Americans’ Views Of The Economy Are Partisan, But They’re Not Immune To Bad News

Welcome to Pollapalooza, our weekly polling roundup.

Poll of the week

“How is the economy doing?” is a question that we have several hard, objective metrics with which to answer. But increasingly, how Americans think the economy is doing seems to depend on their party. But just because views of the economy are polarized, that doesn’t mean they’re stagnant — and that could prove pivotal for President Trump in 2020.

Quinnipiac University’s latest national poll of registered voters, conducted Aug. 21-26, found a huge gap between how Republicans and Democrats perceive the state of the economy: 43 percent of Republicans described it as “excellent,” and another 45 percent described it as “good”; among Democrats, however, just 2 percent described the state of the economy as “excellent” and 37 percent said it was “good.” A plurality of Democrats, 43 percent, thought it was “not so good,” while 17 percent called it “poor.”

In total, 88 percent of Republicans used a positive adjective for the state of the economy, while only 39 percent of Democrats did the same. That 49-point gap in how the two parties perceived the strength of the economy was the second-largest that Quinnipiac has measured in the last three years.10 (The largest gap came this January, amid the government shutdown.)

Granted, it’s not unusual for voters’ politics to color their perceptions of the economy. During the Obama administration, Democrats were generally more optimistic about the economy than Republicans were; that flipped a few months into Trump’s tenure. Since then, Republicans have consistently viewed the economy more positively than Democrats have, but as the chart above shows, the gap between parties has really widened over time.

Most of this is due to Republicans warming to the economy, going from having a slight majority with positive views at the beginning of the Trump administration to near-universal approval today. Under the surface, Republicans haven’t just gotten more likely to feel positively about the economy, they’ve also been reporting more strongly positive feelings. As late as April 2018, only 18 percent of Republicans said the state of the economy was “excellent” and 65 percent said it was “good.” In May 2019, 50 percent rated it “excellent” and 42 percent rated it “good.” That said, a bit of that enthusiasm did wear off in the latest poll, when the “excellent” number dropped by a few points. This could be the result of heightened anxiety that a recession might be around the corner.

But while Republicans’ views on the economy have largely been on a steady march upward, Democrats’ views appear to be more sensitive to events. In February 2018, after Trump’s first State of the Union address and a strong month in the stock market, 62 percent of Democrats had a positive view of the economy, but at moments like the January 2019 government shutdown or this most recent poll, positive views of the economy among Democrats have dipped to as low as 39 percent. But electorally, Republicans’ and Democrats’ views on the economy may not matter all that much; their 2020 votes are largely already set in stone.

So what’s perhaps more troubling for Trump is that independents’ opinions on the economy look a lot like Democrats’ — they often react to current events in a similar way, though their recent baseline is about 20 points higher. While it’s plausible that partisan polarization is so strong these days that even a recession would not change many voters’ minds about Trump, the fact that independents appear persuadable on the economy is a point in favor of the theory that a recession would indeed damage his reelection chances.

Indeed, it might be a coincidence that Trump’s overall approval rating has ticked down in recent weeks amid speculation about a looming recession, but it also comes at a time when only 57 percent of independents have kind things to say about the state of the economy — the lowest number Quinnipiac has found since the summer of 2017.

Other polling bites

  • As long as we’re on the topic of partisan gaps, there is also one on whether the U.S. should spend more money for scientific research. According to the Pew Research Center, 62 percent of Democrats and people who lean Democratic would increase federal spending on scientific research, but only 40 percent of Republicans and Republican-leaning respondents agreed. That said, that’s still more support among both parties than Pew previously found in almost two decades of polling.
  • Marquette University Law School — one of the best pollsters in a crucial swing state — released its latest poll of Wisconsin this week, and residents gave Trump a 45 percent approval rating and a 53 percent disapproval rating. Interestingly, 37 percent of Wisconsinites said they thought the economy has gotten better over the last year, and 25 percent said they thought it had gotten worse — but when asked how they thought the economy would change in the next year, the results were reversed. Just 26 percent thought things would get better, and 37 percent thought they would get worse.
  • Tune in next week for a full preview of the upcoming do-over election in the North Carolina 9th Congressional District, but don’t overlook the special election in the North Carolina 3rd District that’s happening at the same time. That race has gotten less attention, but this week the Republican blog RRH Elections released a poll of that race that gave GOP state Rep. Greg Murphy a 51-40 lead over Democrat Allen Thomas in this bright-red district. (According to FiveThirtyEight’s partisan lean metric,11 it’s 24 points more Republican-leaning than the nation as a whole.)
  • Just in time for CNN’s climate town halls with the Democratic presidential candidates, the Sierra Club and Morning Consult released a poll of “climate voters” — those who say that the candidates’ climate plans are “very important” to their vote. And the results looked a lot like polls of the Democratic field overall, with former Vice President Joe Biden at 30 percent, Sen. Elizabeth Warren at 21 percent and Sen. Bernie Sanders at 20 percent.
  • Turmoil on the Thames! Opponents of a no-deal Brexit successfully took control of the parliamentary agenda away from British Prime Minister Boris Johnson and advanced a bill to prevent the United Kingdom from exiting the European Union without an agreement in place. In response, Johnson wants to call a new parliamentary election, and British voters tilt slightly in his favor. A YouGov poll says 41 percent of Britons want their member of Parliament to vote to hold a new election in mid-October, and 31 percent want their MP to vote against such a measure.

Trump approval

According to FiveThirtyEight’s presidential approval tracker, 41.4 percent of Americans approve of the job Trump is doing as president, while 53.9 percent disapprove (a net approval rating of -12.5 points). At this time last week, 41.3 percent approved and 54.2 percent disapproved (for a net approval rating of -12.9 points). One month ago, Trump had an approval rating of 42.4 percent and a disapproval rating of 53.2 percent, for a net approval rating of -10.8 points.

Generic ballot

In our average of polls of the generic congressional ballot, Democrats currently lead by 6.5 percentage points (46.3 percent to 39.8 percent). A week ago, Democrats led Republicans by 6.7 points (46.4 percent to 39.7 percent). At this time last month, voters preferred Democrats by 6.1 points (46.1 percent to 40.0 percent).

Check out all the polls we’ve been collecting ahead of the 2020 elections.

If Vegas Predicts Your NFL Team For Six Wins, You May Be In Luck

Every year, we look back on preseason win totals produced by forecasters and betting markets and chuckle at some of the more egregious misses. Last season, the Chicago Bears were initially forecast for seven wins by Las Vegas, then traded for Khalil Mack and somehow won 12. The Green Bay Packers’s predicted win total was 10, but they melted down in spectacular fashion and ultimately ended up winning just six games.

We’ve already published our Elo projections, and we think they’re the best we’ve ever produced for the NFL, but there will still be lots of misses to grouse about come January. Forecasting a sport as luck-driven as the NFL is rough that way.

It raises the question: How good are betting markets at predicting team wins? To find out, I got my hands on a tranche of win prediction data stretching back to 1989, courtesy of Sports Odds History, and checked how well Vegas preseason win totals predict actual team wins. While Vegas overall does a good job identifying good and bad teams, it turns out that at the lower end of the range of projected wins, Vegas predictions don’t seem particularly well calibrated — through the confidence intervals at the lower end are large because of the small sample size, so the results aren’t statistically significant.

Projected win totals of six and fewer undersell teams’ prospects by about a win on average, with the exception of teams forecast for five wins.

Win totals don’t change as frequently as the moneyline odds, so we probably shouldn’t take win totals at face value — at least for teams with low projected wins. What does this mean for non-bettors? It should be decent news for the Cincinnati Bengals, New York Giants, Oakland Raiders and Washington Redskins — teams that both Elo and Vegas have pegged for six wins in 2019 — since we should be more bullish on their chances than we currently are.

Optimism for these probable cellar dwellers might feel forced. But we should fight the urge toward overconfidence, especially in the face of history. A few of these teams will end up surprising us — in a good way — at the end of the year for reasons inscrutable to us now.

Which NFL teams might beat expectations?

Average actual wins (1989-2018) by Vegas preseason expected wins, and the 2019 teams at each number of expected wins

wins
Expected Actual 2019 teams
11.0 10.2 New England
10.5 9.9 Kansas City, L.A. Rams, New Orleans
10.0 8.8 Philadelphia
9.5 9.0 Chicago, Green Bay, L.A. Chargers
9.0 8.5 Cleveland, Dallas, Indianapolis, Minnesota, Pittsburgh
8.5 8.9 Atlanta, Baltimore, Houston, Seattle
8.0 7.4 Jacksonville, San Francisco, Tennessee
7.5 7.6 Carolina
7.0 6.9 Buffalo, Denver, N.Y. Jets
6.5 6.4 Detroit, Tampa Bay
6.0 6.7 Cincinnati, N.Y. Giants, Oakland, Washington
5.0 4.6 Arizona, Miami

Sources: SportsOddsHistory.com, Greg Guglielmo, Pinnacle, Betfair, William Hill, Bet365, BetOnline

Well … maybe not the Bengals. Not only is Cincinnati saddled with an injured A.J. Green, who appears to be out until around Week 8, the Bengals have an offense that is bereft of top talent at nearly every position. Cincinnati replaced head coach Marvin Lewis after 16 seasons of on-again, off-again contention and turned instead to Zac Taylor, a coach best known for being friends with L.A. Rams wunderkind Sean McVay. The hope must be that Taylor can revitalize the career of quarterback Andy Dalton, who sports a middling career yards per attempt of 7.2 and is one of the few starting quarterbacks who Vegas believes wouldn’t move a line if he were to be replaced in the lineup. The defense doesn’t offer a compelling reason for optimism: The Bengals ranked 28th in defensive Defense-adjusted Value Over Average (DVOA) last season. Perhaps we should view that as a reason to be bullish on their prospects in 2019 simply due to regression, since defensive performance year to year isn’t terribly stable. If that seems like a bridge too far, magic might be the answer: Taylor may give lip service to the notion that he isn’t trying to be like his mentor McVay, but McVay’s brand of QB sorcery seems like the best hope for the Bengals to crest seven wins this year.

