CivicIQ · Methods

NZ Index methods and non-partisan charter

Phase 0/1 covers Health, Housing, Safety, Affordability and Wellbeing. The public brand is CivicIQ / CivicNZ. Health Index rows are share of the published 2030 target. Other domains remain z-scores against 2015–2024 and are not comparable 0–100s. The composite is experimental and is not a national headline. Interpolated Health, Housing and Safety curves have been removed. There is no debate scoring in this release.

Back to the index

Non-partisan charter

  1. We score against evidence, stated objectives, measurable targets and follow-through. We never score ideology, party or values.
  2. The same rubric applies to every party, minister and agency.
  3. All methods, weights, prompts and code are public. Every number links to its source.
  4. We publish corrections. Model scores are labelled “AI-assessed” and the evidence is quoted.
  5. No party, union or lobby funding. The funder list is published.

How a number is scored

  • Health Index rows are labelled Share of 2030 target: 100 × current / Health NZ’s published 2030 percentage, capped at 100, and never printed as 100 while the official value is still below the target. FSA 66.1% against 95% is 69.6, not a z-score of 100. That number is not called Score.
  • Index Health is four of five GPS 2024–2027 health targets (ED, first specialist assessment, elective treatment, 24-month immunisation). Faster Cancer Treatment (90% by 2030) is on the Health desk only — not silently dropped and not in the Index mean.
  • 2025/26 milestones sit on a chip (met / missed), sourced from the Health NZ Q4 2025/26 factsheet. They do not inflate the share. Elective’s chip is 70% (Health NZ Q4 pack); GPS Table 2 printed p. 50 lists 67% for the same year. 72.6% met both.
  • Housing, Affordability and Wellbeing remain z-scores against their own 2015–2024 baseline, then sign-flipped so up always means better. The 0–100 scale there is 50 + 10z, clipped. 50 is the baseline average, not a value judgement. Those 0–100s are not comparable to Health’s share of 2030 target.
  • Health baseline z (50 + 10z) is methods-only. It uses Health NZ’s decade Q4 sheet, Q4-only, years 2016/17–2023/24, and only when n ≥ 8 and span ≥ 5 years. It is not the public Health grade. A failed gate omits z; it does not stamp the indicator Insufficient.
  • Target-share is Health-only, higher-is-better percentages with a positive official target. Lower-is-better series and a target of 0 are not divided. Housing, Affordability and Wellbeing are not switched to current / target.
  • Domain Health may show the mean of those four shares (today 78.6) labelled Mean share of 2030 target. The NZ Index headline is not the mean of that share and the z-scored domains. Mixed rubrics are not averaged.
  • Indicator trends use the same CivicNZ rule as the desks: official 95% significance and published confidence intervals first, then a like-for-like year-on-year change, then Theil–Sen only when no same-season pair exists and there are four or more official points spanning two years. Labels are Improving, Worsening, Flat or Insufficient. A homemade 0.2-unit / 2% Flat band is not used.
  • Missing values are never imputed or interpolated. A series that is more than 12 months old is marked stale. Components without a real official observation are excluded from the domain mean. Insufficient remains “no scored official series” (Safety). The Health desk is not Insufficient because of this scoring change.
  • Every displayed figure carries its agency source, series code, release date and fetch date. Official series are typically Crown copyright, CC BY 4.0 (NZGOAL).

Published weights

Health 1, Housing 1, Safety 1, Affordability 1, Wellbeing 1. Readers can re-weight in the browser. CivicIQ does not treat those personal sliders as the published index.

The twenty specified indicators

DomainIndicatorSourceSeriesStatus
HealthED stays within 6 hoursHealth NZ / Te Whatu OraHEALTH-ED-6HRScored
HealthFirst specialist assessment within 4 monthsHealth NZ / Te Whatu OraHEALTH-FSA-4MScored
HealthElective treatment within 4 monthsHealth NZ / Ministry of HealthHEALTH-EL-12MScored
HealthImmunisation at 24 monthsHealth NZ / Ministry of HealthHEALTH-IMM-24MScored
HousingPublic housing registerMinistry of Social DevelopmentMSD-HR-TOTALScored
HousingEmergency housing householdsMinistry of Social DevelopmentMSD-EH-HHScored
HousingMedian rent to median incomeMBIE Tenancy Services / Stats NZMBIE-RENT-INCExcluded — no official compiled series
HousingNew dwellings consented per 1,000 peopleStats NZ ADESTATSNZ-BLD-CONSExcluded — no official compiled series
SafetyRecorded victimisations per 1,000 peopleNZ PolicePOLICE-VIC-1000Excluded — no official compiled series
SafetyNZCVS crime prevalenceMinistry of JusticeMOJ-NZCVS-PREVExcluded — survey began 2018; no pre-survey points
SafetyViolent victimisations per 1,000 peopleNZ PolicePOLICE-VIO-1000Excluded — no official compiled series
SafetyYouth proceeded against per 10,000 aged 10–17Ministry of JusticeMOJ-YOUTH-10KExcluded — no official compiled series
AffordabilityMedian house / median HEDIStats NZ Housing in Aotearoa 2025STATSNZ-HIA-HEDIScored
AffordabilityRenters spending 30%+ on housingHUD system indicators / Stats NZ HESHUD-HES-RENT30Scored
AffordabilityHLPI minus LCIStats NZ HLPI and Labour Cost IndexSTATSNZ-HLPI-LCIExcluded — June 2026 point is after the 2015–2024 baseline
AffordabilityHousing costs per $100 disposable incomeStats NZ Household Economic SurveySTATSNZ-HES-HC100Scored
WellbeingOverall life satisfactionStats NZ General Social SurveySTATSNZ-GSS-LSScored
WellbeingLonely most or all of the timeStats NZ General Social SurveySTATSNZ-GSS-LONEScored
WellbeingTrust in parliamentStats NZ General Social SurveySTATSNZ-GSS-TRUST-PARLScored
WellbeingLife worthwhileStats NZ General Social Survey / Ngā Tūtohu AotearoaSTATSNZ-GSS-PURPOSEScored

Ingest starts from compiled official-style fixtures so the page is reviewable offline. Live agency CSVs replace a series when the official file responds.

What this release does not do

  • No Hansard, question or speech scoring.
  • No Attention Gap scatter and no OIA drafts.
  • No Education, Economy, Environment or Fiscal domains yet.
  • No paid endpoint for the composite /index score or the area desks.
  • REINZ and CoreLogic commercial house-price files are not ingested.