The Influence of Lock-in Dynamics and Industrial Policy on the Transition to EV Manufacturing in South Africa

Abstract

South Africa has one of the most developed automotive manufacturing sectors in the Global South. Its plants export mainly to Europe, where demand is moving quickly from internal combustion engine (ICE) vehicles to electric vehicles (EVs). Global regulation, shifting demand, and the risk of losing export access all push the sector to change. The transition, however, has been slow. This thesis asks why. It examines how lock-in dynamics and the industrial policy regime together shape South Africa's transition from ICE to EV manufacturing.

The study applies the six-domain lock-in framework of Helmrich et al. (2023), which distinguishes technological, economic, social, individual, institutional, and epistemic lock-in. Two sub-questions guide the analysis. The first identifies which lock-in mechanisms constrain the transition and how they interact across domains. The second asks how the current industrial policy regime reinforces or challenges these mechanisms. The research is qualitative and document-based. It draws on a systematic literature search, national policy documents, and sector reports. Five policy instruments were assessed: the APDP, SAAM 2035, the EV White Paper, Section 12V of the Income Tax Act, and SAREM. Each provision was coded as reinforcing, challenging, neutral, or silent on lock-in. The scope is limited to four-wheel passenger vehicles.

The first analysis finds lock-in in all six domains. These barriers are not independent. Conditions in one domain sustain barriers in another, forming three chains: one on the production side, one on the demand side, and one running through the policy response itself. The demand chain is the only fully closed loop. Because the chains renew each other, addressing any single barrier in isolation has limited effect. The domains also differ in weight. The institutional and economic domains act as sources of reinforcement, while the individual and epistemic domains receive it and remain largely unaddressed.

The second analysis finds a policy regime in mid-pivot. Its direction has changed, but its weight has not. The older instruments still carry most of the financial support and continue to favour ICE production. The newer instruments challenge lock-in but reach only a narrow part of the sector. The regime also contradicts itself: the state now subsidises EV production while its tax rules still make EVs more expensive to buy than ICE equivalents. Comparing the two analyses sorts the lock-ins into three groups. Some are addressed by at least one instrument. Some are not addressed at all. Some are addressed on paper but persist in practice, showing that policy on record is not the same as policy in effect.

The thesis concludes that the transition is held back by a self-sustaining structure of lock-in that policy has begun to challenge at the edges but not at its sources. No single instrument can carry the transition alone. What matters is whether the actors involved can be brought to act together. Recommendations follow for policymakers, existing automotive companies, and new EV ventures. Future research should examine actor coordination, the implementation of the newer instruments, and whether the same pattern holds in other developing economies.