- Global data traffic has increased one thousand times over the last 20 years.
- Data infrastructure policy helps to make the modern data economy more equal.
- The World Development Report 2021, Data for Better Lives seeks to highlight and address inequalities with regards to data connectivity.
- This includes helping countries have access to modern data infrastructure to ensure reliable and affordable connectivity for their population.
An explosion of data
Global data traffic has increased one thousand times over the last 20 years. Data travel thousands of kilometers along this seamless global data infrastructure supply chain at breathtaking speeds of 200,000 kilometers per second—meaning that digital data can, in principle, circle the globe five times within a second. This explosion of data creates an ever-growing demand for data infrastructure.
Data Infrastructure policy, one of the building blocks of a data governance framework, helps to level the playing field in the modern data economy making things more equitable. Without modern data infrastructure at the national level, countries are unable to provide affordable and reliable connectivity to their population. At the same time, without access to broadband infrastructure at the local level, people cannot gain access to the wealth of data that exists, nor can data about those people be captured—so that their needs and preferences can be taken into account in the design and provision of public and private services. The World Development Report 2021, Data for Better Lives advocates for improving fairness in the global data system by addressing inequities in the availability of data infrastructure, both inequities between people and inequities between countries.
Have you read?
Effective and efficient management of modern data flows requires that countries have access to a modern data infrastructure at the national level. Due to lack of competition and good governance, such modern data infrastructure is much less prevalent in LICs than elsewhere (Figure 1), bringing disadvantages in terms of cost, speed and performance.
Figure 1: The global fiber-optic cable submarine network reaches all corners of the world, but data infrastructure is unevenly developed
The starting point for national data infrastructure is the establishment of Internet Exchange Points (IXPs), which allow domestic data traffic to be exchanged locally, without the need for the data to travel vast distances to reach overseas IXPs, incurring significant costs and time delays in the process. For example, Latin America spends around US$2 billion a year for international bandwidth—a sum that could be reduced by one-third through greater use of IXPs. Overall, countries relying on overseas exchange of data have fixed data charges that are 35 times as high as those with full modern data infrastructure, and mobile data charges that are seven times as high.
While the development of IXPs does not call for huge investments, governance arrangements are critical to their success. Two contrasting stories from East Africa are instructive in this regard. In Djibouti, DJiX acts as a regional hub for data exchange that effectively serves the East Africa region, however, due to a domestic market monopoly, local internet users in Djibouti nonetheless face very high data charges. The situation is different in Kenya, where KIXP operates as a nonprofit with a multi-stakeholder board, attracting a wide diversity of participants including international content and cloud providers that help to improve the affordability and performance of data services in the country.
Once IXPs are established, countries need to create an environment for the private sector to invest in Colocation Data Centers, which—among other things—allow popular internet content from overseas to be stored locally, greatly reducing the cost and increasing the speed of usage. However, at present, there are more data centers in California than the whole of Sub-Saharan Africa. For the private sector to be willing to invest in Colocation Data Centers, a highly stable physical and regulatory environments is required, combined with abundant low-cost renewable energy. These conditions can be challenging for low and middle-income countries to meet.
Cloud computing services overseas are becoming increasingly critical to access state-of-the-art software services and take advantage of artificial intelligence. However, hyperscale global cloud data center providers operate in just a handful of larger emerging markets such as Brazil and South Africa. While it may not be feasible to develop cloud computing in smaller LMICs, the provision of cloud on-ramps—cheap, fast, and secure direct links to cloud computers that bypass the public internet—is emerging as a valuable alternative. Cloud on-ramps do not yet exist in any LICs, but are available in some MICs that have the Colocation Data Centers to host them.
Significant progress has been made during recent years in rolling out coverage of 3G internet services, which has grown from 79% in 2015 to 92% in 2018; even if coverage continues to remain much lower at 71% in Sub-Saharan Africa.
