Measuring Happyiness and Wellbeing in Smart Cities

Lisbon Case Study

Joana Branco Gomes, João Sousa Rego and Miguel de Castro Neto

NOVA Information Management School (NOVA IMS), Universidade Nova de Lisboa,

Campus de Campolide, 1070-312 Lisboa, Portugal

Keywords: Quality of Life, Human Smart City, Urban Data, Data Visualization, Dashboards.

Abstract: This paper presents the results of a data analysis on Lisbon rates of happiness and wellbeing as a measure of

smart cities. To analyse this issue we collected, respectively, objective and subjective data from an open portal

data website and a survey of subjective data filed by the citizens, represented at parish level, using a ranking

of 1 to 5. The 52 datasets of objective and subjective data supported the production of a dashboard at parish

level. The parishes with high performances (Avenidas Novas, Misericórdia, Santo António and S. Vicente)

are all in the centre of the city. One of the possible conclusions is that there is a cluster of higher values in the

city centre, that could be explain for economic reasons and also because to the proximity to city facilities.

1 INTRODUCTION

In recent years, happiness and wellbeing are being

used as a primary indicator of quality of human life

and development. Since 2012, aligned with the UN

and OCED a World Happiness Report (Helliwell, et

al., 2017) has been published with echoes in

government meetings and policies implemented.

Happiness and wellbeing are nowadays

considered a measure of social progress and a goal of

public policy. This information can be used by

governments, communities and organizations, to

enable policies that support better lives. By analysing

several indicators like income, education, health,

among others, it’s possible to have a better inside of

communities welfare, that analysing these indicators

individually.

In February 2017, the United Arab Emirates held

a full-day World Happiness meeting, as part of the

World Government Summit. Now International Day

of Happiness, March 20th, provides a focal point for

events spreading the influence of global happiness

research. The launch of this report at the United

Nations on International Day of Happiness is to be

preceded by a World Happiness Summit in Miami,

and followed by a three-day meeting on happiness

research and policy at Erasmus University in

Rotterdam.

OECD Better Life Index - www.oecdbetterlifeindex.org

Nevertheless, happiness and well-being are a

subjective and complex concept to calculate. There

are several possible methods and data samples to

measure it. Some data is objective, like

unemployment rate, distance from services, scholar

dropout rate, … Other subjective like the perceived

education conditions (that can by higher or lower than

the actual education conditions). There are several

ways of measuring. OECD

, Eurostat

, and World

happiness report (Helliwell, et al., 2017) use different

ways of measuring and divide happiness and well-

being in different categories.

Besides the challenge of calculate subjective data,

there is a technological challenge of keep the data

actual, meanful, and useful in an automatic way.

1.1 City of Lisbon Case Study

Most research on happiness and wellbeing are

country oriented. The World Happiness Report

analysis indicators such as: Log GDP per capita,

Social support, Healthy life expectancy at birth, and

adaptation, Freedom to make life choices, Generosity,

Perceptions of corruption. The results are analysed by

the governments and most measure / public policies

taken into places are introduced as a Smart

Government measure. Such as the creation of

Eurobarameter – Quality of life in European cities -

ec.europa.eu/regional_policy/sources

270

Branco Gomes, J., Sousa Rego, J. and Castro Neto, M.

Measuring Happyiness and Wellbeing in Smart Cities.

DOI: 10.5220/0006771102700277

In Proceedings of the 7th International Conference on Smart Cities and Green ICT Systems (SMARTGREENS 2018), pages 270-277

ISBN: 978-989-758-292-9

websites, portals with public information or digital

services.

In recent years, a growing number of city

governments have been getting into the game of

Happiness and Wellbeing measure. In 2016 In the

European Commission dropped its yearly Flash

Eurobarometer of quality of life in European cities

, a

huge survey of how happy people in hundreds of

cities across Europe are.

The Urban Europe — statistics on cities, towns

and suburbs defines smart cities as a Urban Europe —

statistics on cities, towns and suburbs. In this index,

where Portugal stands in the middle of the boar,

capitals have a clear lead.

