Why Trump’s big claims of voter fraud have yielded so few prosecutions

Never before has the federal government been more focused on finding evidence of mass noncitizen voting. What it’s found is exponentially more modest.

Under the orders of President Donald Trump, the administration has launched a far-reaching effort to find evidence that foreigners are tainting elections. Top officials have claimed that they believe hundreds of thousands of noncitizens are on voter rolls, and they’ve threatened election officials with prosecution for not doing enough to stamp out the supposed problem. But those figures have narrowed significantly under additional scrutiny, and much smaller still is the number of people the administration has charged criminally for noncitizen voting.

The rhetoric continues claims that Trump has made for at least a decade, alleging that the votes of millions of noncitizens were the sources of Democratic electoral victories. But even with hundreds of government employees now investigating the issue, they’ve come nowhere close to supporting it.

“Not only do we have no evidence that noncitizen voting is happening at scale,” said Wren Orey, the managing director of the Bipartisan Policy Center’s Elections Project, “But when there are cases that appear to be noncitizen voting or registration, when you dig into the numbers, very often, they fall apart.”

Part of the issue, according to election officials and other voting experts, is that the administration is relying on flawed data-matching programs to make its initial claims, in a process that it has shrouded with secrecy. The administration has been quick to rush out the maximalist interpretation of what those matching programs show, without any hedging or acknowledgement that the numbers must be further investigated, and that they’ll shrink, once they are.

Homeland Security Secretary Markwayne Mullin claimed in July that his agency had found 250,000 noncitizens on the voter rolls in just four Democratic-led states. Letters his office sent that week to the states in question — California, Pennsylvania, Nevada and New Jersey — were more hedged about the actual number. The agency, in the months since, followed up with some of the individual states with a much smaller list of “high-confidence” matches, and even those lists appear to inflate the problem, according to analysis the election officials have done.

DHS is still claiming that, after a review of more than 100 million voter records across both Democratic- and Republican-led states, it has found 300,000 “potential” noncitizens. As of mid-September, the agency had opened 1,620 investigations based on the review, and so far, the effort has led to around 160 arrests — a number that covers both criminal arrests and arrests for administrative immigration proceedings.

More than 800 DHS employees are involved in the effort to review and process the data-matching efforts, a government attorney said in court this week.

“The number of arrests of aliens who unlawfully registered and voted continues to increase daily as the Trump Administration enforces the law,” a DHS spokesperson told CNN. “Every illegal registration and every illegal vote will be investigated, and those responsible will be held accountable.”

DHS has not been working alone on Trump’s hunt for voter fraud. The Justice Department brings any prosecutions that arise from the DHS investigations, and US Attorney’s offices across the country have devoted significant resources to the investigations as well.

California has been ground zero of this effort. Last week, the agency held a press conference, led by the number two official at the department and the US Attorneys from the state, to roll out two dozen criminal cases accusing noncitizens illegally voting. The announcement appeared aimed at boosting a civil lawsuit the Justice Department has filed, demanding that California elections officials do more to verify the citizenship status of 20,000 individuals on the voter rolls before counting any ballots from them.

“And to be clear, that’s just the floor. We’re doing rolling review – DHS is — so the number is likely to grow,” acting Deputy Attorney General Trent McCotter said. “Imagine … 20,000 citizens, enough to fill a small stadium. You go and talk to every one of them and say, “Well, congrats on voting. But your vote got canceled out by someone who’s not allowed under federal law to vote.’”

But the picture painted by the prosecutions is far different than the narrative put forward by Trump and his top allies that Democrats are purposely flooding the border with undocumented migrants to help them rig elections. With just 23 California cases alleging noncitizen voting in Trump’s second term — and similar or smaller numbers of cases charged in other states — the DOJ’s enforcement efforts so far suggest the impact of unlawful foreign votes is well, well short of the scale of tipping elections. Furthermore, most of the defendants are legal permanent residents, some of them having lived in the United States for decades.

Many are not charged with noncitizen voting. Instead, they are charged with making false statements during their naturalization process, during which they were asked whether they had voted or had been registered to vote.

The office of California Secretary of State Shirley Weber received some 300 grand jury subpoenas since August, lawyers for the Democrat said in court filings last week.

The top Trump-appointed prosecutor in Los Angeles argued in court that DOJ doesn’t have the ability to charge every noncitizen voter, and that prosecutions primarily serve the role of “deterrence.”

