Methodology
The Cities with the Best Work-Life Balance 2021 ranking reveals the
cities with the best social, cultural and structural systems in
place in order to provide their residents with the most well-rounded
work-life balance, in terms not only of work intensity, but also
livability, well-being and rights. As the third yearly iteration
since 2019, and the second since the pandemic began, the study also
takes into account how the shift to remote-working and the impact of
Covid-19 has changed and continues to affect work-life balance in
major cities around the world.
After reviewing hundreds of global metropolises, a shortlist of 50
of in-demand cities with sufficient, reliable, and relevant datasets
were selected. This included cities known for attracting
professionals and families for their work opportunities and diverse
lifestyle offerings. As the third iteration of a continuous study,
this index includes 10 more cities than in 2019.
This index is not designed to be a city livability index, nor is it
intended to highlight the best cities to work in. Instead, it is
designed to be a guideline which supports the fulfillment of
residents’ lives by improving the aspects of life which help relieve
work-related stress and intensity.
Factors and Scoring
The study was divided into three categories - Work Intensity, Society and Institutions, City Livability - comprising the following 18 factors which contribute to work-life balance during and beyond the pandemic:
Work Intensity: Remote Jobs (%), Overworked Population (%), Minimum Vacations Offered (Days), Vacations Taken (Days), Unemployment (Score), Multiple Jobholders (%), Paid Parental Leave (Days)
Society and Institutions: Covid Support (Score), Healthcare (Score), Access to Mental Healthcare (Score), Inclusivity & Tolerance (Score).
City Livability: Affordability (Score), Happiness, Culture & Leisure (Score), City Safety (Score), Outdoor Spaces (Score), Air Quality (Score), Wellness and Fitness (Score), Covid Impact (Score)
Each factor consists of one or more indicators which were scored and averaged. The equation for scoring is as follows:
z-Score =
x - mean(X)
Standard deviation(X)
in short
x - μ
σ
For columns where a low value is better, the score is inverted such that a high score is always better:
z-Score inverted = -1 *
x - mean(X)
Standard deviation(X)
in short -1 *
x - μ
σ
Data is normalized to a [50-100] scale, with 100 being the best score. Therefore, the higher the score, the better the city ranks for that factor in comparison to the other cities in the index. The formula used is min-max normalisation:
score = (100 - 50) *
x - min(X)
max(X) - min(X)
+ 50
The final score was determined by calculating the sum of the weighted average score of all of the indicators. Below you can find a detailed description of each factor within the study, and the source used.
Work Intensity
Remote Jobs (%)
The quantity of jobs that are workable from home as a percentage
of all jobs. Locations with a greater share of teleworkable jobs
(aka remote working) may provide residents with the ability to
comply with strict social distancing requirements while also
maintaining regular employment and income.
Data taken from the "How many jobs can be done at home?" (Dingle &
Neiman, 2020) white paper, which uses job classification survey
data to identify jobs that can be performed from home, and applies
this to job statistics to estimate the share of all jobs these
jobs account for. US cities employ data at a Metropolitan
Statistical Area level, while country-level data was taken for all
other cities. Where data was unavailable, values for some
countries were modeled using GDP per capita and percentage of
population with university degrees.
Sources: Dingel, J. I., & Neiman, B. (2020). How many jobs can be
done at home?. Journal of Public Economics, 189, 104235; World
Bank – GDP per capita, PPP (current international $), latest data;
World Bank – Percentage of population age 15+ with tertiary
schooling. Completed Tertiary, latest data.
Overworked Population (%)
The percentage of full-time employees working more than 48 hours per
working week. The International Labour Organisation (ILO) recommends
a workweek of 40-hours and considers weekly work of over 48 hours
"excessive".¹ The “Overworked Population” is considered to be the
percentage of full-time employees working more than 48 hours per
working week.
For cities in the United States and in the European Union, average
number of hours of work was incorporated into the country-level data
to approximate percentages on a city-level. For all other cities,
country-level data was used to evaluate the average working hours
per week.
Sources: ILO-STATISTICS – Labour force survey, latest available
data; US Bureau of Labor Statistics – Current Employment Statistics
survey (State & Metro Area), 2019; EUROSTAT - Average number of
usual weekly hours of work in main job by sex, age and NUTS 2
regions.
Minimum Vacations Offered (Days)
The minimum number of compensated vacation days an employee is
legally entitled to after at least one year of service. Data was
taken at a national level for a full-time, five-day workweek
(excluding public holidays). In the US, under the Fair Labor
Standards Act, no such federal or state-level regulations exist that
require employers to pay employees for time not worked, including
holidays.² Despite this, time off agreements are often negotiated
between employer and employee. Data for US cities is based on the
average number of reported paid holiday days for a private industry
employee after their first year of service (10 days per annum).³
Sources: International Labour Organisation; European Commission –
EURES Living and Working Conditions; Thomson Reuters – Practical Law
database; Various national labour departments.
