L2 aims to flip this equation:

• success = “real-world outcomes for citizens and society”,

• mindset = “if I pursue agreed outcomes transparently and use evidence, the system supports me even if some risks materialise”.

2. Re-designing the basic unit: the “programme”

We need to reinterpret what a “programme” is. Instead of being mainly a legal-financial container (budget chapter + eligibility criteria), a programme becomes a structured attempt to change reality.

Each significant programme should be required to have:

1. Problem and context analysis

   • What concrete problem is being solved? For whom? With what root causes?

   • How does this programme fit into the relevant mission(s)?

2. Theory of change

   • If we do X (inputs, activities), we expect Y (outputs, intermediate outcomes), which should lead to Z (final outcomes).

   • What assumptions must hold? What could go wrong?

3. Results framework

   • Clear, manageable set of indicators:

     • output indicators (what is delivered),

     • outcome indicators (what changes in reality),

     • possibly impact indicators for longer term.

4. Evaluation plan

   • What will be evaluated (process, outcomes, cost-effectiveness, distributional effects)?

   • When and by whom (internal vs independent)?

   • How will results be used (explicit decision rules for redesign, scaling, or termination)?

5. Risk and learning plan

   • Key risks (legal, financial, reputational, operational),

   • mitigation measures,

   • explicit learning questions: what do we most need to learn from this programme?

This structure should be embedded in guidance from:

• the centre of government (L1)

• and the Ministry of Finance / control bodies.

3. Changing the incentives: how people are judged and protected

The culture shifts only if what is rewarded and punished changes.

Performance assessments of senior managers

• Introduce outcome-related criteria:

  • progress on programme results,

  • quality of evidence and evaluation usage,

  • ability to identify and address structural problems.

• Reduce over-weighting of:

  • pure compliance with process,

  • superficial “activity levels” (number of projects, calls, meetings).

Recognition and career progression

• Create recognition schemes for teams that:

  • redesign programmes based on evidence,

  • transparently admit failure and implement stronger alternatives,

  • pioneer new models (e.g. outcome-based funding, integrated services).

• Link such behaviour to:

  • accelerated promotions,

  • access to leadership training,

  • public recognition (within the administration and, when appropriate, externally).

Protection for responsible risk-takers

• Develop “safe harbour” rules:

  • if a programme has a clear theory of change, risk assessment, legal review, and an approved experimental design, then failures within that design are not treated as personal misconduct, but as learning events.

• Clarify in law and internal directives:

  • the distinction between negligence or abuse vs. good-faith, documented experimentation within approved parameters.

This requires close coordination with:

• internal audit units,

• the Supreme Audit Office,

• inspection bodies,

• and even prosecutors, so that interpretations are aligned.

4. Aligning oversight with outcomes and learning

Oversight bodies can either lock in formalism or enable learning.

Recalibrating audit and control

• In cooperation with NKÚ and internal auditors, shift emphasis towards:

  • examining whether programmes have clear objectives and plausible theories of change,

  • reviewing whether evaluation findings are acted upon,

  • identifying systemic design flaws, not just individual irregularities.

• Introduce “learning-oriented audits”:

  • targeted reviews that look at a cluster of similar programmes across ministries or regions,

  • focus on what works and what doesn’t,

  • produce recommendations for redesign, not just blame and sanctions.

Integrating evaluation and audit cycles

• Create a joint cycle:

  • programme → mid-term evaluation → corrective measures → late-term evaluation → audit of both performance and use of evidence.

• Require audited entities to:

  • show not only compliance,

  • but also how they used analysis and evaluation to improve.

5. Embedding experimentation in law and practice

Innovation cannot rely only on brave individuals; it needs institutionalised experimentation.

Policy labs and pilots

• Each major ministry should have access to:

  • a policy lab: small team with skills in design, behavioural insights, prototyping, experimentation.

• For selected problems, design:

  • pilot projects in a limited number of regions/municipalities,

  • randomised or quasi-experimental evaluations where feasible,

  • a clear plan for scaling successful models and abandoning failures.

Experimental clauses and derogations

• Introduce legal provisions for temporary experimental regimes:

  • allow deviations from some rules or procedures in a controlled, time-limited way,

  • require ex-ante approval, monitoring, evaluation, and ex-post decision on whether to generalise or terminate the experimental arrangement.

Learning infrastructure

• Create a central repository of evaluations and experimental results:

  • accessible summaries,

  • searchable by topic, ministry, region, target group.

• Organise cross-administration learning events:

  • regular conferences and workshops where officials present lessons from reforms and experiments,

  • cross-pollination of ideas between sectors (e.g. health learning from education, social policy from labour market programmes).

6. Daily work practices: micro-level change

Culture isn’t only changed by law; it changes through daily routines.

Some practical shifts:

• Decision templates:

  • require each important decision memo to include:

    • what evidence was considered,

    • what alternatives were evaluated,

    • what risks are recognised,

    • how outcomes will be tracked.

• Meetings and reporting:

  • replace part of the classic “round of status updates” with structured discussions on:

    • what we learned since the last meeting,

    • what’s not working and why,

    • what experiments we are trying next.

• Language and framing:

  • leaders at all levels consistently talk about:

    • solving real problems for citizens,

    • evidence and learning,

    • honest acknowledgement of mistakes.

  • this gradually shifts what people perceive as “normal” and “professional”.

7. Implementation roadmap for L2

This reform is more cultural, but still needs a sequence.

Step 1 – Standards and pilots

• The centre of government (L1) issues:

  • new guidance on programme design (theory of change, results framework, evaluation plans),

  • with pilots in a few ministries and selected regions/municipalities.

• Design training programmes for mid- and senior-level officials on outcomes, evaluation, and experimentation.

Step 2 – Oversight alignment

• Convene a joint working group with NKÚ, internal auditors, MoF, justice ministry and prosecutors to:

  • clarify safe-harbour rules,

  • adjust methodologies to incorporate effectiveness and learning.

• Start with a few learning-oriented audits as demonstration projects.

Step 3 – Legal and HR changes

• Introduce legal changes where needed:

  • experimental clauses,

  • explicit references to evaluation in key laws,

  • adjustments to civil service evaluation criteria.

• Update HR frameworks to:

  • include outcome and learning metrics in performance reviews,

  • create recognition schemes for innovative and evidence-driven work.

Step 4 – Scaling and institutionalisation

• Gradually make the new programme design and evaluation standards mandatory for all major programmes.

• Embed outcome and evaluation requirements in:

  • budget process,

  • EU fund programming,

  • strategic planning cycles.

• Build a public-facing evaluation portal to make results visible and create external pressure for learning.

8. Risks and counter-risks

Risk: Everyone adopts the new language (theory of change, outcomes) but not the substance – “old formalism in new words”.
Counter-risk:

• require at least some independent quality checks of programme logic (by the centre or external experts);

• use audits and evaluations to expose purely cosmetic compliance;

• publicly highlight examples where real learning leads to redesign.

