Demand for Internet Access and Use in Spain

L. Cerno & T. Pérez Internet Demand in Spain Demand for Internet Access and Use in Spain Leonel Cerno Departamento de Economía Universidad Europea ...
5 downloads 0 Views 191KB Size
L. Cerno & T. Pérez

Internet Demand in Spain

Demand for Internet Access and Use in Spain

Leonel Cerno Departamento de Economía Universidad Europea de Madrid C/ Tajo s/n 28670 – Villaviciosa de Odón - Madrid e-mail: [email protected] Tel.: (34) 91 211 56 45

Teodosio Pérez Amaral Departamento de Fundamentos del Análisis Económico II Universidad Complutense de Madrid 28223 - Madrid e-mail: [email protected] Tel: (34) 91 394 23 80

We analyse a new phenomenon, the Internet demand in Spain, using a new high quality source of information and advanced econometric techniques for estimate Internet demand functions, focusing on the socio-demographic characteristics of the population. We begin with a graphic analysis of the data searching for relationships between the different characteristics. Then, based on the evidence found in the previous analyses, we specify and estimate binary Heckit models, ratifying the results obtained in the literature for other countries and verifying that they follow similar patterns of consumption. Keywords: Narrowband Connection, Broadband Connection, Internet Demand, Heckit Models. (JEL C2, C25)

1

L. Cerno & T. Pérez

Internet Demand in Spain

1. INTRODUCTION

Many socio-economic studies of both a theoretical and empirical nature are currently being developed in relation to the phenomenon of Internet service use and high-speed Internet access (called “broadband”) in Spain and other countries. If we analyze this phenomenon from a historical perspective in terms of adoption of a new product, the Internet as such is nothing extraordinary. As happens with any new product, Internet demand gradually grew until, in a very short time, it became an indispensable product. Historically, the concept of connecting and using systems in a shared network that allowed for the connection of two computers began to be developed early in the decade of the 60s. It was only in 1969 that the decision was made to implement an experimental network that would make it possible to exchange information between different computers. The network was funded in USA by the Advanced Research Project Agency (ARPA) – for this reason it was named ARPANET – and it is considered as the embryo of what would later become the Internet. Today practically no activity is conceivable without taking into consideration the information available in this large, worldwide net. What sets the Internet Service apart from other Telecommunications services is that never before has there been a technology that has been so rapidly adopted and that has had such an impact on all walks of life. Some years from now, we will undoubtedly be able to see how the Internet phenomenon has transformed not only the bases of telecommunication technologies, but also the very cultural bases of the developed countries.

Since it is claimed that the Internet phenomenon has changed the habits of families in developed countries, we must ask ourselves what relationships there are between the different factors that lead a family to acquire broadband Internet.

The Case of Spain

In Spain, there are some early descriptive studies carried out on the basis of independent surveys concerning the adoption of Information Technologies and communications in the country. The National Institute of Statistics (INE) began to compile this information

2

L. Cerno & T. Pérez

Internet Demand in Spain

as of 2001. The data in this work were taken from the 2003 INE Survey on Equipping and Use of Information Technologies in the Home (TIC-H).

As a first approximation, we could say that there is a great variability in access to and use of the Internet. The overall figures contained in Graphic 1 show that, while access to the Internet is increasingly common in Spain homes (25.2% of all Spanish homes), the distribution of this access is not homogeneous. Graphic 1: Internet Access in Spain (Households) 35.0

32.7 29.8

30.0

32.2

31.7 26.7

25.8 25.0 20.8

21.4

22.9

22.7

21.7

21.0

%

20.0

26.7

20.7

16.9 14.7

15.0

31.7

29.0

14.3

10.0 5.0

Ri oja Ce uta M eli lla

An da lus ia Ar ag on Ba Astu lea r ric ias Isla Ca n na d ry Isla nd Ca n C t Ca astil abri a sti a la L - L eon aM ac ha Ca tal on i Va a Es len tre cia m ad ura Ga lici a M ad rid M ur cia Ba sq Nav ue a Co rra un try

0.0

This access is more widespread in Catalonia (32.7% of homes), the Basque Country (32.2%), Madrid (31.7%) and Melilla (31.7%). This contrasts with Estremadura (14.3%), Castilla-La Mancha (14.7%) and Galicia (16.9%), where the figure is under 17%.

The main political debate in this area is how to help consumers access the Internet and, in particular, adopt the broadband technology1 . The debate begins by defining the

1

The concept of Universal Service Obligation could also include the Internet Service with broadband technology. This involves a different line of research than the one proposed in this article.

3

L. Cerno & T. Pérez

Internet Demand in Spain

Internet connection and the broadband technology2 . For purposes of simplification, we will say that Broad Band and High Speed are synonymous.

Rappoport, Kridel and Taylor (2003) clearly establish the existing similarities and differences between the conventional telephone service and adoption of some kind of Internet connection within the demand for telecommunications. What these two types of services have in common is that telecommunications are not consumed in isolation. There is a whole network of productive units involved. Aside from the interdependence and externalities that this situation entails, Access and Use are different concepts. The first and foremost of the differences concerns the measurement of output, since in conventional telephony output is measured in minutes, whereas in the case of the Internet it is measured by speed of file transmission. This leads to the second major difference between these two types of telecommunication services. In the case of some forms of Internet access (such as the cable modem), the speed will be affected by the number of individual Internet access lines that are transmitting at any given time, whereas this does not happen in conventional DSL lines.

This study estimates models of Internet access and use and compares the types of narrowband and broadband connections based on the consumer characteristics contained in a survey carried out by the National Institute of Statistics (INE). These preferences will be considered from two standpoints. On one hand, we will take into consideration the individual conditioned by education, experience and income which, in turn, is conditioned by family size. On the other hand, we will study the demand for the Internet service based on the family group that is demanding it and considering the equipment as indicator of one part of household income as no human capital investement (see appendix).

2. EARLY STUDIES

2

Authors such as Owen (2003) believe that it is a true mystery that the debate on the Economics of Broadband has reached the boiling point without yet having agreed on a definition of the term “Broadband” when there is a complete, accepted list of the services that includes its adoption.

4

L. Cerno & T. Pérez

Internet Demand in Spain

A pioneer econometric study on the adoption of the broadband Internet service is the one by Madden, Savage and Simpson (1996), who examined a database of 5,000 survey responses collected in Australian homes. These authors were the first to discover that demographic characteristics are one of the main influences on the individual decision to use the broadband Internet service. For example, they demonstrate, among other things, that people who have not finished secondary school show less of an interest in using the broadband Internet service; people who live in homes with at least one native member from Europe or Asia are more interested; age also influences interest depending on whether the individual is younger or older than 65 years of age.

