BETA
This is a BETA experience. You may opt-out by clicking here

More From Forbes

Edit Story

In The Cloud Era Why Do Government Websites Still Go Dark During Shutdowns?

Following
This article is more than 5 years old.

Getty

Last January during the inaugural government shutdown of 2018 I wrote about how government shutdowns in the US bring about a conflicting patchwork of updated websites, non-updated websites and even disabled websites as the nation’s myriad government agencies respond in myriad conflicting ways to the bureaucratic stalemate. A year later and more than a quarter century after the web’s birth, in an era in which the digital world impacts nearly every facet of our lives, why is it that the government’s web presence still disappears during shutdowns in a way its former paper era did not? Most importantly, in the cloud era when more and more government websites are being outsourced to the commercial cloud, why do they still go dark when their social media accounts do not?

As the longest US Government shutdown in history drags into 2019, myriad agency websites remain either frozen in time or disabled entirely.

Many websites, like those of the United States Geological Survey (USGS), United States Department of Agriculture (USDA) and National Endowment for the Humanities (NEH), remain online, albeit with notices they are no longer being updated during the shutdown. This is precisely how one might expect a government agency to react to a funding lapse: leaving its existing information online and available to the public even if it is no longer able to update it.

Notably, the National Science Foundation (NSF) followed suit this year and has thus far left its website online with a similar message about the shutdown. This is a sharp reversal from last year when not only did NSF take its website offline, but argued the cybersecurity risk was so great that it could not allow its website to be accessible in any form during a shutdown for fear of its being compromised. Ironically, NSF had left its social media accounts accessible, apparently unconcerned with the possibility of their being compromised via social engineering during the shutdown.

At the time, NSF had acknowledged its actions were out of step with its peer agencies and that it was reevaluating its IT policies. NSF’s policies were all the more notable given that its peer agency NEH had an explicit written policy requiring its website to remain accessible, though not updated, during shutdowns. It is a welcome change to finally see NSF join the web era in keeping its website online.

Indeed, during the current shutdown it appears more websites than ever are remaining online and accessible, even if they are not being updated.

The same, however, cannot be said for the vaunted open data portals that were supposed to usher in the governmental data revolution.

The main websites of the Environmental Protection Agency (EPA) and National Oceanic and Atmospheric Administration (NOAA) both remain online. When it comes to their open data missions, the EPA’s data portal is still accessible, 22 days after the shutdown began. In contrast, visitors to NOAA’S open data portal, https://data.nodc.noaa.gov/, are redirected to https://governmentshutdown.noaa.gov/, with a message that the site has been disabled during the shutdown.

Both data portals appear to be hosted on US Government computing systems, rather than outsourced to the cloud. The fact that EPA has been able to maintain access to its site during the shutdown demonstrates that it is possible even for government hosted data portals to continue to function during shutdowns.

Other open data sites like HealthData.gov and USAID’s Development Data Library (DDL) have also remained online and fully accessible. Both are hosted on Amazon’s AWS cloud service.

At the same time, the most famous US Government data portal, Data.gov, has been entirely disabled, despite being hosted on Amazon AWS. Similar to NSF last year, despite disabling its website, Data.gov has elected to keep its Twitter account online.

It seems that in 2019 it doesn’t matter if a site is hosted in the cloud or on a government-run computing system – the determination of whether it remains available during a shutdown seems to be political, rather than technical.

It is surprising that even cloud-hosted sites like Data.gov are being disabled, while their social accounts remain available.

As social media supplants websites as the public’s interface with government agencies, will the shutdown of the future no longer result in lost content as agencies leave their social accounts as-is? The ad supported and surveillance state nature of social media mean government agencies don’t have to pay to maintain their accounts, meaning there is no financial impact of leaving those accounts active during shutdowns.

In the meantime, where is a researcher to go for critical government data during the current shutdown? Sadly, at least one data mirror run by the University of California, Datamirror.org, which had archived 42TB of datasets from Data.gov, is no longer accessible, having been unable to secure the resources to continue its archive just as it would have been most useful.

Thankfully some of the highest traffic government datasets are already mirrored and available on the major commercial cloud platforms, such as a number of NOAA’s datasets. These archives have not only vastly increased their usage but also ensured their continued availability during government shutdowns.

The fact that government data portals can simply disappear during a shutdown is a stark reminder that we need more non-governmental archives of these shared national resources. Having official or at least updated mirrors of platforms like Data.gov on non-government-controlled servers would ensure their continued access during a shutdown.

One option would be for the US Government to partner with a major existing web archive like the Internet Archive, which already regularly archives government websites as part of its general Wayback Machine. What if the government provided funding for an outside agency like the Internet Archive to maintain a live mirror of all US Government data portals, updated daily? During shutdowns any portal that is unable to remain available could simply provide a link to this archived version of its data, ensuring uninterrupted access.

Of course, one might argue that if the government’s physical buildings are shuttered, why should their websites remain available?

The answer is that in the paper era, government information was still available, mirrored in libraries across the country even if the agencies that produced that information were closed. At a research university library, work using government information could continue within disruption no matter whether the government was open or closed.

In contrast, in the web era, government websites and data portals are the sole source of all of that information that was formerly mirrored in libraries all across the nation. When agencies shutter their data portals during a shutdown, that data suddenly ceases to exist, cutting off researchers, journalists and the public from the information they need.

The heavy reduction in IT personnel during government shutdowns also raises critical questions about the nation’s ability to defend against cyberattacks during these vulnerable periods. Intrusions have more likelihood of going unnoticed while a full-scale attack would find most agencies even more ill-prepared than usual to respond.

Putting this all together, it is simply unimaginable that in 2019 our nation’s web presence could go partially dark during a shutdown. While agencies seem to be making strides in keeping their main websites online during shutdowns, data portals are still casualties, cutting off access to critical datasets. The fact that it does not appear to matter whether data portals are hosted on government-owned servers or in the cloud on whether they remain online during a shutdown or are disabled, suggests that the remaining obstacles are political rather than technical.

After all, it was just a year ago that NSF took down its entire website, arguing that cybersecurity risks prohibited it from having any web-accessible content during a shutdown. That NSF has finally entered the 21st century is a welcome sign that federal agencies may be slowly and grudgingly modernizing.

Yet, the fact that data portals like Data.gov still go dark, show that we have a long way to come. The lack of alternative sources for government data during shutdowns also reminds us how vulnerable the digital era has made us. Unlike in the past where government data was mirrored at libraries all across the nation, today when an agency takes its data portal offline, that’s it, that data is gone.

In the end, how long will it take for government to recognize the importance of maintaining access to its vast landscape of critical datasets during shutdowns? Will government find a way to keep its own websites running during shutdowns or is a better approach for it to partner with outside web archives to maintain access during shutdowns? How will the digital world replace the print era’s massive replication of governmental records at libraries all across the country to ensure their uninterrupted access? Only time will tell.