Harris 2024), indicating whether a shift came from Democratic stay-at-home attrition or from Republican voter expansion, patterns later confirmed by Pew’s validated-voter data.
Cross-sector data overlays. Integrates certified vote counts with BEA personal income, per-capita income and regional price indices (all items, goods, rents, utilities, services); U.S. Census diversity and diffusion scores; local property-crime records; Pew unauthorized-immigrant estimates; and Jumptuit’s stability and bilateral stress indices.
Multi-modal reasoning engine. Combines temporal tracking (official county certifications as they are updated), comparative benchmarking (county against state and national baselines), causal linkage and analogical pattern recognition across otherwise different states.
Speed to publication. Working directly from official state certifications allowed Jumptuit to publish comprehensive county-level findings in January 2025, nearly five months ahead of survey-based voter-file validation studies.
Comparative Landscape: Post-Election Research, November 2024 – June 2025
Jumptuit Genesis J2T Datacast (January 16–30, 2025). Certified raw vote counts measured against same-party and same-candidate baselines, for every county in all 50 states. Shows where turnout fell versus where the opposition grew, linked to local economic and social conditions.
Pew Research Center, American Trends Panel (June 26, 2025). Individual survey responses (n = 9,609) matched to commercial voter files, at the national level and for broad demographic subgroups, with local detail limited by sample size. Shows which individuals switched candidates and which abstained. It did not model local macroeconomic variables, regional inflation, crime rates or unauthorized-immigrant population pressures as causal drivers of county-level shifts; it was a retrospective compositional and behavioral audit of the electorate.
A fully validated study of this kind was not possible in January 2025, because it depends on every state adding individual turnout records to its official voter file after certification, on commercial vendors compiling those records nationwide, and on matching them to survey respondents, a process that runs for months after Election Day.
Mainstream media and policy briefs (November 2024 – March 2025). Two-party margin shifts and vote-share regressions for selected counties, single cities or congressional districts. Shows net shifts in competitiveness, often masking raw turnout declines.
Media and Policy Precedents
Outlets such as NBC News (Nov 23, 2024), Semafor (Nov 15, 2024) and The Philadelphia Inquirer (Nov 2024), alongside policy analyses from the Cato Institute (Mar 25, 2025) and Blue Rose Research, documented aspects of the 2024 shift. None conducted a nationwide, multi-state raw-vote audit integrating economic, crime and immigration layers at county resolution.
KEY QUANTITATIVE FINDINGS
1. State and County Raw Vote Shifts, 2016–2024
In large, diverse urban counties, Harris’s raw vote fell 18% to 23% below Biden’s, while Trump’s raw vote expanded steadily across consecutive cycles.
Los Angeles County, CA. Clinton / Biden / Harris: 2,464,364 / 3,028,885 / 2,417,109. Harris −611,776 vs. Biden (−20.2%).
Queens County, NY. Clinton / Biden / Harris: 517,220 / 569,038 / 437,282. Harris −131,756 vs. Biden (−23.2%).
Alameda County, CA. Clinton / Biden / Harris: 514,842 / 617,659 / 499,551. Harris −118,108 vs. Biden (−19.1%).
San Bernardino County, CA. Biden / Harris: 455,859 / 362,114. Harris −93,745 vs. Biden (−20.6%).
Miami-Dade County, FL. Clinton / Biden / Harris: 624,146 / 617,864 / 480,355. Harris −137,509 vs. Biden (−22.3%).
Broward County, FL. Obama 2012 / Clinton / Biden / Harris: 508,312 / 553,320 / 618,752 / 507,328. Harris −111,424 vs. Biden (−18.0%), and below Obama 2012.
Hudson County, NJ. Clinton / Biden / Harris: 163,917 / 181,452 / 144,765. Harris −36,687 vs. Biden (−20.2%).
Bronx County, NY. Trump 2016 / 2020 / 2024: 37,797 / 67,740 / 98,174. Trump +60,377 vs. 2016 (+159.7%); +44.9% vs. 2020.
Miami-Dade County, FL. Trump 2016 / 2020 / 2024: 333,999 / 532,833 / 605,590. Trump +271,591 vs. 2016 (+81.3%); county flipped.
Hudson County, NJ. Trump 2016 / 2020 / 2024: 49,043 / 65,698 / 79,913. Trump +30,870 vs. 2016 (+62.9%); +21.6% vs. 2020.
State of Texas. Trump 2016 / 2020 / 2024: 4,685,047 / 5,890,428 / 6,393,597. Trump +1,708,550 vs. 2016 (+36.5%); +8.5% vs. 2020.
State of New Jersey. Biden / Harris: 2,608,400 / 2,220,713, Harris −387,687 (−14.9%). Trump 2020 / 2024: 1,883,313 / 1,968,215, Trump +84,902 (+4.5%).
Source: state election offices, as displayed in the Genesis J2T interface. Percentage changes computed from the certified totals.

