Category Archives: Campaign 2016

Media Bias, Trump Voters, and Apples vs. Oranges

When I was in graduate school, I think the most difficult concept for any scholar to measure was media bias. How do we know one story is biased while another one is not? Bias is a great and easy to use concept if we are talking about me shooting a bow and arrow as part of summer camp as a kid. There’s an objective target out there. If my shot keeps missing two inches to the left of where I aim, I can try to correct the physical errors in my shot by aiming two inches to the right of the bull’s-eye.

Objective journalists strive for the bull’s-eye every time too. However, there is no stable, fixed target for journalists to shoot at. Writing news stories involves a lot of judgment calls. Is this source credible? How do I portray them? Reasonable people can disagree about where the target should be. This means bias is an inherently relative concept when we are talking about the media. It’s based on our frame of reference. As I wrote years ago, audiences may want something different than news organizations.

Measuring media bias is awfully difficult because the target isn’t fixed. We have to choose a reasonable target – some standard of objective journalism – before we could measure how far off actual news content is. Choosing the target is almost inevitably a value-laden judgment. Even “objective” targets lead to value-laden interpretations of the results. For example, let’s say we want to compare news organizations to members of Congress. It takes a lot of work to going through the Congressional Record to build the target. All the work may not be worth it. Congress isn’t a truly “objective” target because it has a majority party. A centrist newspaper would be to the left of Congress today, even if it is closer to the average ideology of Americans.

Since looking for “objective” targets to measure media bias can be a lot of work for little reward, that’s not how most scholars have tried to study journalistic bias. It’s much easier to look at two different media organizations and measure their content relative to each other. Another favorite strategy is to choose two similar news events, then study how the apple received different news coverage than the orange. These strategies are easy to do. They don’t require special math skills for the audience to follow the argument.

Unfortunately, the “compare apples and oranges” style of study distorts our view of the world. Imagine comparing how Vox Media and Breitbart covered the recent scandals facing the Trump administration. One media organization is on the left and the other is on the right, so obviously they will cover the scandal differently. That doesn’t mean either organization makes for a good bull’s-eye. Both media organizations built their brand on ideological takes; neither is trying to be in the center!

Scholars are human, so we will probably have more sympathy towards either the apple or the orange in an “apples and oranges” study of media bias. If you read one of these studies, you will notice that they look for cases where the differences are relatively large. If the apple and the orange receive similar news coverage, no one cares. Apples and oranges studies are good at describing differences in coverage in relative detail. However, these studies lack an anchor. The only way to say that the apple is biased while the orange isn’t is if we really, really like oranges. (I like trolling with the apples and oranges metaphor because I prefer kiwi!)


Whenever I read a pundit or an academic study comparing the attitudes of Trump voters to the attitudes of Clinton voters, I can’t help but see the same problem. Of course there will be some differences between Trump and Clinton voters. For example, a lot of people have focused on the large difference between Trump and Clinton voters on race. Trump announced his candidacy by disparaging Mexican immigrants as rapists and undesirables. It’s easy to see Trump, see a gap in racial attitudes between voters, and assume this is a story of increasingly racist Republican voters. On the other hand, Trump’s explicit racism and the Black Lives Matter movement may have combined to make White Democrats more progressive on race than they were four years ago.

When we separate everyone in to two categories, all we can do is see the difference between those categories. We can see race polarized voters. However, it’s hard to know just how much racism pushed voters in to Trump’s camp, as opposed to anti-racism pushing other voters to Clinton, just by making a comparison.

The easiest way to write about statistical models actually makes this problem harder. To try and explain this, let’s create a quick and dirty regression model from the American National Election Study. I used age, race (white/nonwhite), gender and having a college degree as control variables. (The publicly available version of the ANES doesn’t have income yet, so take all the results here with a grain of salt. I’m not looking for an optimal model, just a teaching tool.)

Social scientists mainly look at racism via a scale of “symbolic racism.” These are four questions that social scientists started using in the 1990s after it seemed clear that people would not truthfully answer more direct questions about whether Blacks should be discriminated against. The symbolic racism scale comes from a time when social scientists almost exclusively focused on Black-White differences. I will get to Trump’s explicit racism against Mexican-Americans in a separate post. Let’s start with the well-established symbolic racism scale. All these questions are on a 5 point agree to disagree scale:

  • Irish, Italians, Jewish and many other minorities overcame prejudice and worked their way up. Blacks should do the same without any special favors.
  • Generations of slavery and discrimination have created conditions that make it difficult for blacks to work their way out of the lower class.
  • Over the past few years, blacks have gotten less than they deserve.
  • It’s really a matter of some people not trying hard enough, if blacks would only try harder they could be just as well off as whites.

We put these items in a scale, since the average response to four related questions is probably more reliable than a single question. The scale still ranges 1-5. I re-ordered responses so low values are respondents who feel additional steps are necessary to end the systematic disadvantages that African Americans face. High values represent racial resentment. To try and illustrate the strength of symbolic racism, I added a respondent’s political ideology on a 1-7 scale (1=strong liberal, 7=strong conservative). I am using a logistic regression model, where positive coefficients represent groups who voted Trump and negative coefficients represent voting for someone else (not necessarily Clinton). Non-voters are dropped from the sample.

Screen Shot 2017-05-21 at 12.27.32 AM

Screen Shot 2017-05-21 at 12.27.41 AM

If we were interpreting this model quickly, we’d look and see that the coefficient for symbolic racism is relatively large and statistically significant at the .001 level. As a voter’s symbolic racism goes up, the likelihood they voted for Trump goes up by a fairly considerable margin. Since the symbolic racism scale goes from 1 to 5, the difference isn’t as big as the difference between strong liberals and strong conservatives. Nonwhites were considerably less likely to vote Trump, and we could talk about other control variables.

At this point it’s very easy to point a finger at Trump voters and accuse them of being racists. The average Trump voter scored 3.92 on the 1-5 scale for symbolic racism. It’s an open and shut case, right? More racism causes more voting for Trump.

When I ran my regression model this quick and dirty way, I set a very unusual bull’s-eye for comparison. Strong liberals are the baseline for political ideology, because they have the lowest numerical value. However, only 3.6 percent of the respondents identified as strong liberals. Because symbolic racism is coded 1 to 5, the minimum value on the symbolic racism scale is the bull’s-eye for this regression model. It looks like symbolic racism causes an increase in voting for Trump because the value can only go up. Let’s see what happens if we reverse the symbolic racism scale to create an anti-racism index, then use it in the regression:

Screen Shot 2017-05-21 at 12.42.29 AM

All the values in this regression are identical, but the coefficient for the anti-racism index is negative. In other words, as people disagree with the statements in the symbolic racism scale, they are less likely to vote for Trump. The constant changed considerably too. You may have noticed the very low constant in the first regression model. Since logit is a multiplicative model, that low constant actually matters a great deal. However, it’s still very easy to overlook if we are skimming results. I often do!

The bull’s-eye in this model is weird. I’m comparing everyone to a hypothetical strong liberal with the maximum score on the symbolic racism index. That combination probably doesn’t exist in the real world, but that’s not the point. This model exists to show that we could just as easily make a regression model showing opposition to racist assumptions made someone less likely to vote for Trump.

Regression coefficients tell us that there’s a large difference in vote preference between people who scored high and people who scored low on the symbolic racism index. However, regression coefficients can’t tell us which end of the scale is more meaningful. When we compare groups of people, regression models don’t tell us what the bull’s eye should be. We build assumption in to the model, and these assumptions shade how we initially interpret results.

Thankfully, the American National Election Study has asked these symbolic racism questions on a fairly regular basis since 1992. We don’t need to make an assumption about whether symbolic racism surged upwards among Republicans in 2016. We can just check the data. In the interests of space, I will present changes in the index instead of going through each of the four questions. Note that this graph measures party identification not who someone voted for. It includes non-voters. “Independents” who lean towards a party tend to act like partisans, so they are included as partisans here.


Every group reported somewhat lower symbolic racism in 2016 as opposed to 2012’s high water mark. This doesn’t really seem to fit with Trump, his base, or the regression results from earlier. One possibility is that racists are learning “politically correct” ways to answer these questions. This happened with more explicit questions about racism. However, there is little evidence that Trump’s base has tried to hide or give safer socially desirable answers in public opinion polls.

The biggest change in symbolic racism scores for 2016 is White Democrats’ large decrease, moving to nearly the same level as non-white Democrats. The symbolic racism score for White Democrats was relatively unchanged from 1992 to 2012. Trump’s campaign may have polarized voters more than it pulled them towards racism. Explicit racists got more of a mainstream voice in the Republican nominee. At the same time, whites who did not want to endorse racial resentment may have moved to more explicit anti-racist positions.

