Opportunities for short-sea shipping in the Southern African Development Community (SADC) region: evidence based on discrete choice modelling

Doctoral Thesis


Permanent link to this Item
Journal Title
Link to Journal
Journal ISSN
Volume Title
The thesis investigates the development of short-sea shipping (SSS) in the Southern African Development Community (SADC) region by studying the determinants of SSS, the stated choice preference of shippers and freight forwarders and the stated intentions of maritime carriers for SSS. It is purported the introduction of SSS in SADC could reduce socio-environmental problems currently faced such as road damage, road congestion, pollution and transport related accidents. Discrete choice modeling (DCM) is employed as the main methodology to study shipper and carrier behavior. Discrete choice modeling permits the construction of general utility functions incorporating various decision maker characteristics and choice attributes to elicit preference of respondents. The general postulate in DCM is that utility is derived from the properties of things rather than the actual thing per se. A particular benefit of DCM in this study is the elicitation of preference for services and interventions that have not been introduced by SSS. The first step in the study is a theoretical investigation of the potential of SSS in the SADC region. It highlights the policy initiatives, the barriers and enablers related to the development of SSS. The proposed SSS system would have three main roles: to offer an alternative mode of freight transport service between port cities, to serve as the main leg in an intermodal transport network, and to serve feeder services between hub-and-spoke ports. The findings reveal that, SSS has the theoretical potential to work in the SADC region, given the large geographic region, projected freight volumes and customs and trade policies the SADC region is pursuing. The second step in the study involves an a-priori study conducted to develop a general understanding of freight transport in SADC. For this purpose, a uniquely developed online survey was conducted across the SADC region to ascertain in particular: who the decision maker is in terms of freight mode choice; and what the significant attributes that influence freight mode choice are. The results reveal that both the shipper and the freight forwarder are involved in mode choice decisions, however the shipper being the dominant decision maker. Furthermore, the results of the exploded logit model reveal that the top five modal attributes that shippers consider most important are: reliability, transport cost, risk of damage, frequency of service and transit time. These results were subsequently employed to inform the shipper and carrier behavior studies. The third step entails the assessment of shipper behavior, where trip specific mode choice decisions are studied along five intra-urban origin-destination (O-D) paired routes (which would form the study corridors). Three of these corridors considered unimodal SSS, and the two considered intermodal SSS. Unimodal SSS was studied along the following corridors: Cape Town (South Africa)~ Walvis Bay (Namibia), Walvis Bay (Namibia) ~ Luanda (Angola) and Durban (South Africa) ~Beira (Mozambique); and intermodal SSS was studied along the following corridors: Durban (South Africa) ~ Harare (Zimbabwe) and Cape Town (South Africa) ~ Windhoek (Namibia). To develop the choice scenarios, d-efficient stated choice experiments were uniquely developed for each of the corridors with the following key modal attributes systematically varied and analyzed across respondents: service frequency, reliability in terms of arriving on time, expected delay, transport cost and transport time. Subsequently, the following choice models were developed: Binary Logit, Mixed Logit and Integrated Choice and Latent Variable Structure models for the unimodal corridors; and Multinomial Logit, Nested Logit and Cross Nested Logit models for the intermodal corridors. The results highlight that in addition to the modal attributes, mode choice decisions are driven by shipper characteristics and situational characteristics. Moreover, the unimodal SSS study reveals that underlying latent perceptions also influence freight mode choice decisions; while the intermodal SSS study reveal strong correlations in the intermodal SSS alternatives, which requires improved intermodal capability if SSS is to become competitive. The fourth step in the study entail the assessment of maritime carriers preference for SSS given varying levels of maritime conditions that include: dedicated freight volumes, income from freight, port dues discount, terminal handling fees discount and ship registration requirements. The results of an ordered logit model reveal that ship registration provisions and terminal handling charges are the most important to the development of SSS from a carrier side. Moreover, ship registration and maritime cabotage provisions require visitation to boost the participation of carriers in SSS. The last step of the study revisits the modeling results and considers their implications through the estimation of willingness-to-pay and attribute elasticities. The results were then employed to suggest policy actions and interventions to develop SSS.