Space-time modelling of unemployment rate in Namibia
| dc.contributor.author | Amutenya, Fransina | |
| dc.date.accessioned | 2021-07-01T05:43:44Z | |
| dc.date.available | 2021-07-01T05:43:44Z | |
| dc.date.issued | 2021 | |
| dc.description | A thesis submitted in partial fulfillment of the requirements for the Degree of Master of Science (Applied Statistics and Demography) | en_US |
| dc.description.abstract | Among major and burning issues in the developing world is the problem of unemployment. Not being employed does not only have a negative effect on the wellbeing of an individual, but also pose a great concern to policy makers at all levels of government. Namibia like any other developing country especially in Africa, continues to carry the burden of a high unemployment rate, varying across regions. This study aimed at determining whether variation in unemployment rate is due to regional clustering and if covariates such as education, sex, age, population density and time have effects on the regional unemployment rate variations. To achieve these objectives, the study used a Fully Bayesian (FB) spatial smoothing approach with temporal trends to assess regional variations in the unemployment rates obtained from the Namibia Labour Force Survey data of 2014, 2016, and 2018. Results indicated that most of the variation in the average unemployment rate was due to the temporally unstructured effect and not regional clustering, while the effect of age, sex, and population density were insignificant. Educational level on the other hand was found to have a significant negative effect on the variation of unemployment rate. Furthermore, the probability map showed a long-term increasing risk of unemployment rate in regions such as Kunene, Omaheke and Kavango West. In conclusion, results showed that the variations in the unemployment rate in Namibia are not due to regional clustering, nor do covariates such as sex, age, population density have a significant effect on them. However, introducing the the unstructured time trend component in the model significantly explains the variation in the unemployment rate over time and space. Also, educational level has a significant negative effect on the variation of unemployment rate. Lastly, the thesis showed that the best response distribution for modelling rates and proportions is a beta distribution. | en_US |
| dc.identifier.uri | http://hdl.handle.net/11070/2946 | |
| dc.language.iso | en | en_US |
| dc.publisher | University of Namibia | en_US |
| dc.subject | Unemployment rate | en_US |
| dc.subject | Bayesian analysis | en_US |
| dc.subject | Space-time | en_US |
| dc.subject | Labour force | en_US |
| dc.subject | Beta distribution | en_US |
| dc.title | Space-time modelling of unemployment rate in Namibia | en_US |
| dc.type | Thesis | en_US |