Filling a vacuum: Efforts to map and enumerate Lilongwe


My latest for looks at how civil society actors in Lilongwe are generating their own maps and data to understand urban poverty due to a gap in official statistics. Initiatives include participatory community mapping efforts as well as an open geospatial database. But more could be done.

You can read the article here, and join the discussion on visualising poverty around the Global South here.


Lilongwe: Fear in the dark.

This post features edited excerpts from my recent report Survey of Urban Poor Settlements in Lilongwe, produced by ActionAid Malawi and the Lilongwe Urban Poor People’s Network.

Residents don't trust each other in the dense settlement of Mwenyekondo

Residents don’t trust each other in the dense settlement of Mwenyekondo

To get a sense of the level of social capital in the 33 settlements surveyed, we asked respondents about feelings of safety and trust. As the below data shows, while levels of trust are relatively high, the sense of safety in most settlements varies considerably between day- and night-time. In essence, come sundown, you’d better be home.

Respondents were asked to rate how safe they feel in the settlement as well as the extent to which residents trust each other to assess the level of social capital in the settlements. As the below chart indicates, levels of trust are high in the majority (76%) of the settlements – only five settlements indicated low levels of trust. Four out of the five settlements with low trust levels between residents, Mwenyekondo, Mtsiriza, Mtandire, and Tumbwe, are characterized as relatively dense with high a percentage of renters, which may partly account for the lack of trust.

trust level

The sense of safety in the settlements surveyed is starkly divided between day- and night-time. While respondents in most settlements stated they felt safe during the day, they also indicated that they felt unsafe during the night. As a result, the sense of safety in 58% of the settlements is moderate. It is of note however that the sense of safety is low in almost a quarter of the settlements surveyed; 58% of settlements also cite crime as a problem.


Eighty-five percent of settlements indicated that they have good relationships with neighbouring settlements. Many respondents highlighted funerals as common occasions when residents in neighbouring settlements come together to support each other.

On age and land management in Lilongwe’s poor settlements.

This post features edited excerpts from my recent report Survey of Urban Poor Settlements in Lilongwe, produced by ActionAid Malawi and the Lilongwe Urban Poor People’s Network.

Home in Chifumbe, one of the older settlements in Lilongwe

Home in Chifumbe, one of the older settlements in Lilongwe

Since when have you lived here? is not always an easy question. At least, that’s what we found when trying to get a sense of when the 33 poor settlements we surveyed were first established.

A number of reasons make it challenging to collect data on the age of the settlements. Some of the settlements have been in existence for a long period time, meaning current residents are uncertain of the exact date of settlement. Other settlements have experienced regular turnover of residents, diminishing communal memory.

In the below chart, settlements have been grouped in three main categories, which give a rough reflection on the ages of the settlements surveyed. Of note is that the majority (52%) of settlements were established in the post-1950 era, many as a result of evictions in the 1970s as Lilongwe became the capital of Malawi. A third (11) of the settlements surveyed are over a century old.

age of settlements

Data analysis shows that the age of a settlement increases the likelihood that land in the settlement is managed in a customary fashion (as opposed to by the City Council): having settled before 1900 compared to settling after 1951 increases the likelihood of a settlement having customary land management approximately 13.5 times. This suggests that the age of a settlement could be seen as a proxy for the strength of traditional authority, making older settlements more resistant to land ownership claims by the City Council.

On the topic of land management, respondents were asked who the land the settlement is located on belongs to. Responses were largely divided between land belonging to the City Council (55%) and land being held customarily (42%). Only one settlement (Mgona) indicated differently, stating that the land the settlement is located on is railway land.

Railway tracks going through Mgona settlement in Lilongwe

Railway tracks going through Mgona settlement in Lilongwe

It should be highlighted that the responses are based on perception of land ownership; land in some of the settlements who indicated that land is held in a customary fashion may formally belong to the City Council. As such, it is useful to consider how land in the settlements is actually managed, specifically, whether or not Chiefs play a role in allocating land.

