The bus will be quiet. That is the first thing people notice about an electric bus: the absence of the diesel shudder that has been the background music of Karachi’s roads for a generation. It will be clean, air-conditioned and, at least in the photographs, new. Sindh’s People’s Green Transport Project envisages 500 of them, 450 for Karachi and 50 for Hyderabad, and the official architecture is more serious than a vehicle order: charging systems, five depots, 850 shelters and an intelligent transport system. As displayed by the Sindh PPP Unit in July 2026, the project remains at investor-solicitation stage, with a request for proposals dated 30 December 2025, which means the fleet exists so far as an architecture and an invitation rather than a timetable. It still deserves to be taken seriously. This is not a procurement dressed up as policy. It is policy with an operating system attached.
And yet one of the most consequential decisions inside the programme has received far less attention. It is not which buses to buy, from whom, or on what financing terms. It is where to send them.
A bus is a technology. A route is an allocation. The drivetrain decides what comes out of the tailpipe. The route decides who gets to work on time, which crowded van or borrowed motorcycle is displaced, which neighbourhood gets an hour of its day returned, and where the public value of a public asset finally lands. Karachi is about to make hundreds of these allocation decisions, and almost all of the attention is on the hardware. The bus can be green before the route is intelligent. They are separate achievements, and only one of them photographs well.
There is a larger reason to care, and it sits far from Karachi, at the borders of the European market. The EU’s Carbon Border Adjustment Mechanism entered its definitive regime on 1 January 2026. Its current scope covers cement, iron and steel, aluminium, fertilisers, electricity and hydrogen. CBAM does not currently cover textiles, and nothing in this article should be read as claiming otherwise. But it has changed the terms of the argument around carbon and trade. It establishes the principle that embedded emissions can become a condition of entry into a major export market. The EU is also considering an extension to specific downstream goods made using materials already covered by the mechanism. The immediate legal obligation sits in other sectors. The direction of travel does not.
Figure 1 · Carbon at the Border, Pressure in the Factory
Carbon-conditioned trade enters supplier operating reality
Two-lane systems diagram showing CBAM establishing the principle of carbon-conditioned trade while supplier operating reality moves through wider carbon pressure, Scope 3 employee commuting, state-shaped transport data and route allocation as a policy lever.
Trade direction
CBAM establishes the principle of carbon-conditioned trade.
Textiles are not currently in scope.
Supplier operating reality
Wider carbon measurement and reduction pressure
Corporate Scope 3, including Category 7
State-shaped transport and data
Route allocation as a policy lever
Direction of travel and wider market pressure, not a current textile CBAM obligation.
The route question sits inside an industrial and trade architecture, even where textiles are not currently covered by CBAM.
For a labour-intensive export economy, public transport is therefore not separate from industrial decarbonisation. It is part of the infrastructure through which that decarbonisation either becomes possible or remains something suppliers are instructed to deliver alone. Which is why the question of where Karachi sends its electric buses is worth more attention than it has received.
The Route Is the Second Decision
Procurement is the visible half of green transport. It comes with numbers that fit in a headline: 500 buses, five depots, 850 shelters, an estimated 215,000 daily riders, a figure best read as ambition rather than observation. All of it can be announced, budgeted, tendered, audited and inaugurated.
Routing is the less visible half, and it is where much of the environmental arithmetic is settled. An electric bus running full through a corridor where it replaces overloaded vans and ageing motorcycles is doing one kind of climate work. The same bus running half-empty along a corridor already served by other modes is doing another kind entirely. Same battery, same drivetrain, same tender, same press release. Different public outcome.
There is nothing automatic about the word electric, either. Electric buses can reduce emissions. But the reduction is made by a route that is used, that displaces higher-emitting travel and that runs on a defensible electricity baseline. A route that draws passengers out of high-emission, low-occupancy modes is a different investment from one that competes with walking, with cycling or with an existing service. Displacement is a property of the route, not of the vehicle. No specification sheet can supply it.
Figure 2 · The Same Bus, Different Emissions Outcome
Electric is not enough
Symmetrical route scenario diagram showing one identical electric bus producing different emissions outcomes depending on occupancy, utilisation and mode displacement.
Route context A
Fuller bus
Higher utilisation
Displaces higher-emitting travel
Stronger climate outcome
Same vehicle
The drivetrain is identical. The route is not.
Route context B
Weaker occupancy
Lower utilisation
Weak mode displacement
Weaker climate outcome
Illustrative route scenarios, not observed Karachi route-performance data.
Electric is not enough. Route design determines much of the climate value.
The route determines who receives access and who keeps waiting. It determines load factors and dead mileage, and therefore the real emissions cost of every passenger-kilometre the fleet delivers. A fuller bus spreads each vehicle-kilometre across more passenger-kilometres; an emptier one does the opposite. It determines whether the new vehicles extend the corridors, depots and interchanges the city has already paid for, or quietly duplicate them. It determines whether the affordability benefit reaches households for whom transport consumes a serious share of income, or subsidises journeys that would have happened anyway.