Lisbon, as the main city in Portugal, should be

analysed as a smart city that promotes happiness and

wellbeing index.

In this paper, we analysed the city of Lisbon with

data collected from open data platforms and a survey

made by the authors in May 2017.

The data was first analysed with a geographic

information system (ARC GIS) and then visualized

using a dashboarding software (Power BI software).

The objectives of this case study are:

 Create a model to evaluate or Test a

model, usign a mix model from OECD

Beter Life Index 1 and the quality of life

in European cities

parameters

 Identify the parish and categories with

lower and higher performances

 Compare the self-reported results

(subjective data) and objective resuts.

 Identify the most important categories

for the citizens

 Create a hapiness map

 Create an interective hapiness index

The challenge of this paper is collect the data, that

is not available, updated and open. This challenge is

strongly related to open data policies in Portugal.

2 DATA SOURCES

Concerning the objective data of geographic nature it

was imported from the municipality of Lisboa open

data portal

. The no geographic (statistical data) was

retrieved from Instituto Nacional de Estatísitca

and

Eleições – Secretaria Geral da administração interna

and introduced by parish.

The subjective data was collected from a survey

launch online, and answered by 67 individuals (43

Open Data Arc Gis Dataset –

opendata.arcgis.com/datasets/CML

female and 24 male), during 2 weeks in May of 2017.

The age of the persons inquired was mainly between

26-35 (44%) and 36-45 (29%).

This survey has divided in 4 phases. First the

identification (sex, age and postal code); Second the

definition of happiness (select the topics more

important for the personal happiness); Third the city

conditions (9 questions), where the inquires answer

from bad to good in a scale of 1to 5; and Foutrh the

personal life quality (10 questions) where the answer

were also from a scale from 1 to 5.

Both objective and subjective data collected was

introduced in a geographic information system (in

this case ARC MAP), at the parish level.

The data was divided into 11 categories,

according the OECD Better Life Index

and quality of

life in European cities

Figure 1: Old Parish of Lisbon – image from the official

CML website.

Figure 2: Current Parish of Lisbon – image from the official

CML website.

INE – Instituto Nacional de Estatística – www.ine.pt

www.eleicoes.mai.gov.pt

Measuring Happyiness and Wellbeing in Smart Cities

271

Table1: Categories and data.

CATEGORY

MEASURE

OBJECTIVE DATA

SUBJECTIVE DATA

Housing

Housing spending’s

(€/parish)

Overlapping houses

(%/parish)

Self-reported housing

conditions

Jobs

Unemployment rate

(%/parish)

Income

Income / Job security

Education

Distance to public

schools

Illiteracy (%/parish)

Scholar dropout

(%/parish)

Self-reported

education conditions

Health

Distance from public

hospitals and health

centres

Self-reported health

Environment

Self-reported

environment quality

Self-reported cleaning

conditions

Safety

Self-reported safety

outside home

Community

Trust in people

Civic engagement

Voters percentage

(%/parish)

Trust in government /

city hall

Trust in public

services

Work-life balance

Distance to subway and

trains stops

Distance from

commercial areas

Distance from sports

facilities

Distance from

playgrounds

Distance from cultural

facilities

Distance from a green

area

Building degradation

(%/parish)

Working hours

Time devoted to

leisure

City infrastructure

Self-reported public

transport quality

Self-reported sports

facilities quality

Self-reported cultural

facilities quality

Self-reported green

areas and leisure parks

quality

Self-reported streets,

buildings, and public

spaces conditions

Use of green, cultural

and leisure spaces

In 2015 the city of Lisbon create a new map of the

city, redefining the parishes, and decentralization

competences. The parishes pass from 53 to 24. Since

the data collected for this paper refers to data prior

and after 2015, it was used the old parishes limits,

with 53 parishes + 1 (Parque das Nações), the only

new parish created in 2015.

This paper also should considerer social media

data mining in order to find how people feel regarding

some subjects and also regarding their lives.

