Fact-checked by Grok 2 weeks ago

Election recount

An election recount is a second examination and tabulation of votes cast in an election, aimed at verifying the accuracy of the initial count, and is typically triggered either automatically when the victory margin falls below a predefined or upon request by a , voter, or . Procedures for recounts vary significantly by , with some requiring machine re-scans of ballots and others incorporating manual hand counts, often under strict chain-of-custody protocols to maintain integrity. In the United States, state laws govern these processes without a uniform federal standard, leading to differences in eligibility, timelines, and costs borne by requesters in non-automatic cases. Empirical analyses of U.S. statewide general elections from 2000 to 2023 reveal that recounts occur infrequently—only 36 instances out of 6,929 races—and result in outcome reversals even less often, with just three flips, all in contests where initial margins were 0.11% or narrower. These reversals, such as the 2008 Minnesota U.S. Senate election where Democrat overcame a 206-vote deficit to Republican by 440 votes after recount, represent exceptions rather than the norm, as the average margin shift in non-flipping recounts was minimal at about 0.03% of votes, frequently confirming or widening the original lead. Such data underscore the reliability of modern vote tabulation systems, where errors, when detected, stem more from human handling of ballots than systemic flaws. While recounts enhance public confidence by addressing close races transparently, they have sparked controversies, particularly in high-stakes contests like the 2000 Florida presidential election recount, halted by the U.S. amid disputes over standards like "hanging chads," ultimately affirming George W. Bush's narrow victory over . Politicization can arise when losing parties demand recounts without evidence of material errors, yet studies affirm that initial certifications are accurate in the vast majority of cases, with post-election audits further validating results through statistical sampling. This rarity of substantive changes highlights recounts' role as a safeguard rather than a frequent arbiter, promoting electoral stability grounded in verifiable processes over unsubstantiated challenges.

Overview

Definition and Purpose

An election recount is the re-examination and re-tabulation of cast in a specific contest to verify the accuracy of the initial vote count reported by election officials. This process typically focuses on from a single race or ballot measure rather than the entire , distinguishing it from a full canvass or . Recounts may involve machine re-scanning, manual inspection of , or both, depending on jurisdiction-specific procedures and the nature of the equipment used. The primary purpose of an election recount is to identify and rectify potential errors in the original tabulation, such as mechanical failures, human miscounts, or ambiguities in voter intent on ballots, thereby ensuring that all valid votes are accurately reflected in the final results. In closely contested races, where the margin of victory is slim—often defined by statutory thresholds like 0.5% or less—recounts serve to resolve disputes over the winner without necessitating a new , promoting through empirical verification of the physical or recorded votes. This mechanism addresses causal factors like tabulation inaccuracies that could otherwise undermine the causal link between voter preferences and certified outcomes, while fostering by demonstrating in close-margin scenarios. Recounts are not intended to overturn results based on absent evidence in the ballots themselves, but rather to confirm the precision of the through methodical re-processing, which empirical from past recounts shows rarely alters outcomes significantly—changing winners in fewer than 1% of cases across U.S. jurisdictions since 2000. By limiting recounts to verifiable , the process prioritizes factual accuracy over unsubstantiated claims, though it can extend timelines for by days or weeks.

Historical Development

The practice of election recounts emerged in the 19th century as democracies adopted secret ballots and paper-based systems, enabling systematic re-examination of tallies to resolve disputes over accuracy rather than relying solely on witness testimony or parliamentary inquiries. Prior to widespread secret , such as in the before the Ballot Act of 1872, challenges typically involved "controverted elections" handled through petitions to , where committees investigated bribery, intimidation, or procedural irregularities without re-tallying votes, as was often public and verbal. The shift to secret ballots necessitated recount mechanisms to verify counts independently of initial declarations, marking a transition toward procedural verification grounded in empirical re-tabulation. In , formal judicial recounts were introduced in federal electoral law in , authorizing a to scrutinize ballots in ridings where the margin between candidates was one-thousandth of total votes cast or less, reflecting early recognition of in manual counting. This procedure evolved to include provisions for disputed ballots, with the judge determining voter intent, and has been triggered sporadically, such as in close races, to ensure outcome integrity without overturning results in the majority of cases. In the United States, recounts predated statutory codification in many states, occurring informally in local elections during the colonial era and early republic through re-checks by officials or challengers. Formal laws proliferated in the late 19th and early 20th centuries amid reforms standardizing ballots and addressing machine tabulation errors; for instance, statutes enabled recounts from at least 1946 onward, often confirming initial tallies within narrow margins. By the mid-20th century, states varied in thresholds—typically 0.5% or less—leading to procedures distinguishing automatic triggers from candidate-requested ones, with empirical data showing recounts seldom alter winners but frequently affirm vote accuracy. Globally, the saw recounts adapt to technological shifts, from hand-counted paper to punch cards and optical scanners, prompting innovations like risk-limiting audits in the to statistically sample ballots for rather than full re-tallies, prioritizing efficiency while maintaining causal fidelity to original voter intent.

Types and Methods

Automatic versus Requested Recounts

Automatic recounts are triggered by statutory provisions when the initial vote tally reveals a margin of victory below a specified , typically expressed as a or of total votes cast, thereby mandating without requiring a formal request from any party. This mechanism aims to enhance confidence in election outcomes for exceedingly narrow races by proactively addressing potential counting errors or ambiguities in ballots. In jurisdictions with automatic recounts, the process is initiated by election officials upon of preliminary results, often involving manual re-examination of ballots to confirm the accuracy of machine or initial hand counts. Requested recounts, by contrast, depend on an application submitted by a , , or occasionally voters, usually within a short timeframe after results are announced and subject to criteria such as evidentiary grounds for doubt, payment of associated costs (which may be reimbursed if the outcome changes), or restrictions to races with close margins. These recounts allow targeted challenges based on specific allegations of miscounts, rejected ballots, or procedural irregularities, but they place the onus on the requester to justify the need, potentially limiting access due to financial or logistical barriers. In practice, requested recounts can extend timelines and incur expenses borne initially by the petitioner, reflecting a balance between accessibility and prevention of frivolous demands. The distinction between the two types influences election integrity protocols: automatic recounts prioritize systemic safeguards in ultra-close contests to mitigate risks of unaddressed errors, while requested recounts emphasize and accountability, enabling scrutiny where perceived issues arise but avoiding routine re-verification of non-competitive results. For example, , 16 states mandate recounts for margins within predefined limits—such as Georgia's 0.5% or Minnesota's 0.5% for statewide offices—covering federal, state, and local races where applicable. In Canada, federal elections trigger an judicial recount if the margin equals or falls below one-thousandth (0.1%) of total votes cast in the , overseen by a to ensure . Jurisdictions like the and lack automatic triggers, relying instead on candidate requests for recounts in parliamentary or federal contests, often limited to disputed counts and conducted by returning officers or electoral commissions. This binary framework varies by jurisdiction to align with local electoral laws, with automatic provisions more common in systems emphasizing proactive (e.g., certain U.S. states and Canadian races) and requested options predominant where resources are conserved for contested outcomes. Empirical from U.S. recounts since indicate that automatic processes rarely alter results significantly—changing winners in fewer than 1% of cases—but frequently confirm initial tallies, underscoring their role in bolstering amid minimal evidentiary shifts.

