Photo: Pere Virgili
To prevent wars, it is essential that we stress on the prior latent causes of latent conflicts. The use of modern technologies and the analysis of big data are the basis of innovative empirical research into these causes.
A black and white map of ethnic groups in Africa which reflects the continent’s divisions prior to the colonisation, covers the largest wall of her office, since it was through observing the relationship between poverty and wars in Africa that the economist Marta Reynal-Querol (Barcelona, 1973) decided to focus her research on the prevention of conflicts. “Conflicts are still the main problem in many countries and the international efforts made to resolve them have been somewhat unsuccessful. And the saddest part is that, among other ‘illnesses’ of the world, the one which represents conflicts is one of the few that depends on the action of man, but we seem unable to stop it”, she says.
Doctor by the London School of Economics, Reynal-Querol was an economist at the World Bank, and is an ICREA researcher and lecturer at the Barcelona GSE and the Pompeu Fabra University, where she has her office, on the Ciutadella campus; quite a coincidence that it happens to be in a building that was built to serve as military barracks in the mid-19th century. Reynal-Querol explains how they research information in history, and how they trust new technologies and the study of big data to prevent the human tragedy made up by inequality, social protest, terrorism and wars, which according to her are the largest and most serious problems of today’s world.
Avoiding wars wouldn’t seem to be a frequent concern for economists.
Conflicts are as old as the origin of humanity. Philosophers and writers, since ancient Greece to today, have thought and debated about the issue. Until recent years, many economists described empirical research on conflicts as “extravagant” and “extremely risky”. They said that was nothing to do with economy! Theories can be put together, but they then have to be compared with reality, and to do so we need data. I have applied the tools we have in economy to issues that were dealt with in other disciplines, such as political sciences for example. The debates were never-ending, but always based on narrative arguments, never based on scientific data or analysis. We economists are lucky in so far as all the problems in the world can be seen from an economic approach.
Where does this interest come from?
When I was studying in London, I was soon attracted towards economic development, but in economic literature referring to Africa we would always see unexplainable phenomena. Every investment in roads, bridges, hospitals…, was destroyed by the following war. Or we would see that after a war, no individuals would carry out economic transactions with the ethnic or religious groups against which they had fought. That was when I realised that we would never understand why these countries failed to develop without previously understanding the reasons behind so many armed conflicts. When I was working on my thesis, there was only one economist working on empirical conflict-related issues. That was the senior research of the World Bank, Paul Collier, who is now a professor at Oxford. I worked with his team at the World Bank, in Washington DC. That was one of the most enriching periods of my professional career, a perfect combination between research and discussion with people responsible for policies. During this time, economists were beginning to understand the importance of empirical conflict-related issues.
What are the necessary data for understanding and preventing conflicts?
For many years now, countries have been gathering data such as the GDP per capita. And in wars we quantify the number of dead, using international journalists and observers as our source. Big data has an enormous potential and we will need to change what we have been doing up until now… We are still debating on how to carry it out.
But, once you have scientifically defined the conflicts, will it actually be possible to avoid wars and their victims?
Those responsible for policies in the United Nations and other international organisations often ask themselves which will be the next country at war and when will it occur. The answer is always the same: it is possible to say which countries have more probabilities of undergoing conflict, but it is impossible to predict when they will suffer this unexpected clash that will turn latent conflict into widespread violence. Policies cannot be defined based on the clashes that detonate violence, as actions will arrive too late or will be devoid of any possible success. To prevent war, it is essential that we stress on the prior latent causes of the conflicts.
And what are these factors, how can they be measured?
For example, there used to be theories that indicated that the more ethnic groups there were in a country, the more possibilities there were of violent conflict breaking out. Several empirical works contradict this. The probabilities increase when there are fewer groups, but larger in number of followers. The regions with social polarisation are the ones with the highest risk of social unrest. Discussions have now steered away from the theories, many of which are apparently reasonable, and are more focused on their consistency with existing data. Many empirical results have broken down schemes that were well-established, but which had little empirical confirmation.
