Have Your Say

Working for Families at 20: Inside the creation of the biggest welfare reform of the 21st Century

Thomas Coughlan
Political Editor·NZ Herald·
31 Dec, 2025 07:41 AM15 mins to read

Finance Minister Michael Cullen talks with Prime Minister Helen Clark before delivering his final Budget. Photo / NZPA
More than 300,000 families — about half of all families with children — get support from Working for Families, a Helen Clark-era tax credit system that boosts family incomes by an average of about $178 a week. Despite being described by former Prime Minister John Key as “communism by stealth”, it is now an enduring part of the country’s social safety net. As the main part of the scheme turned 20 this year, the Herald looked back on how it came to be.
The two most consequential changes to our welfare system in the last half-century are connected by a handful of people and two whiteboards.
Looking back two decades after the second of these changes, Working for Families, began paying out in April 2005, Peter Hughes, then the chief executive of the new Ministry for Social Development (MSD), told the Herald that for him, the genesis of the tax credit scheme began decades earlier in the early 1990s.
Hughes, who retired from the public service last year after rising to the top job of Public Service Commissioner, was then the acting director of policy at the old Department of Social Welfare, a precursor to MSD.
The Government was planning to cut benefit rates, and Hughes and some other public servants were called over to ministers’ offices to work out the details.

“I can remember being called over to Ruth Richardson’s office – I’m pretty sure it was her office – late at night with Minister Jenny Shipley and she had a whiteboard… with new benefit rates on it,” Hughes said.
Initially, Hughes recalled, he and the other officials were given just 24 hours to make the new benefit rates “happen” in policy terms.

The problem for Hughes was not just the speed of the decision; it was that the new benefits were being made with very little information about whether they would be enough for people to live on.

“There was very little information on adequacy of benefit income – very little at all.
“Benefit rates in those days were much higher relative to wages than they have been since.
“People didn’t see a need to generate that information so we didn’t have it. So it was quite hard doing that work,” he said.
Hughes said he remembered coming back from Richardson’s office and discussing with colleagues “whether we were going to do this or resign”.
“It was quite shocking, the rates that were proposed,” he said.
Shipley, the minister responsible for social welfare, got the team more time. Hughes did the work. He recalls now that the rates ultimately agreed by Cabinet were even lower than the ones he first saw on the whiteboard.
A decade and a half in the future, Hughes would find himself heading up the new Ministry for Social Development, as a new Labour Government planned the biggest intervention in social policy since Richardson’s cuts: Working for Families.
Enter the second whiteboard.

It was a little over a decade later, and a group of public servants was crowded round another whiteboard. This time, Hughes wanted to make sure they’d have better information on income adequacy than he’d had in 1990.

Full employment wasn’t enough
About halfway through her term, Prime Minister Helen Clark had a frustrating problem.
The economy was running well by some measures; New Zealand was one of the fastest-growing economies in the OECD in 2002, growing 4.4% thanks to an export boom and strong domestic demand. Unemployment, at 5.2% — slightly below where it is now — seems high by today’s standards, but Clark recalls it as being quite low, which it was, when compared with the 1990s, when the unemployment rate reached double digits.
Unemployment fell even lower than forecast in 2002 and 2003, plunging to below 4%, where it stayed for most of the rest of Clark’s time in government.

Clark said by the time the Government got to Working for Families, “it was clear that full employment on its own was not going to overcome poverty challenges”.
She recalled to the Herald, “New Zealand has relatively low wages and high housing costs”.
“Families in those circumstances easily fall into poverty, including the working poor,” Clark said.
At some point, early in Clark’s second term, Social Development Minister Steve Maharey came to her and Finance Minister Michael Cullen “with the idea that we should look at cash transfers being stepped up systematically” to address the problem,” Clark recalled.
“That was the genesis of it – full employment on its own in the context of relatively low wages and high housing costs would not address the problem and we had to do something more,” Clark said.
Maharey told the Herald the Government had inherited public concerns about poverty and income adequacy from the 1990s.
During the Government’s first term, ministers talked about “these issues coming into a single package”, but ultimately scrapped the idea because the Government was locked in a long battle with Christine Rankin, the head of what was then the Department of Work and Income.
Maharey also knew that the cost of significant social reform would be high, and he would need a powerful agency to fight his corner.

