Interview | Ferike Thom

On which data do you base the modelling on and what can be calculated? Are there working hypotheses and which criteria are used for the modelling?

I work with CAPRI, an open source model developed on behalf of the European Commission. CAPRI stands for "Common Agricultural Policy Regionalised Impact Modelling System". The model is designed to predict the effects of agricultural and trade policies on production, income, markets, trade and the environment on a regional to global scale. Many scientists from all over Europe are working with it to analyse a wide range of research questions. For example, how a reform of agricultural policy affects the decisions of farmers.
The detailed input data comes, for example, from Eurostat, the statistical office of the European Union) or the FAO (Food and Agriculture Oganization of the United Nations). The model takes into account not only production balances, but also environmental indicators, such as the amount of greenhouse gas emission equivalents generated by beef production. So I do not need to collect my own data.
For my work, I use a scenario for 2030 as a reference or baseline. We postulate the establishment of a new equilibrium resulting from the new EU policy framework that is currently introduced. The scenario for 2030 serves as the status quo for my calculations. In the first part of my project, I compare this with an extreme scenario: No Trade. I simulate what would happen if there were no more agricultural imports by massively increasing import duties. With 400% import duties on soya, it would be four times more expensive and therefore no longer imported. This brings me to a No Trade extreme scenario. For the balance to be restored, something has to change elsewhere. The model finds the best economic way for farmers to compensate for the lack of imports – by increasing soya cultivation in the EU or by importing other protein feeds. Another consequence is a decline in animal products due to this tariff increase. The great advantage of the model is that it takes all known relationships into account and weighs them against each other, and displays the resulting consequences in the modelled system. It would be impossible for an individual person to have a complete overview of this.
However, it is not just one agricultural product – soya in this case – that is to be excluded from trade. I am now facing the challenge of defining the key elements of the No Trade scenario. I am concentrating on food. This challenge is very time-consuming, because the CAPRI model is not really designed for extreme scenarios. Smaller tariff changes of 10 to 20% or import stops of products that are mainly produced within the EU can be represented well with the equations. However, the food4future scenarios I have to approach step by step. At some point I reach the point where the system of equations no longer works because, for example, the demand function cannot be solved. Regardless of price, there is always a "minimum demand". So I am currently working on the solution to represent the demand in a way that allows the model to solve and reflect the extreme scenario.

How do you approach your project and are there first results?

In my project, there is little that I deal with as intensively as with modelling. This is probably both an advantage and a disadvantage of the project. CAPRI has been in existence for many years now and throughout Europe various institutes are working on it with different focuses. The model is constantly being improved and expanded, so that today there is hardly anyone who has an overview of the entire model down to the last detail. So it can happen that no one before me has worked on a particular issue and I am lucky if I get information from others about which part of the code might be relevant. I then sit in front of the computer, click through the code and try to figure out why the error messages appear. I like to compare my work to an escape room: I solve a puzzle and get one step closer to the goal while discovering a new problem at the same time. The goal, analogous to the escape from the room, is the successful modelling of the No Trade scenario. Afterwards the results have to be evaluated. And again, as many questions one can ask the model, as many answers one get. What I personally find very interesting is the question of the scenario's impact on the carbon footprint. In the public debate, regionality is often associated with a positive impact on the climate. In a limited scenario we have stopped imports of soya and cereals. This would result in more soybean being grown in the EU, with lower yields and higher land requirements than in South America. Other agricultural commodities, which actually find optimal growing conditions in the EU, would therefore have to be imported to meet demands. Overall, the model has predicted a significant increase in greenhouse gas emissions as a result. In principle, the result is in line with our working hypothesis, as the soya issue and protein gap have been the subject of a long-standing expert discussion.
What surprised us, however, was the result of the simulations with regards to biofuel with dried distiller's grains [with solubles] (DDGS), which are a waste product of bioethanol production and can be used as animal feed. The simulated import ban on protein animal feed showed that the two are coupled: More bioethanol would be produced so that more feed would be available.

The extreme scenarios - No Land, No Trade - sound exactly like that: extreme. Which conceivable developments could lead to their occurrence?

In purely technical terms, the reasons for the development towards the extreme scenarios are not so decisive for my modelling. For implementation in the model, it does not matter whether the import stop is caused by tariff increases, non-tariff trade barriers or sharp increases in transport costs. 
In my view, there is currently more evidence for a No Trade scenario than for No Land. We are in a global pandemic. Many people are asking themselves to what extent [the EU or Germany] is willing or obliged to depend on global supply chains or to become more self-sufficient.
Another current example is the use of the new CRISPR/Cas9 method, genome editing on crops. The legislation on genetically modified organisms in the EU and in the US differs fundamentally. In the EU it is process-oriented, which means that e.g. soya plants that have been bred using conventional genetic engineering or new genetic engineering methods are always classified as genetically modified. Even if only one letter in the genetic material has been altered, no foreign genes have been introduced and the end product is not distinguishable from soya generated by traditional breeding methods. In the USA, on the other hand, there is product-oriented legislation, so that some varieties produced with the help of "gene scissors" are classified as conventional, because their genetic material is not distinguishable from traditionally generated varieties. However, under EU legislation they would be classified as genetically modified organisms. So, if a trader wants to import non-genetically engineered, but conventionally produced soy from the US into the EU, they cannot know for sure that it could be genetically modified soy according to EU law. If companies do not want to take this risk, it could effectively lead to an import stop.

What benefit can we draw from modelling extreme scenarios and their effects on the agricultural and food economy?

I think it can already be deduced from the No Trade modelling that an isolationist policy, be it of the EU or an individual nation, without simultaneously changing consumption will have negative effects on the climate, above all, through the increase in greenhouse gas emissions. In other words: if isolation were to be politically desired at some point, without further intervention it would have negative effects on the climate footprint. The introduction of a global trade regime, e.g. by introducing a CO2 tax and/or a CO2 cap, and a change in consumption towards a plant-based diet could counteract the negative climatic consequences.

What led you to specialise in agricultural economics?

During my economics bachelor's programme, I was particularly interested in resource issues, which brought me to the Faculty of Agricultural and Horticultural Sciences. There one deals with the question of how to use existing resources most efficiently, whereas at my bachelor’s programme in economics did not really cover that. I was also very interested in development policy and how economic development actually takes place. In so-called "developing countries", the agricultural sector accounts for a much larger share of GDP and employment than in Germany. When one talks about economic development, one must inevitably talk about the economic development of the agricultural sector. In development policy, there is the idea of strengthening economic growth in "developing countries" with a strong agricultural sector through sales markets in the global North, which is also an interesting combination.
Furthermore, agriculture is exciting because – in contrast to many other sectors of the economy – trade barriers are still very present and their changes have a visible impact on trade. So after I completed my Master's degree at the Faculty of Agricultural and Horticultural Sciences at Humboldt University, I quickly realised that I had not yet finished with the topic and that I wanted to stay with it.
And then there are always moments during my work when things make sense. When, after two or three weeks of sitting over my results and not being able to do anything with them, and wanting to throw my laptop out the window, the moment of realisation suddenly comes, that is so rewarding. One then walks through the world with open eyes.

Thank you very much for the interview.

_{The interview was conducted by J. Vogt, IGZ}

Ferike Thom presented her project "The EU as subsistence" at the GFFA Science Slam in Berlin in January this year. Here you can watch the recording: