Reference: 20250878
4 February 2026
Hayden
[FYI request #33291 email]
Dear Hayden
Thank you for your Official Information Act request, received on 16 December 2025.
You requested the following:
Thank you for your response to OIA request 20250861 dated 15 December 2025.
I note your statement that no documents exist within the scope of that request.
Given Treasury's core functions in fiscal planning and policy analysis, I am
making this refined request focusing on baseline statistical values and
methodological documentation that Treasury must maintain to fulfill its statutory
functions.
1. STATISTICAL LIFE VALUE (VSL) - CURRENT PARAMETERS
The Ministry of Transport (2021) states that the Value of Statistical Life is
"estimated at $4.9 million per fatality in June 2021 dollars."[1] This value is
incorporated into economic analyses including suicide cost studies.[2] The
Treasury's CBAx tool contains a database of impact values for cost-benefit
analysis.[3]
Please provide:
1.1 The current Value of Statistical Life (VSL) figure used by Treasury in cost-
benefit analysis (as of December 2025), including:
The methodology used to calculate or update this figure
The year this figure was last updated
Any guidance documents on when/how agencies should apply this VSL
1.2 The age-adjusted VSL calculations, specifically:
VSL values used for deaths at ages 15-24
VSL values used for deaths at ages 25-34
The methodology for age adjustment
1.3 Annual mortality figures used in Treasury's fiscal baseline projections for:
1 The Terrace
PO Box 3724
Wellington 6140
New Zealand
tel. +64-4-472-2733
https://treasury.govt.nz
Suicide deaths (by age cohort: 15-24, 25-34, 35-44, 45-54, 55-64)
Expected deaths from all causes for the same age cohorts
The time period these projections cover (e.g., 10-year forward estimates)
2. FISCAL BASELINE ASSUMPTIONS
2.1 Documentation showing what mortality and morbidity assumptions are
incorporated into Treasury's long-term fiscal projections, specifically:
Whether suicide rates are assumed to remain constant, increase, or decrease in
fiscal models
Whether disability and mental health service demand is modeled to change over
time
2.2 The cost categories Treasury includes when calculating the fiscal impact of
premature death, including whether the following are factored in:
Lost tax revenue (income tax, GST)
Welfare payments to surviving dependents
Healthcare costs prior to death
Coronial and justice system costs
This is not a request for complete analysis - simply confirmation of which cost
categories Treasury considers when evaluating the fiscal impact of mortality.
3. METHODOLOGY DOCUMENTATION
3.1 Any Treasury guidance, frameworks, or standard operating procedures that
outline:
How to conduct cost-benefit analysis for social policy interventions
When agencies should use the VSL in policy analysis
How to calculate the fiscal impact of policy changes affecting mortality or
morbidity
3.2 Confirmation of whether Treasury uses the TAWA (Tax and Welfare Analysis)
model mentioned in my previous request, and if so:
Basic documentation of what the model does
Whether the model can analyze distributional impacts across population cohorts
4. CLARIFICATION REQUEST
Given that Treasury's response stated that no documents exist for any part of my
previous comprehensive request, please clarify:
4.1 Does Treasury conduct any economic efficiency analysis of social welfare
policy?
2
4.2 Does Treasury conduct any cost-benefit analysis when advising Ministers on
policy changes affecting welfare, mental health, or justice expenditure?
4.3 If the answer to either question is "yes," please provide:
Examples of such analyses completed in the last 3 years (titles and dates only)
The framework or methodology documents that guide such analyses
If the answer is "no," please confirm in writing that Treasury does not conduct
economic efficiency analysis of social policy.
5. FISCAL POLICY AND SOCIAL HARM - IMPACT MODELING
5.1 Does Treasury model the health and social impacts of fiscal policy decisions,
specifically:
When advising on interest rate impacts, benefit cuts, or fiscal tightening, does
Treasury assess potential increases in:
Suicide and self-harm rates
Mental health service demand
Family violence and social disorder
Food insecurity and housing stress
5.2 When the Reserve Bank lifts interest rates (OCR increases), please provide:
Any Treasury analysis of observed correlations between OCR changes and:
Suicide rates
Mental health hospitalisations
Domestic violence reports
Financial hardship-related welfare applications
5.3 When Treasury advises Ministers on spending constraints or fiscal
consolidation, please confirm:
Whether Treasury models the fiscal costs of increased social harm resulting from
spending cuts
Whether cost-benefit analysis of spending reductions includes the value of
statistical lives affected by reduced social services
Any guidance on how to account for second-order health and social impacts of
fiscal policy
This is not requesting policy advice - simply confirmation of whether these impact
pathways are modeled in Treasury's analytical frameworks.
