Livability ranking

Where you live shapes the life available to you. It determines the jobs within reach, the people you meet, the places you frequent, how you spend your free time, and what becomes easy or difficult to do repeatedly.

A good city is therefore not simply pleasant. It expands the set of lives you could realistically live.

This is my attempt to measure that.

Overall

Overall measures the quality and breadth of life a city offers without assuming any particular interest. It combines safety, affordability, healthcare, infrastructure, environment, governance, economic opportunity, amenities, urban quality, and connectivity. Serious weaknesses matter — a city cannot compensate for dysfunction in one area simply by being exceptional in another.

USInternational

Rank	City	Neighborhood	Score
1	Boston	Beacon Hill	89
2	New York	Upper West Side	88
3	San Diego	Little Italy	88
4	Seattle	Queen Anne	87
5	San Francisco	Pacific Heights	87
6	Miami	Brickell	86
7	Chicago	Lincoln Park	86
8	Washington, DC	Georgetown	85
9	Austin	Tarrytown	85
10	Denver	Washington Park	84

Rank	City	Neighborhood	Score
1	Copenhagen	Frederiksberg	93
2	Vienna	Neubau	92
3	Tokyo	Ebisu	91
4	Munich	Glockenbach	91
5	Sydney	Surry Hills	90
6	Melbourne	Albert Park	90
7	Zurich	Kreis 4	89
8	Amsterdam	De Pijp	89
9	Madrid	Salamanca	88
10	London	Marylebone	87

Work

Career

Access to professional opportunity, compensation, industry depth, capital, entrepreneurship, and high-value networks.

USInternational

Rank	City	Neighborhood	Score
1	New York	West Village	96
2	San Francisco	Pacific Heights	95
3	Seattle	Queen Anne	91
4	Boston	Back Bay	90
5	Washington, DC	Logan Circle	88

Rank	City	Neighborhood	Score
1	London	Shoreditch	96
2	Singapore	Tanjong Pagar	93
3	Tokyo	Ebisu	91
4	Paris	Le Marais	89
5	Seoul	Gangnam	87

Affordability

The quality of life obtainable per dollar through housing, purchasing power, taxes, and everyday costs.

USInternational

Rank	City	Neighborhood	Score
1	Raleigh	Oakwood	90
2	Austin	Tarrytown	88
3	Dallas	Lakewood	87
4	Minneapolis	Linden Hills	86
5	Nashville	Sylvan Park	84

Rank	City	Neighborhood	Score
1	Madrid	Malasaña	90
2	Lisbon	Campo de Ourique	87
3	Berlin	Schöneberg	85
4	Seoul	Mapo	84
5	Barcelona	Gràcia	83

Mind

Learning

Access to universities, research, libraries, lectures, classes, intellectual communities, and highly educated peers.

USInternational

Rank	City	Neighborhood	Score
1	Boston	Beacon Hill	97
2	New York	Upper West Side	94
3	Washington, DC	Georgetown	92
4	Chicago	Hyde Park	91
5	San Francisco	Pacific Heights	89

Rank	City	Neighborhood	Score
1	London	Bloomsbury	96
2	Paris	Latin Quarter	94
3	Tokyo	Bunkyo	92
4	Berlin	Mitte	90
5	Toronto	Annex	89

Body

Healthcare

Access to exceptional hospitals, specialists, advanced medicine, research institutions, and breadth of care.

USInternational

Rank	City	Neighborhood	Score
1	Boston	Back Bay	97
2	New York	Upper East Side	94
3	Houston	West University Place	93
4	Minneapolis	Linden Hills	92
5	Baltimore	Fells Point	91

Rank	City	Neighborhood	Score
1	Tokyo	Azabu	95
2	Seoul	Gangnam	94
3	Singapore	Novena	94
4	Vienna	Josefstadt	92
5	Zurich	Enge	91

Fitness

Access to training, sports, active culture, outdoor exercise, recreation, and recovery infrastructure.

USInternational

Rank	City	Neighborhood	Score
1	Los Angeles	Santa Monica	95
2	Miami	Brickell	94
3	San Diego	Pacific Beach	93
4	New York	West Village	92
5	Austin	Zilker	90

Rank	City	Neighborhood	Score
1	Sydney	Bondi	95
2	Copenhagen	Vesterbro	93
3	Barcelona	Barceloneta	92
4	Melbourne	Albert Park	91
5	London	Clapham	90

People

Family

Conditions for raising children through schools, childcare, safety, housing, healthcare, parks, and family infrastructure.

