2026-08-20T08:00:29.099Z / https://www.cnn.com/2026/08/20/politics/flawed-census-report-alleged-noncitizen-voting
前普查局领导层、统计学家和投票权倡导者们发出警告，称该局网站上发布了一份不同寻常的无署名新报告，声称有数万名外籍人士在2020年美国总统大选中投了票。
唐纳德·特朗普总统周二在Truth Social平台上宣扬了这份报告结果，称普查局在分析了1.28亿份个人选民记录后，证实有超过2.4万名外籍人士在2020年非法投票。
“我赢得了大选！”特朗普写道，他再次重复了毫无依据的主张，称大规模选举舞弊让他输掉了2020年的竞选。他预测，一旦分析人员处理完剩余的3200万份选民记录，这一数字将“大幅攀升”。
负责监管普查局的商务部部长霍华德·卢特尼克随后发帖称：“这只是开始”，并预测下一次分析“可能会显示又有数万名非法选民”。
即便这份报告准确无误，2.4万张选票也仅占2020年约1.6亿张总投票数的0.02%不到。乔·拜登以超过700万票的优势赢得了全国普选票。
这一调查结果远低于拜登在竞争最激烈的摇摆州的胜选优势：
- 亚利桑那州： 据称有1100张外籍人士选票；拜登以10457票胜出
- 佐治亚州： 据称有400张外籍人士选票；拜登以11779票胜出
- 威斯康星州： 据称有500张外籍人士选票；拜登以20682票胜出
该报告也没有——也无法——说明这些选民支持了哪位候选人，因为选票选择是保密的。
这份引发争议的调查结果发布之际，外界对普查局面临的政治压力日益担忧。该局历来由职业统计学家领导，独立于白宫运作。
三名曾在特朗普政府时期任职的前普查局官员透露，今年早些时候，特朗普政府将一批来自支持特朗普的美国优先政策研究所的党派分析师安插进了普查局。美国优先政策研究所未立即回应置评请求。
普查局、商务部和白宫均未回应置评请求。
特朗普政府加紧努力寻找证据，以支持其长期以来关于大规模选举舞弊的主张。上个月在黄金时段的讲话中，特朗普援引一批新解密的文件，声称美国选举系统存在严重漏洞，未来大选可能面临外国干预和大规模外籍人士投票。
周二发布的文件显示，普查局的分析将商业选民数据库与美国国税局、社会保障管理局、国务院和国土安全部的联邦记录进行了比对。
联邦记录显示，这家商业数据来自弗吉尼亚州的Data-Clear公司，该公司在2024年获得了一份64.95万美元的联邦合同，用于获取选民登记数据。
分析人员使用普查局的记录关联系统，寻找与联邦政府数据库中外籍人士记录匹配的个体选民。
这份被称为“初步分析”的周二报告立即引起了前普查局高级官员的警惕。他们表示，发布此类无署名作者或贡献者的文件极不寻常，这可能意味着职业工作人员被边缘化了。
2016年至2022年担任该局首席科学家的约翰·M·阿博德告诉CNN，该报告的统计数据太小，无法与常规匹配误差区分开来，不能作为外籍人士投票的确凿证据。
他说，即使设计精良的记录关联系统也可能产生错误匹配。在分析的1.28亿份记录中，2.4万份外籍人士匹配结果约占0.019%。
“我在普查局认识或共事过的专业研究人员都不会从这项研究中得出这样的结论，”阿博德说，“专业研究人员会报告这一结果，并给出相应的背景信息。”
前普查局局长罗伯特·桑托斯（2022年上任，在特朗普第二任期初期辞职）将这份报告的发布形容为明显的“白宫劫持”。
“我感到震惊和沮丧，居然会有这类文件——我拒绝称之为报告——出现在普查局的网站上，”桑托斯告诉CNN，“这显然是白宫和商务部分析师的产物。”
他表示，这项分析的另一个“致命缺陷”在于其基于一个错误的假设：联邦、州和地方的记录完全准确。
选民档案通常包含过时或错误的信息。这些问题在新冠疫情期间可能更为严重，当时选举办公室面临关闭、人员短缺和其他 disruption。
“完全有理由相信所使用的数据质量非常可疑，因为这些数据的编制期间并非正常情况，”桑托斯告诉CNN。
前官员们表示，常规的普查局分析会明确列出这些警告和局限性。周二的文件却没有。
卢特尼克的部门在特朗普打击所谓选举舞弊和重塑联邦选举规则的推动中承担了越来越重要的角色。今年3月，特朗普在椭圆形办公室签署一项行政命令，指示政府编制各州公民名单，并要求邮政服务加强对邮寄选票的审查，卢特尼克当时就站在特朗普身边。
上个月，国土安全部部长马克韦恩·穆林表示，他正与卢特尼克合作，利用联邦资金作为杠杆，迫使各州采纳特朗普支持的选举改革。
“这意味着，如果这些州想要获得拨款并获得联邦选举运营的报销，它们就必须实施安全措施，”穆林告诉记者。
政府越来越多地利用联邦数据库来寻找外籍人士投票的证据。上个月，穆林声称国土安全部审查了公开的选民登记名单，发现新泽西州、内华达州、加利福尼亚州和宾夕法尼亚州共有超过25万名潜在外籍选民——这一数字遭到了这些州的一些官员的质疑，他们要求提供国土安全部计算该数字的细节。
“又是老一套，使用有缺陷的数据库和方法来炮制总统会喜欢的故事情节，但实际上与现实不符，”无党派投票权倡导组织“共同事业”的政策主任兼投票与公平代表法律顾问安德鲁·瓦尔丘克说。
瓦尔丘克称最新的普查局报告是一场“闹剧”，旨在破坏人们对选举的信任。
Data-Clear总裁卡罗琳·卡尔森告诉CNN，该公司从各州获取并提供给普查局的选民信息包括可用的出生日期，但不包括社会保险号码。
专家们质疑，对该数据的分析是否包含足够的个体选民具体信息，以防止假阳性结果。
特朗普政府的司法部已起诉30个州，要求获取未编辑的选民记录，包括驾照号码和部分社会保险号码，但多次被法院驳回。与司法部对抗的州援引联邦和州隐私法称，这些法律保护选民的个人信息。
卡尔森表示，该公司无法也未参与普查局对其提供的数据的分析，并将报告结果的相关问题交由普查局处理。
佛罗里达大学选举专家迈克尔·麦克唐纳表示，普查报告中引用的数字可能包括事实上的美国公民。“我们应该将这2.4万人视为指控，而非外籍人士的确凿证据，”他说。
“选举官员知道，数据中存在陷阱，个人可能被指认为外籍人士，但实际上确实是公民，他们可以提供公民身份的证据，”他说。
他补充说，通常情况下，当地选举官员会联系选民，给他们机会提供证据，证明自己有资格留在选民名单上，然后再将其除名。
