The Second Paradox of Economic Development

Public criticism of economic development often assumes that practitioners believe every corporate claim placed in front of them. Florida offers a more credible account. They work in an authorizing environment shaped by elected officials, neighboring jurisdictions, confidentiality rules, announcement politics and a narrow set of performance measures.

The empirical case against the prevailing model is substantial. Cailin Slattery and Owen Zidar found some evidence that a major project increased employment within the industry receiving it, with little strong evidence of broader state or local growth. The average discretionary subsidy in their sample was US$178 million for 1,500 promised jobs. Timothy Bartik reviewed 34 estimates from 30 studies and concluded that typical incentives probably change a location, expansion or retention decision in no more than one quarter of cases. Among studies with no obvious bias, the median estimate was 3.4 percent. On Bartik's reading, at least 75 percent of recipient firms would probably have made a similar decision without the support.

The political return follows a different clock. A deal produces an announcement now. Its counterfactual can rarely be observed. Opportunity costs surface across several budgets and many years. A rigorous evaluation may arrive after the officials responsible have changed jobs.

Florida is therefore right to reject the idea that research can carry the reform on its own. He also gives the profession a role beyond administering a system designed by others. Standards, shared evidence and professional pressure can change which behavior is considered competent. That is a more serious proposition than asking practitioners to become better salespeople or more disciplined analysts.

The compact still needs to separate several transactions that current debate places under one word.

The classic incentive contest concerned a mobile firm choosing among broadly substitutable locations. A government reduced the firm's cost and hoped to alter the choice. Florida's history of industrial location theory explains how that model took hold and why it became disproportionate.

Today's largest packages often sit inside national industrial strategies. Semiconductors, AI infrastructure, critical minerals, clean energy and defense-linked technologies carry goals that extend beyond local employment. Governments are buying production capacity, supply-chain resilience, technological learning and strategic options. Some projects also create genuine local benefits. A strategic purpose may justify public involvement, though the purpose alone says nothing about a reasonable price. Evaluation has to account for both.

UN Trade and Development reported that five strategic sectors accounted for 44 percent of announced global greenfield investment value in 2025, up from 16 percent in 2020. Data centers alone exceeded US$270 billion. As Doyen argued in Ten Rewirings, the entrance fee for these projects has risen far beyond the scale at which most jurisdictions can compete on price.

Even the European Union, which Florida rightly identifies as the strongest existing model for controlling state aid, has adapted its rules to the new industrial contest. The European Commission's Clean Industrial Deal State Aid Framework makes it easier for member states to support clean energy, industrial decarbonization and clean-technology manufacturing through 2030. The EU retains common rules and review. It has not withdrawn from strategic subsidy competition.

A workable compact should distinguish four uses of public support.

Relocation support pays a company to move existing activity across a political boundary without adding meaningful capacity to the wider labor market. This is the cleanest candidate for prohibition.

Expansion support applies when a firm will add activity while moving or enlarging an operation. Public support should attach only to the verified increment. Existing jobs and investment need a baseline, and the agreement should survive creative relabelling.

Strategic-capability support pursues a public objective such as domestic semiconductor capacity, energy security or a critical-minerals supply chain. The level of government that owns that objective should bear the main cost and state it openly. A local government should not finance a national-security premium from its property-tax base.

Shared-asset investment builds infrastructure, workforce capacity, serviced land or research facilities that remain useful across companies and over time. Such spending can support a specific project, but its public value should not disappear if that company leaves.

The distinctions matter because a blanket cap can block useful investment while leaving room for a government to rename a subsidy as infrastructure. A compact needs common accounting and a clear theory of what each payment is purchasing. Otherwise the old contest continues through different budget lines.

Florida argues that the failed HQ2 non-aggression pact suffered from bad timing. Mayors and governors were already inside a highly visible competition, making withdrawal politically costly. He proposes starting with willing jurisdictions and building a coalition before the next contest arrives.