Risk: Experimentation produces some visible failures, causing political backlash.
Counter-risk:

• prepare the public narrative upfront: experimentation is how we avoid huge, nationwide failures;

• be transparent about selection of pilots and protection mechanisms;

• make sure some early pilots produce visible, positive results to build credibility.

Risk: Officials fear that “safe harbour” won’t protect them if prosecutors or media attack.
Counter-risk:

• align legal and oversight interpretations as much as possible;

• back brave officials publicly when they have followed agreed procedures;

• ensure top political leadership consistently supports “responsible risk” rhetoric.

L3 – Making the civil service a high-performing profession

1. What problem L3 is really solving

Right now, the Czech civil service is:

• reasonably capable for routine administration,

• but systematically underpowered for complex transformation.

Root problems:

• Politicisation and instability at the top: senior managers are too dependent on political cycles.

• Rigid, flat pay and career structures that don’t reward performance or scarce skills.

• HR as admin, not as strategic workforce management.

• Weak capability in the very areas needed for transformation:

  • complex policy design,

  • data and analytics,

  • digital & AI,

  • project, portfolio and change management.

L3 aims to turn the civil service into a mission-driven, high-performing profession that can actually power the missions defined under L1.

2. Strategic principles for civil service reform

Four guiding ideas:

1. Insulation from partisan politics, not from reality
   Senior officials should be protected from arbitrary political interference, but not from performance expectations.

2. Differentiated, competitive rewards
   The state cannot compete with the private sector on everything, but it must be competitive on purpose + learning + reasonable pay, especially for critical roles.

3. Mobility and shared culture
   People should move within the system (between ministries, regions, agencies) to build a shared professional ethos and circulate know-how.

4. Capability-building focused on missions
   Training and development must be aligned with the concrete missions and reform agendas, not generic one-size-fits-all courses.

3. De-politicising and stabilising top management

You don’t get a high-performing service if senior roles are reshuffled with each election.

Key measures:

• Independent appointments commission for top posts

  • Define a tier of positions (e.g. directors of departments, heads of agencies, regional directors) that require competitive selection.

  • Establish a commission with:

    • senior civil servants,

    • external experts,

    • perhaps a limited role for cross-party parliamentary representatives.

  • The commission runs or supervises selection processes and proposes shortlists; ministers choose from the shortlist, with written justification if they deviate.

• Fixed, overlapping terms

  • Fix terms (e.g. 5–7 years) for key management posts, not aligned perfectly with electoral cycles.

  • Removal before term:

    • only for specified reasons (poor performance, misconduct, restructuring),

    • with clear, transparent procedures and appeal mechanisms.

• Transparent reporting on appointments

  • Publish appointment and dismissal decisions and basic justifications.

  • Regularly report statistics on:

    • how many contests had only one candidate,

    • how often political appointees moved into senior civil service roles,

    • gender and diversity metrics.

This doesn’t de-politicise policy choices, but stabilises execution leadership.

4. Modernising pay and career structures

A transformation-capable state can’t pay everyone the same.

a) Create differentiated “professional tracks”

• Define several key tracks:

  • General management & coordination,

  • Policy analysis & design,

  • Digital, data & AI,

  • Project and portfolio management,

  • Regulatory design & oversight,

  • Specialised technical areas (health, environment, defence, etc.).

• For each track, specify:

  • competency profiles,

  • progression path (junior → senior → expert/lead → director),

  • salary bands that reflect market realities (especially for digital and data).

b) Introduce market-informed supplements for critical skills

• Within a regulated framework, allow ministries to pay supplements for:

  • scarce digital, data, AI roles,

  • highly specialised technical roles.

• Conditions:

  • transparent criteria (skill type, level, responsibilities),

  • capped but flexible ranges,

  • regular benchmarking with labour market data.

c) Link part of remuneration to performance and mission contribution

• For senior managers:

  • part of pay linked to performance on mission-related indicators and reform milestones (not just process metrics).

• For teams:

  • introduce team-based bonuses or recognition for achieving significant improvements or delivering complex reforms.

The goal isn’t to create huge pay disparities, but to allow differentiation where it matters and to signal that performance and scarce skills are valued.

5. Reinventing HR as strategic workforce management

HR should stop being primarily a rules-and-forms shop and become a strategic function.

Strategic workforce planning

• Each ministry and major agency should maintain a multi-year workforce plan:

  • expected retirements and turnover,

  • critical roles and skills,

  • hiring and development priorities.

• The centre of government should aggregate and analyse these plans:

  • identify cross-government skill shortages,

  • coordinate recruitment campaigns,

  • design shared training programmes.

Modern recruitment and branding

• Introduce:

  • streamlined, user-friendly recruitment portals,

  • proactive outreach (career fairs, universities, online campaigns) with a clear narrative:

“Come work on missions: climate, digital state, health, security – not just paperwork.”

• Allow:

  • more flexible use of probation periods and try-before-hire mechanisms (internships, fellowships).

HR analytics

• Start systematically collecting and analysing data on:

  • time-to-hire,

  • retention rates,

  • reasons for turnover,

  • internal mobility,

  • training effectiveness.

• Use these insights to:

  • identify units with chronic staffing problems,

  • adjust pay and working conditions,

  • target interventions.

6. Building leadership, analytic and delivery capability

The transformation agenda needs people who can design, lead and deliver complex change.

National leadership academy

• Create a Leadership Academy for Public Service:

  • mandatory training for new senior managers,

  • advanced programmes for mid-level leaders,

  • thematic tracks (e.g. leading digital transformation, leading in crisis, leading cross-sector missions).

• Pedagogy:

  • case-based learning using Czech and international examples,

  • simulations and role-playing (negotiation, crisis management),

  • mentorship by experienced leaders.

Policy analysis and data

• Develop a cadre of policy analysts:

  • train them in problem diagnosis, theory of change, cost–benefit analysis, distributional impacts, behavioural insights.

• Build data science roles within ministries:

  • analysts who can work with administrative data, run models, and support evidence-informed decisions (linked to L8).

Project and portfolio management

• Introduce standardised project management frameworks:

  • require major reforms to have project charters, risk registers, milestones, stakeholder maps.

• Train and certify project managers within the civil service:

  • give them career paths and authority,

  • ensure key missions always have professional project leadership, not just ad hoc coordination.

7. Mobility and cross-fertilisation

A siloed bureaucracy learns slowly. Mobility accelerates learning.

Internal mobility

• Make it easy and attractive for civil servants to:

  • move between ministries,

  • take temporary assignments at regions or municipalities (and vice versa),

  • work in cross-government teams for missions.

