DIGITAL DIVIDE SURVEY ANALYSING THE

“.IT” REGISTRAR MARKET

Maurizio Martinelli, Michela Serrecchia and Irma Serrecchia

Institute for Informatics and Telematics, Italian National Research Council, Via G. Moruzzi, 1, Pisa, Italy

Keywords: Diffusion of the Internet, Digital Divide, Domain Names, Registrar, ICT.

Abstract: In this paper the identification of factors contributing to the differences on Internet use (digital divide) in

Italy, during recent years, was reported, focused on the analysis of the market of “Registrars” in the domain

name registration field. In particular a comparison between the results of a similar study, done until

December 2005, and the data of domain names registered by “Registrars” up to December 2010, was carried

out. In Italy, the registration of a new Internet domain name is usually made through a service company

(Registrar) accredited by the National Registry, where all data related to such registration are managed. In

order to analyse the diffusion of the Internet in Italy, the number of Registrars and the number of registered

domains were used as indicators. To define the factors that determine the digital divide the regression

multiple model was utilized, based on the stepwise method. The dependent variable taken into consideration

was the penetration rate at the regional level and the independent variables were regional economic,

cultural, demographic and technological factors. Above all the regions with a low unemployment rate and

with high economic values, such as added value per employee and high per capita income, are more inclined

to use the network. Furthermore, the level of education resulted a decisive factor: as a matter of fact, regions

with a high number of graduates, specialized in ICT fields, are more inclined to utilize the Internet

technology than those that register a number of ICT graduates below the average. Finally, an interesting

result was that, in Italy, a few Registrars register a higher percentage of domain names under the Country

Top Level Domain ".it". The phenomenon of the registration of domain names is concentrated in the hands

of a few Registrars. This aspect was significantly evidenced in 2010.

1 INTRODUCTION

Web has undeniably entered our everyday life

forcefully, ceasing to be an instrument used by a

limited circle of academics to become a new and

versatile example of mass media. According Rogers

(2005) there are five attributes which, on average,

are considered as the most influential for speed of

adoption of a technology across different types of

users: relative advantage, compatibility, complexity,

trialability and observability. Any increase in the

relative advantage over the previous technology, the

compatibility of the new technology with the needs

of potential adopters, the ability of adopters to

experiment with the new technology and the ability

of the users to observe the new technology, are all

attributes that will speed up the diffusion process.

Although Internet has become a worldwide

protagonist of our days, it is not geographically

uniform among countries, as well as within a

country. For example Greenstein and Price (2004)

reported that people living in rural areas might find

greater relative advantages from the use of the

Internet than people living in urban areas.

Furthermore, according to Strover (2001), low

Internet adoption is caused not only by factors such

as lower education and income levels, but also by

socio-economic factors and by low levels of

technological knowledge. Based on the above

consideration, this work wanted to verify if in Italy it

is possible to sub-divide Internet users on the basis

of the categories proposed by Rogers (1995) for the

generic diffusion theory of a technology: innovators,

early adopters, early majority, late majority, and

laggards. For this purpose, the Internet diffusion in

Italy was analyzed by using as indicator the number

of companies (Registrars) that offer, as a service, the

registration of domain names under the country code

Top Level Domain (ccTLD) “.it”. Furthermore, in

this paper the factors contributing to the differences

in Internet use (digital divide) in Italy at a regional

level were defined by verifying the effects on the

593

Martinelli M., Serrecchia M. and Serrecchia I..

DIGITAL DIVIDE SURVEY ANALYSING THE “.IT” REGISTRAR MARKET.

DOI: 10.5220/0003937705930597

In Proceedings of the 8th International Conference on Web Information Systems and Technologies (WEBIST-2012), pages 593-597

ISBN: 978-989-8565-08-2

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

adoption of some local socio-economic hypotheses.

