A successful government should not be judged by how many people it employs, how many programmes it operates, how much legislation it produces, how many consultations it runs, or how much money it spends.

It should be judged by the outcomes it delivers for the people it serves.

Government is ultimately a delivery system.

Citizens express democratic preferences. Political leaders translate those preferences into objectives. Public institutions convert those objectives into services, infrastructure, rules, protections and other real-world outcomes.

A well-designed government delivery machine should therefore be capable of turning democratic intent into measurable outcomes:

competently, fairly, transparently, securely, efficiently and quickly.

The objective is not a small state or a large state.

It is an effective state.

Government should be large enough to perform the functions society asks of it, but no larger or more complicated than necessary to deliver those functions well.

Every pound collected from citizens represents resources that could otherwise have remained with households and businesses.

Every pound spent by government therefore carries a responsibility to create genuine public value.

The fundamental objective

The government delivery machine should seek to:

Maximise the long-term wellbeing, capability, security and prosperity of the population, subject to democratic consent, individual rights, fiscal sustainability and the responsible use of public resources.

This requires more than minimising expenditure.

A government that saves £1 today by creating a £10 problem tomorrow is not efficient.

Equally, a government that spends £10 delivering something that could have been delivered for £1 is not compassionate.

Both represent waste.

The objective should therefore be maximum sustainable public value with minimum avoidable waste.

1. Democratic delivery

1.1 Democratic mandate

Government activity should have a clear connection to democratic authority.

Citizens should be able to understand:

what elected representatives promised;
what government subsequently decided to do;
which organisation is responsible for delivering it;
how much money was allocated;
what outcomes were expected;
what was actually delivered;
and who is accountable for the result.

The chain from democratic mandate → political decision → delivery → measured outcome should be visible.

1.2 Politics sets outcomes; delivery delivers them

The legitimate role of elected politicians is to determine political priorities within the constitutional and legal framework.

The role of the permanent public service is to provide expert advice and then deliver lawful government policy professionally and impartially.

The civil service should not become an alternative political system.

Equally, professional officials should be able to provide honest technical advice without being expected to manipulate evidence to support a predetermined conclusion.

There should be a clear distinction between:

democratic choice and professional execution.

1.3 Political neutrality

Public institutions should serve governments of different political traditions with equal professionalism.

Recruitment, promotion and institutional influence should not depend upon political ideology.

Civil servants should be able to provide robust evidence and challenge poor proposals internally while faithfully implementing lawful decisions once they are made.

2. Outcome-oriented government

2.1 Define outcomes, not activity

Every major government programme should begin by defining the outcome it is intended to produce.

Success should not primarily be measured through:

money spent;
staff employed;
meetings held;
reports produced;
consultations completed;
programmes launched;
or announcements made.

These are activities.

They are not outcomes.

The relevant question is:

Did the intervention actually improve the thing it was intended to improve?

2.2 Measurable performance

Wherever practical, public services and programmes should have measurable indicators covering:

outcomes;
service quality;
delivery time;
unit cost;
administrative cost;
citizen experience;
distributional impact;
failure rates;
long-term effects.

Metrics should support judgement rather than replace it.

Poorly designed targets can themselves create dysfunctional behaviour.

2.3 Clear responsibility

Every important outcome should have an identifiable owner.

Responsibility should not disappear into committees, departments, agencies, contractors and cross-government working groups.

For any significant programme it should be possible to answer:

Who is responsible for making this work?

Authority and accountability should remain aligned.

A person cannot reasonably be held accountable for an outcome if they lack the authority, information or resources required to influence it.

2.4 Accountability when things go right and wrong

Excellent performance should be recognised.

Failure should be investigated honestly.

The purpose of accountability should not be to create a culture of fear. People must be able to experiment, identify problems and admit mistakes.

But persistent incompetence, dishonesty, negligence, obstruction or misconduct should have consequences.

Public institutions should neither protect poor performance indefinitely nor punish people merely because a difficult project encountered legitimate problems.

