California: Back to the Future

I have drawn a number of maps trying to predict California’s 2012 redistricting.  I will not be posting any of them in this diary, as I think it’s just a guessing game trying to figure out how exactly the lines will look.  There are several different ways the lines can be drawn – all adhering to the new commission standards.  (OK, maybe I will post a prediction map at some time in the future, once I feel confident that I am drawing the lines in a purely objective way – which is hard to do perfectly) …

Nevertheless, in this diary I wanted to find a totally objective way in which to predict only what the net effect of such a perfectly non-partisan, non-biased redistricting may be – in a “most likely” scenario.  I thought that one of the best and most objective ways to do this is to look back at the 1992 non-partisan map and see how those districts compare, in partisan terms, to the existing ones.  This entails nothing more than redrawing the 1992 districts into Dave’s Application.  Hence, I will not even post any statewide maps here, as you can see what the districts looked like via this link:

http://swdb.igs.berkeley.edu/m…

Instead of the maps, I have included a chart here showing how each 2002 district voted in the 2008 election and how each of the 1992 districts would have voted had that map been in effect in 2008.  Granted, there were obvious population shifts between the 90’s and the last decade, so the 1992 lines now would produce over-populated or under-populated districts, and the biggest difference is that there were 52 districts back then instead of 53 now.  But the districts can still be compared for the purpose of this analysis — because the point here is not really trying to see how any particular district would change, but only what the net effect would be, ie. more or less Democratic or Republican districts overall ??  As far as the missing 53rd District, I get to that too, towards the end of the diary …

Under this exercise, it appears that had the 1992 lines stayed in place for 2002, there might now be upwards of 40 seats in California that Obama won by at least 6 points (at least a 52-46 margin) as compared to only 34 such seats currently.

As can be inferred from the above chart, a perfectly fair redistricting of California is likely to add several more Democratic districts.  The numbers highlighted in yellow are districts where the partisan balance doesn’t change much when you go from the 2002 incumbent protection map to the 1992 non-partisan map.  These are districts where the change in Obama’s margin over McCain (or McCain over Obama) is no more than 4 points.  The numbers highlighted in red are ones where the margin changes by 5 points or more in the Republican direction, while the numbers highlighted in blue are ones where the margin changes by 5 points or more in the Democratic direction.  

First, we need to take out the districts where the numbers change by 5 points or more in the Democratic or Republican direction, BUT the change is not likely to make a difference because the district is already super-Democratic or super-Republican.  These include CA-7, CA-30, CA-34, CA-35 and CA-37, which are already very Democratic and which might become even more Democratic under a re-map, as well as CA-1, CA-10, CA-23, CA-27, CA-36, CA-43 and CA-53 which become less Democratic, but are still very likely to be retained by a Democrat even if the lines were changed (all except CA-23 and CA-36 would have Obama percentages of at least 62%).  

What’s left are 7 districts where the partisan change may entail a change in party control.  These include 2 districts that move in a Republican direction, and 5 that move in a Democratic direction (I have marked these districts using a bolded outline in the column referencing the 1992 District partisan breakdown).

CA-18 and CA-20 become less Democratic to the point where, at least in CA-20, the Republican candidate would have very likely won last November under the old lines.  It appears that CA-18 would have also been a 50/50 district in last year’s election had it remained under the old lines.  (For what it’s worth, it should be noted that both districts still remain “Obama districts”.)   However, for both of these Central Valley districts, the actual 2012 redistricting may not be as “brutal” if you’re a Democrat because of the VRA.  The commission is likely to feel the need to be quite “VRA-compliant” in this part of the state, and the resulting districts are likely to have more Hispanics than under the 1992 lines, and thus be more Democratic.

The 5 more Democratic districts include CA-3, CA-24, CA-25, CA-26 and CA-42.  All move in the Democratic direction in this exercise.  CA-24 and CA-26 become so much more Democratic, that a Democratic takeover would be likely, while the other 3 are right on the line – with the caveat that Congressmen like Lungren and Miller would not likely be able to hold the districts in a Democratic or even “neutral” year.  McKeon – I’m not sure, but he certainly doesn’t have the moderate reputation of someone like Bono Mack who is able to currently hold a district with a similar partisan makeup.