The Giants are more interesting. After a promising preseason performance by first-round pick Daniel Jones, New York fans are clamoring for a change of the guard at quarterback. As big of a reach as many believed Jones to be, I still see him as a better use of first-round draft capital than “generational talent” Saquon Barkley. Hailed as a potential savior and the missing piece for Eli Manning’s final championship push, Barkley helped the Giants improve from a terrible three-win team in 2017 to a merely bad five-win unit in 2018.

The Giants were second-worst in the league on Expected Points Added per play on first-down play-action passes after adding Saquon to the backfield,1 and prospects for a bounceback in play-action efficiency seem bleak. After trading all-world wideout Odell Beckham Jr. to the Browns, the Giants lost free agent acquisition Golden Tate to a four-game suspension for performance-enhancing drugs and their expected No. 3 wideout Corey Coleman to a season-ending ACL injury. Their best hope for a productive season may rest in ownership’s willingness to bench Manning for good this time.

The other team to somehow accumulate negative value on first-down play action was Oakland. In what seems to be a pattern for teams at the bottom of the win total forecast, Vegas sees Derek Carr as a quarterback worth just 1 point to the spread. The stats back up that view. Carr’s career yards per attempt is, at 6.7, below league average, and his best season as judged by QBR is an anemic 54.6. His weapons are improved from a year ago, but they are volatile. New Raiders wideout Antonio Brown sat out of practice because he wasn’t allowed to wear a helmet the NFL deems dangerous and is now likely to be suspended for some period of time, and Tyrell Williams is a boom or bust weapon who likes to be targeted deep — something Carr may be reluctant to do given his career average depth of target of just 7.7 yards. Meanwhile “Hard Knocks” captured head coach Jon Gruden disparaging “all the football stats and all the fantasy bullshit” in favor of running backs that will “BOOF” the opposing team in pass protection. Of all the six-win teams, Oakland may be the most unpredictable — and that unpredictability could manifest itself in good ways, as well as bad. Brown’s antics could end with a fashionable and safe new helmet, Carr might be coaxed into throwing the deep ball to a talented field stretcher, and Gruden might use rookie running back Josh Jacobs optimally, leading to wins we simply can’t foresee at this point.

The final team projected for six wins in 2019 is Washington, a team that somehow came to the determination that Mark Sanchez and Josh Johnson were better choices than Colin Kaepernick to take over for quarterback Alex Smith when his 2018 season — and perhaps his career — ended with a gruesome leg injury.

In the draft, Washington team president Bruce Allen added Ohio State QB Dwayne Haskins in the first round but then failed to surround him with receiving weapons. Jamison Crowder left via free agency, former first-round bust Josh Doctson was released at the end of the preseason and tight end Jordan Reed suffered another concussion heading into Week 1. Their current starting wide receivers are third-round pick Terry McLaurin — also from Ohio State — and Paul Richardson.

The outlook at running back is brighter with the return of Derrius Guice from an ACL tear that derailed his rookie season, but there is little evidence to suggest they will put him in advantageous spots to run the ball. With the ageless, tackle-breaking cyborg Adrian Peterson in 2018, Washington lined up against neutral or stacked boxes on first-and-10 or second and long 174 times, decided they liked the look and ran right into the scrum 72 percent of the time. But if Washington can flip the script on downs tailor-made for passing and eke out some yards where they should come easy, the duo of Guice and Peterson could be enough to protect current starter Case Keenum or rookie Haskins while he learns on the job — and possibly beat the team’s six-win projection.

Can Smiling Really Make You Happier?

Before we get started, do me a favor and grab a pen or a pencil. Now hold it between your teeth, as if you were about to try to write with it. Don’t let your lips touch it. Sit with it, and pay attention to how you feel. Are you glum? Cheerful? Confused? Is that any different than how you felt before? Do you feel like this weird smile tricked your brain into a slight jump in happiness?

For a long time, psychologists thought exercises like this one did make us happier. If that were true, it would have implications for what emotion is, how we experience it and where emotions come from. Psychologists have believed that “facial feedback” from emotional expressions like smiling (or frowning) gives the brain information that heightens, or even sparks, an emotional experience.

It made so much sense that it was almost too good to check.

But then scientists did check. What they found poked holes in one of psychology’s textbook findings — which raised a whole new set of questions. Now, a huge group of scientists has banded together to try to get to the bottom of smiles, even if it means working with people who think they’re wrong.

The idea that smiling can make you feel happier has a long history. In 1872, Darwin mused about whether an emotion that was expressed would be felt more intensely than one that was repressed. Early psychologists were musing about it in the 1880s. More than a hundred studies have been published on the topic. And it’s a trope of pop wisdom: “Smile, though your heart is aching,” sang Nat King Cole in 1954. “You’ll find that life is still worthwhile, if you’ll just smile.”

In 1988, social psychologist Fritz Strack published a study that seemed to confirm that facial feedback was real. The researchers asked participants to do more or less what I asked you to do earlier: hold a pen in their mouths in a position that forced them either to bare their teeth in a facsimile of a smile or to purse their lips around the pen. To ensure that no one was clued in to the researchers’ interest in smiles, the experimenters told participants that they were exploring how people with physical disabilities might write or perform other ordinary tasks.

When both groups were shown a set of newspaper comics — specifically, illustrations from Gary Larson’s The Far Side — the teeth-barers rated the images as funnier than the lip-pursers did. This was a big deal for the facial feedback hypothesis: Even though participants weren’t thinking about smiling or their mood, just moving their face into a smile-like shape seemed to affect their emotions. And so the finding made its way into psychology textbooks and countless news headlines. Decades of corroboration followed, as researchers published other experiments that also showed support for the facial feedback hypothesis.

But in 2016, all at once, 17 labs failed to replicate the pen study.

The pen study had been solid — until it wasn’t.

Those 17 studies, coordinated by Dutch psychologist E.J. Wagenmakers, repeated the original study as closely as possible to see if its result held up, with just a few changes. They found a new set of cartoons and pre-tested them to check they were about as funny as the old set. They also changed how they checked up on the participants’ pen technique: The original had an experimenter watching over things, but Wagenmakers and his team filmed participants instead.

When all 17 studies failed to replicate the original result, the effect was “devastating for the emotion literature,” said Nicholas Coles, a psychology grad student whose research focuses on the facial feedback effect. “Almost all emotion theories suggest that facial feedback should influence emotions.” While there are plenty of other methods for looking at facial feedback, many of them are more likely to make participants figure out the real purpose of the experiment, which makes their results trickier to interpret. The pen study had been solid — until it wasn’t.

These kinds of failed attempts to replicate other researchers’ results have been piling up in psychology’s “replication crisis,” which has called the reliability of psychology’s back catalogue into question. Past experiments may be unreliable because they relied on small sample sizes, buried boring or inconclusive results, or used statistical practices that make chance findings look like meaningful signals in what is really random noise. The result has been a morass of uncertainty: Which findings will hold up? And when one doesn’t, what precisely does that mean?

Wagenmakers and his team are just one of the many collaborations hoping to reshape psychology in the image of more established sciences like physics and genetics, where huge international consortia are already commonplace. Some collaborations, like the “Many Labs” projects, conduct multi-lab replications similar to the attempt to confirm the pen study and cover a broad swath of famous psychology studies. Others — like the ManyBabies Consortium, which conducts infant research — concentrate on a niche.

Then there’s the Psychological Science Accelerator, which is more focused on creating the infrastructure for collaboration, allowing its members to democratically elect studies to be run across its network of 548 labs in 72 countries. A recent paper by a group of reforming researchers called this kind of crowdsourced science one of the routes to “scientific utopia.”

Across six multi-lab replication projects, each trying to replicate multiple studies, only 47 percent of the 190 original results were successfully replicated. The failed attempt to replicate the pen study is in good company.

But as powerful as multi-lab replication efforts like these are, they aren’t necessarily the last word. When psychology tries to solve its replication crisis, it can sometimes create a crisis of a different kind, opening up a knowledge vacuum where an apparently reliable finding had previously stood.

Fritz Strack, the lead researcher on the original pen-in-mouth study from 1988, doesn’t think that Wagenmakers’s study tells us all that much — the world is constantly changing, and re-running an old experiment could produce new results not because the idea being tested is flawed but because the experiment itself is now out of step with the times. Although he suggested the replication effort himself, and advised on the design and the materials of the study, he refused to be fully involved. Instead, he said, he wanted the freedom to comment on the problems as he saw them without pulling any punches.

“Things are changing — times are changing, the zeitgeist is changing, the culture is changing, the participants are changing. It’s not under your control.”

When the results were released, Strack found plenty of things to critique. He was concerned that newspaper cartoons would not have packed the same humor punch these days that they did in the Midwest of the 1980s. The filming, he said, was another problem: It could be that filming made participants unusually self-conscious, affecting their experience of the task.

Strack thinks that it’s a mistake to focus on testing a method rather than a hypothesis. A method that fails might have been a bad test of the hypothesis, but the hypothesis is really what counts.

In this case, the hypothesis was that facial feedback can create an emotional effect even when people aren’t aware that their facial expression is an emotional one. Perhaps, Strack argued, his exact methods from the 1980s are no longer the best way to test that.

“Exact” replications are impossible, he said. “Things are changing — times are changing, the zeitgeist is changing, the culture is changing, the participants are changing. It’s not under your control.” What if you did the pen study with memes instead of cartoons? What if you didn’t use cameras? What would the differences tell us about facial feedback and when it comes into play?

Strack has been vocally critical of the credibility revolution, arguing that the term “replication crisis” is overblown. He says he prefers to focus on arguments about the quality of the research methods, rather than the statistical framework that is at the core of the credibility revolution’s concerns.

