Sindh's Property Tax Reform: The Test in 20 Councils Outside Karachi

In Hyderabad, a surveyor taps a location on a handheld GPS at the gate of a three-storey house, records the floor area, the number of storeys, the roof type and the current use, photographs the facade, and then knocks on the door. The person who answers hears two words — property tax — and shakes his head. No one in this house has ever received a return form. No one has ever seen a bill. No one knows whether the house exists in the local council's records at all. The house has electricity, water and people living in it, yet in public data it is almost invisible.

People usually call that gap a revenue shortfall. I call it a sentence written by hurried hands: hands that built the registers decades ago and never came back. The Sindh government, with the financial and technical backing of the World Bank, is now trying to rewrite that sentence through a door-to-door survey in 20 local councils outside Karachi, within a total of 45 councils the province is targeting. The striking part is that the centre of the programme is not the tax rate. It is the map.

A tax system built on an incomplete map

Property tax in Sindh, as in most of South Asia, is not levied on market value but on an estimated rental value. The mechanism has clear virtues: stability, ease of collection, fewer disputes. It also has a fatal weakness: when the property market rises quickly while estimated rental values stay frozen, the tax base shrinks over time. A villa in Hyderabad today may be worth many times the figure in the register, yet the bill is still based on the old number.

The second problem is more serious: registers are incomplete. Rapid urbanisation in Sindh has produced new neighbourhoods, new buildings and subdivided flats, but the registration system has not kept pace. The result is a substantial share of property that exists on the ground but not on paper. Karachi accounts for the bulk of the province's property revenue, but even in Karachi, the precedent of digitised land records and a one-window model — known as the CLICK project — shows that when data is cleaned up, the number of discovered properties always exceeds expectations.

Outside Karachi, the picture is far blurrier. Hyderabad, Sukkur, Larkana, Mirpurkhas and Shaheed Benazirabad are cities with different economies, populations and construction speeds, yet they share one trait: local councils lack both the staff and the tools to know how many properties they manage. A council that does not know how many taxpayers it has cannot budget, cannot measure performance, and cannot speak of fairness.

That is why expanding the tax base to 20 councils outside Karachi cannot be merely a payment-awareness campaign. It has to be a census. And every census begins with the hardest question: who is in the register, and who is outside it.

Why a database is a political matter

There is a familiar temptation in every tax reform: to treat it as a technical exercise. Buy GPS devices, hire surveyors, build software, train staff, publish maps. But property tax has never been purely technical. It is a question of power, because it answers a question many would rather avoid: who owns what, and who is paying less than the real value.

A door-to-door survey inadvertently becomes a local inventory of power. It uncovers unregistered properties, buildings constructed beyond permit, land used for the wrong purpose, subdivided flats that were never declared. Every discovery touches a social relationship. The same dataset can therefore be read in two ways: as a collection tool, or as a threat.

This is the point that purely technical analysis never reaches. One can draw a perfect map of the 20 councils outside Karachi, but a map cannot decide whether councils dare to send bills. Data creates capability. Politics decides whether that capability is used.

The door-to-door survey: programme design

The core of the programme is a cadastral survey tied to a geographic information system, carried out door to door. Surveyors record location, area, structure, number of storeys, use and legal status, and take photographs. The data is then entered into a central system and cross-checked against land records held by the Board of Revenue.

The method is not new. It is standard in property tax reform projects across developing countries. What is new is the scale and the organisation: 20 councils at once, with a precise allocation — Hyderabad 9, Sukkur 3, Larkana 4, Mirpurkhas 2, Shaheed Benazirabad 2. Added to the 25 councils in Karachi already on the path, the total is 45.

The figure of 45 deserves a pause. It reflects a political decision: to expand beyond Karachi, that is, beyond the strongest economy, the richest data and the deepest administrative resources. The 20 councils outside Karachi are where capacity is thinnest, data weakest and — paradoxically — where the need for local revenue is greatest.

The allocation by city also tells a story. Hyderabad receives nine councils, nearly the combined total of the other four cities. That reflects the population size and number of administrative units of Hyderabad, Sindh's second city. Sukkur, Larkana, Mirpurkhas and Shaheed Benazirabad receive fewer, but each council there will face the same problem: low registration density, low tax awareness, and an uneven collection history.

The technical crux is this: a cadastral survey does not collect tax. It only creates the basis for collection. The gap between the two — between knowing and collecting — is where every property tax reform in the world either succeeds or dies.

A programme that has all four links — survey, valuation, notification, collection — creates value. Remove one link and the upfront investment becomes a sunk cost. Experience in developing countries shows survey costs often absorb most of the reform budget, while the percentage of success lies in the later, less glamorous, less funded steps.

Four roles, one stage: who does what

A programme involving the provincial government, the World Bank, the Local Government Department and the Board of Revenue raises the question of responsibility allocation — and that question decides success more than any technical factor.

