Tag Archives: ICT4D

They’re Talking. Are We Listening?

Development agencies and programs around the world are staffed with skilled Monitoring & Evaluation professionals, responsible for tracking a program’s progress against its goals through quantifiable results – think back to the days of creating SMART (specific, measurable, actionable, time-based) goals for your professional evaluations or science class hypotheses. The program I work on, funded by the U.S. Agency for International Development, measures our progress against twelve indicators, which include standard measurements such as number of people trained, number of transactions recorded to track tuna catch, and millions of dollars leveraged from the private sector. While all worthwhile measurements that seek to quantify the program’s impact, does this really capture the full program’s full accomplishments (or failures) in way that reflect beneficiaries personal gains and experiences?

With developments in technology, data collection, and increased use of social media in everyday life, the amount of data is ever increasing, along with professions to analyze and assess this data. How can development tap into this potential?

Development agencies and non-profits are increasingly pushing out information using social media. But, should we be spending more time listening?

In response to America’s current political landscape and reduced funding of government agencies, a recent article in the Boston Business Journal poses how urban social listening can help to determine the impact of aid and social programs. In the study, researchers, “…began by listening to the ‘digital crumbs’ generated by collective searches and postings to social media like Twitter.” Urban social listening offers a, “systematic, rigorous collection and analysis of [social media] ‘crumbs…’ offering useful insights into understanding the government’s role in addressing urban problems” (Hollander and Renski, The Boston Business Journal).

It has been shown that there is an evidenced correlation between positive sentiment and health indicators. In a 2015 study, Twitter sentiment was shown to be a better predictor of cardiovascular disease than any conventional indicator like smoking or socioeconomics. Hollander, author of the Journal article, applies urban listening techniques to U.S. Department of Housing and Urban Development’s Community Development Block Grants (CDBG). While he wasn’t able to find significant improved sentiment in cities with CDBG funding, it seems that the community scale being analyzed is too large to yield highly correlated results.

In development, non-profits and government agencies alike are increasingly using social media to communicate with stakeholders and beneficiaries, but listening is a newer trend. Payal Arora and Nimmi Rangaswamy echo this in Digital Leisure for Development: Reframing New Media Practice in the Global South. “…Research is driven by development agendas with a strong historical bias towards towards the socio-economic focus. Data that does not directly address project-based outcomes is side-lined.”

Tracking sentiment through social media is not a new concept, used for years in consumer confidence and political opinion. While certainly more people are tweeting on popular political opinion topics, how can we the development world harness and tailor these listening practices to hear our beneficiaries, and through listening show that their voice counts.

Arora, Payal and Nimmi Rangaswamy. Digital Leisure for Development: Reframing New Media Practice in the Global South. Sage Journals. October 10, 2013.

Hollander, Justin. Viewpoint: As Ben Carson Hearings Get Under Way, Can Social Media Help Identify Hud Impact? The Boston Business Journal, 2 March 2017. http://www.bizjournals.com/boston/news/2017/03/02/viewpoint-as-ben-carson-hearings-get-under-way-can.html

Eichstaedt, Johannes, et al. Psychological Language on Twitter Predicts County-Level Heart Disease Mortality. Sage Journals. Volume: 26 issue. Pages 159-169. Published January 20, 2015. http://journals.sagepub.com/doi/abs/10.1177/0956797614557867

Data show an increase in hate crimes

When, now President, Donald J. Trump was elected in the United States’ elections in November 8, 2016, it ‘whipped’ up a storm of emotions, reactions, and actions. Some were positive, some were negative, and some outright worrisome. In my previous posts I have written about social media engagement through #resist and how social media was used to organize activists and supporters for the Women’s March. Something that I have yet to cover is that many hate crimes and hate speeches have emerged as well.

