This House would keep Microsoft, Adobe and the outsourcing majors suspended from the PERM green-card program until they change their hiring practices.

PRO — Opening · Challenger (openai/gpt-6-luna)

Keep the suspensions until these employers can show they have changed how they hire. PERM is not an automatic reward for sponsoring foreign workers: employers must certify that the job cannot be filled by a U.S. worker. The government alleges that some firms paired U.S. layoffs with continued sponsorship in related jobs and used ineffective recruitment to make jobs appear unfillable (packet articles 1–2). Those allegations are not proven in court—but they are serious enough to justify pausing new and pending certifications while the Labor Department investigates, as its rules allow (packet article 1).

This is a targeted, reversible enforcement measure, not a ban on immigrant workers: H-1B status and the right to work remain intact (packet article 1). And it has leverage precisely because these firms have sponsored at enormous scale: together they received more than 100,000 permanent labor certifications and 230,000 H-1B approvals (packet article 1). Restore PERM access when they can demonstrate genuine recruitment and stop using sponsorship alongside practices that displace U.S. workers.

The concern is not merely theoretical. Lottery-based research finds that additional H-1B workers can substantially crowd out other employment, with at most modest effects on patenting; another model finds native computer scientists can lose wages and jobs even while consumers and the broader economy gain (packet articles 3–4). That is a reason to enforce the program’s labor protections—not to abandon immigration. The right test is simple: firms that meet those protections regain access; firms that have not changed should not keep using the program as if nothing is wrong.

CON — Opening · Champ (deepseek/deepseek-v4.1-flash)

The motion asks you to keep the suspension running until these companies change their hiring practices. That is an indefinite, unfalsifiable condition — and the packet’s own description gives it away: the suspensions “will last as long as needed,” the administration claims “the tools to extend them indefinitely,” and the stated objective is “a change in hiring practices” (packet article 2). Change them how? Measured by whom? Cleared when? Nobody can say. A sanction whose exit condition is never defined is not enforcement; it is a ban wearing a compliance costume.

Now look at what the government’s case actually rests on. Article 2 is admirably blunt: the targeting tool is public data — “announced layoffs, state WARN notices, and published LCA and PERM disclosure data” — and any employer that “has reduced headcount while continuing to sponsor foreign nationals in the same or related occupations presents a comparable profile, regardless of intent” (packet article 2). Layoffs plus sponsorship is a pattern, not a violation. Laying off workers in one division while sponsoring in another is lawful. The motion asks this House to ratify punishment on profile, before adjudication. Note the contrast the packet itself draws: the harsher, defined penalty — debarment, capped at three years, after process — “has not been announced” (packet article 1). The government chose the tool with no clock over the tool with a clock.

Second, the punishment lands on the wrong people. PERM is the worker’s path to a green card, not a corporate perk. Workers already in line lose the priority date that fixes their place, and recruitment windows — “generally 180 days” — expire during the freeze, “forcing restarts” (packet article 1). Article 2 adds the sharper edge: H-1B employees “approaching the six-year limit often rely on a pending or approved labor certification to extend status,” so the freeze converts directly into “attrition risk” for people who did nothing wrong. If your theory is that workers were underpaid, the remedy that locks them into temporary status at the mercy of the sponsoring employer makes their bargaining position worse, not better. And it cannot protect a single U.S. worker: H-1B hiring continues, because H-1B was never suspended (packet article 1). The suspension removes the green card, not the job.

Third, notice that the motion’s own logic runs backwards on its evidence. PRO cites the crowd-out findings in article 3 and the wage effects in article 4. But read article 4 to the end: immigration “increased the overall welfare of US natives,” lowered prices, raised IT output 1.9–2.5%, and consumers and complementary workers gained — it is “a redistribution question as much as a totals question” (packet article 4). Article 3’s own summary of the literature says crowd-out is contested and one robustness check finds far smaller effects. If the harm is distributional, the answer is wage floors, audits, and targeted prosecution of proven fraud — all tools the packet says already exist and are being used, including the DOL Inspector General’s strike team and “ongoing prosecutions” (packet article 2). You do not fix a wage problem by freezing the queue.

And the motion has a cost the other side cannot wave away. The U.S. share of top AI researchers fell from 46% to 34% while China’s rose from 27% to 41%; restrictive visa policy is one of the two named drivers (packet article 5). Article 5’s own framing: keeping researchers out “does not stop them from doing frontier work — it simply changes where that work happens.” An indefinite freeze on the largest sponsors is a machine for changing where that work happens.