The Sindh government is the owner and the political decision-maker. Without provincial resolve, no local council would dare expand the tax base, because expansion means touching locally influential people.

The Local Government Department is the implementing agency. That is administratively sensible, because local councils collect property tax directly and provide urban services directly to payers. But the department is also chronically overloaded and dependent on the provincial budget. Handing it a large-scale survey without permanent capacity is the surest way to turn reform into a one-off campaign that then goes dark.

The Board of Revenue holds land records and transfer transactions. Coordination between the department and the board is a technical condition for survival. Property tax data is only useful when attached to land ownership — otherwise every dispute stops at who owns this land and never reaches how much tax is due.

The World Bank provides financing and technical support. That role brings resources, international standards and reporting discipline. It also brings a familiar risk: when funding ends, the incentive to maintain the system may vanish with the money, unless the system has learned to sustain itself from new revenue.

Finally, there are the Town Citizen Committees. This is the most undervalued link. A committee playing its role properly can turn a state-imposed survey into a local bargain: we pay tax, we see roads, drains and street lights. A nominal committee, by contrast, only adds another layer of procedure.

Where the programme can break

Data never lies. Only the way we read it lies. And the most common reading — treating the number of surveyed properties as the number of properties that will pay — is the most dangerous misreading in any property tax reform.

The first gap is the gap between enumeration and collection. A survey may uncover thousands of properties. But a property only becomes revenue when it is valued, filed, notified, disputed and resolved, and finally paid. Every step can break. International experience shows the conversion rate from surveyed to paid is often far lower than initial expectations, especially in the early years.

The second gap is between data and institutions. A beautiful GIS system does not replace a transparent collection process. If councils still collect tax through negotiation, acquaintance or informal deals, then new data merely makes arbitrariness more sophisticated.

The third gap is valuation and dispute. Estimated rental value is an argumentative number, because it lacks the clear market benchmark of a sale price. When a wave of new tax notices based on new valuations goes out, appeals will surge. Without a fast, fair and independent appeals mechanism, councils will clog at the very first stage, and taxpayers will learn that objection is an effective way to delay.

The fourth gap — the most overlooked — is the impact on vulnerable groups. A door-to-door survey does not distinguish rich from poor. But the tax burden does. A tenant in a working-class district may see the cost passed through into rent. A small household in a subdivided unit may be valued against commercial standards. Without exemptions, thresholds and sufficiently refined use classification, the programme will generate a political backlash from precisely the people whose cooperation it needs most.

The fifth gap is local political economy. Property tax is the most visible local tax — payers see the bill, the number and the due date. No tax generates stronger local political reaction. In small councils such as Sukkur, Larkana, Mirpurkhas or Shaheed Benazirabad, where local networks are dense, a council determined to collect correctly may face political pressure far beyond its capacity to resist.

The cities that will be the test

Hyderabad, with nine councils, is the test of scale. It is where the programme can prove that centralised data management works in a large city outside Karachi. If Hyderabad stumbles, the stumble will spread to smaller cities, because what survives pressure in a large city tends to be easier to run than in a small one, not harder.

Sukkur, Larkana, Mirpurkhas and Shaheed Benazirabad are the test of sustainability. Each has two to four councils, small enough for a local failure to become serious, and far enough from the centre for technical support to become expensive. This is where the question of whether reform survives after the project ends will be answered.

The remaining 26 councils, including 25 in Karachi and the rest of the 45, serve as a benchmark. Karachi has the CLICK precedent, existing data infrastructure and relatively available staff. Comparing Karachi with the 20 councils outside it will show precisely the capacity gap between centre and periphery — a gap that reform reports usually describe with adjectives rather than numbers.

The number to watch is not the tax rate

Tax is not where money speaks, but where fear whispers — fear of losing local control, fear of losing votes, fear of losing the calm of old relationships. That is why property tax reform usually does not fail in the technical room. It fails at the doorstep, in the council chamber, on a phone call from an influential person.

The encouraging part is that this programme is designed against the old habit: it starts from data rather than from tax rates. If 45 councils truly hold a complete, transparent and updated property register, then every future debate about tax fairness, urban services and local budgets has ground to stand on. If, by contrast, the survey only produces a beautiful database that stays in a drawer, Sindh will repeat the old loop: shortfall, calls for reform, survey, shortfall again — with a higher-tech invoice.

A verifiable prediction: over the next 24 to 36 months, the most important indicator is not the new tax rate, nor the number of councils surveyed, but the share of surveyed records that complete the chain of registration — valuation — notification — appeal — payment. If that share is below 30% across the 20 councils outside Karachi, the programme will be read as a failure, whatever the quality of the data. If it exceeds 60%, Sindh will have a rare South Asian precedent: a property tax reform that travels from data to money, rather than from announcement to disappointment.

When the registers close and the meetings disperse, what remains is not a speech. What remains is the numbers on the bill, and the only question every resident asks: I pay tax, so what do I get back.