The Most Livable City

Every year the Economist Intelligence Unit publishes its Global Liveability Index and every year the same cities appear at the top: Vienna, Melbourne, Vancouver, Zurich, Copenhagen. Every year newspapers across the world report the results as though they described something meaningful about what it is like to live in these places. And every year, not a single publication reporting the ranking bothers to look at what is actually being measured.

The index evaluates 173 cities across five categories: stability, healthcare, culture and environment, education, and infrastructure. The categories sound reasonable. The metrics inside them are not. They are designed for a specific audience that has nothing to do with the people who actually live in these cities. The index was built for the human resources departments of multinational corporations. Its purpose is to calculate hardship allowances — how much extra a company like Shell or McKinsey needs to pay an employee to accept a posting in Lagos rather than London. The full dataset costs around eight hundred dollars. The customer is not a curious resident of Vienna. The customer is a compensation analyst in a corporate headquarters.

Once you understand who the ranking is for, the bizarre choices in its methodology stop being bizarre and start being perfectly rational — rational for a purpose that has nothing to do with livability.

Take healthcare. The index measures the availability and quality of private healthcare. Not public healthcare as a system accessible to all residents, but private healthcare available to someone whose employer provides international insurance. Whether the average resident of a city can see a doctor without going bankrupt is not the question being asked. The question is whether the expat's corporate health plan will find adequate private facilities. By this metric, a city with excellent universal public healthcare but few private clinics would score lower than a city with world-class private hospitals and millions of uninsured residents.

Education follows the same logic. The index focuses on private education: whether quality international schools exist for the children of relocated executives. It does not measure equality of educational access, the quality of public schools, or whether a child born in the wrong district has any realistic path to a university. Detroit, a city where the public school system has been in crisis for decades, scores a perfect 100 in education.¹ The metric is measuring something. It is not measuring what the word "education" suggests to a normal person.

The category called "culture and environment" is the most misleading. The name makes you think of parks, clean air, green spaces, biodiversity, the things that actually constitute the physical environment people live in. Instead, the subcategories include: humidity and temperature ratings, discomfort of climate to travellers, availability of sporting events, availability of cultural events, food and drink options, and consumer goods and services. The "environment" in question is not the natural environment. It is the environment as experienced by a business traveller or newly arrived expat: can I find a good restaurant, will the weather bother me, is there a football match to watch on Saturday.

This is the category that separates Vienna from a perfect score. What keeps Vienna from the top, in recent years, is insufficient availability of large sporting events. The implication is that if Vienna hosted a Grand Prix or an NFL franchise, it would be a better city to live in. Anyone who has actually lived in Vienna and suffered through a forty-degree summer because the city replaced its trees with parking spaces might have a different list of priorities.

Stability, weighted at twenty-five percent, measures the prevalence of petty crime, violent crime, threat of terrorism, threat of military conflict, and, notably, the threat of civil unrest. European cities regularly lose points in this category because of protests. The logic is straightforward from the corporate perspective: if your expat's commute is disrupted by a demonstration, that is an inconvenience to be quantified. That the demonstration might be a sign of a functioning civil society, of citizens exercising fundamental democratic rights, of a political culture healthy enough to permit dissent. None of this enters the calculation. A city where protest is impossible because dissent is crushed scores better on stability than a city where people take to the streets when they disagree with their government.

Infrastructure, weighted at twenty percent, treats road networks and public transport as equivalent subcategories. A city with excellent highways but no usable public transit receives the same score as a city with world-class public transit but narrow roads. This is how American cities, built around the automobile, hostile to pedestrians, with public transport systems that function as last resorts for those who cannot afford a car, end up scoring surprisingly well. Houston, a city where you literally cannot survive without a car, scores 89.3 in infrastructure. The metric does not ask whether the infrastructure serves the population equitably. It asks whether the infrastructure exists, which is a different question entirely.

There is no metric for housing affordability. No metric for housing prices relative to local income. No metric for income inequality or wealth distribution. No metric for job opportunities available to residents rather than imported executives. No metric for social mobility, for how permeable class boundaries are, for whether a child born poor has a realistic chance of not dying poor. These are the most significant determinants of whether a city is livable for the people who live in it. They are absent because the person the index is designed for does not need them. The expat's housing is paid for by the company, the income set by headquarters. The expat does not compete in the local job market. The children attend international schools funded by the relocation package. For this person, housing affordability is irrelevant. For everyone else, it is the single most important factor in whether a city is livable or not.

Sydney and Vienna both sit near the top of the ranking. Sydney's housing market is among the most unaffordable in the developed world, with median house prices exceeding fifteen times median household income. For the index, this does not exist. A city where a nurse or a teacher cannot afford to live within an hour of their workplace receives the same consideration as a city where housing is accessible to the working class, because the index was never designed to notice the difference.

Nassim Taleb observed that IQ tests are not particularly good at identifying genius but reasonably good at identifying the opposite end. The same applies here. The liveability index cannot meaningfully distinguish between the top-ranked cities. The difference between Vienna and Melbourne is 0.1 points. The difference between Vienna and Vancouver is 0.1 points. These are not real distinctions. They are noise dressed up as precision. But at the bottom of the ranking, Damascus, Dhaka, Lagos, Port Moresby, the index does capture something real: these are cities afflicted by war, extreme poverty, or institutional collapse. You do not need a ranking to tell you that Damascus in 2015 was a difficult place to live. But if you are British Petroleum and you need to send an engineer there, you do need a quantifiable metric to calculate how much extra to pay him for the displeasure. That is what this index provides.

The ranking is a measure of the bottom, not the top. It can tell you that Tripoli is worse than Tokyo. It cannot tell you anything meaningful about whether Vienna is better than Melbourne or Copenhagen is better than Zurich. The ranking's attempt to order the cities at the top is like using a scale designed to weigh trucks to determine which of two envelopes is heavier. The instrument is not built for that resolution.

And yet I hear this ranking quoted constantly. Vienna, the most livable city in the world. It appears in tourism campaigns, in newspaper headlines, in political speeches. No one who cites it has read the methodology. No one who reports it has examined the metrics. The phrase "most livable city" is taken at face value because it sounds like it should mean something, and questioning it would require the minor inconvenience of reading a methodology document rather than a press release.

Whether Vienna deserves to be called a highly livable city depends on who you are. Some cities are extraordinarily livable if you are in the top one percent of income. Others are bearable at the top and brutal at the bottom. Some cities manage to be livable even for those in the lower half of the income distribution. A ranking that does not account for this variation is not measuring livability. It is measuring comfort for a specific class of person and calling it a universal verdict.

The suggestion that all Vienna needs is a major football league and it would achieve a perfect score is so removed from the reality of the city that it serves as its own critique of the methodology. The things the index cannot see are the things that define daily life. Vienna has barely any trees in its inner districts; in summer the city bakes. Where trees might stand, parked cars line every block. Corruption and nepotism run through its institutions. People are frequently hostile to non-Austrians. Public transport is adequate but far from what a city of its wealth could achieve. The city remains deeply awkward about its antisemitic past, hyper-conservative in its social norms, and segregated by nationality and class in ways that determine who gets which jobs and which schools. It is a great city. But it has a long way to go before earning full marks in any metric that measures what living in a city is actually like.

But don't listen to me. Just send your expats here. They will love it.

Footnotes

1. According to the 2015 EIU report. One hundred out of one hundred. The same score as Vienna, Zurich, and Tokyo. ↩