Can big data help us make better development decisions?
Data-driven decision making can be a powerful tool in the world of international development but it requires careful planning and management.
We are in an era of ‘information overload’. An overwhelming amount of data and information needs to be digested and processed to make a well-informed decision.
Thankfully, there are a significant number of big-data driven decision support systems embedded in many online applications to help us make evidence-based decisions. Better still, many of them have integrated advanced artificial intelligence that makes them capable of providing options and further advice based on users’ ‘patterns and preferences.’
A decision support system is basically a computer-based information system to support business and organizational decision-making by collecting, organizing and analyzing big data to help management, operations, and overall planning. This can be scary and beyond our comprehension, but we must embrace this new way of making informed decisions to help us manage our daily lives.
How do decision support systems work? Put simply, they integrate big data from different sources with smart algorithms that can quickly process millions of calculations and tons of information to come up with alternatives beyond the capability of our brain.
The options provided can be the result of comparison, optimization or contextualization that make them relevant. For example, if we buy online a ticket, a book, a video and the like, their online application systems can provide us with some alternatives and complementary products for us to choose from. In some cases, it can even provide detailed contextual information of the product such as in the housing market for a house with access to good schools and health, etc.
In the development world, how is this technology relevant in project selection and financing operations? In fact, the relevance is compelling. Resources are limited and becoming more so. While we face competing needs across sectors, we must allocate resources optimally.
Projects on road, energy, urban facilities, school and training centers must all be selected and built based on a set of objectives optimized for the welfare of beneficiaries. Using big data to support good decisions can also remove the rent seeking motivation’ and ‘elite capture behavior’ prevalent in many developing countries. It can therefore improve governance and resource allocation—killing two birds with one stone.
How this can be done? All available data and information should be integrated as part of the broader system with governing frameworks to set rules for how things work. This includes measuring and showing progress and making comparisons in many aspects across different sectors and regions.
A good example is the decision support system developed by ADB’s office in Pakistan. It is designed to help Pakistan develop economic corridors. All available data sets related to infrastructure, economic, social and other key variables of Pakistan have been integrated as parts of five different-but-interactive frameworks: sectoral mapping, district development monitoring, special economic zone assessment, economic corridor development navigation, and zonal valuation (which provides an economic valuation of each zone area or piece of land with a clear boundary).
The decision support system is built using a geographic information system interface that uses publicly available computer software. It is user friendly, interactive and dynamic to allow users to conduct various scenario analyses including their consequences. The analyses can be under one framework only or in a combination with other different frameworks.
Pakistan now needs to deal with 23 special economic zones proposed under different schemes. Seven have been notified by the government since 2015 as making no real progress; nine are proposed in partnership with the People’s Republic of China and seven are newly submitted.
The system can assess the economic feasibility of these zones based on the special economic zone development framework integrated in the system. This first assessment can be further strengthened by combining it with a sectoral mapping framework to see the economic feasibility in line with the sectoral development.
Furthermore, the second assessment can be further contextualized by integrating the district development framework to see how the results fit within the district’s development profile. The integration of the zonal valuation framework in the third round provides initial costing of the options and zonal perspectives from different aspects.
Finally, incorporating the economic corridor development navigation framework will add guidelines on prioritization and direction toward corridor development across different existing roads and geographic areas. Likewise, sectoral and district development assessments considering special economic zone development, economic corridor potential and zonal valuation aspects can be conducted.
Alternatively, one can also select any area via the system’s mapping capability to assess the potential of the area in the context of all five frameworks collectively or individually by stages. Results can be benchmarked against the national average or other selected benchmarks. In short, these analytical tools can provide the type of comprehensive assessment often lacking in developing economies.
The decision support system can be used by key institutions such as ministries of finance, planning, trade, as well as by universities and think tanks, for assessment and analyses on key development issues.
From an operational perspective for international development organizations, decision support systems can identify the best location and selection for developing new economic, social and urban projects, comparing different alternatives and considering the entire picture in all aspects. This is of course in addition to the standard procedure for selecting projects.
These systems can tell us which part of the new road will produce the biggest economic impact and drive economic connections. They can suggest the optimum location for a new facility to minimize travel times of potential users.
The list goes on. The only limit on big data-supported decisions is our imagination and commitment.