Some weeks ago, I received the picture at the top of this post. Apart from triggering a light smile, since at the time I was preparing this blog series, it made me think about data’s development impact. Assessing and evaluating impact is necessarily contingent to the purpose and goals of the development intervention, at least if we refer to positive impact. If the expected impact of a blog post is measured by the number of users viewing the post, then the fabricated picture above is definitively right, at least for the time being. However, if the expected impact is not measured by users’ views but other measurements, i.e: citations or share frequency, the conclusion about impact might be very different. Thus, analysing the impact of data in development will necessarily require agreeing on the purpose and goals as well as in the definition of impact.
A good departure point is then to define impact. According to the OECD DAC, impacts are the positive and negative, primary and secondary long-term effects produced by a development intervention, directly or indirectly, intended or unintended. If in addition, we take into account that impact is also one of the five criteria proposed by the OECD DAC for development evaluation, the definition of the criteria provides some more guidance about how to evaluate impact. In this regard, the OECD DAC recommends looking into the effects on the local social, environmental, economic and other development indicators, as well as the positive or negative impacts of other factors.
A few years ago, Stephen Spratt from the Institute of Development Studies at Sussex University, and Justin Baker from the University of Texas, released a paper about the subject: Big data and international development: impacts, scenarios and policy options. In this paper, they review the difficulties in finding a common definition of big data, since it is dependent on which industry or field we approach it from. They choose to define it as ’referring to our growing ability to generate, manage, analyse and synthesise data to create and destroy different forms of value’. Having a definition provides us with a better understanding of the purpose and goals: to create and destroy different forms of value. They also propose a framework for analysing big data’s impact on development. The framework focuses on looking at effects on four categories: economic, human development, rights, and environment, which are in line with the recommendations of the OECD DAC. The result of their analysis for the so called development countries is summarised in the table below, with negative effects coloured in red and positive effects in green. One could argue that just a few years later the table could be updated in the light of the information that The Guardian and the UN Rapporteur on extreme poverty released last week, although for the purpose of this post that might not be necessary.
Just a quick look to the table helps us understanding that the potential impact of big data in development is likely to have both, positive and negative effects. More importantly perhaps is the fact that it helps us understand the extent to which these effects will realise is contingent to other factors. Factors that represent the existence of certain preconditions or follow up actions in terms of regulations, infrastructure, mechanisms or behaviours. Thus, the realisation of positive and negative impacts of big data in development is subject to the context , as well as other factors, where big data is being applied. This reinforces the idea, presented earlier in this blog series, about the necessity to recognise the heterogeneity of the Global South.
The idea of contingent factors is an important insight as confirmed by Jorge Umaña in his analysis of the State of the art of #BigData in Costa Rica to measure the Sustainable Development Goals, for the Big Data for Development Network in Latin America. Umaña identifies the existing potential in the use of big data for monitoring the SDGs. However, he also identifies the need for the so called contingent factors for this to be realised. Taking into account the role of the Costa Rican state institutions in the organisational structure of SDG monitoring, as well as the existing regulatory framework, Umaña identifies two key factors that condition the potential of big data: the availability of data, both in terms of data points as well as longitudinal series, and the need for time and investment for Costa Rica to develop the capacity required for the management and application of this data. The Costa Rican public sector started investing in developing the technical capacities on big data of its work force in 2015, three years down the road this efforts have not had the time to mature to the extent that is necessary for actually applying big data for monitoring the SDGs.
Enthusiast of big data in development have proclaimed in all directions its benefits on the basis of expected positive impacts, however the critical necessity achieving a number of contingent factors as a preconditions for its success is absent from their narrative. Absent from their narrative is also the fact that there are a number of negative impacts which can be expected from the application of big data and that need to be paid attention to and dealt with before and under the application of big data to development. As with the picture heading this post, just because messages and ideas are mediated and re-mediated, it does not mean these are true, this will be dependent on the perspective of analysis adopted.
The analysis of big data impact in development helps also reinforcing some of the warning flags being waived by scholars and development practitioners in the Global South, which this blog series has attempted to bring attention to. First, the current dominant narratives of data in development reflect the dominance of the views and concerns of the Global North and have traditionally excluded the voices of the Global South. Secondly, the impact of big data in development is contingent to the context where it is being applied, thus the heterogeneity of the Global South needs to be recognised and universalist narratives abandoned. Thirdly, given the potential negative impacts of big data for the population, it is necessary to develop a framework for the governance of big data that pays attention to aspects of ownership, access and application. Finally, harvesting positive impacts of the application of big data to development in the Global South will require, on top of what already has been mentioned, time and investments so that countries have the necessary capacities to use big data and its products, and mitigate its potential negative impacts.
Aware of the multitude of questions which I have not touched upon in this series, but having covered some of the critical aspects of the application of data and data-enabled processes to development from an attempted Global South perspective, this post marks the closure of this series.