Here is the test I want this bout fought on: a sanction is legitimate when it has a defined trigger, a defined standard, and a defined end. This one has a pattern, a press conference, and “as long as needed.” If the Labor Department can prove fraud, it has audits, debarment, and criminal referral — and article 1 says Blanche has already raised charges. Use them. Adjudicate, then punish. Do not keep eight employers and thousands of blameless workers in permanent limbo because the government prefers leverage to proof.

PRO — Rebuttal · Challenger (openai/gpt-6-luna)

Con treats “until they change” as “forever,” but those are not the same. The packet says the suspensions are investigative pauses under a rule that allows an initial 180 days and extensions while an investigation or court case is pending; it does not say the firms are permanently barred (packet article 1). And the objective is explicit: change hiring practices. The standard need not be mysterious: conduct genuine recruitment and stop pairing layoffs in related work with sponsorship practices that undermine the required labor-market certification (packet articles 1–2). If the firms already do that, they can demonstrate it. If not, lifting the suspension would make the safeguard meaningless.

Con is right that layoffs plus sponsorship alone do not prove fraud. But it stretches that point into a demand to wait for final proof before the Labor Department can pause cases—even though the cited rule permits a pause when an employer is under investigation for possible fraud (packet article 1). This is a temporary enforcement measure, not a finding of guilt. And Con overstates the alternative: the packet says debarment is available, up to three years, but does not say it must follow a particular adjudication process; nor has it been announced (packet article 1). “Use prosecution instead” does not explain why the Department should keep processing certifications while it investigates.

The worker harms are real, but Con misstates one detail: the packet says workers lose time and recruitment windows may expire; it says the priority date usually comes from when PERM was filed, not that the freeze necessarily erases an already-established priority date (packet article 1). H-1B workers also keep their current status and right to work. Disruption is a reason to resolve the investigation promptly, not to grant employers continued access regardless of their practices. And while PERM suspension does not stop H-1B hiring, it does pause the green-card certification process whose labor-market test is at issue.

Finally, Con stretches article 5: it discusses restrictive visa policies, not this PERM pause, and the packet expressly says H-1B visas are not suspended here (packet articles 1, 5). The AI talent trend cannot establish that a PERM suspension drives researchers abroad. Articles 3 and 4, meanwhile, do not prove every foreign hire displaces an American—but they do show why labor-market protections matter: immigration can depress native wages and employment even while producing broader gains (packet articles 3–4). The answer is not to abandon immigration; it is to keep these firms suspended until they show those protections are real in their hiring.

CON — Rebuttal · Champ (deepseek/deepseek-v4.1-flash)

PRO’s rebuttal quietly conceded the central point. Asked what “change their hiring practices” means, PRO offers “conduct genuine recruitment and stop pairing layoffs in related work with sponsorship.” But that second clause is article 2’s targeting profile — the one that applies “regardless of intent” to “any employer that has reduced headcount while continuing to sponsor foreign nationals in the same or related occupations” (packet article 2). PRO has not defined a standard of compliance; PRO has restated the pattern. And the pattern is lawful. Laying off in one division while sponsoring in another is not fraud, and article 2 says so in plain terms.

Worse, the remedy is circular. The labor-market test PRO wants to vindicate is the PERM recruitment process. “Keep them suspended until they show those protections are real” asks firms to demonstrate compliance by performing the very process the government has switched off. Suspension does not verify recruitment; it prevents recruitment from being examined. If the theory is that the certifications are false, the answer is to audit the recruitment file — which the DOL already has the strike team and “ongoing prosecutions” to do (packet article 2). PRO asks why the Department should keep processing while it investigates. Because processing generates the evidence. A freeze generates delay.

On the clock: PRO says 180 days plus extensions. But article 1 says extensions run “until the investigation and any court case finish,” Blanche says criminal charges are possible, and article 2 says the suspensions last “as long as needed,” with tools to extend them “indefinitely.” PRO never answered the contrast I drew: debarment — a defined penalty, capped at three years — was available and “has not been announced” (packet article 1). The government chose the open-ended instrument over the defined one. And PRO’s claim that debarment need not follow “a particular adjudication process” is argued from the packet’s silence. You cannot read that silence as process for debarment and as no-process for the pause.

On workers: PRO concedes the disruption is real, then waves it off as a reason to hurry. But the six-year link cannot be waved off — article 2 says H-1B employees near the limit “often rely on a pending or approved labor certification to extend status,” and the freeze becomes attrition risk. On the priority date, I’ll grant PRO’s correction and press the point PRO skipped: article 1’s “forcing restarts” means the recruitment window and certification begin again — more years in line for the same worker, at the mercy of the same employer. PRO’s defense, that “H-1B status and the right to work remain intact,” describes a worker with a job, no green card, and no exit. That is not protection. That is leverage.

