What happens when you're forced to spend more than you'll earn

The European Commission has cut growth forecasts for Italy: +0.4% in 2025, +0.8% in 2026. Second-to-last in Europe this year, dead last next year. Meanwhile, Italy's PNRR is proceeding technically well: 337 milestones achieved out of 621, €64 billion already spent, sixth tranche disbursed.

Here's the paradox: you're spending €194 billion to generate 0.4% growth. It's as if someone forced you to invest £100,000 with a guaranteed return of £400. Except you can't refuse, you can't redesign the investment, and you must complete everything by a rigid deadline or lose the money.

The answer to the title question is simple: you optimise for spending, not for earning. And this is precisely what's happening across the EU's largest policy experiment.

Who designed this system

The 2026 deadline wasn't Italy's choice. It's in the EU Regulation establishing the Recovery and Resilience Facility, approved by Parliament and Council in February 2021. It applies to all 27 member states managing €750 billion, the largest stimulus package ever financed in Europe.

The original logic made sense in the 2020-2021 pandemic context. The European Union was raising common debt for the first time in history. It was a historic political gesture requiring rapid demonstration of efficacy. The funds needed to produce fast economic stimulus to counter the COVID recession. Hence the timeline: all funds committed by end of 2023, all payments completed by December 2026.

The problem is that emergency logic, rational for rapid countercyclical stimulus, becomes counterproductive when the stated objective is "structurally transform economies". Real reforms require years of conflictual implementation. Civil justice doesn't reform in five years. Public administration doesn't change organisational culture in a compressed quinquennium.

The European Commission itself has recognised the problem. By end of 2024, only 28% of EU-wide milestones and targets had been completed. Commissioner Fitto declared the deadline "impossible to change" because it would require unanimity of 27 countries plus regulation amendment. A 2028 extension was discussed in June 2025, but the design damage was already done: for four years, every country optimised for "spend by 2026" rather than "actually transform".

Here's the failure architecture: you've designed a system with rapid Keynesian stimulus logic (correct for countering acute recession) and applied it to structural transformation objectives (which require long timescales and political conflict). When you compress structural timelines into emergency urgency, you get rapid spending without transformation.

Why nobody stopped this design

The natural question is: if the design had evident structural problems, why did it pass anyway? The answer reveals how political logics systematically prevail over economic rationality when pressure is maximal.

Many understood it perfectly well. Technocrats who had worked on European Structural Funds 1994-2006 had seen exactly this pattern produce exactly these outcomes: billions spent, minimal transformation. But that expertise had zero political weight at the moment of decision.

In 2020-2021, incompatible logics were operating that all had to coexist in the same instrument:

Northern Europe wanted guarantees. If we're raising common debt for the first time in EU history, we need rigorous conditionality guaranteeing Southern countries won't waste our money. Hence Italy's 528 conditions, the milestones, the verifiable targets. Political narrative internal to net contributor countries required visible strict controls.

Southern Europe wanted immediate liquidity. We need money now to avoid economic collapse post-COVID. Hence the temporal urgency, the necessity of getting funds rapidly. Every month of delay meant deeper recession and growing internal political pressure.

The Commission had to demonstrate political success. This is historic common debt. If it fails, the Union loses credibility for decades and the European project suffers perhaps fatal damage. We must show it works quickly. Hence the necessity for achievable and formally measurable targets allowing success declaration.

National governments had immediate electoral incentives. I must show large numbers to my public opinion quickly, before the next elections. Hence the preference for rapidly visible projects rather than painful reforms producing benefits in ten years but conflict now.

Result: you've designed a system satisfying all these political constraints simultaneously, but which by design cannot optimise for real transformation. The necessary political compromise produced a local optimum on stakeholder satisfaction, not a global optimum on economic outcome.

The invisible principal-agent problem

There's a classic agency problem hidden in the architecture. The Principal (European Union, Northern net contributors) wants real structural transformation justifying the common debt. The Agent (national governments) wants to capture funds with minimal internal political disruption and maximal electoral visibility.

The design had to satisfy both. The technical solution was: formally rigorous conditionality but verifiable primarily on procedures and spending, not on real transformative outcomes. Thus Northern Europe could tell its voters "we've put strict controls" and Southern Europe could tell its own "we've obtained the money without too much interference on national sovereignty".

But verifying "you've approved the civil justice reform law" is radically different from verifying "civil proceedings now actually last half the time". The first is formally and rapidly measurable, perfect for six-monthly tranches. The second requires years of empirical observation and could show failure when politically it's too late to correct.

The system therefore naturally selected for formal conditionality (laws approved, projects started, funds allocated) rather than substantial (measurable economic outcomes, recovered competitiveness, generated growth). This satisfies all actors' political constraints but guarantees real transformation remains optional.

The collective action problem

There's a deeper level. Even those who saw the problem couldn't block it. Publicly admitting "we already know this design doesn't work based on historical precedent" would have been politically impossible whilst trying to sell the Recovery Fund as historic success and European solidarity gesture at the darkest moment.

No single actor (neither country nor institution) could stop the process to say "let's halt and redesign better". Because whoever had blocked would have been accused of sabotaging European solidarity at maximum crisis, of prioritising technicalities over saving economies, of paralysis whilst citizens suffered.

A perverse equilibrium forms where everyone knows the design is suboptimal, but nobody can deviate unilaterally. Proposing substantial modifications would have reopened negotiations requiring months of painful compromises. The risk of collapsing the entire agreement exceeded the benefit of improving the design.

