Industrialisation for structural transformation: Economy-wide impacts of agro-processing development in Tanzania

ABSTRACT

Despite high economic growth rates experienced recently, socio-economic challenges such as poverty still prevail in Tanzania. This has been attributed to unsuccessful economic transformation where labour has moved out of agriculture into mainly relative less productive services and informal sectors, leapfrogging the highly productive manufacturing sector. The government is targeting the expansion of agro-processing activities for economic transformation and four policy options were explored. Results of a computable general equilibrium model indicated that policies such as (1) attracting foreign direct investment to improve productivity should be encouraged, especially if (2) coupled with export strategies. (3) Agricultural expansion is also crucial to expand the input base for agro-processing activities. (4) Non-targeted policies such as improving education may not be sufficient for the sector’s expansion but are crucial for the expansion of the whole economy. Government should be mindful of the varied implications of different policies on factor returns and household incomes.

KEYWORDS:

• Industrialisation
• agro-processing
• development
• CGE

JEL CLASSIFICATION:

• C68
• O10
• O55

Disclosure statement

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

Notes

1 Analysis was conducted prior to COVID 19, and submission during early stages of the pandemic. The baseline therefore does not reflect any potential negative implications of the pandemic. It is recogised that Tanzania would have experienced lower GDP growth in 2020 than what is reflected in the baseline. The paper however focuses on the simulation results’ deviations from the baseline and it is expected that the deviations will not be greatly affected because the baseline and the simulations will change proportionately by the same values when adjusting for COVID 19, thus the general conclusions of the paper will not be greatly affected.

2 Although the model is not repeated in this paper, selected model variables and parameters are included for cross referencing with the detailed model discussion that can be found in Lofgren et al. (2002).