3 DATA ANALYSIS

The data (52 datasets of both objective and subjective

data) was introduced in a geographical information

system (ARC GIS) in order to relate and compare the

data using geoprocessing capabilities. The

representation of the results was made using ARC

GIS for mapping purposes and Power BI for

dashboarding.

To compare datasets from the same category

different representations were used on ARC GIS. The

distance to facilities was created using a buffer

according the values defined by Adrian Pitts in

Planning and design strategies for sustainability and

profit: pragmatic sustainable design on building and

urban scales (Pitts, A., 2004) (on education - 300m

for pre-scholar, 1,5km for 1º and 2º grades and 3km

for 3º and high school; On health 1km for health

centres and 2km for hospitals; For other facilities -

500m for transports, 600m for green areas and

playgrounds, 2km for sports facilities and 2,5km for

culture and commerce).

Regarding the subjective data the correlation was

made by creating and normalizing values, by parish.

It’s possible to identify:

 There is a correlation between house

spending’s and housing overlapping (the

houses are cheaper in parish where we have

more houses with excess of people). This

could be explained by social-economic

factors.

 There is a clear need for pre-schools around

the city

 The city is almost covered for 1º, 2º and 3º

grade schools.

 The area of Monsanto Park isn’t covered in

several facilities buffer, but since is a park

without housing, services or other living

spaces, it wasn’t considered to define the

best and worst performances.

 The city centre is well covered of health

facilities, but the peripheral parish no.

 The city is well covered for commerce,

cultural and sport facilities.

 The public transportation doesn’t cover all

the city, but this paper doesn’t have the bus

SMARTGREENS 2018 - 7th International Conference on Smart Cities and Green ICT Systems

272

network, in this sense, this category is

incomplete.

 The Self Reported Security on Job/Income is

the category with most differences of results,

where we can find parishes with the 5

different scores.

 Civic Engagement has lower results,

followed by Trust in Community and

Environment.

 Work-Life Balance has clearly better

scores. This could be explained because on

this category the best results are the average

score, and people have tendency to not give

the higher and lower score, staying on the

middle. Health and Safety also have higher

scores then the others categories.

3.1 Rates and Performances

The performances rates regarding the categories, the

parishes, the relation between objective and

subjective data among other, is information helpful

for governants and the city hall. With this information

is possible to have measure progress, identify city

problems, support better lives, make better and more

informed managing decisions.

On the geographical information system built it’s

possible to identify and compare the parishes /

categories with higher and lower performances.

These values were defined by finding the difference

between the parish value and the average value of all

parishes. Because all the data has different measures

and scales the value used to define where a parish

performed worst or better in each category was

different. For instant on the education and health, if

the parish has 100% of the area covered by the buffer

the parish performance is Best, if 30% or more of the

parish area is not covered by the buffer is rated Worst.

Regarding subjective data, and since is rated 1 to

5 from the survey, the parish that have a difference of

more than 0,5 or less then -0,5 were classified has

Best or Worst.

3.1.1 Parishes with Higher and Lower

Performances

Looking at the results it’s possible to identify that the

centre has best performs, except on building

degradation. The lower performances are usually on

peripheral parishes.

The parish with lower performances are Carnide,

Beato followed by Ajuda, Alcântara. Marvila, Parque

das Nações and Santa Clara. None of this parish are

in the city centre. The parish with high performances

are Avenidas Novas, Misericórdia, Santo António

and S. Vicente (by order). All of this parish are in the

centre of the city. It’s possible to conclude that there

is a cluster of higher values in the city centre, that

could be explain for economic reasons and also

because to the proximity to city facilities.

This information is represented in ARC MAP (for

work and planning purposes) and Power BI

(dashboard for explaining, correlate, visualize and to

support citizens to understanding).

Figure 3: Dashboard - Parishes with best and worst

performances.

3.1.2 Categories with Higher and Lower

Performances

The categories with lower results are Education and

Health, followed by Civic engagement. The

categories with higher results are Community,

Housing and Jobs. It’s possible to conclude that there

is a need to guaranty a more accessibility to health

and education facilities, specially in the parish outside

the city centre. There is also a need to act on the civic

engagement and in the trust in public services and

people.