3. Zero-waste government

3.1 Every pound matters

Public money should be treated with the same seriousness as if the decision-maker had earned and saved it personally.

Government should continuously ask:

What outcome are we buying, and could we achieve the same or better outcome with fewer resources?

Waste includes more than fraud.

It includes:

unnecessary administration;
duplicated organisations;
redundant reporting;
avoidable consultancy;
unnecessary meetings;
poorly designed procurement;
obsolete IT;
duplicated datasets;
failed programmes;
excessive management layers;
preventable demand for public services;
and processes that exist primarily because another process exists.

3.2 Administrative-cost test

For every public service, government should understand the distinction between:

C_{\text{total}} = C_{\text{frontline}} + C_{\text{admin}} + C_{\text{capital}} + C_{\text{waste}}

The objective should be to minimise:

C_{\text{admin}} + C_{\text{waste}}

subject to maintaining necessary governance, safety, accountability and service quality.

Administrative expenditure should have to justify itself through the value it enables.

A useful measure for every major service would therefore be the administrative burden:

R_{\text{admin}} = \frac{C_{\text{admin}}} {C_{\text{total}}}

and the proportion of total expenditure reaching productive or frontline activity:

R_{\text{productive}} = \frac{C_{\text{frontline}} + C_{\text{capital}}} {C_{\text{total}}}

These measures should not become simplistic targets. Some services legitimately require substantial administration, governance or specialist oversight.

But government should at least know these numbers.

3.3 Frontline-first principle

Resources should flow as directly as reasonably possible towards the people, infrastructure and services that create the intended outcome.

Teachers should teach.

Doctors and nurses should care for patients.

Police officers should police.

Engineers should build and maintain infrastructure.

Scientists should conduct science.

Caseworkers should help people.

Administration exists to enable frontline delivery.

Frontline delivery does not exist to sustain administration.

3.4 Complexity is a cost

Every new:

form;
approval;
committee;
reporting requirement;
management layer;
regulator;
agency;
dataset;
procurement rule;
consultation;
process;
and statutory obligation

creates a cost somewhere in the system.

Sometimes that cost is justified.

Sometimes it is essential.

But complexity should never be free by assumption.

Every additional layer should have a demonstrable function that cannot be achieved more simply.

4. Prevention before remediation

Government should prefer preventing predictable problems where prevention produces better long-term outcomes at lower total cost.

This principle applies across:

healthcare;
infrastructure maintenance;
crime;
education;
housing;
environmental protection;
energy;
welfare;
cybersecurity;
public health;
and fiscal policy.

The relevant comparison is not simply:

“What does prevention cost?”

It is:

“What is the expected whole-system cost of preventing this problem versus allowing it to occur and dealing with the consequences?”

Conceptually, preventative action should be preferred where:

C_{\text{prevention}} < \mathbb{E}\left[C_{\text{failure}} + C_{\text{remediation}} + C_{\text{consequences}}\right]

where the right-hand side represents the expected future cost of allowing the problem to occur.

A stitch in time should genuinely save nine.

5. Excellent people

5.1 Domain expertise

Complex systems should be led and operated by people who understand them.

Government needs excellent:

engineers;
economists;
scientists;
doctors;
nurses;
teachers;
software engineers;
statisticians;
lawyers;
operators;
product managers;
project managers;
procurement specialists;
financial professionals;
diplomats;
security specialists;
and other domain experts.

Management skill matters.

But management should not become a substitute for knowledge.

The objective should be:

domain experts capable of leadership, not bureaucracy as a profession in itself.

5.2 Recruit excellent people

Government should be capable of attracting outstanding people from:

public service;
industry;
academia;
charities;
frontline professions;
entrepreneurship;
technology;
and international experience.

Movement between sectors should be normal rather than exceptional.

Government should actively import knowledge from organisations that solve similar problems better.

5.3 Fair pay for fair work

Public servants should be paid fairly and competitively for the work they perform.

Pay should reflect:

capability;
responsibility;
scarcity of skills;
performance;
and the need to attract excellent people.