Cassel (1999) uses a survey carried out in 1997 with 30,000 US replies. It focuses primarily on the characteristics of individuals with more than one fixed telephone line in the home. This author finds that having one additional fixed telephone line is positively associated with the size of the home, annual income and homes with small children with Internet access via a conventional telephone line. The research shows that people who own an additional telephone line and are interested in 56 kbps Internet access for a price of $10 to $30 a month could eventually be interested in a high-speed connection for $30 to $75. She finds that interest diminishes with prices for all consumers. This result suggests that users require other features (e.g., 24-hour connection) in addition to high speed before they are willing to pay $70 a month.

Goolsbee (2000) also examines the demand for Broadband Internet access with data from 100,000 answers to a survey carried out in 1999 in various U.S. cities. He estimates a Probit model relating the probability of adopting Broadband Internet via cable-modem, the service price and characteristics such as the number of years during which the respondent has been connecting to the Internet, age, income and education. This model shows that the intention to access the Internet via cable-modem increases for agents with lower prices. The price-elasticity of the demand for different ranks ranges from –2.8 to -3.5.

Duffy and Deno (2001) study a sample of US households for the fourth quarter of 1998 11,458 and find that 22% of the sample has two or more telephone lines. They estimate a Logit model of residential demand for additional access lines in the United States. The

5

L. Cerno & T. Pérez

Internet Demand in Spain

price-elasticity of the demand is –0.59, which is more elastic than the primary telephone line.

Rappoport, Kridel and Taylor (2002) estimate the broadband Internet demand in a different context from the ones proposed up to this point. They use a database with demographic information on homes in 10 US cities for the month of august, 2001. All the respondents have some kind of Internet access in the home and so the choice here focuses on what kind of access they prefer, either low, or high-speed. The authors separate the data into two groups – those that choose low-speed Internet and those that choose high-speed Internet – and focus the study on the characteristics that distinguish one group from the other. In other words, they go beyond distinguishing between people who use or do not use some kind of Internet access service. Using a Nested Logit model, they find that the size of the home positively affects the possibility of choosing broadband Internet. Surprisingly, they find a positive correlation between older respondents and the adoption of broadband Internet, more so than for younger respondents.

Other recent studies that make special reference to the Willingness to Pay of Internet users includes the one carried out by Varian (2002), which uses data from a University of Berkeley project called “Internet Demand Experiment” and in which he estimates how many people would be willing to pay for different connection speed levels. He estimates price-elasticities of between –1.3 and –3.1. He demonstrates how users with technical and administrative professions highly value their own time, whereas the rest of the people would not be willing to pay very much for a high-speed Internet connection. He concludes that a large wave of demand for Broadband Internet access is not to be expected, at least in the near future.

Different reports have been written in addition to these studies, such as the one released by the U.S. Department of Commerce (2002) that also considers demographic influences on the adoption of Broadband Internet. This study uses the same model as the one proposed by Madden et al. (1996) but for surveys carried out in the United States and considering actual patterns of adoption. It is not surprising that similar conclusions are reached.

6

L. Cerno & T. Pérez

Internet Demand in Spain

Another report is the one by the Telecommunications Research Group of the University of Colorado (2003), which addresses issues such as how to estimate preferences and profiles of consumers who adopt the Broadband Internet service. The interesting thing about this report is that it examines issues concerning the Broadband Internet market, as things such as doing business via the net, government regulation and the tele-work phenomenon are considered.

Except for the latter report, none of the above mentioned studies address the structure of the Broadband market. They only give indications of the characteristics of the demand, but this says nothing about how the market structural costs are affected when a certain type of Internet connection is chosen.

Within this line of research, there are works, such as the one by Gabel and Kwan (2000) that acknowledge this deficiency in the econometric literature and decide to address the problem by examining the cost of the demand in residential broadband Internet accessibility. Using a stratified sampling method, they select a sample of nearly 300 observations. They adapt a Logit model for estimating the supply and demand with the factors that influence the probability that ADSL or cable-modem connections will be accessible in an area where there is the right hardwiring for this. The authors conclude that if there are enough 30 to 34 year old adults in that area with incomes approaching the mean, then the probability of adoption will increase. The proxy variables that they use to quantify the cost, and that are significant within the model, are tele-density (number of telephone lines per square miles) and an estimated transmission cost of Internet traffic.

Finally, we should mention the report issued by the OECD (2001) which analyzes the adoption of Wideband Internet connections in 30 countries. This report clearly shows that the countries that were and are in the lead in terms of both technology and number of consumers owe their position to the support and regulation of competition between the different networks and networks with different technologies.

3. THE DATA

7

L. Cerno & T. Pérez

Internet Demand in Spain

In this study, we use information provided by the Annual Survey TIC – H 2003 carried out by the INE in all of Spain and on two registers designs – a family register and another for 10 to 14 year old users. Our objective is to provide empirical evidence in the debate on Internet access and use, focusing primarily on the form of connection and the user profile. Adoption of a type of Internet connection, the nature of the type of connection and the use that is made of the service would be endogenous in the model proposed in this work.

Internet Access from the Home

For the first case, there is a sample of 18,949 answers to the question: “do you have some kind of Internet access in the home?”. The frequency distributions are shown in tables and in the appendix Descriptive graphics are provided below.

Graphic 2: Spanish Households with Internet Access 1,8 2 Narrow Band (Telephone) 34,3

Broadband (DSL, Wire) Other 74,6 NR

8

L. Cerno & T. Pérez

Internet Demand in Spain

Graphic 3: Internet Access from the Home by the Habitat and Connection Form 2,000,000 Narrowband (Telephone)

Number of Connections

1,800,000 1,600,000

Broadband (DSL and Wire)

1,400,000 1,200,000

Other Forms

1,000,000 800,000 600,000 400,000 200,000 0 less than 10,000 inhabitants

10,000 to 20,000 inhabitants

20,000 to 50,000 inhabitants

50,000 to 100,000 inhabitants

more than 100,000 inhabitants

Graphics 2 and 3, corresponding to Table 1 in the appendix, refer to access to the Internet service from the home, taking into account the habitat where it is located. We see, for example, that of all homes with Internet access (3.599.04 homes), the majority chooses the narrowband form of connection of a conventional telephone line (74.6%) and only 34.3% opt for the broadband form of connection or other forms of connection (2.0%). We can see that there are families that would choose more than one form of connection because the sum of these percentages does not equal 100. Table 1 also shows that the sample is categorized in accordance with the size of the respondent’s home. See Graphic 4 below:

