Rep. Ro Khanna, joined by Tom Suozzi and Debbie Dingell, introduced the Industrial Bank for American Manufacturing Act on July 23, 2026 — a bill that would redirect up to 50 percent of Section 301 tariff revenue on Chinese imports into a new federal bank authorized to deploy up to $15 billion annually in grants, loans, and equity investments for small and medium-sized manufacturers in de-industrialized regions. The proposal arrives with manufacturing employment at roughly 12.6 million jobs, down from a peak of about 19.6 million in 1979, and 75,000 jobs lost since January 2025 alone, per Bureau of Labor Statistics data.

The bill’s structural topology is hub-and-spoke: the federal bank at the center, Section 301 tariff revenue as the sole upstream input channel, and recipient manufacturers as downstream spokes. That single-channel capitalization is the proposal’s load-bearing design choice — and the source of its most consequential dynamic. The bank’s entire capital base is bound to tariff revenue. Every upstream failure mode that shrinks that revenue — a U.S.-China trade agreement reducing 301 rates, statutory sunset of the tariff authority, congressional appropriation inaction, or administrative non-collection — narrows all spokes simultaneously. This distinguishes the proposal sharply from the Reconstruction Finance Corporation, the New Deal comparator Khanna’s framing invokes implicitly: the RFC drew on $500 million in Treasury stock, $51 billion-plus in cumulative Treasury borrowing, and public bond issuance, reaching $55 billion-plus in wartime lending from a diversified capital base. The Khanna bank, as described, cannot issue its own bonds.

Yet the single-channel dependency is simultaneously the bill’s strategic leverage. Any tariff rollback, under the proposal’s structure, would directly defund a domestic lending program with geographically specific beneficiaries and recorded congressional commitments. The $100 million threshold for mandatory congressional approval is the enforcement mechanism: every large loan requires a recorded vote, converting diffuse tariff costs into concentrated, traceable benefits with individual political accountability. The same lock-in that makes the bank vulnerable to upstream policy shocks also makes it harder for downstream policy actors to reverse.

A design tension between rhetoric and architecture

Khanna’s framing centers on a Cleveland anecdote — a small manufacturer needing roughly $1 million in capital to make a part locally. The statutory architecture, however, authorizes per-loan ceilings of $500 million and an annual cap of $15 billion. The bank would need at least 30 maximum-size loans per year to exhaust its authorization, or many thousands of $1 million loans. The $100 million approval threshold effectively splits the bank into two operational regimes with sharply different political dynamics: sub-$100 million loans requiring no recorded vote and carrying low political visibility, and loans above that threshold becoming public commitments by legislators from affected districts.

Actual allocation practice would determine whether the bank primarily serves small manufacturers — as the rhetoric suggests — or larger, politically traceable projects as the architecture enables. The tension is structural, not rhetorical: the bill creates a vehicle that can operate at very different scales, and which scale dominates will determine both its political sustainability and its economic impact.

The three historical precedents as a coherent framing network

Khanna invokes Alexander Hamilton (federal financial institution-building), FDR and Bill Knudsen in the 1940s (wartime industrial mobilization), and the Marshall Plan (regional reconstruction). These three precedents share a common theme of large-scale state-directed economic intervention under conditions of perceived crisis. They operate at the framing level, not the operational level — they shape how the bill is interpreted, not how the bank would function — but their internal coherence matters strategically. The Marshall Plan analogy in particular frames de-industrialized American regions as analogous to war-damaged foreign economies needing reconstruction, raising the rhetorical stakes for opposition beyond ordinary policy disagreement.

The game: why the bill likely dies in committee, and why that’s not the point

The sequential game has a short-horizon subgame-perfect equilibrium: Democratic leadership introduces; Republican leadership, holding procedural control, buries the bill in committee without a floor vote. Procedural burial avoids both the revenue diversion and any concession of policy credit on manufacturing at no immediate electoral cost, because manufacturing decline is gradual enough that blame remains diffuse. The terminal-node analysis is straightforward — propose, schedule-or-bury, floor-vote-or-defect — and the subgame-perfect strategy for Republican leadership is to never schedule.

That equilibrium is robust only under a short time horizon. Extend the game to repeated electoral interaction across multiple Congresses, and the calculus shifts. If manufacturing employment continues declining and the 2026 electoral cycle intensifies pressure in rust-belt districts, a Republican majority facing credible Democratic challenges may find that co-sponsoring a modified version — smaller authorization, different funding source — yields higher payoffs than blanket opposition. Cooperation becomes rational when the game repeats and the discount rate on future electoral consequences rises.

The alternative framings sharpen this. Individual members from de-industrialized districts — including Republican-held PA-13 (Johnstown, Rep. John Joyce), OH-06 (Lordstown), and PA-01 (Lower Bucks County) — face reputational costs from opposing a bill visibly targeted at their own constituents. The bipartisan geographic spread of named target regions complicates any party-line burial strategy. And if China retaliates against agricultural exports rather than manufacturing — targeting the same districts the bank intends to serve — it creates cross-pressure on members whose farming constituents bear costs from a trade dispute the bank does not compensate.