Machine, Manual, and Risk-Limiting Audits

recounts involve reprocessing ballots through tabulation equipment, often the same s used in the or alternative certified devices, to verify vote totals without human interpretation of individual ballots. This method prioritizes efficiency, allowing large volumes of votes—such as millions in statewide races—to be re-tabulated in hours or days, reducing labor costs compared to full efforts. However, recounts risk perpetuating undetected systematic errors in the tabulators, such as miscalibration or software glitches, if the underlying equipment issues are not addressed beforehand. Procedures typically require secure chain-of-custody for ballots, public observation, and pre-recount testing of machines to ensure consistency, as mandated in jurisdictions like most U.S. states where recounts are the default for non-close races. Manual recounts, by contrast, entail hand-counting ballots by teams of election officials, often in bipartisan pairs, to directly inspect voter marks and resolve ambiguities like overvotes or undervotes. This approach provides higher transparency and can detect errors missed by machines, such as interpretive issues in or stray marks, making it standard for close-margin contests in places like Georgia's 2020 Senate runoffs, where full hand tallies confirmed initial machine results within 0.1 percentage points after examining over 5 million . Drawbacks include substantial time delays—potentially weeks for large elections—and vulnerability to human fatigue or bias, necessitating strict protocols like randomized ballot batches, duplicate counts for reconciliation, and observer access to minimize discrepancies, which studies show average under 0.5% in controlled settings but can exceed 2% without oversight. Risk-limiting audits (RLAs) represent a statistically grounded alternative or complement to full recounts, sampling a subset of paper ballots to assess whether the reported outcome aligns with the physical record, halting early if the sample confirms the tally within a user-defined risk limit—commonly 5%, signifying a 95% chance of detecting an incorrect result. Developed from mathematical foundations in statistical hypothesis testing, RLAs use methods like ballot polling (direct vote interpretation) or batch comparison (comparing machine tallies to manual samples) to achieve efficiency, often verifying statewide results by examining fewer than 5% of ballots when margins exceed 10%, as implemented in since 2017 across all contests. Unlike exhaustive manual recounts, RLAs scale inversely with reported margins, escalating to full hand counts only if discrepancies trigger the risk threshold, thereby balancing verification rigor with resource constraints; empirical pilots in states like and have demonstrated concordance rates over 99% with initial counts when paper trails exist. Adoption requires voter-verifiable paper records, limiting RLAs to jurisdictions without direct-recording electronic (DRE) systems, and relies on for transparency, as validated by peer-reviewed work from statisticians including .

Thresholds for Initiation

In jurisdictions with formalized election recount provisions, initiation thresholds typically distinguish between recounts—triggered by statutory vote margins indicating potential tabulation errors—and requested recounts, which require application by candidates, voters, or parties within specified deadlines and often subject to fees or evidentiary standards. Automatic thresholds are calibrated to activate in races where the risk of miscount outweighs administrative burdens, generally set at margins of 0.1% to 1% of total votes cast between the leading candidates, though fixed vote differences apply in some cases. In the United States, 21 states and the District of Columbia mandate automatic recounts for margins below predefined levels, such as 0.5% in , , and , or 1% in states like and for certain races. These percentages are calculated based on certified results post-canvass, excluding overvotes and undervotes, to prioritize empirical verification in empirically error-prone close contests. Requested recounts, available in 47 states, impose no universal margin requirement but often limit eligibility to the apparent loser or require demonstration of irregularities, with timelines typically 2–5 days post-certification and costs borne by the requester unless the outcome reverses. Canadian federal elections under the Canada Elections Act trigger automatic judicial recounts if the margin equals or falls below one-one-thousandth (0.1%) of valid votes cast in the riding, overseen by a superior court judge to recount ballots manually. Provincial variations exist, such as Alberta's 0.1% threshold for automatic recounts in close legislative races. Requested recounts beyond this threshold may proceed via court application if fraud or error is alleged, emphasizing causal links to vote discrepancies over mere closeness. Internationally, thresholds reflect similar principles but diverge in stringency; for example, recounts occur only upon petition to the Court of Disputed Returns without fixed automatic margins, while some systems like Germany's tie resolved by lot absent recount triggers. These mechanisms underscore a core rationale: thresholds below which human or machine errors statistically dominate outcomes, as evidenced by historical recount reversal rates of about 1 in 100 U.S. cases altering winners.
Jurisdiction ExampleAutomatic ThresholdRequested Option
U.S. (e.g., )≤0.5% marginAvailable to candidates post-certification, with fees
U.S. (e.g., )None statewide; county-level variesPetitioner pays costs unless margin <10% and reversal occurs
(Federal)≤0.1% of valid votesJudicial application for alleged errors
(U.S.)≤0.5% or ≤100 votes (small precincts)Statutory right for losing candidate
![Minnesota 2008 Senate election recount][float-right] This 2008 Minnesota case, triggered by a 0.4% margin under state law, exemplifies automatic initiation resolving a U.S. Senate race after initial counts showed 225-vote gap among over 2.9 million ballots.

Costs, Timelines, and Oversight

The costs of election recounts differ significantly by jurisdiction and type, with automatic recounts typically funded by government entities while requested recounts often require the petitioner to cover or bond expenses unless the outcome changes sufficiently to warrant reimbursement. In the United States, 28 states mandate that the state or local governments bear the full cost of automatic recounts triggered by statutory margins, whereas in 22 states and the District of Columbia, candidates requesting non-automatic recounts must pay deposits or fees ranging from $1,000 to full costs if the results do not alter the winner's margin beyond a specified threshold, such as 0.1% in some cases. For instance, Washington's 2004 gubernatorial manual recount incurred costs exceeding $700,000 across counties, averaging over 30 cents per ballot recounted. Statewide manual recounts in larger jurisdictions can escalate to millions, as seen in Georgia's 2020 presidential audit and recount efforts, which cost eight metro Atlanta counties nearly $1 million in personnel and logistics. Timelines for recounts are governed by statutory deadlines to ensure results align with requirements, generally spanning from a few days for re-tabulations to several weeks for manual processes involving large volumes of ballots. In the , requests for recounts must typically be filed within 2 to 7 days post-initial , with recounts often completed in 3 to 5 business days and manual ones extending up to 10-15 days depending on state law; for example, requires completion within 5 days of a request, while Florida's manual recount provision allows up to 5 days but ties it to deadlines around 10-19 days post-election. In , under the Canada Elections Act, applications for judicial recounts must be made within 4 days after the validation of results, with the recount process—conducted by recounting valid ballots or all ballots—typically concluding within 7-10 days under judicial supervision to meet federal timelines. Delays can arise from disputes over ballot validity or volume, but laws prioritize expedition to avoid protracted uncertainty in seating officials. Oversight of recounts emphasizes and , with by designated election authorities or courts, mandatory presence of party representatives, and provisions for public or independent observers to verify procedures. In the US, recounts are administered by or election boards, often using bipartisan teams for tabulation and allowing designated watchers from campaigns to monitor handling of ballots, as outlined in guidance and state statutes to mitigate errors or claims. The U.S. Election Assistance Commission recommends chain-of-custody protocols and observer access without interference to ensure integrity. In Canada, judicial recounts are presided over by a who directs election officers in ballot examination, with candidates or their agents entitled to attend and object, providing a formal legal against administrative . This structure, while resource-intensive, prioritizes verifiable accuracy over speed in close contests.