9

L. Cerno & T. Pérez

Internet Demand in Spain

Graphic 4: Internet Access from the Home Taking into Account the Home Size and Connection Form 1,200,000

1,000,000 Narrowband (Telephone) 800,000

600,000

Broadband (DSL and Wire)

400,000

200,000

0 1 member household

2 members household

3 members household

4 members household

5 members and more

We can see here that the telephone dial-up connection via modem persists in all the categories considered, with the homes with 4 members prevailing over the rest (40.2%), especially compared to homes with 1 and 2 members (4.3% and 15.6%, respectively). However, we should note that there are considerable differences between the percentages of access via one channel or another depending on the size of the home. For example, the broadband-narrowband (BB/NB – see appendix) ratio for the extreme sizes (1 and 5 members) equals 55.1% and 58.8%, respectively, whereas this difference is less for the middle sizes (46.4%, 40.5% and 49.2% for 2-, 3- and 4-member homes, respectively). This could be explained by the fact that homes with one and two members present two well differentiated typologies: one type of home inhabited by middle-aged or young working people who prefer to access the Internet directly from their workplace instead of from their homes, and another type of home where retired people live. As the latter will have practically no interest in accessing the Internet from their own homes (since they would rely on other centers for this), the average rate of Internet access in general and via a Broadband connection in particular in these one- or two-member homes is less than for homes with more members.

10

L. Cerno & T. Pérez

Internet Demand in Spain

Internet Use

Graphic 5 contains a descriptive analysis of Internet use that relaxes the restriction that access be only from the home. Another thing of note here is that the question is aimed at the individual and not at the home, as it was in the case of access. We can see that of a total of 12,130,100 people who have used the Internet service in the last 3 months, the majority accesses from the home and via one of the types of connection commented on previously. In other words, the majority of the population that uses the Internet (59.7%) does so from the home, followed by those people who connect to the Internet from the workplace (41.3%), from other places (29.3%) and from centers of study (20.4%)3 . Graphic 5: Internet Use in the last 3 Months by Age and Place of Use

Home 2000000

Individuals

Workplace 1500000 Center of Study 1000000

Other Places

500000

0 15 to 24 25 to 34 35 to 44 45 to 54 55 to 64 65 to 74

75 and more

Age

The place of use illustrates the segmentation of the Spanish population in terms of Internet use. If we observe the graphic, we see that quite a uniform group of all ages uses the Internet from the home, but we also see that younger people (15 to 24 years of age) use it a lot in their centers of study and older people use it a lot from the workplace. Another thing that this graph reveals is that the number of people who connect to the Internet decreases as the age increases, although in relative terms the percentage of 3

The percentages do not add up to 100 because we are considering multiple response tables.

11

L. Cerno & T. Pérez

Internet Demand in Spain

people who access the Internet from the home increases (although not observable in the graphic because of the scale used, 72% of the respondents who are 75 yeas of age or older accesses the Internet from the home, compared to the respondents in the 15 to 24 years of age interval, with 56.6% accessing from the home but 46.2% accessing the Internet from the center of study).

Graphic 6: Internet Use in the Last 3 Months by Study Level and Place of Use 3000000

Individuals

2500000

Workplace

2000000 1500000

Center of Study

1000000 500000

Other Places

Un ive rsi ty Tr ain ing

Pr im ary 1s t. S Sc tag ho ol eS ec on Up da ry pe Ed rL uc ev . el Pr ofe ss ion al Ed 2n d. uc Sta . ge Se co nd ary Ed uc .

W ith ou t

Tr ain ing

0

As regards level of studies, we see from Graphic 6 that Internet access will be used to a greater extent by people who have finished upper level studies (33.9%); they are followed closely by those people who have completed the second stage of secondary education (30.4%), and then by 19.4% of people who have completed the first stage of secondary education and 10.6% of people with upper level Occupational Training. We also see that people with upper level studies will connect a lot from their workplace, while this proportion differs in the other categories under consideration. As is to be expected, people who have completed a level of training other than those mentioned above (except for primary school) and people without any kind of training together represent a mere 0.5% of the respondents and are barely perceivable in the graphic. It is obvious that the level of education is very much associated with Internet access and use.

12

L. Cerno & T. Pérez

Internet Demand in Spain

Considering next the working and professional status of the respondents, table 4 in the appendix and Graphic 7 below show, as expected, that the people who use Internet the most are employed workers who connect to the Internet almost indistinctly from the home and from the workplace (63.83% in all); they are followed by students (21.20%) who connect mainly from the center of studies, which is followed closely by access from the home. Unemployed workers account for 6.07%, and the remaining 9% is divided among people who are not part of the working population (housekeepers, pensioners and other types of unemployed).

Graphic 7: Internet Use in the last 3 Months by Work Status and Place of Use 5000000 4500000

Individuals

4000000 3500000 3000000

Home

2500000

Workplace

2000000 1500000

Center of Study

1000000 Other Places

Ot he r

Pe ns ion ers

Ho us ek ee pe rs

St ud en ts

W ork ers

Un em plo ye d

Em plo ye d

W ork ers

500000 0

Work Status

In general, as we could see, more than half of the people who use the Internet are people whose habitat is in big cities (with more than 100,000 inhabitants and provincial capitals). In the case of the size of the respondent’s home, those who use the Internet the most come from three-member homes, followed by those belonging to homes with four or more members, and so on.

13

L. Cerno & T. Pérez

Internet Demand in Spain

After this basic description of the data, next we want to discover and measure the relationships between the variables. We begin to do that using a demand model, that is designed to explain the relationships between the variables of interest.

4. MODELING HOUSEHOLD ACCESS AND USAGE

4.1. Utility Funtions The theoretical evidence on Internet demand suggests that Internet is used to save money and time. Following the framework of Taylor (1994)4 , there are two types of agents: G0 : Subset of agents without access to the net G1 : Subset of agents with access to the net The utility function of individual i is expressed as U i = U i ( x i, δ i qi )

(1.1)

where x i is the vector of the other goods consumed by the i th agent, and the dichotomous variables determine the access status of the agent, i.e.:

q ∀i ∈ G1 qi  0 ∀i ∈ G0 1 if the agent has access (use) δi  0 otherwise The problem of maximizing the utility would then be expressed with individual utility functions for each type of agent as follows: U 1 = U 1 ( x1 , q ) U 0 = U 0 ( x0 )

if δ =1 if δ =0

(1.2)

4.2. Econometric Approach

4

Lester Taylor summarizes of the contributions to the theory of the telephone demand from mid-1970s. He begins with the idea originally outlined by Artle and Averous (1973) where the dynamic aspects of the telephone system as a public good are analyzed, and where the property of acces-no access plays an important role in the analysis. Further, according to these authors, other telecommunication systems will have similar access-no access properties.