We see such a large difference in symbolic racism scores between Republicans and Democrats in 2016 largely because White Democrats scores’ dropped so dramatically. The regression model that says an increase in symbolic racism made it more likely someone would vote for Trump is technically correct. But that’s not the best way to interpret the data, since the average Republican actually had a slight decrease in their symbolic racism score too. People with the most racial resentment were probably voting for any Republican. The real story in the ANES appears to be a group of white voters moving away from racial resentment, and then voting against Trump because of his explicit racism.

The “secret sauce” to quantitative research is being able to make good arguments and inferences from the data we have. It’s not that hard to build a quick and dirty stats model just by relying on our assumptions of how variables “should” work. Imagine that model fits our assumptions perfectly, like the symbolic racism-Trump voter model. It’s awfully tempting not to dig deeper. Apples to oranges comparisons play in to some of our worst cognitive biases, because it’s so easy to treat one group as the default and the other as the weird thing that must be explained. The tribalism and hyper-partisanship of today’s politics only makes this bias worse.

I spent so much time thinking about apples to oranges comparisons of media content when people were searching for evidence of “bias” that I’ve always been a little more skeptical of the logic behind these comparisons. It’s not a skill that I developed specifically to look at partisan politics. It’s not a skill that I developed from years of stats classes either. The ability not to rubber stamp apples to oranges comparisons that fit our preconceptions is something different. In some ways its harder. It takes courage to acknowledge our preconceptions may be wrong.

In other ways, the ability to question our biases is easier to learn. We don’t need special classes in stats to develop this skill. Every year I taught stats there were several students who walked in to the class having mastered it. When I was a teaching assistant, I think the class that actually focused the most on this skill was a freshman level introduction to linguistic anthropology course. We were going over differences between how men and women communicate. We didn’t just cover differences and debate nature vs. nurture. We spent more time going over whether there are certain stereotypical differences that we expect to see, so those are the only differences we remember. Naive assumptions can become self-fulfilling prophecies unless we consciously try to check them. The good news is we can all learn to check them.


When Will Republicans Bail?

With one shocking headline after another, many progressives are starting to wonder if there is anything that would cause Republicans to stop supporting Donald Trump. Historically, partisans tend to rally around their president during a scandal. As the prior two links explain in more detail, Republicans didn’t start jumping ship until near the end of the Nixon investigation. Reagan’s approval among Republicans never dropped below 73 percent. Bill Clinton actually grew in popularity during the Monica Lewinsky investigation, since he was able to portray it as the work of a rabid partisan prosecutor.

It seems unlikely that leading Republicans who had backed Trump in the past will quickly bail on him now. Every prior backing ties these Republicans closer together. Any Republican who breaks ties now will face questions of “why didn’t you do this sooner?” The Republicans who break with Trump over the Comey firing and/or Trump’s ties to Russia will probably be the same Republicans who refused to endorse him in the general election and have otherwise taken a stand.

It looks like we need to wait for Republican voters to turn on Trump. Then Republicans in Congress will jump off the sinking ship. That’s not exactly the most optimistic proposition. After all, Trump faced a major scandal one month before the election when the Access Hollywood tape was released. Trump boasted about committing sexual assault on this tape. Anderson Cooper confronted Trump during the next debate about whether Trump understood what he was admitting to. Many progressives thought there was no way Trump could win after the tape came out…but he won the Electoral College anyway!

Obviously, a lot has happened in American politics since the Access Hollywood tape and the election. Voters who were willing to give Trump a chance could always say enough is enough. But how come voters didn’t reach this conclusion during the election? I thought it would be worth checking the 2016 American National Election Survey. The ANES has two waves, one before the election and one after the election. In the post-election wave, they asked two questions specifically about the Access Hollywood tape:

In October, the media released a 2005 recording of Donald Trump having a crude conversation about women. Have you heard about this video, or not?

95.54 percent of respondents said they heard about the video. There’s little partisan split: 95.11 percent of Trump voters said they heard about the video. So what did they do with the information. The ANES didn’t ask “did the video affect your vote?” Instead, they asked how others should use the information:

In deciding how to vote, how much do you think the information from the video should have mattered to people?

A great deal, a lot, a moderate amount, a little, or not at all?

I’m not an expert on polling language. My naive assumption is that asking should this have mattered to people is better than asking how did this affect your vote. Strong partisans probably made up their mind about the election well before the tape came out. However, strong partisans may also have the strongest feelings about what moderates should do with this information. To start off with, let’s look at some crosstabs for how people answered this question, based on who they said they voted for:

Video Should Matter… Other Voter Trump Voter Overall
A Great Deal 55.46% 1.99% 32.07%
A Lot 20.80% 4.37% 13.61%
Moderate Amount 13.26% 20.67% 16.50%
A Little 6.36% 34.59% 18.71%
Not At All 4.12% 38.39% 19.11%

There’s a definite partisan split here. There are also some differences within each group. Are there any other variables that explain how voters thought people should make sense of the Access Hollywood tape? If there’s something here beyond the basic “who did you vote for?” this may give us some clues about how people will react to Trump’s more recent scandals.

To test this possibility, I ran an ordered logit regression. The ologit model assumes we can create separate bins for each response category and place them in order from left to right. Each independent variable pushes respondents to the left or the right. For example, voting for Trump probably pushes people to the “the tape should not matter at all” side. The ologit model also assumes there are no walls or speed bumps that would keep a respondent in a middle bin instead of going all the way to an extreme if the value on the independent variables is high or low enough.

Here are the independent variables I used:

  • trumpvoter: Did the respondent vote for Trump or someone else? (Non-voters are in the ANES but excluded here)
  • male: Is the respondent male?
  • senior: Is the respondent a senior citizen? (Age 65 or older)
  • collegegrad: Is the respondent a college graduate?
  • nonwhite: Is the respondent non-Hispanic White or not? (I tested more specific racial categories)
  • foxindex: Did the respondent watch The O’Reilly Factor, Hannity or The Kelly File on a “regular basis” the month before the election? I added the yes responses together. The ANES tells respondents to check the box if they saw a particular TV show once in the last month. It’s a very low bar for TV news. Nonetheless, 65.42 percent of Trump voters have a zero here.
  • msnbcindex: Did the respondent watch Hardball, The Rachel Maddow Show or All In With Chris Hayes? I added the yes responses together for a 0-3 scale.
  • anynightlynet: Did the respondent watch any of the network nightly news broadcasts? Since these broadcasts are direct competitors in the same time slot, an index doesn’t make sense here. It’s just a yes/no variable.
  • ideology_post: The respondent’s self-described political ideology on a seven point scale, where 1 is strong liberal and 7 is strong conservative. Respondents who said they haven’t thought about their political ideology much were excluded (when I double-checked in a separate analysis, they were like moderates on this question). Strong liberals are the omitted category.

Here are the results. Positive coefficients push people towards saying the Access Hollywood tape should not matter at all. Sadly I have to screenshot this, so apologies for the mess:

Screen Shot 2017-05-19 at 8.56.54 AM

Screen Shot 2017-05-19 at 8.53.49 AM

Even after I added a bunch of control variables, whether or not someone voted for Trump is still the biggest factor determining how much they think the Access Hollywood tape should have mattered to people. It’s not surprising that Trump voters would back their candidate and tell others to ignore the scandal. Loyalty to a particular candidate (or maybe a party) blows most variables out of the water. The gender difference is so small that it is not statistically significant. Fox News didn’t push viewers further right here, although MSNBC pushed viewers a little further left.

The big surprise is the effect of political ideology. When I first ran this model, I treated ideology as a linear variable. I didn’t expect there to be anything all that dramatic. Using political ideology as a categorical variable was one of those last minute “I better double check everything before hitting publish” situations. In this model, strong liberals are the omitted category. The regression coefficients measure the difference between moderates and strong liberals, strong conservatives and strong liberals, etc.