External mobility

• Create fellowship programmes:

  • civil servants spend 6–18 months in:

    • academia,

    • NGOs,

    • international organisations,

    • private sector (under ethical rules).

  • bring back new perspectives and methods.

• Allow limited inflow of lateral hires:

  • mid- and senior-level roles opened to external candidates with sector expertise,

  • with transparent processes and orientation programmes to integrate them into public sector culture.

Mobility, if structured well, builds a common professional identity: “I am a public service leader working on national missions”, not “I am a department lifer”.

8. Implementation roadmap for L3

Short term (1–2 years)

• Map current senior civil service: roles, turnover, politicisation patterns.

• Establish an interim Senior Appointments Board with advisory powers.

• Pilot:

  • differentiated professional tracks in some ministries,

  • new recruitment campaigns for critical skills,

  • a basic leadership programme.

Medium term (3–5 years)

• Enact legislation for:

  • more independent, merit-based senior appointments,

  • flexible pay bands for critical roles,

  • HR data reporting and workforce planning obligations.

• Fully launch the Leadership Academy and embed it into career progression.

• Roll out project and policy analysis training widely.

Long term (5–10 years)

• Normalise cross-government and external mobility programmes.

• Consolidate HR analytics and adjust policies based on data.

• Build a recognisable “Czech Public Service” brand with strong identity and values.

9. Risks and mitigations

• Risk: Reform captured by existing insiders, used to lock in current elites.
  Mitigation: Transparent criteria, external members on appointment panels, public reporting, regular audits of appointment practices.

• Risk: Pay differentiation creates resentment.
  Mitigation: Clear communication of criteria; ensure base conditions improve for all; focus supplements on roles that clearly deliver system-wide benefits.

• Risk: Politicians resist loss of control over top appointments.
  Mitigation: Offer greater confidence in quality of execution; ensure some structured political input remains (shortlisting vs final choice); build cross-party consensus around depoliticisation.

L4 – Rational multi-level governance & local capacity

1. What problem L4 is solving

Czechia’s multi-level governance is characterised by:

• Very high municipal fragmentation (thousands of small municipalities),

• uneven administrative and technical capacity,

• complex and sometimes overlapping competence allocation between central administration, regions and municipalities,

• fiscal and grant systems that often reward project chasing, not strategic delivery.

As a result:

• national missions (climate, digital state, social services, education, housing) depend on actors that vary wildly in capability,

• the centre faces huge coordination costs,

• implementation is uneven and often slow.

L4 aims to make the territorial governance system capable of delivering national missions consistently, without violating local autonomy.

2. Clarifying roles: who is responsible for what, end-to-end

The first step is to clarify end-to-end responsibility for major policy areas.

For each mission-relevant domain (e.g. primary healthcare, primary education, social services, local transport, housing, digital services), define:

• What is the state’s role (legislation, funding, standards, oversight)?

• What is the region’s role (planning, coordination, specialised services)?

• What is the municipality’s role (front-line delivery, local adaptation)?

This should be:

• written down in a coherent, public framework,

• cross-checked for contradiction with existing laws,

• used to adjust legislation and contractual arrangements over time.

The aim is not necessarily to change constitutional structures, but to:

• remove grey zones where nobody is truly responsible,

• avoid overlapping duties that create blame games.

3. Building shared services and “delivery hubs”

Fragmentation by itself is not fatal if small units can share capacity.

Shared services model

• Identify key functions where small municipalities struggle:

  • specialised procurement,

  • IT and digital infrastructure,

  • strategic planning and project management,

  • analysis and evaluation,

  • specialised social and education support services.

• Encourage or mandate creation of shared service centres:

  • at regional level,

  • or as voluntary/inter-municipal associations with legal personality.

• Typical services:

  • centralised tendering and framework contracts,

  • shared IT platforms and support,

  • “flying squads” of planners and project managers that assist municipalities.

Delivery hubs for missions

• For specific missions (e.g. energy retrofit of buildings, digitalisation of municipal services, community-based care), designate regional or supra-regional delivery hubs:

  • these hubs receive direct support (funding, expertise, tools) from the centre,

  • they work with clusters of municipalities to implement programmes.

Example logic:

• You don’t expect every tiny municipality to design its own complex social care architecture. Instead, the region (or hub) coordinates the model and supports municipalities in implementing and adapting it.

4. Reforming inter-municipal cooperation frameworks

Czechia already has various inter-municipal structures, but they are often:

• underpowered,

• narrowly focused (e.g. waste management, water),

• not fully integrated into national strategies.

Reforms:

• Simplify legal frameworks for inter-municipal cooperation, making it easier to:

  • establish multi-purpose associations,

  • define clear governance and accountability,

  • manage shared budgets and staff.

• Provide financial incentives:

  • extra support or higher co-financing rates for municipalities that join robust cooperation structures aligned with missions,

  • access to specialised grants only through recognised cooperation frameworks (e.g. digitalisation funds only via certified regional IT hubs).

5. Aligning funding with capacity and missions

Money can either reinforce fragmentation or help overcome it.

Mission-oriented, multi-year programmes

• Instead of dozens of small project calls, create large, multi-year programmes for missions:

  • e.g. “Green & Efficient Municipal Buildings 2030”,

  • “Digital Municipal Services”,

  • “Integrated Social Care Networks”.

• In these programmes:

  • municipalities apply through or with their regional delivery hubs or cooperation structures,

  • funding is tied to outcome targets and adherence to shared standards,

  • there is built-in support for weaker municipalities (templates, technical assistance).

Capacity-based differentiation

• Introduce capacity tiers for municipalities:

  • smallest with minimal administrative staff,

  • mid-size,

  • large cities and statutory towns.

• Match responsibilities and co-financing expectations to these tiers:

  • smallest rely more on shared services and hubs, with lower co-financing obligations,

  • largest bear more responsibility and can access more flexible funding.

Predictability over “grant-hunting”

• Shift from one-off calls toward:

  • predictable envelopes for regions/municipalities based on transparent formulas,

  • performance-based top-ups for those that achieve agreed milestones,

  • fewer, larger, simpler funding streams focused on missions.

6. Strengthening regional role as integrators

Regions are natural candidates to be integrators and capacity carriers between central ministries and municipalities.

Regional strategic capacity

• Equip regional offices with:

  • strategic planning and analytics teams,

  • strong project management units,

  • IT and digital specialists,

  • experts on social and health systems, education, mobility.

• Require regions to develop regional mission strategies aligned with national missions:

  • concretise how national goals will translate into regional project pipelines,

  • coordinate with municipalities and other stakeholders (hospitals, schools, NGOs).

Regional–central contracts

• Sign mission contracts between centre and each region:

  • define outcome targets and milestones for the region,

  • agree on funding envelopes and support,

  • clarify mutual obligations (data reporting, cooperation, evaluation).

Regions thus become co-owners of implementation, not just passive recipients.