“The standard keeps shifting. First the claim was that noncitizen voting doesn’t happen. Now that federal prosecutors are charging it in districts across the country, the claim is that it doesn’t happen enough to matter,” a DOJ spokesperson to CNN, citing specific elections that were decided by a small number of votes. “Close elections are not rare, and every illegal ballot cancels out the vote of an American citizen.”

In its litigation with DOJ, California officials have noted that post-election procedures exist for scrutinizing close contests for the potential that fraud swayed the results.

Ryan Germany, a former investigator for the Georgia Secretary of State, noted that prosecutions can be difficult to bring because the law typically requires that the defendant was acting intentionally. In some states, noncitizens have ended up on voter rolls because of glitches in state software, particularly at the DMV.

Often, his investigations encountered people who had not fully realized they had been registered -–perhaps because they had encountered a voter registration drive and they filled out the paperwork just because they were trying to move on from the interaction.

Germany said that, nonetheless, it’s a problem when noncitizens are getting registered, whether accidentally or not, and “I do think that some states could do a better job” of addressing it.

But, he said, we “really haven’t seen a lot of, ‘I am a noncitizen, I am going to try to vote.”

As the administration has railed against Democratic election officials for allegedly not doing enough to prevent noncitizens from registering, it’s touted its work with “cooperative” Republican led states.

That work has largely been done through a data system known as the Systematic Alien Verification for Entitlements or SAVE. The data tool compares registered voters to federal immigration records for evidence of their citizenship status. The system has long been available to election officials to use for auditing voters, but the administration has beefed up the program by expanding the federal records sources it pulls from and by making it possible for states to upload their entire voter lists en masse. (Previously, election officials used SAVE for individualized reviews).

The White House this summer released figures claiming that thousands of noncitizens had been found on voter rolls in states like Georgia, Missouri, North Carolina and Texas. But election officials have cautioned that the SAVE matches must go through additional rounds of verification, and that the numbers tend to get smaller once local officials investigate further.

“What SAVE is telling you, if they can’t confirm the person is a citizen, they’re not necessarily telling you they’re not a citizen,” Germany said. He praised SAVE as a tool but said that vetting its results required “a lot of manpower.”

“The delta between this initial number and the actual of noncitizens registered, in our experience, that’s a pretty big delta,” he told CNN.

Texas was one of the first states to us the expanded version of SAVE and its review last year suggested 2,724 noncitizens could be on the state’s voter rolls. That number shrunk by more than 500 when the administration added passport data to the system, and more still were verified as eligible voters using state records and other means.

Officials in Collin County confirmed citizenship for 73 people of the 109 tagged as potential noncitizens by SAVE by using state records. Bastrop County found that four of the eight people on its initial list were citizens, according to information provided to CNN by the Texas Association of County Election Officials, while two of the six people on the SAVE list Hays County received proved their eligibility with documents showing their citizenship.

By last month, the state elections office had made criminal referrals for a very small portion of the original list — sending just 117 cases to state prosecutors for potential noncitizen voting. The Texas Attorney General’s office did not respond to CNN’s inquiry about the referrals. But the DOJ, under Trump’s second term, has brought just nine cases in Texas alleging noncitizens had voted.

A similar dynamic played out in Maricopa County, Arizona — the second largest voting jurisdiction in the country and a target of Trump’s 2020 election fraud conspiracy theories. Arizona has a proof of citizenship requirement for voter registration, making SAVE a very common tool used by election officials to verify voters’ eligibility.

County Recorder Justin Heap, a Trump ally who oversees voter registration in Maricopa, announced earlier this year a review he did of 60,000 people caught up in a DMV software glitch that raised questions about their eligibility.

Of that batch, 137 were identified by SAVE as potentially being noncitizens, and Heap said that 60 individuals had records showing they had voted in recent elections. He announced the review in a social media post titled, with a siren emoji, “NONCITIZENS VOTING IN ARIZONA.” His language suggested that the results of the matching program were the final confirmed number.

He then sent that information to Maricopa’s top prosector to review. County Attorney Rachel Mitchell said that she ultimately confirmed that 52 of the 60 voters were actually citizens, shrinking the number of alleged noncitizen voters to just eight.

