State capacity is the gap between intent and delivery

About this conversation

This piece draws on an interview with Lauren Lombardo, Policy Director at the Recoding America Fund—a philanthropic initiative focused on rebuilding government’s ability to deliver.

Lauren has led government modernisation efforts across federal, state, and local institutions. Most recently, Lauren served as Deputy Policy Director for the House Committee on Oversight and Government Reform, helping drive the Committee’s state capacity legislative agenda. Prior to that, Lauren worked in the U.S. Senate as a technology policy fellow for Senator Ben Sasse, advised state and local governments on digital transformation, and was a Senior Data Scientist at Nielsen.

The Recoding America Fund describes its mission as organising the field around a practical state capacity agenda: ensuring government can hire and keep the right people, reduce procedural bloat, develop purpose-fit digital and data infrastructure, and leverage test-and-learn frameworks—paired with reforms to civil service systems, digital investment and oversight, and procedures, and a renewed operating model between the legislative and executive branches.

The state capacity problem is a delivery problem

In the abstract, “state capacity” can sound like a big, distant concept: institutions, funding, rules, bureaucracy. In practice, it shows up in a simpler, more human question: can government reliably do the thing it promised people it would do?

When you’ve worked close to lawmaking and oversight, you see quickly that delivery is not a footnote to policy. It is where policy becomes real or quietly fails.

One of the most formative experiences in shaping this view is seeing the disconnect between what Congress intends when it passes a law and what happens when agencies try to implement it. Policy staff can write with limited understanding of what execution truly demands. Agencies, in turn, often have few effective mechanisms to tell Congress when something is not working, or when the statute itself is the problem.

That gap between intent and delivery is not a minor technicality. It is a core source of capacity failure.

Anchor capacity in public value, not internal process

Talk about “capacity” easily drifts into internal discussions: headcount, procurement, compliance milestones, governance ceremonies. Those inputs matter, but they are not the point.

Capacity is best measured in outcomes that people can feel:

• Can people access the benefits they are owed?
• Can government deliver the infrastructure project it promised?
• Can a service be improved without a multi‑year fight through process layers?

Too often, government cannot deliver—not because of politics or funding alone, but because of upstream constraints in workforce, systems, tools, and accountability.

And there’s an important communications challenge here: connecting highly visible failures to the underlying governmentwide dysfunction is what helps reformers stay focused on what matters and why.

Where reforms fail: good ideas that never meet implementation reality

The most common failure mode is predictable. Good ideas are developed centrally—either in Congress or in policy agencies—without real consultation with implementation teams.

The reform then lands on an agency that does not have:

• the people,
• the technology,
• or the institutional support

to implement it.

Worse, nobody is clearly responsible for ensuring the reform actually works end-to-end. Accountability dissolves across organisational boundaries, and what remains is compliance: evidence that steps were followed.

The broken feedback loop between policymakers and implementers

The feedback loop between Congress and agencies is overwhelmingly oriented around process compliance, not outcomes.

Information flows back to policymakers about how something was done, rather than whether it worked.

That produces a predictable pattern:

• Oversight becomes theatre: heavy on reporting, light on learning.
• Policymaking becomes prescriptive: Congress writes detailed requirements because it does not trust agencies to use discretion.
• Implementation becomes rigid: agencies interpret statutes narrowly because they fear getting it wrong.

Importantly, both sides are acting rationally given the information they have. The problem is that neither side has enough of the other’s information.

A healthier system would change the questions Congress asks and the actions it rewards. Instead of “was the process followed?”, oversight should be designed to ask: “did this work for people?”

Procedural burden: when safeguards calcify into paralysis

Many procedures exist for good reasons: to prevent corruption, ensure public input, and protect against waste.

But over time, process layers accumulate and calcify. In aggregate, processes that started as good ideas can end up paralysing the people trying to do the work.

The core judgement is a trade-off:

• What value does the process produce?
• What value does it delay, reduce, or prevent?

Some processes do little to increase accountability or improve services. These should be reformed.

Others may be necessary, but could be executed differently—with modern technology or redesigned oversight mechanisms—to reduce burden without reducing trust.

Evidence-led government, even when politics and timelines are real

When timelines are tight and decisions are political, “evidence-led” cannot mean waiting for perfect data. It has to mean something more practical: information flows that help people make better decisions in the messy middle.

A high‑leverage habit is building regular, structured feedback loops between policymakers and implementers—loops that reveal:

• whether a policy is working as intended,
• what implementation constraints are emerging,
• and what adjustments are needed to deliver outcomes.

Done well, decision quality improves exponentially—not because politics disappear, but because decision-makers can see reality sooner.

Capability building: avoid short-term wins that relieve pressure for structural reform

There is important work in government that delivers short-term wins: improving individual services, fixing hiring processes, shipping visible improvements.

That work matters. But there is a real risk: those wins can relieve the pressure that might otherwise force structural change.

The deeper goal is not to make the current system marginally more survivable. It is to push the operating model so that good delivery becomes the default, not the exception.

Trust and guardrails: optimise for real accountability

Government builds trust by delivering well, and erodes trust by failing to deliver—regardless of how many compliance processes it performs.

Under many current systems, we optimise for the appearance of accountability rather than actual accountability. The time spent producing reports nobody reads, running public comment processes that do not meaningfully inform decisions, and conducting oversight focused on procedural compliance can drain the very capacity needed to deliver outcomes.

New technologies offer real promise here if they are used to increase transparency and accountability without adding burden on implementation teams, and to improve privacy without stalling shared services and cross‑agency data sharing.

Four interconnected competencies for delivery capacity

One useful frame for capacity reform is four interlocking competencies:

• Workforce
• Procedural reform
• Digital infrastructure
• Test-and-learn frameworks

The point is that you cannot treat these as independent lanes.

You cannot modernise digital infrastructure without addressing the hiring challenges that determine whether government can build and maintain it.

You cannot reform hiring without rethinking the accountability and oversight systems that shape how agencies manage the people they hire.

And in practice, prioritisation is often driven by a blend of what is politically feasible, what has the greatest chance of success, and where opportunities are presenting themselves.

A decade view: a new operating model, not incremental adaptation

Looking ahead, high-performing governments will be defined less by isolated “modernisation projects” and more by a fundamentally different operating model—one that:

• hires the right people,
• enables them to do the right work,
• uses the right tools,
• and holds teams accountable to outcomes.

The current system cannot be simply adapted to the challenges of today.

So the question is not how to make it perform a little better. It is how to build a government that can operate differently—and respond to disruption as a matter of course.

The window for this kind of reform is open now. It will not stay open indefinitely.