Patterns of public eService development across European cities

1

2nd International EIBURS-TAIPS conference on:“Innovation in the public sector

and the development of e-services”

Patterns of public eService development across European cities

Davide Arduini, Annaflavia Bianchi, Alessandra Cepparulo, Luigi Reggi and Antonello ZanfeiEiburs-TAIPS Team, University of Urbino, Italy

[email protected]

University of UrbinoApril 18-19th, 2013

•Focus and motivation

•Novelty of this line of research

•Research questions

•Background literature

•Data and indicators

•Cross-country comparisons of eService development

•Analysis of eService development at the city level

• City characteristics and the development of eServices

•Conclusions and implications

Outline

This presentation evaluates public eService diffusion as part of a smart growth strategy in Europe

• eService development across Europe is a key aspect of innovation in the public sector and contributes to EU long term competitiveness.

• extant benchmarking and studies can hardly account for the actual patterns of public eService development for several reasons:- they most often focus on eGovernment, largely disregarding other

web based public activities - even when attention is given to other eService categories, these are

not examined with comparable methods- comparative studies are mostly focused on national patterns, with

limited attention to the regional or local level of analysis

Focus

… and motivation

• A novel dataset (EIBURS-TAIPS Database) providing comparable data at the following levels:

- across four service categories (eGovernment, eProcurement, eHealth and Intelligent Transport Systems)- across countries: EU15 member states- across 229 large and medium sized cities in EU15 countries

• An in-depth analysis of heterogeneity of public eService diffusion at all three levels (across service categories, nations and cities)

• An exploratory analysis of the characteristics of European cities associated with public eService development

•A focus on the links between public eServices and the “smartness” of cities

Novelty of this research line

• How heterogeneous is public sector innovation across EU15 nations in terms of four public eServices ?

•Is heterogeneity higher across countries or cities?

•Does heterogeneity persist when considering clusters of cities with comparable levels of socio-economic development across Europe?

•Are Smart cities also best performers in terms of public eService development?

•What characteristics of European cities are associated with public eService development?

Research questions

Background literature (1/4)• Fast growing literature on the diffusion of public eServices (Arduini&Zanfei 2011)• Frequent use of composite indicators (CI)in this field (European Commission 2001-

2010, UN 2001-2010)• Most studies focus on eGovernment, very few deal with other public e-services

and none assesses national or regional performances across different public e-services.

• While there is a relatively long tradition of CIs designed at the national level, their use at the regional and local level is still largely under-developed.

• Recent exceptions: • EC (2009) for eProcurement, and CapGemini (2010) for eGovernment have introduced

for the first time a regional focus of analysis. • Academic papers have developed CIs using data at the local level, but based on

individual services and with low impact at the policy making level (Baldersheim et al. 2008, Flak et al. 2005, Arduini et al. 2010, Codagnone &Villanueva 2011).

• T

• The focus on individual services at the country level, combined with data constraints encountered at a more disaggregated level, has long led to:

• Erroneously conclude that evidence on e-service diffusion in one area of public sector activity can be used to make inference on e-service diffusion in other areas.

• disregard the extreme heterogeneity of regions in terms of e-service development

Background literature (2/4)• Need to better capture heterogeneity in public sector innovation • Two academic traditions have been feeding the discussion concerning urban innovation

and the role of public sector in it: 1) Systemic theory of innovation; and 2) the literature of Smart cities

1)Systemic theory of innovation

• The systemic theory of innovation was initially formulated at the national levelLundvall (1992), Nelson (1993) and Edquist (2005, 2008)

• Gradual shift towards the regional and local levels(Braczyk et al., 1997; Cooke and Morgan, 1998; Asheim and Coenen, 2005)

• Innovation as an evolutionary process, as a result of complex interactions among different actors, including public institutions

• Public technology procurement literature emphasises the roles of public sector in systemic innovation (Edquist et al. 2000; Hommen and Rolfstam, 2009)

• Purchaser and end users of technology• Catalysers of innovation

Differences in the nature, behaviour and organisation of players involved, including the public sector, combined with the characteristics of technologies determine a high heterogeneity of innovation processes across countries and regions (Fagerberg, 2005; Tether and Metcalfe, 2004)