‘Social Treasury’ versus the actual Treasury
In October 2001, the Government merged the Ministry of Social Policy and the Department of Work and Income — the policy and delivery arms of social welfare — into the Ministry of Social Development.
Maharey said he wanted a “social Treasury” that could stand its own against the real Treasury.
MSD began publishing the “social report”, an annual inventory of social trends, which looked at things like smoking rates and income adequacy alongside broader measurements of a healthy society like voter turnout and “participation in cultural and arts activities”.
The report is not unlike a larger broader version of the Wellbeing Report included in Grant Robertson’s wellbeing budgets.
Nonetheless, the actual Treasury was not convinced.
“Treasury did not like this policy at all,” Maharey said.
“This was a break with the orthodoxy – you should not be doing redistribution of income,” he said.
“Treasury always had neoliberal tendencies, shall we say,” Clark said.
Once it made its way around the Cabinet, “everyone got excited about it”, including Clark herself, Maharey recalled.

But in many ways, the key person to win over was Cullen, the gatekeeper to the large sums of money that would be required to fund the scheme.
“Despite the fact we’d built a Social Treasury, Treasury still holds a lot of sway and it did not like this policy,” Maharey said.
Maharey said Cullen became “interested later on when the package was largely developed”.
‘He became interested in the economics of it,”
With both Cullen and Clark onside, Maharey put the Working for Families package together.

Whiteboard number 2
Over at MSD, Hughes had a lot to prove. He was a new chief executive of a new department — and his minister, Maharey, had put a lot on the line for the scheme.
There was “a lot riding on it”, he recalled.
“This was the first big test for Minister Maharey’s ‘Social Treasury’ vision”.
The merger had been challenging. The policy team and the delivery teams did not “get on”.
He wanted the delivery side of MSD to be involved in the policy design from the beginning, so the agency would have a better idea of whether the policy was practical.
The practicalities of the scheme were worked out using the same technology with which Richardson and Shipley had concocted the 1990s benefit cuts: a whiteboard.
Hughes recalled “about 30” people from the delivery and policy arms of the agency crowded around a whiteboard to build the scheme.
Initially, it was “complete chaos”, but it later “found a rhythm”. Hughes reckoned that eventually it made for a better department, which was better both at policy development and delivery.
When he went into MSD, Hughes said he had his experience with the whiteboard in Richardson’s office “in the back of [his] mind”.
“One of the things I did was a lot of work on poverty measurement and living standards”.
“At a really basic level it comes down to adequacy of benefit income on one hand and wage-benefit adequacy on the other,” he said.
In order to answer Hughes’ question, the department needed to know what it didn’t know when the benefit cuts were made in the 1990s: how much do people need to avoid poverty?

Measuring poverty
As the 1990s wore on, it became obvious that New Zealand had a poverty problem, and it was getting bigger.
However, beyond what was visible on the street, proving and measuring poverty was difficult. New Zealand had been a laggard when it came to measuring poverty rates. Prior to the 1990s, poverty rates were guessed at with small community studies and were often dismissed by the Government as anecdotal — in some cases, because they were.
Poverty isn’t easy to measure. Knowing how many people don’t have enough means first finding what “enough” actually is, and then finding a way to work out how many people fall below that threshold.
In 1992, that all changed with the launch of the New Zealand Poverty Measurement Project (NZPMP) by Charles Waldegrave of the Family Centre Social Policy Research, Bob Stephens of Victoria University’s School of Government, and Paul Frater of economics consultancy BERL.

In 1995, the group launched its child poverty indicators. The indicators followed an international consensus that poverty could be measured by looking at people’s incomes relative to the country’s median income level. The group also took into account housing costs, as the high cost of housing is a key driver of New Zealand poverty rates. This was relatively innovative, with only the United Kingdom at that time, including housing costs in poverty measurements.
Waldegrave’s work was unpopular with the National Government, but was welcomed when the new Labour Government took office in 1999. It provided much of the evidentiary basis for Working for Families, which, at its heart, is about increasing household incomes through tax credits. The creators of Working for Families still credit Waldegrave with providing the evidentiary backdrop for the system.
The NZPMP measurements have had a long life in New Zealand. Prime Minister Jacinda Ardern included seven of them in the ten measurements included in her Child Poverty Reduction Act. They are still reported each year in the Child Poverty Report.