2 METHODS

In order to analyze Internet diffusion in Italy, the

number of Registrars registered up until 31

December 2010 were used as an indicator. The data

were extracted from the databases managed by the

National Registry, at the Institute of Informatics and

Telematics of the National Research Council in Pisa

(IIT-CNR). The database managed by IIT-CNR until

the above date, counted 1,842 Italian Registrars. The

choice of using Registrars as indicators instead of,

for example, the number of domain names, which is

one of the indicators most used in the literature

(Zook, 2000; Bauer et al., 2002) together with the

hostcount (see studies published by Internet

Software Consortium or by RIPE-NCC (Réseaux IP

Européens - Network Coordination Centre) derived

from the fact that the number of Registrars is an

indicator that appears to be more effective in order

to identify the real extent of the Internet

phenomenon in Italy. In fact, Registrars, in addition

to registrating domain names, can supply other

services related to Information and Communication

Technology (ICT), such as Internet connectivity,

selling of hardware and software products, electronic

mail services, website design, and so on. This

research intended to achieve two main goals. The

first, was to analyse the extent of the Internet

phenomenon in Italy, through which it is possible to

compare the Italian situation with the international

one and, therefore, to identify the situation of Italy

within the international ranking of Internet use. The

grouping of data at regional level allows comparison

of the penetration rate of registrars in single

geographical areas and measures the possible

technological gap (the so-called digital divide).

Moreover, it defines the factors that cause the digital

divide in Italy. The second purpose of this research

was to compare the market of Registrars in the

domain name registration field until December 31,

2010, with that resulted at December 31, 2005. The

comparison was made both to verify if, in the

considered 5 years, the market situation of the

Registrar was the same, and to see if there was an

effect due to change in the system of registration of

domain names. In September 2009, the synchronous

system in the registration of domain names ".it"

entered into force. Before September 2009, the

domains were registered with an asynchronous

system. It is necessary to highlight that in this

research only data concerning the number of

registered domain names “.it” were examined. Some

Registrars, as matter of fact, focus their business on

other types of services such as the xDSL access

supply for retail customers, hardware and software

sales, VoIP (Voice over IP) services or e-commerce

and so on. The indicator used to identify in detail the

existence of the digital divide in Italy at the regional

level, was the penetration rate (PR). These entries

allowed to analyze in detail the existence of digital

divide in Italy at local level and also in macro-areas

(North, Center, South). As penetration rate, the ratio

between the number of Registrars and the number of

companies existing in Italian national territory was

used:

PR = (number of registrars / number of

companies) x 10’000

(1)

3 RESULTS AND DISCUSSION

The analysis based on the proposed methods showed

that in Italy, in the analysed period, only some

regions registered the highest penetration rates. In

particular northern and central regions registered

penetration rates greater than one (see Table 1).

Southern region, resulted all below the tenth

position, demonstrating the presence in Italy of a

digital divide.

Table 1: Internet distribution: the first ten Regions are

ordered on the basis of the number of Registrars per

10’000 firms (PR).

Ranking

Region Area Registrars PR

Trentino Alto

Adige

North

60 5.87

2 Lombardy North 459 5.57

Friuli Venezia

Giulia

North

45 4.57

4 Tuscany Centre 166 4.52

5 Umbria Centre 34 4.06

6 Piedmont North 170 4.03

7 Latium Centre 178 3.85

Emilia-

Romagna

North

154 3.59

9 Veneto North 160 3.49

10 Liguria North 49 3.43

Italy 1,842

In order to reach the second purpose of the study,

the concentration of the number of domain names

registered by Italian Registrars in the different

regions, were identified by using two concentration

indexes: the Herfindahl-Hirschman index (HHI)

WEBIST2012-8thInternationalConferenceonWebInformationSystemsandTechnologies

594

(Hirschman, 1964) and the Gini concentration index

(Gini, 1912). The HHI index, widely used in

literature, measures the degree of competition in the

market. Considering an industry with N firms it is

possible to measure the market share of each firm,

HHI is calculated by adding the square of the market

shares of each firm as:

HHI = Σ S

i=1

(2)

where S

are the market share of each firm measured

in percentage terms. For example, in the case of a

market formed by four firms with shares

respectively of 30%, 30%, 20%, 20%, HHI is equal

to 2600 (30

+ 30

+ 20

). The index is

structured in a way that it increases both when the

number of firms in the industry decreases and when

the gap between firm size widens. An HHI index

lower than 1000 indicates a market that is close to a

competitive context. The markets in which HHI

ranges from 1000 to 1800 are usually considered

moderately concentrated. If HHI is greater than

1800, the degree of monopoly power becomes more

significant.. The HHI index, calculated for

Registrars at national level at the end of 2010,

resulted in Italy 1,389.92. The index, compared with

data of 2005, resulted increased, because both the

number of Registrar decreased and the number of

registered domains increased. Therefore it is not

possible to talk about monopoly, and moreover the

number of firms at national level proves to be high

(1,842 Registrars). The Gini concentration index,

unlike HHI, is a standard index, which ranges from 0

to 1. The Gini index is equal to 1 in case of

maximum concentration (this happens when, for

example, considering income distribution in a

country, only one individual earns the entire amount

of income), while it is 0 in a situation of even

distribution (all individuals earn the same level of

national income). Given its feature, that index is

widely used in statistics literature because it renders

better the concentration measurement in concrete

situations and it is specially suitable for comparing

the degree of concentration among heterogeneous

situations. The Gini index at national level was

calculated on the basis of the number of registered

domain names, and it resulted 0.91, and so higher

than in 2005, indicating that it is not possible to state

that, in Italy, only one Registrar registers all the

domain names under the ccTLD “.it”. However, the

value 0.91 is justified by the fact that only 10

registrars out of 1,842 register 63.83% of total

domain names. In 2005, 10 registrars out of 2,552

registered 46.30% of domain names. Therefore both

the indexes show that the concentration of domain

names registered by Registrars, from 2005 to 2010,

was increased. The analysis of concentration

resulted more clear when, in particular, the three

macro-areas North, Center and South were analyzed.

In all three areas, both indexes are increased (Tables

2 and 3. However, especially in the Centre of Italy,

the Gini index and the HHI in 2010 increased

considerably. In particular, HHI increased almost

twice, comparing data of 2010 with those of 2005.

This depends on the fact that, in such period, the

number of Registrars of the Center decreased while

the number of registered domains increased, and

also because the gap among Registrars in the

registration of domain names increased. In fact, in

2010 the Range (the difference between the

minimum and maximum of domains registered by

Registrars) is higher than in 2005 (Tables 2 and 3).

The Gini index of 0.96, in 2010, indicated that only

few Registrars register the total amount of domain

names under the ccTLD .it. As a matter of fact,

analyzing the data at an individual level, in the

Center, only two Registrars out of 427 register more

than 70% of domain names, 74.17% out of the total

amount of domain names. In 2005, two Registrars

out of 561 registered more than half the domain

names, 55.10% out of the total amount of domain

names. The North, on the contrary, is the region in

which there is more competition compared to the

other macro-areas and, as resulted, the two

concentration indexes are lower than in the Center

and in the South: the Gini is 0.80 and HHI is 228.78,

Table 2: Analysis of concentration of domain names “.it”

registered by Registrars 31-12-2005.

Macro

Area

Gini

Index

HHI

Index

Registrar

Domain

names

Range

North

0.78

103.51

1575 334350 14313

Centre 0.93 1838.44 561 544874 210255

South 0.83 940.84 416 155275 39747

Italy 0.87 542.75 2’552 1034499 210255

Table 3: Analysis of concentration of domain names “.it”

registered by Registrars 31-12-2010.

Macro

Area

Gini

Index

HHI

Index

Registrar

Domain

names

Range

North 0.80 228.78 1101 486322 43183

Centre 0.96 3468.80 427 1118659 617753

South 0.86 1190.19 314 181109 42942

Italy 0.91 1389.92 1’842 1786090 617753

in 2010. Even if the HHI results to be higher than

2005, since the Registrars have registered more than

DIGITALDIVIDESURVEYANALYSINGTHE".IT"REGISTRARMARKET

595

Table 4: Determinants of internet diffusion.

