The end of the Statistical Abstract of the United States?

There are rumblings, mostly among librarians, over the end of the Statistical Compendia branch of the Census Bureau, in 2012. The branch has produced the Statistical Abstract of the United States every year since 1878.

Here’s a note from a listserve on the topic:

I couldn’t take the uncertainty anymore so I called the Census Bureau and spoke to Ian O’Brien who’s the Chief of the Statistical Compendia Branch. He said that the 2012 budget doesn’t include funding for his branch, which would mean the elimination of not only the Statistical Abstract, but all titles produced by that branch (State and Metropolitan Area Data Book, County and City Data Book, etc.). No new editions would be produced in print or online.

And from the Census Bureau’s budget proposal for 2012, page 82 [pdf]:

The Census Bureau requests a decrease to terminate the Statistical Abstract program. The FY 2012 budget request is the result of a review of both ongoing and cyclical programs necessary to achieve Department of Commerce and Census Bureau goals and difficult choices had to be made in balancing program needs and fiscal constraints. The availability elsewhere of much of the information in the statistical abstract has led the Department and Census Bureau to the difficult decision to terminate the program.

This will be a $2.9 million cut.

The sense that I get from library science blogs is one of disbelief, but I’m not so sure that the move is that surprising. Government data has been growing more accessible via outlets like and even via Census itself, and while they still need work, the Abstract still feels dated in comparison.

I’ve played with the data a little bit, and it’s all available in lots of tables in Excel and PDF formats. Keep in mind this isn’t the Statistical Atlas, so it’s not graphics or maps. It’s straight up tables, and as Infochimps notes in their entries from the Abstract:

The Statistical Abstract files are distributed by the US Census Department as Microsoft Excel files. These files have data mixed with notes and references, multiple tables per sheet, and, worst of all, the table headers are not easily matched to their rows and columns.

Plus, the data isn’t straight from the Census Bureau. It’s data from other centers and organizations in a centralized spot, which is what is supposed to be working towards.

So while I think it’s too bad that a branch that has been around for so long is going away, it’s not so much a knock as it is a sign of changing data access and availability.


  • I think the disbelief doesn’t come because the information isn’t out there. It comes because finding that information without the help of the Statistical Abstract is a hugely daunting task not just for lay people and for undergraduates (my primary constituents) but also for most librarians who aren’t specifically gov docs librarians or data librarians. So for the vast majority of people, the Statistical Abstract is far more than a place that holds statistics. It’s a tool that points us toward where to find more statistics like the ones it presents. People can look up information on the general topic they’re interested in, and then use the source notes for that table to figure out which agency, project, or whatever-it-is collected and published the kind of information you need more of. Without that, Google is a vast jungle and it’s nearly impossible to know what your options are. (And I say this as a VERY google-savvy searcher.)

    I think it’s this second function, the pointing function, that I’ll miss the most and that isn’t duplicated anywhere else.

    • That I can understand. In that case, I wonder if time and money are better spent building a catalog instead of publishing tables of data.

      • That could be one compromise. And, in fact, I think people will have to start creating those lists on their own once the Abstract is gone, so it’d be great if such a list could be centralized and complete.

        Another function the Abstract provides, though, is pulling together data from various sources and normalizing the units. And beyond that, it’s incredibly convenient to have data from all these places all gathered, normalized, and displayed coherently in one place.

        I know budgets are tight (and probably tighter than tight, really), so I can see prioritizing some sort of catalog of data sources over all the rest of the work the compendia branch does. But really, I think the compendia they produce are key sources in so many ways, both factual and pedagogical, that any of these compromises seem hollow.

      • Also, never underestimate the importance of a well-constructed table of data. Undergraduates have a LOT to learn in that regard. :-)

  • Josh PolyGlam March 24, 2011 at 4:45 pm

    Well, you dont need statistics post-apocalypse! This makes sense.

  • Alesia McManus March 25, 2011 at 5:30 am

    I’m the librarian who created the Save the Statistical Abstracts FB group and the online petition

    One of my colleagues posted this link to the FB group that mentions that may also lose funding.

    I read the comments above. In some ways, the Stat Abs is a catalog. The online version is searchable. Both the print and online lead to the source of the data where more is available. Why create something new when this works well for the general public and layperson?

  • To expand a bit on what Iris said, it’ not just that it points to where to find more statistics, it provides a large amount of stats in one place. Yes, you can find the info elsewhere though you would have to go to several different sources to find the data that is currently in one spot now. When you get a student asking a multi-part data question, it’s very efficient to not to have to search multiple sources. Add to this that the very source you name in your post as a place to find easily accessible government data is now on the chopping block, you can see why librarians are a bit perturbed:

  • Stanley Henshaw March 26, 2011 at 2:12 pm

    The assumption that all the information in the Statistical Abstract can be obtained from other sources is not true. I provide the data on abortion rates and the characteristics of women having abortions, and much of the information is unpublished elsewhere. Also, the historical information is not easily available from other sources. I am sure this is true of other data series’ in the Statistical Abstract.