Economic development’s broken math of buying jobs

Nobody came to a public hearing three years ago on a data center expansion in New York state’s Hudson Valley. 

After 20 minutes in silence, the Rockland County Industrial Development Agency closed the hearing and approved a deal that could be called the country’s most lopsided.

The package for the JPMorganChase Data Center: $77 million in projected sales-tax exemptions, roughly $40 million of it state money, and just one additional permanent job. The site already took millions in earlier breaks, as reported by New York Focus. In total, just 25 people work there.

Ahead of midterm elections, data centers have become a defining issue. Small-town hearings like the one that went unattended back in February 2024 are now packed to overflowing. The quietly powerful state and local organizations that use public budgets to shape their economies are under scrutiny. 

What is finally breaking the 20th century big-game hunting model for a new generation of economic development leaders?

The economic development profession’s organizing body, the International Economic Development Council, or IEDC, this month celebrates its 100th anniversary at its annual conference. Several thousand practitioners will attend in New Orleans. I’ll be one of them. Ahead of that, this is the first of a three-part series looking at what is changing about this important, but obscure, cornerstone of our communities. This first story starts with what’s breaking. Next week we look at what’s changing, and we’ll close with what comes next. 

But first: Why would anybody forgo tens of millions of dollars for their community in exchange for an unsightly warehouse for computers that two-thirds of Americans now say they oppose?

To answer that, it helps to understand the trade’s origins, and the traditional source of its power: the ability to attract jobs.

Why does the government subsidize business at all?

Inspired by FDR’s New Deal, Depression-fighting state economic interventions spread. Mississippi’s 1936 Balance Agriculture with Industry program leased publicly financed factories cheap to Northern manufacturers. After the Second World War’s devastation, lessons from rebuilding the economies of Europe and Japan came home to the US.

By the 1960s, dozens of US states had formed formal economic development authorities, commonly including Mississippi’s strategy to subsidize expansions by established firms. In 1969, New York State authorized its industrial development agencies — one such organization hosted that empty data center expansion hearing in 2024.

Confronting decades of declining manufacturing, high inflation and geopolitical turmoil in the 1970s, newly formed economic development agencies sought new industries to win.

One was consumer credit. A 1978 Supreme Court ruling let banks export their home state’s interest-rate limits nationwide, and Delaware’s 1981 Financial Center Development Act removed the caps to attract them. It’s why so many banks have offices in tiny Wilmington even today (and why so many Americans carry credit card debt).

The bargain made sense for a time. Local leaders needed revenue for roads and schools, and most of it came from businesses. Those businesses were hard to move, with their inventories and their machines, and they employed a lot of people. Forgoing some revenue now for a payroll that would generate more of it later penciled out.

In that pursuit, why risk scarce public dollars to develop homegrown entrepreneurs, with their tiny and fragile new businesses, when you could simply attract the ones that were already growing? 

That is still economic development’s traditional foundation. One current official, who ran his state’s economic development authority before winning statewide office, described the practice to me plainly: “Some people might call it, you know, buying jobs.”

The economy no longer supports ‘big-game hunting’ (if it ever did) 

Cracks started to appear with the Great Recession, sowing anti-corporatist seeds across the political spectrum — as a previous Technical.ly special report documented.

In 2015, creative-class urbanist Richard Florida told me the 20th-century business-attraction practice of “big-game hunting” with tax incentives was dying. If true, it had one last gasp. 

The strategy reached its apex two years later, when 238 state and local coalitions prostrated themselves before one of the world’s largest corporations in the Amazon HQ2 economic development beauty contest. The multi-trillion-dollar public company eventually selected Arlington, Virginia — it’s much scaled back campus is far from the most interesting thing about the urbane and dynamic satellite city.

I can’t imagine so many places enduring another tax incentive version of The Bachelor, trying to win the chance to give away stuff to a corporation. 

What is finally breaking the 20th century big-game hunting model for a new generation of economic development leaders?

Three things: Our economy changed, and what value there ever was has been competed away. Plus, research has stacked up that the strategy was perhaps never as effective as we thought.

First, forgoing tax revenue or granting public dollars to induce a local investment makes a bit more sense when a firm is opening a manufacturing facility with heavy machinery and loads of jobs. But famously American manufacturing as a share of employment has declined for decades. What facilities do remain have survived largely through ruthless efficiency. 

That led to a meaningful shift in the economic developer’s tools of public inducements: from jobs to buildings. Look at that Hudson Valley data center project. Job creation credits are often tied to payroll. But data center projects typically get different subsidies, like property and sales-tax abatements, which are tied instead to capital commitments.

Promise to build more, and many governments will save you more, often creating efficiencies in much larger projects. The AI infrastructure buildout may already have become one of the largest capital investments in American history. 

That’s what brought those economic developers in New York’s Hudson Valley to give up so much for something so many of their residents now despise.

Which brings us to a second reason why big-game hunting has died: It’s gotten more expensive. No company has commanded quite the spectacle that Amazon did, but the practice is a classic zero-sum game. Marketing a place’s strengths (which existing residents benefit from) is one thing; outright subsidizing specific firms is something else entirely.

New York is a good example to review, not because its strategy is unusual, but because it’s unusually transparent.

In May the state comptroller reported that New York’s 106 industrial development agencies granted $2.1 billion in tax exemptions across 4,183 projects in 2024. That may seem small in context: State, county and city annual budgets in New York total more than $400 billion. But the figure reveals a larger point: The field chiefly responsible for growing local economies is still mostly buying jobs. 

