Tomorrow is the special election to replace Tim Murphy in the House of Representatives. Since I've got the machinery to analyze districts, I thought I'd prep some maps to see what to expect. The election will be held using the old districts, not the Supreme Court's new districts, and in any other year the Republican would almost certainly win. Donald Trump carried it by 19.5% just 20 months ago. The district ran 18.8 percentage points more Republican than the state in 2016 and 18.3 in 2014. But recent polls imply that this race is close. I'm not going to narrate, but thought I'd share the plots I made for myself.
I'm profiling each of the State Supreme Court's new Congressional Districts in the Philadelphia area, looking at their voting behavior and their demographics. Today, the new District 04.
District 04 covers Montgomery county, in Philadelphia's suburbs. This county had been among the most gerrymandered in the state, and saw the biggest changes under the Supreme Court's map. It's a politically diverse county, and chopping it up provided a huge boon to the Republicans. It's a swing-y county, and gets national attention as a pivotal suburb that seems to be trending Democratic.
The county combines Democratic neighborhoods in the southeast with Republican neighborhoods in the northwest. However, that doesn't end up being the relevant distinction to make. The northwest neighborhoods are sparsely populated, and represent very few votes. Instead, the most important distinction is between the heavily Democratic suburbs just outside of Philadelphia--Elkins Park, Glenside, Abington--and the marginally Democratic suburbs in the center. The GOP strategy had been to waste the votes of the former by lumping them in with all-Democrat Philadelphia, while distributing the latter with Republican districts to create safe-but-not-too-safe Republican districts.
Here's how the county used to be divided. It includes Goofy's head of the famed former District 7.
The new district is reliably Democratic.
The county is predominately White and higher income. The racial exception is Norristown, and the wealth exception is the more rural area in the northwest.
Turnout in 2016 came disproportionately from the inner suburbs. That's where the population is, but also has the highest turnout per resident.
This November is a Gubernatorial election. The turnout falls, but proportionately less than in the rest of the state. It also falls less in the southeast, so those neighborhoods are *even more* important in elections for Governor.
As I've pointed out in every one of these profiles, the Trump vote closely matches the Sanders vote. The district went 59-41 for Clinton over Sanders, a bigger Clinton win than the state overall. That was largely driven by the southeast.
The racial cross-tabs are less interesting for this district than others, mostly because it's so White. Perhaps most interesting is the stability of Hillary's primary numbers across races; she doesn't seem to have done quite so well in Black neighborhoods in the county as she did in Philadelphia.
This week, I'm profiling each of the State Supreme Court's new Congressional Districts, looking at their voting behavior and their demographics. Today, the new District 05.
District 05 is the first district we're looking at that stretches outside of Philadelphia. In total, 80% of its population comees from Delaware County, 16% from Philadelphia, and 4% from Montgomery. (That area in South Philly is deceiving; much of it is industrial and has no population).
The district contains portions of Bob Brady's old District 1, which used to stretch out to Chester in order to gerrymander Democratic votes together. It is much less gerrymandered now, though still doesn't have any Republican strongholds.
Turnout for the district is high, running six percentage points higher than the state as a whole. That largely is due to the wealthiest suburbs, where 70-80% of the over-18 population votes.
They also fall off less than the rest of the state between Presidential and Gubernatorial elections.
Again, much of the interesting story of the district is in the 2016 Democratic Primary. The district voted largely for Clinton, with a pattern that we saw in other Districts: Black neighborhoods overwhelmingly supported Clinton, wealthier White neighborhoods still supported her by around 20 percentage points, and middle income White neighborhoods and students swung the hardest towards Bernie (though still ended up at close to an even split).
This district displays the largest over-representation of White voters that we've seen so far; they represent 64% of the population, but 69% of the vote in 2014. We will see if that continues in this high-attention Gubernatorial race this November.
This week, I'm profiling each of the State Supreme Court's new Congressional Districts, looking at their voting behavior and their demographics. Today, the new District 03.
The district is the most diverse of Philadelphia's, and perhaps of the state. It combines the affluent neighborhoods of Center City and Fairmount with West Philly, and then reaches up to Germantown, Chestnut Hill and Manayunk.
The demographic hodge podge of neighborhoods has one thing in common: all are Democratic strongholds. District 3 becomes the most Democratic in the state, and would have been won by Clinton in 2016 by 92-8 (yesterday's District 02 would have been second, at 75-25). It isn't just liberal, but also a Party district: Hillary beat Bernie by 64-36, also the largest margin in the state).
The only neighborhoods in which Clinton didn't beat Trump by more than 50 points were Manayunk and the northern parts of South Philly; middle income predominately White neighborhoods that share traits with Trump's base nationwide.
