by Quentin Dumont | 3 min read
In a way, the public enthusiasm over “big data” long revolved around an embarrassing paradox. Despite resounding promises that data analytics would eventually benefit us all, data scientists have (so far) mostly kept themselves busy with the click-through rate of Californian tech giants, or the ad targeting of North American department stores.
Source: Drew Conway
The good news is: this is starting to change. Over the past few years, a blossoming of initiatives have tried their very best to prove that data analytics can improve much more than credit rating and financial models. In many corners of the world, organizations and individuals willing to make a change are now using data to fight crime in their neighborhood, improve national health services or promote human rights worldwide. Some pioneers of data science for social good such as DataKind have helped nonprofits improve their operations while others chose to support local governments, notably revolutionizing building inspections in New York City.
So, how did this all start? Well, as with many trends, especially in the startup world, it is quite hard to know who called dibs. But it seems that national and local governments were instrumental in taking data science out of the glittering hedge funds, and into civil society. For local and national officials, the crucial role of data during the 2008 US political campaign was an epiphany. It showed to policy makers and data scientists alike that the unique analytical skills of data scientists are directly applicable to the (cluttered) world of public policy. This realization that data science was key in solving social problems was slow but potent. After all, governments have a long history of book-keeping and the increasing digitalization of their data bases has unleashed countless opportunities for data scientists to apply their skills to improve people’s lives.
Rayid Ghani’s career embodies the crucial role played by politics in dragging data science outside of the business sector. Trained as a data scientist, Ghani worked for a time in the Accenture Technology Lab before becoming the Chief Scientist in Obama’s 2012 re-election campaign. Realizing the potential of data science for civil society, he went on to found the Data Science for Social Good Fellowship at the University of Chicago. Every summer, the program gathers about forty students to solve social challenges using countless (and often messy) data sets. In 2014, the students tackled issues ranging from preventing the lead poisoning in children to ensuring efficient school budget allocations in Chicago.
Non-profits were also quick to realize that using data is key if they hope to be more efficient and stand out with innovative solutions. GiveDirectly, a non-profit that uses mobile technology to transfer money to the extremely poor in Africa, provides us in this respect with an inspiring illustration. Looking for ways to increase its impact, GiveDirectly partnered with DataKind to create a more cost-efficient process for targeting poverty-stricken areas in need of cash. Gathering volunteer data scientists, DataKind came up with an algorithm that automatically identifies villages in need of cash transfers by their type of roofing, thus improving GiveDirectly’s work at no additional cost.
However, if it is easier for governments, cities and universities to put in place the funds and networks necessary to attract data scientists, nonprofits are often left out. Not only are they usually unable to hand out a sufficient quantity of usable data, but more importantly the pay they can offer to socially-minded data scientists is a far cry from what the latter would get in the private sector for their short-supplied and incredibly lucrative skills.
But naturally, such a burgeoning trend can be hard to navigate and, for those interested in doing or using data science for good, we made sure to draft an article exploring the various ways data scientists and startups can join the dance. Enjoy!
Want to stay in the loop? Follow us on Twitter.
Quentin is a volunteer contributor to DataLook.