“Targeted” Data Retention: our map explained
This online map demonstrates that according to publicly available data, the allegedly geographically “targeted” data retention proposed by a Belgian draft law will in truth cover the entire national territory and the entire population. This is the first time, citizens and Members of Parliament get a rough impression of the law’s impact, since the government does not provide the data necessary. Please find a detailed explanation below the map. (And read more in our press release.)
Red areas are zones of mass surveillance
The Belgian government plans data retention on five levels (Art. 126/1 § 3 from page 330). Our map shows, colored red, the areas affected by surveillance only on the first of these five levels, the so-called geographically “targeted” data retention. In the areas colored red, all connection and location data of all citizens would be retained. With each of the other four levels explained below, more red-colored areas would be added. In practice, the surveillance pressure against the population is even higher than shown on our map.
A new generation of mass surveillance laws
In combination of all levels, the Belgian Government’s plans undermine the essence of the rulings of the EU Court of Justice. The latter stipulates that surveillance is at most possible in temporary, objectively justified exceptional cases and when really necessary. The plans of the Belgian government, on the other hand, are aimed at permanent and comprehensive data retention. This is why this model must not become a blueprint for the EU and its Member States.
Belgian data retention model violates citizens’ rights
The European Court of Justice ruled that general and indiscriminate retention of information on every person’s calls and movements violates the fundamental right to respect for privacy. In a legal opinion published in April, former EU judge Prof. Dr. iur. Vilenas Vadapalas states that targeting all areas with above-average crime rates would not be compatible with the values and case-law available.
Low crime areas covered
According to our calculations, the Belgian average crime rate is of 11 serious offences per 1,000 inhabitants per three years. Compared to the national average, the proposed thresholds (3, 5 and 7) enabling data retention are thus far below the national average, and the proposal would hence even covers low-crime areas, in contradiction with the former ECJ judge’s opinion.
The Belgian Data Retention Model: surveillance on five levels
Level 1: As shown in the map in red is geographically “targeted” data retention in areas with an average of at least three crimes per 1,000 inhabitants over the past three years. The decisive factors are executed and attempted criminal offenses under Article 90ter §§ 2 to 4 of the Belgian Code of Criminal Procedure. The threshold of three offenses per 1,000 inhabitants is set by the government.
Note: The following other four levels are not shown in the map
Level 2: General data retention throughout the national territory for 12 months, starting at level three of the five-level national threat level.
Level 3: Data retention for particularly vulnerable areas for 12 months. This applies to places frequented by everyone: Ports, metros, airports, critical infrastructure communities, etc.
Level 4: Data retention for 12 months in zones with a potential threat to the interests of the country or the population: buildings of economic or scientific scientific importance, highways, public parking lots, city halls, the National Bank of Belgium, etc.
Level 5: Data retention for 12 months in areas with a potential threat to international institutions and includes EU, NATO, UN and other buildings, among other areas.
How we accessed and compiled the data
For this map, we used only public sources, geo data from taxation authority (provided by geo.be), population data from statistics body and over 1.000 PDFs with crime data from the federal police.
We then wrote our own software to scan the PDFs and to free the crime figures and make them accessible for machines. We also generated a list of over 1.000 different sorts of crime contained in the police data and tried to determine, which of them would trigger data retention following the Belgian governments data retention draft.
Putting these data together, we got a list of 13 Belgian judicial provinces (arrondissements judicaires) with the number of crimes and population in that area, which in the end made the map we present here.
We computed for each province/zone:
1000 x number of crimes (according to art. 90ter §§ 2-4) / population
(The number of crimes was the average yearly value from years 2018-2020.)
If the result number for a province goes over 3, the area becomes red. All areas are far above this level. For 2021 no complete data was provided in the reports we got. All data we digged will be available to the public on codeberg.org where you can also read the whole story about the adventures of freeing Belgian crime data.