But similar critiques of massive replications come from inside the movement. Psychologist Tal Yarkoni, an ardent reformer, thinks that large-scale research efforts would do more good if they were used to test a huge array of different ways of getting at a question. A failed attempt to replicate a particular experiment doesn’t really tell you anything about the underlying theory, he said; all it tells you is that one particular design works or doesn’t work.

Wagenmakers doesn’t think his team’s replication is the final word on the facial feedback theory, either. “It’s a sign of good research that additional questions are raised,” he said. But he does think a failed replication like the one he led shifts the burden of proof. Now, he says, proponents of the facial feedback hypothesis should be the ones coming to the table with new evidence. Otherwise, “the replicating team will be like a dog playing fetch,” he said. “A person throws a ball and the [replication] team brings it back, but oh, it’s not quite right! I’m going to throw it in another direction. … It could go on forever. It’s clearly not a solution to the problem.”

Multi-lab studies can look large and impressive, said psychologist Charles Ebersole, who coordinated two Many Labs projects in grad school. Even so, it’s not clear how much confidence people should have in their results — the studies are big, which can improve confidence in their outcomes, but they’re subject to flaws and limitations just like smaller studies are. “Some people do an excellent job of not listening to [multi-lab studies] at all; maybe that’s the right answer? Some people bet a lot on them; maybe that’s the right answer? I don’t know.”

The way out of the replication crisis clearly isn’t brute replication alone.

When Wagenmakers and his colleagues published their replication study in 2016, Coles was digging deeply into the facial feedback literature. He planned to combine all of the existing literature into a giant analysis that could give a picture of the whole field. Was there really something promising going on with the facial feedback hypothesis? Or did the experiments that found a big fat zero cancel out the exciting findings? He was thrilled to be able to throw so much new data from 17 replication efforts into the pot.

He came up from his deep dive with intriguing findings: Overall, across hundreds of results, there was a small but reliable facial feedback effect. This left a new uncertainty hanging over the facial feedback hypothesis. Might there still be something going on — something that Wagenmakers’s replication attempt had missed?

If the studies thrown into the mix aren’t great to start with, the result isn’t particularly reliable — or, as Coles put it, “crap in, crap out.”

Coles didn’t think that either Wagenmakers’s replication or his own study could put the matter to rest. The technique he used, called a meta-analysis, comes with its own problems. Specifically, if the studies thrown into the mix aren’t great to start with, the result isn’t particularly reliable — or, as Coles put it, “crap in, crap out.”

So he set about designing a different kind of multi-lab collaboration. He wanted not just to replicate the original study, but to test it in a new way. And he wanted to test it in a way that would convince both the skeptics and those who still stood by the original result. He started to pull together a large team of researchers that included Strack. He also asked Phoebe Ellsworth, a researcher who was testing the facial feedback effect as far back as the 1970s, to come on board as a critic.

This partnership founded in disagreement is meant to get the game of fetch out of the way before the study even gets off the ground. Coles’s group, called the Many Smiles Collaboration, is far from the only one using this tactic; although some massive collaborations try to replicate old studies as closely as possible, others choose to workshop a new experiment methodology in excruciating detail before pulling the trigger. Ideally, this means that everyone will be convinced by the results, regardless of what they were personally rooting for or expecting. “It isn’t groupthink,” said Coles. “We’re actually trying to get at the truth.”

The Many Smiles Collaboration is based on the pen study from 1988, but with considerable tweaking. Through a lengthy back-and-forth between collaborators, peer reviewers and the journal editor, the team has refined the original plan, eventually arriving at a method that everyone agrees is a good test of the hypothesis. If it finds no effect, said Strack, “that would be a strong argument that maybe the facial feedback hypothesis is not true.”

An early pilot of the Many Smiles study indicated that the hypothesis might not be on its last legs just yet: The results suggested that smiling can affect feelings of happiness. Later this year, all the collaborators will kick into gear to see if the pilot’s findings can be repeated across 21 labs in 19 countries. If they find the same results, will that be enough to convince even the skeptics that it’s not just a fluke?

Well … maybe. A study like Wagenmakers’s sounds, in principle, like enough to lay a scientific question to rest, but it wasn’t. A study like Coles’s sounds like it could be definitive too, but it probably won’t be. Even Big Science can’t make science simple. “I’m still a little unsure, even though I’ve now replicated the effects successfully in my own labs,” said Coles. “I’ll hold my breath until the full data set comes in.”

What Happens When The FEC Can’t Do Its Job?

Welcome to a special edition of FiveThirtyEight’s weekly politics chat. The transcript below has been lightly edited.


sarahf (Sarah Frostenson, politics editor): Last Monday, Federal Election Commission Vice Chairman Matthew Petersen announced his resignation, leaving the FEC effectively shut down, as only three positions on the six-seat committee are currently filled and the agency is legally required to have four commissioners to be fully operational.

What this means is that the agency responsible for both enforcing and advising on the nation’s campaign-finance laws is out of commission for the foreseeable future. And because we’re in the middle of a presidential election … things could get hairy fast.

The FEC has said that it will soldier on, continuing to process filings and other reports, and has called on President Trump to nominate new commissioners and for the Senate to confirm them quickly. But Congress is still in recess and Trump has yet to move forward with appointing new commissioners (remember, there are now three vacancies).

So here with us today to unpack what this could mean for the 2020 election (and campaign finance in general) is Dave Levinthal, editor and senior reporter at the Center for Public Integrity.

Welcome!

dave.levinthal: My pleasure! Thanks for having me.

sarahf: So, first of all, how did we get to the point that the FEC is basically not operational? And just how big of a deal is this for the FEC?

dave.levinthal: You can trace the situation back to 2008, the last time the FEC found itself in semi-shutdown mode because it lacked enough commissioners to legally conduct high-level business.

That year, the FEC went about six months without a quorum of commissioners, until the Senate and President Bush finally struck a deal to appoint new commissioners and get the agency back on track.

clare.malone (Clare Malone, senior political writer): And by “high-level business,” we mean things like opening investigations into possible campaign-finance violations?

dave.levinthal: High-level business would absolutely include investigating allegations of campaign-finance violations. As I wrote last week:

For now, the FEC can’t conduct meetings.
It can’t slap political scofflaws with fines.
It can’t make rules.
It can’t conduct audits and approve them.

sarahf: So … what can the FEC do in its current situation?

dave.levinthal: The most notable thing the FEC will continue to do is carry out its transparency function. That means that political committees, political candidates and so on must still file their periodic campaign-finance disclosures with the FEC — documents that tell the public how much money they’ve raised, spent, etc. — and the FEC staff will still review and post that material.

But if some political committee screws this up, or for that matter acts in a manner that’s potentially in violation of federal campaign-finance laws, they more or less get a temporary pass because the FEC commissioners don’t have the power for now to do anything about it.

clare.malone: Cool. I was interested to learn that there’s been a lot of discord on the commission for a while. A Democrat and an independent on the committee were apparently irritated that certain investigations they deemed worthy weren’t being looked into because the Republican members kept things from moving forward. So in some ways, it sounds like this is the continuation of an already contentious situation at the FEC.

nrakich (Nathaniel Rakich, elections analyst): And Dave, without enforcement mechanisms, there would be no punishment if a campaign does violate campaign-finance laws, right?

dave.levinthal: Generally speaking, the FEC’s law enforcement capabilities are on ice until the Senate approves at least one more nominee to serve on the FEC.

Right now, Trump has nominated one commissioner — a Texas attorney named Trey Trainor who helped stop an anti-Trump movement at the 2016 GOP convention — who he first nominated in September 2017. But the Senate has yet to give Trainor a confirmation hearing, much less going forth and confirming him.

clare.malone: So … why no hearing? I haven’t really seen an explanation for that in everything I’ve been reading.

dave.levinthal: A complex question! My best crack at it: There’s been a tradition — often adhered to, but not always — that the president would nominate FEC commissioners in pairs: one Republican, one Democrat.

But since President Trump has only offered one nominee, the Senate has chosen not to give that lone nominee a hearing. Could it? Sure. Did it need to? Not really — until now — because the FEC has had enough commissioners to at least conduct its high-level business.

nrakich: I find it interesting that, in an age where the Senate has gotten more comfortable with consolidating power within one party (eliminating filibusters for presidents’ nominees, blocking Merrick Garland’s confirmation to the Supreme Court, etc.), that this norm of nominating one Democrat and one Republican simultaneously to the FEC has persisted.

dave.levinthal: Numbers at the FEC commission level have been going the wrong way since the beginning of the Trump presidency. In March 2017, Democratic Commissioner Ann Ravel resigned. And in February 2018, Republican Commissioner Lee Goodman also resigned. That means the FEC has been operating with a bare minimum four commissioners for 1.5 years.

President Trump could have nominated people filled those vacancies at any time, but he didn’t.

clare.malone: Whew.

Dave, champion explanation.

nrakich: Not to be too cynical, but in practical terms, how much of an effect does this shutdown really have? The FEC’s enforcement mechanisms are already pretty toothless and can take years to be resolved anyway.

For example, earlier this summer, now-Sen. Martha McSally was fined for campaign-finance violations she made in … 2014.

She has served two full terms in the House since then.

And despite breaking the rules by taking $319,000 in excess contributions, she was fined only $23,000. So there was no financial disincentive.

clare.malone: Right, and we didn’t learn about Trump’s campaign-finance violations until well after 2016, so there does seem to be a long lag time on this stuff!

dave.levinthal: I’ve heard from more than a few folks who’ve made that very point — that the FEC is already so dysfunctional that there won’t be much difference.

But even if the FEC deadlocks on investigations, even if it’s unable to make affirmative rulings on whether someone broke the law — and this is often the case — at minimum these situations receive a full public airing. For instance, if special interest groups or others vehemently disagree with an FEC ruling, they’ll sue the organization.

