A livelihood comprises the capabilities, assets (including material and social resources) and activities required for a living (Carney, 1998; Carney et al., 2000). Niehof (2004) confirmed that livelihood is a multi-faceted concept, being both what people do and what they accomplish by doing it, referring to activities and outcomes. The livelihood is seen as an open system, interfacing with other systems and using various resources and assets to produce livelihoods, with the households as the locus of livelihood generations.

The sustainable livelihoods framework is a way of looking at the complexity of the livelihood of people that seeks to understand various dimensions of a person’s livelihood; the strategies, objectives pursued, associated opportunities and constraints (Gichure et al., 2020)

From Figure 1 above, the Human resource or capital (H), Natural resource (N), Financial capital (F), Social capital (S) and Physical resource (P) are critical to the livelihood of the people. The framework places people, particularly the poor, at the centre of a web of inter-related influences that affect how these people create a livelihood for themselves and their households. Closest to the people at the centre of the framework are the resources and livelihood assets that they have access to and use. These can include natural resources, technologies, skills, knowledge and capacity, health, access to education, sources of credit, or their networks of social support.

The extent of their access to these assets is strongly influenced by their vulnerability context, which takes account of trends (for example, economic, political, and technological), shocks (for example, epidemics, natural disasters, civil strife) and seasonality (for example, prices, production, and employment opportunities). Access is also influenced by the prevailing policies, institutions, and processes, which affect how people combine and use their assets to achieve their goals or livelihood strategies (Gebbisa and Mulatu, 2020).

The vulnerability context refers to unpredictable events that can undermine livelihoods and cause households to fall into poverty. It is essential to distinguish between shocks originating from outside the community, which affect all people in the same locality, and individual shocks that principally affect only individual households (Ellis, 2000; Gebbisa and Mulatu, 2020). Examples of vulnerability contexts include drought, earthquakes, floods, pest and disease epidemics (insect attacks and diseases affecting crops, animals and people), economic shocks (price fluctuations, markets, employment and purchasing power), civil strife (war, armed conflict, displacement, destruction of lives and property) seasonal stresses (hungry season food insecurity) environmental stresses (land degradation, soil erosion, bush fires) and idiosyncratic shocks (illness or death in family, job loss or theft of personal property).

Livelihood assets refer to the community’s resource base and different categories of households (Inostroza et al., 2016). The DFID (2000) identified five assets (capital) of livelihood, all interrelated in the sustainable livelihood framework. These capital assets are Human capital, Natural capital, financial capital, Social capital and Physical capital. In order to determine the value of assets owned by the urban households, the paper used asset-based measures to compute a wealth index. The wealth index is a composite measure of a household’s cumulative living standard. The wealth index was calculated using easy-to-collect data on a household’s ownership of selected assets, such as televisions and bicycles; materials used for housing construction; and water access and sanitation facilities (Rutstein & Kiersten, 2004; Anderson, 2012). The wealth index was generated by a statistical procedure known as principal components analysis (PCA) and placed individual households on a continuous scale of relative wealth (Inostroza et al., 2016).

Carney (1998) defined livelihood strategies as “the range and combination of activities and choices that people make to achieve their livelihood goals”. Livelihood strategies include how people combine their income-generating activities, how they use their assets, what assets they choose to invest in, and how they preserve existing assets and income. In this paper, the strategies observed were urban street vending, urban cattle ranching, urban agriculture for crops, wage earners, and running a “bodaboda” motorcycle taxi.

Livelihood outcomes are what household members achieve through their livelihood strategies, such as levels of food security, income security, health, well-being, accumulation of assets and high status in the community. Unsuccessful outcomes include food and income insecurity, high vulnerability to shocks, loss of assets and impoverishment (Ellis, 2000; Devereux et al., 2012). Outcomes that households claim to achieve include food security, paying social bills, increasing income, and purchasing assets like livestock and land.

The study applied a cross-section research design where both primary and secondary data were collected using questionnaire interviews, documentary reviews, and Key Informants Interview methods. Primary data were collected from the household of Kizota Ward (Dodoma city). In contrast, secondary data were obtained from Kizota Ward office, Dodoma City Council office and Dodoma Urban Water Supply and Sanitation (DUWASA). The documentary review involved the critical and comprehensive review of documents relating to urban livelihood in Dodoma. Documents include City Council population trends, City Council 5 years achievements report (2015-2019), City Council yearly Progress report and Dodoma City Council Profiles (2019). The key informants included Planning Officers, Rural Water Supply and Sanitation Officers, Trade officers, Town Planners, Ward Executive Officers and Mitaa Chairpersons in the study area.

