POVERTY THE WORLD BANK REDUCTION AND ECONOMIC MANAGEMENT NETWORK (PREM) Economic Premise MAY 2012 • Number 81 JUN 010 • Numbe 18 68665 Why Cargo Dwell Time Matters in Trade Gael Raballand, Salim Refas, Monica Beuran, and Gozde Isik The international community has been increasing investment in projects that promote trade facilitation and improve logistics in the developing world, including in ports. In Africa, a key motivation for such projects has been a presumption that poor infrastructure and inefficient border control agencies are the major causes of extended delays in sub-Saharan Africa (SSA) ports. Based on new data and analysis, this note argues that collusion between controlling agencies, port authorities, private terminal operators, logistics operators, and large shippers is an important part of the problem. Decreasing dwell times in ports requires governments to combat collusive practices between the private sector and public authorities and recognize that large-scale investments in infrastructure are not sufficient to reduce logistics delays. Infrastructure gaps and high transport costs are critical factors this assumption is incorrect in most ports in SSA. The study hindering growth and poverty reduction in SSA. Efficient, low- comes at a time when several investments in container termi- cost transport systems are prerequisites for African countries to nals are planned in SSA. Disentangling the reasons behind car- become competitive in the global market. Hummels (2001) go delays in ports is crucial to understanding (i) if projects by demonstrates empirically that increased transport time dra- the World Bank and other donors have addressed the most sa- matically reduces trade. Without rapid import processes, trade lient problems and (ii) if institutional port reform and infra- based on assembly operations for export is impossible: delays structure investment, sometimes complemented by customs and unpredictability increase inventories and prevent integra- reform, are the most appropriate answers to the problem. tion into global supply networks. The automotive industry in A New Survey of Dwell Times in SSA South Africa listed barriers to reducing inventories as the most important of 12 major impediments to business. Port dwell time refers to the time cargo (containers2) spends Reducing port dwell time is critical. Arvis et al. 2010 dem- within the port or its extension. To separate the components of onstrate that over 50 percent of total land transport time from cargo delays, the study undertakes a comprehensive analysis of port to hinterland cities in landlocked countries in SSA is spent several unique data sets: in ports.1 It has been unclear, however, which port operation i. Data collection in six ports in SSA (Raballand et al. 2012): components contribute most to this dwell time. Over the last Tema (Ghana), Lomé (Togo), Douala (Cameroon), Mom- decade, international donors have assumed that controlling basa (Kenya), Dar es Salaam (Tanzania), and Durban agencies such as customs are primarily responsible for long port (South Africa).3 delays, with infrastructure issues as the secondary cause. Data ii. Firm surveys (manufacturers and retailers) in Kenya, Nige- collected in a new study (Raballand et al. 2012) suggest that ria,4 South Africa, Uganda, and Zambia to assess logistics 1 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise constraints on importers/exporters, large- and small-scale more importantly, the strategies of importers and customs bro- companies, and traders and their demands for port effi- kers. In SSA, importers often have strong incentives to use ports ciency. as storage areas. At the Douala port, for example, storage in the iii. Discussion of results with stakeholders in the selected port is the cheapest option for an importer for up to 22 days (11 countries. days beyond the container terminal’s free time). Firm surveys Raballand et al. (2012) used a mix of databases, individual demonstrate that low logistics skills and cash constraints ex- questionnaires, and aggregated statistics from customs agencies plain why most importers have no reason to reduce cargo dwell and terminal operating companies in eight countries. Except time; in most cases, it would increase their input costs. In addi- for Durban and Mombasa, all of the studied ports are run by tion, collusion of interests may reinforce rent-seeking behaviors private container terminal operators. among shippers, intermediaries, and controlling agencies. Some terminal operators earn large revenues from storage. Cus- Breaking Down Dwell Time toms brokers do not fight to reduce dwell time because the inef- ficiency is charged to the importer and eventually to the con- Dwell time figures have become a major commercial instru- sumer. ment to attract cargo and generate revenues. Port authorities Prevailing market structure also helps explain the durabil- and container terminal operators have increasingly strong in- ity of certain patterns in cargo dwell time. Firm surveys show centives to lower the real figure. The average or mean dwell that companies may use long dwell times as a strategic tool to time has usually been the main target indicator in SSA ports. prevent competition, similar to a predatory pricing mecha- This statistic is easy to compute and easy to understand. How- nism. Incumbent traders and importers, as well as customs ever, because high dwell times are often driven by a minority of agencies, terminal operators, and owners of warehouses benefit problematic shipments, it is difficult to decrease the average/ from long cargo dwell times (two to three weeks), which act as mean dwell time in the short and medium term. In Douala, for a strong barrier to entry for international traders and manufac- example, planners set an objective of 7 days at the end of the turers. Delays at port also may be considered a means to sustain 1990s, but the dwell time remains over 18 days, despite real comfortable rent generation. Cargo dwell times in SSA show an improvements for some shippers. abnormal dispersion, with evidence of discretionary behaviors Cargo dwell times5 in SSA ports are unusually long—more that increase system inefficiencies