None of this is an administrative afterthought to be sorted out once the vehicles clear the port. Routing is the second sustainability decision, as consequential as the first and far less examined. The route is where the announcement becomes policy.
The Factory Gate Is Not the Boundary
I read bus announcements the way a supplier reads them, which is to say with an emissions inventory somewhere at the back of my mind. Under the GHG Protocol, a company’s climate account does not stop at its own walls. When employees travel to work in vehicles the company neither owns nor controls, those journeys sit in Scope 3, Category 7: employee commuting. The category is unglamorous, difficult to measure and easy to postpone. In labour-intensive manufacturing, it is also anything but marginal.
A recent working review with SupplierSays community members who prepare and examine GHG inventories across the Global South found that, in the large-scale textile manufacturing inventories examined, Category 7 generally accounted for roughly 15 to 20 per cent of total Scope 3 emissions. That is not a textile-sector benchmark, and it should not be projected onto other industries. It is a recurring signal from the inventories reviewed.
Figure 3 · The Factory Gate Is Not the Boundary
Employee commuting sits outside the gate but inside the account
Factory-boundary diagram showing residential-origin nodes connected by worker journeys to a factory, with the perimeter labelled Scope 3 Category 7 employee commuting and a caveat that the 15 to 20 percent observation is not a sector benchmark.
Scope 3 - Category 7: Employee Commuting
North-west housing
Informal settlements
Peripheral neighbourhoods
Mixed commuter corridors
Central node
Factory
The emissions story begins before the shift begins.
Working observation
15-20%
In the large-scale textile manufacturing inventories reviewed, Category 7 generally accounted for roughly this share of total Scope 3 emissions.
Working observation from reviewed inventories, not a sector benchmark.
Labour-intensive production depends on journeys the factory does not own.
Large-scale textile production remains unusually dependent on the daily movement of people, particularly where spinning, fabric and garment operations sit within the same manufacturing system. Thousands of people must reach those operations before production can begin, which can make employee commuting materially more significant than it would be in industries operating with much smaller workforces.
In a human-heavy industry, the factory gate is not where the emissions system begins. Thousands of people must reach it first. And as large textile manufacturers reduce emissions from electricity and process heat, the movement of the workforce becomes harder to dismiss within the remaining climate account.
The number carried another weakness, and it is the more instructive one. Several of the calculations behind it relied on UK government conversion factors because the practitioners involved could not locate suitable, reliable, relatable public factors for the relevant geographies and transport modes. The choice is defensible. The UK factors are accessible, documented, updated annually and familiar to anyone who works with a GHG inventory. They are also, by their publisher’s own description, suitable for UK-based organisations and for international organisations reporting on their UK operations. The public-bus factors draw on British transport data, including fuel use, vehicle kilometres, passenger kilometres and average occupancy assumptions. They are not invented. They are carefully constructed estimates derived from British bus operations. But a British bus factor is still a British bus factor. It cannot know the age, maintenance condition, loading pattern, traffic conditions, congestion or fuel quality of the bus carrying workers through Karachi.
So the true commuting total in those inventories could be higher or lower than reported. Actual emissions may be materially different from what the imported factor suggests, particularly where fleets are older, poorly maintained, heavily congested or running at occupancy levels no British timetable would recognise. The size of that difference cannot be responsibly stated without locally relevant data, and suitable public data was precisely what the practitioners could not locate. A practitioner facing a material emissions source has three options. Use an imperfect imported proxy. Build a potentially more representative model whose assumptions are harder to verify independently. Or leave the source poorly estimated. The imported proxy often becomes the least-worst option, because it is at least documented, with a published source and methodology that can be defended.
Figure 4 · The Proxy Problem
Missing measurement infrastructure becomes a supplier reporting problem
Decision-branch diagram showing a material commuting source to estimate, with three equal options: imported proxy, local model, or leaving the source poorly estimated.
Material source
Commuting emissions to estimate
01
Use imported proxy
Documented; defensible; imperfect fit
02
Build local model
Potentially more representative; harder to verify independently
03
Leave source poorly estimated
Weaker disclosure; weak comparability; unresolved uncertainty
This does not make UK factors dishonest, unusable or necessarily high or low for Karachi. It makes the local evidence gap visible.
The problem is not false methodology. The problem is the absence of publicly usable local measurement infrastructure.
This is the quieter failure underneath the reporting requirement. Suppliers across the Global South are increasingly asked to produce internationally credible carbon accounts of the kind that carbon-conditioned trade increasingly demands, while public institutions in many of the relevant geographies have not consistently made the local transport data and emission factors available. Too many suppliers across the Global South are being asked to measure themselves with instruments their own states have not built.