For example?
For many years the predominant idea has been that poverty causes wars. International organisations used to advocate that in order to fight against war it was essential to fight against poverty, implementing well-known measures: this was about granting international aid. But it has been proven that it doesn’t work like that; resolving poverty doesn’t resolve conflict… International aid often makes things worse! It is one thing to determine that poor countries suffer wars and another quite different is to maintain that these wars are caused by poverty. It may be that a country is the victim of a historical inheritance that now generates wars and poverty, or that it suffers due to the actions carried out by a despot leader who applies bad policies for growth and social coexistence, thus generating poverty and war. But the trigger for these wars are economic clashes such as those caused by a fall in the price of the raw material that sustains the country’s economy, or due to a recession…
Photo: Pere Virgili
Once the empirical results have been obtained, what is the next step? Who can prevent the wars?
It would be great if it were that easy! We then implement scientific methodology to study what makes a country suffer more wars. A large proportion of the causes are hard to correct, such as the presence of inappropriate leaders that make the wrong decisions; which, for example, leads us to research why, and in which contexts, unsuitable people are often voted into power. Could this be due to education? Maybe a lack of information, or non-biased information? Hence, we study the behaviour of the people, the role played by the media, how internet access affects them… At present, thanks to big data, we can analyse many more things that we could before.
And in the meantime, there are countries where people are still killing each other…
Yes, in the same way as there are still people dying from diseases for which we have no cure, and that is why we continue researching. We do exactly the same for conflicts and development. We are testing methods that are very new, such as automatic learning, to predict the places where violence may already be imminent, and we collaborate with specialists in data science. There are very few interdisciplinary teams, given that interdisciplinary work is very complicated; but, if we don’t come together, we will never come up with any new solutions. We must innovate and seek solutions that work. For their part, those who have the technology must become more sensitive towards social issues: the social return will be immense.
We talk a lot about countries, but not all conflicts can be explained using borders.
At present, access to the data from very small areas obtained from space enable us to analyse at regional and urban levels. For example, NASA provides information on the density of light during the night in areas as small as one square kilometre. This enables us to make very good approximations, especially in undeveloped areas, for which no official data is available. With help from the supercomputer MareNostrum and technical engineers, for the first time we will be able to obtain detailed and specific data regarding poverty and inequality, regarding the most disadvantaged areas… This will enable us to see whether the conflicts develop with the same dynamics in urban contexts as well as in rural areas, or whether the distribution of groups into neighbourhoods can affect different aspects of urban life… Thanks to this kind of big data, studies have been started up in relation to crime levels in cities. But these are early days.
And then, will we also be able to analyse and act differently in cities?
Bid data allows to perform income distribution measurements that did not exist up until now. Through the use of satellites, we can see the income inequalities in the different neighbourhoods. A small inequality is said to be good as it encourages improvement, but too much inequality proves discouraging. These theories have never been proven as we didn’t have the exact data regarding the distribution of people. But, for example, we can now see the exact moment or social situation as of which conflict begins. Understanding the processes of the conflict allows to predict or prevent it. An example learned from Africa is that if a country doesn’t work because there is a conflict, building roads is pointless, as they will be destroyed; it is equivalent to throwing money away. In some African countries attempts were made to fight unrest between ethnic groups by means of television series showing marriages between people from different ethnic origins, with the aim of promoting and standardising integration. These are political science problems approached using different tools.
The use of data is infinite.
There is great potential. It’s about taking the optimisation criteria and mechanisms used in companies and applying them to development projects. When a natural disaster occurs in a country, we should be able to rebuild it in an optimal manner that will favour subsequent development, but all too often this doesn’t happen. To prove it, we would like to calculate how far away each country is from its optimal infrastructure, verifying the kilometres of roads built and the level of connection infrastructures. This will show that Germany is much closer to its optimal level than Spain. Another example could be seen in the relationship there is between the level of connectivity of the infrastructures and the corruption in a country. Applying the science of data to economic development is exactly that.
What problems do you think may be resolved soon thanks to all this information?
