Tax administration: Where power meets people

This series has examined what is taxed, how it is designed, and who pays. Now we ask: how do people actually comply? Who can enter the system, and who struggles before enforcement even begins?

Tax administration is where the system meets people. It decides who can register, who can file, who can pay. It decides whose time is taken, whose distance is measured, whose work is recognised.

Before any penalty, before any audit, before any enforcement – there is administration. It is here that some taxpayers are designed into ease. Others into struggle.

Access – who is permitted entry — and on whose terms?

Registration is the first gate.

It demands identity documents, stable addresses, formal recognition. A salaried worker with formal employment is often pre‑registered or guided through the process. A small trader, a market vendor, or a cross‑border informal operator must navigate forms, offices, and requirements that may not align with how they live or work.

Registration is digitised. An educated taxpayer with a smartphone can complete the process online in minutes. A rural trader cannot. She has no smartphone, no reliable internet, no digital literacy.

She can seek help from a revenue officer – a free service – but that means travelling to a tax office, queuing, losing a day’s work. Time and distance become barriers. What is free in money is costly in hours.

Registration becomes a barrier. Filing becomes a filter. Entry becomes exclusion dressed as efficiency.

Access is not just administrative; it is socially mediated. Women — particularly in informal economies — are less likely to have formal IDs, registered assets, or recognised business status. Access is therefore shaped by legal identity regimes that historically centre male ownership and formal employment.

Access is not a gateway. It is a gatekeeper.

Who bears the cost of a system that digitises without inclusion?

Information – who is expected to know?

Tax obligations are communicated through forms, notices, and digital portals. They presume a level of literacy – legal, financial, and often linguistic. Instructions may be technically correct but practically inaccessible.

Large firms hire accountants. Professionals access advice. Small business owners rely on fragments.

An online survey I conducted in 2025 of over 100 taxpayers in Uganda found that even among this relatively advantaged group – people who were literate, had internet access, and were already navigating the formal system – over half still found tax rules unclear. A third of women in the formal sector cited lack of information as a primary challenge, notably higher than men.

If the tax system is opaque to those with literacy, digital access, and formal employment, what does it look like to a rural trader with no smartphone, no internet, and no accountant?

Complexity concentrates advantage. It excludes before enforcement ever begins.

Information assumes time, literacy, and exposure. It raises deeper questions: Who is socialised into financial knowledge? Who is excluded from formal financial spaces? Women’s economic activity is often learned informally, yet tax systems privilege formal, technical knowledge.

Information is not neutral — it reflects whose knowledge counts.

Who pays when understanding is a privilege – and even the privileged struggle?

Time: whose time is taken — and rendered invisible?

Filing is not just a task; it is time. Time away from income-generating activity, from care work, from survival.

For some, filing is a scheduled administrative step. For others, it is a day lost — or several.

A woman who runs a catering business must close her kitchen, travel to the tax office or the bank, and queue for hours. She loses income. She loses time with her children.

The system records revenue. It does not record time.

Yet time is what many cannot afford to lose.

In my 2025 research, women described distance not just as geography, but as time they could not afford. They spoke of long journeys to tax offices, of limited time flexibility, of the work required just to make time available — arranging childcare, closing businesses, losing income.

A three-hour journey becomes a full day.

Women’s time is already spoken for — by unpaid care, domestic labour, and community responsibilities.

Tax systems treat time as infinitely available — and economically neutral.

But it is neither.

Time is not a resource. It is where the system extracts — quietly, repeatedly, and without record.

Who pays with hours that never return?

Distance: how far must one travel to be seen as compliant?

Tax offices, banks, and authorised agents are unevenly distributed. Urban proximity enables compliance; rural distance complicates it. Transport costs, queues, and repeated visits turn compliance into a logistical burden.

A rural trader travels hours to the nearest tax office. She loses a day’s income. The distance is not a cost; it is a barrier.

Distance is not just geography. It is how far the system sits from people’s lives.

Tax administration assumes movement — to offices, banks, agents. It assumes the ability to leave work, to travel, to return.

But movement is not equal. It is shaped by safety, by mobility norms, by caregiving constraints.

A trip to a tax office is not the same for everyone.

For some, it is a short errand. 

For others, it must be planned around children, around markets, around the work that cannot be paused.