and total logistics costs. than two weeks on average, compared to under a week in large In Durban, two factors have helped improve dwell time: a ports in Asia, Europe, and Latin America. Excluding Durban strong, domestic private sector with interests in global trade, and Mombasa, average cargo dwell time in most ports in SSA is and a public sector willing to support it. A “penalty storage� fee close to 20 days (table 1). Very long dwell times in SSA ports has discouraged long-term storage at the port and has helped hurt the efficiency of port operations and the economy in gen- Durban maintain a dwell time of three to four days, compara- eral. A common assumption holds that the private sector (ter- ble with ports in Europe and the lowest in SSA. Using Durban’s minal operator, customs broker, owner of container depots, example and simulations of container movements in a port ter- shipper) has an interest in reducing dwell time. But this is not minal, simulations suggest that a reduction of dwell time from always true, and is pointedly not the case in most SSA ports, one week to four days would more than double the capacity of where collusion of interests between controlling agencies, port the container terminal without any investment in physical ex- authorities, private terminal operators, logistics operators tensions (Raballand et al. 2012). (freight forwarders), and large shippers drives up prices for consumers. Storage Dynamics and Strategies of Importers and Poor handling and operational dwell time generally ac- Customs Brokers count for no more than 2 days out of at least 15 days of dwell Case studies and shipment-level analysis show that long dwell time on average. Most delays are due to transaction and storage times are mostly related to factors that depend on shippers. time, resulting from controlling agencies’ performance and, More specifically, the demand by importers for excessive port dwell time seems to be related to the private sector’s inventory management and “business Table 1. Average Dwell Time in SSA Ports (Days) model� (including informal practices). Average Analysis demonstrates that cost-minimiza- Dar es (Durban tion and profit-maximization strategies may ex- Durban Douala Lomé Tema Mombasa Salaam excluded) plain why behaviors that seem irrational, such as leaving cargo in the port, are in fact the best op- 4 19 18 20 11 14 16 tion for an importer. Shippers importing con- tainerized cargo have to choose a logistics path- Source: Kgare et al. (2011) and surveys. 2 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise way, starting with loading containers in the supplier’s facilities conditions allow a company to justify higher markups or hold and ending with unloading them in the customer’s facilities. inventories to speculate on higher sale prices. Among oligopoly The decisions within the pathway are generally based on ratio- cases, in a cartel or leader–follower situation, businesses will nal criteria such as cost, delivery time, frequency and risk, as follow monopolistic pricing strategies. In a price war, the mar- well as some behavioral patterns (for example, repeated buyer ket behaves as if in free competition, and companies try to min- behaviors). imize dwell time. The analysis shows that any market player seeking to mini- Market structure also helps to explain why high cargo mize total logistics costs will try to reduce port dwell time. But dwell times are difficult to reduce. Monopolists and cartels two secondary conclusions are important. One is that long have an incentive to reduce dwell time, but to retain their mo- clearance times encourage shippers to split their annual orders nopoly power, they must discourage worldwide competition. A into smaller, more frequent batches. Another is that companies long cargo dwell time (two to three weeks) is a strong barrier to seeking to minimize total logistics costs will, in some cases, pur- entry by competitors. Thus, in most SSA ports, a vicious circle posely leave their cargo in the port. This happens when the fi- arises in which monopolists favor high dwell times, which dis- nancial cost of clearing cargo immediately from the port is courages global competition and enforces their market power. higher than the potential storage-cost savings in private or Terminal operators’ incentives may also affect dwell time. third-party storage facilities outside the port. In other words, Storage tariffs bring revenues, so the optimal policy for a port shippers may wait to sell the cargo—holding it in relatively inex- operator could be to increase dwell time. Port authorities pensive port storage facilities—before paying the required port might be interested in increasing dwell times because: (i) as to- clearance charges. Thus, port terminals become strategic stor- tal revenues increase, employees receive informal payments, age units. and (ii) higher dwell times provide justification to increase port capacity, which means additional funding for infrastruc- Market Structure and Profit Optimization ture investments. Other observed strategies, including the paradoxical situation This cycle explains why cargo dwell time is so stubborn where shippers seem to be indifferent to long dwell times, can and why the industries that prosper in SSA are not time sensi- be explained using the economic theory of monopoly. Despite tive—such as exports of raw materials or minerals. Those that being a cost setter, a rational monopolist should seek to reduce flounder are time-sensitive, value-added industries. port dwell times to optimize profits because he or she cannot pass on all costs to clients without losing sales. But where de- Implications for Donors in SSA mand is inelastic to price, the monopolist is not affected in the short term by higher logistic costs and will therefore make no These findings may help explain why many trade facilitation effort to reduce dwell times. Such a scenario is likely to happen measures have faced difficulties in SSA. Market incentives are for cyclical demand patterns that are elastic to price only in the too weak for supply-side measures to drive radical changes in long term (food supplies, drugs, equipment, and so forth). trade