The measurement gap runs straight back to the route question. Route allocation determines which modes a new electric bus displaces. Without local transport data and local emission factors, the climate value of any given route stays partially invisible, to the city as much as to the companies reporting alongside it. Public infrastructure decisions create emissions consequences that private reporters are then expected to estimate with foreign proxies.
A public electric bus does not automatically fix any of this, and it would be convenient nonsense to claim otherwise. A supplier cannot book a Category 7 reduction because a green fleet exists somewhere in the city. It needs evidence: employees actually riding, modes actually shifting, occupancy actually measured against a defensible baseline. What a public bus on the right corridor can do is make the reduction possible. It moves the question from whether workers can travel cleanly to whether the route ever reaches them.
And that question is not the supplier’s to answer. A manufacturer can procure cleaner power, electrify process heat and improve its own operations. It cannot lay on a citywide bus network, choose public routes or publish national transport emission factors. Part of the decarbonisation infrastructure sits, and will always sit, with the state.
Workers Are Not a Private Transport Problem
There is a reflex in transport policy that files the industrial commute under private logistics. The destination is a factory; the factory has an owner; let the owner solve it. The reflex is understandable, but wrong. A garment worker travelling across the city to an industrial estate is not an input being delivered to a production line. She is a citizen crossing her own city to reach employment, on the same roads, in the same heat, with the same claim on public infrastructure as a student travelling to a university or a patient travelling to a hospital.
The claim is not abstract. Karachi’s industrial districts and its affordable housing have drifted apart over decades, which quietly converts employment access into a transport problem. Where the distance between home and work is long and the transport filling it is informal, unreliable or unsafe, the costs are specific: hours lost every day, wages consumed by fares, and a labour market that narrows for women, for whom an unsafe commute is often the difference between taking a job and staying home. None of this appears in a procurement document. All of it is decided by a route map.
Industrial time sharpens the point. A factory does not open when the worker arrives; the worker arrives when the factory opens. That inversion makes the industrial commute among the least flexible journeys in the city and the least forgiving of unreliable transport. A missed connection is not always a minor inconvenience. It can mean lost wages, disciplinary consequences and, over time, a less secure job. Transport economists would call this a captive market. It is more accurate to call it a claim.
In an export economy under growing carbon pressure, the corridor that moves an industrial workforce is not only social infrastructure. It is part of the machinery through which trade-exposed industry decarbonises, or fails to.
The counter-reflex deserves its hearing too. Public transport should not quietly absorb costs that employers would otherwise carry. Some firms already run contract shuttles, and a public route that displaces them transfers a private expense to the public purse. That is a fair design question, answerable through co-financing and honest accounting of who benefits. It is an argument for structuring industrial routes carefully. It is not an argument for striking workers off the map.
The Case Against Our Own Case
If the argument so far sounds like a brief for sending buses to factories, it is worth stating plainly what the evidence does not support.
First, Karachi’s planning system has not ignored industrial mobility. The Yellow Line is designed around the Korangi to Landhi employment corridor: a 21-kilometre line from Dawood Chowrangi to Numaish whose current PPP description proposes 256 electric buses across trunk, direct and feeder services. The World Bank describes the corridor as designed for over 300,000 passengers a day, at an estimated cost of USD 450 million including a USD 382 million loan. Whatever else is true, one of Karachi’s major industrial corridors is also the focus of one of the city’s largest transport investments. The honest question is not why industrial workers have been forgotten. It is how the next 450 buses will connect to, extend or complement what the Yellow Line already recognises.
Second, public transport exists for more than employment. Education, healthcare, residential access, women’s mobility across every trip purpose, and the coherence of the network itself are legitimate public claims, not obstacles to be argued past. A city that routed its entire green fleet by industrial logic would simply have inverted the error this article describes.
Third, industrial concentration is not the same thing as bus demand. Shift-based corridors may offer predictable peaks. They may also prove sharply directional, thinly used off-peak and expensive in dead mileage. The available public evidence does not tell us. This research found no published load factors, directional-balance analysis or industrial-worker origin-destination dataset for Karachi. Official route directories list the two Peoples Bus Service routes designed to serve the SITE estate as inactive, and what SITE’s workers use instead is not documented in any public dataset. The gap cuts both ways. It prevents anyone from proving that industrial routes would outperform. It equally prevents anyone from dismissing them without assessment.
Fourth, the border argument must not be overworked. CBAM’s current scope does not touch textiles. The extension now moving through the EU’s institutions concerns specific downstream goods, not garments, and it must not be conscripted as an imminent textile tariff. The case here rests on direction, not on a deadline.
The evidence is not strong enough to declare a winner. It is strong enough to demand a fair test.
