How every number is computed

Min-max normalization

The weakest market in the set scores 0 on that indicator, the strongest scores 100, everyone else lands proportionally between.

score = (value − min) ÷ (max − min) × 100

Scores are therefore relative to this set of markets, not absolute. 100 means best here — not best possible.

Absolute scales

Two indicators are not rescaled against the other markets at all — they are mapped onto a fixed scale, so the score means the same thing regardless of who else is in the set.

health = clamp((lifeExpectancy − 50) ÷ 35 × 100)  ·  crime = clamp(homicidePer100k ÷ 28.6 × 100)

Life expectancy of 50 years reads 0 and 85 reads 100; a homicide rate of about 28.6 per 100k reads 100. Using min-max here would make a market look healthy simply because its peers were sicker.

Inverted indicators

Where a higher raw reading is the worse outcome — cost of living, housing cost, climate risk, inflation, unemployment — the scale is flipped.

score = 100 − normalized(value)

So a better real-world condition always produces a higher score, everywhere on the site. Rows where a low number wins are labelled as such.

Overall score

One number for how attractive a market looks across quality of life, spending power, growth, risk and business conditions.

overall = 0.30·GLI + 0.25·purchasingPower + 0.20·opportunity + 0.15·(100 − risk) + 0.10·businessScore

Risk enters inverted, so a low-risk market gains rather than loses. The five weights sum to 1.

Global Life Index

A broad read on quality of life: the unweighted average of all twenty pillars, each already 0–100.

gli = mean(20 pillar scores)

Unweighted on purpose — weighting it would bake one definition of a good life into the headline. Apply your own weights on the relocate page instead.

Purchasing power

How far local income actually stretches: what people earn, set against how expensive the market is.

purchasingPower = 0.6·incomeScore + 0.4·affordabilityScore

affordabilityScore is cost-of-living inverted, so a cheaper market scores higher. Both inputs are min-max normalized across the market set.

Big Mac affordability

How many burgers an average local salary buys, relative to the other markets — an everyday affordability gauge.

bigMacAffordability = normalize(incomeNetUSD ÷ bigMacUSD)

Both sides are in US dollars, so the comparison holds across currencies.

Opportunity

How promising the market looks for growth.

opportunity = 0.30·gdpGrowthScore + 0.30·businessEnv + 0.25·purchasingPower + 0.15·tourism

The growth term is live real GDP growth (3-year trailing average), min-max normalized. Business environment and tourism are seed values.

Risk

What could go wrong operationally — prices, jobs, climate exposure, a weakening currency, weak institutions and lethal violence.

risk = 0.22·inflation + 0.18·unemployment + 0.15·climateRisk + 0.18·currencyRisk + 0.12·(100 − governance) + 0.15·crime,  where currencyRisk = clamp((0.5 − currencyTrend ÷ 2) × 100)

Inflation and unemployment enter here directly, where a higher reading is itself the downside. Everywhere else they are inverted. Lower risk is the better outcome.

Economic strength

Near-term strength of the local economy: income, jobs, prices and business conditions.

economic = 0.30·incomeScore + 0.25·employmentScore + 0.25·inflationScore + 0.20·businessEnv

Inflation and employment are inverted here, so a calmer economy ranks higher.

Consumer readiness

How ready and able local consumers are to spend, and how reachable they are.

consumerReadiness = 0.40·purchasingPower + 0.30·incomeScore + 0.30·digital

Business conditions

How straightforward it is to operate in the market.

businessScore = 0.60·businessEnv + 0.40·infrastructure

Both inputs are seed values on a 0–100 scale.

Data confidence

How much the underlying data for this market can be relied on. Lower means read every other number with more caution.

dataConfidence = seed value per market (0–100)

Deliberately NOT derived from the other scores — it is a separate judgement about the data, not about the market. A strong market with thin data is not the same thing as a weak market.

Dollar price

The local price converted at the prevailing exchange rate.

usdPrice = localPrice ÷ marketExchangeRate

Currency valuation

Whether a currency buys more or less than burger parity implies, measured against the US benchmark. Positive reads overvalued, negative undervalued.

valuation% = (bigMacUSD − usBigMacUSD) ÷ usBigMacUSD × 100

Beyond ±10% reads as over- or undervalued; inside that band, fair.

Macs per salary

How many burgers a month's net pay buys — the comparison that holds the product constant and lets the wage move.

macsPerSalary = incomeNetUSD ÷ bigMacUSD

Work minutes per Mac

Minutes of work to afford one burger.

workMinutes = bigMacUSD ÷ (incomeNetUSD ÷ 10,560)

10,560 minutes is a 22-day, 8-hour working month.

Scenario spread

How far the optimistic and pessimistic cases sit from the base case. Volatile, poorly-measured markets fan out; stable ones stay tight.

spread = clamp(round(max(0, inflation − 2)·0.35 + |currencyValuation|·0.06 + risk·0.06 + (100 − confidence)·0.05), 3, 18)

This is why a high-inflation market shows a wide band even when its base score is unremarkable.

The three cases

Base is the market's own score; the other two sit symmetrically either side of it.

conservative = overall − spread ·|· base = overall ·|· growth = overall + spread

Scenario weighting

The upside carries more weight where opportunity outruns risk, and less where it does not.

P(base) = 0.38 ·|· P(growth) = 0.62 × (opportunity − risk + 100) ÷ 200 ·|· P(conservative) = remainder