Divvying up Chicago, Blue Wall time bomb

Plus: Assessing Perot's impact on the 1992 election, an aging Congress, Twitter's recommendation algorithm, and the politics of pro sports

No. 261 | April 7th, 2023

🗳️ Elections

Nathaniel Rakich: The 4 Political Neighborhoods Of Chicago (FiveThirtyEight)

"America’s cities are some of its most solidly Democratic areas — but that doesn’t mean they are solidly liberal. Over the past two years, the mayoral elections in our two biggest cities have boiled down to surprisingly tight contests between a moderate Democrat and a more liberal alternative.

Tuesday’s runoff election for mayor of Chicago, the nation’s third-most populous city, is no different. Former Chicago Public Schools CEO Paul Vallas rode his pro-police, tough-on-crime messaging to first place in the first round of voting on Feb. 28 but has faced accusations of being a Republican based on comments he made in 2009. And Cook County Commissioner Brandon Johnson finished second on a platform of taxing the rich to pay for more public services but has had to walk back his past support for decreasing police funding. (Dogged by COVID-19, rising crime rates, personality conflicts and broken promises, incumbent Mayor Lori Lightfoot didn’t even make it out of the first round, finishing third with 17 percent of the vote.)

But with Vallas (33 percent) and Johnson (22 percent) combining for barely half of the vote in February, a lot of voters are up for grabs in the runoff — and most Chicagoans don’t fit into neat ‘moderate’ or ‘progressive’ buckets. Chicago politics has more layers than a deep-dish pizza, and you need a more nuanced taxonomy to understand them — so we made one. “

@cincy9: Maps of the (very) preliminary Chicago Mayor precinct winner (left) and precinct turnout relative to February (right). On the relative turnout map, green = exceeded the 98.7% of Feb citywide overall votes cast so far; purple = Feb > Apr. A few patterns in the next Tweet. (Twitter)

@PatrickRuffini: Precinct-level demographic plurality of those projected to vote in today’s election in Chicago. Purple = Black, Blue = White College, Red = White Non-College, Orange = Hispanic, Green = Asian. (Twitter)

Senior Bread Price Fixing Engineer: Seven Chicagos, Defined (Medium)

“Anyone who knows anything about Chicago has heard some version of the ‘tale of two cities’ metaphor. Some explain it rather curtly (‘here’s an invisible border at Roosevelt Road’), others have described it in much more detail. But the core idea is always the same: there are two very distinct parts of Chicago; one is extremely rich, White and highly-educated, resembling the Upper East Side/Upper West Side of Manhattan, northern San Francisco or northern Atlanta. The other is extremely poor and mostly Black, like Detroit or Cleveland.

My view is that in order to capture the key fault lines of Chicago’s 77 neighborhoods, you need at least 6 or 7 buckets, not 2. So to get an idea of what that would look like, I assembled a summary of the demographic breakdown of each neighborhood using data from the CMAP data hub. I looked at four different demographic axes: race/ethnicity, age, income, and education.”

Michael Baharaeen: The Ticking Time Bomb in Democrats’ Blue Wall States (The Liberal Patriot)

Democrats were understandably relieved—if not thrilled—with the results of last year’s midterms. While conventional wisdom suggested they might lose seats up and down the ballot to Republicans, they held their own in hotly contested races for Congress, statewide constitutional offices, and state legislatures. This included marquee races in battleground states like Michigan, Pennsylvania, and Wisconsin, which collectively make up the party’s “Blue Wall” in presidential elections.

However, those results obscured a nagging problem for the party. In general elections since 2012, when President Obama won his second term, Democrats have not been quite as dominant in those three states’ major urban areas: Detroit, Philadelphia, and Milwaukee. These cities are vote-rich, home to large black populations, and deeply Democratic, but the party’s candidates have mostly underperformed Obama in them while voter turnout there has simultaneously lagged.