The recent history of Kansas and Missouri both supports and complicates that strategy.

In 2019, the two states agreed to stop subsidizing the movement of existing jobs across the state line that divides the Kansas City metropolitan area. Kansas used an executive order. Missouri enacted legislation. Regional economists believed the truce worked for a period.

It began to come apart when Kansas and Missouri competed over the Chiefs and Royals. Missouri's statutory limits expired. In September 2025, the Kansas City Council unanimously removed its own restrictions. Then came a more ordinary corporate project. In May 2026, Lockton broke ground on a US$765 million headquarters in Leawood, Kansas, supported by about US$125 million in tax incentives. The company had 1,500 local employees and was expected to add roughly 500. Kansas treated the expansion as sufficient to keep the deal outside the truce.

The truce had been assembled in quieter conditions, just as Florida recommends. It lasted for six years and showed that cooperation was possible. It also failed under the pressure it was created to contain. A trophy asset opened a political exception, a time limit expired, and an expansion clause provided room for a large cross-border move.

Kansas City offers useful design evidence for Florida's compact. Voluntary restraint needs a durable legal form, a common definition of net-new activity and an all-in measure of public support. Land discounts, utility concessions, tax increment financing, grants, loan guarantees and dedicated infrastructure belong in the same account. Special categories such as sports facilities cannot sit outside the rule simply because their politics are unusually intense.

The coalition also needs ideological breadth from the start. Florida suggests that progressive states and large cities could move first. Fiscal exposure to incentive competition crosses party lines, as does the attraction of a ribbon cutting. A compact associated with one political camp would give the other an easy reason to remain outside it. State budget offices, metropolitan business groups and distressed communities may provide a more durable base than partisan alignment.

Professional standards can support such an agreement. They cannot bind an elected government facing a flagship decision. Federal policy will eventually matter, whether through legislation, caps or eligibility rules attached to discretionary funds. The compact's early members can build the proof and the accounting system that make federal action usable.

Florida proposes redirecting public resources toward four broad areas: workforce and talent, innovation ecosystems, manufacturing extension, and community-enhancing development. Each has a much stronger claim on public money than an indiscriminate tax holiday. His insistence that housing, social fabric and inclusion affect competitiveness is especially valuable. An unaffordable high-productivity city eventually obstructs its own growth.

Here Florida is more categorical than the evidence permits. The label “place-based” identifies where a policy acts; it does not establish that the policy will work. Demand-linked training and manufacturing extension have relatively direct routes to employment and firm performance. Cluster policy depends heavily on a region's existing economic base, related capabilities and connections to outside markets. The OECD's recent framework makes local fit and an institution's capacity to coordinate conditions for success. A compact should avoid replacing faith in firm attraction with faith in any program described as an ecosystem.

Retention varies across every one of these categories.

A region can train people for expanding industries and watch them leave, or find that the jobs are filled mainly through recruitment from elsewhere. Public research can generate intellectual property that is acquired and commercialized outside the region. Manufacturing extension can raise a firm's productivity while the resulting value appears in margins or a sale to an external owner, with little change in local wages or supplier depth. New transit, amenities and housing can raise land values that accrue to incumbent owners while lower-income residents face higher costs.

The original investment may still carry substantial public value. Productive capacity and shared prosperity remain separate results, joined only by the way a place structures ownership, access and distribution.

That is the problem examined in Doyen's The Capture Gap. The report starts from the widening distance between the value a place helps make possible and the value it keeps. Its project ledger records value retained, public cost, distribution and duration. It then asks what claims the public has secured: fiscal, ownership, contractual and institutional.

Consider the new generation of data-center projects. A C$13 billion complex promising 300 permanent jobs looks indefensible when permanent employment is the only benefit. That ratio can also obscure genuine value in tax revenue, grid investment, construction, research demand or a larger computing cluster. The ledger forces both sides of the case into view. It records the public contribution of land, power, water, infrastructure and tax treatment. It identifies where the returns go, how long local benefits last, what obligations survive a change of ownership, and who carries the environmental and infrastructure costs.

