Sweden is the best country in the world at making the web work for its citizens, whilst Yemen and Zimbabwe are the least effective, according to a new global survey backed by the inventor of the world wide web, Tim Berners-Lee.
The web index, as it is known, claims to be the “most comprehensive effort yet” to measure the impact of the web around the world and how it is changing people’s lives.
It ranks countries by how much benefit their citizens derive from the web and how it has transformed areas such as politics, economics and society. It also assesses any barriers a country has to allowing its citizens to access a free and open web, finding that around 30% of countries face some sort of government restriction.
The survey also revealed that just one in three people use the web globally, a number that drops to fewer than one in six in Africa.
Sir Tim is the public face of the new index, but it was economist Hania Farhan who was tasked with building it.
BBC Future caught up with her to find out how it was done and what it all means.
Why build the web index?
Tim Berners-Lee and others at the Web Foundation noticed there were lots of stories and anecdotes about how the web had transformed people’s lives through education, social networks, politics and so on. But there was very little data as such. So they said “why don’t we try to come up with an index to measure the value of the web and its impact on people?” That was where I came in. I had been working on the Ibrahaim index of African governance [which annually ranks the quality of governance in Africa] for three years. My background is as a quantitative policy economist and I had experience of building those kinds of things.
Where did you start?
You start from the basic idea of how do you, as an individual, get value from the web. We broke it down into three basic components – infrastructure, content and impact. The first thing you need is to get on to the web. So, a country needs the communications infrastructure – broadband or a mobile device, electricity, access to computers etc. Then you also want the institutional infrastructure - you need to have a minimum level of literacy, you need freedom of access and you need to have policies that allow that freedom of access. With those two in place you are online. And then what? Then you need relevant content in a language you understand to be able to use it. So if I get on the web and it is all in Russian, I get zero value as I do not understand Russian, sadly. Then finally the third layer, which is arguably the most relevant - and the one we put the most emphasis on in the model - is the impact. That is measured in terms of political, social, economic, developmental impact etc.
We then simply went out and scoured databases - from the likes of the World Bank, the UN, the ITU, the World Economic Forum, Reporters without Borders, Freedom House, etc - to find indicators to populate these various components of the index. In the end we chose 34 indicators from the data we found.
Were there any questions you just couldn’t answer or find data for?
Some of the questions that many people kept asking were ‘how many pages are on the web?’ and ‘how many pages are there in each language?’ Those data do not exist because there is no centralised unit - a web central intelligence unit of sorts. When we looked at other questions, especially in the area of web impact, we could also see that there was a massive data gap. So, I said let’s gather our own data. We identified experts in all 61 countries and asked them about 60 questions. For example, is the web used for political mobilisation in your country? They were all then reviewed by other experts to make sure that the scores were consistent and had evidence to back them up. That gave us another 51 indicators and allowed us to plug some of the gaps in the data.
All of the data was then put through the model - which weights different elements - to come up with the final scores.
What was the result?
In first place was Sweden, followed by the US, UK, Canada and Finland in that order. But taking into account margins of error, the four after Sweden are really not much different from each other. At the bottom, were Yemen, Zimbabwe, Burkina Faso, Benin and Ethiopia.
What does it mean to be the top country or even the bottom country?
That is the question isn’t it? The idea is that if you are at the top, then the people in that country are - according to this index - getting the most value from the web. And at the bottom they are getting the least value. The index is a tool for leaders and policy makers so that they can use the web much more effectively to improve people’s lives, be it economic, social or political. For example, a policy maker sitting in Iceland or Italy may look at Sweden’s ranking and ask why does the web seem to serve the people of Sweden apparently better than it does their own people... and therefore what can we do about it to make our citizens better served by the web?. They can zoom into the indicators and see how they can change their country’s performance. Similarly, Ghana might ask why Kenya ranks higher than it does. The lower you are in the rankings, the more indicators you may want to change to affect the lives of the people in that country.
Some people may look at the index and see countries like China in 29th place ranking higher than they might expect, and countries like Japan and Korea [in 20th and 13th places respectively] ranking lower than you might expect. Can you explain this?
The web index is an average number calculated from lots of underlying indicators. A country could therefore do relatively well in one or more area, say economic impact, but relatively poorly in other areas, say political impact. So, on average, it will score not so badly, but not so well either. In fact, China scores relatively poorly in the political and institutional components, but middling in all other areas. Hence, overall, on average, it ranks in the middle. Also, we need to keep in mind that rankings are always relative to how other countries are doing. They do not just reflect that country's own performance, but its performance relative to how the others are doing.
Why were there only 61 countries included?
One of the reasons was getting consistent and comparable data. But to be honest, we were also limited on resources. Google kindly gave us the initial seed funding, but it is expensive to do this kind of thing, particularly collecting your own data. And so we ended up being able to choose close to 65 but we had to leave out a couple because we weren’t too happy with the results. We had to get a spread of countries across the different continents. So it wasn’t an active decision for instance to leave out Denmark or Saudi Arabia. We just needed to make a decision given the resources and the time limitations we had.
The whole thing is framed in a very positive way, as if the web only has positive impacts. But there are also negative effects, like cybercrime and censorship. Did you take these into account?
The negative impacts are a very important aspect to include. The problem was that there is so little data. There is anecdote and sometimes there is country specific data, but, as far as we could see, there hasn’t been much work done on that. For example, to what extent has criminal activity increased as a result of the web? We could not actually find any data for a large enough country sample for multiple years to be included. So, all we could do was put in a few questions in our own data survey. We have two questions – basically to what extent do you think that the web is making it easier to undertake criminal activities in your country and then the other side which is: are there laws against cybercrime in your country. So in answering that question, any expert needs to consider: do those laws exist and how effectively are they implemented.
Of course we need many more questions answered in this area. Also, for now, the questions are framed in that way because they need to take into account cultural differences. That is a very difficult thing, which you have to handle very carefully. For example, how do you factor in that gambling is illegal in some countries? Framing the questions was actually one of the most difficult parts of designing the survey.
Will you try to fill these gaps in subsequent years?
One of the things I would like to do is for sure add more countries and add more survey questions to cover more of the areas, as well as the negative effects. We also want to work with other organisations that produce data that cover some of the dimensions of the web index and ask them to help us cover more countries. For example, Reporters without Borders produce something called Enemies of the Internet. It is totally descriptive and it covers a relatively small number of countries, meaning we can’t use it. So we need to work with organisations like them to expand the country coverage and try to quantify their results.
Finally, you openly admit you are not a “techie”. Was that a problem?
When I was first asked to construct and produce the index I told the Web Foundation that I wasn’t sure that I was the right person, as I had no clue about the back end of a computer. But they said that was not a problem and, in fact, it was my other skills they needed. I could see things from the point of view of the average user, and the index is targeted at a far wider audience than just techies. I would love to know how to write code and programme, and one day I might learn, but for now, I’m just about managing to keep up with their “special language”.