Study area

The study was done in Nairobi City County, the administrative and economic capital of Kenya (Fig. 1). Nairobi City County falls between latitude 1° 09′ and 1° 27′ south, between longitude 35° 59′ and 37° 57′ east, and an altitude range of 1600 m to 1850 m above sea level. Although the City covers an area of 703.9km2, its metropolitan area has stretched out to approximately 3000km2 in its neighbouring Counties of Kiambu, Machakos and Kajiado to the north, east and south, respectively (Bekker and Fourchard 2013). Nairobi City County falls within agro-ecological zone III and experiences a typically sub-humid climate, with a bimodal rainfall of between 638 and 899 mm on average annually, and an average temperature between 10 and 29°C (GoK 2014). The long rains occur from March to May, while short rains occur from October to December (Amwata et al., 2015).

Fig. 1
figure1

Nairobi City’s population is currently estimated at 4,397,073 people, which accounts for 8% of the nation’s populace (KNBS 2019). It is also one of the fastest-growing and highly urbanized cities in Africa (Ren et al. 2020; RoK 2016). The city was selected as a good representative of many other urban areas experiencing high urbanization, population growth, increased migration and settlement of pastoralists in cities and the emergence of alternative livelihoods besides pastoralism. The city is neighbouring Kajiado and Narok Counties, which are dominated by the Maasai nomadic pastoralist community. Besides, Nairobi City is traditionally a dry-season grazing reserve for the Maasai herders who normally returned home when the wet season sets in (Boles et al. 2019). In fact, the name ‘Nairobi’ is derived from the Maasai words ‘Enkare Nyirobi’, which translates to a place with cool waters (Mwita and Giraut 2020). In the city, pastoralist herders resided in pastoralist bomas (temporary livestock enclosures) and manyattas (permanent settlements) located in Dagoretti, Lang’ata, Embakasi, Kasarani and Westlands Sub-counties of Nairobi. Moreover, Nairobi City has a number of policies and legal frameworks governing rights to use of urban land for connected reasons, among them, the Urban areas and Cities Act, No 13 of 2011 (RoK 2011); the Physical and Land Use Planning Act, No. 13 of 2019 (GoK 2019); and the Nairobi City County Urban Agriculture Promotion and Regulation Act, No. 4 of 2015 (GoK 2015) which promotes inclusion of crop and livestock production in urban farming.

Data collection

Data was collected from five Sub-counties of Nairobi City County, namely Dagoretti, Lang’ata, Embakasi, Kasarani and Westlands (Fig. 1), which were purposively selected due to the presence of pastoralist herds, bomas and manyattas. A reconnaissance survey revealed the existence of 193 pastoralist households mainly from Kajiado and Narok Counties. Upon siting the first boma and household herd, a snowball sampling approach was used to identify the location of other pastoral households and herds in the city. Therefore, each household was targeted with one respondent (above 18 years), to raise a sufficient sample size for this study. Given the high heterogeneity among the bomas and manyattas, the intention was to interview all the households. However, only those whose representatives were available and willing were interviewed. A total of 144 households selected through a proportionate sampling of the sub-counties were interviewed in Dagoretti (13), Lang’ata (97), Embakasi (17), Kasarani (15) and Westlands (2) between February and October 2020. In addition, 16 key informant interviews (KIIs) and 6 focus group discussions (FGDs) were conducted to complement and validate the information from individual households. Each FGD comprised between 6 and 8 participants consisting of a mixture of at least 2 youth (between 18 and 35 years) and 2 elderly men and 2 women (above 35 years). On the other hand, KIIs consisted of the leaders of pastoralist bomas in the City, area administrative leaders, livestock traders and officials from national and county governments.

Data analysis

Data analysis involved both descriptive statistics and regression using STATA version 15. Descriptive analysis was done to generate means, standard deviations, frequencies and percentages of the socio-demographic attributes of the sampled pastoralist households. Binary logistic regression was used to determine factors motivating migration and settlement of pastoralists in Nairobi City.

Description of variables used in the binary logit model

The migration of pastoralists and their herds into Nairobi City was used as the dependent variable (Y) in the model. The variable was categorized into a binary response, namely pastoralists who have migrated permanently (who migrate, settle and occasionally engage in other economic activities alongside pastoralism in Nairobi City) and those who have migrated temporarily (those moving into Nairobi only during the extended dry season and back home soon after rains). The dependent variable was assigned 1 for the permanent migrants and 0 for the temporary migrants. The independent variables hypothesized to influence the migration of pastoral herders included household herd size, household land size, access to pasture and water, alternative markets, pests and diseases, gender of household head, age of respondent, household size, education level and presence of relatives in Nairobi (Table 1).