Just over half (52%) of the settlements indicated that local Chiefs either allocate or sell land in the settlement. Of the settlements in which Chiefs play a role in land management, ten (59%) fall into the Rural category, six (35%) into the Transitional-2 category and only one falls into the Transitional-1 category. The one Rural settlement that did not indicate that Chiefs sell or allocate land was Kasengere, where respondents stated that no land remained for sale or allocation by Chiefs.

Chiefs, by contrast, do not play a direct land management role in any of the Urban settlements and 80% of Transitional-1 settlements. Chiefs do also not play a land management role in 45% of Transitional-2 settlements.

Nevertheless, Chiefs retain authority in all settlement categories: 90% of settlements indicated that Chiefs ‘always’ witness land transactions. The reason Chiefs do not witness land transactions in three settlements (Chifumbe, M’bwetu, Federation) is because land is not sold in these.

Locating the poor in Lilongwe.

This post features edited excerpts from my recent report Survey of Urban Poor Settlements in Lilongwe, produced by ActionAid Malawi and the Lilongwe Urban Poor People’s Network.

Makatani settlement, categorised as Transitional-2

Makatani settlement, categorised as Transitional-2

In country where most people give directions based on landmarks, it can be difficult for an outsider to get a sense of what’s where. When it comes to poor settlements in Lilongwe, most people know they exist near so-and-so, but even people working in the settlements are not always clear which administrative area a particular settlement is located in. This of course says a lot about the relevance of (imposed) administrative boundaries for locals, as well as differences in how space is viewed and experienced by locals and outsiders (such as myself). In any case, part of what we wanted to do through our recent research project was to geo-locate the settlements we surveyed, and use that data in our analysis. Here’s what we found:

Poor settlements exist across Lilongwe City. However, as the below map shows, while most of the areas of Lilongwe are home to only one or two poor settlements, some areas stand out as containing more. Areas 25 and 49 are both home to five of the settlements surveyed, while Areas 50 and 55 contain three each. This is of note given that all four Areas, containing a total of 16 settlements (48% of total), are located in the northern part of Lilongwe and border each other.

The congregation of poor settlements in the northern part of the city is reflected in the distribution of settlements across Traditional Authorities (T/A) in Lilongwe. The clear majority (64%) of settlements surveyed is located in T/A Chitukula, which spans Areas 10, 25, 39, 41, 44, 49, 50, 51, and 55. 24% of settlements are located in T/A Tsabango, while only 2% of settlements are located in T/A Malili and T/A Njewa respectively.

There are however notable differences in living standards across the settlements surveyed. While all can be described as ‘poor’, there are variations in access to basic public services as well as land management practices between the settlements. To give an overall sense of the type of settlements surveyed, each settlement was scored on ten characteristics, and categorised as either Urban, Transitional-1, Transitional-2, or Rural. The categories were defined as follows:

Urban – Settlements are planned and residents pay City Rates. Land is managed by the LCC and Chiefs do not allocate land. House types are mixed or permanent, and the majority has access the LWB water, ESCOM electricity, access by tarmac road, and LCC waste collection.

Transitional-1 – Land is managed by the LCC and residents pay City Rates. All settlements are either fully or partly planned. All have access to LWB water and ESCOM electricity (one does not have electricity access).

– Settlements are unplanned and residents do not pay City Rates. House types are mixed, and some residents have access to tenure security documents. All have access to LWB water and ESCOM electricity.

Rural – Unplanned settlements where Chiefs play a role in land allocation and residents do not have secure tenure nor do they pay City Rates. Very limited access to public services.

The below pie chart shows the frequency of the resulting groups:

settlement typology

The chart indicates that two-thirds of the settlements fall into the Rural or the Transitional-2 category. As the map below shows, Transitional-2 settlements are located throughout the city. Settlements categorized as Rural on the other hand are located in the northern parts of the city; seven are located in adjacent Areas 25, 49, 50, and 55, while three are located in nearby each other in Areas 39, 41, and 44. Only one, Sese, is located in the southern part of the city.