14

L. Cerno & T. Pérez

Internet Demand in Spain

The literature suggests that Internet users differ from other users of telecommunications services in terms of the type of attributes that are relevant. This is supported by Rappoport et al. (2002) when they outline the differences between telephone demand and Internet demand. In accordance with this line of thought, Jackson et al. (2002) use a maximization model of the work-leisure choice, assuming that the agents want to have income, leisure time and also online activity. The theory indicates that the maximization of the agent’s utility having access to the Internet in the home is conditioned by the consumption of other goods and the allocation of time and money. A linear approximation of a conditional utility function would be as follows: U i* = xiT β + ε i

(1.3) i = 1, 2, ..., n.

where U i* is the latent utility that the agent experiences on accessing and/or using the Internet, β is the vector of parameters to estimate and represents the vector of the marginal utilities of each of the regressors that are found in the vector xTi , and ε i as an error term. The parameters of the individual latent utility function (the marginal utilities) are estimated on the basis of information from the responses to the questions posed. In this paper, we use this framework and an econometric specification to analyze the characteristics of Internet users in the case of Spain and to look for relationships that socio-demographic characteristics such as habitat, sex, age, studies completed and others, have with the agents that access the Internet from the home or use the net through centers other than the home, e.g. a center of study, workplace, etc. Taking into account the relationship between access and use of the Internet said above, we can say that a division of the sample of size N is now: P1 : Number of agents with access to Internet from home. P2 : Number of agents with no access from home but with use in other places. P3 : Number of agents without home access and no use. N = P1 + P2 + P3 We cannot measure directly the utility provided by Internet because that information is censored. We take this into account when we specify the model. Following Meng and Schmidt (1985), to obtain efficiency and consistency we estimate jointly an equation of selection to determine who has access and who doesn’t have access to Internet at home, and another equation (outcome equation) to explain what type of connection he has at home, by means of a procedure of maximum likelihood. But to evaluate this likelihood function it is necessary to know the probabilities of each of the observed events. In the pattern that we want to specify they can produce four events, depending on whether the individual is “selected” to participate in the sample (for example, that he belongs to the

15

L. Cerno & T. Pérez

Internet Demand in Spain

group G1 ), and the election that he makes once he has been selected. Gives the four events as we saw, only three are observed, because we don't know which election of use would be made by the non-selected agents:

Table 1: Probabilities of Internet Access and Type of Access Choice at Home Internet Type of Access from Home

Access / No Access from Home

qi = q

qi = 0

δ i =1

P ( q i = q | δ i = 1)

P ( q i = 0 | δ i = 1)

δi = 0

P (q i = q | δ i = 0 )

P (q i = 0 | δ i = 0 )

Whereas we don't have observations to calculate the probabilities P ( q i = 0 | δ i = 1) and P ( q i = 0 | δ i = 0 ) , though due to δ i is a binary variable, P ( q i = 0 | δ i = 1) + P ( q i = 0 | δ i = 0 ) = P ( q i = 0 )

Thereby: P ( q i = q | δ i = 1) + P ( q i = q | δ i = 0 ) + P ( q i = 0 ) = 1

So, the likelihood function will be then:

L = ∏  P ( q = q | δ = 1)  i =1 N

i

i

i i



 P ( qi = q | δ i = 0 )   

q i (1 −δ i )

P (q i = 0 )  

(1− q i )

(1.4)

This bivariate relationship can be considered through a procedure in two stages that allows consistent estimation of the parameters, and controling for the bias due to non random selectivity (Heckman, 1979). Following Sigelman and Zeng (1999), this method is used when the interest centers on the relationship between x and y but data are available only for cases in which another variable, z * , exceeds a certain value. Specifically5 ,

5

Davidson and MacKinnon (1993).

16

L. Cerno & T. Pérez

Internet Demand in Spain  yi*   xi β   ui   *  =  +   zi   wiγ   vi 

(1.5)  σ2 ui  : NID  0,      ρσ vi 

ρσ    1  

The first equation is centered on the behavior about the election, whereas the second one has made the selection for the sample of interest, and ρ is the correlation coefficient between the errors6 . The estimator is based on the conditional expectation of the observed y : E  y i zi* > 0 = xβ + σ 2 ρλi ( − wiγ )

(1.6)

where λi is the inverse of the Mills ratio. As we can see, equation (1.6) implies that the conditional expectation of yi will be xi β only when the correlation among both errors is null. In other cases, it is always affected by the equation of selection. We sketch the decisions of access and use of Internet in Figure 1:

Figure 1: Relationship between Access and Use of Internet

As we can see, Figure 1 suggests that the demand of some connection to Internet occurs if before it is opted by the access from the home. Starting from there we determine a demand for broadband keeping in mind the following attributes:

6

The restriction that the variance of

vi is equal to 1 is imposed because only the sign of zi* will be

observed. See Davidson and MacKinnon (1993).

17

L. Cerno & T. Pérez

Internet Demand in Spain

where each attribute will be quantified as shown below. The model used is based on maximization of the utility of an agent that considers accessing the Internet in a certain scenario.

4.3. Specification Because of the type of data available, we will use a binary probit model with selection bias. Thus, for the first model specified for the i th individual we fit two equations and will have both endogenous variables as binary dummies. The endogenous variable of the equation of broadband demand is CI i ( / H i = 1) = 1   CI ( / H = 1) = 0 i 

if has broadband in home. otherwise.

and are referred to the type of Internet connection at home (=1 if has broadband), but only make sense when the variable Hi (referred to the answer to the question “Do you have Internet access from your home?”) is equal to one. This means that the variable Hi selects non randomly the sample for estimating a demand model to measure the effects in the type of Internet connection at home. The variables that compose the vectors of regressors are explained in the following Table: Table 1. Regressors Variable Name

Definition

Economic o isi

Family income index.

