Introduction

What are the insights of this data story?

       Since the Neolithic Revolution, agriculture feeds the world’s population. Even today, the agricultural sector employs a large part of humanity and is a major part of the economy of developed and developing countries. In the history of mankind, the economic growth of countries has almost always been accompanied by the development of agriculture and the agricultural economy, i.e. production, imports and exports. In this data story we want to explore different links between the agriculture and the economic success of countries. Which crops are produced, and which animals are traded by richer countries? Do they tend to be net exporters or importers of food? Do they tend to be self-sufficient? We try to answer all of these questions and show many interesting insights along the way.

Which important events had a significant influence on the agriculture and the economy for the historical period from 1970 to 2015?

       To be able to properly understand the studied data, we must first look at the historical context of the period we are analyzing. There was the Cold war from 1945 to 1990 between the two economic superpowers (USA and USSR). The USSR was then dissolved in 1991. The Japanese economic miracle occurred from 1945 to 1990 and allowed Japan to come out of the disastrous state in which it was at the exit of WW2 and become one of the world’s largest economies. There have been 2 big oil crises, in 1973 and 1979. There have also been many wars (Middle East wars 1973-2000 e.g. Yom Kippur War 1973, Islamic Revolution in Iran 1979, Iran–Iraq war 1980-1988, Gulf war 1990-1991, Yugoslav wars 1991-2001…). The third Agricultural Revolution (also known as Green revolution) occurred from 1960 to 1990 and improved agricultural productions thanks to fertilizers and chemicals. These historical events were hugely impactful on countries populations,agricultures and economies. We see this in the example of the GDP trend between the USSR, Russia and Ukraine.

Which countries have the highest GDP? How did GDP evolve during the last decades?

       In 1970, the countries with the highest GDP were United States, Germany, Japan, France, United Kingdom and Italy. Japan overtook Germany in 1971. Around the 90s, Italy and UK were fighting for 5th place in the ranking. From 1996 to 2006 the UK surpassed France. China entered the top 6 in 1999 in place of Italy and got 2nd in ranking in 2009. On average, GDP continuously increased during this period.

What is the structure of international trade?

       This network shows just how highly correlated GDP from one country to another is. It’s a visualization of the Pearson Correlation between countries GDPs throughout the years of our dataset. Possibly the most obvious observation from this network graph is the clear distinction between former Eastern Bloc countries and the rest of the world. About half of the years of our dataset are during the East-West divide where there were two clear global trade blocks. Nevertheless, GPD increased the same way almost everywhere and the whole world is close to be a single big cluster. This can be explained by the fact that the global economy is highly interlinked, and countries have strong trading relations which leads to the GDP evolving the same way. Global trade and thus interdependence have been continuously increasing throughout the past decades. There are many institutions put into place for this to be possible. Today, almost all countries worldwide are part of the World Trade Organization, which makes sure that tariffs and other barriers to trade are reduced. In the past years, we’ve also seen that many countries, especially the United States, are becoming increasingly more skeptical of global trade and even engage in trade wars. Some scholars are even speaking of a trend of deglobalization since this decade. What kind of impacts could this have on the future of international trade?

About the dataset

       The data that we use for analysis comes from the “Global Food & Agriculture Statistic”. It is a dataset by the Food and Agriculture Organization of the United Nations and includes extensive information about many factors related to agriculture, such as production output, food prices, trade balances and environmental impacts among others. It includes country-wise yearly values from 1970 to 2015.

Our approach

       How to measure economic success? The first and most intuitive idea is to use the Gross Domestic Product of countries. The second approach we took was price stability, expressed in low levels of inflation and thus small changes in the Consumer Price Index of food. Can we create a model that will tell us which products to produce and which to trade to maximize for GDP or minimize for inflation? As it turned out, price stability did not yield significant correlations to any production or trade. GDP however did…

Most important agricultural and economical features

Which features are the most influential features on the Gross Domestic Product?

       Using a knowledge-based method as well as a more technical, data analysis algorithm, we were able to select specific features. However, as we expect the most influential goods to not only have high correlation with GDP, but to also be globally relevant in production volume, we narrowed down our possible choice of features. With all of this, our model managed to pick out some features that could optimize GDP.

