A new model to monitor food insecurity


Food insecurity affects almost 33% of the world’s population.

Contrary to previous assumptions, a new study forecasts that this percentage will fall by around four percentage points by 2030. But even in this scenario, hunger will remain a problem for about 2.5 billion people worldwide. An interactive visualization of the data is available online at

In a new study, Jesús Crespo Cuaresma, professor of economics at WU (Vienna University of Economics and Business), and his coauthors from Harvard University, the World Data Lab, and the International Institute for Applied Systems Analysis present a statistical model to forecast the future development of global food insecurity.

Following the recent uptick in hunger indicators, current estimates indicate that food insecurity today affects 32.6% of the world population. The researchers predict that this global figure will decrease throughout the 2020s, largely because of economic growth in developing countries, and drop to 28.4% by 2030.

The estimates in the study indicate that currently 30 persons escape food insecurity per minute, but this positive development will start to slow down by the middle of this decade: The researchers forecast that the numbers of people escaping food insecurity will become smaller again by that time.

When looking at how the amount of people suffering from moderate and severe food insecurity is likely to develop in the future,  the researchers found stark differences across the countries and subnational regions of the world. While the overall number of people affected by moderate and severe food insecurity is projected to decrease at the global level, the predictions forecast an increase of these figures in lower income countries.

New results revise earlier predictions

In part, these new results contradict some of the predictions by international organizations such as the United Nations’ Food and Agriculture Organization (FAO), which predicts that undernourishment rates will go up. The novel methods used in this new study, in contrast, project a decrease of the population experiencing severe food insecurity over the same period.

On the one hand, the projections regarding the drivers of food security that are presented in the study give cause for optimism. On the other hand, however, the forecasts also indicate that a total of around 2.5 billion people will still be affected by severe or moderate food insecurity by 2030. This means that we probably won’t achieve the goal of ending hunger around the world by 2030.

New prediction methods

The novel prediction model is based on climatic and socioeconomic trends and employs machine learning tools to forecast the development of food insecurity for countries and subnational regions up to the year 2030.

The statistical analysis conducted in the study reveals that trends in poverty rates, income per capita, rates of stunting, and income inequality indicators can be employed to provide precise, reliable forecasts of food insecurity levels.

The study

Matthew Cooper, Benjamin Müller, Carlo Cafiero, Juan Carlos Laso Bayas, Jesús Crespo Cuaresma, Homi Kharas: “Monitoring and projecting global hunger: Are we on track?” – available at

The study was published in the journal Global Food Security.

Back to overview