Further, the comptroller report flatters the cost by setting one year of exemptions against more than 200,000 net jobs gained across every active project. That’s about $5,500 a job, which would be remarkable efficiency. But it’s only one year’s cost, so the lifetime figure is some multiple of it.

Even that average hides the range. Western New York comes in at $4,098 a job. The North Country pays $15,117. Rockland County, the agency that approved that big data center deal, reported $19.2 million in net exemptions against 624 jobs gained, which is about $30,800 apiece.

Nearly 4 in 10 subsidized projects reported no net job gain.

Those worsening outcomes are in part because of the field’s competitive race to the bottom. 

One 2020 analysis found firm-specific incentives encourage shuffling locations, with weak evidence of broader local growth. Britain’s What Works Centre for Local Economic Growth warns that a subsidy moving jobs one county over has created nothing.

Heidi Knoblauch saw the problem from inside. She now runs Homefield, which sells impact-tracking software to economic development organizations and partially underwrites this newsletter. Before that she chaired the industrial development agency in Troy, New York, roughly 130 miles up the Hudson. Her board once granted a developer a tax break to build housing with ground-floor office space, on the promise of a tenant that would bring jobs. Then the developer moved its own offices in, from a mile down the road.

“No new jobs,” Knoblauch said to me, pausing for effect. “No new jobs.”

Buying jobs is getting more expensive, and we’re getting fewer of them too. The 196,067 jobs New York agencies expect to support is the state’s lowest total in a decade, and the comptroller’s own conclusion is that the benefits “have shifted from job creation toward job retention.” We are increasingly paying to keep what we have.

Jonathon Beschen is the managing director of the tech practice at Ben Franklin Technology Partners, a category-creating, state-backed investor whose mandate is counting jobs in five Pennsylvania counties. He sees the same thing from the other end: “We’re doing a lot more deals for the same number of jobs.” That represents what we’re seeing across the American economy: Surging rates of new businesses from a few giant winners and many more tiny slow-growers, while the mid-sized declines.

Then there’s the question of whether any of it is true. The NY comptroller’s office says plainly that it “reviews but does not independently verify this data.” When auditors go look, here is what they find: one county agency did not verify job creation or retention, relying instead on self-reported information and undocumented discussions. Another missed its job goals on 8 of 15 projects and pursued no recapture. Ten of the 106 agencies reported no clawback agreements (to ask for money back from employers who miss their commitments) on their active projects at all.

That brings us to our third reason the math of economic development inducing established firms is breaking: Mounting research suggests it’s too expensive and ineffective to pursue at all. 

Once all costs are tallied, business tax incentives cost on average $436,000 per job, according to a W.E. Upjohn Institute for Employment Research analysis. The cost of support services for homegrown entrepreneurs, from guidance on manufacturing upgrades to specialized training, averages between $78,000 and $155,000 per job created.

Governments spend about $80 billion a year on the expensive option and about $10 billion on the cheap one.

We haven’t just learned that the old model is expensive — turns out it’s also ineffective at many of those foundational goals.

Job creation? Without new firms there is no net job growth at all. Startups add roughly 3 million jobs a year while every other category of firm, taken together, is a net job destroyer, according to Ewing Marion Kauffman Foundation researchers. They argue that job-creation policies aimed at luring larger established employers “will inevitably fail.”

In short, entrepreneurship is the start of a local economy, and yet most local economic leaders spend all their time and money on the most expensive, least effective prospects — all while relying on volunteers, hobbyists and do-gooders to do the entrepreneurship work.

It takes work to ‘kiss 1,000 frogs’

If the strategy has been breaking for this long, why does it persist?

Victor Hwang ran entrepreneurship at the Kauffman Foundation before founding Right to Start, which lobbies statehouses to spend more on new businesses. Hwang also co-hosts with me Builders Live, a podcast on the pro-entrepreneurship movement. The research says recruitment is a net loss for society, he told me, “but people still do it. It’s hard to give up arms unilaterally.”

Years back, I spoke at a conference focused on tax credits — you’ll be shocked to learn that at such dull affairs, I count as borderline exciting. Though more than one spreadsheet was displayed in a conference room, the confab was hosted at a posh Miami hotel. The parties were good, and attendees did business in similar attire and with a familiar lexicon. In short, it’s comfortable and predictable: Good deal or not, a local leader can pledge to forgo tax revenue they might not have ever gotten otherwise and in exchange they and their local elected officials can announce jobs, buildings and activity.

In contrast, “we don’t know the startup lingo,” as one economic development leader said to me, and “with an entrepreneurship strategy, you have to kiss 1,000 frogs to find a prince.”

Courting Amazon execs at a high-end restaurant, networking at a conference party in Miami and repurposing spreadsheets for what you’ve done for years all seem so much nicer than frog kissing. 

IEDC, the trade group, represents the moment. Their conference this month features plenty of conventional, and necessary, professional development — and also an entrepreneurship-focused summit. Likewise, Hwang’s Right to Start is one of a constellation of former Kauffman Foundation–affiliated groups working to transform economic development. They’ve advanced statewide policy. Next month, the Global Entrepreneurship Network will organize its globe-spanning, week-long celebration of the field, and the recently rebranded ESHIP Alliance is convening next month grassroots organizers.

Whether they reflect lasting and overdue change for an overlooked field, or the fruitless agitations of hopeful outsiders is yet to be seen.

Across the country, most economic development is done the way it’s always been done. Yet more communities are adjusting their strategies, which will be the focus of the next in this series.

 Said Hwang: “We’ll need all of us in the room.”

---