In my earlier post I made the argument that, whatever its other drawbacks and merits, the California redistricting proposal is unlikely to cost the Democrats seats, and may actually win them some additional ones. My reasoning, in brief, is that Democratic voters are already so highly concentrated into safe Democratic seats that redistricting can only have the effect of dispersing them, and thus giving Democratic candidates more chances to win additional seats.
The table that I presented in the previous post contained some prima facie evidence for my argument, because it showed that the majority of Democratic seats are extremely safe, while the majority of Republican seats are only won by comfortable, but not overwhelming margins.
While this evidence is strongly suggestive, it may not be completely convincing, however; if geographic boundaries (such as cities, counties, etc.) contain even higher concentrations of Democratic voters than current Democratic districts, then redistricting by geography could lead to even greater concentration of Democratic voters and thus a loss in the number of Democratic seats.
It turns out that this is not the case, however. The following table shows voter registration for the largest counties and cities in California in September 2004, and the advantage in party registration enjoyed by Democrats in each district (a ‘+’ indicates more registered Democrats, while a ‘-‘ indicates more registered Republicans).
Source: California Secretary of State.
While there are several counties and cities with high concentrations of Democratic voters, rarely does the concentration of Dem. voters in these geographic districts match the extraordinary Dem. concentration in current Congressional districts – where 7 seats were won by 60 point margins, and 19 seats were won by 40 point margins. If Congressional districts were organized by city and town, a few districts might have winning margins of 40 points (and maybe one or two would even have margins of 60 points), but certainly fewer districts than the current 19 would be so lopsided.
This means that organizing Congressional districts by cities and counties will almost certainly end up dispersing the Democratic vote among more districts. And that in turn will mean that Democrats will almost certainly not lose seats, and may in fact gain some.