Think of all the big court cases that have “FEC” in their names, with Citizens United v. FEC the biggest among them. Without a functioning FEC, this process grinds to a halt, for all intents and purposes.

clare.malone: Got it. So the shutdown is really affecting the transparency of the FEC.

nrakich: I find it interesting, though, that voters don’t seem to care too much about campaign-finance violations. There’s some research suggesting that they don’t do as much as, say, sex scandals (probably for obvious reasons — they’re much drier!) to hurt candidates at the ballot box.

And I assume that most candidates will continue filing disclosure reports even if the FEC is shut down at the filing deadline. But what if they don’t? Would they really suffer any consequences with voters? I’m not sure they would, and that’s scary to me.

clare.malone: I think that’s in part because it’s so much a part of American political culture — and the culture at large — to see big money and politics as linked. There’s a perception that there’s a degree of unfettered spending going on.

dave.levinthal: While the FEC doesn’t have a quorum, if a political committee wanted to stop filing campaign-finance reports or otherwise violate campaign-finance rules, the FEC would not be in any position to do something about it.

Now, the FEC may very well pick up the matter once it’s in business again. But for now, political committees don’t really have anyone policing their activity in a way that would lead to some immediate penalty.

Said another way: The cops are at the station, they’re doing paperwork, but they’re not answering emergency calls.

sarahf: So given how critical the situation is, won’t one of Congress’s first orders of business be appointing the one commissioner Trump has already nominated to get the FEC back up and running?

dave.levinthal: Actually, Sarah, I don’t think there is a reason to think that Congress will make the FEC its first order of business. That’s not to say that the Senate won’t act quickly. But this is largely in the hands of Majority Leader Mitch McConnell and, to a lesser extent, Minority Leader Chuck Schumer.

Also, President Trump plays a major role here. He could theoretically nominate a full slate of new commissioners. He could clean house — float six new nominees. But presidents have largely missed opportunities to proactively replace FEC commissioners. The result: The remaining commissioners continued to serve in “holdover status” — serving even though their term has expired.

sarahf: 😱

dave.levinthal: Bottom line? If Trump wanted to defy political convention — something he’s not exactly shy about doing — he could nominate six new commissioners of his choosing.

Legally, he can’t pack the committee with Republicans — no more than three commissioners can be of the same party — but he doesn’t have to nominate Democrats. He could theoretically nominate three Republicans and three … Libertarians. Or independents.

nrakich: Right. I feel like that would be well within character — but also, I feel like Trump doesn’t spend a lot of time thinking about the FEC.

sarahf: Or any president, it sounds like.

So if you’re Petersen, why resign now, knowing it would throw the FEC into chaos? Was his resignation a surprise?

dave.levinthal: No, Petersen almost left in late 2017, when Trump nominated him for a federal district judgeship. But Petersen flamed out of the judicial job when he was unable to answer a string of basic questions at his nomination hearing, and he withdrew himself from consideration shortly thereafter. His consolation prize? He continued to serve on the FEC.

At the time — December 2017 — CPI wrote that even then it seemed like the FEC had a strong possibility of losing its quorum of commissioners.

So it’s not as if this problem somehow snuck up on people.

nrakich: Yeah, I keep coming back to the fact that the public doesn’t seem to care about campaign finance.

It’s a big problem, IMO, since this is one of the main mechanisms by which we keep candidates accountable.

clare.malone: Well, very few people ever end up serving jail time for these violations, for instance. And unlike a sex scandal, a lot of people will be bored reading about campaign-finance violations.

dave.levinthal: But there are some money-in-politics sex scandals! (cough Stormy Daniels cough) And the FEC plays a role in those … or could.

nrakich: Right, and there are some types of campaign-finance scandals, like when politicians (for example, California Rep. Duncan Hunter) use campaign money for their personal benefit, that I think appeal to voters’ instinct against politicians abusing their office. But I think that’s different from, say, McSally’s case, where her only crime was accepting more than the legal limit in donations.

I fear Americans see the latter as just violations of arbitrary bureaucratic rules, rather than as an immoral act.

clare.malone: Well, not everyone sees breaking the law on certain things as a moral violation. The law doesn’t necessarily equate with morality! Lots of people might think the ends (big money) justify the means. They might not think that the moral universe extends to bureaucratic violations.

dave.levinthal: Great point, Clare, and yes — there are wide swings in opinion on whether the Stormy Daniels matter is a campaign law violation in the first place. (Michael Cohen certainly has some thoughts on this.)

clare.malone: What about for media watchdog organizations like CPI and Open Secrets, though, Dave?

How will your work potentially change because of this?

dave.levinthal: Our job is to report about the role money plays in American politics. And while most people often don’t care about the legal or technical particulars of campaign-finance law, I’ve never gotten the sense that they don’t care about campaign money, especially in the context of their favorite candidates raising cash.

There are tens of thousands of people every day who make campaign contributions, millions every year. The sophistication of political fundraising has made it as easy as ever to support a candidate or cause. That’s why, for example, you see presidential candidates — President Trump and the gaggle of Democrats — raising huge amounts of money from small-dollar donors.

nrakich: Right — and voters do seem to care about the source of the money candidates raise.

For example, every Democratic presidential candidate has pledged not to accept money from corporate PACs, and most have pledged not to accept money from the fossil-fuel industry or federal lobbyists. They wouldn’t be doing that if they didn’t think voters cared about those issues.

clare.malone: Have we seen any irregularities from any of these Democratic candidates? Or from the Trump campaign (this time around)?

dave.levinthal: No, but the FEC also won’t be in a position to address some novel questions about how political candidates should act. For example, lots of cities have sent the Trump campaign bills for police and public safety costs — related to Trump campaign rallies — that they believe the campaign should pay. The Trump campaign doesn’t acknowledge these bills and doesn’t list them as debt, or even “disputed debt,” on campaign-finance reports. It’d be the FEC’s job, ostensibly, to figure this situation out. It can’t now.

clare.malone: Oh that’s really fascinating. NYC certainly saw a lot of controversy over the cost of Trump Tower security right after his election in 2016.

sarahf: So to wrap, it sounds like as long as the FEC can still perform some of its basic functions (like getting candidates to file their reports), we might not see Petersen’s replacement for a while, right? Where does this political fight head next?

clare.malone: I guess … nowhere? Dave is certainly the expert on this, but I don’t think there’s much political will to replace the FEC positions. And I say that mostly from the point of view of public pressure — there’s no incentive to change the course of behavior toward the FEC.

nrakich: As Dave said earlier, President Bush and Congress did finally reach a deal in 2008 the last time the FEC went into limbo because it didn’t have enough members. But I agree with Clare — I think this is so far down on the to-do list for both Trump and McConnell.

Maybe the FEC becomes a poor man’s Merrick Garland — no action until one party regains full control of government.

dave.levinthal: I talked last week with Rep. Derek Kilmer, a Democrat from Washington, about the broader issue of the FEC’s role in government and politics. And he made the case that the FEC needs to be fundamentally reformed and given greater independence and strength.

He even has a bill that would make the FEC a five-commissioner body, which would address the issue of deadlocked votes. The bill isn’t going anywhere, but his hope is that Democrats will win everything in November 2020, and come 2021, the FEC will be reformed.

So I’d say keep a close watch on Schumer in the Senate. If he wants to make a big stink about this, he could. But he, too, has been pretty quiet about the FEC lately. I will also be curious to see if this comes up during the presidential debate next week, since several candidates have been very anti-Citizens United and anti-BIG MONEY in their campaign rhetoric.

nrakich: Yeah, Steve Bullock presented campaign finance as his big issue when he launched his campaign. But in general, I’ve felt that the candidates have not done a good job sticking to what was supposed to be their signature issue (Eric Swalwell and guns, Jay Inslee and climate change, etc.).

dave.levinthal: Even though Bullock has made it his signature issue, he’s had his own little bumps in the road.

clare.malone: Gillibrand also made public funding a thing, if I recall. That did not catch on.

nrakich: That said, if the FEC is shut down for a full year or more — say, through the 2020 election — I bet there will be more of an appetite to reform it come 2021 if Democrats are in charge.

Just speculating, but I think an FEC shutdown might be the kind of thing that gets more noticeable with time.

dave.levinthal: An FEC that effectively sat out the 2020 election would be monumental. It’d take us back to a pre-Watergate era of campaign-finance regulation in certain ways. (The FEC was created after Watergate to help defend against campaign money problems, irregularities and potential lawlessness.)

In fact, I’d say it’d be the most incentive Congress has probably had since Watergate to fundamentally change the nation’s campaign-finance regulation regime.

It Doesn’t Hurt To Hire Alumni As Head Coaches. But It Doesn’t Help, Either.

When David Shaw informed his parents that he was Stanford’s new head football coach, there wasn’t a dry eye in the house. “Pride,” he recalled in an interview this offseason, “is such a small word for that feeling.”

Shaw remains at the helm nearly a decade later.

Stanford is where Shaw was a four-year letterwinner and where his dad was nearly elevated to head coach. It’s where Shaw met his wife and where he proposed; where he left and where he returned. The lockscreen on his iPad is a family photo taken among the eucalyptus trees on campus in 1975, then replicated some 35 years later. “To have so much of your life associated with a place,” Shaw told me, “is weird.”

An alma mater, Shaw contends, is an extension of home. In turn, its people — from the dining hall staff to the university board of directors — are akin to family. In 1957, Paul “Bear” Bryant left a winning program in College Station to return to Tuscaloosa, where the Alabama Crimson Tide had endured a fourth consecutive losing season. Why? “I’ve heard mama calling,” he told his players.

Hiring an alumnus1 has been a marketable, low-risk, high-reward strategy for athletic departments for almost a half-century. Unless a splashy move is feasible, if an athletic director seeks to turn around a program — or merely wants to sell more tickets — there are far worse blueprints to follow than returning someone to their roots. Which is perhaps why alumni have permeated the market for much of the modern era. Since 1975, 38 FBS football programs2 hired more than one alumnus as head coach.3 Over that period, there was only one season in which alumni didn’t hold at least 10 percent of the available head coaching gigs.