The resulting negotiation is less vulnerable to vague claims about an ecosystem. A jurisdiction might seek a predictable fiscal stream, an equity or land interest, enforceable local-procurement terms, supplier-development programs, shared research infrastructure, grid upgrades that serve other users, or training tied to actual positions. The right claim depends on the project. The principle is that a public contribution should leave the place with something more durable than an announcement.

This also changes who needs to be in the room. A large package may draw value from a municipal landowner, a regulated utility, a state tax authority, a workforce fund and a federal program at the same time. Each institution can see its own contribution as modest while the company negotiates against the total. The project ledger creates one public balance sheet before the separate concessions become a package no single body intended to offer.

The hinge in Florida's account appears in his description of innovation ecosystems that keep gains local. Local retention cannot be assumed from local activity. It depends on ownership, contracts, institutions and tax design.

The second paradox follows. A jurisdiction can stop wasting money on corporate poaching, invest in every input that sound regional economics recommends, and still build an economy whose most valuable claims belong elsewhere.

Florida's wider definition of economic development presents another practical question. Affordable housing, early-childhood support, education, infrastructure and social cohesion all shape economic performance. That does not mean an economic development agency should administer all of them.

Investment promotion agencies already carry expanding mandates. An OECD survey found that the average agency handled 5.4 mandates in 2025, up from 4.8 in 2017. In real terms, funding dedicated to their core investment mandate fell by 22 percent between 2019 and 2025. Only one in four agencies believed it had enough resources to deliver fully.

Florida's community argument is economically persuasive and institutionally unfinished. An agency can identify housing as the binding constraint on an investment corridor, quantify the cost, bring employers into the planning process and make housing delivery part of the region's economic strategy. Planning authorities, housing agencies, infrastructure bodies and community organizations retain the expertise and authority to deliver it.

This boundary protects the breadth of Florida's insight without turning economic development into a label for every desirable public policy. It also gives the compact a realistic division of labor. Economic developers should help the system recognize how these investments affect productive capacity. They should not inherit each operating mandate that touches growth.

Near the end of the article, Florida asks how research could enter practice more reliably. His analogy is the teaching hospital, where evidence, professional training, live cases and measured outcomes sit inside the same institution.

Economic development has no comparable learning system. Deal files are fragmented across governments and protected by commercial confidentiality. Public announcements remove the assumptions and concessions that would make a case useful. Failed bids disappear. Evaluations examine a program years later, often without the information needed to reconstruct what officials believed when they approved it. Practitioners carry hard-won knowledge from one negotiation to the next, but the field retains little of it when they leave.

The field also struggles to make practitioners feel that they belong to one profession. Its associations, conferences and credentials tend to follow established institutional boundaries. An investment promoter in Singapore, a site selector in Atlanta, an export adviser in Ottawa and a regional strategist in Manchester can face the same questions about public risk and private value. Their common professional conversation remains thin. Job titles inherited from separate public systems divide people who increasingly use the same forms of judgment. Their knowledge travels through personal networks and often leaves with them.

A teaching hospital teaches people what membership in a profession requires. Early-career practitioners work beside experienced ones while real decisions are opened to review. Outcomes feed back into the methods used on the next case. A practitioner develops standing through the judgment they show and the knowledge they return to the field. That standing follows them when they change institutions or cross a border, giving the profession a memory beyond one team's tenure.

For Florida's compact, this professional bond matters. The pressure to approve a weak deal reaches officials who know the evidence. Professional standards gain force when they shape who is trusted, hired and invited into consequential work. Full-cost accounting, restraint, honest counterfactuals and post-deal disclosure become marks of competence. The source of authority survives elections and job changes.

