Middle management is often described as the "bottleneck" of innovation. It's a convenient expression, but a reductive one — and above all an ungenerous one. In my experience, the intermediate layer of large organizations is rarely made of incompetent people or people acting in bad faith (though the possibility can never be ruled out). It resists because it sits at the crossroads between those who decide and those who execute, constantly squeezed by constraints, pressures and contradictions. And its resistance, seen up close, is almost always perfectly rational.
A rational resistance
The middle manager is the transmission belt between strategic governance and the operational base — and also the shock absorber between forces that, without him, would collide catastrophically. He operates inside a web of internal alliances, assigned targets and unwritten constraints, in an environment where relationships matter as much as roles and where the stability of one's own perimeter is often the only identity shield available. He knows perfectly well that a cross-cutting project can break power balances, generate tension between areas, add new tasks without removing old ones, and expose him to new responsibilities.
Add to this the decisive structural factor: the objectives assigned to middle management rarely coincide with those of an innovation project. Often they clash. The project wants to improve the user experience and experiment; the head of a unit is evaluated on deadlines, regulatory compliance, internal efficiency. The perceived risk is losing real points to chase an abstract benefit. And even the manager who privately shares the vision cannot always afford to back it: the culture of keeping your head down is strong, and for good reason — sticking your neck out is risky.
Where hostility, resistance and complexity nest
The O.R.C.A. model gathers the four recurring forces that obstruct innovators. Three of them — hostility, resistance, complexity — find their natural habitat in the intermediate layer:
Hostility. Not always declared, but very much there. The innovation project is perceived as a threat to one's role, one's influence, the balances built over time. Even when the goal is agreeable, the fact that it wasn't decided "from within" can generate opposition that is silent but active. It's territorial defence, not a judgement on the merits.
Resistance. Here it takes its most dangerous form: the passive one. Nobody opposes you head-on; everything slows down. Things get postponed, further analysis gets requested, procedural objections get raised. It's the grey zone of "I'm not saying no, but I'm not doing yes" — extremely hard to expose, because formally it looks like prudence. In reality it's organizational inertia dressed up as caution.
Complexity. Middle management knows the organizational labyrinth intimately — and often uses it as a shield. "Not my perimeter", "we have no cover", "we'd need an official circular". Complexity becomes the perfect alibi for standing still. And as long as the project hangs between vague signals from above and scarce enthusiasm from below, you wander the labyrinth with no destination.
The pincer
Against this kind of blockage I know of only one tool that I've seen work reliably. I call it the pincer (in Italian, la morsa — the vice): a visual metaphor that is perhaps a little brutal, but also a very concrete operational concept. It consists of applying convergent, simultaneous pressure from two directions: from above, top management legitimizing, protecting, authorizing; from below, the operational base participating, recognizing itself in the project, contributing. In the middle, squeezed between the two pushes, sits middle management — which at that point can no longer stand still without exposing itself, i.e. without paying exactly the cost its culture tells it to avoid.
The pincer is not there to punish. It's there to unblock — which, given what a vice usually does for a living, is decidedly counterintuitive. Its purpose is not to change middle management's mind: it is to create a context in which standing still is no longer a viable option. In one sentence: make saying yes easier than saying no.
Of course, if the project's objectives coincide with those of the key players, no persuasion is needed: support arises on its own. That's the ideal condition, and occasionally it even happens. In most cases, though, spontaneous alignment isn't there, and the calibrated push is required. How to build the two arms — sponsorship from above and buy-in from below — is the subject of the three tools of public-sector innovators.
The pincer in practice: the Sirio case
In the Sirio project, INPS's design system, resistance from the intermediate layer was never frontal. Some technical managers immediately grasped the project's value and moved to support it. But for most of them, adopting the design system simply wasn't a priority: services were being built and put online anyway, formally complete. By the logic of compliance, the goal was already achieved — and Sirio risked looking like an unrequested "extra": interesting, maybe useful, certainly "nice". But not urgent. What followed was silent disengagement: delays, empty chairs at meetings, shared guidelines quietly ignored.
The double push was designed, not hoped for. Support from the top prevented the project from being downsized or ignored; active involvement of the base generated informal pressure and operational consensus. And in between:
Some doors opened, not because they were knocked down, but because keeping them shut had become difficult.
The concrete instrument was the cycle of recurring design reviews with the extended group: periodic sessions presenting the work of previous cycles and gathering feedback. Among the participants were the people from internal teams involved operationally — often belonging to the same areas as those sitting at the intermediate level of the hierarchy. So middle management watched the project advance "through its own people", in a setting that neither excluded it nor overexposed it. Involvement was calibrated: formally informed, listened to when necessary, never forced to take a position or contribute at every phase.
This balance produced simultaneous advantages: nobody felt bypassed, because the reviews were part of the process; nobody was forced to pick a side, because the work was presented as a continuous flow; and everybody was put in a position to feel part of the project, even without a direct operational role. A relational architecture that worked precisely because it didn't ask too much, and it excluded no one.
In short
The pincer, despite its menacing name, was never an aggressive strategy. It's a design configuration that keeps a project from stalling where things usually slow down: in the folds of the organization, where nobody is against you, but nobody is truly for you. Middle management should be neither demonized nor bypassed: it should be placed in conditions where supporting the project — or at least not obstructing it — becomes the most convenient path. With the right arguments, the right pressure, at the right time.