7. Capacity-building and peer learning at local level

Not all improvements have to come from the top.

Local leadership programmes

• Develop programmes for mayors, councillors and senior municipal staff:

  • leadership in multi-level governance,

  • financial management and procurement,

  • digital transition,

  • community engagement and participatory planning.

Peer learning networks

• Support networks of municipalities around specific themes:

  • small towns on social inclusion,

  • cities on smart mobility,

  • rural communities on renewable energy.

• Facilitate:

  • joint projects,

  • study visits,

  • communities of practice,

  • shared toolkits.

8. Implementation roadmap for L4

Short term (1–2 years)

• Conduct a mapping exercise:

  • who does what, where are overlaps and gaps,

  • where are the biggest capacity deficits.

• Identify pilot regions and policy domains where:

  • shared services and mission delivery hubs can be tested quickly (e.g. digitalisation, building renovation).

• Adjust a small number of grant schemes to require or reward cooperation structures.

Medium term (3–5 years)

• Update legal frameworks for:

  • inter-municipal cooperation,

  • competencies and delegated tasks,

  • shared services.

• Scale successful pilots of:

  • shared procurement,

  • shared project management units,

  • regional mission strategies.

• Embed mission contracts and funding models into multiannual financial planning.

Long term (5–10 years)

• Consider voluntary municipal mergers supported by generous transitional funding.

• Expand the role of regions as:

  • data and analytics hubs,

  • integrators of health, social, education and transport planning.

9. Risks and mitigations

• Risk: Municipalities perceive reforms as centralisation and loss of autonomy.
  Mitigation:

  • emphasise that shared services support, not replace, elected councils;

  • design governance of hubs to include municipal representation;

  • keep local political accountability for decisions, while pooling technical and administrative capacity.

• Risk: Stronger regions overshadow smaller municipalities politically.
  Mitigation:

  • clear division of responsibilities,

  • transparent participation rules for municipalities in regional planning,

  • regular regional–municipal conferences and joint committees.

• Risk: Shared services become new bureaucratic layers without adding real capability.
  Mitigation:

  • tie funding and existence of shared centres to clear performance metrics,

  • allow restructuring or termination of underperforming centres,

  • regularly assess cost–benefit of shared arrangements.

L5 – Clean, intelligent money: integrity & strategic procurement

1. What L5 is solving

L5 is about how money moves through the system and what it incentivises.

Today, Czech public finances and procurement often:

• treat formal legality as the main success criterion;

• rely heavily on lowest price in tenders, even for complex projects;

• separate anti-corruption from strategic investment thinking;

• and leave large flows (especially at local level and in public companies) in audit or transparency grey zones.

This produces three core problems:

1. Integrity risks (waste, capture, collusion, outright corruption).

2. Systematic underuse of the state’s purchasing power as a strategic tool.

3. A pervasive sense that “they’ll steal it anyway”, which kills political will for large reforms.

L5 aims to rewire the financial and procurement architecture so that:

• honest, high-quality, mission-aligned spending is the easiest option,

• high-risk, low-value, corrupt behaviour is systematically squeezed out.

2. Strategic principles for L5

Four principles should shape the reform:

1. Transparency and traceability of money end-to-end
   Anyone (internally, and often publicly) should be able to see what money is spent, by whom, on what, and with what results.

2. Risk-proportionate controls
   Heavy controls where risk and value are high, lighter where risk is low, instead of blanket bureaucracy.

3. Value and outcomes, not just price and compliance
   For complex contracts, evaluation should favour quality, performance and life-cycle cost, not just lowest bid.

4. Alignment with missions and integrity
   Funding and procurement should channel energy toward agreed national missions, while embedding anti-corruption and integrity in design, not just punishment.

3. Extending and modernising oversight

You cannot control what you cannot see.

a) Complete the audit map

• Extend the remit of supreme or supreme-like audit over:

  • municipalities above a certain size or risk profile,

  • state- and municipality-owned enterprises (transport, utilities, hospitals, development companies),

  • entities heavily dependent on public funds or EU money.

It doesn’t necessarily have to be one institution, but there must be no large blind zones in the public finance ecosystem.

b) Shift part of the focus from micro-errors to systemic design

• Redesign audit methodology so that a significant portion of work looks at:

  • whether programmes have clear objectives and indicators,

  • whether procurement models are suitable for the type of project,

  • whether repeated failures are being addressed.

c) Integrate oversight data

• Build a central data platform where:

  • findings from various audits (supreme audit, internal audits, inspectorates) are standardised and pooled,

  • recurring patterns can be spotted (e.g., similar procurement issues across sectors),

  • ministries and the centre of government can see where structural fixes are needed.

4. Rewiring procurement: from “lowest price” to “best value”

Procurement is where a huge chunk of state capacity actually lives.

a) Categorise procurement by complexity and strategic importance

• Separate procurement into at least three categories:

  1. Simple, commoditised purchases (office supplies, basic goods).

  2. Standardised but significant (typical construction, standard services).

  3. Complex, strategic or innovation-intensive (IT systems, hospitals, integrated services, long-term concessions).

Each category gets different rules and expectations.

b) Keep “lowest price” for simple commoditised items only

• For category (1), lowest price can be appropriate if:

  • quality is easily specified and verifiable,

  • the market is competitive,

  • switching costs are low.

Even there, use framework agreements and centralised purchasing to get economies of scale.

c) Require multi-criteria, life-cycle evaluation for complex contracts

• For categories (2) and (3), require MEAT-type evaluation:

  • quality, performance, maintainability, life-cycle costs, environmental and social criteria where relevant.

• Introduce standard templates and guidance so smaller contracting authorities don’t have to reinvent evaluation criteria.

d) Professionalise procurement as a career

• Create professional procurement roles:

  • clear job profiles and career paths,

  • mandatory training and certification,

  • mobility across ministries, regions and large municipalities.

• Establish central centres of excellence:

  • units that support the design and running of complex tenders,

  • repositories of model contracts and criteria,

  • advisory hotlines for smaller authorities.

5. Anti-corruption architecture that is built in, not bolted on

Integrity must sit inside spending design, not just in police files.

a) Risk-based integrity plans

• For high-risk sectors (construction, healthcare, IT, regional development), require integrity plans that cover:

  • risk mapping,

  • staff rotation policies,

  • conflict-of-interest management,

  • rules on gift, hospitality and secondary employment,

  • reporting channels and whistleblower protection.

b) Whistleblowing tied to money flows

• Connect whistleblowing systems to procurement and budgeting:

  • explicit, safe channels for reporting manipulation of tenders, grant allocation, or budgetary “fixing”,

  • clear obligations to investigate and report back,

  • tracking of follow-up actions (disciplinary, criminal, institutional).

c) Blacklists and real consequences

• Maintain and share nationally:

  • lists of companies and individuals sanctioned for corruption, bid rigging, or serious conflicts of interest,

  • rules for exclusion from future tenders for a given period.

