Between 2010 and 2012 Cooperazione Internationale (COOPI), implemented a World Food Programme’s (WFP) initiative of cash transfers to 2,500 households in Mathare Valley slums, in Nairobi, Kenya. These cash transfer beneficiaries spread across all of the villages in Mathare Valley. Spatial Collective was approached by COOPI in 2012, at the end of the program, to advise on and help develop a secure monitoring and evaluation platform for visualizing longitudinal data-sets of these 2,500 beneficiaries of cash transfers.
During the two years of data collection COOPI created several databases on households in Mathare. Data included extensive longitudinal household information, such as, marital status, number of children, and even individual’s health details and material possessions. They also collected comprehensive behavioral information of these households, for example, how many years each person lived in Mathare, how the individuals moved around, how many times they went upcountry, how much money they spend for transport, etc. The database was enormous but there was one thing missing: the spatial information on 2,500 people included into survey. For this purpose Claudio Torres who was at that time working for COOPI approached Spatial Collective. He was interested whether our company could help COOPI collect spatial information on all of these beneficiaries (their households) and whether we could help them visualize and analyze some of this data.
As soon as the details of the project were established, Spatial Collective went on and designed the process of data collection and visualization. Our team trained COOPI data collectors on how to collect data using GPS units. At the same time data collection forms were created and teams were organized into groups to cover bigger area. COOPI’s only demand was that the GPS data collection moves parallel to COOPI’s household survey so that it has minimum impact on the main survey itself. After a day of learning basic GPS and mapping theory and a day of field training COOPI data collectors embarked on a GPS data collection of households of 2500 beneficiaries. The process of data collection went something like this: the beneficiary was called up over the mobile phone (COOPI had the database of all the individual’s telephone numbers), the meeting place was arranged, the beneficiary then took the data collectors to the household where they conducted the survey and collected spatial information (coordinate). Spatial data was later parsed with the two years worth of data and digital visualization was created under the leadership of Spatial Collective’s technical lead David Kutalek.
A rough look at the data displays how amazing and extensive COOPI’s database is. Looking at the data showcasing the years of residency we see that 45% of the people surveyed (1,897 out of 193,416 which represents approximately 1% of the population) lived in Mathare Valley 5 years or less; 69% of the people surveyed lived in Mathare 10 years or less. As we see from the data set, COOPI actually only managed to get in touch with approximately 1900 people (out of 2500); others moved out, some of them died. This data points to an extremely transient nature of the residents of the slum. The slum is considered to be a temporary step towards a better and more prosperous life. But, as we see from the data, some people do stay. More than 30% of the people surveyed have lived there for more than 10 years, some of them more than 30 or 40 years. Another very interesting indicator is that of tenancy which shows that 90% of the people surveyed are renters. Mathare lies on government land so this is not at all surprising. Also considering the sometimes very tense relationship between the landlords and renters in the area the number clearly points to the lack of security of tenure. These two indicators represent only a small portion of data displayed. As mentioned, the database is extremely detailed.
People often assume things about informal settlements. Slums are transient they say, most people are renters, they say. But to back up these assumptions with data is what really matters in order to help us understand informal settlements not only in Nairobi but broadly. COOPI has done a great job in shedding the light on some of the assumptions about Mathare Valley and Spatial Collective is extremely lucky and grateful to be a part of this process.
 Soft Kenya, Mathare Constituency, accessed March 3, 2013, http://softkenya.com/constituency/mathare-constituency/