Public service should neither depend upon underpaying talented people nor permit compensation, pensions or employment structures that create unreasonable liabilities for future taxpayers.

The objective is:

fair pay for fair work, sustainably funded.

5.4 Knowledge transfer

Government should retain institutional knowledge without becoming institutionally stagnant.

When people leave roles:

important knowledge should remain accessible;
decisions should be documented;
systems should not depend on one individual;
successors should receive structured handovers;
lessons from success and failure should be preserved.

Government should have organisational memory without becoming imprisoned by precedent.

6. A productive working culture

6.1 Meetings are a cost

Meetings should exist to make decisions, solve problems, coordinate genuinely interdependent work or transfer necessary information.

Meetings should not exist because they are recurring calendar events.

People who are not contributing or learning something necessary should be free to leave.

Every meeting consumes public resources.

If $n$ people attend a meeting for $t$ hours at an average fully loaded employment cost of $c$ per hour, its direct labour cost is approximately:

C_{\text{meeting}} = n \times t \times c

That does not mean meetings are bad.

It means they should create value greater than the time they consume.

6.2 Shortest-path communication

Information should travel through the shortest responsible path required to solve the problem.

Communication should not have to climb and descend unnecessary management hierarchies.

A frontline employee who identifies a serious problem should have a viable route to the person capable of fixing it.

6.3 Plain language

Government should communicate in language ordinary people can understand.

Unnecessary:

acronyms;
jargon;
management terminology;
policy language;
institutional euphemisms;
and deliberately obscure technical language

should be eliminated.

Complex ideas sometimes require technical terminology.

Simple ideas should not be made complicated to sound important.

6.4 Common sense

Rules exist for reasons.

Where application of a rule clearly defeats the outcome the rule was intended to produce, there should be a controlled mechanism for escalation and correction.

Good institutions combine:

rules + judgement + accountability.

6.5 Happy and meaningful work

A high-performing public service should also be a good place to work.

People should:

understand why their work matters;
have sufficient autonomy to do it well;
work with competent colleagues;
be treated with dignity;
receive useful feedback;
have opportunities to learn;
and be proud of the service they provide.

Burnout, pointless work and institutional cynicism are themselves forms of waste.

6.6 Friendly public service

Citizens interacting with government should normally encounter people and systems that are:

helpful;
respectful;
competent;
responsive;
clear;
patient;
and solution-oriented.

The citizen is not an inconvenience to the institution.

Serving the citizen is the reason the institution exists.

7. Digital government

7.1 Digital by design

Government should operate as a modern digital organisation.

Paper processes manually reproduced in software are not digital transformation.

Services should be redesigned around:

users;
data;
APIs;
automation;
interoperability;
secure identity;
reusable infrastructure;
and measurable outcomes.

Technology should simplify government rather than merely digitise its bureaucracy.

7.2 Information should be usable

Government possesses enormous quantities of valuable information.

Subject to privacy, security and legal constraints, information should be structured so that authorised people can actually use it.

Government should develop secure, interoperable data infrastructure capable of supporting:

policy analysis;
service delivery;
forecasting;
fraud detection;
infrastructure planning;
operational decision-making;
scientific research;
and evaluation of outcomes.

The same information should not have to be manually collected repeatedly by different departments where lawful and secure reuse is possible.

7.3 People who need to know should know

Information should flow to the people who require it to perform their responsibilities.

Access should be determined by legitimate operational need rather than institutional politics or administrative convenience.

The principle should be:

the right information, to the right person, at the right time, for the right purpose.

7.4 Privacy and security

A digitally capable state must also be a trustworthy state.

Personal information should be:

collected only where justified;
minimised where possible;
securely stored;
appropriately encrypted;
access-controlled;
auditable;
retained only as necessary;
and used only for legitimate purposes.

Citizens should not have to surrender unnecessary privacy to obtain competent public services.

7.5 No preventable data leaks

Government systems should be engineered on the assumption that they hold information capable of causing serious harm if mishandled.