• Tie public funding eligibility to basic integrity standards (beneficiaries without basic compliance cannot receive funds).

6. Mission-oriented public finance and EU funds

Even clean money can be strategically stupid if it’s fragmented and misaligned.

a) Design mission envelopes

• For each national mission (from L1), set up multi-year financial envelopes:

  • combine domestic budget allocations and EU funds,

  • earmark them to clearly defined mission components (e.g. “energy efficiency in public buildings”, “digital services at local level”, “primary healthcare networks”).

b) Replace scattered calls with coherent programmes

• Instead of dozens of separate calls, structure fewer, larger mission programmes:

  • stable rules over several years,

  • clear pipelines of expected projects,

  • technical assistance for municipalities/regions.

c) Introduce performance-based components

• Attach part of funding to results:

  • number of citizens served with digital services,

  • energy saved,

  • reduction in specific waiting times,

  • improvements in access to care.

• Use milestone-based disbursement:

  • early tranches for setup and capacity-building,

  • later tranches conditional on verified progress.

This forces money to pull behaviour towards national missions instead of just feeding existing structures.

7. Data, transparency and public scrutiny

Public pressure is a powerful integrity tool.

a) Radical transparency of contracts and spending

• Publish all contracts above a modest threshold:

  • contract text,

  • winning bid price and evaluation report,

  • key milestones and amendments.

• Standardise formats so they are machine-readable for civic tech and watchdogs.

b) Open data on grants and subsidies

• A central portal listing:

  • all grants and subsidies,

  • recipients, amounts, purposes,

  • progress reports and final outcomes,

  • links to evaluations where available.

c) Analytical tools for detection

• Develop or procure systems that:

  • analyse procurement and financial data for red flags (bid patterns, repeated winners, high share of negotiated procedures without publication, price anomalies),

  • feed signals to oversight bodies and investigative units,

  • can be partially open for researchers and civil society (with anonymisation where needed).

8. Implementation roadmap for L5

Short term (1–2 years)

• Map current oversight gaps across local governments and public companies.

• Tighten transparency requirements for all public entities receiving significant public funding.

• Define categories of procurement and develop initial guidance for multi-criteria evaluation.

• Launch pilot “procurement excellence hubs” in a few ministries/regions.

Medium term (3–5 years)

• Legally extend audit coverage or create complementary supervisory mechanisms.

• Professionalise procurement:

  • certification schemes,

  • standardised training modules,

  • career paths.

• Reconfigure selected EU and national funding streams into mission programmes with clear envelopes and performance components.

Long term (5–10 years)

• Consolidate integrity systems:

  • mature whistleblowing culture,

  • robust blacklists,

  • data-driven red-flag detection.

• Expand mission-oriented finance to most major investment areas.

• Regularly evaluate whether the new arrangements deliver better value and lower corruption risk.

9. Risks and mitigations

• Risk: Over-bureaucratisation – adding layers of compliance without strategic gain.
  Mitigation: Keep risk-based approach; simplify where risk is low; use digital tools to automate controls.

• Risk: Resistance from entrenched interests (local barons, captured companies).
  Mitigation: Combine transparency, enforcement, and positive incentives (easier access to funds and support for those who cooperate).

• Risk: Procurement professionals become a conservative veto point, blocking innovation.
  Mitigation: Train and reward them for enabling strategic procurement, not just preventing risk.

L6 – Democracy with teeth: participation that actually shapes decisions

1. What L6 is solving

L6 addresses a core paradox:

• Formal democracy in Czechia works: multi-party elections, alternation of power, basic rights.

• Substantive democracy often feels weak:

  • citizens have low trust and feel they have little influence;

  • participation channels are either symbolic or ad hoc;

  • complex reforms are made “over people’s heads” and then meet instant resistance.

The result: politics is a sequence of electoral shocks and symbolic fights, not a continuous, constructive process of dealing with collective problems.

L6 aims to build permanent, institutionally grounded participation and deliberation that:

• improves the quality of policy,

• increases ownership of reforms,

• extends the political space for difficult decisions.

2. Strategic principles for L6

Four guiding ideas:

1. Participation must have visible consequences
   People should see clear links from their input → options considered → decisions → outcomes.

2. Deliberation, not just aggregation
   It’s not enough to collect opinions; we need structured deliberation where people learn, discuss trade-offs, and refine preferences.

3. Institutionalisation over one-off projects
   Mechanisms must be recurring, predictable and embedded in law or standing rules, not just “nice experiments”.

4. Inclusion and representation
   Participation should reflect society, not just the loudest or best-resourced groups.

3. Citizens’ assemblies and deliberative panels

These are the backbone for dealing with complex, long-term issues.

Design elements:

• Random selection with stratification

  • participants chosen by lot, corrected for age, gender, region, education, etc.,

  • to approximate the wider population.

• Structured learning phase

  • participants receive balanced information from experts and stakeholders,

  • can question and cross-examine them.

• Facilitated deliberation

  • trained facilitators help small groups discuss respectfully and systematically,

  • participants work through disagreements and trade-offs.

• Collective recommendations

  • assembly produces recommendations, often with documented levels of consensus.

Institutional role:

• Anchored in law or parliamentary rules as advisory bodies on specific topics:

  • climate policy,

  • pension reform,

  • education reform,

  • AI and digital rights,

  • health system transformation.

• Government and parliament must:

  • respond formally within a set time,

  • state which recommendations are followed and why,

  • explain deviations publicly.

The point is not to replace elections, but to add a deliberative layer that can tackle issues regular politics mishandles.

4. Structured consultation and co-creation around laws and strategies

Beyond assemblies, everyday lawmaking needs better participation.

a) Open, transparent consultation process

• For major laws, strategies and mission programmes:

  • publish drafts early,

  • invite written submissions from citizens, NGOs, business, municipalities, academia.

• Provide:

  • clear timelines,

  • structured templates (what problem, what evidence, what alternative),

  • support for smaller organisations to participate (e.g. guides, webinars).

b) Obligatory feedback and reasoning

• Authorities must publish:

  • a summary of consultation input,

  • how it affected the final text,

  • reasons for rejecting major proposals.

• This closes the loop and signals that participation matters, even when suggestions are not adopted.

c) Thematic co-creation forums

• For complex issues, establish multi-stakeholder forums:

  • regular working groups including ministries, local governments, experts, NGOs, industry reps, user groups.

• Their output:

  • practical implementation proposals,

  • guidelines, tools, pilots.

5. Youth participation and civic learning

Czech democracy’s long-term resilience depends on how younger generations are integrated.

a) Institutional youth councils

• Create or strengthen youth councils at:

  • municipal level (linked to councils and mayors),

  • regional level,

  • national level (linked to parliament or government).