After the US elections in 2016, Ushahidi, a crowdsourcing tool, converted their USA Election Monitoring platform to start monitoring post election hate speech, harassment, violence, threats, and protests. By November 18, 2016, the crowdsourcing data collection had received over 800 reports (though some were duplicates). Ushahidi “was developed to map reports of violence in Kenya after the post-election violence in 2008.” The tech organization, headquartered in Nairobi, has since then been used by thousands to raise their voice.

Another, US based organization, Southern Poverty Law Center’s (SPLC) project Hatewatch in collaboration with ProPublica had by February 7, 2017, registered 1,372 post-election bias incidents. The data was collected through their #ReportHate intake page and from news reports. SPLC has partnered with ProPublica in order to better document, verify, and investigate these incidents. The New York police reported that between 2015 to 2016 hate crimes had increased by 31.5%, up from 250 to 328. Hate crimes targeting Muslims are up from 12  to 25, and hate crimes targeting Jews are up from 102 to 111. Overall, according to SPLC’s data, anti-immigrant incidents remain the most reported. Below is a chart of the highest motivational factors for the hate incidents:

Snapshot of data of Southern Poverty Law Center and ProPublica data collection of Hate Incident Motivation post-election in the United States, 2016: https://www.splcenter.org/hatewatch/2017/02/10/post-election-bias-incidents-1372-new-collaboration-propublica   

The authors Read, Taithe, and Mac Ginty examines technological innovation, primarily digital technology, and the promise it shows for improving humanitarian responses (Read, Taithe, & Mac 2016: 7f). The data technologies used by the three forementioned organizations create awareness, and it could be argued that they are empowering. There have been multiple cases of where crowdsourcing of data on violence which have had somewhat different outcomes. For example, the Libya Crisis Map, has involved the coordination and vertical transmission of knowledge of urgent situations to national and international actors and audiences. In another example, in Kenya, the knowledge took a horizontal form, used by locals to alert each other of unfolding situations (Read, Taithe, & Mac 2016: 11).

So how does it work? Ushahidi collects data by using custom surveys as well as crowdsourcing tools. The organization’s tools are not only used for monitoring elections, but also for crisis response, as well as advocacy and human rights. The sources stem from surveys, and third party emails,Twitter, Twillio, Nexmo, and SMS connected with an SMS gateway or SMSsync, amongst others. If subscribed, organizations and users can receive important alerts and notifications while deployed. What is interesting with Ushahidi is that it is not an organization from the global North and thus challenging not only the norm of what a humanitarian tech organization looks like and where it comes from, but that they also become a stakeholder in the power of knowledge (Read, Taithe, & Mac 2016: 12f). Below, is an example from Ushahidi’s site: a Muslim NYC Transit worker who was pushed down flights of stairs and called terrorist.

Snapshot of Ushahidi’s platform to monitor USA post election hate crimes. https://usaelectionmonitor.ushahidi.io/views/map

However, Read, Taithe, and Mac suggests that while collecting data by and for domestic and international development may have several advantages, it also comes with responsibilities and risks. The authors write that the evidence thus far suggests that the information gathering capabilities of some humanitarian actors outstrip their capacity to deal with the information.

Ultimately we conclude that the new aspiration towards hubristic big data processing is just another step in the same modernist process of the production of statistical truth. Where it holds a particularly seductive power is in the promise that it may, somehow, become autonomous of human intervention, magnified in legitimacy and relevance by the new processing technologies (Read, Taithe, & Mac 2016: 13).

The unfolding situation in the United States hits home as my father is Muslim, and my father’s side of the family are too, yet the minority Muslim community is just one of many targeted minority groups in the country; the communities of African-Americans, LGBTQs, refugees, immigrants, and women all over the United States are targeted too. This hits home because I have Muslim friends, immigrant friends, African-American friends, LGBTQ friends, and jewish friends, many of whom are women, myself included. This hits home because the misconceptions, stereotypes, and convictions, of these minority groups are so strong that it will lead to violence, threats, and harassment.