Technologic o ord1i o laptopi o mobilei o

usagini

=1, if has a PC at home; =0 otherwise =1, if has a laptop at home; =0 otherwise =1, if some household member has a mobile phone; =0 otherwise Frequency of Internet use (approximate number of quarterly times):

18

L. Cerno & T. Pérez

Internet Demand in Spain

=70 every day (at least five days at week) =14 every week, but not every day = 3 at least one time by month = 1 Not every month = 0 Never Social and Demographic o ecoursi o levelest i o housemi o habitat i o sexhi o agei o agesqi

=1, if the respondent is studying; =0 otherwise Degree achieved in studies (measured by years of study). Persons residing in the household Population size =1, if the respondent is male; =0 otherwise Age of the respondent Square of age of the respondent

So, the demand of broadband access at home model that interests us finally is: CI i = β0 + β1 isi + β 2 ord1i + β3 laptop i + β4 mobilei + β5 ecoursi + + β 6 housemi + β 7 habitat i + β8 sexhi + β 9 agei + β10 agesqi + IMR ( x ) + ε i

where IMR ( x ) =

φ ( zTγ )

1 − Φ ( z T γ )   

(1.7)

is the inverse of the Mills ratio as we explain it above.

5. ESTIMATION AND DISCUSSION In this section, we discuss the estimation results of the proposed model. As we saw above, the endogenous variable in the outcome equation is the broadband access in the home, and the endogenous variable in the selection equation is just the probability of some access to Internet in a home. As noted earlier, the marginal effect on CI i is composed from the effect of selection equation and the outcome equation. This is to say that every predictor in the model appears not only as the exogenous variable in the outcome equation, but also as a component of the IMR ( x ) . One consequence of this nonlinearity is that the effect on n units of change in the vector of the exogenous variables is not simply n times the effect of one unit of change in this vector. This means that the change in the endogenous variables depends not only on the magnitude of the change, but also on the base from which the change takes place (Sigelman and Zeng, 1999).

5.2. Subscription to Broadband Access in Home Model Results 19

L. Cerno & T. Pérez

Internet Demand in Spain

Table 2 shows the marginal effects of the first model fitted. The complete results of estimation are in the appendix. As we can say, there are some marginal effects with a sign for some type of connection and with the opposite sign for the other type. For example, the personal income level will have a different effect in the demands for Internet, according to the type of connection. For the case of broadband access, the effect is positive. This can be explained due to that the broadband access is still considered a luxury good in the Spanish society. Among the technological attributes we have the connection with PC and with laptop, which are positive although they are bigger for the case of broadband access. Another positive effect in the probability from both types of connections is the attribute to be studying, although the contribution to the probability of narrowband access is much bigger than for broadband access. This could be due to the possibility of high speed access from the center of studies, and in the home the phone connection is used with narrowband access. For the case of the technological attribute as the connection with mobile telephone, we have opposite results for each of the types of connection. Broadband is positive and significant to 95%, whereas narrowband it is negative and significant to 90%. The socio-demographic attributes will have according to what probability of connection we are referring to opposite sings. For example, in the case of broadband access, the habitat has influence negative, (although in a minimal proportion) because in big cities there is more possibility to access to Internet from other places. For the case of narrowband, the effect is the opposite. As for the sex, we see that the masculine sex is prone to the broadband, and less to the narrowband, although this last effect is insignificant. As for the age, just as we could observe in the model specification above, the variable enters in both equations and in levels and squared forms. This causes that, Ceteris Paribus, the marginal effect in the probability to acquire some type of Internet connection is not constant, but a linear function. That is: ∂CI i = β 9 + 2 β10 Agei ∂Agei

(1.8)

The effect of the age in the probability of broadband connection at home, is negative in early ages, but grows until becoming positive in later ages. The inflection point from which said effect becomes positive is around 50 years. This means that the biggest weight in the decision to acquire broadband for the home relapse in the family boss that in turn that his income represent the half or more than half of the family income. It happens the opposite with narrowband access (see appendix), since we observe that it is rather falling with the increase of age. The contribution in early ages will be positive for the same reason: it is more likely that a youth between 14 and 25 years is connected Internet from the home with a connection of Narrowband, whereas a more old people (56,6 years) would not connecting. This can observe it in the graph below. 20

L. Cerno & T. Pérez

Internet Demand in Spain

Table 2: Demand of Broadband in Home, Marginal Effects

Regressors

βˆ

Constant

Pr > | z |

|z|

0.666

8.53

0.000

3.240

598.24

0.000

0.144 0.055 0.052 0.002

4.60 2.23 1.81 9.10

0.000 0.026 0.070 0.000

-0.002 0.002 -0.023 0.009 -0.004 0.00004 -0.212

0.10 0.35 7.60 0.65 1.67 1.70 12.85

0.922 0.728 0.000 0.514 0.096 0.090 0.000

Economic Attributes isi Technological Attributes ord1i laptopi mobilei usagini Socio and Demographic Attributes ecoursi housemi habitat i sex _ hi agei agesqi IMR ( x) Total of Observations: 18,940 Uncensored Observations: 4,470 Log likelihood = -9229.138 (Prob > χ 2 = 0.00) Wald test of independence: 133.12 (Prob > χ 2 =0.00)

5.2. Internet Use Model and Results Now we specify and estimate a second model referred to use Internet demand but relaxing the restrictions that have access in home and taking into account the use in workplace, center of study and other places such as cybercoffees, booths, etc). In this section we analyze a second model with the multinomial logit specification and in which we consider the use of Internet and its determinants, from the home, the

21

L. Cerno & T. Pérez

Internet Demand in Spain

worksite, the place of study and other places (cyber-coffees, booths, etc). The regressors that we use, although they are not exactly the same ones, follow the same philosophy of considering economic, technological and socio-demographic attributes. The dependent variable is a non ordered politomic variable referred to the intensity of use of Internet in the four mentioned places. That is,  = 1 if has use in one place of these considered  = 2 if has use in two places  USEi   = 3 if has use in three places = 4 if has use in the four places considered The multinomial logit specification to explain the different profiles considered and he’s relations will be equal to estimate the follow relations respectively: ln

Pi1 = xiT β1 (1 − Pi1 )

ln

Pi2 = xiT β 2 1 − P ( i2 )

ln

Pi3 = xiT β3 (1 − Pi3 )

ln

Pi4 = xiT β 4 (1 − Pi 4 )

where Pi1 , Pi2 , Pi3 and Pi4 are the probabilities of use Internet in one, two, tree and four places considered. The xTi is a vector of regressors of i , and β is a vector of parameters. As we saw above in the first graphical analysis, the most frequently place of use is the home, followed by the workplace, and the place of study. Indeed, we have a multinomial logit with a endogenous variable with four categories, one that will be equal to one if Internet uses only in one than the four places (may be in the home), another that will be equal to two if it uses in two (may be also in the work or in the school), three and four if uses in almost or all places considered respectively. Furthermore, these places of use are not mutualy excluding. The results can be seen in the Tables 3 below. Table 3: Demand of Internet Use, Marginal Effects