USInternational

Rank	City	Neighborhood	Score
1	Boston	Beacon Hill	92
2	Raleigh	Five Points	91
3	Minneapolis	Linden Hills	89
4	Austin	Tarrytown	87
5	San Diego	La Jolla	86

Rank	City	Neighborhood	Score
1	Copenhagen	Frederiksberg	96
2	Vienna	Hietzing	94
3	Munich	Haidhausen	93
4	Zurich	Enge	92
5	Stockholm	Vasastan	90

Dating

Depth and quality of the dating market through demographics, singles density, social activity, mobility, and opportunities to meet.

USInternational

Rank	City	Neighborhood	Score
1	New York	West Village	97
2	Los Angeles	West Hollywood	94
3	Miami	Brickell	93
4	Chicago	West Loop	90
5	Austin	South Congress	89

Rank	City	Neighborhood	Score
1	London	Shoreditch	96
2	Madrid	Malasaña	94
3	Paris	Le Marais	92
4	Barcelona	Eixample	91
5	Toronto	King West	89

Community

Ease of forming durable friendships, joining recurring groups, developing local belonging, and participating in community life.

USInternational

Rank	City	Neighborhood	Score
1	Austin	South Congress	91
2	Minneapolis	Linden Hills	90
3	Denver	Washington Park	89
4	Boston	South End	88
5	Chicago	Lincoln Park	87

Rank	City	Neighborhood	Score
1	Copenhagen	Nørrebro	94
2	Amsterdam	De Pijp	93
3	Madrid	Malasaña	91
4	Vienna	Neubau	90
5	Melbourne	Fitzroy	89

Scene

Nightlife

Depth, quality, variety, and accessibility of bars, clubs, lounges, live music, and late-night activity.

USInternational

Rank	City	Neighborhood	Score
1	New York	Lower East Side	98
2	Miami	South Beach	96
3	Los Angeles	West Hollywood	93
4	Chicago	River North	92
5	Las Vegas	The Strip	92

Rank	City	Neighborhood	Score
1	London	Soho	97
2	Madrid	Malasaña	96
3	Tokyo	Shibuya	94
4	Barcelona	El Born	93
5	Berlin	Kreuzberg	92

Arts

Access to museums, galleries, architecture, theater, music, cultural institutions, and creative communities.

USInternational

Rank	City	Neighborhood	Score
1	New York	Upper East Side	98
2	Washington, DC	Dupont Circle	95
3	Los Angeles	West Hollywood	94
4	Chicago	Gold Coast	93
5	Boston	Back Bay	90

Rank	City	Neighborhood	Score
1	Paris	Le Marais	99
2	London	Bloomsbury	98
3	Berlin	Mitte	94
4	Madrid	Centro	93
5	Amsterdam	Jordaan	93

Food

Depth, quality, diversity, and accessibility of restaurants, markets, casual food, and fine dining.

USInternational

Rank	City	Neighborhood	Score
1	New York	West Village	98
2	Los Angeles	West Hollywood	95
3	San Francisco	Mission District	94
4	Chicago	West Loop	93
5	Miami	Brickell	91

Rank	City	Neighborhood	Score
1	Tokyo	Ginza	99
2	Paris	Le Marais	98
3	London	Soho	96
4	Madrid	Salamanca	95
5	Singapore	Tanjong Pagar	95

Luxury

Access to premium hospitality, retail, residences, restaurants, clubs, wellness, and private services.

USInternational

Rank	City	Neighborhood	Score
1	New York	Upper East Side	98
2	Miami	Miami Beach	96
3	Los Angeles	Beverly Hills	95
4	San Francisco	Pacific Heights	91
5	Las Vegas	The Strip	90

Rank	City	Neighborhood	Score
1	London	Mayfair	98
2	Paris	8th arrondissement	98
3	Dubai	Downtown Dubai	97
4	Singapore	Orchard	95
5	Zurich	Enge	94

Place

Climate

A broadly desirable combination of mild temperatures, sunshine, outdoor usability, and limited weather extremes. Climate is inherently preferential — this ranking assumes a general preference for mild, sunny weather.

