By Mijan Ahmed
A family arrives at the Bangladesh border after fleeing violence in Myanmar. Crossing is only the first uncertainty. Will they be allowed to enter? If they reach a camp, how long will it take to record their identity? Where will they sleep while they wait? If someone threatens them, who will listen—and what will happen after they speak?
For another family, the danger may begin inside the camp. A child goes missing. A woman needs help after violence at home. A man receives a threat and fears that reporting it will make him a target. Each family needs something different, but all depend on a path from danger to help. That path may involve a neighbour, a Rohingya volunteer, a humanitarian worker, the Camp-in-Charge office, police, a lawyer or a health facility. It works only if people can reach the right person, safely, and receive help before it is too late.
That path is under strain.
In the second quarter of 2026, UNHCR recorded 1,389 people affected by refoulement-related incidents at Bangladesh’s border, an increase of 83 percent from the same period a year earlier. During the same quarter, the agency warned that, without new funding, interruptions could affect registration, community protection, child protection case management, outreach and specialised support for survivors of gender-based violence.
These findings describe different problems. The rise in border incidents cannot simply be blamed on cuts to services in the camps. Nor does a warning about possible interruptions mean every listed service has already stopped. What the findings show together is serious enough: more people are encountering protection risks at a time when the systems meant to identify and assist vulnerable people face financial pressure.
That should change how we talk about humanitarian funding. Protection is often treated as a line in a budget, somewhere beside food, shelter and health care. For a person in danger, it is a sequence of human decisions. Someone must notice the risk. Someone must believe the person reporting it. Someone must know where to send them. And someone must remain responsible after the first referral.
The border is the first test
The figure of 1,389 needs careful reading. UNHCR describes people affected by recorded refoulement-related incidents. It does not say that 1,389 people were all individually deported from Bangladesh. Its monitoring may also miss incidents. Turning the figure into a claim about the exact number of people sent back would misrepresent the evidence.
But caution about the figure should not make the problem disappear. Rohingya continue to flee conflict and insecurity in Rakhine State. When people reach a border, decisions made there can determine whether they find safety or remain exposed to danger. UNHCR has linked the increase it recorded to tighter border enforcement. Its report for the first quarter also described Bangladesh’s restrictive border approach and called for stronger efforts to uphold access to safety.
Bangladesh has hosted Rohingya refugees for years, at enormous cost to communities and public services in Cox’s Bazar. The pressure is real. But so is the danger faced by people fleeing Myanmar. A border policy cannot be judged only by how many people it keeps out. It must also be judged by what happens when a frightened family says they cannot safely go back.
That calls for a process people can understand and use: an opportunity to explain their circumstances, a serious assessment of individual risk, and a way to obtain urgent help. These safeguards matter especially for children travelling without parents, people who need medical care, and those who may face particular threats if returned.
Reaching a camp is not the end of the journey
Even after arrival, safety can depend on paperwork.
During April–June 2026, UNHCR and the Bangladesh authorities biometrically identified 6,651 newly arrived Rohingya. Identification is an important step towards a Joint Assistance Card, which enables access to essential assistance and other services. Yet UNHCR reported that some Camp-in-Charge offices had suspended appointments, reduced processing or introduced their own screening requirements before people could join the identification exercise. It said these measures differed from the nationally agreed policy. The public report does not name the camps or say how many people experienced those delays.
For an official, an appointment may be one part of an administrative process. For a newly arrived family, waiting can shape every part of the day. How do they obtain assistance? How do they explain who belongs to their household? Where do they go if a child is ill or someone is threatening them?
UNHCR also reported that no formal shelters were allocated to newly arrived Rohingya during that quarter. That finding should lead to a direct question for the authorities and aid agencies: where did those families stay, and what support was available while they waited? The published figures alone cannot answer it.
Registration is sometimes discussed as if it were chiefly about managing a population. To a refugee, it can be the difference between being visible to a service provider and having to explain an emergency from the beginning each time. Delays may have consequences well beyond the registration desk.
What happens after someone reports violence?
The need for help is not hypothetical. UNHCR recorded 168 serious security incidents through its own monitoring channels in the second quarter of 2026. Seven Protection Sector partners separately recorded 470 serious incidents across the Cox’s Bazar camps and Bhasan Char. These are figures from different reporting systems; they must not be added together as if they form one total. UNHCR’s recorded incidents included abductions or kidnappings, serious assaults and killings.
A report of violence is only the beginning of a response. If a person fears retaliation, where can they speak in private? If their complaint concerns a powerful person or an armed group, who can protect them afterwards? If a case is referred to police, who will explain the process and follow it up? A family may need help immediately, then again days or weeks later.
UNHCR’s first-quarter report shows how difficult that path can be. It described limited witness protection, fear of retaliation, inconsistent police responses and obstacles to using formal justice systems. Only 27 percent of the reported cases described in that quarter’s access-to-justice section were formally registered. The same report says UNHCR and its partners BLAST and BRAC provided legal assistance to more than 12,000 refugees, including support for victims of serious violence to file police cases. Both facts matter: help exists, and important barriers remain.