“这份所谓基于普查数据的分析经不起推敲，”前司法部律师、现任非营利组织“选举创新与研究中心”负责人大卫·贝克尔说，“普查数据无法可靠地与公开的选民数据匹配，因为它缺乏可与这些公开选民记录进行比对的唯一匹配标识符。”
CNN的伊森·科恩对本文亦有贡献。
Experts slam ‘flawed’ Trump administration Census report on alleged noncitizen voting
2026-08-20T08:00:29.099Z / https://www.cnn.com/2026/08/20/politics/flawed-census-report-alleged-noncitizen-voting
Former Census Bureau leaders, statisticians and voting-rights advocates are raising alarms about an unusual, unsigned new report posted on the agency’s website that claims tens of thousands of noncitizens voted in the 2020 presidential election.
President Donald Trump touted the results Tuesday on Truth Social, saying the bureau had proved that more than 24,000 noncitizens illegally voted in 2020 after analyzing 128 million individual voter records.
“I WON THE ELECTION!” Trump wrote, repeating his unsubstantiated claim that widespread fraud cost him the 2020 race. He predicted the number would “explode” once analysts processed the remaining 32 million voter records.
Commerce Secretary Howard Lutnick, whose department oversees the Census Bureau, later posted, “It’s just the beginning,” predicting the next analysis “will likely show tens of thousands of additional illegal voters.”
Even if the report is accurate, 24,000 votes would amount to less than 0.02% of the roughly 160 million ballots cast in 2020. Joe Biden won the national popular vote by more than 7 million votes.
The findings fall well below Biden’s victory margins in the most closely contested states:
- Arizona: 1,100 alleged noncitizen votes; Biden won by 10,457
- Georgia: 400 alleged noncitizen votes; Biden won by 11,779
- Wisconsin: 500 alleged noncitizen votes; Biden won by 20,682
The report also does not — and cannot — say which candidates those voters supported because ballot selections are secret.
The disputed findings come amid growing concern about political pressure on the Census Bureau, which has traditionally been led by career statisticians and operated independently of the White House.
Earlier this year, the administration installed a group of partisan analysts from the Trump-aligned America First Policy Institute inside the Census Bureau, according to three former bureau officials who served under Trump. AFPI did not immediately respond to a request for comment.
The Census Bureau, Commerce Department and White House did not respond to a request for comment.
The administration has intensified its effort to find evidence backing Trump’s longstanding claims of widespread voter fraud. In a primetime address last month, Trump cited a trove of newly declassified documents to claim the US election system has gaping vulnerabilities that leave future contests open to foreign interference and widespread noncitizen voting.