What Populism Actually Changes
This is where the politics enters, and it enters without needing a villain. No allegation about any official, party or institution is required here, and none is made. What is required is a look at the incentive architecture that any government, of any composition, operates inside.
Public systems reward what is visible, countable, announceable, photographable and attributable within a political cycle. An electric bus scores on all five. It can be counted at the depot, photographed at the inauguration and attributed to the administration that bought it. A route-selection model scores on none. It is a spreadsheet and a set of trade-offs. It cannot be inaugurated. Its benefits mature over years and disperse across passengers who will never know which methodology served them.
The two decisions also run on different clocks. A fleet can be procured, delivered and inaugurated inside a single political cycle. The value of a well-chosen route accrues over a decade, in time saved and wages kept and journeys made possible, long after the ribbon has faded from memory. A system that measures achievement in announcements will always find the first clock more useful than the second. No one has to intend any of this for it to happen.
That is the performative state: not a state that necessarily fails to act, but one whose incentives reward what can be displayed before what can be demonstrated. It can buy the visible green asset and announce it truthfully, while leaving unanswered whether the route reduces the emissions of the country’s trade-exposed productive economy, and while the deeper decarbonisation burden stays where it has always been: with the suppliers.
The consequence is structural rather than conspiratorial. The procurement can produce the photograph before the allocation logic has been made public, and far fewer visible incentives push the other way. This is the precise mechanism through which populist incentives meet sustainability. Populism does not always block sustainability. Sometimes it redirects it toward the camera. The asset is real, the technology is sound, the announcement is accurate, and the deepest question, where the public value of the asset will land, goes unexamined because the incentives to examine it publicly are comparatively weak.
A green asset can be technically correct and still be strategically misallocated. Nothing about the correctness protects against the misallocation. Only the allocation method does, and the allocation method is the one part of the programme nobody photographs.
Figure 5 · The Same Bus, Different Public Value
A public asset needs a public allocation method
Route Value Test diagram showing the same neutral electric bus feeding two illustrative candidate routes, a general urban route and an industrial-worker route, both assessed through eight shared public-value criteria.
Neutral public asset
Illustrative candidate: general urban route
Illustrative candidate: industrial-worker route
Common Route Value Test
Mobility need
Equity and safety
Network integration
Utilisation
Emissions displacement
Economic access
Industrial decarbonisation leverage
Data confidence
Industrial-worker corridors are candidates for assessment, not automatic winners. These are illustrative candidates, not an exhaustive route universe.
The public asset is the same. The route determines who benefits and how much value is created.
Make the Invisible Decision Visible
The constructive demand is modest, which is what makes it hard to refuse. Before the routes for the new fleet are finalised, publish the criteria by which they will be chosen, and assess candidate routes, industrial and otherwise, through the same public framework. Call it a Route Value Test. A Route Value Test is not a case for one destination. It is a published method for deciding where public transport produces the most public value. Eight criteria would carry it.
Mobility need. How many people currently lack reliable, affordable access?
Equity and safety. Who gains access, including women, low-income workers and underserved neighbourhoods?
Network integration. Does the route complement the BRT corridors, existing services, depots and interchange points?
Utilisation. What are the expected operating hours, load factors, directional balance and dead mileage?
Emissions displacement. Which current transport modes would the route replace, and on what electricity baseline?
Economic access. What employment, travel-time and affordability gains could the route create?
Industrial decarbonisation leverage. Could the route measurably reduce commuting emissions in trade-exposed industry, and with what confidence could that reduction be verified? This criterion earns its place beside the others; it does not outrank them.
Data confidence. For each of the above, what is measured, what is estimated and what is unknown?
Industrial-worker corridors should enter this test alongside every other candidate. They should not be excluded as a private concern. They should not be privileged because an exporter’s inventory would look better. They should be assessed, in public, against the same questions as everything else, with the answers published before the routes are locked in rather than reconstructed afterwards.
The border is the reason this should not wait for perfect data. Carbon-conditioned trade is already changing the value placed on credible emissions measurement, even where a particular sector is not yet covered. As embedded emissions become a condition of entry into export markets, the routes that move an industrial workforce become part of trade policy whether or not the transport authority thinks of them that way. A state that wants its exporters measured fairly has an interest in building what fair measurement requires: the routes, the data and the factors.
None of this begins with buying more buses. It begins with making the reasoning behind the existing investment visible, so that the second decision is made as openly as the first. A government confident in its allocation loses nothing by showing the model. A public that has been promised 450 electric buses is owed more than the buses. It is owed the reasoning.
The vehicles will arrive, and they will be electric, and that part of the story will be told well because it is easy to tell. The part that will decide what the investment was for is the map. The politics of green transport were never primarily in the battery chemistry. They are in the distribution of public value across the city, one route at a time. The route is not what happens after the policy. The route is the policy.