3
6. PROJECTION VALIDITY AND FEEDBACK EFFECTS
This section addresses a fundamental methodological question about Treasury's
fiscal projections:
6.1 Self-Invalidating Projections:
Treasury's fiscal advice influences government spending decisions. If Treasury's
advice leads to policy changes that increase social harm (suicide, mental health
crises, crime), this creates fiscal costs that were not included in the original
projections.
Please provide:
Any Treasury analysis of whether fiscal consolidation advice has historically led
to increased social costs that exceeded projected savings
Documentation of how Treasury accounts for "policy-induced harm" in long-term
fiscal projections
Whether Treasury has assessed the accuracy of past fiscal projections where
social spending was reduced
Example scenario: If Treasury advises reducing mental health spending by
$100M, but this leads to increased suicide (each life = $4.969M VSL), increased
ACC claims, increased justice costs, and lost tax revenue - does Treasury model
whether the total fiscal cost exceeds the $100M saved?
6.2 Baseline Assumption Validity:
Treasury's long-term fiscal projections assume certain baseline rates of suicide,
welfare dependency, mental health service demand, and crime. Please confirm:
Whether these baselines assume government policy settings remain constant
How Treasury adjusts projections when policy settings change (e.g., benefit cuts,
service reductions)
Whether Treasury has identified cases where policy changes invalidated previous
fiscal projections by changing the underlying social harm baseline
6.3 Cost-Shifting vs. Cost-Saving:
Please provide any Treasury analysis distinguishing between:
True cost savings (reducing expenditure with no negative social impact)
Cost-shifting
Cost-amplifying
Specific question: Has Treasury identified any historical cases where reducing
preventive social spending led to increased remedial spending that exceeded the
4
original savings?
7. METHODOLOGICAL TRANSPARENCY
7.1 Please provide Treasury's position on the following methodological question:
"If Treasury does not model the social harm impacts of fiscal policy advice, and if
fiscal policy can influence suicide rates, mental health outcomes, and crime -
does this mean Treasury's long-term fiscal projections are systematically
underestimating future social costs?"
7.2 If Treasury does not currently model policy-induced social harm, please
confirm:
Whether Treasury acknowledges this as a gap in current methodological
frameworks
Whether there are plans to incorporate social harm modelling into fiscal analysis
What barriers exist to including these impact pathways in cost-benefit analysis
RATIONALE FOR THIS REQUEST:
This request addresses a fundamental question about the validity of Treasury's
fiscal projections: Can fiscal projections be accurate if they don't model the social
harm impacts of the fiscal policies they recommend?
The concern is this potential feedback loop:
Treasury provides fiscal advice based on projections that assume stable social
harm baselines
Government implements fiscal consolidation (benefit cuts, service reductions)
based on this advice
Social harm increases (suicide, mental health crises, family violence, crime)
This generates fiscal costs (VSL losses, healthcare, justice, welfare, lost
productivity) not included in original projections
Treasury's projections become self-invalidating because the policy changes they
recommended altered the baseline assumptions
If Treasury does not model these feedback effects, how can Parliament and
Ministers rely on Treasury's fiscal advice?
This is not a request for speculative analysis - it is a request for:
Confirmation of whether current methodologies account for these pathways
Documentation of historical cases where this occurred
Transparency about methodological limitations
The previous OIA response claiming no relevant documents exist is particularly
5
concerning given Treasury maintains the CBAx tool with Value of Statistical Life
values specifically designed to monetize mortality impacts in cost-benefit
analysis.
its possible the req was maliciously misunderstood to reduce work load....
This request focuses on basic statistical parameters, methodological guidance,
and public fiscal data that Treasury must maintain to perform its core functions. I
am happy to discuss this request if clarification would be helpful.
Information to be refused and context
To respond to your request the Treasury would need to search through a large amount
of information to find relevant excerpts contained within various reports and internal
documents. This would require a significant amount of time and resources. Accordingly,
this request has been refused under section 18(f) of the Official Information Act –– the
information requested cannot be made available without substantial collation or
research. However, the Treasury has coordinated some explanatory context to each of
your questions, that we trust will be useful to you.