References – imbedded in the text and:

Read, R., Taithe, B., and Mac Ginty, R. 2016: Data hubris? Humanitarian information systems and the mirage of technology, Third World Quarterly, forthcoming.

Big Data Catches: Furthering Development, not the Divide

Women in the port of Bitung, Indonesia prepare the day’s catch for market. 

The global fishing industry is a multi-billion dollar industry, with a recent report released by Stratistics MRC valuing it at $239.8 billion in 2015 — and projecting it to reach $320 billion by 2022. Southeast Asia is one of the largest contributors to this market, accounting for more than 50% of the world catch and with the largest concentration of fishing vessels in the world– over 3 million. More than just an economic powerhouse and one of the globe’s biggest sources of protein, Southeast Asia’s fisheries employ 93% of all fishery and aquaculture employees worldwide, and 10% of the world’s total working population.

The industry, while growing and a profitable employer of hundreds of millions, is also a large developmental concern. Its exponential growth and increased demands that it enjoys negatively impact the health of valuable ecosystems, marine biodiversity, and threaten already dwindling fish stocks. In addition to the environmental impacts, the region is rife with illegal, unreported, and unregulated (IUU) fishing – an issue that is gaining traction in development both because of its environmental impacts, but also because of its serious human welfare implications.

“Data has become increasingly important to the way we think and talk about conflict and our humanitarian responses to it” (Róisín Read, Bertrand Taithe and Roger Mac Ginty). In a world where technology is often looked to to address human problems, how do these issues relate to data and development? A number of non-governmental, governmental, private sector, and industry players are increasingly looking to data to gather information on an industry that, despite its wealth, is incredibly challenged by tracking and sourcing its products, far behind other food sector’s traceability capabilities.

With new U.S. regulations that will require a set of standard data to be submitted with every seafood import to enhance transparency (released in December 2016 and going into effect in January 2018) also serving as a catalyst, fisherfolk and industry must quickly get aboard supply chain data collection efforts, and figure out how to implement these systems in their operations. Data collection is challenged by limited connectivity at sea, hesitancy to invest, unwillingness to share proprietary data (such as location of catch), and complex supply chains.

So, how can development organizations address this challenge to increase sustainability, protect finite marine resources, and address serious human welfare concerns that include limited labor rights, forced labor, and slavery? While more data can equal more intelligent decision making, transparency, and the development of effective policy and regulation, it can also leave beneficiaries behind – often the most critical beneficiaries, those that have sole-source incomes, limited access to technology, and depend on the sector for their livelihood.

Not only do large scale commercial operations need to be addressed, but small-scale fishers – those that often catch only a few kilograms of fish per day and is their sole source of income. How can large scale operators and artisanal fishers, living in small coastal fishing villages comply with regulations and uniformly collect data in a way that doesn’t push small, independent fishers out of the supply chain? With the proliferation of big data, how can development take care to further development goals with technology, and not further segment society into groups that are more or less likely to adopt data technology? “…The way information technology [has] operated in the sector [is] equivalent to ‘buying a state of the art car, driving it into the desert and leaving it there’. More and more money is invested in developing these technologies but their use is often limited, driven not by a clear sense of what is needed to improve response, but by what the advances in technology enable.”



Stratistics MRC. Commercial Fishing Industry – Global Market Outlook (2016-2022). January 2017. http://www.satprnews.com/2017/03/03/commercial-fishing-industry-market-size-share-analysis-report-and-forecast-to-2022/

Global Implications of Illegal, Unreported, and Unregulated (IUU) Fishing. National Intelligence Council. 19 September 2016. http://www.iuufishing.noaa.gov/RecommendationsandActions/RECOMMENDATION1415/FinalRuleTraceability.aspx
Third World Quarterly, 2016  Data hubris? Humanitarian information systems and the mirage of technology Róisín Read, Bertrand Taithe and Roger Mac Ginty Humanitarian and Conflict Response Institute, University of Manchester, UK. http://dx.doi.org/10.1080/01436597.2015.1136208