Background literature (3/4)2) The literature of Smart cities

• One distinctive feature of smart cities is their performance in the field of innovation (Arribas et al. 2012; Deakin, 2012; Capello et al., 2012)

• Smart cities defined as territories combining: the creativity of talented individuals, the development of institutions that enhance learning and innovation, and of digital spaces facilitating knowledge transfer(Eger, 1997; Graham and Marvin, 2001, Komninos, 2006; Sotarauta, 2010, Boulton et al., 2012)

• The support of local government innovation policies is fundamental to the design and implementation of smart city initiatives (Lindskog, 2004; Lepouras et al., 2007; Ingram et al., 2009; Giffinger and Gudrun, 2010)

• City smartness will likely feed-back on local government capacity to innovate their organisation and activities, including their services

Smarter cities are likely to have more innovative public sector and more advanced public services

Heterogeneity of innovation within regions and across cities as a result of different mix of city level characteristics and evolution

Background literature (4/4)

Based on these converging streams of literature we should thus observe:

• A high heterogeneity in public sector innovation across nations

• Heterogeneity in terms of overall eService diffusion• Heterogeneity in terms of eService portfolio and

specialisation

• Heterogeneity is even higher at the local level, reflecting the variety of actors and local government roles

• Cities will differ in terms of eServices reflecting their smartness

Data and indicatorsOur study combines two datasets:

1) EIBURS-TAIPS Dataset (source: University of Urbino)• Data characteristics: information collected by the TAIPS team through

website-surfing to monitor public e-services provided by local public transport companies, municipalities and hospitals at the city level (15-EU). Data refer to availability of eServices in 2012, corrected to account for standard quality measures.

• Sample design: 229 cities representing the EU15 subset of the 322 cities monitored in Eurostat’s Urban Audit dataset

• Variables: info on the provision of 23 eServices classified into four categories ITS/Infomobility (based on ITIC-Between methodology, 2010) eHealth (Based on Empirica methodology, 2008; and Deloitte

methodology, 2011) eProcurement (based on IDC methodology, 2010) eGovernment (based on Capgemini methodology, 2010)

2) Urban Audit Dataset (source: Eurostat) :• Data characteristics: comparable information on 322 cities, out of

which the EU15 sample of 229 cities is derived

• Sample design: cities included correspond to 20% of the national population the geographic distribution of population within the country (peripheral,

central) the size distribution within countries (medium-sized cities having a

population of 50,000 – 250,000 inhabitants, large cities with >250 000)

Time coverage: six waves 1989 - 1993; 1994 - 1998; 1999 - 2002; 2003 - 2006; 2007 - 2009

Variables: demography, social aspects, economic aspects, civic involvement, training

and education, environment, ICT, travel and transport, information society, culture and recreation

Data and indicators

Data and indicators: E-HEALTHUnit of analysis : hospitals

Service list

Videoconferencing/Video consultations between patients and doctors

Dedicated and formal use of facilities such as consultations between patients (either at home or outside the hospital) and hospital medical staff (for clinical purposes)

Electronic Patient Records (EPR)

A computer-based patient record system which contains patient-centric, electronically-maintained information about an individual’s health status and care. The system allows online access to patients

e-booking Electronic appointment booking system

Online clinical testsComputer-based system for electronic transmission of results of clinical tests. The system allows online access to patients

e-referrals Hospitals offering the possibility to external health actors to make appointments for their patients

Telemedicine service (tele-homecare/tele-monitoring)

The provision of social care at a distance to a patient in his/her home, supported by means of telecommunications and computerized systems

Online chronic disease management Home care services using ICT can contribute to the management of long duration/slow progression diseases

Online ticket payment Hospitals offering web based payment systems for visits and clinical tests

Data and indicators : ITS/INFOMOBILITY

Category Unit of analysis : Local public transport companies

Service list

Public Informed Mobility

Online info to users while travelling

Public transport companies providing online information to users (e.g. waiting times, strikes, delays, failures, etc.)

Online time table consultation Public transport companies offering the possibility to consult the online timetable of public transport network

Online travel planning Public transport companies offering timetables with route planning (travel planner) on the web

Online ticket purchase Public transport companies offering web based payment systems

Private Informed Mobility

Info to car drivers while travelling

Public transport companies providing online information to travelers about traffic or parking

Electronic road or parking toll Public transport companies offering a electronic ticketing system of parking spaces

Data and indicators :E-PROCUREMENT Unit of analysis: Municipality

Service list

eProcurement Visibility

Publication of general information on public procurement General information on public procurement made available on the municipality websites

Publication of notices to official electronic notice boards Official electronic board on the municipality websites where procurement notices are made

Link to e-procurement services Link to a web page (owned by the municipality or by external parties) providing eProcurement services

eProcurement (Pre-Award Phase)e-NOTIFICATION Publication of tenders and procurement notices on the web

Online registration of supplier Creation of user accounts and profiles with related roles

e-mail alerts for suppliers Possibility for the suppliers to receive email alerts about forthcoming calls and notices of their interests

e-SUBMISSION

Assistance services to the supplier E-mail, chat, audio/videoconferencing communication for Question and Answer sessions between eProcurement operators and bidders

Online supplier help session help services to assist suppliers in the preparation of online tender

e-AWARDSOnline information about awarded contracts The website publishes the contracts awarded and their winner

e-auctions Availability of tools to carry out real-time price competitions

eProcurement (Post-Award Phase)e-ORDERING

e-catalogues Online order from e-catalogues through eProcurement website

Electronic market Electronic market hosted by the eProcurement website, for online interaction between buyers and suppliers

e-INVOICINGe-invoicing service E-invoicing services managed by the eProcurement website

e-PAYMENTe-payment service Online payment services, managed by the eProcurement website

Data and indicators: EGOVERNMENT

Unit of analysis : Municipality

Service list

Online local taxes Declaration, payment, notification of assessment

Online registration schoolStandard procedure to register children at kindergarden

Online registration of residence Standard procedure to register the residence in a local area of town

On line payment fines Standard procedure to pay fines at municipal police office

Online personal documents Standard procedure to obtain an international passport and an identity card

Online public libraryStandard procedure to consult the catalogue(s) of a public library to obtain specific information regarding a specific carrier (Book, CD, etc)

Online birth/marriage certificates Standard procedure to obtain a birth or marriage certificate

Online registration of a new company

Standard procedure to start a new company

Measuring service availability and quality

CI pillar eService Availability eService Quality

E-HEALTH 8 eServices considered Not measured

INFOMOBILITY 6 eServices considered Presence/absence of quality features including: multi-channel delivery, advanced functions and applications

E-PROCUREMENT 1 eService considered = eProcurement

Presence/absence of quality features associated with each phase (visibility, pre-award, post-award phases)

E-GOVERNMENT 8 eServices considered Interactivity stages, normalized 0-100% (see CapGemini, 2010)

Construction of a Composite Indicator (CI) -1

An indicator of public eServices development is calculated as the average of a city’s performance in the four domains considered (eHealth, Infomobility, eGovernment, eProcurement)

For each domain, we used existing analytical frameworks in order to: (a) define the different dimensions of phenomena studied, including standard measures of quality of eServices available (e.g. interactivity); (b) define the nested structure of the various sub groups that will guide the aggregation ‐process; (c) select the underlying basic indicators

The indicators obtained are then normalized (MIN-MAX method) in order to make the scores of each city in the four domains fully comparable.

Construction of a Composite Indicator (CI) -2

CI pillar Framework Weighting method Aggregation

E-HEALTH Based on Empirica, 2008; and Deloitte, 2011

Multiple Correspondence Analyis (MCA) applied to the basic indicators. Suitable for dichotomous variables (OCED, 2008).

Arithmetic mean

INFO MOBILITY Based on ITIC-Between, 2010

Non-linear Principal Component Analysis (PCA) applied to the basic indicators. Suitable for qualitative variables (OECD, 2008).

