A Voice of the Web

Steph's personal blog on professional matters


This is the blog of Stephane Boyera...Yet another blog about mobile, ICT and Social & Economic Development in the South.

A short blog post for once and not many months after the previous one! I spent most of the last weekend translating in French the World Bank OGD Readiness Assessment Tool. This is a great document, and I recommend its reading to anybody interested in the topic. It is very well structured and covers all major points. I’ve worked on a similar methodology a couple of years ago with my colleagues (see e.g. OGD feasibility study in Ghana), and this document is surely far more developed and covers more dimensions.

But as usual, a good blog post needs to be a bit controversial! Therefore, I wanted to make a few comments on this document, developing a bit further what I wrote in a previous post related to OGD in Developing Countries. The essence of my view is that the types of barriers that can exist in developing countries are slightly different than the ones in Western countries. The differences are centered on 2 aspects:

• The role of an OGD initiative
• The demand side, and in particular the concept of “empowering citizens”, and the role and type of ICT.

On the first point, I think it is essential to separate clearly and differentiate OGD from transparency and accountability (T&A). There is no question that OGD is very helpful and can strongly support T&A. But this is only one aspect of OGD. Other aspects include social and economic development, business opportunities, increased efficiencies of public services etc. T&A is only one benefit or one angle of attack, but usually not the easiest one. When you are in a country that is a recent democracy, that has decades of culture of secrecy, people (politicians, public administration) are quite scared to change completely their way of working, and they are scared about what people could do with the data they would release.

In such context, T&A is perceived as very threatening. It is therefore essential to identify what could be the potentially most impactful theme for OGD (ICT industry development, social and economic development, etc.) and focus on this aspect (the low-hanging fruit). I believe it should be part of the methodology to identify this theme. The focus of the future OGD initiative would have a major impact on how the assessment is structured. In particular, in section 4, depending on the theme, the list of datasets to explore will be very different.  In the same way, the ecosystem and their members will be slightly different. So I believe that identifying the low-hanging fruit is the key activity that should be conducted i.e. identifying the list of areas/dataset that are the most useful for a community interested in the process and that is seen as non-sensitive for the government. Then, when an OGD is already in place, it would be IMHO easier to fight for specific more sensitive datasets such as budgets, etc.

I’m also a bit concerned about what is considered as positive or negative evidences. At a macro level, it is obvious that building an OGD initiative on a strong ground is easier. But it is also obvious that this ground does not exist yet in most of low and middle-income countries (LMICs): take e.g. 40% internet penetration rate, records in electronic format, etc. But those are not negative evidence IMHO, but instead opportunities for an OGD initiative to have a greater impact on society. Ok it will make things longer, tougher, but with a far greater impact. Let’s just take the case of digitalizing records. There are tons of LMICs countries not engaged in this process. This is e.g. a huge opportunity for youth workforce development, and ICT capacities development. Few governments are now following this path like e.g. in Ghana. I’m personally more in favor of looking at the big picture, and look at possible actions that would make an OGD initiative possible instead of just declaring that there are too much negative evidences.

In that sense, it is essential IMHO to sort the evidences. Without political willingness at the top level, without internal champions particularly in the public administration, and without a group of CSOs (with champions) that are willing to work with the government in a constructive way, no real chance to go anywhere. Those are IMHO some kinds of pre-conditions. Then if such pre-conditions exist, then you can evaluate if it is going to be easy or hard depending on the other dimensions.

Now, more specifically on ICT, which is my core domain of expertise compared to OGD, I feel that the assessment is a bit weak. In few words, I feel it is based on the western model of web based access, with a very slight mention of mobile, but in the form of apps (e.g. smartphone). We all know that this is irrelevant to 99% of e.g. the Sub-Saharan countries. The current penetration of smartphone if very low, the current internet usage rate is around 15% (see sources) at the continent level, so there is no way that the proposed model is appropriate. On the other end, we also know the power of mobiles in these countries, and in particular the development of e.g. SMS-based applications or voice-based applications. These types of mobile services are intrinsically very different. The smartphone apps or internet-based apps model is an open model that is well established all over the world: almost anybody can create content and apps and make them available to anybody else in the World by sharing a URI. It is a direct developer-to-consumer link, and the underlying infrastructure (internet) is outside the scope of the developer.
In the case of mobile specific service like SMS, it is a completely different model, where the developer has, in most cases, to take care of the infrastructure (getting e.g. a phone number) and has to negotiate with mobile operators. This is where the situation varies a lot from countries to countries. In some countries, the regulation is in place and ensures that third-party developers and service providers have rights related to service deployment. This relates e.g. to getting cross-operator shortcodes. In some countries, there is a strong mobile community, tech. hubs, etc. and mobile operators are supporting them. E.g. Kenya, Ghana or Nigeria are surely in this category.
In some other countries, it is not the case: operators are maintaining walled-gardens; there is no activity outside them, no third-party independent service developers, no community no tech hubs; no capacity and no curriculum in mobile technologies. In such settings, the setup and launch of mobile services is very difficult and the offer is limited.

From my perspective, it is essential to explore, as part of an OGD assessment, the details of the pre-existing ICT ecosystem, with a specific focus on locally-relevant mobile technologies, beyond just the traditional western model.