logistics efficiency. Rather, actors in the trading, industri- A third behavior observed among monopolistic compa- al, and logistics sectors exhibit risk-averse behaviors because nies is opportunistic pricing. In some cases, adverse logistics they are operating in a context of oligopolistic competition where significant adjustments do not translate into obvious gains for them. An implication is that governments and donors Figure 1. The Vicious Circle of Cargo Dwell Time need to re-think intervention strategies. Table 2 presents policy recommendations to reduce high dwell times in SSA. One of the worst options is to invest in additional storage and off-dock fraud and collusion in Long cargo dwell time customs and port because of mutliple yards where congestion occurs. Structural issues that lead to bargaining processes long dwell times, including demand characteristics, need to be tackled before undertaking costly physical extensions. Effective solutions to high dwell times in SSA ports will revolve around the challenging task of breaking the private sector’s short-term collusive strategies and providing incentives for public authori- Low pressure for Self-selection of ties, intermediaries and shippers to reduce delays. increasing productivity, importers/brokers with a simpli�cation, and rentier objective About the Authors reducing cargo dwell (monopolies/oligopolies times (cost plus margin and informal sector) + Gael Raballand is a Senior Economist, Monica Beuran is a Consul- strategy for companies) barrier at entry for tant, and Gozde Isik is an Economist, all with the Africa Region of competitive private sector the World Bank. Salim Refas is a Transport Economist for the Is- Source: Raballand et al. 2012. lamic Development Bank. 3 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise Table 2. Policy Recommendations Dos Don’ts When facing a port’s capacity shortage, envisage o more optimal use of the When facing port’s capacity shortage, immediately consider capacity existing capacity by targeting long-stay containers/cargo and encouraging extensions. fast clearance through price incentives. Conduct a careful assessment of the way the private sector operates before Necessarily privatize/concession a container terminal to reduce dwell time. investing in port infrastructure: understand demand before changing supply. Inform public decision makers at highest levels (prime minister, ministries of Support measures that create new rents and reduce system transparency, economy and finance) on the need to implement public governance–related such as proliferation of off-dock container yards. actions to build a broad coalition to change the equilibrium. This should include thorough analysis of the economic cost of poor system performance to the national economy. Sensitize the local population and trading communities on the importance of Consider as a given that everybody is aware that transport and port “costs� port clearance performance and the proper calculation of total logistics costs. are high, and address the issue of port delays only from a monetary cost perspective (with no mention of the time and reliability costs). Identify port performance indicators with a benchmark pegged to the most Report averages only with no distinct evaluation of good, average, and poor efficient shippers in the port. performance. Source: Authors’ compilation. Notes 5. This note focuses on import containers because they are im- portant for import-export models, and dwell time is usually 1. Wilmsmeier, Hoffmann, and Sanchez (2006) found that the low for outbound containers. Most boxes stay in port of one to combined port efficiency of the importing and exporting coun- two days to be marshaled before loading. Bulk or noncontainer- tries’ ports has a very strong impact on maritime charges. In- ized general cargo usually fit specific patterns of storage and creasing the indicator for port efficiency by 1 percent reduces loading/unloading strategies. freight charges by 0.38 percent. If the two countries of the sample with the lowest port efficiency improved their efficien- References cy to the level of the two countries of the sample with the high- Arvis, Jean-François, Gael Raballand, and Jean-François Marteau. est indexes, the freight charges on the route between them 2010. The Cost of Being Landlocked. Washington, DC: World would be expected to decrease by 25.9 percent. Bank. 2. Even in SSA, more than half of total imports are container- Hummels, D. 2001. “Time as a Trade Barrier.� GTAP Working ized, and this traffic is growing. Data for containers are more Paper 1152, Center for Global Trade Analysis, Department of systematic and reliable than bulk traffic data. Agricultural Economics, Purdue University. Kgare, T., G. Raballand, and H. W. Ittman. 2011. “Cargo Dwell 3. This study strived to select the largest or among the largest Time in Durban.� World Bank Policy Research Working Paper ports in the four subregions: Durban in southern Africa (which 5794, Washington, DC. is also the largest in SSA), the two largest in East Africa (Mom- Raballand, G., S. Refas, M. Beuran, and G. Isik. 2012. Why Do basa and Dar es Salaam), Douala in Central Africa, and Lomé- Cargo Spend Weeks in Sub-Saharan African Ports? Washington, and Tema in West Africa. DC: World Bank. Wilmsmeier, Gordon, Jan Hoffmann, and Ricardo Sanchez. 2006. 4. Nigeria was selected because although it undertook a major “The Impact of Port Characteristics on International Maritime port reform, it has continued to face long dwell time. Nigeria is Transport Costs.� In Research in Transportation Economics, Vol- also one of the most important African economies. ume 16, ed. Kevin Cullinane and Wayne Talley. Elsevier. The Economic Premise note series is intended to summarize good practices and key policy findings on topics related to economic policy. They are produced by the Poverty Reduction and Economic Management (PREM) Network Vice-Presidency of the World Bank. The views expressed here are those of the authors and do not necessarily reflect those of the World Bank. The notes are available at: www.worldbank.org/economicpremise. 4 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise