By Sandra Sequeira First published on the International Growth Centre website When we think about the costs for the developing world to engage in international trade, what often comes to mind are the cost of tariffs, transport and non-tariff barriers, usually in the form of stringent regulatory requirements that the products they are trying to trade have to meet. However, there are other hidden, less predictable, costs that firms still need to grapple with when trying to move goods across space: corruption at border posts. With the World Bank estimating that the burden to the global economy per year is US$1-trillion, or 3% of its GDP, corruption may play an important role in hindering trade. But how can corruption affect firms’ trading decisions? Costs versus benefits On the one hand, corruption can be beneficial to firms. Corruption, conceived as ‘grease’ to oil the wheels, can allow firms to overcome cumbersome regulations, by providing underpaid bureaucrats with incentives to perform and so improving allocative efficiency. On the other hand, corruption could have a negative impact, acting as a form of distortionary taxation which reduces allocative efficiency via lost revenues for government and the increased transaction costs, uncertainty, and unenforceable contracts which hamper business activity. Business behaviour To understand how corruption is affecting firms’ behaviour, we need to look at how corruption alters the marginal price of the public service to firms, and how their demand for the service changes in response to variations in that marginal price. We examined this question in the specific context of how corruption could increase the cost of international trade in sub-Saharan Africa. To do so, we looked at firms in north-eastern South Africa with a choice between using two ports with very different levels of corruption: Maputo in Mozambique or Durban in South Africa. We drew a random sample of three types of firms: those drawn from an area equidistant to the two ports, an area in South Africa that would be considerably closer to the more corrupt port of Maputo and Mozambican firms located close to Maputo, which did not have the option of shipping through the port of Durban. An important aspect of this setup was that the location of firms and the decision of which product to trade pre-dated the opening of the Maputo port to international business following an almost two-decade long civil war in Mozambique. By collecting information on directly observed bribe payments for the shipment of different types of products, we could then combine this information to see how corruption affects firms’ choice of which port to use. The price of corruption An important finding was that corruption varied significantly depending on the type of product being shipped. In particular, South African products that fell under a high tariff grouping according to the tariff code of Mozambique were more vulnerable to being stopped to pay a bribe when travelling from Maputo to South Africa. Falling under a high tariff grouping according to the tariff code of a neighbouring country was uncorrelated with any important shipment attributes such as its size, value, or level of perishability, and was uncorrelated with whether that product would ultimately pay a high or low tariff once arriving at its final destination in South Africa. This strategy thus allows us to see how corruption affects South African firms’ choice of which port to use, and how firms substitute higher transport costs for higher corruption costs. What we found indicates that firms do respond to the price effects of corruption, organising production in a way that increases or decreases demand for the public service. We found that different border bureaucracies created opportunities for officials to engage in two different types of corruption. First, there was indeed collusive corruption where officials were able to sell tariff evasion, mostly to Mozambican firms importing goods through Maputo. This reduced trade costs and meant there were high rents acquired by private agents (as bribes were often only 0.2% of the total tariff due) and significant losses in government revenue (equal to, on average, a 5% nominal tariff reduction). However, we also observed instances of coercive corruption, with private agents being forced to pay an additional fee to see the shipments go through, thus increasing trade costs for the firm. Pricing peace of mind Of the two types, coercive corruption affected South African firms’ choice of which port to use. Given the chance that an imported good would fall under a high tariff code in Mozambique (and so have a higher chance of being subjected to coercive corruption in Mozambique) firms appeared to prefer to double their transport costs and ship their imports through the port of Durban just to avoid the bribe. In the most extreme case, the cost incurred to reroute via Durban rather than Maputo (where the mean bribe was triple what it was in Durban) was three times higher than the cost of the actual bribe. For South African firms that have 57% lower transport costs to Maputo rather than Durban, 46% still chose to use Durban – that figure jumped to 75% of firms for perishable cargo, and 74% for urgent cargo. This is hard to square with standard price theory and suggests that it is the uncertainty created by corruption that firms strongly dislike. The cost of this diversion to a less corrupt port went beyond the additional transport costs directly incurred by the firms: it created imbalanced flows of cargo through the transport network, introducing significant distortions in local transport markets. Overall, this research suggests that reducing corruption may improve allocative efficiency, but that there could be heterogeneous effects on firm-level trade costs depending on whether the reduction is in coercive or collusive forms of corruption. Three implications for policymakers There are some implications from this we can draw for policy: Firstly, incentives for corruption are partly shaped by the organisational structure of public bureaucracies, as these can create differential structural opportunities for bureaucrats to extract different types of bribes. Two possible options are to reduce in-person contact between private agents and port officials, or reducing the steps in the process of public service delivery (e.g. using online submission of documents and single windows for the submission of clearance documents)Secondly, a better understanding of the rules of thumb used by different officials to identify bribe opportunities would enable more targeted anti-corruption strategies.Thirdly, the distribution of rents between public and private agents will determine the degree of public support anti-corruption policies will have: if private agents are getting large rents by for instance evading tariffs; it will be difficult to rely on them to enforce anti-corruption measures. Numerous factors shape how bureaucrats engage in different types of corruption and these can in turn change the costs and benefits for firms involved in international trade. At a more aggregate level, the effects of corruption can go beyond price effects for firms, to generate imbalanced trade flows and lost government revenues. More effective evidence and research-based anti-corruption strategies provide a great opportunity for governments to start cleaning up corruption at borders, thus providing the developing world with a greater chance at reaping the benefits brought by international trade. Sandra Sequeira is a lecturer in development economics at the London School of Economics and Political Science.