USInternational

Rank	City	Neighborhood	Score
1	San Diego	La Jolla	98
2	Los Angeles	Santa Monica	96
3	Honolulu	Kakaʻako	94
4	Miami	Miami Beach	91
5	San Francisco	Pacific Heights	90

Rank	City	Neighborhood	Score
1	Sydney	Bondi	96
2	Lisbon	Estrela	95
3	Barcelona	Eixample	94
4	Madrid	Salamanca	92
5	Melbourne	Albert Park	91

Nature

Access to exceptional natural environments including mountains, beaches, water, wilderness, trails, skiing, hiking, and outdoor recreation.

USInternational

Rank	City	Neighborhood	Score
1	Honolulu	Kakaʻako	97
2	San Diego	La Jolla	95
3	Seattle	Queen Anne	95
4	Salt Lake City	Sugar House	94
5	Denver	Washington Park	93

Rank	City	Neighborhood	Score
1	Vancouver	Kitsilano	98
2	Sydney	Bondi	97
3	Auckland	Ponsonby	96
4	Cape Town	Sea Point	96
5	Barcelona	Gràcia	92

Urbanism

Quality of daily urban life through walkability, transit, density, mixed use, architecture, public realm, street life, and car independence.

USInternational

Rank	City	Neighborhood	Score
1	New York	West Village	98
2	San Francisco	Hayes Valley	94
3	Boston	Back Bay	93
4	Chicago	Lincoln Park	92
5	Washington, DC	Dupont Circle	91

Rank	City	Neighborhood	Score
1	Amsterdam	Jordaan	99
2	Copenhagen	Vesterbro	98
3	Paris	Le Marais	97
4	Vienna	Neubau	96
5	Tokyo	Ebisu	96

Access

Travel

Ease of reaching other places through international breadth, domestic breadth, nonstop destinations, frequency, geographic position, and airport access.

USInternational

Rank	City	Neighborhood	Score
1	New York	West Village	98
2	Miami	Brickell	96
3	Chicago	West Loop	95
4	Atlanta	Midtown	94
5	Dallas	Uptown	93

Rank	City	Neighborhood	Score
1	London	Marylebone	99
2	Amsterdam	De Pijp	97
3	Paris	Le Marais	96
4	Frankfurt	Westend	96
5	Dubai	Downtown Dubai	95

How it works

The ranking is a constrained, multi-attribute utility model evaluated over city–neighborhood pairs. Raw observations are not aggregated directly. Each is transformed into a normalized utility score, grouped into latent subdimensions, weighted according to the objective being evaluated, and aggregated geometrically.

Utility functions

Let $x_{icn}$ denote the observed value of indicator $i$ for city $c$ and neighborhood $n$ . Each observation is transformed onto a common 0–100 utility scale:

s_{icn}=u_i(x_{icn}), \qquad s_{icn}\in[0,100]

The transformation $u_i$ depends on the economic meaning of the variable rather than its measurement scale.

For monotonic variables with a sufficiency threshold:

u_i(x)=100\cdot\operatorname{clip}\left(\frac{x-F_i}{T_i-F_i},0,1\right)

where $F_i$ is the floor below which the variable contributes no utility and $T_i$ is the point at which additional abundance ceases to materially improve livability. The transformation is reversed for variables where lower values are preferable.

Variables without a natural saturation point use diminishing-return functions. Variables such as temperature, humidity, commute distance, and neighborhood intensity use ideal-point functions in which utility declines with distance from a preferred interval. Risk and burden variables use nonlinear penalty functions where deterioration becomes increasingly costly near unacceptable levels.

The model scores marginal utility, not raw abundance.

The tenth direct international destination may materially improve connectivity — the hundredth unit of broadband speed may not. A temperature can be too high or too low. Crime can become disproportionately consequential beyond a threshold. Treating these relationships as linear would imply preferences the model does not actually hold.

Indicator hierarchy

Indicators are not allowed to acquire additional importance merely because several datasets measure the same underlying phenomenon.

They therefore enter the model hierarchically:

\text{Indicator}\rightarrow\text{Subdimension}\rightarrow\text{Dimension}\rightarrow\text{Composite}

For subdimension $g$ :

S_{gcn}=\frac{\sum_{i\in g}a_i s_{icn}}{\sum_{i\in g}a_i}

where $a_i$ is the measurement weight assigned to indicator $i$ .

