Title: Ten years of the Venezuelan crisis - An Internet perspective
Author：
Esteban Carisimo(Northwestern University),
Rashna Kumar(Northwestern University),
Caleb J. Wang(Northwestern University)，
Santiago Klein(Universidad de Buenos Aires)，
Fabián E. Bustamante(Northwestern University)
Scribe: Mingjun Fang
Introduction:
Over the past decade, the crisis in Venezuela has evolved into a series of escalating challenges affecting various aspects of civil society. Without delving into the root causes, Venezuela’s decline can be attributed to the collapse of oil exports, which subsequently triggered a chain reaction in macroeconomic indicators, impacting the general population. The Venezuelan crisis has drawn international attention, and numerous studies have explored its impact on public health, energy, and water management. While experts from various fields have discussed this economic and political crisis from different angles, the authors of this paper consider a previously underexplored perspective—the impact of the Venezuelan crisis on the country’s internet infrastructure. The authors investigate the crisis’s impact on Venezuela’s internet from the perspectives of network infrastructure, inter-domain connectivity performance, and network performance. They employ multiple complementary datasets to capture a panoramic view of the changes in Venezuela’s network over the past decade. To measure changes in network infrastructure, the authors collected data on submarine cable changes from Telegeography’s Submarine Cable Map and changes in root DNS server routing from RIPE Atlas’s built-in measurements. They explored the crisis’s impact on inter-domain connectivity through changes in AS-level connectivity of CANTV-AS8048 and snapshots of CAIDA’s PeeringDB collected since April 2018. They reflected changes in network performance using bandwidth measurement data from Measurement Lab (M-LAB). To contextualize their findings and provide a benchmark, the authors compared their analysis results with other Latin American countries.
Key idea and contribution:
Impact of the Crisis on Network Infrastructure. The authors discuss the impact of the crisis on network infrastructure from five perspectives: peering facilities, submarine cable deployment, IPv6 adoption, the availability of Root DNS infrastructure, and the deployment and adoption of content delivery networks. This section mainly explores the authors’ measurement methods and conclusions, with specific results presented in the experimental evaluation section. The authors analyzed the development of peering facilities in the country by examining monthly snapshots from PeeringDB since April 2018 and compared them with other countries in the region.
Figure 1 displays the evolution of peering facilities within Latin America through four panels. Out of the 600 facilities in the region, only four are located in Venezuela; two registered in 2021 and the other two in 2023. Brazil’s number of facilities grew from 102 in 2018 to 311 in 2024. Other countries also experienced significant growth during this period, including Mexico and Chile. The results indicate a significant lag in Venezuela. The authors used Telegeography’s dataset to investigate the growth of submarine connectivity in Venezuela and Latin America.
Figure 2 shows the growth in the number of cables connected to countries in the Latin American region. The deployment of submarine cables in Latin America can be divided into two phases: pre- and post-internet bubble. From 2000 to 2024, the number of cables in the region significantly increased from 13 to 54. However, Venezuela did not participate in this second wave of expansion, ranking at the bottom in terms of submarine cable deployment during this period. To study the adoption of IPv6 in Venezuela and the Latin American region, the authors utilized Facebook’s public dataset, which reports the percentage of IPv6 requests registered by Facebook (Meta) by country.
Figure 3 shows the evolution of the percentage of IPv6 requests received by Facebook. Significant progress has been made among the major Latin American countries. The growth in IPv6 in Mexico and Brazil has been steady, exceeding approximately 40% in the latest snapshot, while Argentina, Chile, and Colombia hover around 20%. In contrast, Venezuela failed to participate in the region’s IPv6 wave, remaining at nearly zero levels until 2021, and only reaching 1.5% by mid-2023.