Procedures by Jurisdiction

United States

In the United States, election recounts are administered at the state level under varying statutory frameworks, with no overarching federal law dictating procedures for federal elections, including presidential races. State laws govern the initiation, conduct, and certification of recounts for all contests, from local offices to congressional seats and presidential electors, reflecting the decentralized nature of U.S. election administration. For presidential elections, recounts occur in individual states and can affect the allocation of electoral votes, but they must align with certification timelines to meet the federal "safe harbor" provision in 3 U.S.C. § 5, which urges states to resolve disputes by early December—typically December 8—to ensure congressional acceptance of electors on January 6. Recounts may be triggered automatically in 28 states if the victory margin falls below a predefined , often ranging from 0.1% to 1% of votes cast or a fixed number of votes, such as 0.5% in or 2,000 votes in for statewide races. Candidate-requested recounts are available in 42 states, usually filed within days of the official canvass—e.g., 48 hours in or five days in —and may require a deposit estimated at 10-20% of projected costs in states like . Voters can request recounts in only four states for any election, with 31 states limiting such requests to specific races or ballot measures. In 21 states, requests are confined to margins below certain limits to prevent frivolous actions. Procedures typically begin with a machine re-tabulation of ballots to verify initial counts, followed by inspection of disputed or damaged ballots in many jurisdictions, conducted under bipartisan and with provisions for challenges. Hand recounts are mandatory in some states like for margins under 0.5%, involving precinct-by-precinct reviews, while others like prioritize machine recounts unless irregularities warrant ones. State election boards or secretaries of state oversee the process, with courts resolving legal disputes; for federal candidates, the provides guidance on fundraising for recount expenses but defers operational control to states. Costs are borne by the requesting party in 23 states if the outcome remains unchanged, with potential refunds in 27 states if a recount alters the result or reveals errors exceeding the margin. Timelines vary but are expedited—often completing within 10-30 days—to comply with certification deadlines, such as Arizona's 20-day limit for statewide recounts. A growing number of states, including and , integrate risk-limiting statistical audits alongside or instead of full recounts to sample ballots for verification, enhancing efficiency while maintaining accuracy. These procedures apply uniformly to offices, ensuring state-level consistency in verifying popular vote tallies that determine congressional or electoral outcomes.

Canada

In Canadian federal elections, recounts are conducted as judicial recounts under the Canada Elections Act, primarily to verify vote counts in electoral districts (ridings) where results are exceptionally close. These recounts are overseen by a judge of a superior court and involve either aggregating figures from initial statements of the vote or manually re-examining ballots, including potentially disputed or rejected ones. A judicial recount is automatically triggered if the difference in votes between the leading candidate and the nearest competitor is less than one-thousandth (0.1%) of the total valid votes cast in the riding. The must request the recount from a within four days of validating the initial results, with the judge setting a start date no later than four days after receiving the request. Candidates receive written notice and may appoint up to five scrutineers each to observe the process, alongside recount teams comprising a handler, recorder, and verifiers. Additionally, any elector may apply for a recount within four days of validation by submitting an alleging specific errors, such as miscounts or irregularities in ballot handling, provided the application notifies the and is approved by the . The recount proceeds continuously from 9 a.m. to 6 p.m. daily, excluding breaks, at a location determined by the , typically involving secure transport of ballot boxes and statements from polling stations. Recount first attempt to reconcile totals from statements of the vote; if discrepancies arise or upon decision, they manually count valid , segregating disputed ones for the 's based on voter intent and ballot validity rules. Legal counsel for candidates may argue interpretations, but the judge's rulings are final on the spot. Upon completion, the judge certifies the revised vote totals, which supersede initial results and may alter the winner, as occurred in the 2025 Terrebonne riding where a one-vote Liberal margin emerged post-recount. Costs may be reimbursed to candidates upon application to the Chief Electoral Officer if deemed reasonable. Provincial and territorial elections follow analogous but jurisdiction-specific procedures, often mirroring federal thresholds for closeness (e.g., 0.1% margins triggering recounts in ), with courts overseeing manual verifications, though details vary by provincial election acts without uniform national standards.

United Kingdom

In the , recounts are discretionary processes initiated at the request of a or election agent, typically in response to a close initial vote tally, and must occur before the formal declaration of results by the . These apply to parliamentary elections under the first-past-the-post system, as well as , mayoral, and assembly contests, with procedures outlined in the Representation of the People Act 1983 and supplemented by Electoral Commission guidance. Unlike automatic triggers based on fixed margins, approval rests with the , who assesses whether the recount is warranted to verify the accuracy of the manual count, often after presenting provisional results for review. The recount involves a complete manual re-examination of papers in the affected constituency or ward, supervised by the with candidates and agents present to observe and challenge disputed votes, such as those rejected as invalid. This step aims to correct potential human errors in the initial hand-count, which is standard for polling stations without widespread use of voting machines. If the recount confirms or alters the outcome without flipping the winner, the process concludes; repeated requests are possible but rare, as returning officers prioritize timely declarations to avoid delays in result aggregation. Post-declaration challenges, including allegations of counting errors, proceed via election petitions to the rather than recounts, requiring evidence of corrupt or illegal practices under the Representation of the People Act 1983. Recounts remain uncommon in UK elections due to the transparency of on-site manual counting, which minimizes systematic errors, though they occur in tight races to build confidence in results. For instance, in the May 2025 Runcorn and Helsby parliamentary by-election, an initial count showed Reform UK ahead of Labour by four votes, prompting a Labour-requested recount that ultimately confirmed a six-vote victory for Reform UK's Sarah Pochin. Similarly, the December 2024 Sark general election recount yielded no change in the outcome. In larger contests like the 2024 general election, close margins—such as Labour's 17-vote win in Finchley and Golders Green—did not trigger recounts that altered winners, underscoring the infrequency of outcome reversals. Costs of recounts, including staff and venue expenses, are generally covered by local authorities, with potential recovery from requesting parties if deemed frivolous, though specific allocations vary by election type. No statutory thresholds exist for initiation, reflecting reliance on officer judgment over rigid rules, which critics argue could introduce inconsistency but aligns with the UK's decentralized administration.