联邦政府从未像现在这样如此重视寻找大规模非公民投票的证据。但它找到的证据却远没有想象中那么多。

在唐纳德·特朗普总统的指示下，本届政府发起了一项影响深远的行动，旨在寻找外国人操纵选举的证据。高级官员声称，他们认为有数十万非公民登记在选民名册上，并威胁要起诉那些未能尽力消除这一所谓问题的选举官员。但经过进一步审查，这些数字已大幅缩减，而本届政府因非公民投票而提起刑事诉讼的人数则更少。

这种言论延续了特朗普至少十年来一直坚持的说法，即数百万非公民的选票是民主党赢得选举的来源。但即便现在有数百名政府雇员正在调查此事，他们的调查结果也远未支持这种说法。

“我们不仅没有证据表明非公民投票正在大规模发生，”两党政策中心选举项目总经理雷恩·奥雷说，“而且，当出现一些看似非公民投票或登记的案例时，当你深入研究数据时，往往会发现这些案例站不住脚。”

选举官员和其他投票专家指出，部分问题在于，政府依赖存在缺陷的数据匹配程序来做出初步声明，而这一过程却被政府严密掩盖。政府急于对这些匹配程序显示的结果做出最极端的解读，既没有保留意见，也没有承认这些数据需要进一步调查，而且一旦调查结果出来，数据就会有所缩水。

国土安全部长马克韦恩·穆林7月份声称，他的部门仅在四个民主党执政的州就发现了25万名非公民选民。他办公室当周发给这四个州——加利福尼亚州、宾夕法尼亚州、内华达州和新泽西州——的信函对实际数字则含糊其辞。此后几个月，该部门又向部分州提供了一份规模小得多的“高置信度”匹配名单，而根据选举官员的分析，即使是这些名单似乎也夸大了问题的严重性。

美国国土安全部仍然声称，在审查了民主党和共和党执政州超过1亿份选民记录后，发现了30万名“潜在”非公民。截至9月中旬，该机构已根据审查结果展开了1620项调查，迄今为止，这些行动已导致约160人被捕——其中包括刑事逮捕和因行政移民程序而被捕的人员。

本周，一名政府律师在法庭上表示，超过 800 名国土安全部员工参与了数据匹配工作的审查和处理。

“随着特朗普政府加大执法力度，非法登记和投票的外国人被捕人数每天都在增加，”国土安全部发言人告诉CNN。“每一起非法登记和每一张非法选票都将受到调查，肇事者将被追究责任。”

在特朗普对选民欺诈的调查中，国土安全部并非孤军奋战。司法部负责提起任何由国土安全部调查引发的诉讼，全国各地的美国检察官办公室也投入了大量资源参与调查。

加州一直是这项行动的中心。上周，该机构举行新闻发布会，由该部门二号人物和加州联邦检察官共同主持，公布了二十多起刑事案件，指控非公民非法投票。此举似乎旨在推动司法部提起的一项民事诉讼，该诉讼要求加州选举官员在统计任何选票之前，对选民名册上的两万名选民的公民身份进行更严格的核实。

“需要明确的是，这只是最低数字。我们正在进行滚动审查——国土安全部也在进行审查——所以这个数字很可能会上升，”代理副司法部长特伦特·麦科特说。“想象一下……两万名公民，足以填满一个小型体育场。你去和他们每个人谈话，告诉他们：‘恭喜你投票。但是你的选票被一个根据联邦法律没有投票权的人作废了。’”

但检方的调查结果与特朗普及其主要盟友的说法截然不同，后者声称民主党人故意向边境投放大量非法移民，以帮助他们操纵选举。在特朗普的第二个任期内，加州仅有23起指控非公民投票的案件——其他州的案件数量也与之相近或更少——司法部迄今为止的执法行动表明，非法外国选票的影响远未达到足以左右选举结果的程度。此外，大多数被告都是合法永久居民，其中一些人已在美国生活了几十年。

许多人并非因非公民投票而被起诉，而是因在入籍过程中作虚假陈述而被起诉。在入籍过程中，他们被问及是否投过票或是否登记过投票。

加州州务卿雪莉·韦伯的办公室自 8 月以来收到了约 300 份大陪审团传票，民主党候选人的律师上周在法庭文件中表示。

特朗普任命的洛杉矶首席检察官在法庭上辩称，司法部没有能力起诉每一位非公民选民，而且起诉的主要作用是“威慑”。

“标准一直在变。一开始他们声称非公民投票根本不存在。现在联邦检察官在全国各地的选区提起诉讼，他们又改口说这种现象太少，无关紧要，”司法部发言人对CNN表示，并列举了一些仅以少数选票决定胜负的具体选举。“选情胶着的选举并不罕见，每一张非法选票都会抵消一位美国公民的投票权。”