Regressors

USEi = 1

USEi = 2

USEi = 3

USEi = 4

Constant

-3.80 (20.09)***

-6.57 -9.45 -14.29 *** *** (26.54) (18.90) (10.46)***

Economic Attributes 22

L. Cerno & T. Pérez

isi Technological Attributes b _ ai usori

Internet Demand in Spain

11.84 (16.10)***

21.23 24.98 38.87 *** *** (21.88) (13.30) ( 7.86)***

1.44 (15.01)*** 1.06 (37.08)***

1.74 (16.75)*** 0.75 (19.69)***

2.04 (13.06)*** 0.44 (5.01)***

2.09 (6.19)*** 0.41 (1.59)

0.83 (10.55)*** -2.51 (6.19)*** -0.20 (7.96)*** 0.15 (1.19) 0.41 (7.38)*** -0.05 (21.68)***

1.51 (17.20)*** -5.46 (10.54)*** -0.20 (6.57)*** 0.02 (1.37) 0.58 (8.59)*** -0.06 (20.83)***

2.29 (6.19)*** -6.07 (6.08)*** -0.26 (4.53)*** 0.06 (2.29)** 0.88 (6.88)*** -0.08 (12.83)***

2.88 (7.81)*** -12.65 (4.85)*** -0.28 (2.00)*** 0.04 (0.59) 0.90 (2.81)*** -0.09 (5.13)***

Socio and Demographic Attributes ecoursi studylevi housemi habitat i sex _ hi agei Total of Observations: 18,940 Log likelihood = -9115.67 (Prob > χ 2 = 0.00) Pseudo R2 = 0.4218

We see that for the case of the household income, the marginal effect will be positive, being the biggest contribution to the Internet use from the home, continued by the connection from the work and lastly from connection from other places. The personal income won't contribute in anything to the probability of use from the place of study. This agrees with the descriptive analysis where we saw that most use of Internet is from the home, and it will be bigger in those households with larger income. The only technological attribute that we have considered in this second equation is to have broadband in the home, and we see that it is significant for the case of the use from the home and from the work, but not for the case of the connection from the center of study and the connection from other places. The contribution to the probability of Internet use is bigger for the case of the use at home, being also positive, although smaller, the marginal effect for the use of Internet from the workplace. Obviously there will be more possibility of use of Internet mostly from the home if one has broadband service, and there will be more probability that the use from the work is more labor type than personal. To be studying, as is logical, will have the largest marginal effect on the probability of use Internet from the center of study. Indeed, and takint into account the people that doing two activities (working and studying), the marginal effect gives to be studying in

23

L. Cerno & T. Pérez

Internet Demand in Spain

the probability of use from the workplace it will be negative, giving bigger sense and consistency to what we sustained thereinbefore with the use of Internet from the work. The marginal effects of the social and demographical attributes, as the age for example, in both equations, is positive for the case for the use in the home and in the work, and negative for the use from the center of study or other places. Other social and demographic attributes considered, and they aren’t as explanatory variables in the selection equation (as members in home or habitat), will be positive and significant to 99% for the use from the home and from the center of study and negative for the cases as the connection from the work or from other places (this last one significant to 90%), not being significant for the case of the use in the work or in the center of study. The masculine sex is significant at the 99% and has a positive marginal effect in all the probabilities of Internet use, except in the use from the center of study where the marginal effect will be negative.

6. SINTESYS OF THE RESULTS Below, the Tables 5 and 6 provide an overview of marginal effects for the Internet home access and uses respectively. In general the results are logical and follow the logic of previous results seen in the literature. For the case of the income, we see that it is an important factor in the decision for the type of Internet connection at home, and that high household income increases the likelihood of acquire broadband connection and/or diminish the likelihood of to be connected with narrowband. Following Madden et al. (2004), if the household income is low, tends to subscribe only to fixed-line telephony, and in this sense we can think that the household connect to Internet at home with a phone-connection of narrowband. By contrast, high income households increase the probability to subscribe the Broadband in home. Furhter, broadband connections are considered a luxury good. There are some social and demographic attributes that are insignificant in the probability of the type of connection at home (as the numbers of members in the home, for example). The contrary occurs when we consider the probability of Internet use, where practically every one significant, but with positive or negative marginal effects according to if the social or demographic characteristic increase or decrease the likelihood for the place of Internet use. Thereby we can expect, for example, the fact of be studying increase the probability to connect to Internet in the center of study more than connect at work or other places. The same for the use at home when has broadband connection. Table 4: Overview of Access to Internet from Home Model7 Opposite Marginal Effects According to Type of Connection

Equal Marginal Effects According to Type of Connection

7

Opposite Marginal Effects means that this marginal effect considered in the probability for broadband at home has the opposite sign than narrowband at home. Equal Marginal Effects means the contrary (same sign in parentesis). For more detail see the Tables 2 and 3.

24

L. Cerno & T. Pérez

Internet Demand in Spain

-

Economic Attributes: o Household Income.

-

Tecnological Attributes: o Mobile Connection.

-

-

Tecnological Attributes: o Connection with PC (+). o Connection with Laptop (+)

-

Social and Demographical Attributes: o Be Studying (+). o Members in home (-).

Social and Demographical Attributes: o Habitat. o Sex. o Age.

Table 5: Overview of Use of Internet Model8 Opposite Marginal Effects According to Internet Usage -

Social and Demographical Attributes: o Be studying. o Members in Home. o Habitat. o Sex. o Age.

Equal Marginal Effects According to Internet Usage -

Economic Attributes: o Household Income (+).

-

Tecnological Attributes: o Broadband in Home (+).

7. CONCLUSIONS AND FUTURE LINES OF RESEARCH For conclude, we now present the general conclusions in the following bullets: •

Our intention in this work is to use the teoretical-empirical focuses provided in the literature on the profile of residential Internet access and estimate the consumer demand for use of this service.



The preliminary analysis of the data from a survey carried out by the INE in 2003 suggests that 25.2% of the Spanish population accesses the Internet, and of those a majority accesses through narrowband and only 35.5% through a broadband connection.



A joint analysis is the first tool used to estimate the characteristics that affect use of the Internet service, with a descriptive analysis of the relationship between the

8

Opposite Marginal Effects means that this marginal effect considered in the probability for use Internet at home, at work, at center of study or other places hasn’t the same sign in every equation estimated. Equal Marginal Effects means the contrary (same sign in parentesis). For more detail see the Tables 4 and 5.