Harrison Lavelle, Armin Thomas: Examining Ross Perot’s Impact on the 1992 Presidential Election Results (Split Ticket)

Many political scientists now recognize that Ross Perot’s independent candidacy did not spoil the 1992 presidential election for Republican President George H.W. Bush, yet few comprehensive quantitative analyses exist to prove such conventional wisdom correct. 

While exit polls are by no means perfect measures of electoral preferences, they are often the best resources available for analyzing historical contests. We used the Voter Research Survey (VRS) exit polling data to quantify Perot’s true impact in the 1992 race for the White House while simultaneously examining the demographic composition of the presidential electorate.

@_JeaneLannes: Alright. Here's the big project I finished while I was hone. The 2016 Republican Primary mapped by county by vote share. Early on in the election it was still a competitive two horse race, but one Cruz dipped, Trump became the presumptive nominee winning the RNC and Presidency. (Twitter) 

Alex Samuels: Biden Is Moving Right On Immigration. Will That Hurt Him In 2024? (FiveThirtyEight)

“President Biden’s administration recently announced new asylum restrictions and other proposed stringent initiatives to deter an influx of migrants at the U.S.-Mexico border. But for some Democrats, the latest crackdowns stand in stark contrast to the ‘fair and humane immigration system’ that Biden once promised voters following the harsher policies of his predecessors, Donald Trump and Barack Obama.

A number of reporters (including my colleague Nathaniel Rakich) have speculated that Biden’s pivot to the center represents an attempt to shore up electoral weak spots as he prepares for his imminent reelection campaign; after all, Americans regularly tell pollsters that they trust Republicans on immigration more than Democrats. And, sure, there are likely practical considerations at play, too — the 2022 fiscal year set a record for migrant encounters with border patrol at the southern border with Mexico.”

🗽 Congress

Geoffrey Skelley: Congress Today Is Older Than It's Ever Been (FiveThirtyEight)

“Older members of Congress are notorious for their lack of familiarity with modern technology. Late last month, at least three different representatives in a hearing on TikTok called the popular app ‘Tic Tac,’ a breath mint available in many store checkout lines. This is only the latest in a long line of amusing tech-related congressional miscues: Back in 2006, Alaska Sen. Ted Stevens described the internet as ‘a series of tubes,’ and Senate Majority Leader Chuck Schumer copped to his struggles when in 2022 he held up his flip phone and said he was ‘not very tech-oriented’ during a speech on the Senate floor.

Don’t expect such unfamiliarity to change anytime soon: As it turns out, Congress today is older than it’s ever been. Across all senators and representatives, the median age of the 118th Congress is 59 years old. The median senator is 65 years old, a record high; the median representative is about 58, for the fourth Congress in a row. Congress has notably aged since 2001: From 1919 to 1999, the median senator never eclipsed 60 years old and the median representative never surpassed 55.”

🖥  Digital Data

Twitter's Recommendation Algorithm (Twitter)

“Twitter aims to deliver you the best of what’s happening in the world right now. This requires a recommendation algorithm to distill the roughly 500 million Tweets posted daily down to a handful of top Tweets that ultimately show up on your device’s For You timeline. This blog is an introduction to how the algorithm selects Tweets for your timeline.

Our recommendation system is composed of many interconnected services and jobs, which we will detail in this post. While there are many areas of the app where Tweets are recommended—Search, Explore, Ads—this post will focus on the home timeline’s For You feed.

Katie Harbath: Whirlwind - A look back at the last three months (Anchor Change)

“Ahhhh, I’m so excited and nervous about this newsletter. 😱💖🤓

For the longest time, I’ve been thinking about doing a quarterly slide deck about what’s happened at the intersection of technology and democracy. I wanted to make my version of what Mary Meeker used to do with her internet trends reports or what Bruce Mehlman does to explain what’s happening in Washington.

I finally did it. 

Introducing the first Anchor Change Tech & Democracy Quarterly Report.” 