Table 1 Description of variables and expected influence

Household herd size

Herd size was expected to have a positive relationship with pastoralist migration. It was hypothesized that the larger the number of animals owned by a pastoralist, the faster the depletion of available forage, which increases the chances of migrating permanently into new areas. For the purpose of standardized comparison of livestock numbers, household herds were converted into Tropical Livestock Units (TLUs), a universal unit of measurement of which 1 TLU is equivalent to a 250-kg livestock life body-weight (Abebe 2012) and calculation made using livestock species conversion factor as described by Gietema (2006) and Ducrotoy et al. (2017).Footnote 1

Household land size

Households’ land size was hypothesized to have a negative influence on the migration of pastoralists to the city. It was expected that pastoral households owning small parcels of land were more likely to migrate to the city and stay longer than those with larger land sizes. This is partly because the size of land owned by a household is regarded as an indicator of wealth (Omollo 2017). Wealthy households are therefore not only able to afford the cost of temporary migration both within and between seasons, but also have somewhere to go back to as compared to small parcel owners or landless ones. In this study, household land size was a continuous variable measured in hectare (ha) units.

Access to pasture and water

Access to pasture and water resources was expected to have a negative influence on the migration of pastoralist households. It was expected that pastoralists with limited access to pasture and water at origin were more likely to migrate and stay longer in the city as compared to those having better access. This is because pasture and water are central for pastoral livestock production. In this study, access to pasture and water resources was a dummy variable assigned 1 if respondents migrated in search for pasture and water in the city and 0 if they did not migrate for this reason.

Alternative markets

Availability of alternative markets and income-generating opportunities in the urban areas was considered to be positively correlated to the migration of pastoralists to the city. Poor livestock markets and unsupportive market-based policies have been among the major constraints in the pastoral production sector (Amwata et al. 2015; Brussels 2012). Therefore, perceptions of better market opportunities in urban areas are likely to trigger migration and longer stays in the city by pastoralists who wish to take advantage of trade opportunities to enhance their livelihoods. Alternative markets are opportunities for trade and attractive prices for pastoralists’ commodities in the city not available at origin. Such market opportunities include the sale of live animals, milk, livestock manure, leather products (belts, wallets and sandals), clubs, beadwork, wild honey and traditional medicine in the city. Search for alternative markets by pastoralists in Nairobi City was a dichotomous variable assigned 1 if the respondent moved to seek alternative markets and 0 if they did for other reason.

Livestock pests and diseases

Livestock pests and diseases were hypothesized to have both positive and negative effects on the pastoralists’ migration. This is because occurrence of pests and diseases such as East Coast fever (ECF), Foot and Mouth Disease (FMD), Rift Valley fever (RVF), and Trypanosomiasis among others, undermine health and productivity of pastoral herds and thus is expected to trigger migration to new areas of refuge (D’Alessandro et al. 2015). Pastoralists who have been previously exposed to pests and diseases are more likely to migrate permanently to other areas to evade such shocks. On the other hand, an outbreak of livestock pests and diseases was unlikely to cause migration of pastoralists to the city, since pastoralists are well-known to possess indigenous technical knowledge (ITK) for management of livestock pests and diseases that have previously faced them (Muricho et al. 2018; Oba 2012; Onono et al. 2019). In this study, this was a dummy variable, denoted by ‘yes’ if the respondent mentioned pests and diseases as the reason for migration to the city, otherwise ‘no’ if they did not give that as the reason for migration.

Gender

Gender of the household head was expected to have a positive influence on the settlement of pastoralists in Nairobi City. This is because traditionally, it is men who migrate with herds to distant pastures among pastoralist households, whereas women may just temporarily follow them to supply food. Gender was a dummy variable assigned a value of 1 for the male respondent and 0 for the female respondent.

Age of respondent

Age was expected to have a negative effect on the settlement of pastoralists in the city. It was anticipated that younger pastoralists, being in greater need for employment, are more likely to migrate in search of opportunities in urban areas. Most of the pastoralist youth seek wage employments as security guards, drivers, civil servants, casual labourers, business or petty trade outside herding (Coppock et al. 2017; IOM 2015; Munishi 2013). In addition, the youths are the ones mainly entrusted with herding in pastoral systems and therefore would be the ones to migrate with herds to the city. The respondent’s age was categorized into two: youth (between 18 and 35 years) assigned a value of 1 if they migrated, and elderly persons (above 35 years) assigned a value of 0 if they did not for this reason.