I believe we can even go a bit further, questioning the overall model and role of ICT in an OGD initiative in LMICs. In Western countries, take e.g. US or UK, ICT and the Web is the primary access mechanism to data and services and it is therefore essential to ensure that civic hackers exist to transform open data in nice meaningful understandable services that one can access on the Web. But in countries where the use of ICT (mobile or traditional) is low and where illiteracy is high, it is not a channel that can help empower citizens. So one way to go is just to say that there is no way an OGD initiative is going to make an impact in such countries. But then you remove a big part of the LMICs. Another option is to realize that ICT is not the right channel and look if there are other channels. What is the primary source of information for people? Newspapers, community radios, TV, etc. Are these media free? Are they excited to go into data journalism? If the benefits of an OGD initiative will reach people not through ICT but through traditional media or informal channels (like e.g. NGOs) then the readiness assessment should specifically focus on this dimension. This leads to very different challenges and roadmaps to success IMHO, specifically on the demand-side.

All these dimensions are already in the readiness assessment tools, but there are all put at the same level. It is essential to look at specific dimensions in specific context. In LMICs, investigations on mobile sector, media and infomediaries are surely far more critical compared to Western countries. Each section should therefore have a different weight depending on the local context.

All in one, this is again a great document for me, but I believe that it should be expanded to cover specificities of LMICs, and consider OGD as a potential way to really impact the country at large, and not only its T&A part. It is also essential IMHO to integrate new models of ICT or channels to provide the benefits of OGD to citizens. The models of smartphone Apps/Desktop Web is great for Western countries, but there are other models to achieve a similar impact. Finally, I think that it may be useful to have an overall OGD readiness scale that split the evidences/findings in different groups and identify clearly the pre-conditions, and then the level and type of challenges that are presents that would make the implementation very hard and long or easy and quick. In other words, I believe that the assessment tool might be more efficient if it identifies a few different types of country profiles and instantiate the methodology for the different profiles to highlight the specific challenges that need to be addressed.



2 Responses to “OGD Readiness Assessment in low and middle-income countries”

  1. Dear Steph,
    Thank you very much for your kind words and for your thoughtful comments. They deserve equally thoughtful replies. While I digest these, allow me a few quick reactions:
    1. On T&A not being the same as OD, I fully concur. As a matter of fact, OD goes beyond T&A both in its possibilities and even in the technicalities, in the sense that good OD models are two-way whereas T&A is often one way. Technically, you can be fully transparent and accountable without providing machine-readable data.
    2. I also agree that weak pre-requisites for OD, such a digitalization of data, data quality issues, and many others, represent opportunities for stronger impact from an OD initiative. I also think that most (maybe all) of my colleagues at the Bank would agree.
    On the previous two, we will discuss your comments internally and review the ODRA methodology. It is a living document, so we appreciate such comments.

    On your third point, about adapting the OD initiative to the technological context in each country, I am not sure I agree 100% with your comments. Raising this point is very important and I fully agree that we cannot take for granted things that are taken for granted in more developed countries. I also agree that more emphasis on mobile apps that would work on G2 networks is needed. These too, shall be discussed internally with a view to make necessary improvements to the methodology.

    That said, there are two reasons why I am not in full agreement. First, you mention Africa and quote internet penetration rates in this continent. But bear in mind that the Bank clients include several countries with far higher internet penetration rates that either do not have open data initiatives or where these initiatives are incipient. I am sure you would agree that, in such contexts, it would be senseless not to use the existing internet capabilities in an OD initiative.

    Secondly, even in the poorest countries, internet penetration may be low, but look at its growth. Using your same sources, the number of internet users in Africa has grown from 4.5 million in 2000 to 167 million 10 years later. These are growth rates even higher than growth in mobile G2 users a few years ago. In my opinion, it is highly likely that internet expansion will follow the same patterns of mobile expansion, so planning for an internet-dependent OD while developing mobile-dependent OD seems the way to go in these countries.
    In sum, point well taken that recommendations need to be adapted to the context. We all agree with that.

    Thanks again.
    Amparo Ballivian

    Amparo Ballivian

  2. Dear Amparo,

    Thanks a lot for your comments and detailed analysis.
    I’m very glad to see that we are sharing the same view on almost all the points.
    Now about what you consider as divergent views, I think we are also on the same line. Perhaps I didn’t make it clear, but obviously, where internet penetration and usage is appropriate then it is a far more powerful option than basic mobile services. I’ve been trying to say that the methodology should not be flat but integrates a decision tree where that will specific investigations to be conducted based on the specific context. So if the primary access is basic services, then detailed investigation on that model should be conducted. if you are in a country were the mobile apps/desktop web is relevant, then no need to go for the basic-mobile service investigation.
    Finally, we perhaps slightly disagree on your last point. I’m a bit uncomfortable see governments and others always focusing on the next revolution. Before, we didn’t have any way to deliver ICT services to people. Now, we have basic mobiles, which are still going to prevalent for quite long time accroding to industry forecast, particularly for people at the base of the pyramid. So we don’t have to wait for broadband/internet/smartphone to exploit this opportunity. I think it is essential to build an OGD initiative on an existing ground, and not on a ground that may come sooner or later. However, I’m also not sure that it should be either internet or basic-service model. I’m not going to develop that here, but it is imho essential to integrate in any service development the fact that technology evolves very quickly. It is just a design methodology: one should not develop an SMS service or a voice-based service, but a service that have different channels (sms, web, voice, etc.) and where users can migrate from one to the other as they have access to it. It is obvious that more advanced devices and better connectivity leads to better/more poweful services. It is therefore critical to design services that can take into account these different channels.



Leave a Reply

You must be logged in to post a comment.