Everyone is considerably to the right of strong liberals’ feelings about the Access Hollywood tape. Maybe a better way to put it is strong liberals felt very strongly that the tape should matter a great deal to people. Other respondents had more mixed opinions. After controlling for who someone voted for, there isn’t much of a difference between moderates and strong conservatives. To help make these regression coefficients more concrete, I used Stata’s margins command to give predicted probabilities for some respondents:

Not Trump Not Trump Trump Trump
Strong Lib. Moderate Moderate Strong Cons.
A Great Deal 87.36% 42.40% 4.02% 3.38%
A Lot 7.82% 24.30% 6.66% 5.71%
Moderate Amount 3.52% 18.44% 20.56% 18.46%
A Little 1.00% 6.83% 34.98% 34.58%
Not At All 0.30% 2.22% 33.78% 37.87%

There is a big jump among people who didn’t vote for Trump between strong liberals and moderates. (The median Clinton voter identified as slightly liberal.) There is another big jump between moderates who voted Trump versus moderates who voted for someone else. However, the difference between moderate Trump voters and strong conservative Trump voters is minimal.

Strong liberals rallied around the Access Hollywood tape. I don’t think any of the strong liberals I knew gave Trump a chance of winning after the tape was released. All the criticism of Trump being unqualified because of his incompetence and racism turned in to criticism that Trump is morally unqualified because he bragged about sexual assault. How could anyone but the most committed conservative vote for this man? I was always a bit dubious about this argument. Historically, the United States is a bit of an outlier in expecting moral purity from heads of state. Some voters are deeply affected by seeing someone who bragged about sexual assault in the White House. Other voters are more selfish, and mainly want to know what government will do for them.

Without any polling or data, my guess is that the current scandals surrounding Trump will play differently. Trump fired the head of the FBI and is making sweeping changes to law enforcement philosophy. Trump gave classified intelligence to the Russians and may have deep ties to Putin. This would fundamentally weaken national security. Trump’s current scandals are less symbolic. It’s easier to connect Trump’s latest actions to dangerous policy. If strong liberals focus on the tangible implications of Trump’s scandals – not just the symbolism – it may be possible to pull moderates and weaker conservatives away from Trump.

So Trump Had a Press Conference…

I’ve mainly been working on non-politics projects lately, so I haven’t seen any of the Donald Trump press conference yet. I’m not going to try and assess his performance or the news coverage of that event specifically until I watch the film. On the other hand, I have published on presidential press conferences, so I wanted to cover a few broad rules of the setting that should translate to this week’s press conference or any future press conference:

1) Subsequent news coverage will focus on what Trump said, as opposed to getting a wide balance of opinion, unless a news organization dedicates a lot of space to the press conference

One of the main things I wanted to study in my dissertation was whether politicians could do things that affects the balance of opinion found in news coverage. I looked at press conferences from 1891-2009 since we can compare what was said to the subsequent news coverage, which is rare in news interviews. I measured balance of opinion as a proportion of quoted words from the president to all quoted words in news coverage of an event. The New York Times‘ average proportion was 0.757. In other words, nearly 76 percent of the quoted words came from the president versus 24 percent for everyone else. ABC Nightly News’ proportion was 0.695.

I found surprisingly few variables affect this ratio. As we might imagine, if a president has a joint press conference with another foreign leader they tend to get some quotes in the news as well. However, there was no significant partisan difference after controlling for different ways that presidents set up their press conferences. Approval ratings have tiny, insignificant effects. One peer reviewer was convinced that presidents who had been in the news a lot wouldn’t have as much control. After weeks of gathering additional data I found this had literally 0 impact.

The main thing that leads to more balance of opinion is whether news organizations leave enough space to quote people other than the president. When the Times or ABC only published one short story, that story would only try to summarize the main things the president said. The single most common outcome is a completely one-sided story. More important conferences will get multiple stories, each focused on a different issue or incident. This is when we get more sources.

2) The questions and the subsequent writeup are two different things

Press conferences are a performance. Since presidential press conferences are on TV, there’s more room for journalists to perform, not just presidents. Many journalists want to brand themselves as tough interrogators, holding presidents accountable. With the rise of partisan news, the image of holding people accountable has become less about “I’d grill everyone, even my own mother!” and more about partisan allegiances.

Working journalists know that they need to produce some story at the end. If they keep pressing the president on one topic, he may cancel the press conference or simply not say anything interesting. Since one press conference could lead to five stories on five different topics, there is pressure to move on instead of harping on a particular topic. The more a president opens up about a topic, the more there is for a journalist to write about. Aggressive flashes during the press conference can turn to deference afterwards. On the other hand, exchanges that seem innocuous as they happen may attract much more attention after the conference, when people can review the tapes.

3) The performance is news

Journalists have written about how people conducted themselves in on-the-record presidential press conferences for as long as presidents have held them on the record. I looked at these stories dating back to the Times’ very first story about Eisenhower doing one press conference for print media then inviting news cameras in for a second, shorter conference that repeated most of the same questions. Day-to-day news reporting in the Times and other objective news organizations follows a strict rule of avoiding judgments about a source’s moral character. Press conferences are an exception. Someone’s performance and character is fair game. One of the more common strategies is to have some reporters focus on the literal facts while one reporter focuses on the conference as a performance.

I assume a lot of people are writing about Trump’s performance as we speak. You may think journalists focusing on theatre is another one of those unusual Trump things. It isn’t. Journalists only gave a few years before hammering away at Eisenhower for poor performances in front of the camera:

“One purpose of today’s Presidential news conference was to promote acceptance of the Administration’s new Middle East security policy, but two incidents occurred there that are likely to have the opposite effect.”

Throughout the article the reporter focused on Eisenhower’s communication style: “casual remarks…that implied that the United States might well have to use small atomic weapons” and his “extemporaneous” answer to another question. Five months later, Eisenhower’s “casual dismissal” of Soviet Premier Nikita Khrushchev’s appearance on CBS was used to imply “the tendency of the Administration to continue regarding propaganda as a nuisance rather than as an opportunity.”

4) Public opinion doesn’t affect journalists

My adviser Steve Clayman – along with several colleagues – studied whether unpopular presidents face more aggressive questioning. He found no relationship between approval ratings and any form of journalistic aggressiveness. I found no relationship between approval ratings and how much attention a press conference gets or the balance of opinion in those stories. It’s probably a good thing that the press will not be more deferential to a popular president. On the other hand, this means Trump will not get additional scrutiny just because he is deeply unpopular.

Stepping Away From the Rage

You might be surprised that I went dark for a few weeks after the election, but now I am writing a flurry of posts. I’ve actually had some of these posts written in draft form for weeks. I wanted to wait out the circular firing squad that I’m seeing on left-leaning social media. Unfortunately, it’s December and the circular firing squad isn’t going away. The absurdity is getting a bit much, so I’ve got to say my peace and move on.

As someone who spent years studying how different news organizations gravitated towards different topics in the 2008 election, I expected the same in 2016. Candidates have different things they care about, different priorities for the country. They also gravitate towards issues that poll well for them. We saw this kind of asymmetry in the primaries. Bernie Sanders was all about income inequality and then added political reform to reflect his struggles under DNC rules and new followers’ anger towards the DNC. Hillary Clinton emphasized race and gender to try and capture Obama’s base along with her foreign policy experience. For Republicans, Ted Cruz emphasized local government and Christian values. John Kasich emphasized pragmatic experience. Donald Trump emphasized immigration, trade and being an outsider who could make better deals to clean up DC.

In the general election, both sides converged on the campaign agenda. Trump made his campaign about his own personality, and Clinton agreed to make the campaign about Trump. People talked about race and gender largely via reference to Trump’s behavior and critiques of Trump. That’s why it feels like he has a mandate and why many on the left feel so devastated. I think a lot of progressive put all their eggs in the “make this election a referendum on Trump” basket and it didn’t quite work out. The symmetry of Campaign 2016 is part of why it got so nasty.

I’ll get back to this idea of symmetry and how it is haunting Democrats’ attempts to process the election in a bit. First it is important to recognize just how narrow Trump’s victory was. As a sports fan, I know narrow championship defeats feel much worse than getting blown out. A whole bunch of things had to break in Donald Trump’s favor or against Hillary Clinton. This list is adapted from a longer list David Roberts offered at Vox of possible reasons for Clinton’s loss. I’m not always a fan of Roberts’ work, but he did a good job in the beginning of his essay laying out all the explanations that other progressives have offered:

  • An unpopular candidate in Clinton
  • Poor Clinton strategy on which states to focus on
  • Lack of Clinton outreach to white working class / rural voters
  • Too much emphasis on Trump’s character vs. Clinton’s economic plan
  • The FBI, Russian hackers, Wikileaks all worked against Clinton
  • The media emphasized Clinton’s e-mails over any other story
  • Voter restrictions in states like North Carolina
  • The electoral college made Trump votes more valuable
  • Voters may have disproportionately chose third party candidates over Clinton
  • Racial and gender politics driving at least some voters to Trump as opposed to disqualifying him in voters’ minds.