And recently, hiring from within has been a successful strategy. Alumni accounted for roughly a quarter of the coaches represented in last season’s final Top 25 rankings. This season, alumni will command 18 programs, or 13.8 percent of the market. The fraternity includes long-tenured stalwarts (Northwestern’s Pat Fitzgerald and Air Force’s Troy Calhoun) in addition to the more idiosyncratic personalities in the sport (Michigan’s Jim Harbaugh and Oklahoma State’s Mike Gundy).

Shaw was hired to “take the dips out of Stanford football.” He became the winningest coach in program history two years ago. But was the university seal on Shaw’s diploma in any way predictive of that success? And more broadly, does an alumnus traditionally make for an exceptional hire?


There were 27 head-coaching changes this offseason, but only one — Northern Illinois’s hiring of Thomas Hammock — involved an alumnus returning home. Hammock and 11 other head coaches — all non-alums — made their FBS coaching debuts over the weekend, while Miami’s Manny Diaz made his debut the previous Saturday.

If the recent success of certain high-profile alumni head coaches is predictive, Hammock should lead a successful squad in 2019. Kirby Smart’s Georgia Bulldogs were among the most efficient teams in the country. Fitzgerald led Northwestern to its first appearance in the Big Ten title game a third consecutive bowl game victory. Jeff Tedford coached Fresno State to its best season in school history. Scott Satterfield turned Appalachian State into the darlings of the Sun Belt and parlayed it into a job at Louisville. Bryan Harsin led Boise State to the Mountain West championship game for the second consecutive season.

In total, alumni head coaches went 159-122 (.566) in 2018. According to Sports-Reference.com’s Simple Rating System, the average alumnus-led program was 4.16 points better than the average team in 2018, the highest mark of any season since 1975. However, alumni-coached teams have also seen a broad range success relative to the average team in this time period.

There were 146 alumni head coaches from 1975 through 2018. In total, they won 52.6 percent of their games, and the median alumnus-coached team was 2.2 points better than the average team in a given season. Non-alumni head coaches, of which there were 641, won 51.1 percent of their games and were on average 0.37 points better than the average team in a given season.

That divide has increased over the last 20 years. Alumni head coaches have gone 2469-2029 (.549), while non-alumni have gone 13177-12299 (.349).4

In addition to recent success, there are several qualities that make the alumni coaches unique. Alumni traditionally begin their college head coaching careers where they suited up. Of the 146 alumni head coaches from 1975 to 2018, only 26 (17.8 percent) had previous Division I head coaching experience. That has continued to be the norm. Since 2000, of the 53 alumni head coaches who were hired, all but eight were becoming first-time head coaches.

Hawaii head coach Nick Rolovich knew at some point in his career he’d return to Honolulu. It ended up being where he got his first crack as the man in charge. “One of my real goals in coaching is to repay Hawaii [for] what Hawaii as a state and a university gave me,” Rolovich said. “Which is everything.”

And once installed at the head of the program, alumni seldom leave. Of the aforementioned sample, more than 75 percent didn’t take another Division I head coaching gig after securing the job, and nearly 70 percent spent their entire college coaching careers at their alma mater.

One potential reason for this continuity is that alumni seem to be working with a longer leash, perhaps as a result of performance. Since 1975, more than 80 percent of alumni head coaches lasted at least three years with a program, while the same is true for just 15 percent of non-alumni. While more than 12 percent of alumni last at least 10 years, only 2.2 percent of non-alumni can say the same.5

Some alumni coaches attribute this trend to their one of their strengths: having already demonstrated an ability to represent the university well as a student-athlete. “It’s very easy for me to talk about Boise State,” Boise State head coach Bryan Harsin said. “I don’t need a map.”

“If you can have success, then hopefully they’ll be proud of you,” Fitzgerald said. “That’s the hope, at least.”


That isn’t to say there aren’t obstacles. The allure of returning home is countered by increased expectations, both from inside and outside the locker room. Happiness is fickle and patience wears thin when losses pile up. And sometimes winning isn’t enough, even at your alma mater. At Maryland, Ralph Friedgen was fired after going 9-4 and winning ACC Coach of the Year. He later burned his diploma.

“A lot of times, criticism is a faceless person,” Western Michigan head coach Tim Lester said. “But at your alma mater, sometimes it’s a little bit harder because you do know who they are.”

It can be a challenge to remember that it is, in fact, a job.

“You have to be able to mix the business side with the side where your heart is,” Rolovich said.

Regardless of the campus where a coach cut his teeth, the terms of employment remain clear. In college football, the long-term prognosis isn’t stable. “Either we’re going to get fired or we’re going to leave,” former Tulsa head coach Bill Blankenship said. Blankenship coached at Tulsa, his alma mater, from 2011 to 2014. “The odds of retiring at the school that hired you is a pretty low percentage.”


But is hiring an alumnus predictive in any significant way?

Not especially.

To measure how predictive hiring an alumnus is of on-field performance, I pulled all of the head coaches from 1975 to 2018 as well as their alma maters, their head coaching experience and their team’s SRS (as well as its previous season SRS), removing interims from the sample.6 Then I ran a linear regression. Controlling for the coach’s experience and the previous SRS of the team, coaches at their alma maters are statistically indistinguishable from non-alumni.

However, a team’s previous season SRS (0.72) is far more predictive of on-field performance — each point of SRS in a previous season is worth around three-quarters of a point in the current one. Furthermore, whether it was the first year of a head coach at a given school is also more predictive of on-field performance than whether they were an alumnus; being a newcomer is strongly negatively correlated with performance (-1.53 SRS).

Plus, despite the recent success they’ve had, in terms of historic single-season performance, alumni haven’t produced a ton. Of the 20 best single seasons, as defined by SRS, only one7 was coached by an alumnus.

Phillip Fulmer is the last coach to win a national title at his alma mater. That was in 1998. Since 1949, only Steve Spurrier, Bryant, Ralph Jordan and Frank Leahy can say the same.


When Shaw’s legacy is written, it will no doubt be penned in cardinal red.

“It’s one thing to fit the program, it’s another thing to fit the entire institution,” Shaw said. “Because I went to school here, I understand the ethos, I understand the air of this place.”

But despite the recent tear of success, that indelible connection that he and fellow alumni head coaches have isn’t terribly predictive of future success.

Neil Paine contributed research.

It’s Time To Stop Pretending That The NFL Preseason Isn’t Pointless

The curtain fell on the 2019 NFL preseason Thursday night — and judging by the volleys of rotten produce hurled at it by fans, writers and coaches, the NFL may never want to stage that show the same way again.

For decades, it’s felt like the NFL has had a predictable rhythm to how (and how often) starters play. A little in the first game, then a little more, and then the third preseason game is the “dress rehearsal,” when coaches game-plan, starters start, and the fans who paid full price for tickets get treated to something resembling their team. In the context of meaningless August football, this one game on the preseason schedule was the closest thing fans got to the real thing. The fourth preseason game, in which starters rarely played, has been a forgivable afterthought.

But the ugly, pointless football played Thursday night felt unforgivable — because the players fans pay to see barely played in the first three games, either. Even the last bastion of NFL preseason relevance seems to be vanishing. This year’s “dress rehearsals” hardly lived up to their billing. Carolina Panthers starting quarterback Cam Newton left the game after a minor injury. Almost all of the Green Bay Packers’ starters were held out. Houston starter Deshaun Watson got sacked to start the team’s first possession, the Texans lost starting running back Lamar Miller for the season on the next play, then Watson got sacked again, fumbled the ball and headed for the bench without throwing a pass. Indianapolis Colts starter Andrew Luck retired before taking a single preseason rep.

If teams are comfortable going the entire preseason with their starting quarterbacks barely taking the field, the league’s case for making their fans spend the time and money to watch these games is significantly weakened. Perhaps as no surprise, calls to reduce the number of preseason games are now coming from everywhere, from fans on Twitter to major news outlets. It feels like all of a sudden, the whole NFL-watching world has given up on the preseason.

Of course, calls for a shortened NFL preseason are nothing new. Analyst John Clayton called for it in The Washington Post earlier this month — almost two decades after he wrote for ESPN that players’ union representatives had already been pushing for it “for years.”

The year after Clayton wrote that ESPN article, the NFL expanded to 32 teams. Fourteen of those teams’ eventual Week 1 starters led their squad in preseason pass attempts. Even as their union reps were arguing that a four-game preseason was at least one game more than anybody needed, stars like Tom Brady, Drew Brees, Daunte Culpepper, Brett Favre and Peyton Manning were out there taking more reps than anybody on their team.

During that 2002 season, 34.4 percent of preseason passes were thrown by quarterbacks who would go on to start Week 1. A decade later, starters’ share of the workload was about the same. But from 2012 to last season, their leaguewide share of pass attempts dropped by 36.5 percent.

This season, it fell even more steeply. Projected Week 1 starters (via OurLads) accounted for just 11.7 percent of pass attempts.1 That’s a 43 percent drop in one year, after more than a decade of consistently giving fans at least a decent look at the most important player on the team.

The same pattern shows up when we look at how many starters have led their team in preseason attempts — and how many starters have brought up the rear. In 2002, 14 of 32 teams’ starting quarterbacks led their team in preseason pass attempts, and in 2012, 13 starters still led their team in preseason throws. But since 2015, no more than three have.

For most of the 2000s and into the middle of this decade, the number of starters who threw the fewest preseason passes on their teams stayed in the low single digits. Last year, it was a full half of the league’s starting QBs.

This year, no starters have thrown the most passes of anyone else on their team, and 22 threw the fewest. In just seven years, we’ve gone from almost half the league mostly playing their starters to over two-thirds the league barely playing them at all.

