Data Analytics in social sector – Implementation and bottlenecks

The Social Sector is a vast system with complex interconnections within various stakeholders playing specific roles. It covers all that is essential for human survival and good life, and hence is always monitored to ensure quality of life. It is easy to visualise that there is immense amount of data available to study this sector but in practice the problem is complicated and a simple data collection and analysis procedure does not work well. We will explore in this blog the complexity of Social Sector Data Analytics and compare the pros against the cons.

Energy

Environment

Health

Infrastructure

Poverty

Water

Whenever there is a discussion on development or social sector, information from research becomes relevant in the building a narrative. The sector itself comprises of various areas essential for humans to survive and have a decent life. What this means is each area has to be looked at systematically, data must be collected methodically (like a researcher!) and subsequently this has to be analysed to understand its state, discover facts and implement changes. Data, is the key word here. There are various organisations involved passionately in developing data in this sector and let’s look at how analytics matter.

How is it done ?

Identify areas that require action.

Collection of data and formulation of database

Analysis of data collected to bring out right insights

Output from the analysis in the form of reports, policy related decisions, project planning and more

To give an idea, this is how a data collection process looks like

Let’s understand this with a practical example:

Few Use-Cases

If one has to visualise the changing trends in achieving SDG 6, track developments and filter relevant information to design effective policy, it has to begin with the collection of relevant information and data.

Urban Air Pollution and subsequent health impacts has become an important discussion in recent times. Let’s look at PM 2.5 Annual Mean Data for 3 countries (Source: https://ourworldindata.org/air-pollution) to understand this area and consequences for India. The following figure shows that India’s Mean Annual trendline since 1990 for PM 2.5 has always been above that of China’s and way above the WHO guidelines. In comparison Sweden has maintained its PM 2.5 levels well under the guideline. Of course, the development patterns, population requirements and historical growth affect current pollution but comparing the countries on their levels is not the idea.

The next figure shows Delhi’s monthly average PM 2.5 levels from January to November 2015 and this show how it isn’t just the annual average but the monthly averages are way above the guidelines. This is true for a lot of cities in India and you can get this data from https://data.gov.in/catalog/historical-daily-ambient-air-quality-data to assess the same. This instigates a discussion on enforcing regulations to meet health guidelines as well as a deeper investigation into the various causes of PM 2.5 pollution that impact air quality throughout the year.

One inference from this investigation was that vehicular pollution is a strong contributor and resulted in stricter guidelines for maintenance of diesel vehicles, or the former implementation of Odd-Even rule. It also forms the basis for the initiation of discussion and action for a policy maker.

Energy and state of rural women

This again is a topic that is discussed when addressing fuel access and availability in India. Much like water, electricity etc, fuel access is marred by social and cultural divisions of caste and gender and this can be understood through data. In the 2004 publication by World Bank (World Bank. 2004. The Impact of Energy on Women’s Lives in Rural India. Washington, DC. © World Bank. https://openknowledge.worldbank.org/handle/10986/20269 License: CC BY 3.0 IGO) the salient point of women spending a lot of time in gathering fuel and being exposed to high levels of indoor pollution is highlighted. This shows the lack of development to meet fuel needs through modern means, opportunity cost of women gathering fuel, and health hazards faced by them on a daily basis.

From the ESMAP Energy Survey of 1996, women spend an average of 0.85 hours collecting fuel daily. It is interesting to note that recent data on this topic is not readily available as it is not surveyed for periodically. In the same report, 2001 data on Rural women from Andhra Pradesh inhaled an average of 700μg/m³ of mixed particulate matter in 24-hour time span in the Kitchen area. Whereas EPA guidelines restrict indoor pollution to 50μg/m³. While this is a crucial issue and is data driven, data complications as well as interest in improving the said area impact the quality of all affected people.

Issues related to data compilation

While this article talks about how important Data Analytics is in the Social Sector and illustrates how Analytics is applied, there are various bottlenecks in getting it done in reality. While there are various organisations actively involved in bridging this gap, the gap exists nevertheless.

(Refer: https://ssir.org/articles/entry/code_switching_across_the_digital_data_divide)

Another issue about data in social sector is how it is not always available online or available freely. This data at times ends up having a price tag attached for interested researchers thus hampering access to data, further affecting the action that the research was originally going to implement. But given these minor and major issues in Data there are ways to go around the problem if some pre-requisites are met. In case of partial data, biased or imperfect data, using multiple sources, results, past reports and publications, and applying data triangulation might result in a conclusion better than what either of those sources alone would have led to. Still there is long way to go for Social Sector Data to be available and accessible. And in areas that it is, there is still lot to be done to effectively analyse and subsequently ensure changes.