Democratic Raw Vote, 2016–2024 (Clinton, Biden, Harris): Bronx, Queens and Kings Counties, New York
KEY QUANTITATIVE FINDINGS
2. Black-Majority and Latino-Majority Counties
Across the 162 majority-Black and majority-Latino counties nationwide, Harris received a combined 589,720 fewer votes than Biden, roughly 9.4% of her total 6.25 million national shortfall.
Black-majority counties (n = 76). Harris’s vote fell by 183,281 (−9.1%) against Biden; −14.3% in the median county. Harris trailed Biden in 75 of 76 counties (98.7%).
Indicated mechanism: turnout drop-off. Small rural counties (under 10,000 Biden votes) fell 15.6%, against 8.3% in larger counties. Non-battleground states saw steep drops (Tennessee −18.0%, Mississippi −17.5%, South Carolina −16.7%); battleground Georgia fell only 2.4%.
Latino-majority counties (n = 86). Harris’s vote fell by 406,439 (−17.6%) against Biden; −16.7% in the median county. Harris trailed Biden in 84 of 86 counties (97.7%).
Indicated mechanism: voter switching and urban drop-off. Large urban counties (10,000+ Biden votes) fell 17.8%, against 14.7% in smaller counties. Trump led in 72 of 86 counties (1,978,020 to 1,904,756), though Harris won the major metros (Bexar, El Paso, the Bronx).
National baseline. Harris −6.25 million (−7.7%) against Biden. Pew’s validated-voter data later confirmed that Black voters largely stayed home, while Hispanic voters shifted from 61–36 Democratic (2020) to 51–48 Democratic (2024).

Black-Majority Counties: Change in Harris 2024 Vote vs. Biden 2020

Latino-Majority Counties: Change in Harris 2024 Vote vs. Biden 2020
METHODOLOGICAL PRECEDENT
The 2020 Election Alert and DataCheck
The same-candidate comparable framework used in the 2024 Datacasts was first applied, and validated, by Jumptuit during the 2020 election cycle. On October 22, 2020, twelve days before Election Day, Jumptuit published an Election Alert comparing Trump’s RealClearPolitics (RCP) polling average on October 21, 2020 with his RCP average on October 21, 2016 across eight battleground states.
While national head-to-head polls projected a comfortable Biden victory (RCP +7.2; FiveThirtyEight +8.4), Jumptuit’s baseline comparison correctly anticipated the direction of Trump’s actual vote share in all eight states. It identified gains in seven (Arizona, Florida, Michigan, North Carolina, Ohio, Pennsylvania and Wisconsin) and flagged Georgia as the only state where Trump’s base trailed its 2016 level (−0.5 poll delta; −1.6 points in actual vote share), foreshadowing Biden’s narrow 11,800-vote victory there.

Trump’s RealClearPolitics Poll Average, 2020 vs. 2016: Eight Battleground States, as of October 22, 2020
Six days after Election Day, on November 9, 2020, Jumptuit followed up with a DataCheck Alert, “Failure of Polls: A Close Election, Forecast as a Blowout.” It compared results as of November 8 with the FiveThirtyEight weighted polling averages of November 2 across 12 key battlegrounds. In every one of the 12, the polls underestimated Trump’s support, by 1.87 points in Georgia up to 5.90 points in Ohio (3.9 points on average), and overstated Biden’s margin by an average of 4.4 points, up to 7.7 points in Wisconsin. In Florida and North Carolina the polling average pointed to the wrong winner. The results confirmed what the October Election Alert had signaled: a far closer race than the polls showed.
Polls vs. Results in 12 Battlegrounds
FiveThirtyEight margin on Nov 2 → result margin on Nov 8 (Biden − Trump), with the underestimate of Trump’s support and the overestimate of Biden’s margin, in percentage points:
Arizona: +2.6 → +0.6 · Trump −2.84 · Biden 2.04
Florida: +2.5 → −3.4 · Trump −4.60 · Biden 5.90
Georgia: +1.1 → +0.2 · Trump −1.87 · Biden 0.89
Michigan: +8.0 → +2.7 · Trump −4.70 · Biden 5.30
North Carolina: +1.8 → −1.4 · Trump −2.90 · Biden 3.20
Ohio: −0.7 → −8.2 · Trump −5.90 · Biden 7.50
Pennsylvania: +4.6 → +0.6 · Trump −3.46 · Biden 3.99
Wisconsin: +8.4 → +0.7 · Trump −5.20 · Biden 7.70
Minnesota: +9.1 → +7.2 · Trump −2.70 · Biden 1.90
Nevada: +5.3 → +2.4 · Trump −3.30 · Biden 2.90
Iowa: −1.3 → −8.2 · Trump −5.60 · Biden 6.90
Texas: −1.2 → −5.8 · Trump −3.60 · Biden 4.60
Source: Jumptuit DataCheck Alert, X post (Nov 9, 2020).

Results vs. Polls, 2020: How Far the Polls Overstated Biden’s Margin, by State, as of November 8, 2020
Sources
Jumptuit Genesis J2T Datacasts: Ep. 1, California · Ep. 2, New Jersey & New York · Ep. 3, Texas, Arizona & Nevada · Ep. 4, Florida
Pew Research Center, Behind Trump’s 2024 Victory, a More Racially and Ethnically Diverse Voter Coalition (June 26, 2025) and methodology
NBC News, Where Trump made his biggest gains (Nov 23, 2024)
Semafor, Democratic turnout plummeted in 2024, but only in safe states (Nov 15, 2024)
The Philadelphia Inquirer, Philadelphia voter totals (Nov 7, 2024)
Cato Institute, Naturalized Immigrants Probably Voted Republican in 2024 (Mar 25, 2025)
Jumptuit, Election Alert, X post (Oct 22, 2020); DataCheck Alert: Failure of Polls, X post (Nov 9, 2020)
Vote counts: state election offices (certified results); context layers: U.S. Bureau of Economic Analysis, U.S. Census Bureau, Pew Research Center unauthorized-immigrant estimates