Australia

In Australian federal elections, recounts are governed by the Commonwealth Electoral Act 1918 and administered by the . There are no statutory thresholds mandating automatic recounts based on vote margins; instead, divisional returning officers (DROs) may initiate them at their discretion in cases of close results, apparent discrepancies, or errors during the initial , while candidates can formally request one before results are declared, subject to officer approval. For seats, employing instant-runoff (preferential) voting, recounts entail a full re-examination of all papers, of voter , and re-distribution of preferences to determine the two-candidate preferred outcome. The process is observed by party-appointed scrutineers to maintain and , with occurring at divisional offices or central facilities. recounts, under sections 278 and 282 of the , similarly allow re-counts before declaration to resolve ties or determine election order among quota-attaining candidates in the system, often involving exhaustive re-scrutiny of thousands of ballots. Recounts typically conclude within days to weeks, depending on volume, with costs borne by the and no direct candidate fees unless frivolous requests are deemed such. Oversight emphasizes procedural rigor, including post-recount audits where needed, though outcomes rarely reverse initial tallies absent clear errors like misplaced . Prominent cases illustrate application: In the 2016 Herbert () division, a DRO-ordered recount after an initial margin under 100 votes confirmed Labor's win by 37 votes following eligibility challenges to army absentee . The 2007 McEwen () recount narrowed the Liberal victory from 31 to 12 votes amid postal vote scrutiny. The 2025 Bradfield () recount, triggered by a razor-thin flow, finalized a 26-vote win for independent Nicolette Boele over Gisele Kapterian. An exceptional instance occurred in the 2013 Western Australia election, where 1,375 missing papers prompted a court-ordered full re-scrutiny, reallocating two seats from Greens to and Nationals candidates due to quota recalculations. State and territory jurisdictions operate analogously under local laws, with commissions like New South Wales Electoral Commission empowered to order recounts pre-declaration if advisable, as per section 172 of the Electoral Act 2017, focusing on manual verification without fixed automatic triggers.

Other Countries

In India, election recounts are initiated upon written application by a contesting candidate or their agent to the returning officer after the initial declaration of results but before the final announcement, typically when the margin of victory is narrow or discrepancies are alleged in specific rounds of counting electronic voting machine (EVM) data. The Election Commission of India (ECI) oversees the process under Rule 56D of the Conduct of Elections Rules, 1961, allowing recounts of particular rounds or the entire vote if justified, with observers from all parties present to verify seals and totals. Such requests must specify grounds like unsealed EVMs or arithmetic errors, and the returning officer decides within hours, often completing recounts on the same day to minimize delays. In , the (INE) mandates partial recounts in up to 9% of precincts nationwide if turnout exceeds 60% with anomalies, or in districts where vote margins fall below 1% or show statistical irregularities like improbable turnout patterns. For the 2012 presidential election, recounts covered ballots from over 54% of polling stations within 72 hours following complaints of inconsistencies, involving manual verification by citizen councils under INE supervision. Challenges to results can escalate to the Federal Electoral Tribunal, which may order full recounts in contested districts, prioritizing paper ballots and cross-checks against precinct records to resolve disputes empirically rather than automatically. Germany's system emphasizes rapid local counting at polling stations, with recounts triggered primarily through formal election protests filed by candidates or voters to district returning officers or, for federal elections, the Federal Constitutional Court, alleging procedural errors or miscounts. Under the Federal Elections Act, initial verifications occur locally before aggregation, and successful protests can lead to targeted recounts of ballots, as seen in post-2017 Bundestag election challenges where courts ordered partial re-examinations in specific constituencies without altering overall outcomes. The process prioritizes evidence of concrete irregularities, such as damaged ballots, over automatic thresholds, ensuring recounts remain exceptional to preserve election finality. In , electronic voting via direct-recording machines dominates, with no routine full recounts due to the absence of printed ballots until recent legislative pushes; instead, audits involve statistical sampling and parallel vote tabulation by parties post-election. The (TSE) handles disputes through , as in when fraud allegations prompted verification of machine logs and random audits across 5% of urns, confirming results without systemic changes. A 2025 bill approved by the aims to mandate vote printing for manual recounts in future cycles, addressing transparency concerns in close races.

Notable Recounts and Outcomes

United States Cases

The 2000 presidential election recount in Florida drew national attention due to its decisive impact on the Electoral College outcome. On November 7, 2000, Republican George W. Bush initially led Democrat Al Gore by 1,783 votes out of nearly 6 million cast, prompting an automatic machine recount under Florida law that narrowed the margin to 327 votes. Gore then requested manual recounts in four Democratic-leaning counties, where punch-card ballots exhibited issues like "hanging chads," but the process faced legal challenges over varying standards for vote tabulation. The U.S. Supreme Court halted the recounts on December 12, 2000, in Bush v. Gore, ruling that inconsistent county-level criteria violated equal protection principles and that completing the recount before the Electoral College deadline was infeasible, certifying Bush's 537-vote victory and securing his presidency. In the 2008 U.S. Senate election in , a recount reversed the initial results between and Democratic challenger . Initial tallies on November 4, 2008, showed Coleman ahead by 215 votes out of over 2.9 million, triggering an automatic recount due to the margin falling below 0.5 percent. The recount, completed by December 2008, shifted the lead to Franken by 225 votes after resolving disputed ballots, including absentee and rejected ones, through a bipartisan board. Coleman contested the results in court, alleging errors in ballot acceptance, but the upheld Franken's certification on June 30, 2009, by a 312-vote margin, prompting Coleman's concession the following day. Allegations of voter irregularities, such as double-counted absentee ballots in Democratic areas, surfaced but were not deemed sufficient to alter the outcome by . The 1960 presidential election in Illinois featured Republican demands for a recount amid claims of in 's Democratic machine, where defeated by 8,858 votes statewide. Nixon conceded on November 9, 1960, without pursuing a full recount despite irregularities like higher turnout in some precincts exceeding registered voters, opting instead to prioritize national unity. Partial Republican-led recounts in downstate counties confirmed Nixon gains but insufficient to overcome the margin, and no statewide recount materialized, leaving Kennedy's electoral votes intact. Subsequent analyses have debated the fraud claims, with some evidence of padded rolls but no consensus on outcome-altering scale. Other notable recounts, such as those in Georgia's 2020 presidential certification and various state races, have generally affirmed initial results with minimal vote shifts, underscoring the rarity of outcome reversals even in margins under 1 percent. Empirical reviews indicate that while recounts detect clerical errors, they seldom flip elections due to the inherent accuracy of modern tabulation systems, though pre-2000 punch-card methods amplified discrepancies as in .

International Cases

In Canada's 2025 federal election, recounts were triggered in four ridings due to margins under 0.1% of valid votes cast, as per rules. In Terrebonne riding in , a judicial recount flipped the result by a single vote, awarding the seat to the candidate and increasing their total to a slim majority under . In Terra Nova-The Peninsulas in , the recount resolved 1,041 disputed ballots, confirming a Conservative victory by 12 votes over the incumbent after an initial tie. Overall, these recounts netted the Conservatives two additional seats while the Liberals gained one, demonstrating how manual re-tabulation can resolve ambiguities in paper ballots but also highlighting procedural delays of up to two weeks. In Romania's 2024 presidential election first round, an unprecedented full recount ordered by the following allegations of irregularities confirmed the original tallies with negligible differences, preserving ultranationalist Călin Georgescu's 23% lead amid claims of foreign influence favoring him. The recount, covering all ballots, rejected most challenges and attributed minor discrepancies to clerical errors rather than , though the court later annulled the round entirely on December 6 due to declassified intelligence on covert online campaigns, not recount findings. Georgia's , , parliamentary saw opposition demands for a nationwide recount amid accusations, but the Central conducted a partial recheck at 364 polling stations (12% of total, representing 14% of votes), which upheld the party's 54% victory and reduced the margin only slightly from 163,000 to 152,000 votes. observers from the OSCE noted procedural flaws like vote-buying but did not dispute the recount's affirmation of results, underscoring limited scope in targeted re-examinations compared to full audits. In Iraq's 2018 parliamentary election, a manual recount of 10% of boxes in disputed areas, prompted by tampering claims against systems, confirmed initial outcomes with minimal adjustments, allocating seats proportionally under the and averting major shifts despite militia influence allegations. The process, overseen by the Independent High Electoral Commission, rejected most appeals but exposed vulnerabilities in , contributing to subsequent electoral law reforms.

Controversies and Criticisms

Effectiveness and Rarity of Outcome Changes

Empirical studies of statewide election recounts from 2000 to 2023 reveal that outcome changes are exceptionally rare. Out of 6,929 such elections, only triggered recounts, representing approximately one per 192 contests, typically in races with margins under 1%. Among these, just three resulted in a reversal of the winner—a rate of one per 2,310 elections—all occurring in initial margins of 0.06% or less. The average vote shift during these recounts was 551 votes, or 0.03% of total ballots cast, with shifts more often widening the original margin than narrowing it sufficiently to flip results. In cases where recounts have altered outcomes, such as the 2008 Minnesota U.S. Senate race—where Democrat Al Franken overcame a 215-vote deficit to Republican Norm Coleman by 225 votes after recount—the initial margin was extraordinarily narrow at 0.007%. Similar flips include the 2004 Washington state gubernatorial race and the 2006 Vermont state auditor's race, both with margins under 0.01%. However, broader data from earlier periods, such as New Hampshire elections from 1946 to 2002, confirm that discrepancies between initial tallies and recounts average around 1-2% in close races but seldom exceed random error levels sufficient to reverse winners without systematic tabulation failures. The limited effectiveness of recounts in changing outcomes stems from the nature of errors identified: typically small, unsystematic deviations from optical scanners, hand-counting inconsistencies, or provisional resolutions, rather than widespread or . A of Wisconsin's 2011 Supreme Court and 2016 presidential recounts found error rates of 0.21% to 0.86% at the level but net shifts too minor to affect winners, with margins of 7,004 and 22,748 votes respectively remaining intact. These findings underscore that while recounts enhance verification—detecting errors in about 1 in 116 to 475 ballots—they rarely produce decisive shifts beyond 0.024-0.03% of votes, even in contests decided by fractions of a percent. Internationally, patterns align with U.S. data, though fewer comprehensive studies exist; for instance, recounts in Canadian provincial elections and contests similarly confirm results in over 95% of cases, with flips confined to margins below 0.1%, as seen in rare instances like Quebec's 1998 by-election reversal by 11 votes. This rarity fuels on recounts' utility, as the procedural costs and delays often outweigh the infrequent corrections, prompting calls for targeted automatic triggers in ultra-close races paired with risk-limiting audits to prioritize efficiency over exhaustive re-tabulation.

Potential for Political Abuse and Delays

Election recounts carry inherent risks of political exploitation, as losing candidates or parties may request them strategically to postpone , impose financial burdens on election authorities, and foster about results even when margins exceed typical thresholds for meaningful reversal. In the United States, statutes in most states permit recounts upon request if the margin falls below a specified , often 0.5% or less, but broader provisions allow challenges on grounds of or , enabling extensions beyond strict recount triggers. Such maneuvers can serve aims, as evidenced by post-2020 analyses indicating that unsubstantiated contestations erode voter confidence without altering certified tallies, with experts warning that repeated invocation without constitutes an designed for accuracy assurance rather than adversarial prolongation. Delays from recounts frequently arise due to the labor-intensive nature of manual reexaminations, particularly in large jurisdictions, where verifying millions of ballots can span weeks or months, impeding timely transitions of power. The 2000 Florida presidential recount, initiated after an initial machine recount showed leading by 537 votes, escalated into machine and partial hand recounts across counties, culminating in a U.S. decision on December 12, 2000, to halt further proceedings and affirm Bush's victory, thereby delaying finalization and national resolution by over a month. Similarly, the 2008 Minnesota U.S. recount between and , triggered by a 215-vote initial margin, required multiple phases including absentee ballot challenges and a , extending from 2008 until certification on April 13, 2009, which postponed seating and legislative functionality. These instances illustrate how recounts, while verifying , can cascade into legal disputes that amplify uncertainty, with empirical data showing over 90% of U.S. recounts since 2000 confirming original outcomes, underscoring the disproportionate disruption relative to reversal frequency. In jurisdictions beyond the U.S., such as , recount provisions under the Canada Elections Act allow judicial recounts for margins under one-thousandth of votes cast, but recent municipal cases, like Edmonton's 2025 ward recount flipping a result after initial tabulation, highlight delays from resource strains and challenges, though less prone to national-level abuse due to centralized oversight. Australian federal elections employ preference distributions over complex counts that can extend days but rarely trigger abuse via recounts, as disputes route through courts rather than reballoting. Overall, while safeguards like cost recovery from requesters mitigate frivolous uses, the potential for delays to undermine institutional trust persists, particularly when politicized narratives exploit procedural timelines absent evidence of irregularity.

