IN Balochistan, during a provincial food security initiative, a seven-year-old girl was given a food box at school.

Instead of opening it, she placed it in her bag and said she would take it home for her younger sister who had nothing to eat.

This single moment captures the human dimensions of food insecurity better than many policy statements.

It also captures an important policy truth: food insecurity is not only about production.

It is about how climate stress, weak governance and unequal access to support are experienced across households.

Nowadays, Pakistan’s food security challenge is often framed as a climate problem. That is only partly correct.

Climate change is intensifying stress across agriculture, but the scale of the crisis is being shaped by weak institutions, policy incoherence and a political economy that continues to reward short-term extraction over long-term resilience.

This challenge cuts across the full agricultural landscape.

It affects canal-irrigated farming systems in Punjab and Sindh, riverine and deltaic zones as well as rain-fed areas in Potohar and Khyber Pakhtunkhwa (KP), dryland systems in Balochistan and interior Sindh, and glacier-fed farming systems in the northern area.

Climate change is functioning as a stress multiplier within an already unequal agrarian structure.

Smallholders, tenants and women farmers remain more exposed because they have fewer assets, weaker access to formal services and less room to absorb losses.

When a crop fails or livestock die, they are more likely to rely on debt, distress sales or reduced consumption.

Adaptation policy that does not address these structural inequalities may improve selected practices, but it will not reduce vulnerability at scale.

National policy responses only result in enhanced local risks, warns Imran Saqib Khalid

Water governance sits at the centre of this problem.

Yet, irrigation remains inefficient, groundwater extraction is weakly regulated, and politically difficult reforms are repeatedly postponed.

This is not because the problem is poorly understood.

The state is aware of canal losses, wasteful irrigation practices and pressure on aquifers.

What it has not done is to align policy incentives with hydrological reality.

We continue to protect production systems that are inconsistent with water availability.

Sugarcane is the clearest example. The point is not to blame farmers for growing what the system rewards.

It is to recognise that the state continues to support patterns of production that deepen water stress, and then presents the consequences as unavoidable scarcity.

Unfortunately, the same problem appears in the debate on crop diversification.

This is why climate-smart agriculture is often misunderstood in Pakistan.

It is treated as a collection of practices rather than as a restructuring of the agricultural policy environment.

Improved seed, better irrigation, conservation agriculture, agroforestry, crop rotation and integrated pest management are all important.

But adoption depends on whether farmers are able to take on the risk of change.

Where water is underpriced, markets are distorted and extension is weak, climate-smart agriculture remains confined to pilot projects and donor-supported demonstrations.

Climate-smart agriculture works when institutions lower the cost and risk of transition.

The lesson for Pakistan is not that foreign models should be copied directly.

It is that adaptation cannot be built on fragmented governance and policy formulation alone.

In Pakistan, indigenous systems of water and land management have often been more closely aligned with ecological limits than the policies that replaced them.

The karez systems of Balochistan are one example. These were not simply traditional technologies.

They reflected a collective logic of groundwater use that limited over-extraction and distributed access through shared management.

Their decline was the outcome of policy choices that encouraged private pumping.

A similar lesson can be found in mountain areas and rain-fed zones.

In the northern valleys, local water-sharing systems, such as kuhls in Gilgit Baltistan (GB), built around community rules and labour, have long helped manage scarcity.

In barani areas, farmers continue to rely on fine-grained knowledge of soils, slope, moisture and local weather signals when making decisions about sowing and inputs.

Such knowledge is not a substitute for science, but neither should it be ignored.

Climate-smart agricultural policy should build on practices that already work under local conditions.

This brings the discussion directly to agricultural extension. If there is one institution that should anchor climate adaptation in agriculture, it is the extension system.

That is where scientific knowledge, local conditions and farmer decision-making should meet.

Yet, extension in Pakistan has been allowed to weaken to the point that it often cannot perform this role effectively.

Departments are under-resourced, staffing is uneven, field outreach is limited, and institutional priorities are often administrative rather than problem-solving.

This is one of the main reasons adaptation remains thin on the ground.

Pakistan has treated extension as a residual service when it should have been treated as core climate paradigm.

A more serious adaptation strategy would treat universities and research bodies as operational partners in resilience planning, not simply as producers of technical outputs.

Climate adaptation in agriculture requires working links between research, irrigation, extension, local government, disaster risk management and rural finance.

Yet, these sectors continue to operate on separate administrative tracks.

Coordination is often discussed, but rarely institutionalised. Local government is the weakest link in this chain since climate impacts are experienced locally.

Federal government or provincial departments cannot manage local micro-level tasks effectively from a distance.

However, local governments continue to be treated as political orphans. This has direct consequences for equity.

The small farmer must remain at the centre of policy design. Climate-smart agriculture that is accessible mainly to larger landholders will not resolve food insecurity.

A policy framework that claims to build resilience while reproducing exclusion is not adaptation. It is selective delivery.

Pakistan already has many of the elements needed for a stronger approach. Farmers are observing climatic shifts and adjusting practices. Researchers are generating useful knowledge.

Some provincial and local programmes have produced practical lessons.

What is missing is not awareness. It is the willingness to confront politically protected inefficiencies, invest in public delivery systems, and devolve real authority to local institutions.

A more serious agenda would begin with five changes. Water governance must be treated as a political reform issue, not only a technical one.

Cropping incentives must be aligned with ecological limits, especially in water-stressed zones. Agricultural extension must be rebuilt as a core climate institution.

Universities and research bodies must be integrated into local and provincial adaptation systems.

And local governments must be given stable authority, resources and clear mandates to act as frontline institutions of resilience.

This approach should underline how Pakistan governs and associates itself with agriculture.

The country needs a shift from paper-adaptation to on-ground, governed adaptation.

Until that happens, climate-smart agriculture will remain narrower than the challenge it is meant to address, and food insecurity will continue to deepen through a system that still confuses policy presence with policy performance.

The cost of that failure is visible not only in damaged crops, weak irrigation systems and falling water tables.

It is also visible in the quiet decision of a seven-year-old child who takes her school food box home because her younger sister has nothing to eat.

The writer works on the intersection of climate, water, ecology and society. He can be reached at: imranskhalid@gmail.com