One of the issues we are working on is the internet’s true capacity for reducing inequalities. There are areas with educational inequalities which become polarised when internet is introduced: some people make use of the access to the net for obtaining better information, while others only use its entertainment functions. We now focus the research on finding out the levels of education as of which the polarisation is emphasised. Everyone knows that education is essential but will now be able to know when and where it needs improvement, if we can prove that internet, without the educational element, can lead to social isolation.
Another of your lines of research is the relationship between the education of the leaders and the economic development of a country.
The idea that leadership affects countries has been around for a long time, but researchers have spent little time analysing the personality of the leaders. What differentiates a good leader from a bad one? Do their level of education and social background bear influence? These questions are essential when understanding why some countries sometimes adopt mistaken policies. If companies demand impeccable curricula when employing their staff, why doesn’t the same apply to political leaders? To begin the research, we required a database of leaders that didn’t exist, so we went about searching for biographies of all the political leaders of the world from 1875 until 2004 and we compiled a database with all the available information on their education, environment, family and personality. The first thing we found was that those with a higher level of education made decisions that led to increased growth in their country.
How did you prove this scientifically?
In order to establish a causal link and not only correlation, we must seek or create experiments, as carried out in medicine to prove whether a drug works or not. In wars it is very hard or impossible to generate random experiments and what we do is to seek what we refer to as “natural experiments”. In the case of leaders, we seek unexpected and external changes (such as those caused by accidental deaths”, not caused by poor management, and we analyse the situation of the country before and after this change. We ignore the exact mechanics of the relationship between the economic development of the country and the level of education of its leader, but it is most probably related to the fact that a better educated leader tends to surround him/herself with better people and, therefore, their decisions are more appropriate.
One of the historic moments on which we focus now is that of the colonisation of the American continent, which provides a unique opportunity to advance on the debate regarding the role of human capital as a factor for economic and institutional development. In less than 50 years, the colonisers founded cities inland, spreading over a very large and diverse area, ranging between northern Mexico and Buenos Aires. The key factor is the period immediately after the initial colonisation (1492-1540), when the settlers established themselves in the territory in a manner completely by chance. They had no idea where they were going. And that is a perfect example of natural experiment. These first settlers, merchants or breeders knew nothing about the lands they conquered. When comparing the settlements, we can see if the ones where better educated people arrived to worked better than others. And the answer is yes, they did.
And what should we do with this information? Should we choose our leaders by reading their curricula? Could this lead us to a new era of enlightened despotism?
Each profession has its requirements, and leading a country is no different. In a company you are required to have experience, training. This is not antidemocratic at all. Leading a country requires specific training. We could define what the task of a leader consists of and, based on this, the training that is required.
What can we do to improve the situation of the leaders in our own country?
The level is very low, possibly due to how one manages to become a candidate, and this may be discouraging for well-educate people. Leaders come from pools for which no selection process has been carried out, and this should be performed by the parties. This is the problem inherent to not having an open list system.
In Catalonia there is a situation of social unrest. How would you analyse it?
There is obviously social unrest in Catalonia. But I prefer to leave this analysis to international experts who don’t have a biased view as one may have from the inside. And a fact that surprises my friends from the outside is that here there is no specific violence among groups.
And why do you think there is no violence?
We are unaware of the triggering factor that leads people to kill each other. It may be that here we have a much more educated population, one that understands it must contain itself. The unrest has not been resolved, it is latent, but violence has been avoided. What is needed are political measures that keep the different groups happy; that is exactly what negotiating is about.
Are there societies which are more susceptible to conflict than others?
In Africa there is a persistent war culture: the areas where there were wars before the arrival of the Europeans are still areas in which the tendency is to resolve unrest through violence. We now carry out studies on this exact phenomenon in Latin America, to see how violence was tolerated in the pre-colonial societies: in some of these, the conflicts were resolved through negotiation; in others, they were resolved using violence. By means of persistence studies, we want to establish what values are transferred from fathers to sons and how education can help to straighten out the current situation.