Reforms and Empirical Lessons

Empirical studies of U.S. statewide recounts conducted between 2000 and 2023 reveal that changes in election outcomes are rare, with winners flipping in fewer than 1% of cases, even among contests decided by margins under 1% of votes cast. In most instances, recounts yield minor vote adjustments—typically 0.1% or less of total ballots—attributable to human tabulation errors, ambiguous marks, or provisional ballot resolutions, but these seldom alter the leader. A notable exception occurred in the 2008 U.S. race, where a recount reversed an initial 215-vote margin favoring Republican , securing victory for Democrat by 225 votes after reviewing over 2.9 million ballots. Such reversals highlight recounts' potential to detect localized errors, yet comprehensive data underscores their confirmatory role over transformative impact, with over 90% of close-race recounts affirming original results. Recounts also demonstrate inefficiencies in resource allocation and timing. Full manual recounts, as in Florida's 2000 presidential contest, can extend certification timelines by weeks, incurring costs exceeding $10 million in some states due to , , and legal challenges. from post-2000 reforms indicates that machine-assisted initial counts with paper trails reduce recount discrepancies, but persistent overvotes and undervotes—often 1-2% of ballots—underscore the need for verifiable records to enable accurate re-tabulation without exhaustive manual review. Lessons from these processes emphasize causal factors like voter error and equipment variability over systemic , with statistical models showing that pivotal vote shifts are probabilistically negligible in large electorates. In response, U.S. jurisdictions have pursued reforms prioritizing statistical efficiency over exhaustive re-counting. Risk-limiting audits (RLAs), adopted in over a dozen states by 2025, sample ballots probabilistically to achieve a predefined risk limit—typically 5-10%—of failing to detect an incorrect outcome, often verifying results by inspecting 1-5% of ballots compared to 100% in traditional recounts. Pilots, such as Colorado's statewide RLAs since 2017, confirm reported tallies with high statistical assurance while reducing costs and duration; for example, a 2020 audit examined under 1% of ballots to affirm results statewide. These audits leverage mathematical guarantees absent in fixed-percentage recounts, addressing causal vulnerabilities like tabulation errors more scalably. Procedural reforms include tighter recount triggers and mandatory paper ballots. By 2025, nine states amended laws to adjust automatic recount thresholds—e.g., narrowing from 0.5% to 0.25% margins—or mandate before recounts, aiming to curb frivolous requests that delay certification without evidentiary basis. Post-2000 federal incentives via the Help America Vote Act promoted optical-scan systems with audit trails, empirically correlating with fewer disputed ballots in subsequent close races. Recommendations from analyses advocate automatic recounts only in sub-0.1% margins to balance verification against abuse, while expanding RLAs nationwide for routine post-election checks that preempt costly full recounts. These evidence-based shifts reflect a that targeted auditing enhances causal confidence in outcomes more effectively than reactive, labor-intensive re-canvassing.

