THE KPI THAT COULD MAKE HOMELESSNESS DISAPPEAR (FROM THE SPREADSHEET)
Reducing the number of emergency-housing grants looks like progress on a spreadsheet. Whether it represents fewer people in crisis is another matter.
For someone turned away from emergency housing, the consequence is not an abstract “service-delivery outcome”. It can mean a car, a couch, a shelter queue, a night outside, or a scramble to keep children safe. That is the human test now sitting behind a very tidy piece of bureaucracy: Ministry of Social Development staff and managers were measured, in part, against the number of people in their region receiving emergency-housing grants.
The metric was revealed by 1News/Q+A on 28 June. Documents obtained by the programme showed “multiple tiers of MSD staff” were on performance agreements covering 11 measures, including “the number of people in their region who receive emergency housing grants”. Staff received regular grading on the measures, and managers were told that where performance did not meet expectations they would be required to develop targeted improvement plans.[1]
That is the sort of sentence that sounds harmless in Wellington and lands very differently at a Work and Income counter. Fewer emergency-housing grants can mean fewer people need emergency housing. It can also mean fewer people are approved for help. The dashboard does not, by itself, know the difference.
The policy chain begins with a real problem. Emergency housing, especially motel-based emergency housing, has been widely criticised as costly, unstable and unsafe. An MSD Emergency Housing Deep Dive report says the Government announced in April 2024 a target to reduce the number of households in emergency housing by 75 percent by 2030.[2] Ministers have argued, fairly enough, that motels are no place for people to live. The hard question is whether the machinery built to meet the target measures actual housing security, or merely a smaller grant ledger.
This week, that question broke through. OpenBrief Press Lens recorded 16 news items on “emergency housing performance targets” in the week to 4 July, from a prior average of zero, led by RNZ, the Otago Daily Times and TVNZ.[3] Parliament also took it up: Hansard records a 30 June item titled “Ministry of Social Development—Staff Performance Measured on the Number of People in Their Region who Receive Emergency Housing Grants”.[4]
MSD has not denied the existence of the metric. Graham Allpress, MSD group general manager of client delivery service, confirmed to Q+A that staff were assessed against emergency-housing targets, while saying it was “just one in a range of metrics” in performance agreements.[1] That is MSD’s central defence: this was not a lone red button labelled “deny motel”. It was one measure among many.
But performance systems rarely need red buttons. They work by nudges, rankings, expectations and conversations with managers. 1News separately reported that managers were warned of consequences, including warnings or poor appraisals, if their KPIs did not meet expectations.[5] If a regional manager knows one of the numbers being watched is the number of people receiving emergency-housing grants, the incentive is obvious enough. A bureaucrat does not need to be told “say no” to understand which way the wind is blowing.
Community agencies saw the danger immediately. Auckland City Missioner Helen Robinson told Q+A the targets were “extraordinarily difficult to swallow” and said people denied emergency housing had little other choice but to sleep on the street. “I have never seen this scale of homelessness in my 13 years,” she said.[1] Christchurch Methodist Mission’s Jill Hawkey was blunter: “I don’t think anybody should be rewarded for denying somebody their human right to shelter, and that’s essentially what’s happening.”[1]
MSD chief executive Debbie Power later defended the approach in an RNZ interview republished by Scoop. She said the ministry was implementing government policy and that the Government had set a target to reduce the number of people in emergency housing. Power said the targets were “part of our performance suite”, but insisted eligibility rules had not changed: if someone sought emergency housing and was eligible, “they would be granted it”. She also said every decline was reviewed by a manager to ensure decisions were made against policy.[6]
That assurance matters. If the legal test has not changed, eligible people should still receive help. But the public still needs to know how the pressure operates in practice. A manager reviewing a decline is a safeguard only if the review is independent, well documented, and not conducted inside the same performance culture that rewards lower numbers.
The ministerial accountability line is also messy. Associate Housing Minister Tama Potaka, whose roles include Associate Minister of Housing with responsibility for social housing, reportedly said he was not comfortable with the targets and that they were something he would ask the MSD chief executive about.[5] RNZ reported him saying it was up to MSD’s chief executive how she managed staff performance.[6] Louise Upston is listed by MSD as Minister for Social Development and Employment, while Chris Bishop is Minister of Housing and Potaka is Associate Minister of Housing.[7] The result is the familiar public-sector fog machine: ministers set the target, MSD designs the lever, and when the lever looks ugly everyone points to the level above or below.
The numbers show why this cannot be settled by rhetoric. RNZ reported that households in Emergency Housing Special Needs Grants fell from 2,880 in January 2024 to 498 in December 2024, with the figure at about 471 by May 2026.[6] That is a dramatic fall. It may include genuine exits from motel dependency. It may also include people diverted, deterred, declined, or pushed into arrangements that do not show up in the emergency-housing count. The public data cited so far does not tell us enough.
The missing information is the story. Where are the monthly applications, approvals, declines, withdrawals and repeat applications by region? What happened to people who left or were refused: public housing, transitional housing, private rentals, whānau, cars, shelters, the street, or “unknown”? Did refusal rates change after performance agreements were introduced? Did Māori, Pacific people, disabled people, families with children or people fleeing violence experience different outcomes? Without that, the Government can claim success while the rest of us are asked to admire a shrinking number in a box.
One Q+A-reported instruction adds to the concern: documents said 95 percent of people leaving emergency housing should not receive a repeat grant for at least six months.[1] Reducing repeat grants is sensible if people are moving into stable homes. It is very different if they are simply being told not to come back.
The wider policy setting makes the distinction sharper. The same weekly scan that picked up the emergency-housing KPI also found a spike in coverage of “move on orders” for rough sleepers.[8] Taken together, the risk is not hard to see: emergency accommodation becomes harder to access, visible homelessness becomes easier to move along, and the state celebrates the miracle of fewer people appearing in the category it chose to count.
There is a better scoreboard. Measure verified exits to safe, stable housing. Publish decline and review data. Track repeat need, rough sleeping, shelter demand, time in cars, and outcomes for children. Report regional and demographic patterns. Audit decisions independently. If emergency housing falls because people are safely housed, that is success. If it falls because the door got narrower, that is not solving homelessness. It is spreadsheet magic, and the person paying for it may be sleeping outside.
References
1News/Q+A — “‘Perverse incentive’: MSD staff metrics include emergency housing grants”
Ministry of Social Development — Emergency Housing Deep Dive Report, July 2024
OpenBrief Press Lens — emergency housing performance targets
1News — “Hipkins says he will scrap ‘morally bankrupt’ MSD staff metrics”
Scoop/RNZ — “MSD Boss Defends Linking Staff Performance With Cutting Emergency Housing”