[EL] Apportionment Analysis critique

The short answer to Mike's analysis is that both of us are right and both of us are wrong. But, more importantly, it would be irresponsible to claim New York will only lose one seat when it is just as likely the state will see a loss of two congressional districts. We prefer to point out the potential problem so that politicians and others can get prepared, rather than follow the path of the ostrich and stick their head in the sand.

By way of a longer answer, the following is what happened and details the facts that are before us. Election Data Services was provided new 2010 state population estimates by ESRI. The initial claim was that these represented Census day (April 1) estimates. We did not receive 2009 estimates. If we had, we may have also done a proportional allocation of the data across time once we later learned the 2010 estimates reflected July 1. We have done such "linearly interpolating" in the past when we were comfortable that the process to produce the two end points of data were comparable.

However, there are indications that this is not the case with the ESRI data for 2009 and 2010. In discussions with demographers at ESRI it appears there were methodological changes made between the two years so that they could more completely understand and track the economic recession/depression and foreclosure crisis in the nation. The fact that additional efforts were taken made the ESRI data much more valuable in assessing potential populations by state. But it also raises questions whether "linear interpolation" should be done. If Mike had taken more time to understand this rather then seek a headline, or even talked with us, he probably wouldn't have made his mistake.

If one does attempt to create estimates for April 1 through "linear interpolation" and run the resulting numbers through the apportionment formula, Mike is correct to note that New York would only lose one seat. However, he fails to note that the state's 28th district would be seat number 435, the very last seat to be handed out in Congress. Being in this position is the most precarious of all, as the slightest change in any state's population number could send New York's last seat over the edge and to another state. The fact that the military overseas numbers has yet to be added to the estimates are also likely to change the last one or two seats, as they did in 2000 when a seat disappeared from Utah and ended up in North Carolina.

Finally, adjusting the "linear interpolation" method reveals that New York's flip to a second seat loss doesn't wait for the middle of summer, but instead occurs at June 1, when the Census was still being taken. This points out the underlying fact in the data, that New York is heading in the wrong direction to keeping it's representation. As time goes on the state will definitely lose that second seat and maybe more.

There is one point Mike and I can agree upon ... the final arbitrator will be heard from in less than three months, as the Census Bureau announces the official 2010 population for the 50 states. Stay tuned.