Models Variables Coeff

statistic

Standard error Sig. R

Economic (Model 1)

Per capita income 0.001 4.452 0.000 0.000 0.703

Total added value 1.954E-05 2.272 0.000 0.036

Cultural (Model 2)

Number of ICT graduates 0.003 3.249 0.001 0.004 0.370

Demographical (Model 3)

Unemployment rate -0.599 -5.969 0.100 0.000 0.733

Population 7.119E-07 3.768 0.000 0.002

Technological (Model 4)

Firms ICT every 1000

inhabitants

7.606 4.998 0.000 0.000

0.581

40% of domain names with respect to the situation

in 2005, we cannot speak of a monopoly situation as

in the Center. The data observed at an individual

level, show that the first two registrars of the North

ccTLD .it, while in the Centre the two first registrars

Table 3). In conclusion, in Italy a few Registrars

the ".it". The phenomenon of the registration of

domain names in Italy is concentrated in the hands

of a few Registrars, and such aspect increased

significantly in 2010.

3.1 Factors that Cause the Digital

Divide

To define the factors that cause the digital divide, a

multiple regression model was used, taking into

consideration the penetration rate as a dependent

variable (PR) at regional level, and economic,

cultural, demographic and regional technological

factors as independent variables. In this work, four

models were defined, called Model 1, Model 2,

Model 3 and Model 4, which take into consideration,

as independent variables, economic factors, cultural

factors, demographic factors and technological

factors, respectively. The variables taken into

account in this analysis were extracted from various

sources (ISTAT - Italian Statistics Institute, G.

Tagliacarne Institute and so on). The economic

factors taken into consideration in Model 1 were:

added value per employee; total added value; total

income; per capita income; total amount of tourist

businesses; firms with 250 employees or more;

patents every 100 firms; entrepreneurial density

every 100 inhabitants. In Table 4 are reported the

only two economic variables that express the linear

relation with the penetration rate: the total added

value and per capita income. The remaining

variables do not express in a significant way a linear

relation with the penetration rate). This means that

regions with a high per capita income are in the first

positions in the Internet use ranking. Model 1

explained approximately 70% of total variability (R

= 0.703). The fit to the model proved to be good and

significant. Model 2, based on the use of cultural

factors as independent variables, took into

consideration the number of ICT graduates; the

number of graduates; the number of ICT graduates

every 100 graduates; and the amount of employees

involved in research and development. At the

regional level, the only variable of Model 2 that

expresses in a significant way the linear relation

with the penetration rate resulted the number of ICT

graduates (Table 4). However, the fit to the model is

very weak and it explains approximately 37% of the

variability of Internet diffusion at a regional level

= 0.370). Demographic factors taken into

consideration in Model 3 were: population;

percentage of men and women (to verify for

example if regions with a higher percentage of men

are more inclined to use the Internet than other

regions that have a higher percentage of women);

population density per Km

; the total amount of

foreign people at regional level; and unemployment

rate. The statistical analysis showed that in Model 3,

the significant independent variables are constituted

by population and unemployment rate. The other

variables analyzed in the model have been

eliminated as scarcely significant. The results

obtained are shown in Table 4. As expected, the

correlation,

β, between penetration rate and

unemployment rate proves to be negative and rather

different from zero (

β = – 0.750), in accordance with

the economic literature (Bimber, 2000). In order to

define the technological factors that cause the digital

divide in Italy, in Model 4 different indicators were

taken into consideration: the degree of digitalization

calculated as weighted average by the indexes of

territory coverage of infrastructures and connectivity

services, such as optical fibre and broadband

(ADSL, HDSL, SHDL) (Assinform, Milano); the

degree of ICT specialization measured through the

ratio between the concentration of ICT employees

and the concentration of employees of all productive

WEBIST2012-8thInternationalConferenceonWebInformationSystemsandTechnologies

596

fields (Iuzzolino, 2001); the employees

concentration index measured through the ratio

between the number of ICT employees of the region

and the number of national ICT employees

(Iuzzolino, 2001); investments in information

technology (IT) made by the regions, IT expenditure

on the regional added value, IT expenditure per

employed person. As shown in Table 4, in Model 4

the only significant variable resulted to be the

“number of firms ICT every 1000 inhabitants”. Such

model expresses approximately 58% of Internet

diffusion variability.

