Analysis of W3C Office Coverage in Europe

Project acronym: QUESTION-HOW
Project Full Title: Quality Engineering Solutions via Tools, Information and Outreach for the New Highly-enriched Offerings from W3C: Evolving the Web in Europe
Project/Contract No. IST-2000-28767
Workpackage 3, Deliverable D3.1

Project Manager: Daniel Dardailler <danield@w3.org>
Author of this document: Ivan Herman <ivan@w3.org>

Date: 29 November 2001.

Introduction

This report gives an overview of the current coverage of W3C presence through the network of Offices, and compares it with relevant statistical and other data. The goal of the report is to work out the detailed plans within the QH project directed to a better overall coverage within the European Union. The QH project envisages two parallel actions in this respect (Workpackages 3 and 4), namely:

• Regionalization of current offices
• Establishment of up to four offices, two in the current EU countries, and two in the candidate states

As a result of our analysis concrete plans for both Workpackages is now available, outlined below, which will be implemented in the following phases of the QH project.

Collected data

The following data contributed to the final decisions:

Current membership in W3C per countries

The data reflect a snapshot of the W3C membership database of autumn 2001; although variations occur, they do not alter the overall conclusions

Statistical data on the Internet/Web presence in various countries

The analysis aimed at using publicly available data on:

• Number of PC's per country
• Number of hosts per country

I was felt that these data give an adequate measure of the internet 'activity' in an area. The various sources which were considered to find these data included:

• RIPE network coordination centre (http://www.ripe.net)
• Internet Software Consortium (http://www.isc.org)
• Nua Internet Surveys (http://www.nua.net/surveys/)
• Netsizer (http://www.netsizer.com/daily/TopCountry.html)
• Various statistics published by the ITU (International Telecommunication Union)

Unfortunately, it is notoriously difficult to collect and compare these types of data, because the statistical methods used by various organizations differ. Finally, the ICT statistical data, published by the ITU, were retained for the purposes of this analysis. These reports have indeed the advantage to collect both types of data together, ie, the statistical methods used are (probably) similar. The two documents (freely available on the net) are:

• Internet Indicators at Glance, 2000
• Internet Indicators at Glance, 1999

Using both the 1999 and 2000 data allows to map some trends, too.

GDP data (GDP per capita, GDP growth)

These data could be collected from several sources. The data published by the World Bank, at http://www.worldbank.org/data/countrydata/countrydata.html was used in the present report. (To normalize the results, the country population data was also used, available from the same source.)

It must be noted that the GDP and Internet data stem from the period before the current economical downturn and they certainly do not reflect the possible consequences of the September 11 events. However, it is believed that the main trends and results remain valid on long term (and setting up new offices and/or regionalizing a current office is certainly a long-term investment).

The data themselves

The following table shows all the collected statistical data:

The table includes the collected data for all EU countries plus the strongest candidates to the EU at the time of writing this report. Note that, although a participant of the QH project, data for Israel are not listed; due to the present political situation it is indeed impossible to include Israel in any kind of regionalization project. Countries where an office or the European host operate have been highlighted in the table.

The following few charts, extracted from the table, allow for a better appreciation of some of the conclusions:

Figure 1: Number of W3C members in the various European countries.

Figure 2: Number of PC-s per 100 inhabitants in various European countries. The top and the bottom row for each country represent the data for 1999 and 2000, respectively.

Figure 3: Number of Internet hosts per 10,000 inhabitants in various European countries. The top and the bottom row for each country represent the data for 1999 and 2000, respectively.

Analysis

Although some of the trends, reflected in the data, are well known (eg, the strong internet presence of the Nordic countries), some conclusions are important from the project's and W3C's point of view.

• There are a number of countries where the local internet activity measured, for example, by the number and the evolution of internet hosts and/or PCs is high but the number of W3C is comparatively quite low. In other words, these countries are "underrepresented" in W3C. The clear examples are: Austria and Denmark, but even Finland, Ireland, or Belgium fall into this category.
• The presence of an office or the European host has a clear positive effect on the evolution of membership, hence the presence and influence of W3C technologies. In spite of its undeniable industrial strength one could argue, for example, that the large number of UK members is an "overrepresentation". Obviously, this is only for the sake of the argument; the goal is not to reduce those numbers but to increase the numbers in other countries!
• Some countries, mainly Spain and Portugal, are still on the lower end of the scale in terms of Internet presence. This is in contrast with the relatively large growth in GDP of both countries. It could be expected that a strong local office might play a beneficial role in the local evolution in web technologies.
• Among the potential new EU members, the level of Estonia and of Slovenia in terms of those statistics are both impressive; they score better than some of the current EU countries. A relatively low population in these countries does not make them a proper choice for a full-blown office, but it is worth noting this strength nevertheless; a conclusion might be that it is beneficial to include them in regional initiatives, if possible.
• Looking at the larger candidate countries, the Czech Republic and Hungary score relatively well, with Poland a bit behind them. The GDP evolution of Hungary and of Poland are impressive, and this makes them good targets for further "investment". However, only Hungary has a stable W3C member; there have been Polish and Czech members in the past, but none of them stayed too long. (The Hungarian member, SZTAKI, has been a stable member ever since the creation of W3C.)