There’s also reason to believe that the decline of starting QB reps across the league this preseason is not a coincidence. The NFL and its players’ union have begun negotiations for their next collective bargaining agreement, and truncating the preseason is reportedly a major negotiating topic. In that context, coaches and players are incentivized to force the owners’ hands.

Last week, with Luck likely to sit out the third preseason game, Colts head coach Frank Reich and Chicago Bears head coach Matt Nagy texted before the game and reached a mutual I-won’t-play-my-guys-if-you-won’t truce. A similar detente was reached between Philadelphia Eagles head coach Doug Pederson and Baltimore Ravens skipper John Harbaugh after their week of joint practices.

In fact, joint practice sessions seem to be where the starters are getting all the reps they’ve been giving up.

“I think [joint practices are] the trend. I think that’s where we’re going. I think that’s the way the league is heading,” Pederson said in a recent press conference. “As coaches, we get to set the situation and control the environment, and sometimes you don’t get those in games. You don’t get that situation in a game, and this way we can control that and work on specific things and get some really good work done with our starters.”

That all makes sense: If a coach really wants to work on the two-minute offense, a preseason game offers no guarantee that a team will even get in a two-minute situation. The same is true for any other situation, matchup or personnel package. What doesn’t make sense, though, is charging fans full price to watch an uncontrolled scrimmage between a bunch of players who likely won’t even make their respective teams.

On Monday, Texans head coach Bill O’Brien suggested that fans could whet their appetite for starter-on-starter action by the league televising joint practices in lieu of two preseason games:

If the coaches, players and fans all feel like the risk of injury has outstripped the value of playing the games, there’s no viable path forward for the four-game preseason. Only one question remains: Whether the coaches, players and fans can persuade the owners to get on a different path.

A Peaceful (But Not Peaceful) Transition Of Power In Riddler Nation

Welcome to The Riddler. Every week, we offer up problems related to the things we hold dear around here: math, logic and probability.

Before we puzzle this week, some Riddler news: This is my final Riddler column as editor. I’m stepping away from FiveThirtyEight to spend a year researching how journalists should investigate and write about advanced artificial intelligence. Algorithms and robots, beware.

Editing this column for the past four years has been a pure joy, thanks entirely to you, its readers. You submit the innovative puzzles, you labor over the elegant solutions and you power this whole inventive mathematical enterprise — the exciting, challenging and educational numerical community that has become known as Riddler Nation. I have been merely a curator of your quantitative creations. Thank you.

And fear not. Riddler Nation shall persist.

I’m thrilled to introduce FiveThirtyEight’s new puzzle editor, Zach Wissner-Gross.

Zach studied physics and biology at MIT before completing a doctorate in physics at Harvard. These days he leads a team that’s developing a K–12 math curriculum, and by night he’s a proud member of Riddler Nation. I’m excited for his leadership of this column and community.

Regular readers of this column will already recognize the contributions Zach has made as a citizen of Riddler Nation. He’s authored puzzles about robot pizza cutters, misanthropic settlers and swindling car salesmen, and his many beautiful solutions have tackled alien invasions, railroad construction and martini glass spillage, to name but a few. Who knows where he’ll take us next?

You can find Zach on Twitter, and do be sure to email him all your favorite puzzles. As for me, I may no longer be editing but I will surely be reading — and solving. (Or, at least, always trying to do the latter.)

This week’s Riddler

For a long time, my favorite Riddler has been the Battle for Riddler Nation, first suggested a couple of years ago by Joel Baker. It brings together all of what makes this column what it is: a fantastical story, mathematics, large-scale interaction, the power of computer simulation, game theory and fierce (yet good-natured) competition. So let us battle once more. Here are the rules:

In a distant, war-torn land, there are 10 castles. There are two warlords: you and your archenemy. Each castle has its own strategic value for a would-be conqueror. Specifically, the castles are worth 1, 2, 3, …, 9, and 10 victory points. You and your enemy each have 100 soldiers to distribute, any way you like, to fight at any of the 10 castles. Whoever sends more soldiers to a given castle conquers that castle and wins its victory points. If you each send the same number of troops, you split the points. You don’t know what distribution of forces your enemy has chosen until the battles begin. Whoever wins the most points wins the war.

Submit a plan distributing your 100 soldiers among the 10 castles. Once we receive all your battle plans, we’ll adjudicate all the possible one-on-one matchups. Whoever wins the most wars wins the battle royale and is crowned king or queen of Riddler Nation!

For your strategizing benefit, you may consult Riddler Nation’s chosen attack distributions from the first, second and third sanctionings of this battle royale, which had the same rules. A summary of them is in the table below. Adapt and conquer!

Troop deployment strategies

How Riddler Nation distributed its forces in each battle royale

Average number of soldiers sent to castle …
Battle 1 2 3 4 5 6 7 8 9 10
May 2019 2 3 4 7 9 12 13 17 17 16
May 2017 3 4 6 8 10 12 15 17 14 12
Feb. 2017 3 3 4 7 9 13 16 19 16 11

Finally, it should be noted that your work is royally cut out for you. There is currently a two-time defending champion, Vince Vatter. Can anyone dethrone the king?

Submit your plan

Solution to last week’s Riddler Express

Congratulations to 👏 Stephen Wilson 👏 of Nashville, Tennessee, winner of last week’s Riddler Express!

Last week found you employed as an expert counterfeiter, specializing in forging $100 bills. You knew from experience that the bank could only spot your fakes 25 percent of the time, and that trying to deposit only counterfeit bills would be a ticket to jail. However, if you combined fake and real notes, there was a chance the bank will accept your money. You had $2,500 in real hundreds and an unlimited supply of counterfeits. The bank scrutinized cash deposits carefully: They randomly selected 5 percent of the notes they received, rounded up to the nearest whole number, for close examination. If they identified any note in a deposit as fake, they would confiscate the entire sum, leaving you only enough time to flee. How many fake notes should you have added to the $2,500 in order to maximize the expected value of your bank account? How much free money were you likely to make?

You should add 55 fake notes to your 25 real notes for a total deposit of $8,000. Your expected gain is $1,256.

The full counterfeit solution (counterfeit bills, that is … the solution itself is very real) comes to us from this puzzle’s submitter, Jason Ash:

Your expected gain is the weighted average of expected profit from successful deposits and expected losses from bank seizures. With 55 fake notes, there is a 47 percent chance — a combination of the chances of the bank selecting our fake notes for examination and of the examination revealing any fake notes — we avoid detection and collect a profit of $5,500, the value of the fake notes we were able to sneak into circulation. (Remember we started with $2,500, so only the fake notes count as profit.) There is a 53 percent chance we are caught by the bank and lose the $2,500 in real dollars we used as decoys. Therefore, our expected profit is (0.47)(5,500) – (0.53)(2,500) = 1,256.

How can we be sure no other strategy produces a higher profit? For example, suppose we combine 30 fake notes with 25 real notes instead. The bank will select three notes for its audit — that is, 5 percent of 55, rounded up to the nearest whole number. Depending on our luck, the bank could choose all three fake notes, all three real notes, or some combination in between.

Let’s assume the bank randomly chooses two fake notes and one real note for its audit. This occurs with roughly 41.5 percent probability, given by the formula (3 choose 2)(30/55)(29/54)(25/53). Each fake note is detected 25 percent of the time, which means at least one fake note from the pool of two is detected \(1-0.75^2\) = 43.75 percent of the time. We use similar logic to solve for the likelihood and detection rate of the audits with three fake, one fake and zero fake notes. The overall detection rate is equal to the weighted average across each potential audit.

For the 30 fake, 25 real strategy, the probability of success is 64.3 percent — again, we need to either avoid their selecting our fake notes for examination or, if they do select them, avoiding identification of them as such — and the probability of detection is 35.7 percent. Therefore, the expected profit is (0.643)(3,000) – (0.357)(2,500) = 1,038. That’s a decent payday, but we can do better. The chart below shows the profit for strategies with up to 200 fake notes, and it illustrates that the maximum is achieved when we use 55 of them.

We can see two patterns above. First, as we move to the right, we enter the “greedy danger zone,” in which the bank becomes more likely to discover our fraud and seize the starting capital, resulting in larger and larger expected losses. Second, we see “sawtooth” behavior caused by the bank’s practice of auditing 5 percent of deposited notes. The effect is significant: 55 is the ideal answer because if we deposit 80 total notes, the bank will audit four. If we use 56 fake notes for a total of 81, then the bank audits five notes instead, which cuts the expected gain in half!

A life of crime only pays if you’re good with numbers. But with great power comes great responsibility, Riddler Nation.

Solution to last week’s Riddler Classic

Congratulations to 👏 Jake Wiley 👏 of Wilmington, Delaware, winner of last week’s Riddler Classic!

Last week also found you employed in your second job as the resident mathematician of the Puzzling Pizza pizzeria. Your boss had purchased the equipment necessary to design the machine below.

Your job was to calculate the exact path and flow rate the sauce-dispensing arm should use to fill a 12-inch circular pizza with sauce as fully and evenly as possible. The dough sat on a platform rotating at a constant speed, and the arm traveled in a straight line across the pizza, distributing sauce in a circular shape with a 0.5-inch diameter. Once the arm started pouring sauce, it couldn’t stop until the pizza is covered, and the sauce can only be poured onto the dough. At the end, the pizza was meant to have an even layer of sauce on as much of its surface as possible. What path and flow rate should the sauce-dispensing arm take to give you the best pizza?

The full solution comes to us from this puzzle’s submitter, Tyler Barron:

Let’s start with the path the arm would take. The best method is to start at the middle and move outwards at a constant velocity until you reached the end. (It’s also possible to travel this route from the outside in.) This shape it would create is known as an​ ​Archimedean spiral​. Saucing a pizza like so would give you a path that looks like this — the spiral stops where it does simply because we’ve run out of room for more sauce:

This curve should move outward at a rate of 0.5 inches per rotation for 23.5 rotations, until the arm is 11.75 inches out. To introduce some formal notation to our pizzeria, this can be written as \(r(\theta)=\frac{0.5}{2\pi}(\theta)\).