References

  1. [1]
    Election Recounts - National Conference of State Legislatures
    Recounts are conducted after an election either automatically when the margin of victory for a race was narrow, or because someone (usually the losing ...
  2. [2]
    An analysis of statewide election recounts, 2000-2023 - FairVote
    Oct 28, 2024 · States should require automatic recounts in close races: When the margin is very close, a recount should occur automatically without a candidate ...
  3. [3]
    Recounts - Minnesota Secretary Of State
    A recount is a process to precisely determine the vote count between two candidates whose vote margin is extremely close. During a recount, election officials ...
  4. [4]
    Recount Guide - Maryland State Board of Elections
    A recount is the process of resolving a challenge to the final vote count reported for an election. A recount is limited to votes counted in a single contest.
  5. [5]
    [PDF] CONDUCTING A RECOUNT - U.S. Election Assistance Commission
    A recount provides an op- portunity for an election official to ensure that all the ballots cast are counted accurately and that the correct candidate or ballot ...
  6. [6]
    Controverted elections: how disputed results used to be part and ...
    Nov 13, 2020 · Disputed parliamentary election results – often taking months to resolve – were a frequent feature of English political culture before the reforms of the 19th ...
  7. [7]
    Judicial Recounts - Canadian Parliamentary Review - Article
    Judicial recounts involve having a judge review the ballots to determine the election results in a riding. The process first appeared in federal electoral ...
  8. [8]
    [PDF] recounts.pdf - Andrew Reeves -
    affect the voter and the democratic process. Thad Hall is an associate professor of political science at the University of Utah and a.Missing: notable | Show results with:notable
  9. [9]
    Election recount laws and procedures in the 50 states - Ballotpedia
    In 31 states voters can request recounts in only some elections or just for ballot measures. Can a candidate or voter request a recount? State, Can a candidate ...50-state overview of recount... · State-specific summary of...
  10. [10]
    Canada Elections Act
    ### Summary of Judicial Recount Provisions (Canada Elections Act, Section 300)
  11. [11]
    Recount procedures - Electoral Commission
    Dec 20, 2023 · ... request a recount on their appointment) may request to have the votes recounted or, following a recount, recounted again.2. The GLRO has ...
  12. [12]
    Complex count scenarios explained - Australian Electoral Commission
    May 7, 2025 · This page aims to identify and explain, using plain language, a number of potential scenarios that could occur during a federal election count.
  13. [13]
    Audits vs. Recounts: What's the Difference? - Verified Voting
    Nov 18, 2020 · An audit is a routine part of the post-election process designed to ensure systems worked as expected. A recount is performed in response to something that may ...
  14. [14]
    Post-Election Audits - National Conference of State Legislatures
    A post-election tabulation audit—often known as a post-election audit or PEA—checks that the equipment and procedures used to count votes during an election ...
  15. [15]
    [PDF] Post-Election Tabulation Audit Guide
    Nov 12, 2024 · Welcome! This is an introductory guide to post-election tabulation audits. It is designed primarily for election officials who want to begin, ...
  16. [16]
    Risk-Limiting Audits - Verified Voting
    A risk-limiting audit (RLA) checks a random sample of voter-verifiable paper ballots, giving strong evidence to support the reported election results.
  17. [17]
    [PDF] A Gentle Introduction to Risk-Limiting Audits
    Risk-limiting audits provide statistical assurance that election outcomes are correct by manually examining portions of the audit trail—paper ballots or voter- ...
  18. [18]
    Risk-Limiting Audits - National Conference of State Legislatures
    Risk-limiting audits are one version of a postelection audit, and a version that is growing in interest among election officials and legislators.What Are Risk-Limiting Audits? · Components of an RLA · Types of RLAsMissing: explained | Show results with:explained
  19. [19]
    [PDF] Risk-Limiting Post-Election Audits - Berkeley Statistics
    Stark, University of California, Berkeley. Page 2. The Risk-Limiting Audits Working Group is a self-organized group of election officials, researchers, and ...<|separator|>
  20. [20]
    State laws determine how and when recounts occur.
    Oct 26, 2020 · Twenty-one states and the District of Columbia provide for automatic recounts, based on the vote margin between the two candidates receiving the highest number ...
  21. [21]
    Here's how presidential election recounts work in key swing states
    Nov 5, 2024 · As Election Day arrives, polling still shows razor-thin margins between Vice President Kamala Harris and former President Donald Trump in ...
  22. [22]
    Recount Law Database - Verified Voting
    The Verified Voting State Recount Law Database is a resource that describes state laws, regulations, and procedures for post-election recounts.
  23. [23]
    Judicial Recount Handbook – Elections Canada
    Apr 30, 2025 · A judicial recount is a formal means of verifying the count of the votes cast for an electoral district. It is presided over by a judge of a superior court.
  24. [24]
    Canada Elections Act ( SC 2000, c. 9) - Laws.justice.gc.ca
    310 (1) After a recount, a candidate may make an application to the Chief Electoral Officer for reimbursement of his or her costs in respect of the recount, ...
  25. [25]
    Procedures to Request and Conduct a Recount
    This outline summarizes Texas election law with respect to recounts as currently codified. We have made every effort to insure the accuracy of this summary ...
  26. [26]
    Presidential election recounts cost more than a million dollars for 10 ...
    Dec 8, 2021 · Media interview Cobb County elections director during a hand recount in of the 2020 presidential elections. Photo was taken in November 2020 and ...
  27. [27]
    Election Results, Canvass, and Certification
    Have you ever wondered what happens to your ballot after you get your I Voted Sticker? After the polls close, ballots and electronic vote records are ...<|separator|>
  28. [28]
    Explaining how recounts and contested presidential elections work
    Oct 26, 2016 · Gore helped to settle a controversy over an automatic recount in Florida that decided the national presidential election. The initial vote tally ...
  29. [29]
    Candidate | Recounts and contested elections - FEC
    Federal Election Commission guidance for federal candidate committees on raising donations and making disbursements in connection with an election recount ...
  30. [30]
  31. [31]
    Provisional results and recounts - Electoral Commission
    Mar 14, 2024 · You should be satisfied that the number of votes for each candidate is accurate before proceeding to a provisional result.
  32. [32]
    How are general election votes counted and what are the recount ...
    Dec 12, 2019 · There are two stages to counting votes at UK general elections. ... What are the general election recount rules? A recount is allowed if ...Missing: United | Show results with:United
  33. [33]
    Recounts | Electoral Commission
    Feb 1, 2024 · Any candidate or election agent present when the counting of votes is completed will be given the opportunity to request a recount of the votes.Missing: procedures | Show results with:procedures
  34. [34]
    Challenge an election result: How to make a challenge - GOV.UK
    To challenge an election you must apply to the Election Petitions Office. This is called issuing an election petition.
  35. [35]
    Challenge an election result: When you can make a ... - GOV.UK
    You may be able to challenge the result of an election if you think it was not run properly, for example the votes were not counted correctly or a candidate ...
  36. [36]
    Runcorn by-election goes to recount as just four votes separate ...
    May 2, 2025 · This is the first major electoral test for the Labour government, as well as Kemi Badenoch and Nigel Farage, after the general election last ...
  37. [37]
    Blow to Sir Keir Starmer as Reform takes by-election by six votes
    May 2, 2025 · Reform UK's Sarah Pochin has won the Runcorn and Helsby by-election in a blow to Prime Minister Sir Keir Starmer's Labour Party by six votes.Missing: United | Show results with:United
  38. [38]
    'No change' after Sark's general election recount - BBC News
    Dec 13, 2024 · 'No change' after Sark's general election recount ... A birds-eye view of Sark. To the left is the land and on ... The government in Sark said there ...Missing: elections | Show results with:elections
  39. [39]
    Smallest majorities: The seats won by fewer than 100 votes - BBC
    Jul 5, 2024 · The smallest majority of 2024. It was a tense night for the candidates in Hendon, the closest constituency of the 2024 election. Labour ...
  40. [40]
    COMMONWEALTH ELECTORAL ACT 1918 - SECT 282 Re - AustLII
    COMMONWEALTH ELECTORAL ACT 1918 - SECT 282. Re - count of Senate votes to determine order of election in other circumstances. (1) Where the scrutiny in ...
  41. [41]
    we've seen some agonisingly tight elections, but Bradfield may ...
    May 25, 2025 · A recount in the Victorian seat of McEwen in 2007 gave victory to the Liberals by 12 votes. It's one of the three closest results in recent history.
  42. [42]
    Bradfield recount finalised - Australian Electoral Commission
    Jun 4, 2025 · The recount in the electoral division of Bradfield (NSW) finalised today. The final margin is 26 in favour of Independent candidate Nicolette Boele.
  43. [43]