4 CONCLUSIONS

This research identifies a crucial issue: the

continuity of a digital divide at local level. The

penetration rate, calculated in relation with the

number of Registrars, appear to confirm this trend.

As in the previous period analyzed, few regions

above all, for regions with a low unemployment rate

and with high economic factors, such as high added

value per employee and per capita income.

Furthermore, regions that have a high number of

firms specialized in ICT fields in relation to the

number of inhabitants, are more inclined to use the

Internet. Therefore, regions that are in the last

positions in economic terms are also in the last

position in technological terms. This is probably due

to the fact that low economic development is also

associated with a lower interest in new technologies

and their adoption. Furthermore, our research shows

that the concentration of domain names registered

under the ccTLD “.it” is very high, in fact only few

Registrars register the total amount of domains at

national level, especially in 2010. Moreover a high

rate of registrations results in those areas where

competition is not present. The absence of

competition was measured by the portion of domain

names owned by Registrars in a given geographical

area (North, Centre, South) and by the number of

actors that supply ICT services (Registrars) in a

given geographical area. Considering the macro-area

level, the study shows that in the Centre, where the

competition level is lower than the North and South

(HHI index, the index that measures competitiveness

in a territory, is greater in the Centre than in the

North and South) the number of registered domain

names is on average greater than the national

average. These results appear not in line with

Greenstein & Prince (2004) which affirm that is

absence of competition, Internet Service Providers

are less motivated to intensify their services.

However, it is necessary to remember that, here,

only the domain name registration service were

taken into consideration. Registrars could be

specialized in other types of services, like Internet

connectivity, electronic mail services, website

design, etc. Moreover, it must be underlined that

areas that are more specialized in domain name

registration are those presenting a competitive

advantage, in terms of economics, culture and

technology, over the others.

REFERENCES

Bauer, J. M., Berne, M., & Maitland, C. F., 2002. Internet

access in the European Union and in the United States.

In Telematics and Informatics, No.19, pp 117-137.

Bimber, B., 2000. Measuring the gender gap on the

Internet. In Social Science Quarterly, No.81, pp 868-

876.

Gini, C., 1912. Variabilità e mutabilità. Reprinted in: E.

Pizetti, T. Salvemini (Eds.), In Memorie di

metodologica statistica. Libreria Eredi Virgilio Veschi

(1955), Rome.

Greenstein, S., & Prince, J., 2004. The geographical

diffusion of the Internet in the United States. In M.P.

Singh (Ed.), The Practical Handbook of Internet

Computing (pp. 56.1-56.17). Chapman & Hall/CRC

Press, Boca Raton, FL.

Hirschman, A. O., 1964. The Paternity of an Index. In The

American Economic Review, Vol. 54, No. 5, pp. 761-

762.

Iuzzolino, G., 2001. Struttura dell’offerta e divari

territoriali nella filiera dell’information and

communication technologies in Italia. Economic

working papers n. 421, Economic Research

Department, Bank of Italy, Rome.

Rogers, E. M., 1995. Diffusion of innovations (4th ed.).

Free Press, New York.

Strover, S., 2001. Rural internet connectivity. In

Telecommun. Policy Vol. 25, No. 5, pp. 331–347.

Zook, M. A., 2000. Internet metrics: Using hosts and

domain counts to map the internet globally. In

Telecommun. Policy, Vol. 24, No. 6-7, pp. 613-620.

DIGITALDIVIDESURVEYANALYSINGTHE".IT"REGISTRARMARKET

597