There is some general context that may help in understanding some of our responses.
Agencies provide the Treasury with fiscal forecasts, appropriation details, Performance
Plans and Budget initiatives for review, collation and consolidation within the fiscal
forecasts depending on the nature of the information. That means the Treasury does
not routinely initiate and conduct specific evaluations of social spending.
The Treasury produces two fiscal projection models, the Fiscal Strategy Model (FSM)
and the Long-term Fiscal Model (LTFM). These models project spending out beyond
the fiscal forecasts. These models are Treasury generated analysis and are designed
to project out the government’s accounts under a set of assumptions around key
macroeconomic influences, such as economy-wide growth and the ageing structure of
the population.
The purpose of the fiscal projections is to understand the sustainability of fiscal policy
and fiscal dynamics over the longer term. To this end they contain no data at the micro
level, including at the cohort level. Because of this they can only project out influences
built into their macro inputs, such as those built into Stats NZ’s population projections
or Treasury’s economy-wide forecast of the unemployment rate.
Meanwhile, the Tax and Welfare Analysis (TAWA) model discussed below has been
designed to cost individual policies or events that impact on a relatively small section of
the population. It is a specialised microeconomic model that can be focused on specific
policy areas, which involves data and assumptions at a micro level.
Response to part 1 –Value of Statistical Life (VSL) current parameters
Response to Question 1.1 In the latest version of the CBAx model on the Treasury’s website (updated October
6
2025), the current values of statistical life (VoSL) in the database are $16,615,947
(midpoint) and $10,767,134 (low point).
These estimates use values from NZ Transport Agency/Waka Kotahi data in 2021 and
are adjusted for inflation, giving values adjusted to 2026. The reference for the mid-
point estimate is
(Social Cost of Road Crashes and Injuries - Te Manatu Waka 2022
refers) and the reference to the minimum estimate is
(Waka Kotahi - Monetised
Benefits & Costs Survey 2023 (P13) refers).
There is no specific Treasury guidance on how to use VoSL, as it depends on the
circumstances of proposals and how to appropriately incorporate the concept in cost
benefit analysis. In addition, there is no requirement in cost benefit analysis to use any
particular estimate for VoSL. However, we expect any estimate/value used to be
justified.
There is general guidance available on Treasury’s website regarding the CBAx model
(CBAx Tool User Guidance refers), and for cost benefit analysis in general
(Guide to
Social Cost Benefit Analysis refers).
Response to Question 1.2
VoSL is not age-adjusted by cohort; it is the same for all age groups. If you would like
more information on how VoSL is derived, we suggest you contact the New Zealand
Transport Agency - Waka Kotahi directly.
Response to Question 1.3
The Treasury’s fiscal projection models use Stats NZ’s
(National Population Projections
refers) as demographic inputs. Specifically, the 50th percentile (median) set of
population projections are used as inputs to the baseline projections of the models.
Stats NZ projections include changing assumptions, over the projected years, about life
expectancy at birth and at age 65 years for both males and females. These
assumptions wil have been influenced by data on several aspects of demography,
including annual mortality figures over past years. More information on the
assumptions that go into these population projections would need to be sought from
Stats NZ, but the Treasury age and gender group projections do not specifically factor
in any assumptions around things like premature deaths or suicide rates, beyond what
has been built into Stats NZ’s projections.
Response to part 2 – Fiscal Baseline Assumptions
Response to Question 2.1
As outlined in the answer to 1.3, the Treasury’s fiscal projection models use Stats NZ’s
median
(National Population Projections refers) as demographic inputs. Information
about what assumptions Stats NZ use in these projections regarding annual numbers
of deaths and life expectancy at birth, for both males and females, can be found in the
tables downloadable from
(this site refers).
7
Response to Question 2.2
For reasons explained above, the Treasury’s fiscal projection models do not include
calculations of the specific fiscal impacts of premature death.
Response to part 3 – Methodology Documentation
Response to Question 3.1
Social policy interventions are subject to scrutiny when proposed and when in place:
• When proposed:
• If there are regulatory implications, agencies complete a Regulatory Impact
Statement that test the intervention rationale, options and the pros and cons
of each, before arriving at a preferred approach.
• If there are fiscal implications, they are typically set out in agency
submissions and considered as part of the Budget process, which subject
proposals to value for money scrutiny. To this end, for capital proposals, there
is the 5-case model (strategic, economic, finance, management and
commercial cases). While operating proposals are subject to the ABCDE
framework (Alignment, Benefits, Costs, Delivery and Equity).