The Census Bureau analysis compared a commercial voter database with federal records from the IRS, Social Security Administration, State Department and Department of Homeland Security, according to the document released Tuesday.
The commercial data came from Data-Clear, a Virginia-based firm that received a $649,500 federal contract in 2024 to obtain voter registration data, federal records show.
Analysts used the Census Bureau’s system for linking the records, searching for individual voters who match noncitizen records in the federal government databases.
Tuesday’s report, dubbed a “beginning analysis,” immediately raised red flags among former senior Census Bureau officials. Publishing such a document without named authors or contributors is highly unusual, they said, and could indicate that career staff were sidelined.
John M. Abowd, the bureau’s chief scientist from 2016 to 2022, told CNN the report’s statistical finding is too small to distinguish from routine matching errors and claim as definitive proof of noncitizen voters.
Record-linking systems can produce false matches, even when they are well designed, he said. The 24,000 noncitizen matches represent about 0.019% of the 128 million records that were analyzed.
“None of the professional researchers that I knew or worked with at the Census Bureau would make such a claim from that study,” Abowd said. “Professional researchers would have reported that result and given it some context.”
Former Census Bureau Director Robert Santos, who held the role from 2022 until resigning in the first days of Trump’s second term, described the publication as an obvious “hijack by the White House.”
“I was shocked and dismayed that a document of this sort – and I refuse to call it a report – would appear on a Census Bureau website,” Santos told CNN. “It is clearly a product of White House and Department of Commerce analysts.”
He said another “fatal flaw” of the analysis is that it rests on a faulty assumption: that federal, state and local records are completely accurate.
Voter files often contain outdated or incorrect information. Those problems may have been worse during the Covid-19 pandemic, when election offices faced shutdowns, staffing shortages and other disruptions.
“There’s every reason to believe that the quality of the data that’s being used is highly suspect, because it was not business as usual when those data were being developed,” Santos told CNN.
A typical bureau analysis would spell out those caveats and limitations, the former officials said. Tuesday’s document did not.
Lutnick’s department has taken on a growing role in Trump’s push to root out alleged voter fraud and reshape federal election rules. He stood beside Trump in the Oval Office as the president signed an executive order in March directing his administration to compile state-by-state citizenship lists and ordering the Postal Service to step up scrutiny of mail ballots.
Last month, Homeland Security Secretary Markwayne Mullin said he was working with Lutnick to use federal funding as leverage to pressure states to adopt Trump-backed election reforms.
“Meaning that if these states want a grant and they want to be reimbursed to run federal elections, they’re going to have to implement security issues,” Mullin told reporters.
The administration has increasingly tapped federal databases to bolster its search for noncitizen voting. Last month, Mullin claimed DHS had reviewed public voter rolls and found more than 250,000 potential noncitizens registered across New Jersey, Nevada, California and Pennsylvania — a figure some officials in those states disputed, demanding details about how DHS calculated it.
“It’s the same old story of using flawed databases and ways to generate a storyline that the president will like, but actually doesn’t align with reality,” said Andrew Walchuk, Policy Director and Counsel for Voting and Fair Representation at Common Cause, a non-partisan voting-rights advocacy organization.
Walchuk called the latest Census Bureau report a “farce” intended to undermine trust in elections.
The voter information that Data-Clear obtained from states and provided to the Census Bureau included dates of birth where available but not Social Security numbers, Data-Clear’s President Carolyn Carlson told CNN.
Experts questioned whether the analysis of that data contained enough specific information about individual voters to guard against false-positive results.
Trump’s Justice Department has sued 30 states in an effort to obtain unredacted voter records, including driver’s license numbers and partial Social Security numbers, but has been repeatedly rebuffed by the courts. The states fighting the DOJ cite federal and state privacy laws that they say protect voters’ personal information.
Carlson said the firm had no control or input into how the Census analyzed the data the firm provided and referred questions about the report’s findings to the agency.
Michael McDonald, an elections expert at the University of Florida, said the numbers cited in the Census report likely include people who are in fact US citizens. “We should treat these 24,000 as allegations, not as definitive proof of noncitizens,” he said.
“Election officials know that there are pitfalls in the data where individuals may be alleged to be a noncitizen but are actually indeed a citizen, and they can provide proof of their citizenship,” he said.
Typically, local election officials will reach out to the voters to give them an opportunity to provide evidence that they are qualified to remain on the voting rolls before removing their names, he added.
“This alleged analysis based on Census data fails the laugh test,” said David Becker, a former Justice Department lawyer who now oversees the nonprofit Center for Election Innovation & Research. “Census data cannot be reliably matched to public voter data, as it lacks any unique matching identifiers to compare to those public voter records.”
CNN’s Ethan Cohen contributed to this report.
发表回复