加州官员在与司法部的诉讼中指出，选举后有相关程序来审查势均力敌的选举，以确定是否存在舞弊行为影响了选举结果。

曾任佐治亚州州务卿办公室调查员的瑞安·格曼指出，由于法律通常要求被告人是故意行事，因此起诉往往十分困难。在一些州，由于州政府软件（尤其是车辆管理局的软件）出现故障，导致非公民最终出现在选民名册上。

他的调查经常会遇到一些没有完全意识到自己已经登记的人——可能是因为他们遇到了选民登记活动，填写表格只是为了尽快结束这次互动。

德国方面表示，尽管如此，非公民被登记注册仍然是一个问题，无论是有意还是无意，“我认为一些州可以做得更好”来解决这个问题。

但他表示，我们“真的很少看到有人说，‘我不是公民，但我打算尝试投票。’”

政府一方面抨击民主党选举官员，指责他们没有采取足够措施阻止非公民登记投票，另一方面又吹嘘自己与“合作的”共和党领导的州的合作。

这项工作主要通过一个名为“外国人资格系统核实”（Systematic Alien Verification for Entitlements，简称SAVE）的数据系统完成。该数据工具将已登记选民与联邦移民记录进行比对，以验证其公民身份。选举官员长期以来一直可以使用该系统对选民进行审核，但本届政府通过扩大其所依赖的联邦记录来源，并允许各州批量上传其全部选民名单，从而加强了该项目的功能。（此前，选举官员使用SAVE进行的是个别审查。）

今年夏天，白宫公布的数据显示，在佐治亚州、密苏里州、北卡罗来纳州和德克萨斯州等州的选民名册上发现了数千名非公民。但选举官员提醒说，SAVE 系统匹配到的信息必须经过多轮核实，而且一旦地方官员进行深入调查，最终数字往往会减少。

德国表示：“SAVE系统无法确认某人是否为公民，但这并不意味着他们就一定不是公民。” 他称赞SAVE系统是一个有效的工具，但也指出核实其结果需要“大量人力”。

“根据我们的经验，最初的统计数字与实际登记的非公民人数之间的差距相当大，”他告诉 CNN。

德克萨斯州是最早使用扩展版“选民登记系统”（SAVE）的州之一，该系统去年的审查结果显示，该州选民名册上可能有2724名非公民。在政府将护照信息添加到系统后，这一数字减少了500多人，此外，通过州政府记录和其他途径，还有更多非公民的选民资格得到核实。

科林县官员利用州政府记录，确认了SAVE系统标记的109名潜在非公民中的73人的公民身份。据德克萨斯州县选举官员协会向CNN提供的信息，巴斯特罗普县发现其最初名单上的8人中有4人是公民，而海斯县收到的SAVE名单上的6人中有2人提供了证明其公民身份的文件。

截至上个月，州选举办公室仅对最初名单中的一小部分人提起刑事诉讼——仅将117起案件移交州检察官，指控他们可能存在非公民投票的情况。德克萨斯州总检察长办公室没有回应CNN关于这些移交案件的问询。但自特朗普第二个任期以来，司法部在德克萨斯州仅提起过9起指控非公民投票的案件。

类似的情况也发生在亚利桑那州马里科帕县——该县是美国第二大投票辖区，也是特朗普2020年大选舞弊阴谋论的目标。亚利桑那州要求选民登记时提供公民身份证明，因此SAVE系统成为选举官员核实选民资格的常用工具。

负责马里科帕县选民登记的县记录员贾斯汀·希普是特朗普的盟友，他于今年早些时候宣布，他对因车辆管理局软件故障而受到影响的 60,000 人进行了审查，该故障引发了人们对他们资格的质疑。

在这批人中，SAVE 组织认定有 137 人可能为非公民，希普表示，其中 60 人的记录显示他们曾在近期选举中投票。他在社交媒体上发布了一篇题为“非公民在亚利桑那州投票”的帖子，并配上了一个警笛表情符号，宣布了这项审查。他的措辞暗示，匹配程序的结果就是最终确认的数字。

随后，他将这些信息发送给了马里科帕县首席检察官进行审查。县检察官雷切尔·米切尔表示，她最终确认60名选民中有52名实际上是公民，这意味着被指控的非公民选民人数减少到仅8人。