Interestingly, ten of the eleven settlements categorized as Rural are located in T/A Chitukula, implying that there is not only a significant amount of poor settlements in the area, but also that those poor settlements are some of the worst off among those surveyed. Indeed, 71% of settlements in T/A Chitukula fall into categories Rural and Transitional-2, indicating a concentration of poverty in the area.

Only five settlements fall into the Transitional-1 category. Three of these are located close to each other in rapidly gentrifying Area 49. Similarly, only six settlements are classified as Urban. Apart from Mchesi, these were all established after Lilongwe became the capital of Malawi, i.e. post-1970. They are all also located in the vicinity of trading centres, which may account for their better than average access to public services as well as the prevalent formal land management structures.

A house in Sector 7, a settlement categorised as Urban

A house in Sector 7, a settlement categorised as Urban

urban_net #5: urban planning and managing access to water

nov meet

The fifth urban_net meeting took place at the ActionAid Malawi office this morning, featuring two presentations. First up was Asayire Kapira from WES Network, speaking about Water Users Associations (WUAs) in Lilongwe. WUAs are cooperative societies where communities establish a legal business entity and register it with the government to operate all water facilities – usually water kiosks – in a designated area. Through its Tilitonse Project, WES Network works with a number of WUAs mainly in Lilongwe’s peri-urban areas, seeking to improve accountability of the WUA model through participatory tools and approaches. While the project has met successes, including establishment of a WUA network and de-politicisation of existing associations, there have also been some challenges. A key challenge has been lack of responsiveness and even resistance by duty bearers, namely the Lilongwe Water Board, to be accountable to their customers. You can download Kapira’s presentation here.

The second presenter was John Chome of UN-Habitat in Malawi, though speaking in his private capacity. Chome delivered a thought-provoking historical presentation on urban planning in Malawi, and the failures of both past and current planning policy to address the challenges of urbanisation (and take advantage of the opportunities). He concluded that urbanisation in Malawi now occurs “outside of planning” and challenged the participants to deliberate on whether current planning in Malawi is addressing the real issues facing the country, and if the appropriate models to address these issues are being used. This lead to a lively discussion on the anti-urban bias at the government level, the need for Lilongwe City Council to take on a leadership role when it comes to city development, and how Lilongwe as a city needs to densify and grow upward, not continue to spread and sprawl. You can access Chome’s presentation here.

The next urban_net meeting will take place on Dec. 4. Join us!

A standard for cities?

The International Organisation for Standardisation (ISO) has issued an international standard for cities. Entitled ISO 37120:2014 Sustainable development of communities — Indicators for city services and quality of life, it features 46 performance indicators, mainly on access to public services. Cities are expected to track these, although compliance is not compulsory.

According to proponents, such as Neal Peirce of Citiscope, the implications of the “city ISO” are dramatic. Having set indicators will allow for cities to be compared based on the same objective indictors, “[a]nd just as significant, the people of cities — civic, business organizations, ordinary citizens — will be able to access the same new global standards. This means they can ask city leaders tough questions, stoking debate about their own city’s performance on the basis of verified measures ranging from education to public safety to water and sanitation.”

Sure, better data and strengthened accountability. Sounds good. But what is an international standard?

According to ISO, “[a] standard is a document that provides requirements, specifications, guidelines or characteristics that can be used consistently to ensure that materials, products, processes and services are fit for their purpose.”

So if a city doesn’t rank well on the ISO standards it’s not fit for purpose? Sounds rather odd. I’m also curious about how developing country cities are expected to provide reliable data such as “Assessed value of commercial and industrial properties as a percentage of total assessed value of all properties”, and how much all this data gathering might cost them.

Thankfully there’s a chance I might get some answers: Meeting of the Minds, Citiscope, and the World Council on City Data are hosting a webinar on the topic on Nov. 11. You can register for it here.