After giving his long list, Roberts offers one of the more insightful things I have seen any Democrat say after the election:

Like everyone, I buy some of these more than others. But there are bits and pieces of evidence for all of them. Some of them don’t hold up on their own — voter suppression probably didn’t swing the election, nor did third-party candidates — but all of them plausibly played a role or have some grain of truth.

Even before giving the list of what went wrong, Roberts explains what went right. Hillary Clinton won the popular vote. By the time all the votes are counted, Clinton should have the second highest vote total of anyone who has run for President of the United States. Trump earned 2.4 million fewer votes and squeaked through the Electoral College. Since Trump had such a narrow victory, it’s easy to say that Clinton would have won if just one thing broke differently.

I think this is why there has been so much unproductive finger pointing on the left. Clinton came so close to winning that everyone can convince themselves “I held up my end! It’s that other part of the Democratic coalition that let us down!” For starters, here’s Bernie Sanders:

You may recall that Sanders’ biggest problem in the beginning of his campaign was emphasizing economics almost exclusively. He said racial inequality was secondary to class. Black Lives Matter activists targeted his rallies for protests, charging the stage a few times. Sanders moved towards the left on race for the primary. But now that the Democrats lost the general election, Sanders is saying he was right all along. He hasn’t exactly been received warmly by other progressives.

Roberts posted these tweets in his Vox piece and then piled on:

There isn’t a ton of evidence that an economically populist message — divorced of appeals to xenophobia or white resentment — moves the WWC. In fact, as Andrew Prokop notes, “In two of those crucial Midwestern states that flipped to Trump, Democratic Senate candidates [Russ Feingold and Ted Strickland] campaigned on economically populist platforms — but they did notably worse than Hillary Clinton.”

Why is that?

Perhaps because politicians know, though won’t say, that appeals to xenophobia and white resentment work. If I may coin a phrase: It’s the white resentment, stupid.

I’m not sure I would want to build any kind of broad, sweeping theory over two data points. Feingold and Strickland were both politicians who tried to restart their careers after being voted out of statewide office. They both lost to incumbent Republicans. Defeating an incumbent is difficult! It’s hard to ignore incumbency and these Democrats’ previous losses unless you are so convinced race is the only answer that you ignore every other possibility. I have to admit, this is where I found Roberts very hard to follow. It feels like he’s ready to concede Ohio, Wisconsin, and every other state with a significant rural white population. Can that be right? Sorry, but I’m not one to throw in the towel.

Maybe there’s a weakness in the progressive coalition. Once someone is convinced that there is one best explanation, there’s a tendency to point fingers at every progressive who offers a different top priority.

Most of us have a wide range of priorities. If a pollster asked “what is the single biggest problem facing America today?” I could give a clear answer. Give me a minute and I can give issue #2 and #3. We prioritize how much we care about politics too. We can prioritize what kinds of things are dealbreakers on a first date and what we want most out of a career. Setting priorities is one of the basic things we do as people. It’s a basic thing for successful organizations as well. But when progressives say they care about one issue more than another, other members of the progressive coalition attack them for “selling out” part of the group.

The Republicans have their nasty fights too. How often has Ted Cruz accused someone of “selling out” conservative principles? However, the large Christian conservative faction embraced Donald Trump, a man who repeatedly worked with Playboy (should have checked his IMDB page). At the end of the day, most Republicans prioritized winning the White House over anything else, so they voted for Trump. After the election, Democrats are currently fighting a nasty battle about whether to prioritize broadening the party to win elections or ideological purity over race and racism. Anyone who wants to do any kind of political advocacy is going to face questions of how much to focus on ideological purity versus attracting a broad audience. These aren’t easy questions, and the answer probably depends on what you want to accomplish.

Progressives are currently fighting over one of the toughest issues they have ever confronted: how to argue against racism. There is no research that suggests calling someone a racist will cause them to say “aww, shucks, you were right!” Professors who study political persuasion tend to agree that immediately dismissing someone as racist, sexist or homophobic may be one of the least persuasive arguments out there. As I was writing this, this very premise of making political arguments that could persuade other people came under attack in certain academic circles:

Taking these tweetstorms at face value doesn’t add up. My hunch is that there is just enough of an unspoken difference in priorities for people to get angry with each other. Goff wants to focus on more active racists and define a set of behaviors or attitudes as unacceptable. Singal’s priority is thinking of how to best communicate with people who don’t explicitly endorse racism but voted for Trump anyway. Reading these back and forth tweets as an outsider, I think everyone is working with a different definition of what “persuasion” means.

When I taught students about political persuasion, I tried to move them away from the idea of flipping someone’s opinion on an issue. I knew the students who would care most about my lecture were people who felt very strongly on one side of an issue or issues, and they hated people on the other side. So I started the lesson by explaining that if someone is strongly committed, they are going to be incredibly difficult to budge. However, most people aren’t that strongly committed on most issues. The less committed can be nudged and prodded to care about the issues we care about. Instead of trying to convert the worst of the worst, I would try to prod the less committed to say “oh, ok, I’ll do it your way if I get a little something in return.”

There’s an interesting parallel to protests here. You might think protesters are all ideologues jumping up and down. It turns out that successful protests are focused on particular issues and tolerant of ideological differences. Setting priorities is critical. Positive priorities – explaining to people what you want to do – are far more successful than protesting just to critique someone. Effective protesters get friends and family in the door via social networks and worry about ideology later. Ziad Munson, who studied the pro-life movement, argued the right is far more successful at this. Think about all the sniping coming from different corners of the left. It’s a minefield! How many people do you think would look at the circular firing squad and say sign me up?

Other successful organizations tend to focus on their strengths and mitigate weaknesses. Let’s think of the Trump campaign. Trump excels at putting on a show, conveying emotion, and provoking people. He is weak at policy. He made his campaign about conveying emotions and provoking opponents to near constant distraction. To borrow from sports, Donald Trump set the tempo for this campaign. Hillary Clinton and other progressives were happy to play at this tempo, waging their own emotional campaign. I feel like I have to re-state this for emphasis. Donald Trump waged a highly asymmetrical campaign, but Clinton and other progressives decided to engage on Trump’s terms. It sort of worked too. Clinton was relatively weak at conveying emotion. She would have fared better if people focused on policy. And she still got more votes!

But Hillary Clinton is moving on. It’s time to think of Democrats’ strengths. This is more than a bit nerve wracking. Democrats’ main strength in this election was getting more aggregate votes for the presidency and the Senate, but Trump won the Electoral College and Republicans control the Senate. Republicans control the House and dominate at the local level. Yes, the electoral map favors the Republicans. But the only way to change the system is to get large majorities within the current system.

With all the anger surrounding the election and post-election squabbling, I started dreaming of a very different kind of government. Imagine a city where there were concerts outside city hall every week during the summer. It’s not a partisan rally. It’s not even a political thing. City Hall just happens to have a nice courtyard that makes for a great, low cost outdoor venue. Some people wander by for a few minutes before going to one of the local bars or movie theaters. Other people bring their lawn furniture. This is a regular summertime event next to one of my favorite coffee shops. When I dropped by to check out one of the larger concerts, I kind of got the feeling some people go every week. The band is great. Everyone feels welcome, like they belong in the community.

In 2016, the Democrats are the only major political party with the potential to imagine a society where everyone has the potential to make a valuable contribution, and everyone feels like they have the right to belong. Who wouldn’t want to live in a positive place where we all work hard to support each other and lift each other up?

Trump held a lot of rallies, but there was always suspicion and demonization of interlopers. He campaigned on the idea that no matter how much our neighbors have sacrificed to make America great, only a certain group of loyalists count as “true Americans.”

Our country was founded on an idea that people were created equal, with certain unalienable rights. We founded a democracy under the idea that giving people the right to choose their leaders was the best way to promote life, liberty and the pursuit of happiness. That doesn’t mean everyone was granted full citizenship right away. Every generation had its doubts about whether certain “undesirables” deserved the right to vote. Over time, we let more and more people vote and become full members of the community. There was fear. There was anger. But each time America emerged full of vibrant life, with liberty and happiness for more of our community.

I understand why people feel angry and outraged. I understand why people think resisting Donald Trump’s massive disregard for the Constitution requires picking fights at every turn. Remember, picking fights at every turn is Trump’s strategy. Are there progressives out there who think the best plan is to try and beat Donald Trump at his own game? If you oppose Trump’s attempts to deny Americans equality and unalienable rights, you can’t campaign on a message that many Americans are too stupid or biased to possibly be redeemed. Those emotions are too discordant. Democrats are the only political party that could imagine a future full of life, liberty and happiness for everyone currently living in the United States. That vision only works if everyone thinks they could be welcomed in to our community.


