California is a hard state to analyze redistricting for several reasons.
First, there is the sheer number of districts. With 52 districts (down from 53) there are a lot of choices to make.
Second, besides the large population, you also have a very diverse population. The state as a whole is a minority-majority state. And there are enough Latino, African-American, and Asian-Americans that map makers have to consider whether it is possible to create influence districts (or even majority districts) for each of these groups.
Third, California has a rather unique method for choosing the members of the Citizen Commission that draws the lines. Individuals interested in serving apply to the auditor’s office which has panels review the application to narrow the pool to twenty Democrats, twenty Republicans, and twenty “others.” It is only then that politicians get to narrow the pool to twelve from each group. Then the initial group of commissioners are chosen at random from each of these groups and these commissioners fill out the commission from the remaining people in the pool. The end result is that the Commission is not necessarily wedded to the map from the last cycle. And the California Constitution bars considering the existing incumbents in drawing the lines. The last cycle (the first with the Citizen Commission) produced a map that differed substantially from the last legislatively-drawn map. Whether this cycle will produce a map with minor changes from the 2011 map or something that looks very different is unknown.
Fourth, while Los Angeles is the largest single metropolitan areas, there are a lot of different regions of California. Currently, Los Angeles County has nine whole districts and eight partial districts. And if you add Orange County, you get another three whole districts and one new partial district. But even with all of these districts, you are still talking about only 40% of the districts in the state. But the LA and Orange County districts represent a total shortfall of around 550,000 people. So it seems likely that they will lose a district between those two counties.
Fifth, while respecting county lines and respecting communities of interest are a criteria that the commissioners are supposed to consider, the 2011 lines definitely split more counties than was necessary and had fewer single county districts than were possible. Again, it remains to be seen if the new maps give more weight to minimizing multi-county districts.
Sixth, if you use the current map as a base, the current districts have shown drastic changes in population. On the one hand, the First District in northern California is approximately 50,000 people short of the population needed. On the other hand, the Forty-second District in Riverside County to the east of Los Angeles is approximately 55,000 people over the target population. So, there will be significant boundary shifts even if you start with the current maps.
Finally, the current maps show a slight preference for maximizing the competitiveness of districts. On paper, there are thirty-one solid Democratic seats, four safe Democratic seats, three lean Democratic seats, six seats that are the edge between lean Democratic and toss-up, one pure toss-up that favors the Democrats, one pure toss-up that favors the Republicans, one seat that is on the edge between toss-up and lean Republican, four Lean Republican seats, one safe Republican seat, and one solid Republican seats. Clearly, there has been very little packing of Republican votes in the current map. It is debatable about whether the desire to make districts competitive has resulted in some cracking of Republican votes as ten of the sixteen potentially competitive seats favor the Democrats. However, the Repbulicans currently hold nine of those sixteen seats.
For drawing the maps, I assumed that the Commission would follow the rules. In other words, I did not look at partisan lean of precincts in drawing the map since partisan benefit is not a consideration. I also did not look at the residences of the incumbents as the provisions governing the Commission bar drawing lines to protect incumbents. Finally, to the maximum extent possible, I drew lines that respected county boundaries. (Since the requirement is that the numbers start in the northeast corner of the state, I started there and tried to keep districts as compact as possible).
For the top line numbers, I ended up with twenty-six solid Democratic seats, seven safe Democratic seats, six lean Democratic seats, three seats that are on the border between lean Democratic and toss-up, two toss-up seats that favor the Democrats, one toss-up seat that favors the Republicans, two seats that are on the edge between toss-up and lean Republican, and five lean Republican seats. In other words, a slightly more competitive map than the current map and the nominal loss of a Democratic seat. But given that the Republicans hold three nominally Democratic seats, it is unclear who will end up with the advantage under this map. The number of competitive seats increases from sixteen to nineteen. Under this map a 52 to 0 Democratic sweep is possible although not likely.
Key changes by county, starting in the north:
Sacramento County — current map has one whole district and three partial districts; my draft map had two whole districts and one partial (which joined with San Joaquin County to make a whole district)
San Joaquin County — current map has two partial districts; my draft has one partial district
Santa Clara County — current map has one whole and three partial districts; my draft has one whole and two partial districts
Madera County — current map has two partial districts; my draft has one partial district
Fresno County — current map has four partial districts; my draft has one whole district and one partial district
Tulare County — current map has three partial districts; my draft has two partial districts
San Luis Obispo County — (the rare county in which my draft actually has more splits than the current map) current map has one partial district; my draft has three partial districts
Kern County — current map has three partial districts; my map has one whole district and one partial district
Ventura County — current map has three partial districts; my map has one whole district and one partial district
San Bernadino County — current map has one whole district and three partial districts; my map has two whole districts and one partial district
Los Angeles County — as noted above, current map has nine whole districts and eight partial districts; my draft has thirteen whole districts and two partial districts
Orange County — current map has three whole districts and four partial districts; my draft has three whole districts and two partial districts
Riverside County — (another of the rare counties that is more divided under my draft than the current map) the current map has three whole districts and one partial district; my draft has two whole districts and two partial districts.
As far as voting rights and minority representation is concerned; the current map has thirty minority-majority districts with ten in which Hispanics are the majority, three in which Hispanics are the largest share of eligible voters, and three in which Asian-Americans have the largest share of eligible voters. Minorities have at least one-third of the eligible voters in forty-eight districts. In my draft map, thirty-one districts are minority-majority districts. The number of districts with a Hispanic majority goes down to five but the number of districts with a Hispanic plurality increases to seven. The number of seats with an Asian-American plurality goes down to one, but there are probably some adjustments that can be made within the Los Angeles County districts to improve things. Overall, the number of seats in which at least one-third of eligible voters are members of minority groups increases to forty-nine with only the three districts in northern California having lower numbers of minorities.
In short, there are neutral maps that can be drawn up in California that would increase competitiveness. Given the overall lean of California, an emphasis on competitiveness would probably benefit the Democrats. But the process in California makes it hard to predict where the lines will be. And one thing to remember about California is its primary system.
The top two primary can create weird results. The majority party can be shut out in a primary for a winnable seat (either an open seat or one held by the other party) if it has too many primary candidates. Running two primary candidates is good (in most circumstances) as — if your party has the majority of voters — at least one of your candidates will advance and there is a chance at locking the other party out of the general. But if a party has three candidates running, there is always the chance that they will split the vote in a manner that results in that party being locked out of the general. There have certainly been times over the past decade when Democrats have failed to gain a winnable seat because too many Democrats ran and we got locked out of the general election. So, just having a good map does not guarantee winning a winnable seat.
Earlier this week, the Census Bureau announced that it will be releasing the block level data (the numbers used to actually draw the new lines) on August 12 (a couple of days earlier than the previous estimate). So the guessing about what the local numbers that have driven this series will be over and we will have the actual numbers. Of course, because, unlike in past cycles when a handful of states were released each week, all fifty states are being released at the same time and the process will begin immediately, we have used the estimates to write about approximately half of the states over the past three months. My general read is that the net results of the redistricting will probably move the median Congressional seat even further to the right of the country as a whole. This past decade, by the numbers, the Democrats needed to beat 53% in the total national vote to take control of the House. I would not be surprised if that number goes up to 54%. There are some states — New York, Illinois, and Colorado for example — in which lines could be drawn to make it easier for Democrats to win seats, but there are even more states in which Republicans are in complete control in which the lines could be drawn in a way to allow Republicans to gain more seats with a lower percentage than their current numbers in that state.
