The Bribe You Can't See

1. Corruption has a familiar shape

Imagine two price tags hanging from the same government service.

The first says:

Official fee: Rs1,000

The second is invisible.

It says:

Come again tomorrow.

or:

Bring another document.

or:

Find someone who knows the officer.

or:

Pay Rs3,000 and the file will move.

The visible price is Rs1,000.

The effective price may be Rs4,000 plus a day's lost work.

That difference is the hidden tax of administrative dysfunction.

It is why corruption cannot be separated completely from bureaucracy.

A complex process creates more discretion.

Discretion creates more points of friction.

Friction creates opportunities for shortcuts.

Shortcuts create markets for intermediaries and unofficial payments.

Not every bureaucratic delay is corruption.

Not every unofficial payment comes from an official.

And not every person who uses an agent is bribing someone.

But the structure creates opportunities for all of them.

2. What Pakistanis said in 2025

Transparency International Pakistan's 2025 National Corruption Perception Survey was conducted in September 2025 and covered 4,000 respondents across 20 districts in Punjab, Sindh, KP and Balochistan.

Respondents identified:

The survey also reported that 77% of respondents were dissatisfied or had low satisfaction with government efforts to combat corruption.

That is a public-perception signal, not a measurement of every corrupt act.

But perception matters.

When a large share of citizens believes corruption is entrenched, it changes how they approach government.

They may assume:

“The normal process will not work.”

Once people assume the official route is unreliable, middlemen become more attractive.

That can create a self-reinforcing system.

3. Perception is not incidence

This distinction is essential.

A survey asking:

Which sector do you think is most corrupt?

measures perception.

A survey asking:

Did you pay a bribe to access a service?

measures self-reported experience.

An audit measuring:

How much money disappeared from a contract?

measures a different phenomenon.

We should not blend all three.

PPRI should therefore use three labels:

Public perception

What citizens think.

Reported experience

What citizens say happened to them.

Documentary evidence

What records/audits/law-enforcement material establish.

This separation is especially important when writing about politically sensitive institutions.

4. The 2025 survey contains a paradox

The same survey found that a majority of respondents did not report paying a bribe to access a public service during the preceding 12 months.

That matters.

It prevents the lazy conclusion:

“Everyone in Pakistan has to pay a bribe.”

Clearly that is not established.

At the same time, a substantial minority did not fall into the “did not pay” category, and the survey shows strong concern about corruption and weak confidence in government anti-corruption efforts.

The better conclusion is:

Bribery is not universal, but it is widespread enough to shape how citizens perceive the state.

That's a much more defensible finding.

5. What the 2023 survey tells us about the price

The 2023 NCPS offers selected reported average bribe amounts.

Examples include:

• Punjab police: about Rs21,186
• Sindh land administration: about Rs6,426
• KP judiciary: about Rs162,000
• Balochistan health: about Rs160,000

These numbers need to be read carefully.

They do not mean:

“A police bribe in Punjab costs Rs21,186.”

They mean:

respondents who reported such a payment reported an average amount around that level.

The distribution underneath an average can be enormous.

One person may report Rs500.

Another Rs50,000.

Another may report a much larger amount.

The median, frequency and service-specific distribution are therefore as important as the average.

This is exactly where raw survey microdata would be valuable.

6. Why corruption is a regressive tax

Suppose two households face an unexpected Rs5,000 unofficial payment.

Household A has a monthly disposable income of Rs100,000.

Household B has Rs30,000.

The nominal payment is identical.

The economic shock is not.

For Household A:

Rs5,000 = 5% of monthly income.

For Household B:

Rs5,000 = 16.7%.

This is a simple illustration, not a measured national ratio.

But it explains why corruption is particularly damaging to poorer citizens.

It takes a larger share of the budget.

And unlike taxes, the payment often comes at precisely the moment when the citizen needs a government service.

7. Corruption can be a tax on urgency

Some government services can wait.

Others cannot.

Consider:

• a passport needed for work abroad;
• a land record needed for a transaction;
• a death certificate needed for inheritance;
• a police report needed for legal action;
• a health-service approval;
• a business licence;
• an identity document.

The more urgent the service, the more valuable speed becomes.

If a corrupt system allows someone to buy speed, then the process creates a market:

delay for those who wait normally

versus

speed for those who can pay.

This makes corruption a mechanism for sorting citizens by their ability to pay.

8. The hidden role of intermediaries

Intermediaries are not all corrupt.

Agents can perform legitimate work.

Many people genuinely need help navigating complicated systems.

The policy question is:

Why does a citizen need a private intermediary to navigate a government service that should be intelligible without one?

This is where corruption and administrative design meet.

When government creates a simple, transparent process, there is less value in selling access to the process.

When the process is opaque, the value of insider knowledge rises.

9. Why simplification can be anti-corruption policy

Suppose a government service requires:

10 documents
3 offices
4 signatures
2 physical visits

and has no application tracking.

Now redesign it:

3 documents
1 service point
digital record
automatic tracking
published deadline

The reform doesn't merely make life easier.

It removes opportunities for discretionary obstruction.