Arithmetic mean

E-PROCUREMENT Based on IDC, 2010

Equal weights Arithmetic mean

E-GOVERNMENT Based on Capgemini, 2010

Weights proportional to eServices interactivity as in Capgemini, 2010

Arithmetic mean

Final aggregation = Arithmetic mean

Cross-country comparisons of eService development

Country index vs EU average index

0

0.2

0.4

0.6

0.8AT

BE

DK

DE

IE

EL

ES

FR

IT

UK

NL

PT

FI

SE

E-services index

EU15

Heterogeneity across eServices and across countries

00.20.40.60.8

1eGOV

Infomob

eHealth

eProcurement

Sweden

SE

EU15

00.20.40.60.8

1eGOV

Infomob

eHealth

eProcurement

Denmark

DK

EU15

Front runnersGroup of countries with all or most e-services supplied above the EU average

00.20.40.60.8

1eGOV

Infomob

eHealth

eProcurement

United Kingdom

UK

EU15

00.20.40.60.8

1eGOV

Infomob

eHealth

eProcurement

Luxembourg

LU

EU15

00.20.40.60.8

1eGOV

Infomob

eHealth

eProcurement

Netherlands

NL

EU15

Good PerformersGroup of countries with one or two e-services supplied above

the EU average

00.20.40.60.8

1eGOV

Infomob

eHealth

eProcurement

Ireland

IE

EU15

Heterogeneity across eServices and across countries

0

0.2

0.4

0.6

0.8eGOV

Infomob

eHealth

eProcurement

Germany

D

EU15

0

0.2

0.4

0.6

0.8eGOV

Infomob

eHealth

eProcurement

Spain

ES

EU15

Group of countries with one e-service supplied above the EU average

-0.1

0.1

0.3

0.5

0.7eGOV

Infomob

eHealth

eProcurement

Belgium

BE

EU15

0

0.2

0.4

0.6

0.8eGOV

Infomob

eHealth

eProcurement

France

FR

EU15

0

0.2

0.4

0.6

0.8eGOV

Infomob

eHealth

eProcurement

Finland

FI

EU15

00.20.40.60.8

1eGOV

Infomob

eHealth

eProcurement

Portugal

PT

EU15

Lagging behind Group of countries with average or below average performance

0

0.2

0.4

0.6

0.8eGOV

Infomob

eHealth

eProcurement

Greece

GR

EU15

0

0.2

0.4

0.6

0.8eGOV

Infomob

eHealth

eProcurement

Italy

IT

EU15

Analysis of eService development at the city level

26

Comparing heterogeneity across countries and across municipalities

EU average

To compare the municipalities in terms of their e-service diffusion and sophistication, we need to refer to clusters of homogenous municipalities

To do so we follow a three step procedure:

1. Drawing data from Urban Audit, we use PCA to identify a few “summary variables” (components) that can be held to be representative of different aspects of municipalities

2. We identify the clusters of municipalities based on the above mentioned components

3. Using Eiburs-TAIPS data on eGov, Infomobility, eProcurement and eHealth at the city level we illustrate how clusters can be characterised in terms of eService development

These comparisons are possible for 148 cities only, due to data constraints

Comparing eServices across cities

First step: Principal Component AnalysisDemographic characteristics:

Percentage of residents over 65Population density: total resident pop. per square km

Infrastructural characteristicsLength of public transport network / land area Percentage of households with Internet access at home

Civil societyParticipation rate at city elections Number of female elected city representatives

Human capitalProp. of working age population qualified at level 5 or 6 ISCED

Economic Characteristics: Gross Domestic Product per inhabitant in PPS of NUTS43 Unemployment rate

Sectoral specialization:No. Manufacturing (and service?)Companies (all sectors?)Number of persons employed in provision of ICT servicesProp. of employment in financial and business services (NACE Rev.1.1 J-K)

Environmental sensibility:Annual amount of solid waste (domestic and commercial) that is recycled

Attractiveness:Total annual tourist overnight stays in registered accommodation

Validated PCA -2

Cluster No. obs

characteristics municipalities

1 7 -Industrial and infrastructural development: Medium high- Share of financial and business service employment: Low

ValenciaSevillaLas PalmasPalma de MallorcaPamplona/IruñaPortoHelsinki

2 1 -Industrial and infrastructural development: High- Share of financial and business service employment: High

Stockholm

3 49 -Industrial and infrastructural development: Medium-low- Share of financial and business service employment: Medium

Bonn,Karlsruhe,Mainz,Kiel,SaarbruckenPotsdam,Koblenz,Rostock,StrasbourgNantes,Lille,Montpellier,Rennes,OrléansGrenoble,Aix-en-Provence,MarseilleCatania,Cremona,Trento,PerugiaAncona,L'Aquila,Campobasso,CasertaCatanzaro,Reggio di Calabria,SassariFoggia,Salerno,Apeldoorn,Malmö,Linköping