Below is the full list of the selected features, sorted by order of importance attributed by the model:

       The first thing we notice right away is that the selected features are all very common to our own eating habits. They form integral parts of Western diets that are familiar to us. Many of them are also linked to animal agriculture. Take soybeans for example: this crop is one of the most produced worldwide. The main exporters are the US, Brazil and Argentina and the main importer is China. Around 330 million tonnes of soybeans were produced in 2018. It is mostly used to feed animals. What is striking about the result is not only which features we see, but which we don’t see. Many of the most commonly eaten foods worldwide, such as rice or fish are not present on this list. Of course, the way our model selected for features, this shouldn’t come as a surprise. The United States and Western Europe have some of the highest GDPs in the world, so its natural that these diets will be more represented. Additionally, since producing animal products needs a lot more ressources and is more expensive than plant based foods, the link with GDP is intuitive.

Now that we’ve established which agricultural production and trade features are best at predicting a high GDP, we want to look further into which countries produce and trade them and try to derive further conclusions from that.

Global production

Which countries produce the features of interest?

X
X
X
X
X
X
X
X
X

       Globally, all features that we identified seem to follow a common trend. The biggest producers are in almost all categories China, the USA and Brazil. Russia, France, Spain, Germany seem to get closer to their volumes. The rest of Europe is situated in the second or third tier of producers as well as the other developed countries (Australia, Canada, south American countries). In most categories, north and south African countries are situated in the same orders of magnitude as developed countries. A general trend to observe is that central African countries seem to produce very few resources in comparison to the rest of the world. The only feature going against this general description seems to be cattle. The production of cattle appears to be very well distributed throughout the world. The production of pigs also seems well distributed throughout the world except for countries excluding pork from their diet for religious purposes. The phenomenon might be explained by the increased difficulty in transporting these goods. Live animals and meat transport are much more complicated than grain, vegetables or forage. Meaning that the countries would generally produce what they need in terms of meat and rather import crops or vegetables.

Are big producers wealthier?

       The general trend we observe is that developed countries are bigger producers. This makes sense considering how we selected these features. Our regression model gives us the features that relate to a high GDP thus the features selected will be markers of rich countries.

Is the production of important items stable?

       What can be found surprising is the stability of the producer ranking throughout the years. The top producers 50 years ago are still the top producers nowadays. It is also worth noting that globally, even though the ranking does not change much, the sheer production volume follows a growing trend. As an example, we want to look at wheat production among the top 10 producers worldwide, as an average of the years studied. Throughout the years, with the population growing, the production is also growing steadily.

Global Trade Balances

       In this part, we want to pick out a few of the identified features and look at where in the world they are most exported from and imported to.

What is green maize?

       The feature that our model puts out as the highest weighed for predicting GDP is green maize. What is it? It’s maize that is mainly harvested so it can be either directly fed to animals or conserved through “silage” and fed to them in winter. Maize is native to central America and thanks to genetic selection and hybridization it can now be grown everywhere in the world. Among the forage foods that are grown for animal feed, they’re very high energy and easy maintenance, since they only need to be harvested once. They also require high fertilizer, herbicide and pesticide levels. Especially with the commercial adoption of genetically modified maize that is herbicide and pest resistant during the 90s, it has emerged as a very important animal feed.

Which countries are net exporters of green maize?

       We see that until the 80s no country in the world was a strong net exporter of green maize. It was barely even exported or imported in any country. Starting from the mid-80s the United States emerge as the main net exporter. During the 90s and into recent years, some more countries emerge that are net exporters, mainly in Europe. Globally, the United States stays the single most important exporter of this agricultural product throughout the years, which is probably why our model identified it as the strongest predictor of GDP.

Which countries are net importers of tomatoes?

       Being a tomato importer, but also a tomato producer seems to be a good predictor of wealth for the state. Tomatoes are one of the most important vegetable crops produced worldwide, only potatoes have a higher production. Looking through the timeline, we clearly see that almost all highly industrialized nations are continuously importing tomatoes. Especially the United States, France and Germany are importing large quantities each year. It’s not a big surprise, since tomatoes are a popular ingredient in many modern western recipes but require a climate that is not well suited for most of the United States or Northern Europe.

What can we say about the other trade features of interest?