The votes are not evenly distributed with the population. Interestingly, the Black neighborhoods in Southwest Philly, Overbrook, and North Philly turned out strong in 2016, voting in numbers similar to their wealthier counterparts in Center City.
Those neighborhoods also fall off less between Presidential and Gubernatorial elections. The neighborhoods that vote for the President but don't for the Governor are the young, new-to-the-city neighborhoods: University City, Manayunk, and Penn's Landing/Northern Liberties. Chestnut Hill, Mount Airy, and Cedarbrook by far do the best in maintaining their voting through Gubernatorial years.
The diversity of the district played out in the 2016 primary. Manayunk, Queens Village, and University City all voted for Hillary. There's a clear racial divide, though not a complete one: all of the Black neighborhoods voted strongly for Hillary, while the White neighborhoods appear to be split, with gentrifying White neighborhoods swinging towards Bernie (thought Hillary still eked those out) and wealthier, longer-White neighborhoods voting decisively for Clinton.
Black residents are the majority of the residents of the district, and they are a similar majority of the voters. They vote the most Democratic (though everyone in the neighborhood votes D at over 87%), and supported Hillary strongly over Bernie. White, non-Hispanic residents are disproportionately more likely to vote, and thus make up more of the electorate than they do of the population, though that's mostly true in Presidential elections, when the younger neighborhoods turn out.
This week, I'm profiling each of the State Supreme Court's new Congressional Districts, looking at their voting behavior and their demographics. Today, the new District 02.
The new District 02 covers Northeast and North Philly; basically the entire city above Race/Spring Garden and East of Broad. This is a diverse swath of the City, combining some of the poorest sections of North Philly with gentrifying Fishtown, with the Trumpy "Middle Neighborhoods" in the Northeast.
Overall, the district is decisively Democratic. It voted 75-25 for Clinton in 2016, and 80-20 for Wolf in 2014. It also voted for Clinton over Sanders in the 2016 primary, 62-38, by more than the state as a whole.
That decisive Democratic 2016 victory was the function of a Democratic sweep of North Philly, combined with basically a even split between Clinton and Trump in the Northeast.
The Democratic map almost perfectly lines up with the racial Demographics. Those Trumpy neighborhoods were also the White neighborhoods (with the only exception of Northern Liberties at the bottom), while the Black and Hispanic neighborhoods voted decisively for Clinton.
The areal maps can mislead about proportionality; North Philly is *much* denser than the Northeast. Those votes carried the day. Notice that even though the Hispanic section around Erie Ave have the population density, their low voting rates mean that their votes per mile is lower than their neighboring Black neighborhoods.
Of course, 2018 is a Gubernatorial election, not a presidential election. Turnout is much lower in these elections, and different people vote. This district sees its votes fall by half, in line with the state overall.
And neighborhoods fall disproportionately too.
Among the Democrats, which neighborhoods are the Berniest? Which are the Hillary-est? While Hillary swept the District, she decisively won in the Black and Hispanic neighborhoods. This matching between Sanders strongholds and Trump strongholds plays out all over the city, and across the country.
Finally, we can cross-tabulate votes by race. We don't have voter-level results by race, so I approximate it by aggregating votes to Census Block Groups, and allocating votes within them. This isn't perfect (see Ecological Fallacy), but block groups are small enough and Philadelphia segregated enough to make this a very good proxy for how different racial groups voted. (Note, these percentages are the two-candidate vote. This mostly effects the Primary results, if Clinton won 45% to Sanders' 40%, she would have won 45 / (45 + 40) = 53% of the two-candidate vote).
White Non-Hispanic residents in the district are slightly over-represented among voters (43% of votes versus 40% of the population), and voted the least Democratic (though still being far from Republican).
Last weekend, the State Supreme Court announced the new Congressional boundaries for the 2018 election. Even though the GOP immediately challenged them in federal court, these boundaries appear fairer than any of the other maps proposed--by Republicans or Democrats. (NB: I purely define "fairness" as producing districts with partisanship close to the popular vote. I think compactness and county alignment are well-meaning but bad standards).
This week, I'll be digging into each of the districts in the Philadelphia area, looking at their map, their voting behavior, and their demographics. But first, here are some high-level summaries of the Court's districts. I'm not going to narrate, but mostly planting the images here for posterity. (Note, these were tweeted out by the @sixtysixwards twitter account. Follow that for the hottest real-time analysis.)
Governor Wolf has responded to the GOP's proposed congressional districts by proposing a map of his own. How does it compare?