    Count, count again: The closest votes in Australian history
    Oct 3, 2013 · If the Australian Electoral Commission orders one in Western Australia it will be the first ever for a Senate election;. • Why this time? The ...
  44. [44]
    [PDF] Recount policy | NSW Electoral Commission
    Feb 16, 2023 · 2.1. Section 172 of the Electoral Act provides: At any time before the declaration of an election result, the Electoral Commissioner may, if he.
  45. [45]
    Lok Sabha Elections 2024: When does a recount happen?
    Mar 26, 2024 · The process for a vote recount in the Lok Sabha elections is governed by specific procedures outlined by the Election Commission of India (ECI).
  46. [46]
    How Are Election Votes Recounted in India? Here's The Process
    Nov 11, 2020 · Earlier this week there was a lot of buzz about the US Presidential election results. We, in India, were watching the results unfold with a ...
  47. [47]
  48. [48]
    Election recount begins at more than half of Mexico polls
    Jul 4, 2012 · Ballots from more than 54% of polling places will be recounted within 72 hours, the Federal Electoral Institute (IFE) said. The figure marks a ...
  49. [49]
    [PDF] Mexico Relevant Aspects 2018 Electoral Process
    ∗ FPTP candidates appear on the obverse side of the ballot paper, while the party lists are on the reverse. ∗ Voters mark, solely, their preference for an FPTP ...
  50. [50]
  51. [51]
    [PDF] Federal Electoral Regulations - Die Bundeswahlleiterin
    Apr 19, 2002 · If more than one Postal Ballot Board is established for a constituency pursuant to section 8 subsection (1) of the Federal Elections Act, or if ...
  52. [52]
    Brazil election: Do voting machines lead to fraud? - BBC
    Oct 3, 2022 · President Bolsonaro says the voting machines used in Brazil are prone to fraud, but hasn't provided evidence.
  53. [53]
    [PDF] Practical guide : 2022 brazilian elections - TSE
    Aug 15, 2022 · If the polling stations have 30 to 99 voters, they cast their vote on a ballot. For places with more than 100 voters, electronic voting machines ...
  54. [54]
    Chamber of Deputies approves printing and recounting of votes
    Sep 24, 2025 · The Constitution and Justice and Citizenship Committee of the Chamber of Deputies approved, this Wednesday (11), Bill No.
  55. [55]
    US Presidential Election Florida Recount | Research Starters - EBSCO
    Bush and Democrat Al Gore. The election was marked by an extremely narrow margin in Florida, where Bush won by just 537 votes, triggering a recount due to the ...
  56. [56]
    Bush v. Gore | 531 U.S. 98 (2000) - Justia U.S. Supreme Court Center
    Florida did not need to complete a recount in the 2000 presidential election because it could not be accomplished in a constitutionally valid way.
  57. [57]
    On this day, Bush v. Gore settles 2000 presidential race
    Dec 12, 2023 · On December 12, 2000, the Supreme Court ended a Florida vote recount in the presidential election contest between George W. Bush and Al Gore.
  58. [58]
    [PDF] MINNESOTA'S HISTORIC 2008 ELECTION
    In the race for U.S. Senator, unofficial election results showed a vote margin of only 215 votes between incumbent Norm Coleman and candidate Al Franken—far ...
  59. [59]
    Coleman concedes race to Franken - POLITICO
    Jul 1, 2009 · Republican Norm Coleman has conceded to Democrat Al Franken in the Minnesota Senate race, ending one of the longest Senate races in American history.
  60. [60]
    Al Franken May Have Won His Senate Seat Through Voter Fraud
    Jul 20, 2010 · Minnesota Democrat Al Franken's narrow, 312-vote victory in 2008 over incumbent Sen. Norm Coleman may have come as the result of people being allowed to vote ...
  61. [61]
    Here's a voter fraud myth: Richard Daley 'stole' Illinois for John ...
    Aug 8, 2017 · The 1960 presidential election, when Chicago Mayor Richard J. Daley allegedly “stole” Illinois for the Democratic presidential campaign of John F. Kennedy.
  62. [62]
    Recounts almost never work. Except these three.
    Nov 28, 2016 · Initial result: On election night, Sen. Norm Coleman (R-Minn.) was ahead of challenger Al Franken (D) by just 206 votes out of more than 2.9 ...Missing: notable | Show results with:notable<|control11|><|separator|>
  63. [63]
    Canada's Liberal Party wins one more seat after Quebec recount
    May 11, 2025 · Supporters react at an election night event for Canadian Prime Minister and Liberal Party leader Mark Carney. A single vote in a Quebec riding ...
  64. [64]
    How a judge handled an 'unprecedented' 1,041 disputed ballots in a ...
    May 27, 2025 · A person walks over an Elections Canada sign. A recount for Terra Nova-The Peninsulas started on May 12. Conservative candidate Jonathan Rowe ...
  65. [65]
    Conservatives secure 2 more seats after tight federal election recounts
    May 23, 2025 · After a recount process that took nearly two weeks, Elections Canada announced Friday that Conservative Jonathan Rowe had defeated Liberal ...
  66. [66]
    Romania's vote recount confirms pro-Russian presidential bid
    Dec 2, 2024 · An unprecedented presidential vote recount in Romania has shown no marked difference from the original results, confirming the lead of pro-Russia nationalist C ...<|separator|>
  67. [67]
    The Romanian 2024 Election Annulment: Addressing Emerging ...
    Dec 20, 2024 · On December 6, Romania's Constitutional Court annulled the outcome of the country's first round of presidential elections.
  68. [68]
    Final recount confirms Georgia ruling party victory, says electoral ...
    Oct 31, 2024 · The central election commission told Agence France-Presse on Thursday that a recount at about 12% of polling stations, involving 14% of the vote ...
  69. [69]
    Recount Will Test the Integrity of Iraq's Elections
    Jun 11, 2018 · These incidents took place as the government looked into accusations of tampering with electronic voting machines, ballot box stuffing, voter ...
  70. [70]
    Why recounts rarely change the results of U.S. elections - Reuters
    Nov 12, 2020 · U.S. President Donald Trump hopes a recount of votes will help keep President-elect Joe Biden out of the White House, but as common as ...
  71. [71]
    Using Recounts to Measure the Accuracy of Vote Tabulations
    PDF | The 2000 presidential election exposed a surprisingly high level of inaccuracy in the tabulation of ballots. Differences between total vallots.<|separator|>
  72. [72]
    [PDF] Learning from Recounts - DSpace@MIT
    We compare the results of two recent statewide recounts in Wisconsin—the 2011 Supreme Court election and the 2016 presidential election.
  73. [73]
    What Happens if a Candidate Contests 2024 Election Results? | TIME
    Nov 4, 2024 · ... abuse of the system or abuse of the process. It also is potentially undermining confidence in the election for no legitimate reason.” Former ...
  74. [74]
    The Florida Recount Of 2000: A Nightmare That Goes On Haunting
    Nov 12, 2018 · The weeks-long battle over "hanging chads" that ultimately landed the fate of the presidency in the U.S. Supreme Court, continues to cast a ...Missing: notable | Show results with:notable
  75. [75]
  76. [76]
    Despite Inevitable Calls for Election Recounts, Research Shows ...
    Oct 28, 2024 · One thing that voters can inevitably count on after Election Night 2024 is that regardless of who wins, there will be calls in key ...
  77. [77]
    Election recounts rarely change the outcome of close races
    Oct 26, 2024 · Republican Sen. Norm Coleman led Democrat Al Franken by 215 votes in the initial count, out of more than 2.9 million ballots cast. After a hand ...
  78. [78]
    The Topline: How often do recounts change election results?
    Nov 18, 2024 · An election observer monitors two poll workers processing absentee ballots in the Milwaukee Election Commission warehouse on Aug. 13. Photo by ...
  79. [79]
    Recounts rarely change election outcomes, even in closely divided ...
    Nov 14, 2024 · (CBS 58) -- In a state that's evenly divided, it's not unusual to have close elections where officials are asked to recount the votes. However, ...
  80. [80]
    [PDF] Counts, Recounts, and Election Contests - Scholarship Repository
    Recommended citation: Steve Bickerstaff, Counts, Recounts, and Election Contests: Lessons from the Florida Presidential Election, 29 FLA. ST. U. L. REV. 425 ( ...Missing: "peer | Show results with:"peer
  81. [81]
    Learning from Recounts | Election Law Journal - Mary Ann Liebert, Inc.
    Abstract We compare the results of two recent statewide recounts in Wisconsin—the 2011 Supreme Court election and the 2016 presidential election.Missing: lessons | Show results with:lessons
  82. [82]
    [PDF] The Empirical Frequency of a Pivotal Vote
    Of course, a voter typically does not know the precise election returns before he casts his ballot, so a pivot election is only a subjective probability from ...
  83. [83]
    Nine states have changed their recount laws in 2025
    Oct 3, 2025 · In the 2026 midterms, some elections could be decided by thin margins. In 2024, three U.S. Senate races and six U.S. House races were ...
  84. [84]
    How each of the 7 key swing states would handle a 2024 recount
    Oct 31, 2024 · Some states automatically launch a recount if the results are within a certain margin, while others rely on candidates to ask for one.
  85. [85]
    [PDF] Conducting a Recount - Quick Start Guides
    As states can use hand-marked paper ballots, ballot marking devices, or DRE voting machines created by different companies, it is challenging to determine a ...Missing: oversight supervision
  86. [86]
    [PDF] Election Auditing: Best Practices and New Areas for Research
    The best-known type of election audit is the post-elec- tion tabulation audit that confirms the accuracy of electoral contests by ensuring that the votes were.