How Distracting “News” Works

As you may remember, Mike Pence went to the theatre a few weeks ago. Normally we wouldn’t care if a prominent politician went to see a Broadway play. However, Pence was booed when he walked in the theatre. Who would have thought that one of the most anti-gay politicians in America would be persona non grata on Broadway? It wasn’t “Chase Utley breaks Ruben Tejada’s leg then goes to Queens” booing. Nonetheless, news that Pence was booed on Broadway spread like wildfire. Then we got the seemingly inevitable backlash. Why were more people talking about Pence getting booed while watching Hamilton than Donald Trump settling the Trump University case by paying a $25 million judgment?

I’ll start with the most basic question: why were so many people talking about Pence getting booed? People can read the story and offer their snap judgment in under 30 seconds. There’s nothing more to the story than vice president-elect booed at theatre. People don’t need to know the plot of the play or Pence’s specific policies to ring in. The story encourages people to take clear sides: is booing the vice president-elect at the theatre a breach of decorum?

Because this is an entirely moral question, there isn’t a real risk of looking like a fool because we are poorly informed. I may not want to tweet about the Trump University case because I don’t know whether $25 million represents a large or small settlement compared to other cases like this. I don’t want to act like one of those PhDs who insists I know everything about everything! But you don’t need a PhD or even a high school degree to give an opinion on whether booing in the theatre is wrong. Middle schoolers could give a presentation to the class about it.

Debates over whether booing Pence or settling the Trump University case are more important say something more fundamental about news preferences. We normally think of news preferences as a list. So the list of things I thought were newsworthy that weekend went something like this:

  1. Trump settles Trump University case
  2. Trump nominates National Security Adviser whose rhetoric on Muslims fits ideally in to ISIS recruitment (they love the idea of an ideological holy war).
  3. UCLA is playing USC in the Crosstown Rivalry (* during the game this jumps to #1 through 100)
  4. It’s raining in LA and everyone forgot how to drive
  5. UCLA basketball looks good
  6. Trump’s other nominations
  7. I’m getting crushed in my NCAA confidence pool this week
  8. [Long list of stuff I don’t really care enough to offer in detail. I wrote the list down at the time, then wrote the rest of the post over a week later.]

Some of my list is pretty idiosyncratic. People who don’t care about college sports will cross off several items from this list. Since it’s a weekend with college football and basketball, both sports are well represented in 8-20 on my list. You probably have a different list of most newsworthy things from the weekend. That’s good! We are different people. We probably have different hobbies, different political priorities, and different emphasis on politics vs. other things in life. Maybe it’s a little frustrating that none of my Facebook friends enjoy the Piesman trophy. But life would be almost intolerably boring if we all agreed on everything.

Social Media Lists of Newsworthiness are Weird

You might assume that all of us have one internally consistent list of things we care about. In other words, the list of things we read, the list of things we share, and the list of things we comment on would all look the same. It turns out that may not be the case. How often have you liked or commented on someone’s post before actually reading the link? It turns out this happens a lot. People in new media production have realized there’s relatively little correlation between what people read and what people share or comment on. The list of things we find newsworthy enough to discuss on social media is not the same as the list of things we consume for our personal reading. A year ago, Atlantic writer Derek Thompson went through his top 100 tweets with links, trying to compare how often someone actually clicked on the link vs. other Twitter activity.


The scales on the graph are a bit wonky. I’m pretty sure the x-axis is a ran order from highest click through rate (a whopping 6 percent!) to his 100th best tweet (around 1% of readers clicking on the link). Thompson said his overall click-through rate was 1.7 percent. Now let’s focus on the big picture. If people had the same priorities for their personal reading and their social media discussions, these two lines would go in parallel. They don’t. This suggests there are some stories that people really enjoy talking about on social media, even if they wouldn’t dig deeper. Mike Pence getting booed during Hamilton is a perfect example. People can take sides without having to dig deeper.

Aggregating Individual News Preferences Makes Things Much Weirder

If we add all of our priorities together, you may expect another list: here’s what people think is the most important story, here’s number 2, and so on. That’s why we debate whether or not Hamilton is #1 as opposed to #2 or 5 or 25. If you look at how often people talk about each story, our group behavior is probably closer to a pyramid than a rank order list. A few events are near the top of enough individual lists so they shoot to the top of the news pyramid. Think of Super Bowls and presidential debates this year. It’s not just that the leader is #1 and the next most popular story is #2. The leader gets vastly more attention than #2, then the distance between 2 and 3 is smaller, and so on. Eventually we get to the point where the group is split. There may be a tie instead of a clear #5 story. The further down we go, the more we get to things that are on a bunch of lists but fairly low down, or things that are high on a few individual lists but don’t register on anyone else’s in the group.

If this seems a little abstract, let’s take a detour to the music industry. Every year, there is one artist who is the top seller. They usually have a crossover hit that fits in to some genre, pop, Top 40, and maybe some secondary genre. Then you have a few other hit albums that get crossover appeal. Then there is the larger group of albums that only sell in their genre, or maybe they have a few huge singles but can’t sell the album. At the bottom end, there are lots of people recording music at home who don’t earn any money.

News has the same unequal distribution, but the only way to notice is if we have a database of all the potential news stories that didn’t get in a news outlet. Most of us have never thought of news this way because journalists don’t exactly leave lists like this lying around. If you thought people second guess the media now, imagine what would happen if the New York Times published a list of “here are 50 stories we decided not to pursue further.” They’d never hear the end of it! We can come up with our own lists of stories we’d like to see and feel are left out just by using our imagination. That’s why we assume that no matter how many people we put together, we’d still get a rank order list. When I was a journalist, I assumed if I put every journalist together I’d get a clean rank order list, not a pyramid.

For my Master’s Thesis I wanted to look at whether politicians’ attempts to duck a question by talking about something else had any influence on what journalists wrote. I knew I couldn’t just look at news coverage to answer this question. I had to look at what a politician said and then what got published to make a before and after comparison. Because journalists won’t just hand over interview transcripts, I decided to use presidential press conferences as a dataset. The only way to make sure I separated the quotable statements from the unquotable statements, I had to analyze every single statement! Going through four conferences – a total of 1743 statements – took months. I wanted to code things like was the president criticizing someone, is he ducking a question, is he talking about the military vs. the economy, etc. Because any newspaper can be idiosyncratic, I used the New York Times, Washington Post, Los Angeles Times and USA Today. (Newspapers only is a bit old school, I know.) Out of those 1,743 statements, how many do you think were quoted in each newspaper?


{Here is some empty space so you can play the guessing game if you want to}


The correct answer is four! Leading newspapers tended to have some variety. Only 25 statements were quoted in three of the four newspapers. The full distribution looked like this:


I know, this may not be what you were expecting. When I presented my MA at the leading sociology conference, the veteran professor presiding over my panel didn’t understand this kind of statistical distribution. To be fair, I had to teach myself how to run statistical models with this kind of unequal distribution because it wasn’t covered in my grad level stats classes.

Here’s why it is so rare to have complete agreement on a quote. Part of it is New York Times and Washington Post allocated more space to presidential press conferences. The main explanation is that four writers covering the same story rarely choose the same quote if they have options to choose from. When George W. Bush spent five minutes criticizing Saddam Hussein, two reporters may choose criticism A and two may choose criticism B. This gets us to the upper-middle level of the pyramid. Presidents rarely say anything as direct as “We do not recognize the outcome of the election [in Zimbabwe] because we think it’s flawed.” When Bush said this in 2002, it was clearly a better quote than anything else he said about that election and each newspaper quoted it. Other scholars have seen similar unequal distributions with how often a particular person (van de Rijt et al 2013) or group (Amenta et al 2009) gets quoted in the news.

Mike Pence getting booed when he went to see Hamilton has a natural advantage for vaulting to the top of the pyramid. Pence didn’t say much of anything himself. Trump posted one tweet about how he was offended. The Hamilton cast read one reaction from the stage. Everyone talking about what happened will refer to the same very small set of information. I know, it seems like such a small thing that it’s not worth harping on. If you saw this story and thought “who cares if Pence got booed?” then me talking about a small set of shared information won’t make you care. It doesn’t make me have strong opinions on the morality of booing. The reason I’m talking about this small detail is because it has a surprisingly powerful effect in focusing aggregate behavior. When I looked at the most common phrases on the Internet in 2008, I had to control for the titles of songs, movies, TV shows and even video games. Since people repeated the title over and over, it biased my results for how often particular websites talked about politics compared to other topics.