Finding the flow rate is more difficult. To start, let’s get a baseline of what our sauce would look like if we distributed it at a constant rate through the entire path. If we did this, we would get a pizza that looks like this:

As you can see, the sauce is thinner on the edges and thicker in the center. Why? Imagine the final rotation where the arm travels around almost the entire circumference, compared to one earlier where it travels through the middle of the pizza. The distance traveled per rotation decreased but the amount of sauce per rotation stayed the same, so we’re left with a higher density of pizza sauce on the inner ring.

Therefore, we need to adjust the flow of sauce to the speed of the pizza traveling under the arm. This can be found by taking the derivative of the arc length equation. That beast — specifically, the​ ​Archimedean spiral arc length equation​ — is written mathematically as \(s(\theta)=\frac{b}{2}(\theta \sqrt{1+\theta^2}+\sinh^{-1}(\theta))\). Its derivative gives us the rate that the length changes as we rotate, and that is \(\frac{ds}{d\theta}=b\sqrt{1+\theta^2}\).

Finally (we promise) this can be converted to the relative flow rate by dividing the derivative by the rate of change at the edges, giving you the saucer’s arm speed compared to its maximum speed. That formula is: \(F(\theta)=\frac{b\sqrt{1+\theta^2}}{b\sqrt{1+(47\pi)^2}}=\frac{\sqrt{1+\theta^2}}{147.66}\).

That gives us a (nearly linear) increasing sauce-flow rate that looks like so. As the dough moves through its rotations, we increase the flow from almost nothing at the center to full strength at the edge:

Using this flow rate will give you an even spread of sauce on the pizza, which means we should program the pizza maker to use a path of \(r(\theta)=\frac{0.5}{2\pi}(\theta)\) and a relative flow rate of \(F(\theta)=\frac{\sqrt{1+\theta^2}}{147.66}\). The final result is the rather delicious-looking image below.

Now that’s a pizza Archimedes would be proud of.

Want more riddles?

Well, aren’t you lucky? There’s a whole book full of the best puzzles from this column and some never-before-seen head-scratchers. It’s called “The Riddler,” and it’s in stores now!

Want to submit a riddle?

Email Zach Wissner-Gross (!) at [email protected]

Significant Digits For Friday, Aug. 30, 2019

You’re reading Significant Digits, a daily digest of the numbers tucked inside the news. This is my final column as your digit wrangler. It’s been a pleasure — thank you for reading.


408,000 men and women

A new study in the journal Science — the “largest ever to analyze the genetics of same-sex sexual behavior” — examined genetic data from 408,000 men and women and found that genetics does play some role in determining who has same-sex sex. It’s not the result of a single gene, but many genes that contribute some small effect. The Times also reported that several scientists who are part of the LGBTQ community at the Broad Institute, which led the research, “said they were worried the findings could give ammunition to people who seek to use science to bolster biases and discrimination against gay people.” [The New York Times]


8 pages

During its failed establishment of a second headquarters in New York City, Amazon reportedly kept a “burn book” — “an eight-page, bullet-pointed, Calibri font testimony” of what the company saw as insults from politicians and labor leaders who opposed the deal. Maybe I’m a typeface nerd, but the thing that most bothers me about this news is the use of Calibri. [The Wall Street Journal]


405 police departments

Speaking of Amazon, Ring, its home-security company, published a map this week showing the police departments across the country with which it has signed deals. Ring has acknowledged active partnerships with 405 such departments. The deals vary, and include such things as police offering rebates to people in the community who buy Ring products and granting police “access to a portal they can use to request footage from Ring camera owners in their communities.” [Gizmodo]


3,000 applications

The Morris Animal Refuge in Philadelphia put out a call online recently for adoptees of a 26-pound (!) cat named BeeJay, a.k.a. Mr. B. Needless to say, given the collective preferences of the internet, the call went viral and some 3,000 applications were submitted. Mr. B did find a home, and the shelter is working with the foster family “to help resolve the cat’s health and behavioral issues so that they might eventually become his permanent home.” Good luck, big fella. [NBC News]


140 home runs

Records continue to fall in baseball’s latest home-run era. Most recently, the Minnesota Twins broke the record for most homers on the road, smacking two in the third inning yesterday against the Chicago White Sox for a total of 140. That previous record was 138, set by the San Francisco Giants in 2001. [Associated Press]


From ABC News:
SigDigs: Aug. 30, 2019


2, 3, 4, 8, 9, 20 and 30

For decades, an Edmonton, Alberta, man named Bon Truong played the exact same lottery numbers, derived from important dates in his life: 2, 3, 4, 8, 9, 20 and 30. And last fall, he finally won — $60 million. Truong said he has waited nearly a year to claim his winnings because he was “overwhelmed by the size of his luck.” I, for one, am overwhelmed by the size of his persistence. [The Washington Post]


Love digits? Find even more in FiveThirtyEight’s book of math and logic puzzles, “The Riddler.”


What If The Third Debate Were Based On Different Polls?

Welcome to Pollapalooza, our weekly polling roundup.

Poll(s) of the week

The deadline to make the third Democratic primary debate has passed, and thanks to harder qualifying rules, just 10 candidates made the stage. This, of course, was unwelcome news among candidates such as billionaire activist Tom Steyer and Hawaii Rep. Tulsi Gabbard, who were on the cusp of making the debate. Steyer needed just one more qualifying poll, and Gabbard needed two.

And this got us thinking: What would the debate stage look like if the list or type of eligible polls were different? Gabbard, in particular, has taken the Democratic National Committee to task for the specific pollsters included in its list of approved polling organizations, arguing that had the list of pollsters been expanded, she would have had at least 2 percent support in more than 20 polls conducted during the third debate qualification window. And in fairness to her, understanding how the DNC determines its list of approved polling organizations can be confusing. Gabbard did hit 2 percent in YouGov’s latest national survey sponsored by The Economist, but it didn’t count toward qualifying for the debate.

So to better understand how including different pollsters or relying on different pollster methodologies could affect who made the debate stage, we checked to see who would have qualified if:

  1. all polls had been counted;
  2. just polls from pollsters with a grade of at least B- or better, according to FiveThirtyEight’s Pollster Ratings (as this grade captures a mix of high-quality phone polls and respected online polls); and
  3. only live phone interviews polls, which are often considered the gold standard in polling.

And in this thought exercise, we also kept many of the DNC’s requirements for the third debate, meaning we also included only polls released between June 28 and Aug. 28 — and only counted a candidate as qualified if he or she hit 2 percent support in four polls and attracted the support of 130,000 individual donors (including at least 400 individual donors in at least 20 states).2 We also adhered to the DNC’s rules that limit qualifying polls to national and early-state surveys3 and that said two polls by the same pollster in the same geography can’t be counted.4

OK, so first up: Who would have made the stage in our most generous scenario where all polls are counted? Well, maybe not as many candidates as you’d expect given the parameters. Gabbard and Steyer would make the stage with nine and seven polls, respectively. And author and motivational speaker Marianne Williamson comes a little closer to making it with two qualifying polls. (She also has met the donor requirement.) But this still leaves out seven candidates that FiveThirtyEight considers “major” as well as the candidates who have dropped out since the second debate.

What if all polls had been counted for the third debate?

Candidates who would have qualified for the third debate had the DNC used all polls* in the FiveThirtyEight database released from June 28 to Aug. 28

Met DNC criteria Met hypothetical
Candidate 130k+ Donors 2% in four polls All polls
Joe Biden
Cory Booker
Pete Buttigieg
Kamala Harris
Amy Klobuchar
Beto O’Rourke
Bernie Sanders
Elizabeth Warren
Andrew Yang
Julián Castro
Tulsi Gabbard
Tom Steyer

For candidates considered “major” by FiveThirtyEight.

We adhered to the DNC’s donor requirements and polling support threshold. To qualify for the third debate under the DNC’s rules, a candidate had to reach 2 percent in at least four national or early-state polls from qualifying polling organizations and needed at least 130,000 unique donors, including at least 400 donors in at least 20 states.

*We excluded polls conducted by partisan pollsters, head-to-head polls, polls with open-ended questions and polls in the same geography by the same pollster.

Sources: Polls, Media reports

So, OK — what about the scenario in which we limit our scope to pollsters with at least a B- grade, according to our pollster ratings? It makes sense that the DNC would want to limit at least some of the pollsters included. So we chose pollsters that are still high quality, but our list of pollsters ends up being a little more expansive than the list of DNC-approved pollsters. And under this scenario, the same 12 candidates would make the stage as in the “all polls” scenario, but it’s a much closer cutoff — Steyer would have ended up with exactly four qualifying polls and Gabbard five — just one fewer than former Secretary of Housing and Urban Development Julián Castro.

Only pollsters with a grade of at least B- counted?

Candidates who would have qualified for the third debate had the DNC included pollsters that FiveThirtyEight has given a grade of at least B-, with polls* released from June 28 to Aug. 28

Met DNC criteria Met Hypothetical
Candidate 130k+ Donors 2% in four polls Pollster rating of at least B-
Joe Biden
Cory Booker
Pete Buttigieg
Kamala Harris
Amy Klobuchar
Beto O’Rourke
Bernie Sanders
Elizabeth Warren
Andrew Yang
Julián Castro
Tulsi Gabbard
Tom Steyer

For candidates considered “major” by FiveThirtyEight.

We adhered to the DNC’s donor requirements and polling support threshold. To qualify for the third debate under the DNC’s rules, a candidate had to reach 2 percent in at least four national or early-state polls from qualifying polling organizations and needed at least 130,000 unique donors, including at least 400 donors in at least 20 states.