How it worked
In practical terms, the idea of Working for Families owed a lot to the United Kingdom, where Chancellor Gordon Brown made a Working Families Tax Credit a centrepiece of his 1998 Budget (Maharey says the name Working for Families came from a staffer at MSD, possibly inspired by the British antecedent).
Maharey also brought a senior analyst from the British civil service over to advise officials here.
“It made good policy sense – I wish people would do this [take foreign expertise] more often… So many policies I see look like they’re really not well done because they’re not drawing on expertise like they ought to,” he said.
In December 2003, the Government’s Budget Policy Statement signalled a “Future Directions” package which would be at the “centrepiece” of the Budget, which would come in May the next year.
Unlike today’s Budget Policy Statements, which tend to be very light on what is in the Budget, Cullen was fairly explicit about what he was going to announce.
“It will significantly increase direct income support and incentives to move from welfare benefits into paid employment, and will make housing more affordable for low-income families and single adults,” he wrote.
On 27 May 2004, Budget Day, Cullen revealed the details: a series of tax credits that would roll out from 2004 to 2007, costing about $1.1b a year ($1.7b in today’s prices) once in force.
The scheme was less revolutionary for what it did than it was for the scale on which it did it.
A income-boosting family tax credit scheme wasn’t entirely new; the main tax credits already existed. What Working for Families did was immensely increase their eligibility and the amount the credits paid.
That came at significant cost.
On its own, the scheme was about two-thirds the size of 2024’s tax cut package.
However, because Working for Families was targetted only at low-and middle-income families with children, the income boost to eligible households was larger. Initial estimates reckoned about 300,000 families would get the tax credits. Some families saw their household incomes increase by $270 a week in today’s money.
In October 2004, Family Assistance abatement thresholds were increased, allowing people receiving those tax credits to keep more of what they earned as they took on additional work.
In April 2005, Family Support, a Lange Government-era means-tested tax credit paid to low-income families, both in work and on a benefit, was increased by $25 a week for a family’s first child and $15 a week for each subsequent child.
In 2007, the credit was given its modern name, the “Family Tax Credit” (FTC).
The accommodation supplement, a benefit designed to help with high housing costs, was increased in high-income areas.
In April 2006, the Child Tax Credit, a Jim Bolger Government tax credit that topped up the income of families that worked, was scrapped and replaced with the In-Work Tax Credit (IWTC). It paid $60 a week for families with three children and $15 a week for every other child.
Both the FTC and the IWTC would begin abating from a new, higher threshold of $35,000 (abatement means that once a family’s income reaches a certain level, the tax credit starts to reduce). This replaced the two lower thresholds of $20,356 and $27,481, and had the effect of massively increasing the amount of tax credit families would keep as they earned more money.
A new abatement rate of 20% was introduced, meaning that for every dollar earned above the abatement threshold, the tax credit was reduced by 20 cents. The higher abatement rate had been 30%, meaning under the old system families lost their tax credits faster, giving people less of an incentive to get extra work.
In 2007, the final elements of the scheme were put in place. Family Tax Credit rates were increased by $10 a week.
Inflation adjustments were included, with the tax credit amounts and the income thresholds for the people getting them adjusted regularly to maintain their real value.
The sums of money were vast.
In 2004, the median before-tax household income was $46,000.
A low-income household with two children earning $30,000 a year would be better off by $115 a week by the time the full package rolled out in 2007. A family earning $37,500 with childcare costs would be better off by $163 a week.
Across the 300,000 households receiving the credits, the average gain was $66 a week, with households earning between $25,000 and $45,000 getting $100 a week.

What happened?
The package was incredibly successful at bringing down rates of child poverty initially.
A 2010 MSD report found that between 2004 and 2008 the number of children living in poverty by an income measure fell by 8 percentage points due to Working for Families. By another, non-income measure, it fell by 11 percentage points over the same period.
Another study, focused on poverty rates in ‘working’ households, found the child poverty rate in 2007 had halved from where it was in 2004, to 11%.
In subsequent years, the system was pared back to rein in its cost.
During the Global Financial Crisis, Treasury warned that the Government faced a severe structural deficit, something not foreseen during the late 2000s, when the fiscal picture looked far more secure. The scheme was actually expanded in response to the GFC in order to give families out of work some security of income. But as projections of the fiscal situation worsened, the scheme was pared back.
Abatement thresholds had been indexed to inflation, which meant that the credits remained targeted to the same cohort of people, even as wage inflation gave them nominal pay increases. This was removed to cut costs, and meant that the number of families eligible for the credits slowly reduced.
There was an ideological element too. Scrapping inflation indexing meant many upper-middle-income earners would earn themselves out of the scheme, which meant that over time it would become more targeted to lower-income earners (and cheaper too).
National and Labour Governments have both increased abatement rates too, meaning families lose their tax credits much faster.
Twenty years on the promise and peril of Working for Families endures. With nearly half of all families with children receiving at least some of the tax credits, Former Prime Minister John Key’s allegation that it is “communism by stealth” isn’t entirely unfair.
For the left, it remains one of the key levers of poverty reduction, particularly among the working poor.

Note there are pics that are in the original article that are not reproduced.