And here is what PRO still has not answered: not one U.S. worker is protected, because H-1B hiring continues — the packet says so explicitly (packet article 1). PRO’s reply is that PERM’s labor-market test is paused. That is the circularity again. You do not protect U.S. workers by declining to run the test that protects them; you protect them by running it, and by punishing those who lie on it.

Article 5 is not severed, it is welded. PRO says the trend concerns “restrictive visa policies” generally. A PERM suspension is a restriction on the visa path, and article 2 supplies the transmission belt: workers near the six-year limit depend on the labor certification to stay. Cut it, and the retention risk article 5 names becomes concrete. PRO’s own packet joins the two.

Articles 3 and 4 remain what they were: evidence that protections matter, which I grant — and no evidence that this remedy works. PRO proved a harm and asserted a cure.

The test stands unanswered. Defined trigger: none. Defined standard: the pattern, “regardless of intent.” Defined end: “as long as needed.” Adjudicate, then punish. Do not hold eight employers and thousands of blameless workers in limbo because leverage is easier than proof.

PRO — Closing · Challenger (openai/gpt-6-luna)

CON established real costs: workers face delay and possible attrition, and immigration can benefit the wider economy. Those are reasons to resolve investigations quickly, not to keep processing certifications while the reliability of the required labor-market test is under investigation. H-1B status and work rights remain intact; this is a pause on PERM, not a ban on immigrant workers (packet article 1).

The layoff-and-sponsorship pattern is a reason to investigate, not proof of guilt or a rule that layoffs alone bar access. The relevant standard is the one PERM already requires: genuine recruitment and a truthful certification that the job cannot be filled by a U.S. worker. CON’s demand to audit and prosecute rather than pause does not explain why new certifications should proceed before that test is trusted. The Department’s rule permits pauses during investigations of possible fraud, initially for 180 days and longer while an investigation or court case continues (packet article 1). And the stated objective is a change in hiring practices—not punishment without an exit condition (packet article 2).

So the decisive point survives: this is a targeted, reversible enforcement lever aimed at employers with enormous sponsorship volume, while the underlying allegations concern whether their hiring practices satisfy PERM’s labor protections (packet articles 1–2). The strongest case for PRO is simple: when the gatekeeping process may be undermined, pause access to that process while investigating; restore it when employers demonstrate genuine recruitment and truthful compliance. CON showed that delay hurts and that enforcement must be bounded. It did not show why these firms should retain PERM access before they demonstrate the practices the program requires.

CON — Closing · Champ (deepseek/deepseek-v4.1-flash)

The bout turns on a question PRO never answered: what does compliance look like, and who certifies it? PRO’s closing says the standard is “genuine recruitment and a truthful certification.” But when pressed, PRO defined the violation as “pairing layoffs in related work with sponsorship” — which is verbatim article 2’s targeting profile, the one that applies “regardless of intent” to “any employer that has reduced headcount while continuing to sponsor foreign nationals in the same or related occupations” (packet article 2). That is lawful conduct. PRO never separated the pattern from the violation, because the government’s own stated method doesn’t.

Then the circularity, which PRO’s closing restates instead of resolving. “Pause access to that process while investigating” — but the process is the evidence. PERM recruitment generates the file that shows whether the certification was truthful. Switching it off doesn’t verify recruitment; it prevents recruitment from being examined. The Department already has the strike team, the audits, and “ongoing prosecutions” (packet article 2). PRO asked why certifications should proceed during an investigation. Because a freeze produces delay; processing produces proof.

The exit condition is the cleanest kill, and it survived to the last speech. Article 1 says extensions run “until the investigation and any court case finish”; article 2 says suspensions last “as long as needed” with tools to extend them “indefinitely.” Debarment — a defined penalty, capped at three years — “has not been announced” (packet article 1). The government chose the open-ended instrument over the bounded one, and PRO’s answer that debarment might also lack process is an inference from silence, which cuts equally against PRO.

On workers, PRO conceded the disruption and offered “resolve promptly.” But article 2’s six-year-limit dependency and article 1’s “forcing restarts” mean years re-added to a queue for people who did nothing wrong — and H-1B hiring continues, so not one U.S. worker is protected by this (packet article 1). PRO’s best line, that PERM’s labor-market test is at issue, is the circularity again.

Articles 3 and 4 prove a harm. They do not prove this cure. Article 5’s retention risk is made concrete by article 2’s six-year link.

The test I set in opening still stands with nothing on the other side of it: defined trigger, none; defined standard, the intent-blind pattern; defined end, “as long as needed.” If the Labor Department can prove fraud, it has audits, debarment, and criminal referral. Adjudicate, then punish. Do not hold eight employers and thousands of blameless workers in limbo because leverage is easier than proof.