How we calculate the numbers

Version: 2.1 Last updated: April 2026 Published by: Beneath the Surface AU

This page documents the data sources, assumptions, and formulas behind every number on the Fuel Risk WA dashboard. Our goal is to make the methodology transparent enough that anyone can check our working.

Data sources

Fuel prices come from FuelWatch, the WA government's official daily fuel price service. We use their REST API (v3), which replaced the earlier RSS feed in March 2026. We track unleaded petrol (ULP) and diesel across seven regions. Users can toggle between fuel types on the dashboard. E10 is not tracked — see the limitations section below.

Minimum wage is the current Australian national minimum wage set by the Fair Work Commission — $24.10/hr for the 2025–26 financial year.

Rent is based on REIWA (Real Estate Institute of WA) median weekly rent figures by region. These are updated as REIWA publishes them. When live data is unavailable, we use recently verified fallback figures — see the rent section below.

Commute distances come from the ABS Census 2021 Journey to Work data (Table W06), using SA3-level median commutes by region. This is the most recent Census available; actual commuting patterns may have shifted since 2021.

Fuel efficiency is drawn from the ABS Survey of Motor Vehicle Use (Cat. 9208.0, 2020 edition — the final edition before the survey was discontinued). State-level averages by vehicle type. These figures may slightly overstate fuel consumption due to newer vehicles entering the fleet since 2020, which makes our cost estimates conservative.

Workforce estimates use ABS Small Area Labour Markets (SALM) employment data from September quarter 2025, combined with ABS Employee Earnings and Hours (Cat. 6306.0) to estimate the proportion of workers on or near minimum wage.

Tax rates use the ATO individual income tax schedule for 2025–26.

Baseline date is 28 February 2026 — the day US–Iran conflict began disrupting global fuel supply chains. All "increase since crisis" figures use fuel prices on that date as the comparison point.

Assumptions

The model represents a single-earner household with one vehicle — the most financially exposed common configuration for a minimum-wage worker.

Commute and fuel efficiency by region

Regional centres use shorter commute distances because the urban areas are smaller. Remote towns use the shortest figure for daily commuting, but this understates real transport costs — residents in Broome and Port Hedland often travel long distances for medical care, supplies, and services that aren't available locally. That cost is not captured here.

All calculations assume a standard 38-hour, five-day working week.

Calculations

Daily fuel cost

daily_fuel_cost = (commute_km ÷ 100) × fuel_efficiency × fuel_price_per_litre

This uses region-specific commute and efficiency figures, not a single state-wide average.

Weekly fuel cost

weekly_fuel_cost = daily_fuel_cost × 5

Net weekly income

Rather than applying a single flat tax rate, we use the actual ATO 2025–26 marginal tax brackets:

For a full-time minimum-wage worker (38 hrs × $24.10 = $47,685.60 gross per year):

• Tax on $0–$18,200 = $0
• Tax on $18,201–$45,000 = $4,288
• Tax on $45,001–$47,685.60 = $806
• Total tax ≈ $5,094, effective rate ≈ 10.7%

net_weekly_income = (annual_gross − annual_tax) ÷ 52  ≈  $818.88/week

A note on LITO and Medicare Levy: We omit two small tax adjustments — the Low Income Tax Offset (LITO, which would reduce tax by ~$566/year) and the Medicare Levy (2%, which would add ~$954/year). Together, these would reduce net weekly income by about $7.46 — less than 1% of take-home pay. The impact on burden ratios is under 1 percentage point. We exclude them for simplicity; the direction and size of the error are documented here so readers can judge for themselves.

Risk model

Essentials burden ratio

The core metric on this dashboard is the essentials burden ratio — the share of net weekly income consumed by rent and commute fuel combined.

weekly_essentials = median_weekly_rent + weekly_fuel_cost
essentials_burden_ratio = weekly_essentials ÷ net_weekly_income

This is based on the residual income method used in housing affordability research. The idea is that what matters isn't just rent-to-income in isolation — it's what's left after paying rent and getting to work, because that remainder has to cover food, utilities, medical costs, and everything else.

A ratio of 0.55 means 55 cents of every dollar earned goes to housing and commuting, leaving 45 cents for all other expenses.

