Tuesday, December 29, 2009

Live by the Sword

Speculation, elation and hope were rampant Thursday when a forecast from Election Data Services (EDS) predicted that Washington would gain a 10th Congressional seat due to probable results from the 2010 Census.

(The U.S. House of Representatives is set at 435 members. Population growth results in reallocation of those seats among gaining and losing states to achieve equal representation.)

Such enthusiasm (despondency elsewhere) is most understandable given the power of Congressional representation and the allocations of federal dollars based on populations.

However, a quick review of the EDS forecast and the underlying U.S. Census's American Community Survey (ACS) engenders concern. The EDS news release's tables contain the notation: "No Military Overseas factored in." The issue of this seemingly obscure note is that the Census Bureau allocates the count of deployed soldiers to their home of record, an adjustment that takes place later in the population allocation.

Fort Lewis currently has about 18,000 soldiers deployed, returning in June, 2010, past the April 1, 2010 Census count date. Those soldiers and about 1.23 dependents each could count for Washington's lead of 24,592 persons as forecast by EDS.

Much of the state's populations gain, especially of recent years (that have propelled it past contender Oregon for a Congressional seat gain), has come with the incremental realignment of about 13,000 soldiers, plus families, to Fort Lewis since 2004. Since speculation focuses on the north Thurston County area as a rapidly growing location, note should be made that soldiers and their families are increasingly choosing that area for off post housing.

The cause of concern is that Census uses deployed soldiers' home of record for allocating population. Text in the EDS news release addresses this issue by saying military reallocation shouldn't change the results. What's left unclear is if ACS or the EDS deducted deployed soldiers from local posts from their respective estimate and forecast. That potential reduction occurs along with the unknown factor of whether soldiers' families remained in this area during their breadwinner's deployment or chose to return to the extended families of either spouse.

Before we all cry foul, we should remember that Washington's 9th Congressional seat was won during the 1990 Census because of soldiers deployed from Massachusetts and were not counted there. The Supreme Court eventually upheld the Census rule to allocate by home of record to Washington's gain and Massachusetts' loss.

1 comment:

Rob Allen said...

The EDS press release doesn't mention the margins of error in their various estimates, or aknowledge the fact that Census estimates through both the Current Population Survey and American Communities Survey both have small sample sizes at the state level. It would be interesting to know if the changes they predict are within those margins of error.

Regarding where soldiers are counted, that seems a double-edged sword. While many soldiers are based here and will be counted in their home state, how many soldiers will be counted here that are located somewhere else? We should understand the trade-off before crying foul.