A citizen no longer needs an intermediary to explain where the file is.

An official no longer has the same opportunity to say:

“The file is with another department.”

The transaction becomes more observable.

Transparency increases because the process itself becomes simpler.

10. Pakistan already has examples of this approach

Punjab's e-Khidmat model is explicitly built around:

• one-window facilitation;
• simpler procedures;
• easier access;
• standardized services;
• accountability and transparency;
• reduced opportunities for corruption.

That does not prove the model has eliminated corruption.

But it demonstrates an important design principle:

Anti-corruption does not begin only with punishment. It can begin with redesigning the opportunity to demand a bribe.

11. Complaint institutions reveal where friction survives

The Wafaqi Mohtasib received more than 226,000 complaints in 2024.

Punjab's Ombudsman recorded thousands of complaints across sectors including:

• Local Government and Community Development;
• Revenue;
• Police;
• Health;
• Housing and urban development.

These cases should not all be interpreted as corruption.

Many are maladministration:

• delay;
• wrong decisions;
• failure to act;
• procedural problems.

But that is precisely the point.

A citizen does not need to experience a bribe for government to become costly.

Maladministration itself has an economic price.

12. Corruption can hide inside time

This is the part of the problem that is not captured by a bribe survey.

Imagine there is no unofficial payment.

A government office still delays an application for three weeks.

The applicant:

• makes four visits;
• spends Rs1,500 on transport;
• misses two workdays;
• pays a consultant Rs2,000 to understand the process.

Total private cost:

more than the official fee by a large margin.

Nothing appears in the corruption statistics.

Yet the citizen has still paid for weak governance.

This is why PPRI should study:

bribery + administrative burden

together.

13. The citizen's real transaction cost

We propose:

Citizen Cost of Public Service (CCPS)

For a service s:

CCPS = Official fee + travel + documentation + intermediary cost + lost-income cost + unofficial payment + repeated-visit cost

Each component should be measured separately.

This creates an important ethical advantage.

PPRI does not have to guess that:

“Corruption costs Pakistan X% of GDP.”

Those kinds of macro estimates can become highly speculative.

Instead, we can show something more concrete:

A citizen attempting Service X paid Rs___ officially and incurred Rs___ in additional costs.

That is understandable.

14. A survey PPRI can actually run

We propose a national citizen-friction module. Ask users who recently dealt with government:

Service

What did you need?

Official cost

How much did it officially cost?

Time

How many hours/days did the process take?

Visits

How many physical visits?

Travel

How much did travel cost?

Missing documents

How many times were additional documents requested?

Intermediary

Did you use a private agent?

Payment

Were you asked for an unofficial payment?

Outcome

Was the service eventually obtained?

Complaint

Did you complain?

Satisfaction

Would you describe the process as fair, predictable and understandable?

This would allow PPRI to produce something Pakistan currently lacks:

a citizen-level picture of the cost of bureaucracy.

15. How sentiment should be measured

PPRI should resist calling political speeches “public sentiment.”

A better hierarchy is:

Level 1 — Survey sentiment

Directly ask citizens.

Level 2 — Complaint sentiment

Analyse citizen complaints for recurring frustrations.

Level 3 — Media discourse

Analyse how journalists and columnists describe corruption.

Level 4 — Political rhetoric

Treat it as political rhetoric.

These are four different datasets.

For example:

“Everyone has to pay a bribe.”

is a statement.

It is not a survey result.

A research institute should not turn it into one.

16. A useful future PPRI text-analysis project

Once PPRI has gathered enough material, we can conduct a controlled text analysis of:

• Citizen Portal complaints;
• ombudsman complaint descriptions where available;
• newspaper reporting;
• court decisions;
• public consultation submissions.

Instead of only producing “positive/negative sentiment,” classify the type of frustration:

Delay
Confusion
Bribery
Harassment
Non-response
Discrimination
Unclear rules
Repeated visit
Missing infrastructure
Service failure

This is much more useful than a generic sentiment score.

A sentence being “negative” tells us little.

Knowing that 24% of complaints in a service are about delay tells us something actionable.

17. Corruption and local government

This debate connects directly to decentralization.

A local government can reduce corruption in some services because decisions become:

• geographically closer;
• more visible;
• easier to scrutinize.

But local government can also create opportunities for local elite capture.

Therefore the question is not:

local government = less corruption.

It is:

Which institutional design makes decisions visible, contestable and auditable at the level where services are delivered?

That means:

• published budgets;
• procurement transparency;
• digital payments;
• public registers;
• complaint systems;
• local oversight;
• audit;
• disclosure of projects.

18. Procurement is the other half of the problem

The 2025 NCPS identified tendering/procurement as the second most-corrupt sector by perception.

That matters because procurement is where public money becomes:

• roads;
• schools;
• hospitals;
• water systems;
• municipal services.

A corruption problem here is therefore not just about money disappearing.

It can produce:

higher prices
lower quality
delayed projects
fewer projects
maintenance failures.

Imagine a road contract worth Rs100 million.

If corruption adds 10% to the effective cost, the public loses the equivalent of Rs10 million.

