Archive for the 'Behind the Scenes at Data Vendors' Category

Each May, the mad rush begins.

As the rest of the world stops to smell the growing roses, the demographic data vendors of the world are all working feverishly to be first to market with their latest and greatest updates and projections.

What exactly are updates and projections, and why do they all start coming out about the time when the last snows have melted (at least in Syracuse!) and spring fever hits?

Keep in mind that the major vendors all work with the same base for their calculations: the US Census. They all add in various sources of data to make their own flavors of demographic soup. Each year out from the Census, 2001, 2002, 2003, etc. the data for small levels of geography really gets stale. So these vendors make their money by creating “estimates” for the current year. Most companies in this business have a fairly rigorous model that involves lots of local source data plus additional current population surveys. Add to that list counts from large list brokers and you have a pretty decent formula for estimating the current population.

Projections on the other hand can range from 5 years to 10 years out, and some brave companies (Woods and Poole for example) produce projections for as far out as 20 years. So much can change in five years — much more so in 20 — and it is widely acknowledged that it is difficult to produce projections with great precision that far out. However, for some companies, a very, very educated guess is well worth the price — and so, long term projections are essential data points.

But the question remains, why May? Why not January? After all it is already 2006, why do they wait until the middle of the year to get out new numbers?

I’ve always been told that it takes a long time to gather all the required information into the models, tested, then released. I think that with the 2010 Census changes, we’ll see a very different update schedule in the coming decade. Data can be transformed so much easier, workstations have incredible power, too, and therefore numbers can be crunched and tested so much faster.

Perhaps some day, we’ll be saluting the new numbers with the champagne left over from New Year’s festivities?


Author: Data Diva