What manufacturing projects are eligible? The four tests to pass

A factory build, expansion or machinery purchase does not automatically qualify for manufacturing investment incentives. Eligibility depends on the state, the precise activity, advance approval, documented need for support and continued delivery against agreed milestones.

Foxconn came to Wisconsin with a promise big enough to make the rest of the story almost beside the point: 13,000 jobs and major incentives.

Then the project downsized.

The Associated Press reported on the changed project after the original ambition had become a national symbol of industrial policy. Foxconn’s Wisconsin experience does not prove that incentives fail. It proves something more useful. A large announced package is not the same thing as a benefit that will be earned, and an incentive is not the same thing as a factory.

That distinction matters to every manufacturer considering a new line, expansion, machinery purchase or site move. The tempting question is: “We are investing. What can we claim?” The better question is: “What have we committed to, what must happen first, and what would make the benefit disappear?”

Foxconn’s downsized Wisconsin project shows why large incentives need enforceable performance conditions. Photo: Chad Davis from Minneapolis, United States / Wikimedia Commons, CC BY 2.0.

The package was never the asset

Most people approach manufacturing investment incentives as if they were a discount at the checkout. Spend money on a building or equipment, fill in a form, collect the tax benefit.

That is usually the wrong mental model.

Many state programmes are discretionary, meaning they require an application and official approval. They are conditional bargains. The company receives support only if it makes a project fit a defined public purpose, such as creating jobs, increasing payroll, making a particular type of product or completing construction on time.

The Biggins Lacy Shapiro guide to US site selection describes the issue through the “but-for” test. A business may need to show that, but for the incentive, it would not have selected the same location or proceeded on the same terms. Sign a lease, announce the location or begin work too early, and that argument can become difficult to make.

This is where most people stop looking. They see the headline percentage or maximum credit. The real economics sit in the conditions underneath it.

Foxconn is the broad warning. The state programmes below show how those conditions work in practice.

Wisconsin. Photo: Wikideas1 / Wikimedia Commons, CC0.

Illinois: $10 million is only the entry ticket

Illinois created its Advancing Innovative Manufacturing, or AIM, programme in 2025 under Public Act 104-0006. At first glance, the capital threshold looks straightforward: at least $10 million of new investment within five years.

But the Illinois Department of Commerce and Economic Opportunity makes clear that size is not enough. The business must be primarily engaged, more than 50%, in research and development manufacturing of “critically needed goods”, including semiconductors, aerospace and robotics. The investment must also be new, and the company must apply before assets are placed in service, meaning before they are ready for their intended use.

The credit rises with investment: 3% on $10 million to $50 million, 5% on $50 million to $100 million, and 7% above $100 million.

That sounds generous. Yet a conventional factory upgrade of the same size could still miss the programme entirely if it does not fit the advanced-manufacturing and research-and-development purpose. Illinois is not offering a reward for spending money. It is trying to buy a particular kind of industrial activity.

Illinois AIM rewards a tightly defined advanced-manufacturing investment, not every large industrial upgrade. Photo: Unknown authorUnknown author or not provided / Wikimedia Commons, Public domain.

Imagine two companies each spending $20 million. One makes a qualifying critically needed good and applies before its assets go into service. The other expands a different industrial operation and applies after equipment is already operational. The budgets may look identical. Their eligibility does not.

New Jersey’s small but decisive line

New Jersey’s Next NJ Manufacturing programme opened on 25 September 2025 and is scheduled to run until funds are committed or 1 March 2029. Its headline maximum is up to $150 million.

That number deserves a raised eyebrow.

According to the New Jersey Economic Development Authority, a business needs at least $10 million of investment and 20 new full-time jobs. More than 50% of its activity must be qualifying manufacturing. The programme excludes refurbishing, packaging, software, agriculture and similar activities from that definition.

The credit is also not simply a percentage a company chooses from a brochure. It is the smallest of three figures: 0.1% multiplied by capital and job count, 25% of capital, or the $150 million maximum.

Here is the twist: a project can be large, employ people and have a serious capital plan, yet fail because too much of its actual activity is warehousing, packaging or another excluded function. The decisive work may be less glamorous than site selection. It may be the careful classification of what happens inside the building.

Put Illinois and New Jersey side by side and a pattern appears that neither programme states so plainly. Governments do not subsidise the word “manufacturing”. They subsidise their own definition of the future economy.

Pennsylvania measures a different promise

Pennsylvania’s Manufacturing Tax Credit does not begin with a capital-spend threshold in the supplied programme rules. It begins with payroll.

The Pennsylvania Department of Community and Economic Development says a business must increase annual taxable payroll by at least $1 million above its base year. The credit equals 5% of that increase, with a maximum credit of $4 million per fiscal year. It is also first-come, first-served.

This changes the decision a management team must make. In Illinois, the central question is whether the investment and activity are sufficiently advanced and targeted. In New Jersey, it is whether the business clears both investment and job hurdles while fitting the legal manufacturing definition. In Pennsylvania, the key forecast is the payroll increase.

A company can buy expensive machinery and still not create the payroll growth Pennsylvania requires. Another business, particularly one planning a labour-intensive expansion, may find payroll is the part of its story that matters most.

