When the UN set the Sustainable Development Goals (SDGs) in 2015, the pledge to “leave no one behind” became a central goal in all aspects of development. Yet, we know that progress cannot happen without the data to know how, why and what to develop. Effective action requires a precise understanding of the target population in distinct geographies. That is why accurate and inclusive data which can measure several aspects of development in a country is crucial to achieve this promise. COVID-19 has aggravated these concerns and increased the importance of integrating multiple data sources. As such, big data from mobile network operators, or mobile positioning data (MPD) can make a significant contribution to the SDGs.
SDGs provide a blueprint to address key global challenges, aiming to achieve a better and more sustainable future for all. However, reaching the targets set in each goal depends heavily on our ability to monitor the progress universally for all populations. For many SDG indicators, data is readily available but major data gaps remain in terms of accessibility, coverage, timeliness, and granularity that cover income, sex, geographic location, and other characteristics. Yet, there is a sense of urgency for finding an innovative way to track our progress – we have fewer than 10 years left to achieve the goals.
Big data, big impact
Mobile location data or mobile positioning data (MPD), gathered by processing calls, internet connections and signaling records, offers a solution for filling data gaps. As mobile phones tend to be carried around in our pockets, MPD has great potential to provide information on human mobility and a variety of population-based characteristics.
In the past decade, MPD-based solutions have been developed in various domains such as tourism statistics, population statistics, and mobility. Here are some key aspects in which MPD can impact monitoring the progress of SDGs:
- Sampling coverage – More than five billion unique subscribers worldwide, and more than seven billion people live within the coverage of mobile networks. This ensures that MPD has one of the most extensive sampling coverage among other data sources.
- Timeliness – MPD can be generated in near real-time and offers a better sampling time period, allowing robust historical assessments of a phenomenon over time.
- Data quality – The digital nature of MPD also means that a certain degree of quality assurance can be built into data collection and analysis.
- Granularity – MPD also offers a better degree of data granularity in terms of the time period (daily/monthly/yearly) and geographical location (urban/rural). Moreover, when combined with other reference datasets, MPD could offer other disaggregated data. For example, based on gender and income level.
How MPD will leave no one behind?
These key advantages make MPD an attractive solution compared to other big data sources for monitoring the SDGs. In fact, at least 23 indicators can be directly or indirectly measured using MPD. Positium, a private company specialized in MPD-based official statistics, is constantly thinking of the ways MPD can get us closer to achieving the Sustainable Development Goals. Within the global community, we each have something to contribute to achieving the targets. Positium’s role is to help provide reliable statistics from MPD that lead to improved decision-making.
In collaboration with the International Telecommunication Union (ITU), Positium has demonstrated the value of MPD in developing a standardized methodology and process to produce two SDG indicators (link to case study): 17.8.1 (Proportion of individuals using the Internet) and 9.c.1 (Proportion of population covered by a mobile network, by technology). The methodology was then applied by the national statistical offices (NSOs) in Indonesia and Brazil, and the output of the MPD analysis was compared to the results of an ICT household survey.
Note that the survey could provide the data on the state and city level, while MPD supplied the necessary information on internet access and population coverage at the district level.
Map of the proportion of individuals using the internet via mobile phone in Rio de Janeiro, Brazil. Source: IBGE
This project not only provided experience in applying mobile big data analytics but also contributed to building the capacity of NSOs to produce results for SDG indicators based on alternative data sources. The analysis conducted by the Brazilian Institute of Geography and Statistics (IBGE) shows that the MPD-based results were very close to those of the household survey. For the total area of study, a insignificant difference of 1.04 percentage points (pp) in the Rio de Janeiro metropolitan area and 0.70 pp in the city of Rio de Janeiro was found. The results suggest the method is robust and can be used to produce official statistics. Now, a protocol has been made for the future incorporation of MPD in Brazil’s statistical production pipeline.
|Comparison of MPD and survey results. Source: IBGE||MPD||Survey data||Difference|
|Total study area||93.31||93.89||0.02 pp|
|Rio de Janeiro metropolitan area||95.04||94.01||1.04 pp|
|City of Rio de Janeiro||94.87||95.57||-0.7 pp|
In addition to the geographical area level, MPD can also provide statistics based on different types of technology (2G/3G/4G) accessed by the users. Are you surprised that in some areas of Rio de Janeiro, 2G was still the most common technology used to access the internet? Moreover, the digital nature of MPD means it is generated constantly and is therefore very well suited for monitoring SDG targets over time. MPD can also shed light on social disparities that were previously hidden. Nevertheless, it has been noted that despite its potential, the use of MPD has remained largely in the pilot phase – a situation common to many cutting-edge fields.
Innovative data sources like data from mobile networks can offer a cost-efficient complement to traditional data collection methods and provide a fresh, more precise perspective. Mobile network data can provide the key to informed decision-making around the world, regardless of region.