The districts are compact (as were the GOP districts), but treat Philadelphia and suburbs very differently from the GOP's map, and slightly reduces the wasted vote gap between parties.
To see how Wolf's districts would change congressional voting, below is a plot of the current districts' results in the 2016 Presidential Election, and what it would have been under the proposed districts.
Similar to the GOP's map, Wolf's largely preserves districts' current partisanship. But there are notable exceptions.
District 6, in Delaware and Chester county voted barely for Clinton in 2016, 48.1% to 47.6%. Under the GOP's districts, it would have been a Trump 52-48 win (I do all of my projections using only the two-party vote). Under Wolf's districts, it would have been a Clinton 56-44 win.
Wolf's plan also changes District 15, which covers Reading and Allentown, from Republican to Democratic. In doing so, it makes Bucks County's District 8 more Republican; the district narrowly crosses from Blue to Red in the 2016 vote tally.
District 7, Patrick Meehan's district that the GOP plan sacrifices, similarly becomes a Democratic stronghold under Wolf's plan. That's the comically gerrymandered "Goofy kicking Donald" district, and any redistricting that values compactness will almost certainly swing it hard to the left.
Wolf's map also makes a number of Republican districts much safer. Districts 12 and 16 move from 10-point Republican wins to over 20-point wins, with only District 18 swinging from a 27-point Republican win to a 14-point win.
Overall, Wolf's map brings districts closer to 50-50 splits. And in doing so, it moves one district from Republican to Democratic. The current map has 12 districts that voted for Trump, 6 that voted for Clinton. The GOP's maintains the exact same split. Wolf's map would shrink that slightly to 11-7, or 61% Republican. This in a state that voted 50-50, with Donald Trump winning by less than a percentage point.
So what are we to do? Wolf's map may be as fair as we can expect while being tied to outdated notions of geographical representation and compactness. Maybe, some day, proportional representation.
Pennsylvania Republicans submitted a new map of proposed Congressional Districts to Governor Wolf on Friday. He has until Thursday to approve it and send it to the State Supreme Court.
I've downloaded the boundaries from pubintlaw, and created an interactive map. In order to project how each district would perform, I've aggregated the results of the races for President in 2016 and Governor in 2014. Let's dig in.
You can visit the full-scale interactive maps here.
The GOP's proposal clearly creates compact boundaries that align with county boundaries much more often. While this is what most people focus on with gerrymandering, I think the compactness standard is nonsense: it is completely possible to waste a party's votes with compact boundaries if perniciously drawn. In fact, the concentration of Democrats in cities means that even naive compact districts will almost certainly favor Republicans.
The new districts maintain the exact same Republican advantage as the prior ones. To assess this, let's look at the outcome of the 2016 Presidential election. Donald Trump narrowly won Pennsylvania, with 48.2% of the vote to Hillary Clinton's 47.5%. However, he won 12 of the Congressional Districts to Clinton's 6 (and in total 13 Republican House members won to Democrats' 5). This imbalance between topline vote and district count was achieved by packing Democrats all into a few districts, to waste their votes in landslide elections, leaving comfortable but not wasteful Repubican wins in the rest of the state.
With respect to the 2016 Presidential election, the new boundaries look a lot like the old ones:
Before we look at the district, notice the asymmetric skew between Republicans and Democrats. Republicans don't have a single district where they won by more than 50 points (thereby wasting their votes); Democrats have two. Given this, Trump would still have won 12 of the 18 districts under the proposed boundaries. The Democrats are still given the two most wasteful districts: PA-01 and PA-02, both in Philadelphia. They would continue to comfortably win PA-14, Pittsburgh, and PA-13, in Northeast Philadelphia.
District PA-07 swings hard to the Democrats. That district, famously gerrymandered as "Goofy kicking Donald," is a convenient sacrifice for Republicans as Representative Pat Meehan has already stepped down from the race amidst sexual misconduct allegations.
By sacrificing PA-07, Republicans make PA-06 a more likely Republican seat. The district, covering the farther-West Philadelphia suburbs, had been narrowly won by Clinton, but would now be a 4.4% Trump win.
PA-08 remains the most evenly split district. It covers the suburbs to Philadelphia's Northeast: Doylestown, Newtown, et al. It was and would remain essentially 50-50. Clinton would have won it by the slimmest of margins in 2016.
A strong Democratic 2018 could still swing seats. Districts 12 and 17 would have been Trump wins by 10 percentage points under these boundaries. These are the types of seats that we would expect to see competitive in a Democratic wave election. In fact, the logic of gerrymandering may mean that wave elections have more power: gerrymandered seats should be safe but not too safe, to avoid wasting votes, which may put a lot of them in play in a 10-20% Democratic swing.