Most of the categories have differences between

parishes, except City Infrastructure, where the

parishes are balanced between each other. It’s

possible to find clusters of parishes in individual

categories (for example people on the centre feel

more safe them people from peripheral parish).

Figure 4: Categories with higher and lower performances.

Measuring Happyiness and Wellbeing in Smart Cities

273

3.1.3 Self Reported vs Objective Data

It’s important to correlate the self-reported /

sentiment about a category and the objective data of

the same category. This can only be done for

Housing, Education, Health, Civic engagement and

City Infrastructures, because of the lack of available

data. It’s possible to conclude:

 On Housing its possible to conclude that the

housing spending’s and the housing

overlapping are not the most important

feature that contribute for the individual

sentiment of happiness regarding housing.

 On Education, as opposite, there is a

correlation between objective data and

subjective data. People living on Ajuda,

Alcântara, Beato, Campolide and Carnide

parishes felt more unhappy related to

educations and also have lower

performances in scholar dropout, illiteracy

and area in the parish outside the school

influence ratio. On the other hand people

living in Misericordia and Avenidas Novas,

have higher performance and also fell more

happy about the education conditions.

 On Health the people from parishes of

Avenidas Novas, Misericordia, S. António

and S.Vicente with best performances

(closest to health facilities) feel healthy. But

in other hand the parishes of Penha de

França e S. Maria Maior, that also are close

to health facilities, but where people don’t

fell healthy. Since our survey group is

young and mostly healthy is difficult to find

a correlation between this data.

 On Civic Engagement there isn’t a

correlation between the voters percentage

and the trust in public services and people.

The same happens in city infrastructure.

It’ possible to conclude that the survey should

have more respondants in order to be more accurate,

as also it should have more people with different ages.

3.1.4 Categories more Important to People

On the survey people were asked to identify the 3

most important categories for their one happiness:

 Work-live balance was one of the

categories most important, with 48% of the

inquires choosing this category.

 Health with 44%,

 Housing with 39%

 Income 38%.

The Work-Life Balance is composed by the

average of two questions in the survey: How many

hours a day do you work, and how many hours a day

you spend on leisure and sport. Only 23% of people

answers that work between 4-8h and only 15%

answered they spend 2-4h in leisure and sport. The

percentage of people working more that 8h a day was

72% and spending less then 2h in leisure and sport

was 85%. Its possible to conclude that the target

group works more hours that usually and spend less

hours that it supposed to in leisure. This could be

explained by the age of the group, that are young

people making an effort to grow in their jobs. The

parish more balanced are Carnide, Marvila, Olivais

and Parque das Nações, probably because people on

this neighbourhood have more income, and can afford

to work less hours, and because of the opposite, we

have people precarious part-time jobs that gives them

more time available for leisure. The parishes with

lower performance are Ajuda, Alcântara, Santa Maria

Maior, Santo António e S. Vicente, located on the

center and the ocidental part of town.

Figure 5: Work-life balance performance.

The people inquired rate Health as the second

most important category, but in general they feel very

healthy. This could mean that although they are

healthy (probably because of their young age), they

think there is a need to improve in health care

systems. There are parishes with no access to both

health centres and hospitals: Beato, Benfica, Marvila,

Olivais, Parque das Nações and Santa Clara. This

parishes are not located on the city centre, and there

is a predominance of parishes in the occidental part of

town.

SMARTGREENS 2018 - 7th International Conference on Smart Cities and Green ICT Systems

274

Figure 6: Health performance.

Figure 7: Housing performance.

Figure 8: Income / Jobs performance.

Regarding to housing people feel the housing

conditions are good, but we identify some parish with

overlapping housing conditions, that perform well

regarding the housing spending’s. This is correlated

with lower income groups and social

neighbourhoods.

Income / Jobs is the last category with relevant

value to be analysed in detail. The Unemployment

Rate and also the Security in Their Jobs is lower in

parish with lower income groups. Its possible to find

a relation in most of the parish between jobs and

income, that means people that live in parishes with

low unemployment rate, feel more secure about their

income.