This is the larger case for Doyen. Doyen uses economic acceleration to name the field that links local development, investment attraction, trade promotion, commercial diplomacy, site selection and the advisers around them. These roles share an object: shaping the conditions under which productive activity forms, moves and stays. Together they constitute a profession whose institutional life has lagged behind the scope of its work. Doyen is being built as its modern professional home.

This community is international, and its careers cross institutional lines. Doyen is being built to give it a continuous professional life. A report enters an argument that practitioners test against active cases; the exceptions feed into later work. During a workshop, the reasoning behind a live decision can be exposed while there is time to improve it. Digital case records preserve knowledge that often disappears when someone leaves, and AI can surface relevant cases and peers when a similar problem appears. Someone who enters through export promotion can see how their experience carries into site selection, investment strategy, commercial diplomacy or agency leadership. The platform makes this wider profession visible to itself. As those careers unfold, contribution builds standing alongside title and seniority.

That identity changes the incentives around a difficult deal. Officials usually answer to the institution and political leadership immediately in front of them. Doyen gives them a professional reference group that travels across agencies, firms and borders. Choices made under pressure remain available for peer review. Shared methods move through the field as people change roles. The compact then rests partly on a professional culture whose members expect sound judgment from one another.

The operating core of this system is a common case record. Before approval, that record would describe the project's objective, realistic alternatives, estimated counterfactual, total public contribution, expected local claims, distribution and principal risks. The same file would follow the project through construction and operation. Employment, wages, procurement, tax receipts, ownership changes, automation, energy use and public obligations would be updated against the original case.

Confidential details could remain protected while a negotiation is active. Standardized and anonymized cases could still enter a shared evidence base. After an agreed period, the public should see enough to judge performance. Losing bids belong in the system as well, since they reveal how firms rank locations and whether the eventual winner paid for an advantage it already possessed.

Case review would bring researchers and practitioners together before doctrine hardens. Researchers would gain better data and a clearer view of the actual decision. Practitioners would receive usable findings when another project presents the same problem. IEDC and other professional bodies could organize standards across the network, while journals establish sound methods for comparison. Doyen occupies the continuous professional layer where evidence meets live judgment, cases enter a shared memory, people develop their careers and contribution earns recognition. Florida's teaching hospital gives that layer an institutional form.

A compact suited to the current investment cycle would make four commitments.

First, members would classify the purpose of public support before negotiating the amount. Relocation, incremental expansion, strategic capacity and shared assets would face different rules.

Second, they would use a common ledger for the full public contribution. The calculation would include foregone revenue, financing risk, land, utilities and dedicated infrastructure. It would also identify costs carried by municipalities or public bodies outside the agency announcing the deal.

Third, support above a modest threshold would require enforceable public claims and a stated account of distribution and duration. Job promises would remain relevant. They would sit beside fiscal returns, local procurement, ownership, infrastructure access, workforce pathways and the obligations that survive a sale.

Fourth, each major project would contribute a standardized record to the teaching system. Results would be reviewed against the original counterfactual, not against the press release.

These commitments would give practitioners political cover. A governor could tell a company that the jurisdiction cannot conceal support in a utility contract, count transferred positions as new employment or approve a large package without a durable claim. Competing jurisdictions inside the compact would face the same constraint.

The compact would also become useful beyond the United States. Incentive competition now runs through national industrial strategies, free zones, sovereign investment programs and the race for AI infrastructure. The governing scales differ. The need for a common record, enforceable claims and institutional learning does not.

Florida's paper deserves attention because it places economic developers inside the political system that shapes their choices. It gives the profession standing in the reform and asks researchers to share responsibility for implementation.

Its proposed compact can address the first paradox, in which every participant understands the incentive game and still finds it rational to play. The second requires a further step. Places need to know what they will retain from the economic activity their people, institutions and public balance sheets help produce.

If the next Amazon contest arrives before that machinery exists, cities will again understand exactly what is happening. This time, they should have something sturdier than a nod of agreement.