A salaried man can step away and return. 

A woman managing a business and a household must reorganise her entire day — or lose it.

Distance stretches time. It raises cost. It disrupts income.

The system measures proximity in kilometres. It does not measure what it takes to move.

Who can reach the system — and who cannot?

Digitalisation: Inclusion for whom?

Digital systems promise efficiency. They deliver speed for some — and exclusion for others.

Tax administration increasingly assumes digital access: online registration, e‑filing, mobile payments. It assumes devices, stable connectivity, and digital fluency.

But access is uneven.

In my 2025 survey on just taxation, men reported substantially higher access to digital systems (53.2%) than women (25.3%).

A woman with a smartphone files online. 

A woman in a rural town with poor internet travels to a cybercafé.

Systems accelerate. Access does not.

Digital progress without inclusion redraws inequality.

Digitalisation does not remove effort; it redistributes it. Those already equipped move easily. Others navigate unstable connections, unfamiliar platforms, or intermediaries — often at a cost.

While compliance costs were not cited as a major barrier by most respondents, they affect more than twice as many women (16.7%) as men (6.7%).

Digitalisation creates new burdens. Those burdens fall unevenly.

Who pays when technology leaves you behind — and when even the costs of coping are gendered?

Legibility: whose work is seen — and whose is erased?

Tax administration depends on what it can see.

It is built on records — traceable income, documented transactions, verifiable accounts. Work that is recorded becomes legible. Work that is not is harder to recognise, harder to process, and often treated as non‑compliant.

Formal businesses are structured around record‑keeping. Informal actors often are not. The requirement is not neutral — it reshapes how economic activity must be organised.

A female consultant with invoices is legible. 

A woman who runs a small guesthouse, paying workers in cash and receiving payments partly in cash, is less legible. Her income is layered. The system records some of it. The rest is invisible.

But the system still taxes her spending — her transport, her supplies, her inputs.

Much economic activity operates this way: informal, cash‑based, seasonal, embedded in relationships. Transactions are recorded in notebooks, in memory, or not at all.

The system reads one form of work easily. It struggles with another.

To comply, people must translate their livelihoods into formats the system can read. That translation takes time, knowledge, and resources — not everyone has them.

What cannot be recorded becomes difficult to declare. What cannot be declared becomes vulnerable to error, dispute, or exclusion.

The system measures what is written down. It does not measure what is done.

What kinds of work become legible — and what kinds are rendered non‑compliant by design?

Liquidity: what if ‘on time’ is a luxury?

Tax administration assumes money is available at specific moments.

Payment systems are built around timing — when tax is due, how it is collected, and in what form. Pay‑as‑you‑earn aligns with regular salaries. Advance taxes and lump‑sum payments do not align with irregular incomes.

A salaried worker has tax deducted from each paycheck. 

A trader must assemble a large sum by a fixed date.

Her income fluctuates. Her expenses are daily.

The system measures compliance against the calendar. It does not measure the rhythm of income.

When money comes in unevenly, “on time” becomes uncertain.

To meet a deadline, some must borrow, sell stock, or divert money meant for household needs. Payment is made — but at a cost the system does not see.

Timing is not neutral. It privileges stability.

Who is already out of step with the system — before any penalty is even applied?

Support: who can afford help — and who must navigate alone?

Tax administration is not self‑executing. It is mediated through support — accountants, tax agents, and advisory services that interpret rules, complete filings, and reduce uncertainty.

But support is not evenly available. It depends on networks, income, and proximity to formal systems. Access is uneven and often costly.

A large firm has a dedicated tax department that manages compliance as routine. 

A small trader asks a friend, guesses, or leaves the form incomplete.

One navigates through structured expertise. The other navigates through uncertainty.

Support becomes a resource that can be purchased, inherited through networks, or simply not accessed at all.

Is compliance an individual responsibility — or a service that must be bought to exist within the system?

Who can afford help — and who must navigate alone?

Design bias: who is forced to adapt – and who is excluded?

Tax systems assume stable income, formal employment, individual taxpayers, and linear time. These assumptions shape every other layer of compliance.

Administrative systems are built around a particular model of economic life: stable income, formal employment, and a single, fixed identity.

But real lives do not follow that script.

