Horizon blog: Demystifying the data - the story behind the International Agri-Food Trade Network Model

AHDB, in partnership with Harper Adams University, is using a new model to predict the impacts of new and potential trade agreements on UK agriculture, supply chains and the economy. In this guest blog, Dr Daniel May, Lead Researcher and Analyst from Harper Adams University tells the story behind its development.

With the Government setting out an ambitious vision to position the UK as a champion of free trade after EU exit, AHDB rightly felt new work was needed to inform the industry of the likely effects of our new trading relationships on our farming sectors - the first being its newly published analysis on the UK-Australia deal. Over the past few months, we at Harper Adams have been using our newly developed economic model and worked with the AHDB to quantify these potential impacts.

Modelling can be a really helpful tool to try to understand how changes in policy, such as the striking of new trade deals, may affect farmers and growers and the prices they get paid for their products in future. But when we looked at existing models, we found they did not account for the complexities of global agri-food trade and came to the conclusion that we needed to look at things from a different perspective.

This is because international trade of agricultural and food products is characterised by a number of factors.

A good example is the influence of intermediaries, such as food processors, on other participants in the supply chain: in a buyers’ market, these intermediaries have power over the prices they pay to farmers. Competition with intermediary companies from other countries is also a source of influence, resulting in the supply chains between countries becoming highly integrated. This means that anything that affects the market in one country, such as signing a trade agreement, has a knock-on effect to the other countries it trades with.

So, for example, if Country A increases exports of a product to Country B, this increases the level of competition in Country B, negatively affecting the profits of competitor intermediaries in that country. As a result these companies adjust to compensate for their losses. They reduce the amount of product sold in their own country and divert trade to third-party countries. Equally, the costs of intermediaries in Country A may rise, reducing their own profits, as they have to pay farmers more to source the products for export.

Another important factor in international agri-food trade is the existence of policy biases between countries. Often governments develop policies to assist certain groups, maybe under the influence of powerful lobbying organisations, creating distortions in trade flows.

A third factor is ‘intra-industry trade.’ This is where countries both export and import the same product. Here ‘product differentiation’ comes into play if consumers do not consider domestic and imported goods as the same in terms of attributes like quality.

Existing trade modelling approaches do not normally consider the factors outlined above, as they have a tendency to assume that markets operate under idealised conditions. While this gives them the advantage of being relatively straightforward to work with, and they can certainly be of use, it doesn’t reflect what goes on in the real world. So we needed to come up with a different way of doing things.

Using our newly created International Agri-Food Trade Network Model (IAFTN), which is an alternative and novel modelling approach designed to accommodate all the important market characteristics described above, we worked closely with market analysts at AHDB to provide a more realistic assessment of the likely effects of trade agreements. Anyone interested in digging deeper into the methodology behind the work can read the full technical report.

Like all predictive models, the IAFTN model has limitations as it predicts trends based on historic data. As such, we are dependent on the availability and quality of that data. Also economic modellers don’t have access to a crystal ball so we can’t anticipate unexpected shocks to the economy (such as a trade collapsing between countries due to political disagreements). It may have its caveats, but the IAFTN model has provided deeper, more realistic insights into the impacts of trade agreements on our food and farming sectors than has been previously possible and its capacity will continue to be developed.