Why Instigators Can Be So Much More Effective Today

If any simple story with a limited number of facts can rise to the top of our aggregated preferences, moral outrage is the rocket fuel. Stories like Pence getting booed encourage people to take sides. Any cultural instigator has a huge advantage in getting attention. They just need to make one offensive statement and it distracts people on both sides. Donald Trump used this strategy since the early days of his campaign. People were so busy talking about Trump and how offensive he was that they lost focus on how they would help voters’ lives. Republicans got caught in the web one by one. Hillary Clinton ran an almost entirely negative campaign about how Trump isn’t qualified instead of emphasizing how she would try to make people’s lives better.

It’s important to remember that Trump isn’t the first instigator to use these distraction tactics. The National Rifle Association has used them successfully for decades. NFL commissioner Roger Goddell deflects negative attention from team owners as they demand public resources for private stadiums.

It’s easy to blame Facebook, particularly with their fake news scandal. However, my press conference research largely predates Facebook. That being said, social media makes these problems worse because it aggregates individual decisions. Any story with least common denominator appeal will do better with the aggregate herd than with any individual. Stories that let people take sides on social media and advocate for their “team” have an even more powerful effect than what I saw from journalists. Individual journalists make relatively independent decisions. People on social media may get even more pleasure by talking about the thing that everyone else is talking about.

So How Can We Avoid Being Suckered in to These Stories?

It’s easy to point the finger at media organizations and say they should have better priorities, or point the finger at Facebook’s mysterious news algorithm. But that’s not really satisfying. I want to be able to take some control. Hopefully you do too. Well good news! One of the best ways to avoid being suckered in to low level scandals is to have strong independent judgment on what kinds of things matter. No one else is going to be the reliable voice saying “who cares about this nonsense?” The only people who bring up a story like Pence getting booed are going to be people who have strong opinions…or someone like me who really misses teaching from time to time.

The other key is to recognize that some people in your social media feeds will probably get suckered in to just about every minor story that lets them share their moral beliefs. No matter what happens, it just seems “too important” not to say how outraged they are. The people with the most influence over the process, starting the cascade of posts, are the people who tend to show the least restraint. Over the last few days I have talked to progressive, moderate, and conservative friends who have all said they are leaving social media. I didn’t even ask. They brought it up. Political posts keep rolling through their feed like crashing boulders, and they want to duck out of the way.

This is a good principle for self-preservation, but the only way to have meaningful change is to convince people to show some restraint before they push the petty outrage of the day on to the rest of us. I understand it’s hard to show restraint and let some things go. But I also know that if people can’t prioritize and let some stuff go, their social media feeds will turn in to a broken car alarm. It’s not a perfect metaphor. When it comes to politics some people actually love the sound of broken car alarms. They love outrage news and think every single action helps move a broader movement! But there are lots of people like me who don’t enjoy outrage news. Persuasion is about connecting with other people, not jumping up and down proclaiming how your morals are superior to everyone else. There’s little upside in going online to vent about every single petty grievance. If you voted against Trump because he can be taunted with a tweet, hold yourself to the same standard.


Scammers, Camelot, and the demand for fake news

Last week I walked past the single worst Black Friday ad I ever saw. A local clothing store advertised “openning” 10-6 on Thanksgiving. Yikes! Was this the worst Black Friday behavior I have ever seen in my neighborhood? It’s close, but I saw worse at the local Best buy a few years ago. While I was shopping for a laptop bag, a woman was complaining to customer service that she was locked out of her accounts. Apparently someone texted this woman claiming that she had won an award. She just needed to send her banking information to get paid. Of course, it’s a scam. I turned around to look at the low level best Buy employee and saw the quick look of terror on his face. How does he explain that this potential customer just got scammed and there’s nothing he can do to help? How can he be sympathetic to this confused woman walking in to Best Buy with her child instead of wanting to scold her for falling for such an obvious deception?

If you’re an old enough Internet user, you probably remember scams involving “Nigerian princes.” In case you forgot, this was a scam where someone sent bulk spam email claiming to be a Nigerian prince who has to move money offshore due to political unrest. If you give your bank account info, they would wire $10,000 to your account. Most people realized this was too good to be true, even before the scam became publicized and tech firms dedicated resources to blocking these spam e-mails. However, there were some people who desperately wanted to believe there was a Nigerian prince who would make them rich. Selfishness and laziness beat suspicion and careful research. The selfish and lazy might be pretty easy to exploit.


If you follow tech news over the last week, you saw headlines that the top 20 most shared stories on Facebook had more fake news stories than real ones. Google and Facebook both blocked fake news sites from their advertising sales networks this week – now that the US Presidential election is over. Facebook has always had an unusual set of “community standards” for regulating content. Visual depictions of violence and sexuality are generally banned. The company frequently claims it has tweaked its “News Feed” algorithm to show “higher quality content” as opposed to clickbait. However, Facebook has always strenuously objected to the idea that it is a media company with a profound influence on journalism. Here’s Mark Zuckerberg, trying to answer questions about whether his company helped Donald Trump win the 2016 election:

“Personally I think the idea that fake news on Facebook, which is a very small amount of the content, influenced the election in any way — I think is a pretty crazy idea. Voters make decisions based on their lived experience.”

I’m not sure anyone really believes this, even Zuckerberg. Facebook acted against fake news sites four days after Zuckerberg’s quote (and after Google’s ban). It’s a field day for people who want to blame Facebook’s lack of transparency – or social media more broadly – for Americans’ declining interest in facts and evidence. As much as people have a right to be frustrated that Facebook didn’t do anything about fake news until after the election, it’s not like Facebook was the birth of online scams. People have tried to use the Internet to try and exploit selfish and lazy users for decades. They used other technologies before the Internet. Instead of blaming Facebook, we should ask why would people want to spread misinformation with their friends and family?


My take here is probably different than most people because I actually covered a secession campaign. In 2002, the San Fernando Valley wanted to secede from the rest of Los Angeles. The secession movement started as part policy oriented and part symbolic. Every public school in Los Angeles County is in one school district. Valley voters wanted to break away from the unwieldy behemoth. They also believed their tax dollars were being used to subsidize the poorest neighborhoods of Los Angeles. On a symbolic level, Valley residents felt like they were taken for granted by a remote city hall. It wasn’t a rural / urban divide like presidential elections. If Valley voters formed their own city of Camelot, it would have been the seventh largest in the country! (Sidenote: Camelot won the vote for what to call the new city.)

By the time I started covering the story, Valley secession leaders had already conceded on their largest policy grievance. Camelot would still be part of the LA Unified School District. I talked to voters who said “what’s the point of seceding if we’re still tied to LAUSD?” Other reporters thought there was no way secessionists could get enough votes. They needed to run up enough votes from within the Valley to get a majority citywide. Secessionists didn’t fully want to campaign on keeping Valley tax revenue in the Valley either. The county ruled that Valley residents would have to financially compensate the rest of Los Angeles if secession passed, to make up for lost tax revenue.

With only a weak policy case, potholes became a major campaign issue! Some Valley residents saw every pothole as a reminder that their area didn’t get a “fair share” of city services. Mayor Hahn dispatched construction crews to smooth over problems, both literally and figuratively. (At this point I am obligated to say I don’t live in the Valley and my street gets enough flooding to become one lane only during moderate rain.) Valley secession leaders wanted voters to feel like City Hall was remote. They also reminded people that most local media organizations were located in the older area of the city and not the Valley, so they were biased against secession. The Los Angeles Daily News – which was based in the Valley – was decidedly pro-secession.

There are several things that make San Fernando Valley secession different than the 2016 presidential election. While the Valley has more Republicans than the rest of Los Angeles, it is still a majority Democratic area. Every voter is urban. The two sides had relatively similar arguments about what would happen if the Valley seceded. Valley secession leaders did minimize the potential disruptions. LA’s black neighborhoods emphasized how they would lose out financially if the Valley seceded, but there were few accusations of racism. In the end, the pro-secession movement was even more based on emotion than Donald Trump’s campaign. Trump promised to make American great again. He made incredibly vague policy promises. Trump’s promises may not be credible. But Valley secession leaders openly said they couldn’t deliver on their initial promises of divorcing LAUSD and keeping all the tax revenue.

A slight majority of San Fernando Valley voters still believed secession was a good idea and voted to leave the city! Why would they believe the promises of the secession campaign, even if all the evidence said secession wouldn’t provide tangible benefits for Valley residents?