Cluster No. obs


4 1 -Industrial and infrastructural development: Very High- Share of financial and business service employment: Low

Barcelona

5 2 -Industrial and infrastructural development: Medium- Share of financial and business service employment: High

Frankfurt am MainLuxembourg (city)

6 9 -Industrial and infrastructural development: Medium High- Share of financial and business service employment: Medium High

Wien,Bruxelles / Brussel,Hamburg,MünchenNapoli,Torino,Firenze,Amsterdam,Lisboa

7 30 -Industrial and infrastructural development: Medium low- Share of financial and business service employment: Medium High

Köln,Antwerpen,Essen,Stuttgart,Leipzig,Dresden,Dortmund,Düsseldorf,Hannover,Nürnberg,Wiesbaden,Lyon,ToulousePalermo,Genova,Bari,Bologna,VeneziaVerona,Trieste,Pescara,Potenza,CagliariPadova,Brescia,Modena,'s-GravenhageRotterdam,Utrecht,Göteborg, Köln

Cluster No. obs


8 49 -Industrial and infrastructural development: Medium low- Share of financial and business service employment: low

Charleroi,Liège,Brugge,NamurAarhus,Aalborg,Bochum,Halle an der SaaleMagdeburg,Moers,Trier,Freiburg im Breisgau,Málaga,MurciaValladolid,Vitoria/Gasteiz,Oviedo,Alicante/Alacant,Vigo,Saint-Etienne,Le Havre,Amiens,NancyMetz,Reims,Dijon,Poitiers,Clermont-FerrandCaen,Limoges,BesançonAjaccio,Saint Denis,Fort-de-FranceTours,Lens – Liévin,NijmegenBraga,Funchal,Coimbra,Setúbal,Ponta Delgada,Aveiro,Faro,Tampere,TurkuJönköping,Belfast,Derry

Comparison of the municipalities across and within clusters in terms of

eServices supplied

Clusters index vs Cluster average index

Cluster 3 – “medium low industrial/infrastructural development and medium low share of business services”

Ranking of cities in terms of eServices

Southern Europe

Northern Europe

Southern Europe

Cluster 6 – “medium high industrial/infrastructural development and medium high share of business services”


Correlation among E-services index- Smart index and its components

Smart index

Smart economy

Smart living

Smart people

Smart governance

Smart mobility

Smart environment

eServices index

0.4* 0.4* 0.47* 0.5* 0.3 0.5* -0.25

Source: European Smart Cities ( Vienna University of Technology , Delft University of Technology and the University of Ljubljana).

*significance level 5%

Given the results of correlations we search for drivers of public eServices based on the

empirical literature on determinants of Smart city development

cf. European Smart Cities (2007), Caragliu, Del Bo, Nijkamp(2011), Caragliu & Del Bo (2012)

In search of determinants of eService index

Determinants of the development of smart cities

- Gross Domestic Product of city/region/country (Euro) - New business that have registered in the reference year*- Self-employment rate- Proportion in part time

- Proportion of population aged 15-64 qualified at tertiary level (ISCED 5-6) living in Urban Audit cities - %- Total book loans and other media per resident*

- Length of public transport network per inhabitant- Number of stops of public transport- Number of deaths in road accidents

- Number of tourist overnight stays in registered accommodation per year per resident population- Total number of recorded crimes per 1000 population- Number of hospital beds - Cinema attendance (per year)*- Theatre attendance (per year)*- Number of museum visitors (per year)

Smart economy component

Smart mobility component

Smart people component

Smart living component* Large number of missing values

Correlations check among our index and the potential determinants

ρGdppp 0.2958*

Log(Selfemploy) -0.4156*

Sqrt(Propparttime) 0.3981*

Sqrt(Bedhospital) -0.2487*

Log(Museum) 0.3502*

Isced56 0.3143*Log(Tourist) 0.2033*

Log( public transport stops) 0.2344*

1/sqrt(Lenght transport per inhabitant) 0.108

Crime 0.1616*Sqrt(Road accidents) 0.0874

iij

ij

ij

i itySmartMobileSmartPeoplgSmartLivinmySmartEconoESindex 4321

Our research question thus translates in an empirical model of the form:

where ESindex is the composite indicator of eService developmentsubscript i refers to cities and supra-script j refers to the a specific component of city smartness