That is an illustration, not an estimate of Pakistan's actual procurement loss.

The larger point is:

Procurement corruption can turn into a service-delivery problem.

19. Corruption and inequality reinforce one another

There is a vicious cycle:

poor service
↓
citizen needs private solution
↓
those with money can buy the solution
↓
those without money wait
↓
unequal access
↓
greater inequality

Corruption can therefore be both:

a consequence of weak institutions
and:
a mechanism that deepens inequality.

This is why anti-corruption policy should not be isolated from social policy.

20. Why punishment alone cannot solve the problem

Suppose a corrupt official is prosecuted.

Good.

But if the next officer inherits a system with:

• unclear rules;
• manual files;
• no tracking;
• many approval points;
• weak supervision;
• no published service standard,

the opportunity for corruption remains.

The sustainable reform is:

punishment + process redesign + transparency + digital traceability + accountability.

21. A better anti-corruption question

Instead of:

How corrupt is Pakistan?

ask:

Where does a citizen have the greatest opportunity—or need—to pay to make government work?

That question produces a research agenda.

For each service:

official process
actual experience
complaint data
bribery report
time burden
service outcome

Then compare.

22. The role of provincial reform

The four provincial governments are responsible for a huge share of the public services citizens actually encounter.

That makes provincial institutions central to the corruption story.

A province can improve access by:

• shifting services to district level;
• building one-window systems;
• empowering municipalities;
• digitizing records;
• publishing service deadlines;
• decentralizing approvals;
• creating local complaint channels.

But decentralization must include resources and oversight.

Otherwise an officer closer to the citizen may simply be a local gatekeeper with the same discretionary power.

23. The case for transparent local budgets

Imagine every district government publishes:

This district received Rs10 billion.

Then citizens can see:

Education — Rs2.1bn
Water — Rs1.4bn
Roads — Rs1.6bn
Health — Rs1.2bn

and every scheme is shown on a map.

Then citizens can ask:

Why is my road here and not there?

Why did the project cost this much?

Why isn't the work finished?

Transparency becomes practical.

It turns the budget into something citizens can interrogate.

24. What a good system should do to corruption

We propose a six-part framework:

Simplify

Reduce unnecessary steps.

Publish

Make fees and procedures clear.

Track

Allow citizens to see application status.

Digitize

Create an auditable record.

Decentralize

Move decisions closer where appropriate.

Audit

Make spending and performance visible.

This is not a magical formula.

It is a systems approach.

25. What the existing evidence tells us

The current evidence supports several conclusions:

1. Corruption remains a major public concern.
2. Police, procurement and judiciary are prominent areas of perceived corruption.
3. A majority of respondents in the 2025 survey did not report paying a bribe to access a public service during the relevant period; corruption is therefore not universal.
4. A substantial public trust problem remains, with 77% reporting low satisfaction with government anti-corruption efforts.
5. Reported bribe amounts can be substantial in some contexts.
6. Maladministration and service delays create costs even when no bribe is paid.
7. Administrative simplification can reduce opportunities for discretionary obstruction, but it needs monitoring.

26. What the evidence does not yet tell us

We still do not have:

• a national estimate of total household bribe expenditure;
• a representative estimate of total hours lost to corruption;
• a nationwide service-by-service administrative burden measure;
• reliable causal evidence that decentralization by itself reduces corruption in Pakistan.

Those should be research questions, not invented conclusions.

27. The PPRI Corruption Access Map

PPRI could eventually build an interactive product showing, by service and district:

reported corruption perception
complaint volume
administrative burden
official fees
processing time
local-government structure

This would be far more informative than simply colouring Pakistan:

“red = corrupt.”

Corruption isn't a weather system.

It arises through specific interactions between citizens and institutions.

We should map those interactions.

28. The deeper argument

The common anti-corruption debate begins with morality:

Officials should not take bribes.

Of course.

But a research institute should go further.

Ask:

What institutional conditions make bribery useful?

When the official process is:

fast + predictable + transparent

a bribe is less valuable.

When the official process is:

slow + uncertain + discretionary

a bribe becomes more valuable.

The most durable anti-corruption strategy therefore changes the economics of the transaction.

Conclusion: the invisible bill

A citizen receives a government service and sees an official fee.

But the real bill may contain:

Rs1,000 official fee
Rs800 transport
Rs1,500 lost wages
Rs500 documents
Rs3,000 intermediary
Rs5,000 unofficial payment

The state records perhaps Rs1,000.

The citizen experiences Rs11,800.

That difference is the hidden cost of government.

Not every service has that problem.

Not every official is corrupt.

Not every citizen pays a bribe.

But the existence of large complaint volumes, substantial corruption concerns and reported unofficial payments shows why the problem deserves measurement rather than slogans.

PPRI's goal should therefore not be another paper saying:

Pakistan is corrupt.

Everyone already knows the phrase.

The useful question is:

Which government processes impose the highest hidden cost on citizens, why does that cost arise, and which reforms actually remove it?

That is measurable.

And once it is measurable, it becomes much harder to ignore.