  • Disrespect: Various activists had discussed secession for decades. They felt they were not given a full share of public services, even though they paid a disproportionately high bill compared to the rest of Los Angeles. I don’t remember anyone counting what percent of potholes were unfilled in the Valley as compared to downtown. “We aren’t getting a full share” may be one of those things that is entirely symbolic and not based on rational calculation.
  • If the moral claim rings true, people may turn a blind eye to empirical fact: The facts were not on Valley secession’s side. The county set the terms for secession after certifying that the Valley would be a viable independent city. The terms were not ideal for secession leaders. One balked and abandoned the campaign. The rest held on to their cause, despite mounting empirical facts about how any new city could not do what they wanted to.
  • People emphasized moral claims when policy issues were murky: In 2002 the Los Angeles Unified School Board was so dysfunctional that creating a new Valley School Board with new criteria and policies seemed a lot easier than fixing LAUSD. School board governance played a major role in the 2005 mayoral race and Antonio Villaragosa’s first two years as mayor. No one really knew how to solve the giant mess. No one had a good policy idea. Voters may have been quicker to embrace the symbolic politics of secession because no one offered a policy solution to the tangible problem of underperforming schools.

Now let’s think about what we know of Trump voters:

  • Disrespect: Definitely. Trump voters tend to say the federal government has forgotten them. The political class may not focus on rural areas. Remember, disrespect is a feeling that may or may not have a basis in fact. So is neglect. It’s entirely possible for multiple groups to feel disrespected by the power structure, and for those groups to hate each other.
  • If the moral claim rings true, people may turn a blind eye to empirical fact: Trump offered a wide range of campaign promises, and voters didn’t seem to mind when some of these contradicted each other. In particular, Trump’s tax policies clearly favored the rich. When I think of moral claims ringing true, I think back in the first primary debate. Trump was asked if he would back the Republican nominee no matter what. He refused to say yes. Later on he attacked other Republicans for being in lobbyists’ pockets and bragged about buying influence. The moral claim was very clear: “every politician is a self-serving asshole, but I’m the only one who is honest about being an asshole.” That’s when I thought Trump had staying power.
  • People emphasized moral claims when policy issues were murky: A lot of industrial workers are facing downward mobility. There’s a tendency to focus on Trump voters not being the absolute poorest, then discounting any kind of economic argument. But Trump voters are disproportionately older. Regardless of partisanship, most of the parents I have met care deeply about their children’s opportunity to have a good life. I spent my teenage years in an area that transformed drastically from farmland to suburbia. I left for college before the transformation was finished, so I’d go home and my parents were going to new malls that didn’t exist when I was growing up there. All of this was positive economic growth, but it’s still alienating. This affects how I think about declining industrial towns? Has any politician really offered a solution for declining industrial towns over the last 20-30 years? There isn’t a good plan for how to help workers whose skills are less valuable today, or the potential alienation of economic change. We got nostalgia and moralizing about trade instead of real solutions.

I can see reasons why the people who supported Trump would also be likely to buy in to fake news. I can see why people would believe the feelings contained in these stories and want to share them widely.


At this point, I think other left-leaning writers would just look down at Trump voters and stop writing. Buzzfeed’s story about fake news getting more Facebook engagement than real news feels plausible. It’s very plausible if you don’t know many Trump supporters and you’re looking for some explanation of how they got “fooled.” I retweeted the story without thinking twice. I didn’t even read the study! A friend of mine who voted for Trump posted a critique of the Buzzfeed study design over the weekend. I read the critique, then read the Buzzfeed methodology. Hate to say it, but Buzzfeed fooled me.

Here’s the short version of what Buzzfeed did wrong. They looked up the top 20 fake news stories and the top 20 real news stories. Top 20 lists are highly unequal. The #1 hit is far more popular than #2, but the gap between #2 and #3 is smaller, etc. One huge hit skews the entire set. Remember how Buzzfeed generated massive traffic by posting a dress where people disagreed on what color it was? To make things worse, fake news should have a natural advantage in this metric. If someone is creating fictional news, it is by definition a unique story. Legitimate news outlets don’t get many exclusives. Let’s say 10 people share a fake news story about a Trump-Clinton debate, 5 share the Washington Post’s lead story, 4 share their B story, and 3 more share their third story. More people shared information from the Washington Post than a fake news site, but the fake news site has the biggest single hit. In reality, the one fake news site is competing against dozens of high profile real media organizations and getting swamped in the total volume of Facebook engagement.

So why would people believe the Facebook fake news story?

  • Disrespect: Yes. Democrats’ general election campaign was mainly an argument that Donald Trump doesn’t represent the characteristics we want in a leader. Hillary Clinton won the popular vote but lost the election. Disrespect for progressive values may be an understatement for how progressives feel today, particularly if they are focused on identity politics.
  • If the moral claim rings true, people may turn a blind eye to empirical fact: The main people who bought in to Buzzfeed’s fake news story are Democrats who feel big media organizations didn’t do enough to clamp down on Trumpism. Remember how people bought the myth that network TV news was avoiding “issue” coverage – another study based on terrible methods. There that many sophisticated methodologists in the world. I can’t really fault people for not understanding the weird statistical distributions that biased the Buzzfeed study when they were barely mentioned in my years of graduate level statistics classes. (It’s going to take a separate full length post to explain in detail.)
  • People emphasized moral claims when policy issues were murky: In this case we need to delete “policy details” and replace it with “research methodology.” It’s easy to do a simple study of media content that compares apples to oranges. It’s much harder to compile a list of potential stories or sources and then analyze how different media organizations treat them. But this is the only way to see how rare breakout hits are and how much more attention they get than any other story or source. The statistical skills needed to analyze these data sets are also extremely rare. When I gave talks at academic conferences, most of the other panelists didn’t know what a negative binomial regression was, let alone the audience. That’s why the burden is on me to explain.

We might all be vulnerable to news that is factually incorrect but feels true. I bought in to a flawed study in an area where I spent a decade becoming an expert on that methodology, because I didn’t bother to check the facts! How can we avoid falling victim to our own weaknesses?

I think the key is to look for potentially disconfirming evidence. I wrote about this at more length yesterday, but here’s the short version. As a person sitting in my apartment writing a blog, I don’t think I have the power to affect the supply of fake, misleading and manipulative political posts. Sure, I could jump up and down blaming Facebook. But here’s the good news. Each of us has tremendous power over our own demand for fake and misleading news. I know exerting this power isn’t easy. It’s tempting to just look for evidence that supports our claims. If you are a debater or a lawyer, you want to make an argument based on the evidence that is the best possible interpretation for your side. Evidence that we might be wrong stinks. Most of the time we don’t want to look for it, and we feel like fools if we actually find it. So why bother looking for evidence that would hurt our side?

In the end, it comes down to a question of whether we want to focus on our own feelings or convincing other people. If you just want to make yourself feel good, there isn’t much of a reason to look for disconfirming evidence. Just remember that someone out there is going to recognize your laziness. If you’re lucky, they’ll just want to embarrass you for buying in to a myth. If you’re unlucky, they’ll see you as a mark, willing to hand over political power in exchange for the right set of feelings. The best way we can empower ourselves is by looking for disconfirming evidence. We can keep ourselves from being fooled. We can’t rely on others to do it for us. At the same time, knowing other belief systems is normally a pre-condition for persuading other people.

Forget “fake” news. Focus on “news” that lacks fact.

In the last week many of my academic friends who don’t study news have gotten a lot more interested in “fake news” as a potential social problem. First there was the Buzzfeed story claiming that the top 20 fake news stories during the last few months of the election got more Facebook engagement that the top 20 “real stories.” (I’ll talk more about that study itself separately.) Then there was the NPR story on a psychology experiment showing that students couldn’t adequately separate more and less credible sources of information. Since most people don’t have direct experience teaching about news media, let alone this new issue of fake news, there’s a lot of things I could share. It’s hard to know where to start. I’m going to write a few separate posts.

Let’s start with the most basic idea of how to stop fake news. Why not just hand students a list of websites and say “these websites are fake!” Google and Facebook blacklisted a set of websites after the election, trying to keep fake news from benefitting from their advertising networks. Giving people a list of websites makes me think of giving someone a fish versus teaching someone how to fish. Is fake political news written by teenagers in Macedonia more important than politicians selectively giving information to manipulate prestigious reporters? What about the growth in political memes that have no facts, only feelings? I think this New Yorker cartoon sums it up well:



Caption: “I’m sorry, Jeannie, your answer was correct, but Kevin shouted his incorrect answer over yours, so he gets the points.”

I don’t know if the cartoonist is aware of the irony here. There’s a big banner saying “FACTS DON’T MATTER.” There’s a caption that captures part of the reason progressives resent Donald Trump and his electoral victory. But there are no actual facts in this political cartoon. We don’t know what Jeannie or Kevin said so we can judge the answers for ourselves. Of course, giving facts is not what political cartoons do. They try to present clever mockery. I understand why the cartoon feels true. But the key to distinguishing fake news and other kinds of emotional manipulation is being able to separate literal fact from arguments that feel true or symbolize truth without containing verifiable facts.

Just to give a preview of where I’m going, every style of telling stories has strengths and weaknesses. For my final in Sociology of Mass Communication a year ago I asked students how they would write about the Trump campaign. Why did they choose that style? What are the strengths and weaknesses of that approach? If a student couldn’t talk about the negative ramifications of their decision, they couldn’t get an A. I knew that students could think about the cons of each approach, they could walk out of my class knowing how people may try to use that style of story telling to manipulate them. Sometimes it’s politicians using the media. That’s been one of the main themes of this blog in the past. But since people are mainly worried about fake news today, I’m going to focus more on media producers trying to manipulate our emotions.

I didn’t teach about “fake news” per se when I taught sociology of mass communication a year ago. I’d only change one thing if I taught again this Winter or Spring, and it’s something much broader than fake vs. real news. Fake news is just one kind of manipulation. The New Yorker cartoon is another. Donald Trump dragging out the hot takes over Mike Pence getting booed while watching the play Hamilton is a third kind of manipulation. I’m less concerned about the differences between flat-out lies, tactical half-truths and people relying on emotional arguments because they don’t have facts or evidence to back up their claims. Each form of manipulation can be resisted by asking the same set of questions instead of taking content at face value.

1) How much does the article rely on factual claims or evidence? Some articles are entirely descriptive, trying to relay a set of facts for an audience who didn’t directly see what happened. Other articles rely heavily on moral claims and interpretation rather than empirical evidence. That’s a different type of argument and we’ll get to it in a bit. Traditional interview-based journalism offers quotes as a kind of evidence. Direct evidence from documents is rare.

2) If someone relies on factual claims, where does the evidence come from? Can you clearly trace how evidence traveled from the original source to intermediaries to your brain? For example, reporters interview people and quote them. We don’t know what the reporter decided to quote and what was left out. But we can be fairly certain that the person being quoted said those literal words. Then we can evaluate the reputation of the person being quoted. We can also evaluate the reporter’s reputation. If we can’t clearly trace the flow of information from one person to the next, they may want to hide something. Whistleblowers need anonymity for protection. However, a wide range of political operatives seek anonymity to promote half-truths and misinformation.

3) What emotions is someone trying to convey? Are they trying to make you feel a certain way about things? Some writers and meme creators’ goal is to convey a certain set of feelings. Think of the New Yorker cartoon: it wants to convey outrage. Most newspaper and network TV reporters work very hard not to convey any of their own emotions about the stories they are reporting on. This kind of stoic emotional restraint is pretty rare in other kinds of storytelling. My goal is make sure we all pause and consciously understand where a writer is coming from and what they want us to feel. I don’t want to get my emotions pushed around by anybody, even people I tend to agree with. Any time someone is giving us a set of feelings that we want to believe, we are at risk of not checking their facts and evidence as closely as we should.

4) How honest and up front is the writer about why they are doing what they are doing? I actually haven’t taught this before, but I think it’s an important follow-up to question 3. Every argument has assumptions. If someone is making an argument, how much are they willing to clearly state “here are my goals and my assumptions.” When someone is giving an interpretation of evidence, do they explain why they gave this interpretation? Do they acknowledge other potential explanations and make a case for why their interpretation is better? If someone can’t acknowledge that other interpretations exist, it’s probably because they can’t make a good case for why their interpretation is the best.

My own sense from spending my entire adult life working in or studying journalism is that it’s hard for a writer to excel at giving both factual evidence and feelings. I think it’s particularly hard to combine the two when writing about politics, since the main feelings people convey are moral outrage and judgment. It’s much easier for me to try and combine facts and feelings when recapping a baseball game than in writing this post. That’s why there are tradeoffs. No writer can be consistently good at everything. No one is perfect.

When I teach about journalism, my main goal is to get students to acknowledge these tradeoffs, then ground them in specific examples. Elite media organizations that have access and avoid reporters’ personal judgments tend to defer to sources in order to protect access. When Trump lied on Twitter about the popular vote, many leading news organizations copied his claim in the headline without any critical skepticism. Large news organizations tend to fear an inability to prove an elite source is lying more than they fear publishing an elite’s statement that is probably false. Nixon campaign aides first took advantage of this in 1968 (see Crause’s Boys on the Bus), and it’s been a staple in political operatives’ playbook ever since.

Writers who emphasize moralistic takes and emotion have more incentive to hide, selectively misinterpret or fabricate factual evidence. Someone who really wants to convince me that Trump voters are all racist isn’t all that likely to bring up other reasons why they support Trump. On the other hand, Trump voters who are not explicit supporters of the KKK are trying to emphasize all the non-racial reasons why someone would vote for Trump. Three weeks after the election and I still see both messages from friends, like it was the day after the election. That’s why things like the New Yorker cartoon stick out to me. They encapsulate outrage and victimization, but are not going to persuade anyone who doesn’t already have those feelings deep in their heart before reading the cartoon.

Being self-conscious about what we want from news is hard. I think it has always been hard. We don’t want to acknowledge that every genre of media is imperfect. People who want all-facts news may not want to acknowledge how reporters can be manipulated by powerful sources. People who want a certain set of emotions, moral stance or political ideology may not want to acknowledge there are times they put feelings before facts. Trump exploited this his entire campaign, skewing this election almost entirely towards emotion. Remember that Democrats’ main campaign theme was that Trump was emotionally unstable and personally unqualified for office. We took the bait instead of focusing on a positive message. It’s easy to see something on social media, get agitated, and react right away. This is a more pervasive and bipartisan problem than “fake news.” For another example of what makes this so hard, let’s take the following passage from the end of an article Dara Lind wrote for Vox:

Journalists have long been sensitive to the prevalence of misogyny on social media. In 2016, they’ve become alarmed by anti-Semitism on social media as well. Journalists know and work among women; they know and work among Jews.

Many of them don’t know and work among many people of color. The amount of attention paid to racism on social media (or in real life) among journalists is, accordingly, often disproportionately small — or delayed.

Let’s think about how to factually evaluate the claim that journalists pay more attention to misogyny and anti-Semitism than racism. We could try to construct a database with a list of misogynist incidents, religious bigotry and racism, and then construct a second list of writers and what they wrote about. How often did a particular group of writers tackle a particular topic? This is incredibly difficult, painstaking work. I spent years working on a project like this dealing with media and blog posts from the 2008 election as part of my dissertation. In the end, I could produce criteria for defining statements on race, gender, religion and a large number of other topics. I could produce data on how much a particular set of news organizations preferred or dispreferred phrases on each topic, relative to any other topic.

Even if I provide facts, there is no factual basis for saying whether enough attention is being paid to racism or misogyny or any other issue. The current level of attention is a fact. The ideal level of attention is a feeling. It’s a moral stance. I chose this passage from Lind’s somewhat unrelated article because it does a great job of showing how 100% fact based story telling only goes so far. A description of how much attention is currently being paid to race with no moral claim about how much attention should be paid to race will be unsatisfying for many audiences. Trust me, I have the negative reviews to prove it. On the other hand, moralistic claims that lack evidence are unsatisfying for a different audience. Lind claimed that journalists pay a disproportionately small amount of attention to racism. The rest of the article doesn’t give any additional evidence to support this claim since it is largely focused elsewhere. Unfortunately, that means I have no idea if Lind is trying to manipulate me or not.

If you want to teach critical thinking about media content and not just give students a list of fake news sites, you have to empower students to offer different moral priorities than you. I know some students in my last class liked outrage news a lot more than I did. I assume some of my readers will like it more than I do. That’s fine. My students also liked Buzzfeed’s day-to-day content a lot more than I did, while I wrote lectures while listening to three hours of college football podcasts. I made a point to illustrate a few of my eccentric non-political preferences to set a tone that I didn’t expect them to copy my political, moral preferences either. My goal is to help people make better informed choices about what media we read so we can be aware of what we are not getting. I want us to be able to protect ourselves from all different kinds of deception in the news – particularly from writers we tend to trust the most.