See If You Can Get It On the Paper

New at Reason: Jeff Taylor says pre-digital government isn't worth the paper it's printed on.

Comments

[david byrne]

Let me be the first to congratulate Mr. Cavanaugh on an excellent Talking Heads reference. Some civil servants are just like my loved ones.

[Mandrake Ethos]

The article skips lightly over the primary reason government, and most businesses, still ultimately rely on paper. It works.

More importantly, paper is enduring. This might sound strange, considering we often use paper as a metaphor for insubstantial or fragile, but it's true. Audiotapes, videotapes, digital recordings--nothing has the material or practical permanence of paper.

Look at that situation from a records manger's perspective or an archivist's perspective (over the course of a career, especially). My own state government houses quite a few hundred cubic feet of digital records that can no longer be read. The equipment and software to do so simply isn't made anymore, and it's been years since it was used.

For proponents of an open government (shouldn't that be all of us?) this combination of planned obsolence for digital technology should be disturbing when it comes to government records.

There are also questions about provenance and authentication where paper has the advantage, but that's probably enough archivology for now.

--ME

[jc]

Chasing each succeeding wave of technology will cost gobs and gobs of taxpayer money every half-dozen years or so.

[garym]

Echoing Mandrake Ethos -- and speaking as someone who works in the field of digital archiving -- I agree that digital data poses a greater risk of catastrophic loss, not to mention tampering, if not handled carefully.

Obsolete formats and media are one issue. The durability of the physical media is another.

Digital media can also offer advantages. Information can be backed up at multiple sites, making it less vulnerable to flood and fire damage. More information can be stored per cubic foot, decreasing pressure to discard old records. But it's very easy for old information to become obsolete, if no one is actively maintaining it.

Imagine someone in 2104 going through old records and saying, "A CD-ROM? What in the world is THAT?!"

[Mo]

I'm with gary and ME, we need to have, at the very least, paper backups of records. Just like I am not happy with voting at a voting machine w/o a paper backup, I am uncomfortable with a government without the same. That said, we need our government to catch up technologically. Let's remember that in a hundred years, anyone that finds a CD could think it's a Frisbee (it'll be just as useful, too).

[shanep]

I hope government sticks to its archaic guns.

We'll continue to move forward and find alternative ways of organizing collective decision-making voluntarily thru choice despite the old coercive regulatory welfare state.

I hope the trend stays that way.

[Shannon Love]

I think part of the hesitation in allowing oneself to be recorded springs from a fear that carefully selected phrases will be extracted publicized to one's determent. Only bits and pieces make it to the larger public debate.

Now that the internet allows me to read actual transcripts and papers, I am struck by how poorly the media captures the essence of what people original said or wrote. It must be disturbing for a pubic figure to contemplate having their every utterance or memo infinitely parsed by those hostile to them seeking to find one little piece that might sound bad.

[Matt XIV]

While making hard copies of certain important records is worth the effort, in most cases human-readable paper backups are a waste.

First, most data isn't worth keeping. Why should we pay the costs of foward conversions or paper archives of data that will be largely useless to future generations? Do our descendents really need to know if I paid my parking tickets on time or what address my tax rebate was sent to? When something isn't converted, it's typically because it's this sort of trival poop that composes the bulk of digital records - the high volume, low value stuff that wouldn't be worth recording at all if it weren't so cheap to do.

This leads to my second point that requiring paper archives may actually result in less data being preserved. Since it's cheaper to record data on digital media than paper, more data is recorded when digital means are available. So even if we lose some to unconvertable formats, we can record much more detailed records than if paper copies were mandated.

Third, digital records may often be individually less robust, but it's much easier to make copies. For the price of one book, data could be backed up thousands if not millions of times over in digital formats. Spread the locations of these copies out and your data is much less vunerable to random events or sabotage.

Finally, creating digital records that would be as permanent as human-readable print wouldn't be difficult if the desire existed. Digital data could be printed at a much higher density on paper than readable characters and as long as the specifications for reading it were preserved, a optical scanning device to extract it could be built at any time in the future.

[Matt XIV]

I suppose I should disclose a bias - I work for an company that produces electronic medical records software, so I'm not exactly a neutral party in the filing cabinets vs ferromagnetics debate.

[Phil]

Let's remember that in a hundred years, anyone that finds a CD could think it's a Frisbee (it'll be just as useful, too).

Just like they do with phonograph records now.

Wait -- no they don't.

[garym]

One more thought here: Getting the activity underway to refresh decades-old data can be difficult. It's easy to justify an ongoing cost of storage space for old printed records. But for someone to realize that old magnetic tapes, or CD's. or Zip disks, or whatever are reaching the end of their life cycle, for resources to be allocated to move the information to fresh media, and for possibly obsolete equipment to be found to enable the transfer is a significant barrier. It's hard to justify the cost and effort when what's at stake is old data that people rarely care about.

[Larry A]

I suppose there are reasons to be concerned about researchers who, in 50 years or so, might have trouble accessing today's government records. What I liked about the Estonia's experiment, though, is that it seems a heck of a lot easier to find out what government did yesterday.

[thoreau]

I wonder if some sort of "smart-archiving might be done." Keep new records in whatever the current digital format is, and maintain usage logs for various records. As new technologies come into play, selectively archive records based on their access history. Anything that was repeatedly accessed over an extended period would be copied to a more enduring medium (e.g. paper). Anything else would be kept in its old format, and in time would only be accessible to a historian who can find appropriate technology at the antique shop.

And it should go without saying that some records might be deemed sufficiently important to keep regardless of their recent access histories. But for anything that's in question, base it on access histories.

[thoreau]

Another thought experiment:

The next time a government agency is moving a bunch of data to archives, it might be interesting to poll some leading historians and ask them what they think future historians will find most useful in that data. Put their answers in a safe place, and in 25 years ask the leading historians which data from that archive interests them the most. Compare.

[MALAK]

I am periodically writing down my thoughts and ideas in a ringed notebook so that, when I am gone, my daughter will be able to read about how I felt about her, and about my life, as she was growing up. With the rapid changes in technology these days, this seemed the best way of having a lasting record that she would be able to 'access' no matter what. I know this does not impact on what may be best for government record keeping but, as always, I despise government at the best of times, and the fewer records they have of my existence, the better.

[Douglas Fletcher]

MALAK, did you ever see the movie "The Little Girl Who Lived Down the Lane?"

[Douglas Fletcher]

I'd get myself digitized but I'm afraid some puritan will come along & delete by naughty bits.

[db]

One of the major problems with digital archiving, other than obsolescence of media/extraction tools, is that the rate of decay of the media is higher than paper.

Large companies that keep track of millions of financial transactions have problems keeping up with the rate of media decay. Same too for NASA, which is faced with the extremely difficult decision of what to throw away. Much of the scientific information gathered during the early days of space exploration can be very valuable given new techniques of data mining. However, the degradation rate of the media on which those data are stored is becoming a problem.

As media begin to degrade over time, and the total amount of archived data increases, it will take more and more time simply to continue backing up the data. At some point (this was addressed in a paper I read last year sometime, don't know the title or author), the rate of data loss becomes higher than the possible rate of backup, and you have to decide what's worth keeping and what's worth throwing away.

[Russ D]

Seeing schools blow money on PC upgrades every year, I'm hestitant to see perpetual technological upgrades just for archived data. On the other hand, I'm all for government losing as much data as possible. Damn paradoxes!