Vacations Taken (Days)
The average number of used paid vacation days offered to full-time
employees in a single year. City-level data was used where
available. For US cities, data was calculated by subtracting the
unused vacation days from the average number of days offered. The
percentage of unused vacation days in the US was sourced at a
state-level. For non-US cities, country-level data was taken on the
number of vacation days used.
Sources: US Travel Association – State-by-State Time Off, 2019;
Expedia – Vacation Deprivation study, 2018/17; UBS – Prices and
Earnings study, 2018.
Unemployment (Score)
The unemployment rate for the metropolitan area or region in the
first quarter of 2021. This factor is expressed as a score where the
higher the score, the lower the unemployment rate. For cities that
have not published their unemployment rate, the rate was estimated
using the quarter-to-quarter trend of the country. In rare
instances, national figures were used. Unemployed persons are
considered those of the labour force who are jobless, looking for a
job, and available for work.
Sources: Official statistical websites of each metropolitan
area/region/country.
Multiple Jobholders (%)
The percentage of employed people holding more than one job at any
one time. The holding of more than one job at a time can be a sign
of engaging in precarious work, and is often used as a financial
coping tactic for those in economically vulnerable positions,
including minorities. Multiple job-holding can also expose workers
to longer hours, lower wages, and compromise their labour
protections. The ILO has voiced concern about incidence of multiple
job-holding, describing it as a possible “sign of persons engaged in
irregular low-productive work, with an overlap to working poverty
and an inability to earn sufficient income on the main job alone.”⁴
Unfortunately, detailed geographical data on the number of
multiple-jobholders is underreported and not regularly published.
However the latest available data where possible was compiled from
official statistics and independent research. All US and Canadian
data is at a state and province level, respectively, while other
cities use national data. Values for Hong Kong and Bangkok are
modelled estimates using national figures for the percentage of
part-time workers as a proportion of the workforce.
Sources: Eurostat – Job-holder survey, 2018; Bureau of Labor
Statistics – Multiple-jobholding rates, 2015; Statistics Canada –
Multiple jobholders, 2019; Singapore Ministry of Manpower – Labour
Force report, 2018; Australian Department of Social Services – HILDA
survey, 2019; Stats NZ – Household labour force survey, 2019;
Statistics Korea – Economically Active Population survey, 2019; The
World Bank – Malaysia Economic Monitor, 2019; Japan Labor Issues
Journal – Atsushi Kawakami: “Who Holds Multiple Jobs?...”, 2019;
Latin American Perspectives – “Precarious Work in Argentina
2003-2017”, 2020; United Nations – Economic and Social Council
Brazil report, 2001.
Paid Parental Leave (Days)
The number of paid family leave days from work afforded to employees
by law. The sum comprises the legislated number of days for paid
maternal, paternal and parental leave, and reflects the number of
days compensated, regardless of benefits provided or level of
compensation. At the federal level, the US does not mandate paid
leave for parents, but some states have recently passed relevant
legislation (these include the states of California, New York,
Hawaii, and the District of Columbia). National data is used, except
for US cities, which use state-level data.
Sources: OECD – Employment statistics database, latest available;
data, ILO – Maternity and paternity at work study, 2014; Thomson
Reuters – Practical Law, 2020; Official local government websites.
Society and Institutions
Covid Support (Score)
The degree of income support provided by governments to workers
affected by the economic effects of Covid-19. This factor is
expressed as a score where the higher the score, the greater the
support. It takes into account government programmes to replace
income lost due to Covid, length of unemployment benefits, consumer
confidence, household spending and general wage levels; as well as
overall spending by governments to dampen the impact of Covid on the
economy. In addition, the level of covid cases and deaths was taken
into account.
Sources: Thomas Hale, Noam Angrist, Rafael Goldszmidt , Beatriz Kira
, Anna Petherick, Toby Phillips, Samuel Webster, Emily
Cameron-Blake, Laura Hallas, Saptarshi Majumdar, and Helen Tatlow.
(2021). “A global panel database of pandemic policies (Oxford
COVID-19 Government Response Tracker).” Nature Human Behaviour
https://doi.org/10.1038/s41562-021-01079-8; IMF - Fiscal Monitor
Database of Country Fiscal Measures in Response to the COVID-19
Pandemic (April, 2021): Above the line spendings Additional -
spending or foregone revenues - Non-health sector (as % of gdp);
OECD - Household Dashboard: Real Household Final Consumption
expenditure per capita, Consumer Confidence; year to year comparison
(Q2 2020/Q2 2019 and Q3 2020/Q3 2019); OECD - Benefits and wages
database, SSA Country profiles, local authorities: Length of
unemployment benefits for a 30 year old single; no children; work
prior to unemployment; five years of contribution; average salary,
full-time; contract terminated because of shortage of work; ILO -
Mean nominal monthly earnings; Worldometers - Total Cases per
million, total deaths per million.