25

L. Cerno & T. Pérez

Internet Demand in Spain

socio-economic typologies of the users and the type of connection used. It was observed that there are links between these characteristics and the Internet connection used. •

In addition to access from the home, these socio-demographic characteristics are even more pronounced when measuring use of Internet, and we have seen that this will be very dependent on the level of studies, habitat and age, and the results are quite consistent with the results of other countries.



We then specified and estimated two-stage Heckman model, accounting for the selection bias, for individual cross-section data, and we confirmed that the Internet service is positively related to the level of education and negatively linked to age into the other thinks.

In future research, it would be advisable to have data on the respondent’s income (personal income) and on the prices of the services in order to estimate a broadband demand function for the different regions of Spain, always accounting for the selection bias. Finally, we believe that in future research and as more comprehensive information becomes available, it may be interesting to work with panels of individuals over several periods, in order to evaluate the dynamics affecting the demand for the Internet service in Spain.

REFERENCES 1. Artle, R., Averous, C. “The Telephone System as a Public Good: Static and Dynamic Aspects” Bell Journal of Economics and Management Science, Vol. 4, No. 1, 89-100, spring 1973. 2. Cassel, C., “Demand for and Use of Additional Lines by Residential customers”, in Loomis and Taylor (eds.) The Future of the Telecommunications Industry: Forecasting and Demand Analysis, Kluwer Academic Publishers, Boston, 1999. 3. Davidson, R., MacKinnon, J., “Econometric Theory and Methods” Oxford University Press, 1993. 4. Duffy-Deno, K. T. “Demand for Additional Telephone Lines: An Empirical Note”, Information Economics and Policy, 13, 301-309, 2001. 5. Friedman, M. “A Theory of the Consumption Function”, Press for the National Bureau of Economic Research, 1957.

Princeton University

26

L. Cerno & T. Pérez

Internet Demand in Spain

6. Gabel, D. and Kwan, F. “Accessibility of Broadband Telecommunications Services by Various Segments of the American Population”, Telecommunications Policy Research Conference, August 2002. 7. Goodman, A., Kawai, M. “Permanent Income, Hedonic Price and Demand for Housing: New Evidence”, Journal of Urban Economics, 12, 214-237, 1982. 8. Goolsbee, A. “The Value of Broadband and the Deadweight Loss of Taxing New Technology”, Mimeo, University of Chicago, 2000. 9. Heckman, J. “Sample Selection Bias as a Specification Error” Econometrica, Vol. 47, No. 1, January 1979. 10. Jackson, M.; Lookabaugh, T., Savage, S., Sicker, D., Waldman, D. “Broadband Demand Study: Final Report” Telecommunications Research Group, University of Colorado, 2003. 11. Madden, G., Coble-Neal, G., “Australian Residential Telecommunications Consumption and Substitution Patterns” preliminary draft, 15th International Telecommunications Society Meeting, Berlin, Germany, 4-7 September 2004. 12. Madden, G., Savage, S., Simpson, M. “Information Inequality and Broadband Network Access: An Analysis of Australian Household Survey Data” Industry and Corporate Change, Oxford University Press, 1049-1056, 1996. 13. Mc Fadden, D. “Conditional Logit Analysis of Qualitative Choice Behavior” in Zarembka (ed.) Frontiers in Econometrics, Academic Press, pp. 1117-1156, 1974. 14. Mc Fadden, D. “Econometric Analysis of Qualitative Response Models” in Z. Grilliches and M. Intriligator (eds.) Handbook of Econometrics, Amsterdam, North Holland, pp. 1376-1425, 1984. 15. Meng, C., Schmidt, P. “On the Cost of Partial Observavility in the Bivariate Probit Model” International Economic Review, Vol. 26, No. 1, february 1985. 16. Organization for Economic Cooperation and Development, “The Development of Broadband in OEDC Countries”, October 29, 2001. 17. Owen, B. “Broadband Mysteries” in Crandall, R. and Alleman, J. (ed.) Broadband: Should we Regulate High-speed Internet Access? AEI – Brookings Joint Center for Regulatory Studies, 2002. 18. Pérez Amaral, T., F. Alvarez and B. Moreno “Business Telephone Traffic Demand in Spain 1980-1991: an Econometric Approach”, Information Economics and Policy, 7, 115-134, 1995. 19. Rappoport P., Taylor L., and Kridel D. “The Demand of Broadband: Access, Content, and the Value of Time”, in Broadband: Should We Regulate High-

27

L. Cerno & T. Pérez

Internet Demand in Spain

Speed Internet Access?, ed. by R. W. Crandall and J. H. Alleman, AEIBrookings Joint Centre for Regulatory Studies, Washington, D.C., 2002. 20. Sigelman, L., Zeng, L. “Analyzing Censored and Sample-Selected Data with Tobit and Heckit Models” Political Analysis, 8:2, The George Washington University Working Papers, December 16, 1999. 21. Taylor, L. D. “Towards a framework for analyzing internet demand”, Manuscript, U. of Arizona, 2000. 22. Taylor, L. D. “Telecomunications Demand in Theory and Practice” Kluwer Academic Publishers, 1993. 23. U.S. Department of Commerce, and National Telecommunications & Information Administration, “A Nation Online: How Americans Are Expanding Their Use of the Internet”, February 2002. 24. Varian, H. “The Demand for Bandwidth: Evidence from the INDEX Project”, Mimeo, University of California, Berkeley, 2002.

28

L. Cerno & T. Pérez

Internet Demand in Spain APPENDIX I

Ratio Broadband/Narrowband for Household Size 70 60

55.1

50

58.8 46.4 49.2

40 %

40.5

30 20 10 0 less than 10,000 inhabitants

10,000 to 20,000 inhabitants

20,000 to 50,000 inhabitants

50,000 to 100,000 inhabitants

more than 100,000 inhabitants

29

L. Cerno & T. Pérez

Internet Demand in Spain Appendix II Internet Access in Home : First Model9

Outcome Equation:

Internet Access Demand in Home Regressors: Constant

Broadband ( BBi )

Narrowband ( NBi )

0.61 (0.08)***

0.35 (0.07)***

-0.17 (0.07)**

0.14 (0.07)**

0.17 (0.03)*** 0.08 (0.03)*** 0.07 (0.03)**

0.10 (0.03)*** 0.03 (0.23) -0.05 (0.03)*

0.01 (0.02) -0.0014 (0.01) -0.02 (0.003)*** 0.03 (0.01)** -0.01 (0.00)**

0.03 (0.02)* -0.00006 (0.01) 0.01 (0.003)*** -0.02 (0.01) 0.01 (0.00)***

0.000055 (0.000025)**

-0.000056 (0.000022)**

Constant

-4.05 (0.06)***

-4.03 (0.06)***

Household Income ( isi )

20.67 (0.38)*** 0.25 (0.01)***

20.49 (0.38)*** 0.26 (0.01)***

Economic Attributes Household Income ( isi ) Technological Attributes Connect with Personal Computer ( ord1i ) Connect with Laptop ( laptopi ) Connect with mobile telephone ( mobilei ) Social and Demographic Attributes Be studying ( ecoursi ) Members in home ( housemi ) Habitat of origin ( habitati ) Sex of the respondent ( sex _ hi ) Age of the respondent ( agei )

Age of the respondent square ( agesqi ) Selection Equation:

Internet Access in home (ACCHOME i) Regressors:

Internet Usage ( usagini )

9

Dependent variable is a dummy equal to 1 if the respondent has broadband/narrowband at home, and 0 otherwise. Standard errors are in parentheses. * , ** , *** statistically significant at 10%, 5% and 1% respectively.