       Just as we’ve seen for the production, the trade balances of our features also follow some broader trends. To start off, without even looking at the maps yet, but at the identified trade features, we see that many of them are directly or indirectly related to animal products. Imports of live pigs, turkeys and cattle are some of the highest on the list. Other than that, as we’ve already established, green maize as well as soybeans are mainly grown for animal feed. In addition to those, even for oats, the primary use was for livestock feeding (70% of global production in 2009, according to the FAO). What about the others? Even though we might not think about these crops primarily as feeding stuff, regular maize and wheat are also significantly used for this purpose. They are indeed both major staple foods for people, but especially in developed countries, they are also a major feedstuff. According to the FAO, in 2007 around 16% of global wheat production was used to feed animals, whereas this number was 42% for the EU.

Sheep on ships?

       Even though many animal rights activists are strongly campaigning against it, millions of live animals are being transported in gruesome conditions around the world every year. Their destination is usually the slaughterhouse. When countries income level rises, the demand for meat increases with it. Meat consumption is regarded as a symbol of wealth across cultures. The economic benefits of transporting the animals alive is twofold: on one hand, the costs are reduced, since the it is more expensive to keep the meat frozen during transport and on the other hand, consumers usually prefer their meat as fresh as possible. The live trade industry is booming so it’s no surprise that many of the strongest features identified by our model come from it.

Who imports and exports them?

X
X
X
X
X
X
X
X
X

       What other trends can we notice from the trade balances? At first sight, it might seem that for most of the features, some of the biggest global economies have inconsistent trade balances and thus it is surprising we can find correlations at all. Let’s take pigs as an example: in 1998 France was one of the most important exporters of live pigs, whereas in 1999 it was suddenly one of the most important importers. From 2007 to 2008, we see an equally big shift for Spain. Nevertheless, this leads to an interesting observation: even though developed countries are not being consistent in the directions of their trade balance, they still tend to stay among the most trading countries by volume.

Self-sufficiency

       We identified agricultural features as markers of economical growth, we identified the producers as well as the traders for these features. We now want to look which countries are the most indenpendant regarding these very features. To do so, we calculated the self-sufficiency score defined by the fraction of its consumption that each country is able to produce.

How is the self-sufficiency score distributed for the selected features?

       Unlike the production ranking, the self-sufficiency score seems to be rather unstable. It is interesting to note that the countries which were by far top producers do no stand out early on. For instance, China does not appear in the highest scores until 1990. The most surprising results came from African countries such as South Africa, the United Republic of Tanzania, Nigeria and the Ivory Coast that constantly have a score competing with the richest countries which was unexpected. However this might be explained by the presence of green maize within the calculation of this score and if you recall the producers map presented previously, all these countries were producing pretty large amount of green maize and exporting most of them which would considerably impact their scores. Even though the general trend tends to show that richer countries are more independent, there are outliers and the score seems to have a pretty high variance as well as being fairly evenly distributed.
We also attempted to correlate the self sufficiency score with both the gross domestic product (GDP) and consummer price index (CPI), however we did not find any relevant results. The features that we selected did not include only production data but also trading data which might bias the way we score each countries, in particular import quantities. It would make sense for rich countries to be able to import bigger quantities therefore lowering their scores pretty significantly.

Conclusion

       China, USA, Brazil. These are the countries that feed the world. They are the biggest producers in almost all categories.

       We observe a strong correlation between wealth and production in general. There is probably also a mutual causality between the two observations, meaning that the more you produce the wealthier you are and the wealthier you are then the more you can produce. But more interestingly, we also observe that the amount of agricultural trade a country engages in has an even higher positive impact on their GDP. Most of the features selected are about trade: whether it’s about exporting maize or importing tomatoes, being an important part of the global interconnected economy turns out to be positive for economic growth. Considering recent trends by especially the US to become more skeptical of global trade, it could be relevant to keep in mind what consequences might come from this.

       Another important conclusion we make is the prominent role of animal agriculture. The highest valued feature is green maize, which is only used to feed animals. Many other features are live animal imports and or otherwise linked to crops, which are to a significant extent raised for animal food as well. It takes a lot more land, energy and resources to produce food from animals, so it makes sense that mostly richer countries can afford having a diet with many animal products.