Paper : Ten years of the Venezuelan crisis - An Internet perspective
Authors : Esteban Carisimo, Rashna Kumar, Caleb J. Wang, Santiago Klein, and Fabián E. Bustamante.
Presenter : Xining Yuan, Tufts University.
Guest of Honor : Esteban Carisimo, Northwestern University.
Q: What motivated you to look into Venezuela’s internet infrastructure?
A: I’m Argentinian, so I was already familiar with regional issues. Venezuela’s situation is both personally saddening and academically unique. It used to be the wealthiest country in Latin America with good infrastructure, but its economic collapse presented an opportunity to study how infrastructure deteriorates when investment stops. This research can help in planning Venezuela’s future internet reconstruction.
Q: Have you considered looking into similar issues in Argentina or Brazil?
A: Yes. I have a 2020 paper on Argentina. While Argentina has high inflation and economic issues, the impact on infrastructure has been milder compared to Venezuela. Brazil, on the other hand, has had a massive expansion in internet services—growing from 1,200 to 12,000 ISPs over the last decade. It’s currently in a very strong position, both in terms of infrastructure and resilience.
Q: What are the key challenges—especially technical ones—you faced in conducting this research?
A:
- Data Availability: We weren’t sure if there were enough user probes in Venezuela to measure DNS root server visibility, but thankfully there were.
- Quality of Experience: We used M-Lab speed test data, but initially didn’t know if Venezuelans used it. Fortunately, they did.
- Scale of Data: With 15 years of speed test data across 27 Latin American countries, we had millions of samples. Managing and comparing these large-scale statistics posed a huge technical challenge.
Q: Do you have any follow-up research directions after this study?
A: Yes. I presented the work to the Internet Society, which works on global internet access. Venezuelans themselves asked about alternatives like Starlink. Starlink could bypass local infrastructure, but it’s very expensive and officially unavailable in Venezuela. Still, people buy dishes in Colombia and use Colombian credit cards. Studying Starlink in Venezuela is an exciting follow-up direction, although conducting that research is challenging.
Q: Have you looked into domestic network access in Venezuela, as opposed to international connectivity?
A: Good question. We added some analysis in the appendix. Unlike China or the US, Venezuela doesn’t host most of its own content. Popular services and even Venezuelan newspapers are hosted abroad. So even domestic users have to rely on international infrastructure, which hasn’t been upgraded since 1998. Even if the internal network improved, they’d still face poor performance because content isn’t locally served.
Q: Starting from around 2022, there seemed to be some improvement in Venezuela’s internet. But that timeframe also overlapped with COVID-19. How did COVID-19 affect the internet crisis in Venezuela?
A: COVID-19 made the Venezuelan government realize how essential the internet was for remote work and education. As a result, they issued new ISP licenses to local providers. While these modern ISPs helped improve connectivity, they remain unaffordable for many, creating a growing access gap. Despite this inequality, the overall internet situation started improving.
Q: What about the role of US sanctions during this period?
A: The US had imposed sanctions on Venezuela, which affected various sectors including internet infrastructure. During COVID-19, some of these sanctions were partially lifted. While it’s hard to prove causality, the timing overlaps with the internet improvements, though the main factor seems to be the emergence of new ISPs.
Q: If the country wants to upgrade devices like routers, switches, or cables, how does it source them, especially under sanctions? Has there been any study on the diversity of networking devices used in Venezuela in the past decade?
A: Yes, the team was curious about that. Even assuming no sanctions and enough money, Venezuela still struggles with basics like repairing optical fiber due to a shortage of specialized welding equipment. These aren’t extremely expensive, but are hard to replace or fix locally, which causes long outages when cables are cut.
Q: For graduate students new to internet measurement, where should they start? How can they get hands-on experience given the massive amount of data and tools?
A: Start with simple, well-scoped projects. For example, observe changes before and after an event like the deployment of a local ISP or Starlink station. Use basic tools like DNS resolution, bandwidth tests, and traceroute. Over time, you build up your understanding and develop a broader view of internet infrastructure.
Q: What do you think is the next important research direction in internet measurement?
A: A key challenge is understanding whether governments and policymakers are implementing effective strategies for internet improvement. Many governments aim for better internet, but may lack proper tools, policies, or execution. Studying the impact of such efforts across regions and levels of government—federal, regional, local—will be critical.
Q: Any final thoughts on designing network solutions globally?
A: Every region has unique characteristics—geographical, political, social. There’s no universal solution. For instance, mountainous towns might need different technologies than flat regions. We must approach internet problems holistically, considering local contexts.