The authors analyzed the availability of infrastructure running critical internet services, represented by DNS, in the country. As shown in Figure 4, the local hosting of Root DNS servers within the LACNIC region has significantly increased since measurements began in 2016. During this period, the total number increased from 59 to 138, growing by 2.34 times. In contrast, Venezuela experienced a decline in this area. Initially, there were two Root DNS servers in Caracas, one L (ccs01.l.root-servers.org) and one F (ccs1a.f.root-servers.org), which later disappeared from measurements and were replaced by another L DNS server located in Maracaibo (aa.ve-mai.l.root). The authors also analyzed changes in Venezuela’s network infrastructure from the perspective of content delivery networks. Figure 5 displays four time series representing Google, Akamai, Facebook (Meta), and Netflix, showing the percentage of user groups connected to off-net servers for each country between 2013 and 2021. Over these eight years, on average, Venezuela ranked relatively low among the providers: Google at 19/27 (56.88%), Akamai at 18/22 (35.74%), Facebook (Meta) at 21/25 (28.33%), and Netflix at 23/25 (5.87%).
Impact of the Crisis on Interconnection. The authors first discuss CANTV’s dominance in the telecommunications market, illustrating that it is the main Internet Service Provider in Venezuela. The authors substantiate CANTV’s dominance by comparing the proportion and total amount of address space held by CANTV and its main competitor, Telefonica de Venezuela (also known as Movistar, AS6303), as shown in Figure 6. Turning our attention to IXPs, IXPs provide interconnection hubs that enable direct network traffic exchange, reducing reliance on upstream providers and lowering transmission costs. While IXPs have flourished globally and in Latin America, significantly facilitating local traffic exchange between CDNs and networks, Venezuela remains one of the few countries without any deployed IXPs as of 2024. The authors attribute the lack of IXPs to the dominance of CANTV in the internet ecosystem, coupled with the ongoing economic and political crisis, which leads to a lack of attractiveness for new participants and the abandonment of IXP deployment.
Evaluation：
Impact of the Crisis on Network Performance. The authors investigated the evolution of bandwidth in Venezuela and other Latin American countries. Lagging behind other peer countries in this aspect may indicate that Venezuelans struggle to interact normally with increasingly high-bandwidth online services. The three panels in Figure 7 show the evolution of median download speeds in Venezuela. The top panel provides a comparison of Venezuela with other countries in the LACNIC region (with comparable peers highlighted in vibrant colors) and the aggregated regional average. The lower left panel focuses on Venezuela, displaying only the country’s download speeds, while the lower right panel shows Venezuela’s download speeds standardized relative to the aggregated regional average. While download speeds across the region have been steadily increasing, Venezuela’s download speeds have stagnated below 1 Mbps over the past decade. However, countries like Uruguay, Brazil, Chile, Argentina, and Mexico achieved download speeds of 47.33, 32.44, 25.25, 15.48, and 18.66 Mbps respectively by July 2023, whereas Venezuela’s median speed significantly lagged at only 2.93 Mbps. Venezuela’s median download speed at this time was equivalent to the speeds reached by Uruguay and Mexico in November 2013, Chile in June 2017, Argentina in April 2018, and Brazil in September 2019. This result is significant because it highlights the crucial link between economic and political stability and internet network performance, revealing the impact of the Venezuelan crisis on citizens’ everyday digital interactions from the perspectives of stagnating bandwidth growth and high latency compared to other Latin American countries.
Q & A：
Question: You’ve done a nice work, but I’m a bit curious about how you ensure the quality of the data in your article. In other words, how do you guarantee accuracy?
Answer: That’s a great question. Due to the crowdsourced nature of data sources like SpeedTest, the number of tests per day and across networks can vary by country, which might introduce potential bias. However, we have used multiple data sources and methods to minimize these biases.
Personal thought:
This article reflects the political and economic turmoil Venezuela has experienced over the past decade from a new perspective—the internet. The authors strive to comprehensively investigate and analyze the impact of Venezuela’s political and economic crisis on the daily digital activities and internet infrastructure of its citizens from multiple angles, including network infrastructure, inter-domain connectivity, and changes in network performance. The extensive data collection and statistical analysis performed in this article to derive meaningful results left a deep impression on me. However, I believe the coverage of the platforms used in this study is limited, and the test results may be biased due to geographical proximity. As the authors stated, “Although these platforms are relatively evenly distributed within the region, some countries do not have local servers.” In addition, I think it would be beneficial to conduct random sampling in different geographic locations, including cities and rural areas of Venezuela, to check their network latency and upload/download speeds to more comprehensively reflect the impact of the Venezuelan crisis on the daily digital interactions of its citizens.