• Give them:

  • agenda-setting rights on youth-relevant topics (education, climate, housing, digital),

  • the right to be heard in committees,

  • modest budgets for projects and outreach.

b) Project-based civic education

• Make schools places of real democratic practice, not just theory:

  • students work on real problems of their school or municipality,

  • present solutions to local councils or ministries,

  • get feedback and see whether anything changes.

• Combine with media/digital literacy (L7) so students learn:

  • how to critically assess information,

  • how to argue, negotiate and compromise,

  • how institutions work in reality.

6. Participatory budgets and local-level mechanisms

Democracy with teeth must be felt locally.

Participatory budgeting (PB)

• Municipalities and, where appropriate, regions:

  • allocate a portion of their budget (even 1–5%) to participatory processes,

  • citizens propose projects and vote on them.

• To avoid superficiality:

  • involve citizens in project design and feasibility assessments,

  • integrate participatory projects into long-term planning, not as random extras.

Neighbourhood forums and assemblies

• Regular local meetings where:

  • citizens discuss specific issues (transport, safety, public spaces, services),

  • municipal officials present plans and constraints,

  • possible compromises are explored on the spot.

Online platforms

• Use digital platforms to:

  • submit ideas and complaints,

  • follow their processing,

  • participate in surveys and consultations.

• But integrate them with offline processes to avoid reinforcing participation gaps.

7. Civic tech and feedback loops

Participation must be traceable and visible.

Civic tech infrastructure

• A national or shared platform that:

  • lists all ongoing consultations, assemblies, PB processes and forums,

  • allows citizens to track issues and provide input,

  • provides accessible summaries of complex topics.

Feedback on impact

• For each major participation process, publish:

  • what was asked,

  • what input was received,

  • what decisions followed,

  • what outcomes were observed.

• Over time, these become case studies:

  • “Here is how this citizens’ assembly influenced climate law X.”

  • “Here is how PB changed mobility in city Y.”

This builds a meta-narrative: participation is not theatre; it changes things.

8. Integration with representative institutions

Participation mechanisms must strengthen, not weaken, parliament and elected bodies.

Parliamentary committees and assemblies

• Committees receive:

  • citizens’ assembly reports,

  • summaries of consultations,

  • analysis of PB outcomes.

• They:

  • debate them publicly,

  • invite participants to hearings,

  • incorporate them into legislative reports.

Formal status for outputs

• Recognise certain participatory outputs as:

  • official advisory opinions,

  • triggers for mandatory government response,

  • or, in some cases, as direct inputs into secondary regulations or guidelines.

This ensures participation is not just “around” the system, but inside it.

9. Implementation roadmap for L6

Short term (1–2 years)

• Pilot one or two national citizens’ assemblies on clearly defined topics, with strong communication and transparent follow-up.

• Map existing consultation and participation practices; identify good examples and gaps.

• Support more municipalities in adopting PB and simple local forums.

Medium term (3–5 years)

• Enact a participation framework law or set of binding rules:

  • minimum consultation and participation standards,

  • rules for citizens’ assemblies and youth councils,

  • requirements for government responses.

• Integrate participatory outputs into parliamentary procedures.

• Develop the national civic tech platform.

Long term (5–10 years)

• Normalise citizens’ assemblies as a regular tool, not a novelty.

• Expand PB and local participation into most municipalities above a certain size.

• Evaluate and fine-tune mechanisms based on experience and research.

10. Risks and mitigations

• Risk: Participation captured by organised minorities, not representative of broader society.
  Mitigation: Use random selection for deliberative bodies; design outreach that targets underrepresented groups; mix open participation with representative mechanisms.

• Risk: Participatory outputs are systematically ignored, leading to more cynicism.
  Mitigation: Legally require formal responses; communicate clearly when recommendations are not accepted and why; ensure some early cases show tangible impact.

• Risk: Participation slows decision-making and adds complexity.
  Mitigation: Use participation primarily for big, long-term choices where delay is acceptable and beneficial; set clear timelines; design processes that feed into, not block, institutional decision cycles.

L7 – Information resilience & media/digital literacy as infrastructure

1. What L7 is solving

L7 tackles the fact that:

• The information environment is chaotic: social platforms, algorithmic feeds, AI content, disinformation, propaganda, clickbait.

• Citizens’ ability to navigate this environment is very uneven.

• Institutions (media, schools, government) are not yet organised to systematically strengthen information resilience.

Consequences:

• It is hard to build shared factual baselines for any serious reform (climate, pensions, health, EU, defence, AI).

• Foreign and domestic actors can weaponise grievances and erode trust.

• Policy debates get stuck in conspiracies, tribal narratives, and misinformation.

L7 is about treating information resilience like critical infrastructure – not a side project for NGOs and “awareness campaigns”, but a coordinated, state-level system.

2. Strategic principles for L7

Four core principles:

1. Pluralism and freedom, not censorship
   The aim is to increase citizens’ competence and strengthen trustworthy institutions – not to suppress dissenting views.

2. Systemic, not ad hoc
   Media literacy, public media, counter-disinformation, and strategic communication must form one architecture, not scattered initiatives.

3. Co-production with independent actors
   The state cannot and should not do this alone. Public service media, universities, NGOs, civic tech, and platforms must be part of the design.

4. Resilience-by-design
   Build resilience into:

   • school curricula,

   • media funding models,

   • platform governance,

   • crisis management plans.

3. School system as the backbone of media & digital literacy

Schools are the only universal, structured place where future citizens can be reached.

Curriculum integration

• Include media and digital literacy as a transversal competency across:

  • language and literature,

  • social sciences,

  • ICT,

  • civic education.

• Key components:

  • how information is produced and distributed,

  • how algorithms shape feeds and visibility,

  • recognising manipulation, emotional triggers, and cognitive biases,

  • reading statistics and charts critically,

  • understanding AI-generated content and its limits.

Pedagogical approach

• Move beyond theoretical “media education” to practical, context-rich work:

  • analyse real posts, videos, memes, influencers, conspiracy theories, AI outputs;

  • run classroom simulations of misinformation campaigns and fact-checking;

  • let students create their own media content and reflect on its impact.

Teacher support

• Design specialised modules and materials for teachers:

  • ready-made lesson plans adaptable to different ages;

  • curated examples (periodically updated);

  • guidance on managing sensitive topics (politics, war, health, conspiracy theories).

• Provide:

  • continuous professional development courses,

  • online communities where teachers exchange practices,

  • partnerships with journalists, fact-checkers and researchers.

Without teacher support, curriculum changes are cosmetic. With it, schools can become systematic resilience factories.

4. Strengthening and de-politicising public service media (PSM)

Public service media are a key piece of democratic infrastructure.

Funding model

• Ensure stable, predictable, multi-year funding:

  • ideally tied to an independent mechanism (e.g. licence fee, indexed transfers) rather than year-to-year political bargaining.