Cybersecurity and data protection are therefore operational requirements, not compliance exercises.

Material breaches should be investigated, causes identified and systems improved.

8. Transparency and confidentiality

8.1 Radical transparency of public activity

Citizens should be able to see, subject to legitimate privacy and security limitations:

where public money goes;
who received it;
what it purchased;
what programme it supported;
who authorised significant decisions;
what outcomes were expected;
what outcomes occurred;
what contracts exist;
how programmes are performing;
and where major failures occurred.

Government should make transparency the default rather than requiring citizens and journalists to reconstruct the state through Freedom of Information requests.

For public expenditure, the desired chain should be traceable:

taxpayer → budget → department → programme → supplier/service → expenditure → measured outcome

8.2 Confidentiality where necessary

Transparency does not mean publishing everything.

Some information must remain confidential, including:

personal data;
active security operations;
sensitive diplomatic communications;
legally privileged material;
commercially sensitive information;
and legitimate private policy deliberation.

But confidentiality should have a defined purpose.

It should not be used to conceal embarrassment, incompetence or misuse of public resources.

8.3 No unauthorised leaking

Officials entrusted with confidential information should protect it.

Government cannot function effectively if confidential discussions are routinely leaked for political, personal or media advantage.

The corresponding obligation is that official transparency mechanisms must be strong enough that leaking is not required to expose ordinary government performance.

Protected whistleblowing concerning wrongdoing should remain distinct from unauthorised leaking for political or personal purposes.

9. Learn from the best

Government should not assume that government already knows the best way to perform a task.

For every major capability it should ask:

Who in the world does this exceptionally well, and what can we learn from them?

Relevant knowledge may come from:

other countries;
local government;
businesses;
universities;
charities;
the military;
engineering;
manufacturing;
software;
logistics;
finance;
aviation;
healthcare;
or citizens themselves.

Government should use best-in-class techniques wherever they are applicable.

The public sector should not be technologically or organisationally isolated from the society it governs.

10. Fast enough to matter

A correct decision delivered after the opportunity has disappeared can still be a failure.

Government processes should operate on timescales appropriate to the systems they govern.

Decision time should therefore be measured.

Unnecessary delay should be treated as a cost.

A useful representation is:

T_{\text{delivery}} = T_{\text{decision}} + T_{\text{approval}} + T_{\text{procurement}} + T_{\text{implementation}}

Each component should be visible and capable of challenge.

Where a decision takes months or years, government should be able to explain:

why that time is required;
which activities consume it;
which risks are being controlled;
and whether those controls remain proportionate.

The objective is not reckless speed.

It is the elimination of unnecessary latency.

11. Diverse needs, simple interfaces

Citizens are not homogeneous.

People differ in:

income;
age;
health;
ability;
family circumstances;
geography;
education;
digital access;
language;
employment;
risk;
and the support they require.

Government should accommodate this diversity without creating unnecessary complexity for everyone.

The internal system may be sophisticated.

The citizen-facing experience should be as simple as possible.

Good government should manage complexity internally rather than exporting it to citizens.

12. A strong country that protects the weak

Efficiency is not the ultimate purpose of government.

Efficiency creates the capacity to achieve worthwhile outcomes.

A capable state should be strong enough to:

protect vulnerable people;
maintain the rule of law;
provide essential services;
defend the country;
respond to emergencies;
invest in future generations;
protect individual rights;
support economic opportunity;
and act when markets or institutions cannot adequately protect people themselves.

Strength and compassion are complements.

A state must be capable in order to protect effectively.

13. Economic and fiscal health

Government should preserve the long-term productive capacity of the country.

Policy should consider effects on:

productivity;
investment;
entrepreneurship;
employment;
infrastructure;
skills;
innovation;
energy security;
fiscal sustainability;
inflation;
debt;
and intergenerational fairness.

Today's political promises should not quietly become tomorrow's unfunded liabilities.