But we shouldn't lose sight of just how troubling the baseline assumption in that logic is: Democrats basically need a wave election just to break even. In the dead-even state-wide election of 2016, Republicans won 13 seats to Democrats' 5 (Trump won only 12; another district split). Just because Democrats may be able to overcome the disadvantage to barely break even doesn't mean that the districts are fair.
Just to check that 2016 wasn't a fluke, here is the same plot for the 2014 Governor's race.
The Governor results are all shifted more Democratic; Wolf won the state with 54.9% of the vote to Corbett's 45.1%. Despite the rout, Corbett *won more* congressional districts: he won more than half the vote in 10 of 18 districts. The proposed boundaries do nothing to change that. PA-15, representing Allentown, becomes much more Democratic while PA-06 again becomes much more Republican.
The wasted votes are particularly breathtaking here: Corbett wins 10 of the 18 seats in both scenarios despite losing by 10% overall. And his largest win in a single district is by 18% in the new PA-04; Democrats would win by that much in *four* districts. Their votes are being concentrated, and wasted.
A note about my methodology:
To predict results within the new boundaries, I need to rely on data calculated from previous boundaries.
I am using election data from the amazing folks at openelections.org. I begin with election results at the smallest available unit of geography, the State House Districts. When a district is entirely contained within one of the new Congressional Districts, I simply add the votes. When the state house district is split between two or more of the congressional districts, I allocate its votes based on the 2010 census population-over-age-18 of the overlapping areas. Often, analyses apportion these voters based on area, but I expect using population-weighted measures to be much more accurate. Code available upon request (or once I organize my github repository).
US Congressman Bob Brady stunned Philadelphia yesterday when he announced that he would not run for reelection. This puts Congressional District 1, which he has held since 1998, up for grabs. Immediately, the scramble among Democrats to pursue the seat began.
In November, the seat will almost certainly be won by the Democrat. But who can win the Democratic Primary in May? Is the district ripe for pick-up by a progressive change candidate? Or will Brady, who is also president of the Democratic City Committee, be replaced by another Party stalwart?
What should we expect from the district? Where do its votes come from, and what do we know about those voters? Let's look at some maps. (You can visit my interactive maps, too).
EDIT: This may obviously all be moot if the district changes in response to the State Supreme Court’s map overturn...
District 1 stretches from Northeast Philadelphia, through South and Southwest Philly, and down into Media. The gerrymandering is obvious: the southeast stretch encompasses Chester and Swarthmore, which get grouped in with Philadelphia in hopes of concentrating Democratic votes and thus wasting them.
The district is a completely safe Democratic seat. Clinton won it in 2016 with 80% of the vote. Below, I've aggregated votes to State House districts. Interestingly, the River Wards and parts of South Philly that may be considered Brady's base were the Trumpiest Democrats, swinging sharply for the Republican.
So the Democrat will almost certainly win in November. But what kind of Democrat will it be? Does the district seem to prefer establishment, Clinton types, or leftist, Sanders types? The voters are actually quite diverse in this regard, combining Black and Hispanic neighborhoods that voted entirely for Clinton in North and West Philly and Chester, with River Wards and South Philly neighborhoods that voted for Sanders. (I continue to find fascinating the overlap between Sanders and Trump voters, suggesting that voters may not have been Left or Right so much as Anti-Clinton. I leave it to you to speculate on the reasons.)
However, the neighborhoods that most strongly supported Clinton are also the neighborhoods that voted the most. A huge majority of the District's votes come from Center South Philly, North Philly, and Southwest/West Philly. These overlap with some of Clinton's strongest neighborhoods, especially in Southwest and West Philly.
Since most of the votes come from the city anyway, it's useful to look at the other competitive Democratic Primary we recently had: the 2017 Philadelphia race for District Attorney. This will exclude the suburban voters in District 1, but they are unlikely to carry the District. Those Center City and South Philly neighborhoods that supported Clinton came out for Larry Krasner one year later, as did the Sanders supporters in the River Wards.
District 1 is racially diverse in a way little of America is. Of course, this top-line diversity does not actually represent neighborhood diversity, but is achieved by aggregating largely segregated neighborhoods. Below is a map of census tract race and ethnicity from the American Community Survey for 2012-16. Hispanic North Philly combines with White South Philly and River Wards and Black West Philly to create a patchwork representation. The map of race also lines up with the gerrymandered carve-outs of the district: District 1 hooks out to include predominately White Fairmount in Brady's district, while it notably excludes predominately Black Point Breeze.