Healthcare (Score)
The measure of a city’s healthcare system based on access, quality
and satisfaction. Country-level data was obtained from the Universal
Health Coverage (UHC) index for access and quality indicators, while
US cities also incorporate state-level data from the Health Access
and Quality (HAQ) study. Additional data was taken from healthcare
access indexes developed by the World Health Organisation and the
European Commission. Satisfaction survey results were taken at a
city level.
Sources: Lozano, R., Fullman, N., Mumford, J. E., Knight, M.,
Barthelemy, C. M., Abbafati, C., ... & Cárdenas, R. (2020).
Measuring universal health coverage based on an index of effective
coverage of health services in 204 countries and territories,
1990–2019: a systematic analysis for the Global Burden of Disease
Study 2019. The Lancet, 396(10258), 1250-1284; Fullman, Nancy, et
al. "Measuring performance on the Healthcare Access and Quality
Index for 195 countries and territories and selected subnational
locations: a systematic analysis from the Global Burden of Disease
Study 2016." The Lancet 391.10136 (2018): 2236-2271; European
Commission — DRMKC - INFORM Risk Index (‘Access to health care’
indicator); 2021., WHO - World Health Data Platform, Universal
Health Coverage Index; latest data; Numbeo – Healthcare Index; data
as of April 2021.
Access to Mental Healthcare (Score)
The accessibility and effectiveness of governments in implementing
mental health policies aimed to care for individuals with mental
health illnesses. This factor uses national data on governance,
access to treatment, and the environment necessary for treatment.
This factor also incorporates suicide rates and city-level survey
data on healthcare quality.
Sources: EIU/Jannsen – Asia- Pacific Mental Health Integration
Index, 2016; EIU/Jannsen – Europe Mental Health Integration Index,
2014; Institute for Health and Metrics Evaluation – Health Access
and Quality Index, 2016; Numbeo – Healthcare Index, 2020; Local
statistics departments.
Inclusivity & Tolerance (Score)
The degree to which a city supports gender and LGBT+ equality, inclusivity and tolerance through legislation and opportunity. The score combines the following ‘Gender Equality’ (degree of gender parity), as well as the ‘LGBT+’ (inclusiveness and tolerance) factors:
Gender Equality:
Gender equality scores were developed using data on the level of
difference in economic opportunity and participation, educational
attainment, health, and political empowerment between men and
women. City-level data was used for US cities, with country-level
data used for non-US cities.
Sources: Economist – Glass Ceiling Index, 2020; World Economic
Forum – Gender Gap Index, 2020; Council on Foreign Relations –
Women's Workplace Equality Index, 2020; OECD – Social Institutions
& Gender Index, 2019.
LGBT+
For LGBT+ scores, the comprehensiveness of equality and protection
(an emphasis on work rights) legislation, health access, as well as
political representation for the LGBT+ community were examined. The
percentage of the population that identifies as LGBT+ was also
included, as environments in which a higher number of citizens feel
comfortable openly identifying as a minority is also a potential
indicator of a tolerant and supportive community.
Sources: SPARTACUS – Gay Travel Index, 2020; Gallup – Daily Tracking
polls, 2015/2017; Out Leadership – State LGBT+ Business Climate
Index, 2019; Local statistics departments, latest available data.
City Livability
Affordability (Score)
Monthly living costs as a proportion of the average household
income, after tax. A basket of estimated monthly costs includes:
rent, basic utilities costs, groceries, internet connection,
leisure activities, clothes, and eating out. A higher score
indicates a higher level of remaining monthly income (if any)
after accounting for these deductions.
Sources: OECD – Employment Database, 2018; Numbeo – Cost of Living
Index, data as of April, 2021.
Happiness, Culture & Leisure (Score)
The degree to which residents are able to enjoy their environment after office hours, measured through the average perceived level of happiness as well as the accessibility and variety of a city’s cultural and lifestyle offerings. The score combines both of the following ‘Happiness’ and ‘Culture & Leisure’ factors.
Happiness
Score includes the average perceived level of happiness at a city
level. In the rare absence of city-level data, national data was
used. The score is calculated from survey responses evaluating the
perceived happiness with one’s own life, as well as the degree of
positive and negative effects a respondent experiences.
Sources: Sustainable Development Solutions Network – World
Happiness Report, 2020; Walethub – Happiest Cities, 2019.