Household size

Household size was expected to have a positive effect on the migration of pastoralists to the city. This is because large households with readily available labour are likely to migrate and stay longer than their counterparts with no or less herding labour. Herding labour is a critical production factor in extensive livestock production systems (Roessler et al. 2016). Respondent’s family size was a continuous variable measured as the total number of individuals in a household, consisting of the household head, spouse(s) of the head, children, relatives and employed labourers.

Education level

Education plays a critical role in influencing social networks, access to information and several employment opportunities (Kibera 2013; Ochieng and Waiswa 2019). The respondent’s education level was hypothesized to be positively related to the settlement of pastoralists in the city. Pastoralists with higher education level were more likely to access a variety of livelihood opportunities and stay longer in the city as compared to those with no or low education. In this study, education level was measured as the number of years spent in school and assigned four levels, namely 0 if not educated, 1 for primary education, 2 for secondary education and 3 for pastoralists with tertiary education level.

Relatives in Nairobi

Pastoral communities rely on kinship ties especially when faced with shocks such as droughts, and as a result, individuals will tend to gravitate back to the family and clan bonds during hardships. It was hypothesized that the presence of relatives in Nairobi has a positive influence on pastoralists’ settlement in the city. Pastoralists with relatives in the city are usually assured of assistance in times of crisis and therefore likely to migrate and settle permanently in the city as compared to the ones without relatives in Nairobi.

Specifications of the binary logit model

The binary logit model was used to determine the factors that influence the migration of pastoralists to Nairobi City, given that the nature of the dependent variable elicited dichotomous responses of ‘permanent’ and ‘temporary’ migration. Binary logistic regression (BLR) was selected over ordinary least regression (OLS) because it accommodates categorically measured variables, non-linear relationships and non-normally distributed residuals (error terms). The BLR model was chosen after the statistical test for normality confirmed that the error terms were logistically distributed at p < 0.05.

The logit model was represented as follows:

$$ Logleft[frac{{mathrm{P}}_1}{1-{mathrm{P}}_1}right]= Logitleft({mathrm{P}}_1right)=alpha +{beta}_i{x}_i+{upvarepsilon}_t $$

(1)

$$ Y= Inleft[frac{{mathrm{P}}_1}{1-{mathrm{P}}_1}right] $$

(2)

The regression model for pastoralist migration was specified as follows:

$$ Logleft[frac{{mathrm{P}}_1}{1-{mathrm{P}}_1}right]=alpha pm {beta}_1mathrm{HDSZ}-{beta}_2mathrm{LASZ}-{beta}_3mathrm{APW}+{beta}_4mathrm{ALTM}pm {beta}_5mathrm{LPD}pm {beta}_6 GEN-{beta}_7 AGEpm {beta}_8 HSZpm {beta}_9 EDLpm {beta}_{10}mathrm{REL}pm {upvarepsilon}_t $$

(3)

where:

P1 is the probability of migrating permanently, (1-P1) is the probability of migrating temporarily, ( left[frac{{mathrm{P}}_1}{1-{mathrm{P}}_1}right] ) is the odds ratio, Y is the dependent-categorical variable, xi is the ith predictor variable, α and βi are the estimated coefficients for predictor variables and εt the error terms in the model.

The predictor variables in Eq. 3 are specified as HDSZ = household herd size, LASZ = household land size, APW = access to pasture and water, ALTM = alternative markets, LPD = livestock pests and diseases, GEN = gender of the household head, AGE = age of respondent, HSZ = household size, EDL = education level and REL = presence of relatives in Nairobi.

Multi-collinearity statistical test

To ensure the non-correlation assumption is not violated in the binary logistic model, a multi-collinearity test was carried out in order to establish the relationship between explanatory variables. The variance inflation factor (VIF) method for multi-collinearity detection was preferred since it provides both magnitude and acceptable collinearity limits in the model.

The VIF equation was specified as follows:

$$ mathrm{VIF}=frac{1}{1-{R}_i^2} $$

(4)

where Ri2 is the R-squared value of the regression with ith predictor variable as a dependent variable.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Disclaimer:

This article is autogenerated using RSS feeds and has not been created or edited by OA JF.

Click here for Source link (https://www.springeropen.com/)