Both descriptive and quantitative data analysis techniques were employed for data analysis. Descriptive analysis was used in those objectives intended to statistically describe, aggregate, and present the constructs of interest or associations between these constructs. This analysis helped summarise and support assertions of underlying facts in the specified objectives. On the other hand, the quantitative analysis involved logistic regression analysis, Household Vulnerability Index, Wealth Index, Quintiles and Principal Component Analysis (PCA). Besides, in cases that required understanding the relationship between categorical variables, a chi-square test was used, as shown in Table 1.

It was revealed that there had been a spurt in population growth for the past six years (2015-2021) with an average of 2.7 per anum and increasing other urbanisation indicators. This is a 15.8 percent increase in population, which greatly adds pressure on land, and other urban resources (Table 2). Within the same period, it was observed that deliberative efforts had been made to increase other services associated with population growth.

Source: a = National Bureau of Statistics (2013), b = Dodoma City Council (2020), c = Dodoma Urban Water Supply and Sanitation (DUWASA).

Statistically, on an average annual basis, health services were increased by 9.5%, water supply increased by 17.8, housing permit services increased by 114.8% and sanitation increased by 8.1%. Other services, including an increase in small and medium enterprises and increased business licenses, were 16.9% and 15%, respectively. It is iterated here that despite such efforts, further studies are needed to ascertain how many segments of the urban population have been reached with these services and explore the gap.

The vulnerability of Dodoma city households was meant to be assessed by the use of three variables, namely “Exposure”, “Sensitivity”, and “Adaptive capacity”. To obtain the values of the variables affecting vulnerability, each indicator from each variable was analysed using Principal Component Analysis (PCA). Out of 7 adaptive strategies, it was found that the number of owned urban plots and land for urban agriculture had higher values implying that owning land in urban areas had the greatest impact in increasing the adaptive capacity of urban households, reducing the impact of vulnerability (Table 3).

Regarding “exposure,” eight variables were used to generate factor scores (Table 3). It was revealed that household size, structure, distance to market, and access to lifelines carry more considerable weight in the model. This implies that the nucleation of households in terms of the number of members and potential supports required increases the urban household vulnerability. On the other hand, lower values of these variables reduce the risk of the vulnerability of households. Furthermore, dependency of household income solely on the street vending and salary was among the six sensitivity indicators having greater scores. This implies that these factors reduce the risk of vulnerability among urban households while low values increase the vulnerability risk.

Factor scores were further analysed to reveal the indices of adaptive capacity, exposure, and sensitivity to compute the vulnerability index. The index was computed using the formula [Vulnerability index (VI) = AC − (E + S)]. It was revealed that the average vulnerability index of urban households is 2.06, which is positive and moderate (Table 4).

This finding implies that efforts should be made to ensure that the adaptive capacity of urban households is enhanced to the level that will reduce their vulnerability. Areas of focus should, among others, by reducing the price of owning plots in the city, allocating land and promoting urban agriculture, promoting household livelihood diversification, promoting social and economic cooperative groups, easing access to loans, ensuring access to safe water and sanitation and improve street roads which affect the sustainability of human urban settlements. Further analysis reveals that 155 (72.09%) households had a vulnerability index below the mean, and only 60 (27.91%) had it above the mean.

Principal Component Analysis (PCA) method was used to assess assets indicator

weight and household asset index to generate factor scores from 13 commonly owned household assets. The results indicate that most households owned television, music system, electricity, solar and table. Few households had vehicles, motorcycles, bicycles, refrigerators, electric cookers or gas cookers. In sharp contrast, having a television and music system is weighted more heavily.

The findings exposed that the wealth index of households in the study area was 0.00549, which indicated that the majority of the households are at an above-average value of the Wealth Index. Further analysis indicated that the Wealth Index is skewed to the left with the highest positive value, implying that the wealth index distribution is much more concentrated on the right side than the normal distribution, as indicated in Figure 2 below.

The quintiles of wealth were also computed based on the index to assess the characteristics of the poor and rich (Table 5).

The values between the lowest and second quintiles are negative, implying that the poor households have much less wealth than others in the study area. At the other end of the distribution, the values between the fourth and the highest quintiles are positive, indicating that households within this quintile possess more assets and are well off than those in the lowest and second quintiles.

For comparison of what people have in the study area, the assets used to construct the wealth index were tabulated according to the quintile of the wealth distribution. The percentages of urban households that have these assets by quintile are shown in Table 6.

It can be concluded that most households owned television, music system, motorcycle, bicycle, and solar assets, mostly owned by the wealthiest people in the study area. It is worth noting that the poorest people in the study area were not found to own vehicles, motorcycles, or refrigerators. Surprising, whether the household was found within the poorest or wealthiest quintiles, they were found to have electricity in their house, and few of them had an electric cooker and or gas cooker.