Improper solid waste management is one of the global development agendas dealt with through different perspectives. Waste disposal is a critical issue for the environment and public health. There are several initiatives to turn this threat into wealth through responsible consumption and production framework of sustainable development.

Solid waste management is one of the key services a city requires, and this is more so with the rapid growth of Addis Ababa. Residential solid waste constitutes a sizable share of this municipal waste, and therefore we need to dispose of it properly. According to Diriba & Xiang-Zhou (2021), households contribute 76% of municipal waste. For urban waste, residential areas contribute considerably and having behavioural and practical issues addressed at the household level could make a difference in waste disposal impacts. Moreover, instead of blanket approaches across the city, pragmatic decisions based on the existing data could make the city’s administration prioritize issues to address. This approach will also economize resources through an improved use efficiency while tackling solid waste-related challenges.

Our exploration in Adama and Bishoftu (Oromia regional towns) brought us research questions that we may need further study in specific cases. In our exploration, we have understood that the demographic status of residents could make a difference in solid waste management practices at the household level. Municipality experts reported the comparably low economic class residents solid waste segregation practices over the other segments of the community. Such economic class-based differences in waste management practices within the same town caught our attention to do an in-depth study in the capital city on geographically contrasted locations. Contrasting locations are also proxy indicators of economic classes through residential differentiations. Surveyed locations are indicated in the following map (Figure 1).

(1) AU area is a neighborhood located in the African union head quarter, which is characterized by congested slum, with poor basic facilities (2) CMC area is situated in a transitional zone, between the central business district and the peripheral location, which is characterized by low rise and in most cases better-off housings (3) Yeka Abado, is a peripheral location for low cost and government constructed residential condominiums

For this analysis, we identified contrasting neighbourhood locations in the city from the centre, periphery and transitional neighbourhoods. The neighbourhoods could represent different socio-economic classes based on residential housing types. We selected the households randomly (199 in the central, 209 in the transitional, and 267 in the periphery). We wanted to compare waste management situations and test if the amount of waste generated from a household and the household’s distance from a waste transfer station affects behaviour when it comes to practising improper disposal. Under our context, we defined waste transfer stations as the site where the neighbourhood’s waste is collected until it is picked up and transported to the landfills. We mapped the transfer stations to identify their status and location.

Comparative observations showed that transfer stations are closer to households in central and peripheral than in transitional neighbourhoods. On the contrary, the waste generation per household per week is higher in transitional neighbourhoods. Across all neighbourhoods, there was no recognizable difference in knowledge on waste segregation practices. Improper disposal was higher in the central neighbourhood, followed by the transitional neighbourhoods (Figure 2).

The central neighbourhood (AU area) is dominated by low-rise congested slum housings. In this area, solid waste transfer stations are closer to the residing households than in the next neighbourhood in the transitional zone (CMC area). On the other hand, transfer stations in the peripheral area have the shortest household-to-transfer-station distance. This neighbourhood is dominated by densely populated high-rise condominiums constructed by the government for the medium class residents. Therefore, there is a high likelihood of accessing a transfer station closer to their residential blocks as the housing arrangement is in a vertical layer (high-rise) buildups (Figure 3). Household waste generation per week also seems to vary across neighbourhoods, though variations significance needs further test. Households located in the transitional zone of the neighbourhood seem to generate a higher volume than the remaining locations. This neighbourhood is composed of low rise residential houses owned by those that are from comparably better-off socioeconomic classes (Figure 4).

In addition, we also tested the hypothesis to see if solid waste volume and physical distances from transfer stations could affect households decision to dispose of their waste improperly. Therefore, we have assumed that households that reside the farthest from the transfer station and generate a higher volume of waste would likely dispose of their waste improperly. Improper waste disposal practices (such as on streets, open fields, and open canals) were linked to the households that participated in the survey.

Using the existing distances between households from transfer stations and the waste amount produced by the household, a simple logistic regression method was applied to run the test. The model was used to analyze if the aforementioned factors determine a household’s behaviour on how waste is disposed of and to predict the probability of a household’s tendency to dispose of waste improperly. This information could give insight to the local waste administration unit to either reorient the transfer stations distributions; or set priorities to tackle important factors affecting households’ behaviour. Therefore, improper waste disposal is captured as a binary category with a maximum likelihood of one.

Where P is the likelihood of improper waste disposal behaviour, x is the independent variable explaining the occurrence of improper waste disposal, α is an intercept, and β is a coefficient for independent variables. Then the Analysis of maximum likelihood estimates could lead us to what factor could be determinant for improper waste disposal for households. Analysis of results returned coefficients of intercept =-3.8961024, Distance from transfer station=0.0007044, and Waste size=0.1834206 and their significances < 2e-16 ***, 0.711, and 3.59e-14 *** respectively. Results depict that it was only the size of the waste, which could be said statistically important factor, but not the household’s distance from a transfer station. Therefore, the occurrence of expected behaviour (improper waste disposal) could be explained as : 

From this simple analysis, we have learned that a household’s distance from its nearby transfer station does not determine the household's behaviour when it comes to the proper disposal of waste. The size of the generated waste could also affect a household’s decision on whether to dispose of it in the assigned transfer stations or to practice improper disposal. Therefore, we have concluded that our alternate hypothesis on the improper disposal of solid waste is partially accepted, as solid waste size determines improper solid waste disposal. Based on these findings, the local solid waste administration could focus on areas that generate a larger quantity of waste per household per week and develop a set of interventions that enhances proper waste disposal. Beyond this piece of insight, an emulation of similar experiences on data-driven decisions could be applied to analyze multiple suspected factors that are affecting certain situations in a development context, similar to waste disposal behaviour. This approach could help local authorities and development organizations be pragmatic and manage for impacts, instead of having blanket interventions that sometimes result in wasted time, financial resources and low results.