However, internet coverage does not necessarily translate into internet use. A staggering 40% of the world’s population live within range of a mobile signal but nonetheless fail to make use of it , and this share (which is known as the “usage gap”) has stubbornly persisted over time even as coverage has increased (Figure 2). The situation is even worse in some developing regions, with this usage gap rising to 49% of the population in Sub-Saharan Africa, and as high as 64% of the population in South Asia—the latter being equivalent to 1 billion people in all. There is also an important gender dimension to internet uptake with globally 250 million fewer women than men using the internet, many of these concentrated in Asia. Econometric analysis found that educational attainment and relationships with friends who are already online are the strongest drivers of internet uptake.
Figure 2: The percentage of the population with an internet signal that fails to make use of data services (usage gap) has remained stubbornly high particularly in certain developing regions
Recent survey evidence from more than 20 low- and middle-income countries across Africa, Asia, and Latin America suggests that in many cases digital literacy is the single largest barrier to the uptake of broadband connectivity , with almost 70% of people overall indicating that they do not understand what the internet is or how to use it. While this has not typically been a focus of universal access policies in the past, some countries are beginning to take serious action to improve digital literacy. In Rwanda, the government trained 5,000 youth as Digital Ambassadors to go out into the rural areas with the aim of training 5 million people over a four-year period 2017-21.
After digital literacy, affordability is the second largest barrier to uptake cited by about 15% of those surveyed. Efforts are also underway to reduce the cost of purchasing a smartphone, which can absorb as much as 80% of the monthly budget of poor households. Ethiopia has formed a partnership with Chinese manufacturers for local assembly of smartphones to reduce their cost. More widely in Africa, MTN launched a low-cost smartphone costing just $20, while in India Jio had managed to drive costs as low as $9. Some countries, such as Costa Rica and Malaysia, are even allowing Universal Service Funds to be used to subsidize the purchase of smartphones for disadvantaged groups, as opposed to the more traditional practice of subsidizing network rollout by operators.
Even those who do use data are consuming too little to reap the full value of it. In low-income countries, average data consumption stands at just 0.2 gigabytes per month compared to just over 7 gigabytes per month in high-income countries (Figure 3). This has prompted extensive debate regarding how much monthly data consumption is “enough” to meet basic needs or mounting social expectations. Prior to the COVID-19 pandemic, A4AI had argued for a subsistence consumption threshold of 1 gigabyte per month, but this falls far short of what may be needed to support resilient remote working and schooling arrangements during crisis situations. Research undertaken for the WDR 2021, suggests that under normal circumstances, 0.6 gigabytes per month is adequate to support basic functions such as newsgathering, financial and commercial transactions, and accessing government services. However, adding even a modest amount of video streaming whether for education, employment, or entertainment rapidly takes the requirement up towards 7 gigabytes per month.
Figure 3: Inequities in mobile data consumption across country income groups and regions are huge
Competition among mobile operators remains one of the most powerful tools to promote access. Both Cambodia and India have witnessed a dramatic expansion of 4G penetration following the competitive entry of mobile network operators into the market. Intensifying competition triggered a huge drop in prices from initial levels of $4-5 per gigabyte per month to under $0.20 per gigabyte per month in both of these countries. More affordable prices led to a surge in demand for data services with consumption soaring from less than 1 gigabyte per month to around 7 gigabytes per month; comparable to levels in high-income countries.
Achieving equity through infrastructure
Going forward, global efforts on universal access—based on goals set by the UN Broadband Commission—will play a critical role in making the data economy more equitable by ensuring that all have affordable, reliable, and high-speed internet service. To achieve this vision, efforts will need to expand beyond the traditional focus on rollout (and continuous upgrading) of mobile networks into remote rural areas; important as that continues to be. Increasingly, policy-makers will need to pay closer attention to demand-side barriers, such as digital literacy and handset affordability, as well as ensuring that the whole system is underpinned by modern national data infrastructure to support the efficient exchange, storage, and processing of data.