The 27 percent figure is from January–March, not the second quarter. It cannot tell us what happened to every complaint later in the year. It does, however, show why counting reported incidents is not enough. A system should also be judged by whether people can move from reporting to protection, investigation and a fair outcome.
This is where an apparently small loss of capacity can matter. An outreach worker may be the person a family trusts enough to approach. A legal adviser may help turn an account of violence into a complaint that can be pursued. A caseworker may notice that a person remains at risk after the immediate emergency has passed. Without knowing exactly which posts or services have changed in each camp, we cannot claim that any one case failed because funding was cut. We can say that these functions are central to making protection work.
Women and children need more than an open office
Some people face additional barriers to seeking help. A woman experiencing violence at home may live with the person harming her. A child may be afraid to name an adult. Both need a safe way to speak, trained people who understand the risk, confidential referrals and continued support.
UNFPA reported that recorded gender-based violence cases rose by 10 percent in April–June compared with the previous quarter. It also reported a $10.7 million funding gap for the Gender-Based Violence Sub-Sector. At the same time, UNFPA-supported programmes continued to reach large numbers of women and girls with prevention and response services. The picture is therefore neither one of total collapse nor one of assured safety. Services are operating while the wider network on which they rely is under pressure.
An increase in recorded cases does not automatically mean violence increased by exactly the same amount. More outreach can help survivors come forward; reduced access can have the opposite effect. Counts from different agencies may also reflect different reporting channels. The essential question for this op-ed is what happens after someone asks for help. Can they get private case management, health care when needed, and a referral that leads somewhere?
UNFPA has described work that shows what a functioning response can look like: confidential spaces in health facilities where women and girls can receive case management, referral and care, alongside trained community volunteers supporting prevention and psychosocial activities. Such services depend on trained staff, safe spaces and links between organisations. Keeping one door open has limited value if the next service in the referral chain cannot respond.
Children, too, need support that goes beyond a single report of harm. A caseworker may need to find a separated child’s family, assess whether a home is safe or arrange care over time. UNHCR has specifically warned that child protection case management and outreach could face interruptions without fresh funding. The risk is not only that fewer cases will be recorded. It is that children already known to be at risk may lose the steady attention their cases require.
What does a funding cut remove?
The revised 2026 humanitarian plan reflects a response trying to cover urgent needs with less money. The UN’s overall $710.5 million appeal was 26 percent lower than the 2025 appeal, even as it sought to support refugees and affected Bangladeshi host communities. According to UNHCR’s second-quarter reporting, planned protection-programming costs were 23 percent lower than in 2025. These are comparisons of plans and budgets. They do not mean that precisely 23 percent of protection services have already shut down.
That distinction is important because a useful public debate needs more than a large percentage. We need to know what the shortfall means at camp level. Are registration appointments taking longer? Are outreach workers covering more households? Is a specialised caseworker still available when a child needs one? How long does a survivor wait for a referral? Which services have reduced hours, and which are at risk but still running?
The Protection Sector publishes lists of available services by camp, including an August 2026 listing. These are valuable starting points. But a service appearing on a list does not tell a refugee whether someone will be there when they arrive, whether an interpreter is available, or how quickly a difficult case will be followed up. Those questions require direct checks with service providers and the people trying to use them.
It would also be unfair to place every failure at the door of aid workers. Refugee protection in Bangladesh involves authorities, police, humanitarian organisations, community workers and donors. A caseworker cannot replace a police investigation. A donor cannot make a border decision. A refugee volunteer should not be left to manage a threat that requires professional and official action. Each part of the system has its own responsibility.
Make the route to help visible
Bangladesh’s authorities and humanitarian organisations should give refugees clear, current information about where to seek help in each camp, including after ordinary working hours. People should know how to report a threat privately, where to obtain legal or medical assistance, and whom to contact if a referral does not lead to help. Information must be understandable to the people using it, including new arrivals.
The authorities should explain how people fleeing danger are assessed at the border and how delays in biometric identification are being addressed. UNHCR and Protection Sector partners should state, as precisely as safety permits, which services remain available, which have been reduced, and where funding gaps may cause a break in care. Donors should ask whether support reaches people through the whole response—from first contact to follow-up—rather than counting only how many cases were opened.
Rohingya refugees must also be able to describe what happens when they seek assistance. Their accounts cannot be reduced to a satisfaction figure. A mother who tried to report her child missing, a survivor who sought confidential care, or a newly arrived family waiting for identification can show where written procedures meet daily life. Those accounts must be gathered with consent and without exposing people to further danger.
We already know enough to ask these questions urgently. The second-quarter figures point to risks at the border and serious incidents in the camps. The funding warnings point to possible gaps in the services people depend on. They do not, by themselves, prove exactly which individual has been denied help or why.
That is the next task for reporting—and the immediate task for those responsible for protection. A refugee facing danger should not have to understand an organisation chart, chase a referral across the camp or risk retaliation simply to ask for help.
Protection succeeds when that person reaches someone who can act, and when help continues after the first conversation. At a time of rising pressure, that is the standard by which the system should be judged.
Mijan Ahmed is a Researcher and Writer.