*We excluded polls conducted by partisan pollsters, head-to-head polls, polls with open-ended questions and polls in the same geography by the same pollster.

Sources: Polls, Media reports

On the other hand, what if the DNC had been more — not less — strict with its requirements? For instance, what if the DNC had chosen to just use pollsters that use live phone interviews? Yes, these polls are facing many challenges right now, including low response rates and high costs, but they remain the best performing type of poll. So if the DNC had limited qualification to these types of polls, the number of debate participants would have actually shrunk from 10 candidates to nine. The odd man out would be Castro, who would have ended up with only three qualifying polls, ahead the two for Gabbard and Steyer.

Only polls conducted by telephone counted?

Candidates* who would have qualified for the third debate had the DNC only included pollsters that do live telephone surveys, with polls* released from June 28 to Aug. 28

MET DNC Criteria Met Hypothetical
Candidate 130k+ Donors 2% in four polls Live phone Polls
Joe Biden
Cory Booker
Pete Buttigieg
Kamala Harris
Amy Klobuchar
Beto O’Rourke
Bernie Sanders
Elizabeth Warren
Andrew Yang
Julián Castro

For candidates considered “major” by FiveThirtyEight.

We adhered to the DNC’s donor requirements and polling support threshold. To qualify for the third debate under the DNC’s rules, a candidate had to reach 2 percent in at least four national or early-state polls from qualifying polling organizations and needed at least 130,000 unique donors, including at least 400 donors in at least 20 states.

*We excluded polls conducted by partisan pollsters, head-to-head polls, polls with open-ended questions and polls in the same geography by the same pollster.

Sources: Polls, Media reports

So big picture, you could say the exact DNC rules don’t make a huge difference — most of the same set of candidates makes it on stage regardless. Of course, for the individual candidates on the edge of qualification, that give or take is everything. Suffice it to say, the rules matter quite a bit to them. And in this case, there’s an argument to be made that the DNC’s list of eligible pollsters helped make or break qualification for those candidates on the bubble — Gabbard and Steyer in particular.

Other polling bites

  • New polling from Politico/Morning Consult suggests that Democrats prefer “Medicare for All” to building on the Affordable Care Act. The survey found that 65 percent of Democratic primary voters were either “much more likely” or “somewhat more likely” to back a presidential candidate who supported a single-payer health care system like Medicare for All over “preserving and improving” the ACA. Just 13 percent said such a position would make them “much less likely” or “somewhat less likely” to support such a candidate. Among all voters, 53 percent supported a Medicare for All-type health system compared to 34 percent who opposed it.
  • A new report from the Pew Research Center found that 61 percent of Americans think abortion should be legal in all or most cases, while 38 percent said it should be illegal in all or most cases. And though Pew found stark partisan divides over abortion policy, it did find evidence that there was more support for policies advocated by the Democratic Party (42 percent) than the Republican Party (32 percent) – though 24 percent said they don’t agree with either party’s policies.
  • In a recent report on trust, media and democracy, Gallup and the Knight Foundation found that Americans are at least somewhat worried about local news organizations being consolidated under large media companies, especially if the parent company had strong political views. Sixty-six percent said they would be “very” concerned that the political views of the parent company “would influence the fairness of news coverage,” and 26 percent said they would be “somewhat” concerned. Large majorities also said they were worried about the inclusion of more news from outside the local area and less investment in news gathering and reporting.
  • Sen. Bernie Sanders said last week that he would hold companies accountable for their role in climate change, and YouGov Blue/Data for Progress found that about 62 percent of voters would support holding energy producers legally liable “if it could be proven that they misled the public about the consequences of climate change.” Another 20 percent opposed the idea. And perhaps unsurprisingly, support for the idea fell along partisan lines — 77 percent of Democrats supported the idea as did 63 percent of independents, while 39 percent of Republicans supported it.
  • Ahead of Labor Day, Gallup released a survey on labor unions in the United States, finding that 64 percent of Americans approve of labor unions; that is one of the highest approval ratings in the past 50 years. Since the late 1960s, approval of labor unions has mostly hovered below 60 percent.
  • Pro-Brexit Prime Minister Boris Johnson recently announced that the United Kingdom’s parliamentary session would be suspended until mid-October, not long before the Oct. 31 deadline for the U.K. to agree to a managed transition to leave the European Union. Johnson’s move gives members opposed to exiting the EU without an agreement less time to maneuver against a “no deal” Brexit, and a new poll from YouGov found that 47 percent of Britons oppose Johnson’s decision, while 27 percent support it.

Trump approval

According to FiveThirtyEight’s presidential approval tracker, 54.2 percent of Americans disapprove of the job Trump is doing as president, while 41.3 percent approve (a net approval rating of -12.9 points). At this time last week, 41.5 percent approved and 54.0 percent disapproved (for a net approval rating of -12.5 points). One month ago, Trump had an approval rating of 42.5 percent and a disapproval rating of 53.4 percent, for a net approval rating of -10.9 points.

Generic ballot

In our average of polls of the generic congressional ballot, Democrats currently lead by 6.7 percentage points (46.4 percent to 39.7 percent). A week ago, Democrats led Republicans by 6.3 points (46.2 percent to 39.9 percent). At this time last month, voters preferred Democrats by 5.9 points (46.1 percent to 40.2 percent).

Why Running To Win Women Didn’t Work For Gillibrand

The 2020 Democratic primary’s historic field of women candidates just got a little smaller. On Wednesday, after failing to qualify for the September debate, two-term New York Sen. Kirsten Gillibrand announced that she was ending her candidacy.

It’s not hard to see why Gillibrand dropped out — the writing was on the wall. She still hadn’t met the donor threshold for the September debate, and had only hit 2 percent in one qualifying poll (she needed three more). Her debate performances didn’t do much to help her stand out from the other candidates — even on women’s issues, which she had made the centerpiece of her campaign. And although her name recognition rose over the course of the campaign, she didn’t become better-liked.

Her poll numbers barely shifted, too. In the month after she announced she was running for president, she hit a high of 3 percent in one February poll, but she never reached 3 percent again.

In the end, Gillibrand just couldn’t convince women voters — or most voters for that matter — that she was their candidate. But why her candidacy never picked up steam was always a little bit of a mystery. Of course, she had some hurdles to overcome. Like the other women running for president, she faced voters’ biases against women candidates. She also had the baggage of sparking Democratic Party heavyweights’ ire after she called on former Minnesota Sen. Al Franken to resign when sexual harassment allegations against him came out in 2017.

On paper, though, Gillibrand’s campaign didn’t seem especially quixotic. She was on the national stage for more than a decade before throwing her hat in the ring, and established herself as a strong advocate for women’s rights issues such as paid family leave and sexual assault in the military. She was also explicitly pitching her candidacy toward groups like white college-educated suburban women, whose political enthusiasm had just helped sweep a record-breaking number of women into office in the 2018 midterms.

So Gillibrand’s biggest problem may have simply been that there wasn’t a clear base for her in the Democratic electorate — at least not one for which there wasn’t also fierce competition in the rest of the primary field. After all, she was running against a number of other women who are also strong on issues like abortion rights and equal pay. Without another signature issue to help her stand out, she often got lost in the melee of the primary.

For instance, when several states passed laws dramatically restricting abortion in May, Gillibrand seemed like she could have had a breakthrough moment. She even traveled to two of the states to hold rallies in support of abortion rights, and she called for a federal law that would stop state legislatures from passing limitations on abortion — but so did Elizabeth Warren, Kamala Harris and even Cory Booker. In the second debate, Gillibrand tried again to seize the spotlight by taking Joe Biden to task for his position on a childcare tax credit in 1981 — but unlike Harris’s attack on Biden for his stance on school busing a month earlier, the moment didn’t really land.

In those moments and others, her rivals seemed to harness policies that were key to Gillibrand’s candidacy more effectively than she did. It was Harris, not Gillibrand, who grabbed headlines for her plan to penalize companies for failing to pay men and women equally. And in a recent Politico/Morning Consult poll, respondents said that Warren was best qualified to address gender equality, followed by Biden, Sanders and Harris — Gillibrand didn’t even crack the top 10.

In some ways, Gillibrand’s campaign may have also shown just how tricky outreach to women voters can be, even in a year where issues such as abortion and the #MeToo movement are prominent. Women make up about 60 percent of the Democratic base, but there isn’t a lot of evidence that they gravitate automatically toward female candidates because of their shared identity, or even because of shared priorities. In that Politico/Morning Consult poll, for instance, only 5 percent of Democratic women voters said that gender equality was a top voting priority. And Warren and Harris appear to be polling only very slightly better with women than men; that gap is actually bigger for Biden.

Finally, although Gillibrand said she had no regrets about calling for Franken’s resignation, it may have hurt her among donors and party elites. Her campaign at one point suggested that anger over her role in Franken’s departure was hurting her among major party donors. That may be true: Despite having been a formidable fundraiser in the past, she raised substantially less than others in the field. Moreover, unlike her fellow senators, who all drew the backing of political influencers from their home states, Gillibrand netted only one endorsement.

As the first woman to leave the race, Gillibrand’s departure is noteworthy, particularly since she could, in theory, have stuck around and tried to make it into the October debate — after all, the criteria to qualify isn’t changing. So her decision to drop out now may signal some strategic decision-making for candidates who are prominent within the party. and therefore have more to lose by staying in the race too long. It’s possible that even if the White House isn’t in the cards for Gillibrand this time, she may be thinking about running for another office, like governor of New York, and doesn’t want to fall out of grace with the party. Or she may be withdrawing with the goal of helping to pave the way for Harris or Booker — both of whom are friends.

The question now is whether other candidates follow Gillibrand’s lead. She is the sixth candidate to have dropped out this summer, and it’s possible that her departure could be a harbinger of more winnowing. Particularly for anyone else who is thinking about running for office in the future, and wants to stay on the Democratic Party’s good side by helping to narrow the field.