• When in place:
• Parliament approves funding through appropriations. Appropriations (such as
for social spending) generally set out what is intended to be achieved and
indicators that should be used for assessing appropriation performance.
This is included in the Estimates and then reported against in annual reports
and other end-of-year performance reporting. All of this information is subject
to Parliamentary scrutiny, such as through Select Committees.
• Agencies are expected to self-evaluate the effectiveness of spending against
their objectives and priorities as part of their planning processes and/or
against specific requirements set by Cabinet where decisions qualify for
‘major Budget decisions’ monitoring and reporting.
As discussed above, there is no specific guidance/framework for considering proposals
that have VoSL, mortality, or morbidity implications. Where appropriate these factors
should be included in analysis within the general frames.
Response to Question 3.2
TAWA is used to analyse distributional impacts across population cohorts. This can be
applied in the Budget process to estimate the costs and distributional impacts of
potential personal tax and welfare policies. It is also used in estimating future child
poverty rates as part of child poverty reports, targets, and research, and has provided
advice on welfare reform, including changes to Working for Families and child support
pass-on. It has also been used to understand other policy issues. TAWA has also been
used by other government agencies and in academic collaborations on tax and welfare
issues. This link provides some extra information
(https://www.treasury.govt.nz/information-and-services/financial-management-and-
advice/revenue-expenditure/tax-and-welfare-analysis-tawa-model refers)
8
Documentation on TAWA can be found here
(https://www.treasury.govt.nz/sites/default/files/2024-08/tawa-model-
methodology.pdf.refers)
Response to part 4 – Clarification Request 4.
Response to Questions 4.1 - 4.3
The Treasury is not lead policy advisor on social welfare, mental health or justice
matters. Responsibility for this sits with other government departments such as the
Ministry of Social Development, the Ministry of Health and the Ministry of Justice.
Generally speaking, the Treasury’s role in these areas focuses on providing second-
opinion advice to Ministers on policy or spending proposals from the relevant Minister
or department. This is discussed in 3.1 regarding proposals going through the Budget
process. As such, the Treasury generally does not conduct cost-benefit or efficiency
analysis in these policy areas, but we may provide Ministers with our views on any
such analysis undertaken by the relevant lead government department.
Response to part 5 – Fiscal Policy and Social Harm – Impact Modelling
Response to Question 5.1
When advising on the appropriate fiscal settings, Treasury considers likely effects on
general social outcomes and monetary conditions, for example when advising on the
size of allowances.
This includes modelling the effects of fiscal policy on variables such as income
distribution and income levels. However, we do not do model the specific health and
social impacts of different fiscal policy options and decisions.
Response to Question 5.2
We do not analyse the social impacts of changes in the Official Cash Rate (OCR). Our
models incorporate changes to the OCR, and the effects of the OCR on various
factors, including consumption and income, but we do not look at the impact and/or
correlation of OCR shifts on social outcomes.
Response to Question 5.3
When providing advice about fiscal projections and achieving the Government’s fiscal
strategy goals, including the pace and composition of fiscal consolidation, our analysis
considers the broad trade-offs, including likely effects on aggregate demand, the
provision of government services, and the associated potential impacts on wellbeing at
a macroeconomic level.
However, the Treasury does not model the effects of spending restraints on welfare
and/or societal harm at a microeconomic level.
9
Response to part 6 – Projection Validity and Feedback Effects
Response to Question 6.1
The Treasury’s fiscal models for the FSM and LTFM produce projections, not forecasts.
Forecasts, both economic and fiscal, attempt to predict future outcomes as accurately
as they can, given the information available at the time that they are produced. To do
so, forecasters use wide-ranging resources, comprehensive modelling and expert
opinion and knowledge.
Meanwhile the fiscal models’ projections, which arise from and are heavily influenced
by their forecast bases, are potential paths under a specific set of assumptions. These
paths are based on trends or long-run averages for growth rates or levels of key
economic, fiscal, and demographic variables, and generally assume no policy changes
beyond those built into their forecast base.
The FSM shows how the paths of key fiscal indicators, like the government’s operating
balance or net debt, might evolve under specific expenditure control and a set of other
relevant economic, fiscal and demographic assumptions. The LTFM shows how the
fiscal pressures caused by an ageing population structure, climate change, replacing
old infrastructure, responding to external chal enges, and other factors, wil require
policy change to keep the fiscal position stable and sustainable for current and future
generations.
