The scale of the issue around direct provision is now being contested at the level of official records. John Harding, head of operations at the International Protection Accommodation Services, says claims that large numbers of people leaving Ipas accommodation are ending up homeless are resting on self-reporting rather than verified data.
Verified fact: Harding raised the issue in correspondence dated January 21st, 2026, to Rosemarie Tobin, principal officer for homelessness policy at the Department of Housing. Informed analysis: the dispute is not only about numbers, but about whether the public is being shown a complete picture of how people move from state accommodation into emergency services. The phrase direct provision sits at the center of that debate.
What is being counted, and what is not?
Harding said reports of large numbers presenting for homelessness services after leaving Ipas accommodation were based on people telling emergency services that they had come from Ipas centres. He warned that “narratives” were being created that work against, in his words, the objective of providing clear information and informed decision-making.
That matters because the current figure set is already being used in public discussion. Mary Hayes, director of the Dublin Region Homeless Executive, told the Public Accounts Committee that more than 1, 300 people over the last two years who had left direct provision sought emergency accommodation in the capital. She said that between 2020 and 2025, 1, 362 single adults entered homeless services in Dublin from Ipas centres, alongside 127 families, bringing the total to 1, 758 people.
Hayes also said leaving direct provision in the preceding six months was the largest driver of single-adult homelessness in 2024 and 2025, accounting for 25 per cent of all new single entrants. Those figures have become central to the public argument. Harding’s response is that they need verification against Ipas records before they are treated as settled fact.
Why does Ipas want the data checked?
Harding asked that all local authorities provide Ipas with details of households in this category in 2025, including the name and date of birth of the primary applicant. He said Ipas then proposed to examine its own records, validate the data, and identify trends within the cohort.
He also argued that the majority of people who left Ipas accommodation in 2025 did not use homeless services. Harding said 4, 300 people recognised as having a right to remain in Ireland left Ipas accommodation that year, and on that basis “the vast majority” appear to have sourced accommodation without emergency support.
Verified fact: the correspondence was released under Freedom of Information legislation. Informed analysis: Harding’s push for verification suggests Ipas wants to distinguish between those who truly moved straight into homelessness and those whose later housing problems may have different causes. That distinction is crucial if public policy is to target the right point of failure in the transition out of direct provision.
Who is shaping the public narrative around direct provision?
The people sounding warnings include Simon Harris, the Tánaiste, and Mary Hayes of the Dublin Region Homeless Executive. In December, Harris drew criticism after asserting that a “significant number” in emergency accommodation were immigrants who “don’t have housing right. ”
That intervention widened the political stakes around direct provision. Harding’s objection is not framed as a denial that some former residents are entering homelessness services. Instead, he questions whether the size of the problem is being overstated through incomplete or self-reported information.
Tobin replied that her staff were also considering a qualitative piece of research on households presenting as homeless from Ipas centres. She added that Harding’s suggestion for local authorities to contact Ipas to verify claims would likely require a data-sharing agreement between the Department of Justice and the local authorities.
The implications are practical as well as political. If the figures are accurate, they point to a serious failure in the move from state-supported accommodation to stable housing. If they are overstated, then public debate may be building policy pressure on the wrong foundation. Either way, the need for validated data is now part of the story.
Accountability watch: the next test is whether local authorities, the Department of Housing, and Ipas can establish a shared method for checking claims before more public narratives harden around direct provision. Without that clarity, direct provision will remain a disputed gateway between accommodation policy and homelessness policy, rather than a system whose outcomes can be measured with confidence.
