How Numbeo, Mercer, ECA, and The Economist build their indices — and why your real cost of living may be 20% higher or lower than what they show.
Cost-of-living indices are one of the most-used and least-understood tools in international financial planning. The major indices — Numbeo, Mercer, ECA International, The Economist's Big Mac Index — are cited constantly in articles comparing cities, in salary negotiations, and in expat package calculations. But the indices differ significantly in methodology, coverage, and accuracy, and using them without understanding how they are built leads to expensive mistakes. This guide explains how each major index is constructed, where they agree and disagree, and how to use them correctly.
Numbeo is the most widely cited cost-of-living index on the internet, primarily because it is free and covers more cities (over 9,000) than any competitor. Numbeo's data is crowdsourced — visitors to the Numbeo website contribute price reports for their city, and the indices are calculated from these reports. The advantages: comprehensive coverage, free access, frequent updates. The disadvantages: data quality is inconsistent (anyone can submit, and there is limited verification); the sample is biased toward internet-connected, English-speaking contributors; and the categories are relatively narrow.
Numbeo's headline index uses New York City as 100, with other cities expressed as a percentage of New York. Categories include rent, groceries, restaurants, transportation, utilities, and clothing. The rent index is particularly useful — Numbeo's rent data is more reliable than its other categories because rent is a discrete, easily-verified data point.
Mercer's survey is the gold standard for corporate expatriate compensation. Mercer collects data on a defined basket of around 200 goods and services in 227 cities, using a consistent methodology. The basket is tilted toward expatriate consumption patterns — it includes imported Western goods, international school fees, and other items that local residents may not consume. The data is collected by Mercer researchers, not crowdsourced, which gives it higher accuracy but also higher cost (the survey is not free).
Mercer's index is used by most large multinationals to set expatriate compensation. If you are negotiating an expat package, the company is likely using Mercer data internally. The Mercer index tends to show higher cost-of-living differentials than Numbeo because the expat consumption basket is more expensive than the local consumption basket.
ECA International is the other major paid cost-of-living index, similar to Mercer in methodology but with somewhat different basket composition. ECA's index is widely used by Asian and European multinationals, while Mercer dominates the American market. The two indices generally agree on broad rankings but can differ significantly on specific cities.
The Big Mac Index is a light-hearted but surprisingly useful currency valuation tool published annually by The Economist. The index compares the price of a McDonald's Big Mac across countries, with the implication that a Big Mac should cost roughly the same everywhere in purchasing-power-parity terms. Differences in Big Mac prices indicate currency over- or under-valuation.
The Big Mac Index is not a comprehensive cost-of-living index — it tracks a single product — but it is a useful sanity check on the more comprehensive indices. If Numbeo says a country is 40% cheaper than the US but the Big Mac Index shows Big Macs at the same price, something is off.
The fundamental question is what "cost of living" means. There are three distinct concepts that are often conflated:
This is what most indices measure: the cost of buying the same goods and services in different cities. The basket is fixed (e.g., 1 kg of rice, 1 litre of milk, 1 month of rent for a 1-bedroom apartment in the city centre, etc.) and the index compares the cost of this basket across cities. The basket is typically weighted toward Western expatriate consumption — it includes imported goods, international brands, and Western-style housing.
The strength of this approach is comparability — the basket is the same everywhere, so the comparison is apples-to-apples. The weakness is that the basket may not reflect what people actually consume. A basket that includes imported cheddar cheese and wine is relevant for British expats but irrelevant for locals in most Asian cities, who consume different products.
This concept recognises that the consumption basket varies by location — a "comparable lifestyle" in Bangkok does not include cheddar cheese but does include local street food, Thai-style accommodation, and local transport. The cost of an equivalent lifestyle is harder to measure because it requires subjective judgement about what counts as "equivalent".
This is what most individuals actually want to know: "If I move from London to Bangkok, how much do I need to maintain my lifestyle?" But this is not what most indices measure, which leads to systematic errors.
This is the cost of living as a local — consuming what locals consume, at local prices. This is the lowest cost of living in any given city, but is typically not relevant for expats (who are not willing to live in shared accommodation with multiple families, eat only the cheapest local food, and use only public transport). It is relevant for understanding local poverty lines and for migrants integrating into local society.
Five systematic errors affect cost-of-living indices, particularly the crowdsourced Numbeo:
Numbeo's contributors are self-selected — they are people who choose to contribute price data to a website. This population skews toward English-speaking, internet-connected, younger, and more educated demographics. In cities where this demographic is unrepresentative of the broader population (most cities), Numbeo's prices may be higher than the local average because contributors shop at more expensive outlets.
