Defining “document” used to be easy – it was a single paper or set of papers put together for a specific purpose. The limited amount of information one sheet could hold hadn’t changed much over millennia. 

But now that governments are moving to electronic documents to leverage efficiency and public benefit, the amount and type of information a single document can contain can be huge and widely varying. What might at one time have been hundreds of documents might now be a single electronic file, and people have the flexibility to combine multiple documents into a single one – for example, multiple email messages and attachments might be combined into a single PDF file.

Data-packed electronic documents replacing many pages of paper documents would have the tendency to cut the number of documents needed to respond to a request. But other forces, which we’ll examine below, have tended to significantly increase the quantity of documents provided – and they all stem from the fact that governments are increasing by orders of magnitude the amount of data and documents they are collecting in a multitude of formats.

The pace at which governments are generating new electronic documents, data, and records is difficult to get your head around. The variety and quantity is staggering – imaged and digitized paper documents; born-digital documents; records of communications in a plethora of media; video and audio recordings governments have created themselves or collected from other sources; photos; GIS and other mapping data – the list goes on and on. As the State of Washington puts it “One of the consequences of technology is the ability to generate unprecedented volumes of information and data – exabytes [a billion gigabytes], anyone?”  Or as California describes it “the explosion of electronic data generated in recent decades.”

While GovQA has cited the reality that governments are drowning in public records requests, one article attributes much of the pile-up to email:

“…We are drowning in email….We use email for everything.

“Everything.

“We have long-winded discussions. We have short, pithy exchanges. We transfer documents and photos. We copy our bosses and blind-copy anyone with a pulse. We send HTML newsletters and promote activities through third-party email services. We forward and reply. We have auto-replies and vacation notifications. We have email addresses and inboxes designated for specific services. We create folders (and forget what the folders contain). We send and receive email from our phones, our tablets, our laptops.

“Not to mention: we now have 15, 20, maybe even 25 years’ worth of archived email messages.”

While for many governments it’s relatively easy to do system-wide searches for keywords and date ranges that requesters are interested in, the volume of results these searches generate can be overwhelming. The article cited above continues:

“Why? Because often [many] people are involved in the discussion, and each of those people have been busy replying, forwarding and cc’ing. Because when someone replies, forwards or cc’s, often they are including all of the replies, forwards and cc’s before it.”

So searches will end up producing the same email chains over and over again, and without special technology there is no practical way to mark a chain as having been previously reviewed so staff don’t have to waste time reviewing it over and over as it pings back and forth accumulating new replies and attachments.

Washington State requires that if certain kinds of duplicate records are pulled from different sources they still have to be provided (e.g. the same email exchange produced by two different sources) – but the reality is that even for jurisdictions that don’t need to do this it’s not practical to spend time weeding out duplicates. And for jurisdictions like Virginia that charge requesters for review time, this kind of winnowing down just increases costs.

Another pitfall of email is hidden/inaccessible personal archives. Microsoft-based email systems automatically shunt emails to users’ local hard drives as .pst files, which don’t get caught in system-wide searches. 

Governments are creating masses of new records through their enthusiastic adoption of social media. Since it documents government activities, social media fits the definition of public record and is covered in some form by the public records laws of all 50 states – and even sometimes by open meeting laws.  

As Florida law states, a public record is any material “regardless of the physical form, characteristics, or means of transmission, made or received pursuant to law or ordinance or in connection with the transaction of official business by any agency.” Many states, like California, provide specific guidance for determining what social media activity must be preserved under document retention requirements.

Examples of public records requests involving social media include:

• A request to the City of Santa Barbara, CA, for all social media postings related to a gun buy-back event sponsored by the police department
• A Colorado agency received a request for any and all communications, including on social media, related to oil drilling and fracking 
• A request to a North Carolina town for any and all discussion, comment, or feedback regarding plans for a proposed bridge.

At the same time that political and social developments have spurred citizens to demand greater access to the workings of government, technology has made the information governments generate much more shareable. These two forces come together in open government – making information available on agency websites or dedicated open government portals without it needing to be requested. Proactively posting information and documents is an investment, especially if information that is exempt from disclosure has to be redacted; but some governments are betting that doing so will promote efficiency as well as increase transparency and public trust. The State of Wisconsin, for example, posts investigative files, photographs, and audio and video recordings related to Officer Involved Critical Incidents – for example the August 23, 2020 incident in Kenosha in which police shot an unarmed citizen. 

In theory, proactively making records accessible through self-service (whether using request deflection tools like those offered by GovQA or Open Data portals like those which can be integrated with GovQA) should reduce the number and scope of requests – and thus the quantity of documents needing to be processed and provided. 

The City of Lakewood, CO, for instance, has implemented the Lakewood Speaks interactive platform to boost public participation in city council, planning commission and other meetings where issues are debated and decisions are made. The site makes it much more convenient for citizens to review meeting agendas and submit comments before meetings and take part in the meeting itself. According to this article “The results of the site, which has been active in Lakewood for three years, are staggering. Beyond increasing overall participation more than eightfold, participation by 35-54 year olds is more than ten times higher and the number of participants under 35 is more than 100 times higher online than in-person.”

Governments have provided public meeting documents on a proactive basis for years, but Lakewood Speaks and similar platforms are also creating new types of documentation (including around site users) that are subject to public records requests and privacy considerations. 

For example, comments submitted through Lakewood Speaks are published on the site along with the commenter’s full name, but commenters must also submit their street and email addresses, which become part of the public record.  As the site advises: “Be aware that all information submitted via email or through our websites may be subject to a public records request pursuant to the public records statutes.”

There will always be new technology and new record formats that have the potential to increase or decrease the quantity of response documents assembled for a typical public records request. But trying to measure the impact of these innovations on response document quantity is complex – because many, like the Lakewood Speaks site, could lead to both increases and decreases that are difficult to parse out. And it can be like a game of whack-a-mole – solving one problem or creating a new benefit with one mechanism can just create a different – or entirely new – problem. 

The State of Washington asks “Technology was supposed to make things easier and better, right?” To which one data manager responds: “information is a frenemy” –  just as data and records have great value, they can also create huge burdens. Balancing these factors will be critical as document counts continue to rise.  And technology certainly has a role to play in simplifying the management of the increases in response documents.