30

L. Cerno & T. Pérez

Internet Demand in Spain

Study Level ( levelest i ) Selectivity Correction ˆ (x) IMR Number of observations Censored observations Log likelihood Wald test of independence ( H 0 : ρ = 0 )

-9.62 (0.20)***

-9.52 (0.21)***

-0.19 (0.02)*** 18,940 14,330 -9,071.8 139.04***

0.07 (0.02)*** 18,940 14,330 -8,681.

Table 2: Marginal Effect of Household Income Broadband 2.8108 (0.0048)***

Narrowband -0.9665 (0.0018)***

0.0060 0.0050 0.0040 0.0030 0.0020 0.0010 0.0000 -0.0010 -0.0020 -0.0030 -0.0040 -0.0050 74

69

64

59

54

49

44

39

34

29

24

BB' NB'

19

14

Contribution to the Probability

Graphic 8: Impact of Age in the Probability of Choosing some Type of Internet Connection (*)

Age

The continuous line means that the probability to have broadband in home grow with the age, whereas contrarily occurs with the pointed line referred to have narrowband in home, with the inverse impact in probability.

31

L. Cerno & T. Pérez

Internet Demand in Spain Appendix III Demand of Broadband in Home

32

L. Cerno & T. Pérez

Internet Demand in Spain APPENDIX IV

A Construction of Proxy for individual income

When we estimate demand functions for Internet, it becomes necessary to use variables that make references to the economic attributes. If we referred to the product attributes, this one is the price, and if we referred to the economic agent attibute, these attributes are generally those to make references to the individual or household income. Since this information is not available in our sample, and taking into account that the decision to contract an Internet access for home, for example, according to the price but also according with the household budget constraint, it is necessary find some variable that represent this topics mentioned in the strict sense that they represent, is this to say market power in the case of price, and purchasing power in the case of income’s types. Even so: in this case the variable price is an product attribute gives the different alternatives of connection to Internet (connection via conventional phone, DSL, ISDN and wire). We have opted to omit it considering that when using a sample like crosssection, and when being the regulated telecommunication’s prices for all Spain, this it would be a variable with little or no variability. In the case of income: we will use the concept of permanent income (Friedman, 1957), which may be more useful than the annual average income to analyze the election carried out by the consumer on the access and use of the Internet service. To obtain an indicator that serves as a proxy of the permanent household income, and given the availability of information contained in the sample survey, we have opted to use the method that Goodman and Kawai (1982) applied to the housing market. They use a model of human capital where the income is determined by the investment in human capital and non-human capital. In our case we consider all the observations in the sample (18,948 individuals interviewed in 2003 all over Spain), being considered an index by means of the following specification:

isi = 0.625 [ % HK i ] + 0.375 [ % NHKi ]

where: isi : Household income index. HK i : Investment in human capital of the i th agent. NHKi : Investment in non-human capital of the i th agent.

33

L. Cerno & T. Pérez

Internet Demand in Spain

1) Variables considered in the construction of HK i : This includes: a) Being studying at the moment. b) The level of education. These variables are quantified through seven dummy variables: Table A. Variables for Human Capital Variable Ecursa 1st level 2nd level 3th level 4th level 5th level 6th level

Definition =1 if the person intervewed are studying at the moment; =0 otherwise =1 if the person interviewed doesn’t have complete studies; =0 otherwise =1 if the person interviewed has completed the primary education; =0 otherwise =1 if the person interviewed has completed the first stage of secondary education; =0 otherwise =1 if the person interviewed has completed the second stage of secondary education; =0 otherwise =1 if the person interviewed has completed curses of vocational training; =0 otherwise =1 if the person interviewed has completed an university career gives more than three years; =0 otherwise

We have considered that on average, it takes 16 years from entering the first grade of primary school until obtaining a university degree. Then, when considering six levels of studies, and giving the same weight to all of them, we will suppose that they take 2.66 years for each conditional level. Accordingly with that, each level will be weighted with the following coefficient:

πi =

2.66 j 16

j = 1, 2, ..., 6.

Then, the percentage of human capital will be given by: HK i = 0.5 ecursa + 0.5 π i

2) Variables considered in the construction of NHKi : Give the data in the survey, and the purpose of the model, we will consider the investment in non-human capital as

34

L. Cerno & T. Pérez

Internet Demand in Spain

the investment in technology. This investment will be quantified in nine dummy variables such as we can to see in the Table B of variables:

Table B. Variables for Non-Human Capital Variable Definition =1 if the person interviewed receives channels through parabolic x1 antenna; =0 otherwise =1 if the person interviewed receives channels through cable; =0 x2 otherwise =1 if the person interviewed possesses a personal computer; =0 x3 otherwise =1 if the person interviewed possesses laptop; =0 otherwise x4

x8

=1 if the person interviewed possesses a fixed phone number; =0 otherwise =1 if the person interviewed possesses mobile telephone; =0 otherwise =1 if the agent interviewed possesses high fidelity music or musical chain; =0 otherwise =1 if the agent interviewed possesses video; =0 otherwise

x9

=1 if the agent interviewed possesses dvd; =0 otherwise.

x5 x6 x7

Later on, and following the same approach that, the measure of non-human capital it will be:

9

NHK i =

∑x h =1

h

9

h = 1, 2, ... 9

The weights used can vary according to different approaches. We have preferred to leave all the magnitudes inside each term with the same weight, although we have given more importance to the human capital on the whole. The human capital we can considerate more permanent than non-human capital because the years of study reached, for example, remain with the person for ever and have a direct relation with the job-market earning, though the variables considered here for measure the non-human capital are ever associate with consumer goods with quickly expiration or change date.

35

Suggest Documents