At the simplest level, the state's long-run fiscal position must recognise:

R_t - E_t = B_t

where:

$R_t$ is government revenue in period $t$ ;
$E_t$ is government expenditure in period $t$ ;
$B_t$ is the resulting fiscal balance.

Persistent deficits accumulate into liabilities that must ultimately be serviced by future productive activity.

Borrowing may be justified where it creates durable productive capacity or responds to exceptional circumstances.

It should not become a substitute for confronting structural inefficiency.

14. A capable country in the world

Government should represent the country competently internationally.

The country should aim to be:

a reliable ally;
a credible trading partner;
capable of defending itself;
constructive in international institutions;
respected for its competence;
strong enough to assist weaker partners;
attractive to talent and investment;
and confident in defending its legitimate interests and values.

Domestic institutional competence is itself a source of international strength.

15. Continuous improvement

Government should never assume that an institution, programme or process deserves to exist forever simply because it exists today.

Every significant activity should periodically face four questions:

What outcome is this intended to produce?
Is it producing that outcome?
Is there now a better way to produce it?
What would happen if we stopped doing it?

Institutions should be capable of:

improving;
consolidating;
automating;
shrinking;
transferring responsibilities;
and, where their purpose no longer exists, closing.

Successful institutional leadership should not be measured by how large an organisation becomes.

Sometimes the greatest success is making an organisation's intervention less necessary.

16. The government-as-control-system principle

Government can ultimately be understood as a feedback system.

Society has a measurable state:

x_t

Democratic processes establish desired outcomes or a target state:

x_t^{*}

The difference between the desired and observed state is the system error:

e_t = x_t^{*} - x_t

Government has a set of available interventions or control inputs:

u_t

A competent government should select proportionate interventions in response to the observed state and the causes of deviation:

u_t = \pi(x_t, x_t^{*}, e_t, d_t)

where $d_t$ represents external disturbances and changing conditions that government does not directly control.

The consequences of intervention then affect the future state of society:

x_{t+1} = f(x_t, u_t, d_t)

Government should therefore continuously:

observe the state of the country;
compare measured outcomes with democratically determined objectives;
understand the causes of important deviations;
select proportionate interventions;
act;
measure what happened;
learn;
adapt.

This creates a continuous feedback loop:

observe → compare → decide → act → measure → learn → adapt

This is fundamentally different from assuming that writing a policy, passing legislation, creating an organisation or allocating a budget has solved the problem.

The output of government is not policy.

The output of government is the real-world outcome produced by policy.

17. The overall optimisation problem

Taken together, the government delivery problem can be expressed conceptually as:

\max_{\{u_t\}} \; V_{\text{public}}

where $V_{\text{public}}$ represents sustainable public value across outcomes such as:

wellbeing;
security;
prosperity;
health;
opportunity;
infrastructure;
environmental quality;
freedom;
resilience;
and fairness.

Subject to constraints including:

C_{\text{government}} \leq C_{\text{sustainable}}

R_{\text{admin}} \rightarrow \min

C_{\text{waste}} \rightarrow \min

while maintaining:

democratic legitimacy;
individual rights;
constitutional constraints;
service quality;
national security;
privacy;
resilience;
fairness;
and long-term fiscal sustainability.

The purpose is therefore not simply:

minimise government spending

nor:

maximise government provision

but:

maximise sustainable public value from the resources available to society.

The ultimate test

A successful government delivery machine should make it possible for a citizen to ask:

What did government try to achieve, what did it spend, what did it do, what happened, and who was responsible?

And receive a clear answer.

It should attract excellent people without becoming a protected professional class.

It should retain expertise without becoming resistant to change.

It should protect confidentiality without hiding failure.

It should use data without abusing privacy.

It should regulate without suffocating innovation.

It should protect the vulnerable without creating avoidable dependency.

It should invest for the future without wasting the resources of the present.

It should be capable of doing difficult things quickly when necessary, while remaining constrained by democracy, law, evidence and accountability.

Above all:

Government should exist to deliver outcomes for people — not to perpetuate the machinery of government itself.