A dimension $k$ is then constructed from its constituent subdimensions:

D_{kcn}=\frac{\sum_{g\in k}b_g S_{gcn}}{\sum_{g\in k}b_g}

This prevents correlated indicators from being counted as independent sources of utility. Five measures of labor-market strength can improve the estimate of labor-market strength — they do not make labor markets five times more important.

Interest scores

Each interest is represented by a predefined weight vector over the variables relevant to that opportunity set.

Let $I_j(c,n)$ denote the interest-specific composite for interest $j$ . The published interest rankings allocate 85% of the objective to the interest itself and 15% to general city quality:

R_j(c,n)=100\exp\left[0.85\ln\left(\frac{\max(I_j(c,n),1)}{100}\right)+0.15\ln\left(\frac{\max(Q(c,n),1)}{100}\right)\right]

The general-quality term acts as a regularizer. It prevents a city from ranking first in an interest solely because it has an exceptional specialized ecosystem while being materially dysfunctional as a place to live.

The interest term remains dominant. Fitness measures the opportunity to maintain an exceptional fitness-oriented lifestyle — not obesity prevalence. Healthcare measures access to exceptional medicine — not life expectancy. Learning measures access to intellectual institutions and activity — not average educational attainment. Travel measures the destinations and frequencies available to a resident — not airport passenger volume.

Overall score

Overall does not assume a specialized objective. It estimates the breadth and quality of the opportunity set available to a general resident.

The composite incorporates safety, affordability, healthcare, infrastructure, environment, governance, economic opportunity, human capital, amenities, urban quality, and connectivity.

Let $D_k(c,n)$ denote the normalized score for dimension $k$ , with weights $w_k$ satisfying:

\sum_{k=1}^{K}w_k=1

The overall composite is:

C(c,n)=100\prod_{k=1}^{K}\left(\frac{\max(D_k(c,n),1)}{100}\right)^{w_k}

The geometric form is intentional. An arithmetic mean permits near-perfect substitutability between dimensions. A geometric mean does not.

A city with exceptional career opportunity, restaurants, and connectivity cannot completely compensate for severe deficiencies in safety, affordability, healthcare, or infrastructure. Conversely, a city does not rank highly merely because it is inexpensive, safe, and administratively functional if its broader opportunity set is shallow.

Strengths increase a city's score. Weaknesses constrain it.

Constraints

Hard constraints are evaluated before utility maximization.

For a user-specific constraint set $\mathcal{H}$ :

E(c,n)=\prod_{h\in\mathcal{H}}\mathbf{1}\left[x_{hcn}\in A_h\right]

A neighborhood that violates any hard constraint leaves the feasible set. This applies to requirements such as residency eligibility, maximum housing cost, minimum safety, climate bounds, language requirements, tax constraints, maximum commute, or proximity requirements.

A dealbreaker is not represented by an unusually large negative weight. Doing so would allow unrelated strengths to compensate for something the resident has already defined as non-negotiable.

Neighborhood optimization

A city is not represented by a single predetermined neighborhood. Let $\mathcal{N}_c$ be the candidate neighborhood set for city $c$ . The feasible neighborhood set is:

\mathcal{N}_c^{*}=\{n\in\mathcal{N}_c:E(c,n)=1\}

The reported city score is then:

C_c=\max_{n\in\mathcal{N}_c^{*}}C(c,n)

and, for interest $j$ :

R_{jc}=\max_{n\in\mathcal{N}_c^{*}}R_j(c,n)

City-level variables remain constant across neighborhoods. Local variables — housing cost, crime, walkability, transit access, amenity density, commute structure, parks, and social texture — vary with $n$ .

The model therefore asks whether a sufficiently good configuration of that city is realistically available to the person evaluating it rather than whether the average resident of a metropolitan area has a good experience.

Sensitivity

The displayed score is a point estimate produced by one specification of weights, thresholds, utility curves, and observed inputs. Small differences should not be interpreted as exact differences in latent livability.

For uncertain parameter vector $\theta$ , the ranking is more properly written:

C_c(\theta)

Sensitivity analysis perturbs plausible values of $\theta$ — including weights, thresholds, rubric scores, and uncertain observations — and recomputes the ordering.

Two cities separated by one or two points should generally be treated as effectively tied unless the ordering remains stable across plausible specifications.

The useful distinction is not between a score of 87 and 88. It is between results that are structurally robust and results that depend on fragile assumptions.