MODEL TRANSPARENCY

How Lakive works

Every number on this site comes from a defined formula applied to public data. This page explains what each indicator measures, where the data comes from, how often it is updated, and where the model falls short.

We do not publish the complete scoring weights. Releasing exact weights would allow actors to game the data — tuning listings or postings to score well rather than be genuinely affordable or in-demand. What we do publish is every indicator, its data source, and the direction of its effect on the score. That is enough to understand and challenge any result.

Section 1

Indicator Definitions

Lakive tracks two housing-affordability metrics and six city-quality indices. The two affordability metrics are occupation-specific and property-type-specific. The six city indices are city-level — they do not vary by occupation.

Housing Affordability Indicators

HPI YearsReference price ÷ gross annual salary. Answers: how many years of full pre-tax income to buy this property type? — Lower = less burden
Reference priceCity-specific benchmark by property type (1BR · 2BR · 3BR · Townhouse · Detached). Canada: CMHC / CREA composite benchmark. US: Zillow ZHVI / local MLS median.
Salary denominatorOccupation median salary sourced from Statistics Canada LFS (CA) or BLS OES (US). Fixed per city–occupation pair; does not reflect individual variation.
CA benchmark salaryCAD 75,000 — used when no occupation data is available for a city, as a conservative floor.
HPI Level bandsL1 ≤ 5 yrsL2 ≤ 8 yrsL3 ≤ 12 yrsL4 ≤ 18 yrsL5 > 18 yrs
RPI %Average 2-bedroom rent × 12 ÷ annual salary × 100. Answers: what share of gross income goes to rent in a standard 2BR? — Lower = less burden
RPI Level bandsL1 ≤ 25%L2 ≤ 30%L3 ≤ 38%L4 ≤ 50%L5 > 50%
Pressure Levelmax(hpiLevel, rpiLevel). The overall housing pressure rating is driven by whichever of the two dimensions is worse — you cannot escape a severe RPI with a good HPI.

City Quality Indices (city-level, not occupation-specific)

EOI — Employment Opportunity IndexQualitative assessment of how well the city's economic base generates demand for the specific occupation. Rated High / Mid / Low; converted to numeric (High = 78, Mid = 62, Low = 45) for scoring. — Only manually-set index — updated annually
TAI — Transit & Access IndexComposite of transit mode share, commute times, walkability, and inter-city connectivity. Sources: GTFS feeds, Census journey-to-work, Walk Score API.
HAI — Healthcare Access IndexPhysician-to-population ratio, hospital bed density, and wait-time indicators. CA: CIHI. US: HRSA Area Health Resource Files.
EQI — Environmental Quality IndexAnnual AQI average, green-space coverage, and extreme-weather event frequency. Sources: EPA AirNow (US), Environment Canada AQHI (CA), NASA MODIS land cover.
TCI — Tax & Cost IndexCombined effective tax rate on median income including provincial/state income tax, sales tax, and property tax. Inverted — lower tax burden → higher index.
PSI — Public Safety IndexViolent crime rate and property crime rate per 100,000. Sources: Statistics Canada UCR Survey (CA), FBI UCR / NIBRS (US). Inverted — lower crime → higher index.
Section 2

Data Sources

All source data is public or licensed from the issuing agency. No proprietary data is purchased from third-party vendors.

Canada

Housing pricesCMHC Housing Market Information Portal · CREA MLS HPI
Rental ratesRentals.ca National Rent Report · CMHC SMAI
SalariesStatistics Canada Labour Force Survey (LFS), Table 14-10-0064-01
Employment dataStatistics Canada LFS monthly release · Job Bank postings
HealthcareCIHI Health System Use by LHIN/Zone · CMA physician registry
EnvironmentEnvironment Canada AQHI historical · ECCC weather normals
Tax ratesCRA tax brackets · provincial finance ministry schedules
SafetyStatistics Canada UCR Survey, Table 35-10-0177-01

United States

Housing pricesZillow Home Value Index (ZHVI) · local MLS median
Rental ratesHUD FMR · Apartment List National Rent Report
SalariesBLS Occupational Employment and Wage Statistics (OEWS)
Employment dataBLS monthly CES release · CareerOneStop postings
HealthcareHRSA Area Health Resource Files · CMS hospital compare
EnvironmentEPA AirNow annual AQI summary · NLCD land cover
Tax ratesTax Foundation state-local tax burden estimates · IRS SOI
SafetyFBI UCR / NIBRS · city open-data crime portals

City-level indices (TAI, HAI, EQI, TCI, PSI) are reviewed annually. Housing and salary data are updated on the schedule described in Section 4.

Section 3

Scoring Formula

The Lakive Fit Score is a composite of two sub-scores: Housing Affordability and City Quality. Both are computed at the occupation level and property-type level; the city indices apply city-wide.