What the bill actually does (and does not do) for manufacturing employment

The bank addresses only the capital-access component of a multi-factor manufacturing employment decline. It does not address trade policy, automation, supply-chain restructuring, or the relationship between tariff policy and consumer prices. Khanna’s claim that “all the capital is going to build AI, technology apps and financial firms” operates as correlational framing — it describes a pattern of capital concentration alongside a manufacturing capital deficit, but does not demonstrate a diversion mechanism, and the bill does not reallocate capital away from tech. It creates a separate funding stream.

The credibility landscape

Nearly everything about the proposal is, at this stage, cheap talk. The “boldest industrialization proposal since FDR” claim has no commitment device. The $15 billion authorization and $500 million per-loan cap describe a future institution that does not yet bind. The promise that 50 percent of Section 301 revenue will be available annually has no statutory entitlement backing it without enactment. Republican leadership’s threat of procedural burial, by contrast, is credible: the institutional means exist, the immediate cost is low, and no concentrated countervailing pressure has materialized.

The tariff-stickiness threat toward China is cheap talk today — without the bank, no sunk-cost investment in maintaining tariff revenue exists. If the bank were created, that would constitute a sunk-cost commitment making the tariff threat more credible. The $100 million congressional-approval threshold would function as a second-order commitment device: any legislator voting yes on a large loan would subsequently face a concentrated political cost if tariffs fell, because the loan’s viability depends on continued tariff revenue.

Strategic recommendations from the game structure

Four moves emerge from the analysis:

First, treat the bill as a focal-point anchor, not as legislation to pass this Congress. Reintroduce it after each procedural burial, linking each reintroduction to the latest BLS manufacturing-employment data point. Each cycle raises the narrative cost for Republican incumbents in rust-belt districts who must explain why they declined to act while jobs continued leaving. The objective is to shift the repeated-game payoff structure until a future Congress faces a different equilibrium.

Second, seek a multi-year tariff authorization tied to the bank’s capital — making tariff revenue less discretionary upfront. This would front-load the commitment device currently deferred to bank enactment, shifting the information structure from incomplete (unobservable commitment) to verifiable (statutory obligation) on the tariff dimension.

Third, pre-file a list of specific projects in qualifying districts before the bank opens, forcing members to vote publicly on tangible loans. This would front-load the reputational costs that would later protect the funding stream, converting a cheap-talk promise into specific, traceable commitments that any tariff rollback would have to reverse visibly.

Fourth, form a cross-party coalition of manufacturing-district members grounded in the Republican-held seats in targeted regions. The geographic spread of named target regions already substantiates a bipartisan-constituency claim; activating it as an organized voting bloc would alter the payoff structure for Republican leadership by making procedural burial electorally expensive for individual members rather than costless at the party level.

The structural finding that binds the analysis together

The bill’s structural flaw — single-channel capitalization tying lending capacity to tariff politics — is also its strategic leverage. The same dependency that makes the bank vulnerable to upstream policy shocks (a trade agreement reducing tariffs would defund the lending program) also raises the political cost of tariff reduction (any tariff rollback encounters concentrated opposition from bank beneficiaries and legislatively traceable loan commitments). This tension persists regardless of whether the bank itself is ever created. The proposal’s design encodes an implicit threat: creating this institution would lock tariff policy into an industrial-finance logic that outlasts any single administration’s trade-policy preferences.

Whether that threat is credible depends on factors the bill’s text does not address: the strength of the political-feedback coupling between large-loan votes and tariff-regime sustainability; whether $15 billion annually can meaningfully affect a multi-factor employment decline across automation, trade, and supply-chain restructuring; and whether the full bill text contains contingency provisions — bond issuance, Treasury borrowing, counterparty guarantees — that would decouple the bank from tariff dependency. The Guardian report from July 23, 2026, supplies the originating account; the full bill text was not located for this analysis.

Additional considerations

The analysis surfaces several open questions the proposal does not address. Political viability of redirecting tariff revenue from the Treasury’s general fund is a fiscal-policy question the bill’s framing bypasses. The relationship between tariff policy and consumer prices — a trade-policy impact dimension — is absent from the proposal’s pitch. The proportionality of $15 billion annually to the scale of a manufacturing employment decline driven by automation, trade, and supply-chain restructuring is a labor-economics evaluation for which no Census Annual Capital Expenditures Survey comparison was supplied. Labor unions are a natural coalition partner absent from Khanna’s framing; major manufacturers with global supply chains would likely lobby against reshoring subsidies that disrupt their cost structures; and agricultural exporters in retaliatory crosshairs are not modeled in the bill’s base case. These missing players would shift the expected outcome if activated.

Analytical techniques used in this piece

This analysis applies the methods below. Each links to a short, plain-English explainer you can read and reuse.

Relationship Mapping
Extracts the network of ties among people, institutions, and entities.
Stakeholder Mapping
Charts the parties to a situation — their interests, power, and alignments.
Strategic Interaction (Game Theory)
Models a situation as a game — players, moves, payoffs, and likely equilibria.