Culture & Leisure
The vibrancy and variety of cultural and lifestyle offerings in a
city. The score combines cultural city rankings, the number of
persons employed in the cultural and creative industries, and the
amount of leisure facilities and activities available, such as the
number of sports stadiums, restaurants, parks, shops, entertainment
and nightlife venues per capita. Cities with an exceptional number
of activities received supplementary points.
Note: all data collected for this factor reflects pre-pandemic
conditions where cultural and lifestyle offerings were available
without restriction. This was designed to measure the vibrancy of a
city's offerings under normal circumstances, with the hope that the
existing cultural framework of a location will allow it to return to
a similar level in the future.
Source: US Bureau of Economic Analysis – State Arts and Cultural
Production Employment, 2016; European Commission – Cultural and
Creative Cities Monitor, 2019; Mori Foundation – Global Power City
Index, 2018; TimeOut – ‘48 best cities in the world in 2019’;
Wallethub – Funnest Cities in the US rankings, 2019; OSM Overpass
Turbo API – Searches included: bars; clubs; pubs; restaurants;
cafes; galleries; museums; and cinemas, latest data; TripAdvisor –
Searches included: Nightlife, Museums, Concerts & Shows, Outdoor
Activities, Nature & Parks, latest data; World Stadiums – Database,
latest data.
City Safety (Score)
The degree of safety provided by a city in more than a dozen key
areas, including environmental, social and infrastructural security.
Indicators include statistics on injuries and fatalities, damage
caused at an economic level, public opinion data, and data on the
vulnerability of a location to particular hazards.
Sources: Germanwatch – Global Climate Risk Index, 2021/2020; United
Nations Office on Drugs and Crime – database; Economist Intelligence
Unit – Safe Cities 2019; European Commission/Disaster Risk
Management Knowledge Centre – INFORM RISK report 2021; Igarape
Institute – Fragile Cities index, 2017; Numbeo - Crime Index; Vision
of Humanity – Global Peace Index, 2020; World Health Organisation -
Global Health Observatory database, latest available data.
Green Spaces and Weather (Score)
The prevalence and accessibility of a city’s urban green
infrastructure as a score, including its proximity to residents and
the percentage of land allocated to green space. Data on weather and
daylight conditions that could affect the use of public outdoor
spaces was also incorporated. This includes average temperatures,
the annual number of rainy days, annual sunshine hours, and
cloudlessness.
Significant weighting is placed on the green spaces indicator, as
the existence of favourable weather alone is not a condition for a
good score in this section. Data is collected at a city level.
Sources: United States Forest Service – iTree survey tool; The Trust
for Public Land – ParkScore index, 2020; OECD – Green area survey,
2018; European Environmental Agency – Urban green infrastructure
database, 2017; Weather Spark – Weather analysis data, 2020.
Air Quality (Score)
Annual median particulate matter (PM2.5/PM10) pollution for the year
2019, represented as a score. Data from pre-pandemic conditions was
used in order to assess a city’s air quality under normal
circumstances, with a view that pollution levels may return to
similar levels in the future should measures not be taken to reduce
them.
Daily average data was taken across all days of a single year, with
the median pollution level representing the overall score. Data was
taken at a city level.
Sources: AQICN – Air Quality Index historical database, 2019; World
Health Organisation – Global Ambient Air Quality Database, 2018.
Wellness and Fitness (Score)
The general state of a community’s physical fitness and health as
represented by a population’s average life expectancy, as well as
levels of inactivity, obesity, and the number of fitness studios and
gyms per capita. National data was used for life expectancy at
birth, while US cities use city-level data. Adult obesity rates and
the prevalence of physical inactivity were taken at a national
level, with US cities using state level data. Data on the number of
gyms per capita is taken at a city level.
Sources: World Health Organisation – Global Health Observatory data
repository, latest data; US Center for Disease Control and
Prevention – Adult Physical Inactivity Prevalence, 2020; Opportunity
Insights – US life expectancy data, 2016; The State of Childhood
Obesity – Adult Obesity Rates, 2019; OSM Overpass Turbo API –
Searches included: ‘leisure=fitness_centre’ and
‘leisure=sports_centre’.
COVID Impact (Score)
The degree of social and economic impact on account of a location’s
Covid-19-related response. This factor is expressed as a score where
the higher the score, the lower the impact. Three dimensions of the
impact were taken into account: public health, economic and social.
The impact on public health is quantified through cases and deaths
relative to population; the impact on economy through year-on-year
GDP growth in 2020 and 2021; and the impact on society through the
severity of limiting measures put in place to contain the pandemic,
as well as changes in mobility patterns as a measure of the effect
of these restrictions.
Sources: Oxford COVID-19 Government Response Tracker, International
Monetary Fund; Apple – Covid-19 Mobility Trends Reports.