Measuring Happyiness and Wellbeing in Smart Cities

275

3.2 Happiness Map

The happiness map that results from this work can be

seen as a tool for governments, city halls, and also the

citizens by representing all the information collected

(subjective and objective).

This map takes into consideration the categories

that are more important to people. The data was

related by a ranking system by dataset, in order to

compare the several datasets, with different scales

and information’s.

The results in warm colours represent higher

rankings and with cold colours low ranks. It’s

possible to identify that the centre performed better,

and the north and east parts of city performed worst.

These could be explained by social economical

motives as well as the distance to city facilities.

Figure 9: Happiness map.

Clustering the results it’s possible to identify more

clearly the areas in the city, were is necessary to

invest and reduce the difference of happiness and

wellbeing among the citizens:

 The city centre with higher performances

 The parishes of Charneca, Ameixoeira,

Lumiar and Carnide at the north of the city

 The parishes of Marvila, Beato e São João,

to the east of the city.

Figure 10: Happiness map.

3.3 Happiness Index

To provide some added value to the project an

happiness index was created. This index is a tool

meant to be used by the citizen in order to allow them

to create their one happiness dashboards / maps

according to the categories that are more important to

them. This tool could be made available on public

municipal platforms, allowing the citizen to choose

the parish that suits more his personal needs, in order

to achieve higher personal happiness and well-being.

Figure 11: Interactive Index.

Furthermore, this tool allows for the user to

specify which are the variables that he values the most

(subjective and objective) and the results will adjust

automatically.

4 CONCLUSIONS

This project can benefit the city hall, politicians and

decision makers, making the city smarter, happier,

and with less inequality. It also benefits the citizens

since they can access a dashboard simple do

understand and visualize, making their life decisions

easier (example choosing a place to live).

SMARTGREENS 2018 - 7th International Conference on Smart Cities and Green ICT Systems

276

The city centre has the parish with higher

performances, and the parish of Beato, Carnide and

Ajuda are the parish with lower performance.

This study suggest an investment on:

 Children facilities: pre-schools and

playgrounds around the city

 Primary health care centres (centros de saúde).

 Investment on city infrastructures in the

parishes of Carnide, Santa Clara, Beato,

Marvila and Ajuda.

The open data provided by the municipality of

Lisbon demonstrated the enormous potential in its

use, both in the evaluation of public policies and in

the development of solutions for valuing certain

neighbourhoods or even the real estate sector.

However the study allowed to identify at least 50

datasets were required to achieve higher quality

precluding its use in its fullness. Regarding the

survey, some fragilities were identified in the sample

of the answers collected, it would be benefit to have

more answers and also to a more diverse group (about

age, sex and income).

The development of this study would make sense

through a partnership with the municipality, allowing

full access to all information and providing solutions

for the use of open municipal data. In this way a

greater approximation between the released data and

the users of the data is promoted.

This paper will also benefit with data mining

sentiment analysis, besides the surveys.

Finally this study will also benefit if collecting the

data, processing and storing was done in a dynamic

and automatic way.

REFERENCES

Pitts, A., 2004. Planning and design strategies for

sustainability and profit: pragmatic sustainable design

on building and urban scales. Elsevier.

Helliwell, J., Layard, R., Sachs, J. 2017. World Happiness

Report 2017, New York: Sustainable Development

Solutions Network.

Ortiz-Ospina, E., Roser, M., 2013. Happiness and Life

Satisfaction, Rotterdam: Worls Database of Happiness.

Deaton, A., Eisenhower, D., 2014. Using happiness data in

policy making.

Tabor, D., Stockley, L., 2017. Personal Well-being in the

UK.

Veenhoven, R., Rotterdam: World Database of Happiness.

ArcGis – Open Data

J. Sterne, 2010, Social Media Metrics: How to Measure and

Optimize Your Marketing Investment.

Measuring Happyiness and Wellbeing in Smart Cities

277