Ebola, Poverty and Education : A data scientist’s view on a social disaster

As a data scientist, I am interested in understanding the different propagation models that are floating around on the internet. From the type of data used (travel, migration and more) to the assumptions underlying the different scenarios.
A conversation with my 8 years old daughter made me look at the disease from a different perspective. The propagation of the disease is due to a large extend on the fact that it is a manifestation of poverty and an attack on the fabric of the West African culture.
I blame poverty and access to clean water for making it near impossible for at-risk population to take the necessary steps to avoid spreading the disease. In a community where 3–5 people share the same room, where washing your hands multiple times a day with soap becomes a very expensive habit and the reflex of seeking medical help when sick turned into seeking medical help when dying due to what I call the fear of getting slapped with a prescription you can’t afford.
Poverty is the mother of all evils. It keeps people from enjoying nutritive food, clean water, which in turn reduces life expectancy which in turn leads to more poverty since parents don’t have enough productive time to plan their children ‘s future.
I blame poverty for the high illiteracy rate and for reducing the expected value of an education. When survival (finding food, clean water, decent housing and being healthy) becomes a challenge and threatens the existence of the very future education is supposed to improve, investing in acquiring new knowledge for future use does not make sense from an economic perspective. What I have heard through my social work is “Am I supposed to plan for a hypothetical future when my survival today is quasi uncertain?”. Education builds knowledge out of information. Knowledge leads to adequate actions and efficiency. This is the basis of data science. While the ebola prevention strategies have mainly revolved around providing information to the at-risk population, I am hypothesizing that the results are not as good as what we would have expected because of 2 major factors, one of them being poverty
In addition, the disease is plotting to destroy the very fabric of the affected region’s culture. A culture where community trumps individuality, where eating as a group, from the same plate is a social currency, where caring for the sick and burying the deceased heightens one’s social status. Slowing down the disease basically means reducing one’s social status. It means increasing the social and geographic distance between members of a community that is eager to remain tight. Slowing down the spread could be summarized in a few actions:
1. Not sharing clothes, meals, etc
2. Staying away from the sick and outsource their care to someone who is trained in infectious diseases handling – which in the social /cultural language means outsourcing the care to a stranger
3. Let someone else/ a stranger give the last rituals and bury the dead

Those 3 points, taken out of contest, would be frown upon by most people in the communities the ebola virus is destroying. This is not who they are. This is not what they do. But in light of this terrible disease, they have to become someone else. We have to become someone else. Or at least, as African, find a different way to be who we are. Because, right at this moment, the old way of being ourselves is killing us. Literally.
How can data science turn this around? How can information technology and knowledge creation make this better?
I strongly believe that reducing poverty and altering behavior would lead to exceptional results. The poverty aspect is to me quite clear. Changing behavior will be more challenging.
Every action we take has a social value. It might have no market value, but it has a social value. In order to keep one social status, we need to create opportunity for member of the community to replace a risky behavior that has a high social value with one (or a set of actions) of equal social values.
So here is the first set of tasks:
1. Listing all the behavior/actions that are putting a population at risk as well as the most common behaviors that are not deemed risky
2. Quantifying the social value of the actions – And I am looking toward the non-market valuation economists
Once we have the information, designing a platform where one can alter her/his behavior without losing their social status would help with future outbreak.
This disease is terrible mainly because it goes against who people are and how they were raised.



Leave a Reply

Your email address will not be published. Required fields are marked *