But, what defines a neighborhood?
The way neighborhoods are defined have a profound impact on their residents. The shape and size of a single neighborhood has far-reaching impacts on how its residents are measured, and accordingly the respective allocation and management of public facilities within them.
One of the 3RF strategic pillars is “improving services and infrastructure.” To be able to look at services and infrastructure in a people-centered manner we must take a data driven reading of the way neighborhoods are clustered by way of their inhabitants' movement and interaction with the urban environment, and not only by way of maps.
Officially, Beirut comprises 10 areas at the lowest administration level. However, these neighborhoods are oftentimes disproportionately sized. Ras Beirut, for example, encompasses an area of just 2.5 km2, whereas areas such as Mazraa and Moussaytbeh are almost double in size. Yet, Ras Beirut holds three of Beirut’s major hospitals, whereas only one major hospital is located in Mazraa and Moussaytbeh each.
In view of adopting data-driven approaches to understanding Beirut’s neighborhood, we consider an alternative method of defining what a neighborhood is. In particular, we consider the effect of different neighborhood sizes with respect to Beirut’s population.
Fig 7.a Population data distributed within clusters based on the Beirut road network
Our analysis clustered Beirut’s road network and identified 30 unique neighborhoods within the city. These areas each have dense, interlinked road connections that, as a whole, can be considered a single neighborhood. Considering these new boundaries in line with Beirut’s population, our analysis shows two highly disparate population distributions across the city.
As aforementioned, population distribution across Beirut’s official census delineations shows a large skew due to large discrepancies between municipal area sizes. The size of these boundaries insofar does not seem to be linked to measurements of population density. In Beirut Central District, a population count of 102 is noted; whereas, Moussaytbeh and Mazraa hold a population of approximately 127,000 and 140,000, respectively (ref. Figure 7.a).
In contrast to this, population data distributed within clusters based on the Beirut road network sees a lower variance in population distribution. The smallest delineated area, Cluster 25, holds a population of around 1,700; with the largest area, Cluster 30, attributed to a population size of approximately 57,000. The clusters derived from Beirut’s road network suggest that a more even distribution of population within the municipality is possible.
Fig 7.b Current population distribution using official municipal area boundaries
These findings suggest the importance to explore how resources are allocated currently to each Beirut's 10 municipal areas, to ensure equity in the distribution of essential urban facilities and services within these municipal areas. Similarly, disaggregating Beirut’s boundaries allows a more nuanced view of where specific demographics are located. This may perhaps be a worthwhile endeavor to better understand the hyperlocal spatial distribution between vulnerable group and those with better economic, physical and social mobility particularly given the large number of residents already spread in these areas.
Fig 7.c Bar-graph displays a more even distribution of population reflected by Beirut road network based clusters
Fig 7.d A generative visualization which shows population data distributed within clusters based on the Beirut road network