A woman’s income may come from a shop, rental property, and freelance consulting. She moves across roles, incomes, and responsibilities. Her economic life is layered, shifting, and non-linear.

The system forces this complexity into simplicity. It asks for one identity, one income stream, one structure that can be declared and verified.

What does not fit must be compressed, translated, or excluded.

Those who cannot easily fit the model are penalised — not because they are non‑compliant, but because their lives are not easily made legible to the system.

Who pays when the system was not designed for you?

Representation: who designs the system — and who is it designed for?

Tax administration is shaped not only by systems and rules, but by the people who design and operate them.

Who administers tax shapes how it is administered.

When systems are designed by those who have never run a small business, blind spots become rules. What is not understood is not prioritised. What is not experienced is often not designed for.

This is reflected within institutional structures themselves. Uganda Revenue Authority data from a 2025 online survey I conducted shows that while women hold 40% of board positions and 45.8% of Assistant Commissioner roles, they are absent at the top executive tier. Core technical functions also remain male‑dominated — IT (30.3% women), Customs (36.0%), and Tax Investigations (39.4%).

Representation therefore thins out as decision‑making becomes more technical and more powerful — precisely where system design choices are made.

This shapes what is treated as normal in tax administration: how digital systems are built, how compliance is defined, and how enforcement is prioritised.

Design is never neutral. It reflects who is present in decision‑making spaces — and who is structurally absent from them.

Who pays when the people who make the rules do not live the reality they regulate?

Fragmentation: who experiences the system as many — and who as one?

Tax administration does not arrive as a single system. It arrives in parts.

It collects through fragments — fees, licences, levies, permits, charges — often across multiple authorities, offices, and payment points.

For those inside formal structures, these fragments are coordinated. They are consolidated into predictable obligations, managed through systems, schedules, and accounts.

For others, the system is not one system at all.

A woman running a small shop pays taxes, then market fees, then local levies, then transport-related charges just to move goods. Each payment is small on its own. Together, they accumulate into a heavy, continuous burden.

None of these payments translate into rights, services, or protection in a visible or guaranteed way. They register as obligations, not reciprocity.

Fragmentation does not make the system lighter. It makes it harder to see — and harder to contest.

Burden increases without fairness increasing alongside it.

Who experiences the system as many systems layered on top of each other — while others experience only one?

Data: what is not measured is not addressed

Tax administration runs on data — records, returns, filings, classifications. What is not captured in data is rarely designed for in policy.

But data is selective.

Informal work, unpaid labour, and much of lived economic reality often sit outside formal records. They are fragmented, inconsistently captured, or not captured at all.

Invisibility in data becomes invisibility in design. If something does not appear in the system’s evidence base, it rarely shapes how the system is built, resourced, or adjusted.

This is not neutral absence — it is structured omission. Entire forms of economic life are rendered statistically thin, and therefore administratively secondary.

Policy then reflects what is measurable, not necessarily what is lived.

Who pays when the system does not even know they exist?

Efficiency: when systems become efficient, who do they work for?

Tax administration is increasingly measured through efficiency — faster processing times, higher digital uptake, improved revenue yield, reduced administrative delays.

But these are system‑facing metrics. They describe how the machine performs, not how it is experienced.

They rarely capture confusion, repetition, failed attempts, multiple visits, or quiet withdrawal from the system altogether.

A system that processes a return in two minutes may still take a trader two days to complete — navigating forms, seeking clarification, finding access to a device, or recovering from errors.

Efficiency at the centre can produce inefficiency at the edges.

What becomes easier for the system does not automatically become easier for the user.

In some cases, acceleration intensifies pressure: fewer steps on the system side can mean more responsibility shifted onto the taxpayer.

When systems become more efficient, the question is not only how fast they work — but for whom they become easier, and for whom they become more difficult to navigate.

Who benefits when efficiency is defined from the system’s point of view?

The administrative question beneath the technical system

Tax administration is often framed as neutral infrastructure. In practice, it distributes burden through its design: through who it recognises, how it communicates, where it is located, and what it assumes about people’s lives.

Compliance is structured before it is enforced. Some taxpayers are designed into ease. Others into struggle.

The system creates unequal capacity to comply.

Next: Tax Enforcement – who is punished and who escapes? Harassment, extortion, impunity; harsh on the powerless, lenient on the powerful.