Score structure (Model v4)

HOUSING SCORE (52% of final score)
housingScore = hpiScore × 0.55 + rpiScore × 0.45
hpiScore: 92 (≤6 yrs) → 82 (≤8) → 70 (≤10) → 58 (≤12) → 45 (≤16) → 30 (> 16)
rpiScore: 90 (<25%) → 82 (<30%) → 72 (<35%) → 60 (<40%) → 48 (<45%) → 35 (≥45%)
CITY SCORE (48% of final score)
cityScore = EOI×0.22 + TAI×0.20 + HAI×0.20 + EQI×0.14 + TCI×0.12 + PSI×0.12
EOI is the only index that varies by occupation — the others are city-wide constants.
FINAL SCORE
fitScore = round(housingScore × 0.52 + cityScore × 0.48)
Clamped to [10, 99]. A score of 10 indicates severe housing or city pressure; 99 indicates ideal conditions for that occupation in that city.

Score weights reflect the view that housing affordability is the primary reason people leave a city, while city quality determines whether they want to stay once they can afford to. We do not publish the precise sub-weights of the six city indices to avoid gaming. The direction of each index is unambiguous: higher EOI / TAI / HAI / EQI / TCI / PSI always improves the score.

Score interpretation

85 – 99Highly Recommended — strong affordability and city quality for this occupation
70 – 84Recommended — good overall balance; may carry housing pressure caveat
55 – 69Moderate — viable but requires planning; likely elevated housing cost
40 – 54Challenging — significant housing or city-quality constraints
10 – 39Not Recommended — severe affordability or opportunity mismatch
Section 4

Update Schedule

Lakive follows official release calendars wherever possible. Scores are recalculated after each data update — there is no manual override of individual city scores between scheduled updates.

LFS / BLS monthlySecond Friday of each month (CA) · Third Friday of each month (US). Employment, unemployment rate, and labour force participation.
Housing pricesCMHC: monthly, ~15th of following month. Zillow ZHVI: monthly, ~mid-month. CREA MLS HPI: monthly, ~15th of following month.
Rental ratesRentals.ca: monthly National Rent Report, ~end of following month. HUD FMR: annually, October. Apartment List: monthly.
City indicesTAI, HAI, EQI, TCI, PSI: annually, Q1 of each year (January – March), incorporating prior-year data.
EOI (Employment Opportunity)Annually, reviewed with LFS Annual Survey data. Significant sector shifts may trigger interim updates.
Fit scoresRecalculated automatically whenever an underlying data input changes. No scheduled cadence — always reflect the latest available data.
Section 5

Model Limitations

No model captures everything. The following are known structural limitations of the current version. We prefer to be explicit about them rather than let users discover them accidentally.

⚠ Single benchmark property type

HPI Years uses one property benchmark per city — a 2-bedroom condo composite in most Canadian cities, or ZHVI for a given tier in US cities. It does not reflect the full spread of prices within a city. Neighbourhoods vary widely.

⚠ Pre-tax salary denominator

RPI and HPI Years use gross annual salary. After-tax take-home varies by province / state and individual circumstances. The TCI index partially accounts for tax burden at the city level, but individual tax situations differ.

⚠ No household-size adjustment

All indicators assume a single earner. Dual-income households, families with dependants, and single parents face materially different affordability curves that the model does not capture.

⚠ EOI is qualitative and lagged

The Employment Opportunity Index is the only index not derived from a formula. It is based on sector analysis and reviewed annually. Sudden sector expansions or contractions (e.g. tech layoffs, new facility announcements) may not be reflected until the next annual update.

⚠ No personal circumstances

The model does not account for immigration status, prior local work experience, language proficiency, professional licensing portability, or personal credit history — all of which materially affect housing access and job placement.

⚠ City-level aggregation

All indices are city-level averages. Within a city, neighbourhood quality, micro-market rental rates, and local employer concentration vary substantially. The score should be read as a city-level signal, not a neighbourhood-level one.

⚠ Data lags

Official statistics are published with a lag of 4–8 weeks for monthly data and up to 12 months for annual data (crime, healthcare). Scores may not reflect very recent market shifts (e.g. a sudden rate change or economic shock).

Section 6

Version History

Model v4 · Oct 2026Current version. Single Source of Truth refactor: all scores computed from lib/fit-engine.ts; FIT_MATRIX (hardcoded values) fully removed. Housing weight raised to 52% (from 50%). RPI score table re-anchored to 35% threshold.
Model v3 · Jul 2026Added US city coverage (11 cities). EOI_MATRIX extended to US markets. BLS OES substituted for LFS for US occupation salaries. Zillow ZHVI integrated as housing price source.
Model v2 · Mar 2026Split housing score into HPI Years (purchase) and RPI % (rental), weighted separately. Introduced property-type selector (1BR / 2BR / 3BR / Townhouse / Detached). Added price-adjustment scenario multipliers.
Model v1 · Jan 2026Initial launch. Single housing affordability metric (price-to-income ratio). Five Canadian cities. City score based on four indices (TAI, HAI, EQI, PSI only).

Questions or corrections → hello@lakive.com
Last updated: October 2026 · Lakive City Intelligence Platform v5.0

Rankings →Career Guides →