There are numerous reasons why projections wil not accurately predict future
outcomes. Actual outcomes wil diverge from the projections’ backdrop of an economy
growing exactly on trend, with unchanging unemployment, inflation and interest rates.
Outcomes, both fiscal and economic, wil also be affected by events and policy
changes that occur after the projections are produced.
The kind of averaging assumptions that need to be made for projections, in areas like
economy-wide productivity, average tax rates and adequately funding competing
expense areas like welfare, health and education, wil be a greater source of error in
the projections than a specific area like the lack of social harm feedback loops.
Modelling social harms is a specialised task best addressed in specialised models with
micro level data inputs, and not the generalised, macro level fiscal projection models.
The Treasury has been publishing the results of its LTFM since 2006. In figure 10 in the
following link
(He Tirohanga Mokopuna - Long-term Fiscal Statement 2025 -
September 2025 refers) there is review of the accuracy of government expense
projections.
Response to Question 6.2
The general rule applied in the fiscal projection models is that they project out either
existing policy or policy change that has been agreed and built into the forecast base of
the projections.
The Treasury’s fiscal projection models project the accounts of the New Zealand
government, and it is the current administration which decides and enacts public policy.
10
Consequently, it would be generally inappropriate for the Treasury to “second guess”
policy changes for social policy beyond the forecast years that serve as the base of the
projections.
For further information regarding the assumptions that underpin the FSM and LTFM
you could review the following documents
(Budget Economic and Fiscal Update 2025
of the Fiscal Strategy Model (FSM) - 22 May 2025 and Background Paper for the 2021
Statement on the Long-term Fiscal Position: Demographic, Economic and Fiscal
Assumptions and Logic in the 2021 Long-term Fiscal Model | The Treasury New
Zealand refer)
Response to Question 6.3
As part of their Budget submissions, agencies are expected to identify where there
would be costs (fiscal and/or social) arising from policy proposals (including possible
cost shifting or amplification). Those would be assessed and considered as part of the
Treasury's advice. This requires initiatives to demonstrate alignment with Government
priorities, clearly identified and (where possible) quantified benefits and costs, robust
delivery arrangements, and consideration of equity and distributional impacts. These
assessments are supported by evidence, assumptions, intervention logic and, where
appropriate, cost-benefit analysis modelling, and are assessed and released (where
appropriate) through the Budget proactive release process.
The Treasury does not keep a consolidated or systematic historical record of cases
where reductions in preventive social spending led to increased remedial spending that
exceeded the original savings. In practice, detailed evaluation is undertaken within
sector agencies, and Treasury’s involvement is typically limited to providing second-
opinion advice on analysis prepared by those agencies.
Response to part 7 – Methodological Transparency
Response to Question 7.1
The Treasury does not accurately know how social costs wil unfold in the future. The
purpose of the Treasury’s fiscal projection models is to display potential paths of key
fiscal indicators under a specific set of assumptions. The FSM and the LTFM are not
designed to accurately predict future costs in specific areas.
Response to Question 7.2
The Treasury is always seeking to improve its model ing and wil continue to do so (see
below) and is transparent about the assumptions and limitations of its fiscal projection
models (see response to part 6.2). The FSM is routinely updated and published.
There are areas that a model that is designed to operate at a whole-of-economy level,
using macro data inputs and aggregate assumptions, wil not be able to deal with
adequately. For this reason, some of the areas that have been asked about, such as
the wide-ranging types of costs involved with the relatively small proportion of the
population who die prematurely due to il ness, accident or the malevolent actions of
others, cannot be adequately addressed by these models.
11
In 2025, the LTFM was complemented by introducing a newly developed overlapping
generations (OLG) model. The OLG model adds an intergenerational perspective by
simulating how households and businesses respond to changes in government policy
over time. For more on this
(Long-term fiscal position | The Treasury New Zealand
refers)..
Please note that this letter (with your personal details removed) and enclosed
documents may be published on the Treasury website.
This reply addresses the information you requested. Under section 28(3) of the Act,
you have the right to ask the Ombudsman to review any decisions made under this
request. The Ombudsman may be contacted by email at:
[email address] or by calling 0800 802 602.
Yours sincerely
Jess Hewat
Acting Senior Manager, Public Finance and Value for Money
12
