The 2026 Census: From Counting People to Linking Their Records—Where Is It Headed?

S. Rouzbeh – The second stage of the 2026 General Population and Housing Census is scheduled to begin on 1 Aban, corresponding to 23 October 2026. But this time, the census began long before officials started visiting people: data held by government agencies are being processed and matched, and, unlike traditional censuses, the main method relies on existing administrative records held by various agencies. Gholamreza Goodarzi, head of the Statistical Centre of Iran, has said that more than 27 government agencies are taking part in the process, and that people’s self-reporting is no longer the principal basis of the census.

Before it begins, a simple question arises: if much of the information is no longer collected directly from people, who decides what is known about them, where that information is gathered from, and who is allowed to access it?

The principle of conducting a census is not in question. No country can plan for housing, education, healthcare, employment, migration and future infrastructure without accurate statistics. A register-based census is not an unfamiliar method, either. Several European countries use it today. But the experience of those very countries shows that the success of such a system depends on the quality of the records, data-protection laws and public trust. The United Nations Economic Commission for Europe (UNECE) also identifies public mistrust of linking information from multiple sources as one of the challenges of this method.

The question for Iran, then, is not why this method was chosen; it is what information is being collected, how it is being linked, and who will be able to access it.

During the pilot census in Sabzevar, forms containing 56 data items were discussed, 45 of which could be directly compared with data in administrative records. Another report on the same project gave the figures as 48 control items and 50 field items. Earlier, the deputy head of the Statistical Centre had said that the databases of 13 government agencies had been examined and that 46 data items could be extracted from their administrative records. There may be a technical explanation for these differing figures; but the discrepancy raises a simple question: why has the precise, final list of these items, together with a clear explanation of the differences, not been made available to the public before the census begins?

In a traditional census, the questionnaire was placed in front of the citizen. At least they could see what they were being asked. Under the new method, many of these questions may not be asked at all, because the information already exists in administrative systems.

This in itself is not necessarily a problem. The government needs data to run the country. The Civil Registration Organisation needs to hold population information. The concern begins when scattered data are brought together and create a far more extensive picture of a person’s life. Another question is how far these data will be accessible to security forces.

To understand the scale of this, it is enough to consider what officials at the Ministry of Health have said: population information on around 86 million people is held in the Ministry’s databases, and some of the data required by the Statistical Centre could be supplied from these databases. The Ministry of Health has also emphasised that information about people’s illnesses and other confidential health details will not be shared with the Statistical Centre, and that it has not been asked for such information.

If information about identity, place of residence, property ownership, education, insurance, employment and family is linked through a single identifier, we are no longer dealing with several separate pieces of data. The central question becomes: who is allowed to see this more complete picture, under what authority, and for what purpose? In my view, one reason may be people’s lack of trust in the government when completing paper questionnaires.

Officials at the Statistical Centre say census information will be confidential. Such an assurance matters, but is it enough to allay concerns?

The reason for asking this is not merely political mistrust. In a March 2024 report, Amnesty International reported that vehicle number plates had been recorded through surveillance cameras or reports from officers, and that in some cases vehicle owners had received warning messages or orders to hand over their cars. Meanwhile, Citizen Lab, a University of Toronto research centre focused on digital security and surveillance, examined an interface called SIAM (Siam) in internal documents concerning Iran’s telecommunications infrastructure. According to those documents, the interface was used to search subscribers’ information and link identity data—such as names, telephone numbers, passport numbers, postcodes and addresses—with telecommunications data. This evidence proves that the data from the 2026 census will also be used in this way; but it does explain why simply hearing the word ‘confidential’ may not be enough, or currently acceptable, for some in society.

The recent ‘Janfada’ initiative raises a similar question, although it is unrelated to the census. Iran International has reported that some citizens received text messages about the start of ‘military and emergency-response training’ courses without having registered beforehand. It is still unclear how these people’s telephone numbers entered the system. What connects this example to the census debate is not the project itself, but the question of where personal data come from and how they are transferred: if a person does not know how and under what authority their number was entered into a system, how can they be expected to trust, without question, a much broader project involving the linking of data?

I mention religion and religious denomination here only as an example: an entirely personal, individual and ideological matter that shows just how sensitive the collection of certain information can be. Previous censuses have asked about religion, but the official categories have not necessarily reflected the full diversity of belief in Iranian society. Shia and Sunni Muslims have been placed under the broad heading of ‘Muslim’, while groups such as the Bahá’ís, Yarsanis and Mandaeans, as well as people without religious beliefs, may not have a clear place in the official categories. The issue here is not only confidentiality; the statistics themselves may also present an incomplete picture of society.

Years ago, I took someone close to me to hospital in an emergency. On a form they gave me to complete the patient’s details, they had asked for the patient’s religion and denomination. I said: ‘Is the patient’s treatment based on their religion and denomination? If medical care is being provided to a human being, why is it necessary to know their religion?’ The staff member in charge thought for a few moments and said: ‘You ask too many questions; this form has already been printed—go and ask them.’ That answer mattered more to me than the question itself: information was being collected, but even the person completing the form did not know why it was necessary.

The problem is not limited to deliberate misuse, either. A simple administrative error can persist for years: a person’s address may have changed, property may have been transferred, or information in two systems may not match. Linking databases can turn the same old error into something that looks ‘official’. That is why citizens must be able to find out what data have been recorded about them and have a clear way to correct them.

This issue is directly connected to public trust. The less people trust the way their information is used, the less incentive they have to update their data, and incomplete information in turn reduces the quality of the census. Public trust is therefore not a side issue; it is part of the credibility of the statistics themselves.

The Statistical Centre can ask people to co-operate, but trust cannot be demanded. Trust takes shape when it is clear, before the work begins, what information will be collected, which agencies it will come from, who will have access to the microdata and how long it will be kept. If citizens have no confidence in the government, the outcome will be negative and hundreds of questions will remain unanswered.

“Iran has numerous regulations on confidentiality and the protection of personal information, and the Statistical Centre Act also prohibits the use of census data by judicial, administrative and tax authorities. But the central question is: which independent body will oversee compliance with these restrictions, and how can citizens defend their rights if their information is used for purposes other than those declared?”

If the 2026 census is genuinely intended to understand society accurately and plan the country’s future, these questions must not be answered only after it has been carried out.