• Link part of funding to:

  • quality and diversity indicators,

  • investment in digital formats and youth-oriented content,

  • collaboration with other public-interest media.

Governance and independence

• Design governance to minimise partisan capture:

  • plural, staggered appointments to supervisory boards,

  • clear, transparent criteria and procedures,

  • explicit legal duties to impartiality, pluralism and editorial independence.

• Introduce:

  • transparent performance reports,

  • codes of ethics,

  • complaint and ombuds mechanisms accessible to the public.

Mandate

• Explicitly task PSM with:

  • in-depth, explanatory journalism on public policy;

  • coverage of missions (climate, digital, health, etc.) in ways understandable to non-experts;

  • media literacy content for different age groups, including children and teenagers;

  • collaboration with schools (content, workshops, educational series).

They become a bridge between government, experts and citizens, without serving as government mouthpieces.

5. Counter-disinformation and hybrid threat units

Information resilience also has a security dimension.

Integrated threat analysis units

• Create or strengthen units that:

  • monitor information operations (foreign and domestic),

  • analyse narratives and networks,

  • identify vulnerabilities in specific groups (e.g. veterans, minorities, regions).

• They should be:

  • connected to security agencies and ministries;

  • but also able to cooperate with researchers, civil society and platforms.

Strategic communication capability

• Develop a strategic communication team at the centre of government that:

  • coordinates messaging on key issues across ministries;

  • prepares proactive narratives, not only reactive rebuttals;

  • uses insights from behavioural science and data on how people actually consume information.

Ethical red lines

• Clear legal and ethical rules to:

  • prevent propaganda by the state;

  • protect civil liberties;

  • ensure transparency when the state communicates or cooperates with platforms.

Goal: increase transparency and resilience rather than manipulate citizens.

6. Platform governance and cooperation

The state cannot control global platforms, but it can:

Frameworks and agreements

• Engage systematically with major platforms:

  • share analysis of malign campaigns, hate speech spikes, coordinated inauthentic behaviour;

  • request transparency on political advertising, recommendation algorithms, and enforcement of their own rules.

• Prefer:

  • structured agreements (e.g. codes of conduct, MoUs),

  • aligned with EU digital regulations (Digital Services Act, etc).

Transparency and accountability

• Advocate for:

  • data access for vetted researchers;

  • greater transparency on content moderation decisions;

  • clear labelling of state-controlled and AI-generated content.

The aim is not for the Czech state to “control algorithms”, but to ensure some accountability and visibility into the systems that shape public discourse.

7. Civic tech, fact-checking and independent intermediaries

Information resilience is stronger when multiple independent actors are involved.

Support for fact-checking and investigative journalism

• Provide:

  • stable, transparent funding mechanisms (e.g. media funds, grants) that are insulated from political discretion;

  • training and collaboration opportunities for journalists covering complex topics like climate, AI, defence, health.

Civic tech ecosystems

• Encourage projects that:

  • visualise public data (spending, procurement, lobbying, media ownership, misinformation narratives);

  • help citizens verify claims (browser plugins, bots, verification tools);

  • track political promises and their implementation.

Research and monitoring

• Support universities and think tanks to:

  • monitor information ecosystems;

  • evaluate the effectiveness of interventions (media literacy programmes, fact-checks, platform changes);

  • provide independent analysis to policy-makers and the public.

8. Crisis information resilience

Crisis situations (pandemics, wars, energy shocks) are stress tests.

Pre-crisis planning

• Integrate communication and information strategies into:

  • national crisis management plans,

  • sectoral emergency protocols (health, energy, defence).

• Use scenario-based exercises:

  • simulate disinformation campaigns during crises;

  • test coordination across ministries, PSM, security services, platforms.

Crisis communication principles

• Rapid, consistent, transparent communication:

  • admit uncertainty and unknowns;

  • correct mistakes openly;

  • use trusted third-party voices (scientists, professional bodies, NGOs) as amplifiers.

• Multi-channel approach:

  • TV, radio, online platforms, SMS or app alerts, local-level communication.

A state that communicates well in crises earns trust that it can later spend on long-term reforms.

9. Implementation roadmap for L7

Short term (1–2 years)

• Develop a national information resilience strategy aligning education, media policy, security and platform governance.

• Start revising school curricula and teacher training programmes for media/digital literacy.

• Set up or strengthen an integrated hybrid threat and strategic communication unit at the centre.

Medium term (3–5 years)

• Reform PSM funding and governance; clarify mandate.

• Establish long-term funding for fact-checking, investigative journalism and civic tech projects.

• Run nationwide media literacy campaigns linked to school initiatives and PSM content.

Long term (5–10 years)

• Regularly evaluate resilience (trust levels, misinformation vulnerability, performance in crises).

• Adjust strategies based on new technologies (generative AI, deepfakes, new platforms).

• Normalise cooperation between state, PSM, platforms, civil society and research as a permanent ecosystem.

10. Risks and mitigations

• Risk: Measures are perceived as censorship or propaganda.
  Mitigation:

  • focus on literacy and independent intermediaries;

  • keep PSM independent;

  • ensure transparency and legal safeguards.

• Risk: Politicisation of PSM and strategic communication units.
  Mitigation:

  • robust, cross-party governance models;

  • independent oversight;

  • professional codes and accountability.

• Risk: Fragmentation of efforts across ministries and actors.
  Mitigation:

  • central strategy and coordination;

  • clear division of roles;

  • shared data and evaluation.

L8 – Evidence & experimentation as the default

1. What L8 is solving

L8 is about the knowledge engine of the state.

Current problems:

• Evidence is used selectively and ad hoc, often to justify decisions already made.

• Evaluations are:

  • rare,

  • underutilised,

  • or commissioned as rituals with little impact on design.

• Experiments and pilots happen, but:

  • without rigorous design,

  • without systematic learning or scaling.

• Data are fragmented, hard to access, and not used strategically.

L8 aims to make evidence and experimentation the default at every major stage of policy-making.

2. Strategic principles for L8

Four guiding ideas:

1. Evidence is a process, not a magic paper
   It’s about how you define problems, test solutions, and learn – not just about one “impact study”.

2. Experimentation to reduce risk, not increase it
   Pilots and experiments are tools to avoid big, expensive failures, not to play with citizens’ lives.

3. Institutional architecture > heroic individuals
   Evidence use must be built into structures, rules, and routines – not depend on a few motivated people.

4. Transparency and external scrutiny
   Evidence used (or ignored) should be visible to parliament, media and society.

3. Central evidence & evaluation unit

At the core, you need a central unit that sets standards and supports ministries.

Mandate

• Set government-wide guidelines for:

  • problem diagnosis,

  • impact assessment,

  • evaluation design,

  • use of administrative data.