Brady's district is predominately poor, typical of the City as a whole. Of course, while the population may be poor, the high-income tracts in Center City are also the places where most of the votes come from, so the average voter is wealthier than the district's average resident.
What should we expect in May?
A ton of candidates are going to run for the House seat. Whoever wins the Democratic primary is going to win in November. Philadelphia voters constitute the vast majority of the district's Primary turnout, and Philadelphians will decide the election. Given the neighborhoods that typically vote in primaries, the race will probably be decided by the White voters in Center City, South Philly and the River Wards along with the Black voters in West and Southwest Philly. Those voters split between Clinton and Sanders in 2016, but voted strongly for Krasner in 2017. If a candidate can manage to consolidate the left, especially in a likely high-candidate field, the race should be theirs.
The 2016 election was expected to split the Republican party. Donald Trump coalesced a base of voters fueled by cultural displacement and racial resentment, saying out loud what had only been hinted at. This seemed like it could threaten the alliance between cultural and fiscal conservatives that made up the Republican Party. Leading into November 2016, Democrats hoped that the fault line would turn off "establishment" conservatives and lead to a Hillary victory. Alas.
Ahead of the 2016 election, analysts predicted that Pennsylvania would come down to Philadelphia's suburbs. Voters there had traditionally voted Republican, but swung for Obama in 2012 and were the exact well-educated, wealthier Republicans expected to be turned off by Trump's rhetoric. Many of these districts did, in fact, vote for Clinton, but not strongly enough to swing the state in her favor.
How did these districts split?
Where the districts split
Last week, I posted interactive maps aggregating Pennsylvania's recent elections. I just added to them the ability to map "split districts", in which Clinton won the presidential vote but the Republican won the local race, or vice versa.
Pennsylvania's State House has 203 members, who are voted on every two years. In 2016, 113 Republicans and 77 Democrats won. The remaining candidates were listed on the ballot as a third party, often D/R; when they are added based on their caucus, the overall Republican House victory was 121 - 82. Republicans thus won 59.6% of the House seats; thanks to gerrymandering and asymmetric concentration of votes, they did this with only 52.7% of the total State House vote.
In the plot below, I compare each district's vote for President with its vote for State House. Points on the diagonal line represent districts where Clinton received the same percentage of the vote as the Democratic Representative (and Trump the same as the Republican). Points to the right of the plot voted for Clinton, to the left voted for Trump. Points to the top voted for the Democratic Representative, to the bottom the Republican.
Among the 190 districts that had at least one major party candidate, Clinton outperformed the Democratic Representative in 110, and Trump outperformed the Republican in 80. This suggests an anti-Trump sentiment, which was nonetheless overcome by the overall Republican vote.
Many of these districts still elected representatives of the same party. However, the points in the top left quadrant (above the x-axis and to the left of the y-axis) and the bottom right quadrant (belowthe x-axis and to the right of the y-axis) are districts that actually elected different parties in the topline results. Some 18 contested districts that Clinton won were won by the Republican Representative, compared to 10 districts that Trump won going to Democrats.
The districts at the very top and bottom of the plot are districts where the State House race was uncontested, and thus the gap was 100%. Amazingly, five districts with uncontested Democrats voted for Trump, and one district with an uncontested Republican voted for Clinton.
In the Philadelphia region, the only splits that occurred were voters selecting Clinton but a Republican representative. And there were a lot of them. Check out the interactive maps for more exploration.
What should we expect from the splits?
Democrats would need to win 20 seats to take control of the State House. What should we expect from these 18 that voted for Clinton but a Republican representative? That's an open question.
One line of reasoning is that these districts will continue to support their local Republicans. They clearly dislike President Trump, but even in 2016 that wasn't enough for Democratic representatives to win. In 2018, when Trump isn't even on the ballot, surely voters will continue to support their local, traditional Republicans.
On the other hand, support for Trump has set record lows for Presidential approval (only 39.5% of Americans approve as of this writing), and it appears that voters' distaste may be spilling over to the rest of his party. As Republican officials have fallen in line behind Trump during his first year, they've fallen decisively behind in generic ballots, with unnamed Democrat beating unnamed Republican by 8 points. Voters in the Philadelphia suburbs may have expected Hillary to win, and voted for the Republicans they were familiar with as the Republican Party would inevitably reject Trumpism. Instead, Trump won, and his takeover of the Party may sour these voters on their local Republicans, too.
There were 34 split districts in Pennsylvania in 2016, and even more that just barely missed the cut. Democrats need to gain 20 seats to take over the house. That gap, which once would have seemed insurmountable, may be in play.