What’s interesting, at first, is that Calvert’s district does not appear to change much in the chart.  Here’s where the “53rd District” comes in … Out of the 52 districts under the 1992 plan, two stand out like a sore thumb, in terms of over-population (I am using today’s population numbers, but under the 2000 population numbers they would have been over-populated in a similar fashion).  Both the western and eastern halves, respectively, of Riverside Co. (which under the 1992 map were labeled CA-43 and CA-44 but now more closely correspond to CA-44 and CA-45) are over-populated by over 300,000 persons, or something like 150% (Calvert’s area is at +367,000 while Bono Mack’s is at +321,000) … No other districts in the state even come close (every other is under 200,000, in most cases 100,000 or less) … Taken together, the +367,000 and +321,000 over-populations add up to almost exactly one new Congressional district in Riverside Co.  Granted, we’re using 2006-2008 population estimates on 1992 lines.  Nevertheless, even between the 1990 and 2000 Censuses, Riverside Co. had the highest population growth in the state (other than three relatively small counties – San Benito, Placer and Madera).  Thus, if the 2002 map was truly “non-partisan” the 53rd district created may have been centered not in the Central Valley, but in Riverside Co.  

To see how this plays out in a partisan way, I drew 4 different possible scenarios of how Riverside may look if it contains 3 districts of equal population (with each district corresponding to an ideal district size using population estimates for 2006-2008) …

The first map below is simply the districts under the 1992 map; the old CA-43 is in magenta while the old CA-44 is in blue:

The second map shows how the districts may look if Riverside Co. is divided into 3 districts of equal population.  In this scenario, equal numbers of people are taken out of both the western (magenta) and eastern (blue) halves of the county to create a new district in the middle.  I tried to draw so that cities are not split up by the district lines.  The new district, in orange, encompasses Riverside, Moreno Valley, Calimesa, Beaumont and Perris, and its partisan breakdown is 56 Obama – 43 McCain (the western district in magenta is 47 Obama – 51 McCain, while the eastern district in blue is 51 Obama – 48 McCain).

The third map below shows how the map may look if, instead, Riverside, Moreno Valley and their immediate environs are put together into one very compact district (almost rectangular) while the magenta and blue districts adjust accordingly for equal population.  In this scenario, the new orange district is 58 Obama – 40 McCain, the magenta one is 46 Obama – 52 McCain, while the blue one is 50 Obama – 48 McCain.

The fourth map has Riverside, Norco and Corona all in one orange district at 53 Obama – 45 McCain (the magenta district is 50 Obama – 48 McCain, while the blue one is 50 Obama – 49 McCain).

The last map tries to keep most of Moreno Valley together with Palm Springs (as they are currently); while Norco and most of Corona stay with Riverside.  The orange district here is 55 Obama – 43 McCain; magenta one is 44 Obama – 54 McCain and the blue one is 55 Obama – 43 McCain.

As you can see, any reasonable way you slice and dice Riverside, you’re creating at least one quite Democratic district (and possibly two) – which, like CA-3, CA-24, CA-25, CA-26 and CA-42 above would likely move in the Democratic direction (in scenario #4, the one that produces the “weakest” partisan shift for Democrats, still has the new Riverside Co. district at 53 Obama – 45 McCain which is a 7 point improvement in the Obama – McCain margin over the existing CA-44).

You can see from this exercise that the 2002 plan, even though it was an incumbent-protection map, has benefitted Republican incumbents significantly more than Democratic incumbents.  If something truly non-partisan had been created in 2002 Democrats would likely have at least several more members currently.  With 2012 redistricting being non-partisan it is therefore likely (though never guaranteed ofcourse) that the map will increase the number of Democrats in California.  And, as you can see from the “53rd District” scenario above, even in Republican districts which have or are experiencing the highest population growth in California (like Riverside Co.) the growth has been in Republican districts, but among Democratic constituencies.  

To sum up, therefore, I think that a non-partisan map will ultimately be more of a plus for Democrats than Republicans.  If drawn in a neutral manner, a new CA-18 and CA-20 might be marginal to leaning Republican — but the VRA is likely to “save” those districts for the Democrats.  Meanwhile, non-partisan criteria will more likely than not enable the creation of more Democratic districts in geographic areas currently corresponding to CA-3, CA-24, CA-25, CA-26, CA-42 and CA-44.  The new Democratic-leaning districts may even not correspond to the districts listed here — as this exercise for me really tried to gauge a net effect rather than trying to predict individual districts — but they will appear somewhere in California (mostly in the southern part of the state).  Several of these new districts will likely go Democratic in 2012, while others may be more marginal but may nevertheless provide Democrats at least an even chance of takeover (later if not sooner, as demographic change progresses) — which is better than what the existing map has to offer.  If you’re a Democrat, fortunately we no longer will have the self-defeating Democrats of 2002 to draw the map for us. Instead, we should look forward to the Commission’s work in the Golden State.