VariablesGdppp: Gross Domestic Product purchasing power paritySelfemploy:Self-employment ratePropparttime:Proportion in part time on total workforcebedhosital:Number of hospital beds Museum:Number of museum visitors (per year)Isced56: saher of population qualified at tertiary level (ISCED 5-6)Tourist:Number of tourist overnight stays in registered per yearPublic transport stops(Stopbsn):Number of stops of public transportLength transport: km of public transportation network per resident populationCrime: Total number of recorded crimes per 1000 populationRoad accident: Deaths in road accidents per year

Testing determinants of e-services index -1

* p<0.05, ** p<0.01, *** p<0.001p-values in parentheses R-sq 0.937 0.935 0.940 N 110 110 110 (0.005) prop 0.0600**

(0.990) self 0.000344

(0.001) (0.000) (0.013) isced56 0.00544*** 0.00684*** 0.00410*

(0.000) (0.000) (0.000) stopbusnm 0.0445*** 0.0462*** 0.0399***

(0.522) (0.802) (0.085) bedhosnm -0.00903 -0.00428 -0.0272

(0.091) gdppps 0.00000183 index_a~e index_a~e index_a~e (1) (2) (3)

* p<0.05, ** p<0.01, *** p<0.001p-values in parentheses R-sq 0.937 0.938 0.937 N 113 113 113 (0.769) prop 0.00630

(0.242) self -0.0252

(0.027) (0.022) (0.022) isced56 0.00402* 0.00397* 0.00413*

(0.000) (0.000) (0.000) stopbusnm 0.0406*** 0.0518*** 0.0401***

(0.011) (0.006) (0.024) crime 0.000890* 0.000928** 0.000875*



(0.014) self -0.0585*

(0.014) (0.087) (0.012) isced56 0.00441* 0.00297 0.00449*

(0.445) (0.114) (0.453) stopbusnm 0.0104 0.0223 0.0103

(0.015) (0.001) (0.033) logmuseu 0.0217* 0.0305*** 0.0214*



(0.135) self -0.0318

(0.001) (0.000) (0.004) isced56 0.00513** 0.00568*** 0.00492**

(0.000) (0.000) (0.000) stopbusnm 0.0454*** 0.0600*** 0.0421***

(0.965) (0.764) (0.706) tour -0.000715 0.00470 0.00597


Testing determinants of e-services index -2

CONCLUSIONS• This presentation fills three gaps:

– Coverage of public eServices beyond eGovernment with comparable data

– Comparing eService development across countries and cities– Linking eServices with smartness of cities

• Heterogeneity in Public eService development is high across countries and across service categories

• Heterogeneity is even greater when examined at the city level and across clusters of relatively homogeneous cities

• Cities from nordic and central European countries are largely ranking high, but there is heterogeneity also across these cities a regional and sub-regional approach needed

• “Smart cities” also exhibit high levels of eService development

• Smart city characteristics that are most associated with public eService diffusion are: human capital and transportation infrastructure development

Thanks

City sample

Code Tot cities50 000 – 250 000

ab. > 250 000 ab.

AT 5 3 2BE 7 4 3

DK 4 2 2

DE 40 18 22

IE 5 4 1EL 6 7 2

ES 23 7 16

FR 35 15 20

IT 32 20 12LU 1 1

NL 15 11 4

PT 9 8 1

FI 4 3 1SE 8 5 3

UK 30 12 18

TOT 229 122 107

Data and indicators

Descriptive statistics

eprocurement 229 .6388128 .2440753 0 1 ehealth 229 .187901 .2463106 0 1 infomob 229 .4696298 .2191175 0 1 egov 229 .6156161 .2136404 0 1 Variable Obs Mean Std. Dev. Min Max

PCA -3

Step 2: cluster analysis

Cluster 7 – “medium low industrial/infrastructural development and medium high share of business services”


Central Europe Northern

EuropeSouthern Europe

Cluster 5 – “medium industrial/infrastructural development and high share of business services”


Ranking: single point clusters: 2 and 4

Northern Europe

Southern Europe

Cluster 1 – “medium high industrial/infrastructural development and low share of business services”


Ranking of outliers

Southern Europe

Central Europe

Documents

Patterns of public eService development across European cities