"Perfect is the enemy of done" becomes the mantra. Better an imperfect plan approved in six months than a perfect one discussed for three years whilst economies sank. And that pragmatic 2021 decision locked everyone into suboptimal design for five years, because modifying it ex post requires the same impossible unanimity that generated it.

The political meta-Goodhart

There's a final level. When "approve the Recovery Fund quickly" becomes the dominant political target (because European unity symbolism was needed at COVID's darkest moment), technically optimal design necessarily yields to politically feasible design.

The success metric became "we've approved a historic €750 billion plan in record time" rather than "we've designed a mechanism that will structurally transform European economies". Once that first metric becomes the target, that metric ceases to measure design quality.

You have a case where individual political rationality of each actor (approve quickly to show success, capture funds with minimum internal political risk, declare victory on formal conditionality) produces collective economic irrationality (design that cannot generate real transformation).

It's not incompetence. It's not lack of expertise. It's the structure of political incentives in acute crisis that systematically selects against economically optimal design. When time is the critical variable and political consensus is fragile, mechanism design quality becomes secondary to agreement feasibility.

They knew. But they were trapped.

The invisible mechanism

When you design a funding system with tight deadlines and rigid conditionality, you're designing an incentive set. Every actor in the system (ministry, region, municipality, supplier) will respond rationally to those incentives. The problem is that local rationality produces systemic irrationality.

The PNRR's dominant incentive isn't "transform the country". It's "don't lose the billions". This apparently subtle difference changes everything.

When the primary risk is fund loss, every administrative level goes into defensive mode. The bureaucracy optimises for "satisfy minimum required to avoid losing funding" rather than "maximise transformative impact". It's perfectly rational if you're a director with responsibility for timely completion. It's disastrous if the objective was changing the country.

What game theory calls a perverse Nash equilibrium forms. Each actor, acting rationally according to their own incentives, produces a collectively terrible outcome:

The ministry proposes easily completable projects. Renewing the existing beats risky structural reforms. The region chooses low-risk interventions it already knows how to manage. Administrative innovation requires time, better traditional infrastructure. The municipality prefers standard projects already used elsewhere. Experimenting means risking delays. The supplier offers already-tested solutions. Customising for real transformation lengthens timescales.

Result: mountains of spending, minimal systemic transformation.

When the target becomes the problem

There's an economic principle explaining this mechanism: Goodhart's Law. When a measure becomes a target, it ceases to be a good measure. In the PNRR's case, "spending within deadline" became the target. Therefore that metric ceased to measure what it should: the country's transformation.

The existing bureaucratic machine, unreformed upstream, has an automatic bias towards resource allocation on rapidly spendable but low systemic impact projects. Because changing oneself is painful, risky, slow. Much safer to capture funds channelling towards what one already knows how to do.

This isn't cynicism. It's applied behavioural economics. Loss aversion (avoiding losses) is a more powerful driver than seeking transformative gains. When the incentive setup puts fund loss as primary risk, you've designed a system rewarding administrative conservatism.

The pattern isn't new

I've seen this film twice before. European Structural Funds 1994-2006: billions spent, minimal transformation of target regions. Obama Stimulus Package 2009: rapidly allocated, limited structural impact. All "use it or lose it" funding schemes produce the same outcome.

The problem isn't the quantity of money. It's the temporal architecture and incentives. When you compress implementation times and put spending velocity as primary conditionality, you've guaranteed the system will optimise for that variable at the expense of all others.

Real reforms (those unlocking growth) require conflict, negotiation, overcome resistance. They require time. A system designed for spending velocity automatically selects against painful but necessary reforms.

How it should have been designed

The alternative existed. Instead of "how much you spend how quickly", the system should have been structured on "which systemic blocks you've removed".

Verifiable transformative output metrics: average time for building permits reduced 50%, administrative litigation halved, days to open business below European threshold. Payment linked not to spending but to achieving these structural unblocking indicators.

Extra-bureaucratic implementation agencies: teams with temporary authority to bypass existing administrative pathology. Not reforming bureaucracy from within whilst it must implement. Creating parallel structures with skin in the game on transformative results.

Asymmetric risk incentives: double premium for high transformative risk projects successfully completed, minimal penalty for well-documented failures. Thus you select for calculated risk-taking rather than defensive conservatism.

The problem is this design requires accepting possible initial delays. It requires explaining to Brussels that real transformation doesn't follow linear timelines. It requires political courage to say "I prefer spending less but better" when the political incentive is "show large numbers quickly".

Where we're heading

Based on analogous historical patterns, the probable outcome is: by mid-2026 we'll have spent almost all funds, reached nearly all formal targets, celebrated technical implementation success. And we'll have marginal structural growth, because funds will have gone predominantly to renewing existing infrastructure, digitalising procedures that remain complex, financing projects that could have started anyway.

Real transformation (functioning justice, non-blocking bureaucracy, system attracting investments through intrinsic competitiveness) will be postponed. Again.

It's not a question of people's competence. It's a question of incentive architecture. You've designed a system rewarding behaviours producing the opposite stated outcome. Then you're surprised when it functions exactly as designed.

The lesson for anyone having to design large funding programmes is simple: before asking "how much money", ask "which behaviours does this incentive system naturally select". Because you'll get exactly those behaviours. Always.

When you're forced to spend more than you'll earn, the rational choice is to stop optimising for earnings. The EU's Recovery and Resilience Facility has demonstrated this logic applies even to entire national economies.