🤖 Artificial Intelligence 

@emollick: The new BloombergGPT AI may be harbinger of the next wave of corporate AI. Current AIs are trained on web data (though firms can add their own training). BloombergGPT is 52% either proprietary data or cleaned financial data. And it shows signs of being better at financial tasks. (Twitter)

@emollick: New randomized trial with early (pre-ChatGPT) AI, and another result showing unprecedented increases in performance. Salespeople with AI assistance doubled customer purchases & were 2.33 times as successful in solving questions that required creativity. Top agents gained most. (Twitter)

📈  Data Science

Emma Remy, Gonzalo Rivero: How we review code at Pew Research Center (Pew Research Center)

“At Pew Research Center, we work hard to ensure the accuracy of our data. A few years ago, one of our survey researchers described in detail the process we use to check the claims in our publications. We internally refer to this process as a ‘number check’ – although it involves far more than simply checking the numbers.

As part of our quality control process, we also verify that the code that generates results is itself correct. This component of the process is especially important for our computational social science work on the Data Labs team because code is everywhere in our research, from data collection to data analysis. In fact, it is so central that on Labs we have spun off the review of our code into two distinct steps that precede the number check: a series of interim reviews that take place throughout the lifetime of a project, and a more formalized code check that happens at the end. This is our adaptation of what software developers call ‘code review.’

This post describes that process – how it works, what we look for and how it fits into our overall workflow for producing research based on computational social science.”

📰 Media Habits

@WillFeltus: @MikePShannon and I updated our sports bubbles. The NBA and WNBA are to Democrats as NASCAR and rodeo are to Republicans. Tennis is to Democrats as golf is to Republicans. Note for political advertisers: fans of almost every major sport are more likely to vote. (Twitter)

👫 Demographics

Nicole Acevedo: Large share of Latinos don't identify with current race categories, Census Bureau numbers show (NBC News)

“A large share of the U.S. Latino population doesn't identify with any of the current racial categories in the census, according to new 2020 Census Bureau data that shows "major shifts" in how Americans who identify as Hispanic report their race.

While almost 60% of the 54.6 million Americans who identified as Hispanic reported belonging to one racial group, such as white or Black, over a third (35.5%) of Latinos chose “Some Other Race" alone. This category is currently not recognized as a race by the federal government. The Census Bureau said a combined 43.6% of Americans who self-identify as Hispanics either reported being of ‘Some Other Race’ (35.5%) or did not respond to the race question in the 2020 count (8.1%).

The findings come as the Biden administration is considering allowing Americans to check off ‘Hispanic or Latino’ as their race as well as their ethnicity as part of new proposed classifications for the next census.”

Ryan P. Burge: Gen Z and Religion in 2022 (Religion in Public)

For the last nine months straight, the most read post on Religion in Public is: Gen Z and Religion in 2021. Well, the Cooperative Election Study just released the raw data from their 2022 collection wave which happened in October and November of last year. That means that we can see just how much religious affiliation has shifted between 2021 and 2022. 

What’s nice about the 2022 Cooperative Election Study is its consistency. Same questions. Same response options. Same survey mode. This last collection included 60,000 survey respondents  –  more than enough to do this type of fine-grained analysis. 

Alex Fitzpatrick, Kavya Beheraj: See the fastest growing (and shrinking) U.S. states (Axios)

“Idaho, Montana and Florida saw the highest population growth among U.S. states between 2020-2022, per new U.S. Census Bureau data, while New York, Illinois and Louisiana suffered the most shrinkage.

The big picture: The past few years have been especially turbulent for population trends, with the COVID-19 pandemic affecting birth and death rates, interstate and international migration, and more.

By the numbers: Idaho's population grew by nearly 4.9%, while that of Montana and Florida grew by 3.3% and 3.0%, respectively. Utah and South Carolina came in just a hair under 3%.”

🏀 Sports

@JeffPassan: The first four days of the MLB season, by the numbers: (Twitter)