Two-layer risk display

The dashboard separates the total burden into two parts:

Structural burden — what the region looked like before the conflict:

structural_burden = (median_weekly_rent + baseline_weekly_fuel_cost) ÷ net_weekly_income

Crisis surcharge — what the fuel price increase has added on top:

crisis_surcharge = (current_weekly_fuel_cost − baseline_weekly_fuel_cost) ÷ net_weekly_income

Together, structural burden plus crisis surcharge equals the total essentials burden ratio. Showing both separately makes the causal story visible: how much was already a problem before February, and how much the conflict added.

Risk thresholds

The dashboard displays this on a 0–100 scale for readability.

Workers at risk

A worker is counted as "at risk" when their essentials burden exceeds 50% — meaning more than half of net minimum-wage income goes to rent and commuting.

We estimate the number of minimum-wage workers per region by combining ABS Small Area Labour Markets employment data with the national proportion of workers earning minimum wage or award base rates (approximately 10.6% of all employees, per ABS Cat. 6306.0). The headline figure on the dashboard sums across all seven tracked regions.

This is an estimate, not a census count. The 10.6% national rate is applied uniformly across regions. In practice, remote areas likely have a higher proportion of low-wage workers; inner Perth likely has a lower one. We'll incorporate ABS Census income distribution data at a finer geographic level in a future update.

Important caveat: This calculation uses a hard threshold — if a region's burden ratio crosses 50%, all estimated minimum-wage workers in that region are counted. If it's just below 50%, none are counted. This means the headline figure can jump sharply with small price changes, especially for Perth Metro (which has the largest workforce). Read it as "the number of minimum-wage workers in regions that have crossed the severe stress line" rather than a count of individuals in crisis. We use this approach because we don't have data on how rent varies within each region — only the median — so we can't model what proportion of individual workers exceed the threshold.

Composite risk index

Some dashboard views show a 0–100 "risk index." This is a weighted composite of three factors: rent burden (40% weight), fuel price increase (35%), and grocery basket increase (25%). Each factor is normalised to a 0–100 sub-score, then combined.

This index predates the essentials burden ratio and is retained for chart continuity. The essentials burden ratio (rent + fuel as a share of net income) is the primary metric — it's simpler, more transparent, and directly interpretable.

The risk index weights reflect editorial judgment about relative importance, not a statistical model. It should be read as "higher is worse," not as a calibrated probability.

Rent data

We use REIWA median weekly rent as the primary source. When live data isn't available, we fall back to recently verified figures:

When rent data is stale (7+ days without a live update), the dashboard displays an indicator next to the affected figures.

Essentials basket

We track a 20-item grocery basket scraped daily from Woolworths, covering staples like bread, milk, eggs, rice, chicken, vegetables, and fruit. The full item list is published in our source code (pipeline/market-basket.json).

For each item, we take the lowest available price. The weekly basket cost is the sum of these cheapest prices across all 20 items.

What the basket is: A cost-of-survival indicator — tracking what it costs to buy basic food and household essentials at a mainstream supermarket.

What the basket is not: It's not comparable to the ABS Consumer Price Index (which uses a much broader, expenditure-weighted basket) or the Henderson Poverty Line (which includes non-food costs and is CPI-indexed). Our basket is narrower and simpler by design.

Current limitations:

• Only Woolworths is scraped (Coles and Aldi support is planned)
• Prices are national (online) — regional price variation, especially freight premiums in remote WA, is not captured
• Items are unweighted (bread counts the same as tea bags)

The basket cost feeds into the "combined burden" metric shown on some dashboard views. It does not feed into the primary essentials burden ratio (which uses rent and fuel only), because regional grocery price data is less reliable than rent and fuel data.

What this model doesn't show

Every number on this dashboard is conservative. The model deliberately limits itself to two costs — rent and commute fuel — because those are the ones where we have reliable, current, publicly sourced data.

The following factors are excluded, and each represents real additional financial pressure that isn't captured here:

• Food and groceries
• Utilities (electricity, gas, water)
• Childcare
• Vehicle maintenance, registration, and insurance
• Medical and dental costs
• Phone and internet
• Clothing and household basics
• Existing debt repayments
• Irregular or casual employment (the model assumes a stable 38-hour week)
• Long-distance travel for remote residents accessing services not available locally
• Disability, chronic illness, or carer responsibilities that constrain earning capacity
• E10 fuel type (E10 is not tracked. FuelWatch coverage for E10 in WA is sparse.)

The gap between what the dashboard shows and actual financial pressure on minimum-wage renters is real, and it's not small.

Questions or corrections

If something here looks wrong, reach out through our website. We'd rather fix an error than have it circulating.