And timing still matters. Pennsylvania’s rules, as described by the DCED, require an application process and allocate the credit first-come, first-served. Hiring first and treating the credit as a retrospective rebate is a risky way to discover that the allocation has gone elsewhere.

Washington shows what “conditional” really means

Washington’s sales and use tax deferral for manufacturing is often easier to understand once you focus on the word “deferral”. It is not a waiver. Tax is postponed subject to conditions.

The Washington State Department of Revenue says an application must be submitted before construction begins or equipment is purchased. In the rural-counties programme, available to counties with fewer than 650,000 people, the deferral can be up to $400,000 per applicant. The recipient must begin construction within two years, complete the project within five years and report for up to seven years.

Fail those conditions and repayment or penalties can follow.

The interesting part is not merely that Washington demands paperwork. It is that the state continues to care after the ribbon-cutting. The benefit depends on a chain of events: apply before starting, start on time, finish on time, then keep reporting. Break one link and the cash-flow value changes.

That is why a deferral should not be placed in a financial model as if it were cash already sitting in the bank.

Why investment committees mark the number down

Founders often ask why an investor or lender refuses to give full value to a potential tax credit. “It is in the programme,” they say. “We meet the threshold.”

An investment committee sees three separate risks.

First, eligibility risk. The project may not satisfy the relevant activity definition, job count, payroll test or timing rule. New Jersey’s manufacturing boundary and Illinois’s qualifying-goods requirement are not administrative details. They decide whether the credit exists.

Second, delivery risk. A credit tied to future jobs, payroll or construction is only as reliable as the operating plan. If hiring is delayed, capital spending changes or a project scope is reduced, the benefit may be smaller than forecast.

Third, cash-flow risk. Some support arrives only after performance. Others can be repaid if conditions fail. Washington’s deferral illustrates the second possibility. The Vermont Employment Growth Incentive, or VEGI, programme illustrates the first: the New York State Department of Taxation and Finance’s review describes a structure in which incentives were earned year by year after firms met annual performance criteria, rather than paid upfront and later recaptured.

That is an important distinction. VEGI’s design can reduce the risk of receiving a benefit and later owing it back. It does not make the future benefit certain. It means the company must perform before it gets paid.

A committee therefore discounts contingent credits for the same reason a prudent household would not spend an annual bonus before it is earned. The headline amount may be real. But it is not yet available to service debt, absorb a cost overrun or rescue a weak operating plan.

GI Network’s view: The right way to present an incentive is not as a single headline number. GI Network would separate approved support from unapproved opportunities, test every milestone against the operating plan, identify where repayment exposure sits, and show lenders and investors the project’s cash needs if the incentive is delayed, reduced or never earned.

The exception that proves the rule

Not every tax benefit works this way.

Foley & Lardner distinguishes discretionary incentives from statutory credits, which are fixed in law and broadly available to taxpayers who meet their rules. The research brief identifies statutory research and development tax credits as an example that can be used retrospectively and does not require pre-approval or a but-for test.

That is the exception founders often have in mind when they assume every incentive is automatic.

It is not a general manufacturing answer. A statutory research and development credit is different from a negotiated or programme-based incentive attached to a particular factory, location, workforce commitment or construction timetable. Treating the two as interchangeable is how optimistic models get built.

What operators should do before committing

Before a site decision, build an incentives file. It should state the legal entity, candidate locations, precise activity conducted at the site, capital plan, equipment purchases, hiring plan and expected taxable payroll.

Then test each programme against the actual project, not the marketing description. Read exclusions. Check whether the application must precede equipment orders, construction or place-in-service. Identify whether benefits depend on jobs, payroll, capital expenditure or all three.

Create one calendar for approval, construction start, equipment purchase, project completion and reporting. If the project changes, revisit the award terms before changing the scope.

That discipline also belongs in wider capital planning. A promised workforce expansion is only credible if the company can staff it, which is why underwriting workforce risk before talent constrains capex matters before a job commitment enters an application.

Finally, map each incentive to a specific cost or outcome. Large projects can sometimes stack discretionary incentives, Foley & Lardner notes, but only when structured carefully. Do not assume two programmes can reward the same cost, job or payroll increase without restrictions.

The one-page test

For investors, the practical question is not “How big is the incentive?” It is “What has to go right before this becomes usable money?”

Experienced investors look for the difference between an approved award and a possible award; between a benefit earned on completion and one earned annually; between a tax deferral and a permanent saving; and between a forecast job count and a binding performance condition. They also ask whether the project still works if the incentive is delayed or reduced.

That is the same habit that makes a financing case more credible in a boardroom. As how investors read a pitch deck explains, evidence should answer the next diligence question, not merely make the opportunity sound larger.

Use the Before, Definition, Delivery test:

1. 1.Before: Was the company approved before an irreversible commitment, such as construction, equipment purchase, lease signing or assets entering service?
2. 2.Definition: Does the actual activity, not the company’s label, meet the programme’s manufacturing, technology, capital, jobs or payroll rules?
3. 3.Delivery: What must happen after approval, when will value be received, and what repayment or penalty applies if the plan changes?

If one answer is unclear, the incentive is not yet an asset. It is a possibility with conditions.