• Provide:

  • methodological support to ministries and regions,

  • quality review of major assessments and evaluations,

  • synthesis of evidence across sectors.

• Coordinate:

  • foresight exercises,

  • cross-government research priorities,

  • partnerships with academia and think tanks.

Positioning

• Located within or closely linked to:

  • centre of government (L1),

  • Ministry of Finance (budget and spending decisions),

  • possibly with dual reporting lines.

Staffing

• Economists, statisticians, social scientists, data scientists, policy analysts.

• People with experience in:

  • evaluation methods (RCTs, quasi-experimental designs, cost–benefit analysis),

  • working with administrative data,

  • communicating findings to non-experts.

4. Impact assessments that are real, not ritual

Regulatory impact assessment (RIA) and ex-ante impact analysis are often formalistic.

Reforms:

• Require impact assessments for:

  • major laws,

  • large spending programmes,

  • mission-critical strategies.

• Content:

  • clear articulation of the problem and baseline;

  • options considered (including “do nothing”);

  • expected impacts (economic, social, environmental, distributional);

  • risks and uncertainties;

  • how evidence and stakeholder input were used.

• Quality control:

  • central unit reviews major assessments,

  • returns weak ones for revision,

  • publishes its review comments.

• Transparency:

  • make impact assessments public when laws are proposed;

  • allow external critique from researchers, NGOs, business, etc.

Impact assessments become the place where politics meets analysis, instead of a post-hoc justification.

5. Evaluation culture across the policy cycle

Evaluation must be part of a continuous loop:

1. Ex-ante (design)

2. Mid-term (formative)

3. Ex-post (summative)

4. Meta-evaluation (across programmes)

Standardising evaluation practice

• Require each significant programme to include:

  • an evaluation plan with budget allocation (e.g. 1–3% of programme cost),

  • defined evaluation questions aligned with theory of change,

  • a timeline (mid-term, ex-post).

• Encourage:

  • a mix of external and internal evaluations,

  • peer review of major evaluations.

Using evaluation results

• Set decision rules:

  • programmes with consistently negative evaluations must be redesigned or terminated;

  • programmes with positive evaluations can be scaled or used as models.

• Present evaluation outcomes to:

  • parliament (committees),

  • the public (accessible summaries),

  • internal learning forums across ministries.

This creates a feedback loop where programmes are not static; they evolve.

6. Policy labs and experimentation

Experiments are how you learn what really works.

Policy labs

• Establish policy labs in key ministries:

  • small teams with skills in design thinking, behavioural insights, experimentation, data analysis.

• Their tasks:

  • transform vague political goals into testable interventions;

  • design pilots with control groups or comparison conditions;

  • run and evaluate experiments;

  • translate results into policy recommendations.

Experimental frameworks

• Legal provisions for sandboxing:

  • allow temporary derogations from some rules in defined pilots (e.g. alternative benefit delivery, new digital processes, new service models).

• Conditions for experiments:

  • informed consent where needed,

  • risk assessment,

  • ethical review for sensitive areas (health, social policy, justice),

  • clear exit strategy.

Scaling and mainstreaming

• Create mechanisms for:

  • scaling successful pilots into regulations and programmes;

  • capturing knowledge from failed experiments to prevent repetition.

Experiments thus become a normal part of improvement, not rare exceptions.

7. Data infrastructure and access

Evidence is impossible without good data.

Administrative data integration

• Identify key datasets needed for missions:

  • health records (properly anonymised),

  • education outcomes,

  • labour market data,

  • social services usage,

  • environmental and infrastructure data.

• Create:

  • secure, privacy-respecting data integration platforms;

  • clear governance (who can access what, under which conditions).

Data sharing and standards

• Develop:

  • data standards for interoperability between ministries, regions, municipalities and public enterprises;

  • guidelines on anonymisation and re-use for research and evaluation.

Open data for society

• Publish high-value datasets (with protections) as open data so that:

  • researchers, NGOs, journalists and businesses can build analyses and tools;

  • the state is not the only actor who can create evidence.

8. Science–policy partnerships

The state cannot generate all necessary knowledge internally.

Long-term partnerships

• Framework contracts between ministries and:

  • universities,

  • research institutes,

  • think tanks.

• Purpose:

  • commissioned research and evaluations,

  • modelling and forecasting,

  • expert input into impact assessments and missions.

Embedded researchers and fellows

• Programmes where:

  • researchers spend 6–24 months embedded in ministries or agencies;

  • civil servants spend time in research centres working on relevant projects.

This builds a shared epistemic community: people who speak both “science” and “policy”.

9. Embedding evidence and experimentation in political routines

Even a perfect technical system fails if politics ignores it.

Parliamentary scrutiny

• Committees require:

  • impact assessments and key evidence summaries for major laws,

  • evaluation reports for renewal of large programmes.

• Use:

  • hearings with evaluators and researchers,

  • independent expert panels.

Government routines

• Cabinet rules:

  • submissions for major decisions must include evidence sections;

  • where strong evidence is lacking, reasons must be explained along with plans to fill gaps (via pilots, data collection, etc.).

Public communication

• Government, PSM and others:

  • explain not only what decisions are made, but why and with what evidence;

  • highlight when policies are changed because evidence showed they didn’t work.

Over time, this creates a political norm: ignoring evidence is something you must publicly justify.

10. Implementation roadmap for L8

Short term (1–2 years)

• Create or empower a central evidence and evaluation unit.

• Issue updated guidelines for impact assessment and evaluation; pilot them on selected laws/programmes.

• Start building one or two policy labs in priority ministries (e.g. health, education, digitalisation).

Medium term (3–5 years)

• Make impact assessment and evaluation requirements mandatory for major initiatives.

• Develop data integration infrastructure and standards.

• Launch embedded researcher and fellowship programmes.

• Train a critical mass of civil servants in evidence use and experimentation methods.

Long term (5–10 years)

• Normalise experimentation and evaluation in all major policy domains.

• Integrate evidence cycles with budget cycles and mission reviews.

• Build a culture where “what does the evidence say?” is a default question in political debate.

11. Risks and mitigations

• Risk: Evidence is weaponised politically (cherry-picking, attacks on inconvenient results).
  Mitigation:

  • strong independence of evaluation processes;

  • transparent publishing of methods and data;

  • building a community of respected evaluators and researchers.

• Risk: Over-bureaucratisation – evidence processes slow everything down.
  Mitigation:

  • proportionate requirements (more for big reforms, less for small tweaks);

  • streamlined templates and support from the central unit;

  • pre-agreed “fast lanes” for urgent decisions with post-hoc evaluation.

• Risk: Evaluation fatigue and cynicism (“we evaluate but nothing changes”).
  Mitigation:

  • hard decision rules (terminate or redesign failing programmes);

  • public and parliamentary scrutiny;

  • clear examples of policies changed because of evidence.