Long and short-distance internal migration motivations in post-apartheid Namibia: a gravity model approach

Abstract

The paper estimates a gravity model to analyse the region-level differences that explain internal migration in post-apartheid Namibia, with the specific aim of understanding whether there are differences in motivations for long and short-distance migration. Given Namibia’s history of apartheid-era segregation, the sample is later restricted to African-language speaking migrants to determine whether the distances travelled differ from that of the full population. A zero-inflated negative binomial model is applied to estimate the effects of constituency-level economic indicators, labour market conditions, agricultural activity, and built amenities on migration flows. Regression analysis shows that analysing internal migration flows in Namibia without accounting for distance-related differences in migrant motivations may produce misleading results. Disaggregation of migration flows by distance reveals that for both the entire population and the restricted African-language speaking sample, constituency differences in amenity quality are predictors of intermediate-distance migration volumes. Per capita income differences in favour of the receiving constituency increase long-distance migration volumes. For all distances moved, previous migration in the sending constituency is a strong positive predictor of migration volumes. Migration volumes also increase when the sending constituency shares a border with another country.

Keywords:

• Internal migration
• urbanisation
• gravity model
• zero-inflated negative binomial model
• Namibia

Subject classification codes:

• O15
• O18
• R23

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 The migration distance intervals were constructed to roughly coincide with the bottom 25 percent, middle 50 percent and upper 25 percent of the actual migration distance distribution. Most rural-rural moves in Namibia are over relatively short distances, while rural-urban moves require moving longer distances on average.