One has to emphasize, though, that planning for new offices and/or for the regionalization of existing offices cannot be done based exclusively on cold statistical data. A number of other factors, not expressible by numbers, influence the decisions:

Local contacts.

An office cannot operate without the dedication of local persons and without a local institution ready to take the possible load of an office. Actively seeking out such institutions have been part of the work reported here, and has greatly influenced the current decisions.

Linguistic issues

The role of an office is, also, to be a local "point of contact" for W3C in the country. This is a major constraint on regionalization.

Cultural, political issues

W3C has to consider possible cultural incompatibilities, historical grievances, etc. W3C certainly wants to avoid finding itself in the midst of such problems. Consequently, during the past period, W3C has contacted the representatives of all W3C members, but also other organizations, private contacts, etc, to address these issues (eg, "how would the Irish community feel about being represented from Oxford...").

Conclusions, Plans

Based on all considerations, the plans for the QH projects are as follows.

Regionalization Plans (Workpackage 3)

With a start in 2002:

UK & Ireland office, extending the current UK office

The current office host (RAL) has already good contacts with the academic circles in Ireland (eg, at the Trinity College), which is a good basis for start. Discussions with the representatives of Irish W3C members showed a great interest for such an extension of the UK office.

Office for Germany and Austria, extending the current German office

The contacts between the German and Austrian technological and industrial communities are traditionally good, and various professional societies, like for example the German GI (Gesellschaft für Informatik) have regularly meetings with their Austrian colleagues. As the statistics shows, Austria is clearly underrepresented in W3C, but the relatively small population of the country does not make it a primary target for a separate office. In view of the linguistic facilities, creation of a joined German and Austrian office is the best solution. Note that the final name of the office is still not decided.

Benelux office or Dutch-Flemish office, extending the current Dutch office

That there is a need for a presence in Belgium is clear from the statistics, and the closeness of the Dutch office leads to the obvious choice for regionalization in the area. The relationship between Dutch and Flemish industry, academics, are traditionally excellent.
At the moment of writing this report it is still not clear whether the Dutch office could be extended to the whole of the Benelux; the problem is obviously the linguistic issue. The Dutch office is currently exploring the possibilities, based on a dedicated cooperation on the matter with the European (ie, French) host. If the final answer is not to establish a full-blown Benelux office, then the plan is that the French host will engage into a more active representation in Wallonia and in Luxembourg.

With a start in 2003:

Scandinavian office (ie, Sweden, Denmark, and Norway), extending the current Swedish office

Cooperation based on Scandinavia is already traditional, and this therefore is an obvious choice for regionalization. The very strong presence of the Nordic countries in the internet area makes it clearly necessary to have a regional presence of W3C. Although Danish, Norwegian, and Swedish are not identical, these languages are close enough to make such a regional office viable. However, at this moment, the Swedish office is undergoing some internal changes, and it is not feasible to start this action in 2002 already.

Greek and Cyprus office, extending the current Greek office

Linguistic and cultural bounds make this regionalization step quite natural. However, the Greek office is currently busy reinforcing its presence in Greece; it is therefore not feasible to start this action in 2002.

New offices (Workpackage 4)

In current EU countries

Finland

Although a relatively small county in term of population, Finland is clearly an important player in the Internet area. Linguistically "isolated", it is not feasible to include Finland into a regional office of some kind, eg, a Scandinavian office. Hence the plan to establish a separate office there. Note that the first contacts in Finland also included the possibility to establish a regional office from the start, including Estonia, too: the two languages are almost identical, and there has been traditional bounds between the two countries, mainly in the last 10 years. In view of what has been said about Estonia, this would clearly be the best solution.

Spain

As described above, this country is clearly underrepresented, not only in the W3C are but on the Internet, too. The first discussions with local contacts have revealed that linguistic issues play an important role in this, insofar as the local community is not fluent enough in English to closely participate in the usage and the dissemination of W3C technologies. Coupled with a strong translation programme, a Spanish office could play a very important role for W3C, with an effect in the whole Latin-American community, too.
It can also be envisaged to enlarge a Spanish office later into an Iberian office, although the feasibility of such a move (in view of some existing animosity between Spain and Portugal, as well as linguistic differences) is not clear.

At this moment it is not clear which of these two offices will be established in the first phase of the project, and which in the second. W3C is working on both, trying to find the local contacts and institutions which might be willing to engage into this activity, but there is no conclusive result yet.

In the candidate EU states

Hungary

As shown by the statistics, this is one of the strong countries in the area, with a stable W3C member that has already expressed its interest in establishing an office. Similarly to Finland, the linguistic isolation of the country makes it difficult to include Hungary in any larger regional office. There are good chances for an office to be established in 2002.

Czech Republic, Slovakia, and/or Poland

At this moment it is not clear which of these countries could host an office in the second part of the project. There are pro and contra arguments for all three of them (with the case of Slovakia being a bit weaker, again looking at the statistics) and, at the end of the day, the quality of local contacts and possible institutions may become the decisive factor. Linguistically, the three countries are not very different (people mutually understand one another), so a regional office would not be excluded either. This will depend on the suitable institution coming forward.

If these plans are implemented, a close to 100% "coverage" of today's European Union, as well as some of the new members possibly joining the union in 2004, will be achieved.