Overall progress towards sustainable development in Colombia masks a story of inequality: poverty rates can vary considerably even within the same community and, when data are not representative of specific communities, statisticians and policy makers can overlook people who need help. When Colombia’s national statistical office (DANE) joined forces with PARIS21 and World Data Lab to use satellite imagery alongside traditional data collection methods, a deeper, more detailed picture of Colombia began to emerge: one that can show policy makers who they should target in development policy making.
Inequality hides poverty and disadvantage in Colombia
Colombia has been one of the countries worst affected by COVID-191 2. In 2020, the economy contracted by 6.8%3, 2.4 million people lost their jobs4 and four in ten Colombians are now living below the poverty line5. Poverty and socioeconomic factors are strongly correlated to higher mortality rates from COVID-19, and so understanding who is most vulnerable and why is paramount to tackling the crisis.
Poverty affects communities very differently. Colombians describe their country as being “many Colombias”, due to the multiple socio-economic realities that can be found within one same geographic area. Despite generally strong economic growth and poverty reduction in the years prior to the COVID-19 crisis, the country’s inequality remained high: in 2019 the average income for people in the highest-earning decile was over 36 times that of the lowest-earning decile and the difference between the rate of poor people in the richest region (Cundinamarca) was 48 percentage points less than that of the poorest region (El Choco).
As stated in the National Development Plan 2018-2022, eradicating poverty with an inclusive approach is one of the three main objectives of Colombia in order to ensure the country’s development is built on sustainable foundations6. This is also a priority for DANE, Colombia’s national statistical office. They note that leaving no one behind in statistics is at the centre of their work, and they are looking to increase the amount of data they have on vulnerable communities in order to guide policy makers’ interventions.
Traditional data collection risks leaving out many groups
Up until recently, DANE had been reliant on traditional data collection methods, such as censuses or surveys. In Colombia, as in many other countries, poverty and wealth can sometimes exist side-by-side within the same region, city or district. When the data is not sufficiently granular or disaggregated, pockets of poverty are invisible. Ethnic minority groups, women and girls, displaced people and disabled people are particularly at risk of being left out of the picture, and indigenous households have the highest poverty rates. Income inequality also remains high in Colombia, with an average wage gap of 2.3 minimum wages between women and men.
How new technology is helping to find people left behind
Recent technological breakthroughs have allowed DANE to start filling in some of their data gaps and identify people at risk of being left behind despite an aggregate picture that shows the country making progress overall towards its development goals.
DANE has partnered with PARIS21 since 2015 to incorporate innovation into national statistical systems. One such project, between DANE, PARIS21 and World Data Lab, is using satellite imagery to give a far more detailed picture of poverty, combined with traditional data modelling methods to ensure quality and accuracy. The project also uses remotely sensed data, such as satellite imagery or nighttime lighting to show granular information on the size of agglomerations. The project can help in several ways, for example by providing more timely data. Prior to the project, DANE relied to a large extent on census data, but this creates a time lag: looking backwards instead of showing the real situation.
These new techniques have given a much clearer picture of Colombia. In the Bogotá D.C. municipality, for instance, census data suggested an average population density of 46 persons living in a range of 100×100 meters. However, the new method calculates population density by using satellite imagery to show the density of physical structures. The new method revealed differences in population density within the region: notably between rural and urban areas, ranging from one to 999 people per 100×100 meters.
In Medellín, the capital of the Antioquía region, the availability of more granular data revealed that areas of the city had far higher poverty rates than had previously been picked up. Nighttime lighting is closely linked to poverty rates, since poorer areas tend to emit less light at night. A poverty estimate was developed by using daytime satellite imagery to estimate nighttime light intensity and then comparing this to actual nighttime light intensity. Grid level data enabled the poverty estimates to be much more granular. This method saw poverty estimates increase 70 fold, from 1 123 data points to 78 000 data points. This information allows the government to formulate and implement policies that focus on the most vulnerable people.
The SDG framework puts the emphasis on ensuring that everyone benefits from development, and this means measuring progress beyond economic indicators, which do not give a full picture. By putting “leave no one behind” at the core of their work, DANE is gathering the data and statistics that can guide policy makers to ensure that progress is felt by all and is more sustainable. As Colombia begins preparations to host the World Data Forum in 2024, the country is showing how partnering with other organisations, such as PARIS21 and World Data Lab to pilot new technology to obtain more data can be a powerful engine for progress.