Mercer and ECA, which use professional researchers, have less sampling bias but have their own biases — researchers collect prices at specific types of outlets (typically those used by expatriates) rather than at outlets used by locals.
The basket of goods used by Mercer and ECA is explicitly designed for expatriates — it includes imported goods, international brands, and Western-style services. This basket is appropriate for setting expatriate compensation but overstates the cost of living for someone willing to integrate into local consumption patterns.
For a nomad considering a move to Chiang Mai, the Mercer index (which includes imported cheese and international school fees) is irrelevant. The right index depends on the planned consumption pattern.
"Cost of living in Bangkok" is a meaningful concept only at a very aggregate level. Within Bangkok, costs vary dramatically by neighbourhood — a 1-bedroom apartment in Sukhumvit costs 3–5× what the same apartment costs in Lat Phrao. City-level indices average across these neighbourhoods, which can be misleading for individuals with specific neighbourhood preferences.
For most cities, the practical response is to look at neighbourhood-level data rather than city averages. Numbeo and similar sites provide neighbourhood-level data for some cities; for others, you have to do your own research on local rental sites.
A "1-bedroom apartment in the city centre" in Mumbai and a "1-bedroom apartment in the city centre" in London are not equivalent goods. The London apartment will have better build quality, more reliable utilities, better security, and more space than the Mumbai apartment at the same nominal price. Indices that treat these as equivalent understate the cost of an equivalent lifestyle in lower-cost cities.
This is a particularly significant issue for healthcare and education, where quality differences are large and not captured by headline prices. "Private healthcare" in Bangkok is not equivalent to "private healthcare" in London at the same nominal price — the London healthcare is likely higher quality, even though the Bangkok healthcare is excellent by Asian standards.
Cost-of-living indices are typically updated annually, but prices can change much faster — particularly in economies with high inflation (Argentina, Turkey, parts of Africa) or rapid currency depreciation. An index based on prices from 6–12 months ago may significantly understate the current cost of living in these economies.
For high-inflation economies, supplement index data with recent on-the-ground research (recent forum posts, recent rental listings, recent visitor reports). Numbeo's frequent updates make it the most timely of the major indices for these economies.
Despite their limitations, cost-of-living indices are useful tools if used correctly. The practical guidance:
Compare Numbeo, Mercer (if available), ECA (if available), and the Big Mac Index for any city you are considering. Where the indices agree, you can be reasonably confident in the relative cost. Where they disagree significantly, investigate the cause — usually it is a difference in basket composition or methodology.
Indices are based on a typical basket that may not match your consumption. If you do not eat imported cheese, do not send children to international school, and are willing to use public transport, your cost of living in a "high cost" city may be 20–30% below the index. Conversely, if you require Western-style accommodation and imported goods in a "low cost" city, your cost may be 20–30% above the index.
For most people, the three largest cost categories are housing, food, and transport. These three typically account for 60–75% of total spending. Get these three right and the rest is detail. Use our Cost of Living Comparator for a category-by-category breakdown across cities.
Indices are starting points, not final answers. Before making a major decision based on cost-of-living differences, validate with recent research: current rental listings on local sites, recent visitor reports, current grocery prices from someone on the ground. The combination of index data plus recent validation gives a much more accurate picture than either alone.
Cost-of-living indices are most useful in salary negotiations. If you are considering a job offer in a higher-cost city, the index differential gives you a defensible number for the salary increase required to maintain your lifestyle. "Numbeo shows Singapore is 35% more expensive than Dubai, so I would need a 35% salary increase to maintain the same lifestyle" is a more compelling argument than "I think I should be paid more".
For lifestyle decisions — where to live, what to spend — the indices are less useful because they do not capture quality-of-life differences, personal preferences, and the value of social networks. A city that is 20% cheaper but where you are unhappy is not a better choice.
Cost-of-living indices are useful but imperfect tools. The major indices — Numbeo, Mercer, ECA, Big Mac — use different methodologies and basket compositions, and they can disagree significantly on specific cities. Understanding the methodology behind each index is essential for interpreting the numbers correctly. Use multiple indices, adjust for your consumption pattern, focus on the largest cost categories, validate with recent research, and use the indices primarily for salary negotiation rather than lifestyle decisions. The Cost of Living Comparator on this site uses a Numbeo-style methodology with category breakdown — use it as a starting point, but supplement with your own research for any major decision.
All three are tax-free Gulf hubs with strong expat packages, but housing, schooling and leisure costs vary eno…
Spend fewer than 183 days abroad and you may still be tax-resident back home. Here is how the major countries …
Account-opening friction, fee structures, mobile app quality and remittance integration vary enormously betwee…