THE YEAR OF THE LIBERAL: The Washington Post gets it right!

MONDAY, JANUARY 4, 2016

Part 1—Surrounding that, the deluge:
The Washington Post performed a major service during 2015, a year which ended last week.

Even more improbably, the Post performed a major journalistic service. It compiled a list of fatal shootings by American police over the course of the past year. The paper's voluminous database can be found at this highly informative site.

By almost any definition, the compilation of information is a journalistic service. In the realm of fatal shootings by police, the Post compiled a lot of information during this year-long effort:

According to the Post, 980 people were fatally shot by police during 2015. At least 490 of those people are identified as "white" by the Post; at least 255 are identified as "black." (The "race" of 30 others is listed as "unknown.")

Of the 980 people shot and killed, the Post says 91 were unarmed; that's a bit more than nine percent. In addition to those 91 people, the Post says that 33 others were shot and killed while carrying a "toy weapon"—the kinds of "toy weapons" which are extremely hard to distinguish from the real thing.

Of the 980 people shot and killed, 18 were younger than 18 years of age. For further data, and for capsule accounts of each shooting, you can visit the Post's site, the creation of which was a major journalistic service.

Why are we praising the Post for developing this information? In part, because this type of journalistic effort is exceedingly rare.

In the current state of American journalism, our big newspapers almost never develop information. In the current state of play, our big papers are as likely to hide basic information as they are to develop it.

More often, our big newspapers traffic in narrative and script, along with endless forms of entertainment, distraction and piffle. This unfortunate cultural fact may help explain the problems which surround the Post's important journalistic act.

To what sorts of problems do we refer? For one small example, consider what happened last week:

On Saturday, December 27, the Post published a lengthy, overview report about these fatal shootings. The front-page report, by Kindy and Fisher, ran slightly more than 3200 words.

In some ways, the Post was getting ahead of itself in that front-page report. The year in question wasn't over yet, but the paper was already summarizing the findings of its year-long effort.

That said, it's as we've often told you: increasingly, our nation's successor to journalism tends to be narrative all the way down. That may help explain a problem which appeared at the start of the Post's report.

As they started, Kindy and Fisher reported the Post's basic findings. But just like that, in paragraph 5, they penned a statistical blooper, which we highlight in the text shown below.

Truth to tell, Kindy and Fisher had made at least one peculiar journalistic choice before they committed their blooper. This is the way the Post reporters began their front-page report:
KINDY AND FISHER (12/28/15): Nearly a thousand times this year, an American police officer has shot and killed a civilian.

When the people hired to protect their communities end up killing someone, they can be called heroes or criminals—a judgment that has never come more quickly or searingly than in this era of viral video, body cameras and dash cams. A single bullet fired at the adrenaline-charged apex of a chase can end a life, wreck a career, spark a riot, spike racial tensions and alter the politics of the nation.

In a year-long study, The Washington Post found that the kind of incidents that have ignited protests in many U.S. communities—most often, white police officers killing unarmed black men—represent less than 4 percent of fatal police shootings. Meanwhile, The Post found that the great majority of people who died at the hands of the police fit at least one of three categories: they were wielding weapons, they were suicidal or mentally troubled, or they ran when officers told them to halt.

The Post sought to compile a record of every fatal police shooting in the nation in 2015, something no government agency had done. The project began after a police officer shot and killed Michael Brown in Ferguson, Mo., in August 2014, provoking several nights of fiery riots, weeks of protests and a national reckoning with the nexus of race, crime and police use of force.

Race remains the most volatile flash point in any accounting of police shootings. Although black men make up only 6 percent of the U.S. population, they account for 40 percent of the unarmed men shot to death by police this year, The Post's database shows...
That didn't take long! In the passage we've highlighted, the Post commits a basic statistical blooper in just the fifth paragraph of a major front-page report.

Below, we'll take a look at that blooper. First, though, let's consider Kindy and Fisher's basic framework as they start their report.

The reporters start by noting an incontrovertible fact—the vast majority of fatal police shootings last year did not involve a white police officer shooting an unarmed black male.

It's somewhat odd to start a report by noting that underwhelming fact. But as the two reporters note, such cases have dominated public attention in recent years. It's easy to develop a mistaken idea about the overall shape of the data.

As the reporters continued, they offered a formulation which struck us as odd. The formulation appears in this quote:

"The Post found that the great majority of people who died at the hands of the police fit at least one of three categories: they were wielding weapons, they were suicidal or mentally troubled, or they ran when officers told them to halt."

"They ran when officers told them to halt?" It's odd to see people who ran from the police—or people who were mentally troubled—linked in this particular way to people who were actually "wielding weapons" when they were shot by police.

Presumably, Kindy and Fisher don't mean to suggest that running from the police would typically justify gunfire on the part of police. (Presumably, few police offers would make such a claim.) Still, their formulation almost seemed to make that suggestion.

To our ear, that was an odd formulation. Two paragraphs later, it was followed by a flat-out statistical blooper, which we highlighted in the text we posted above.

For the record, each statistic in the passage we've highlighted seems to be perfectly accurate. Black men do constitute (almost exactly) 6 percent of the total U.S. population. It's also true that black men "accounted for 40 percent of the unarmed men shot to death by police" in 2015, according to the Post's data.

(According to the Post's research, 88 unarmed men were shot and killed by police in 2015, including 37 black men.)

Each statistical claim by the Post is accurate. But the comparison the Post created is bungled. Here's why:

In its first statistic, the Post compares the black male population to the total national population, male and female. In its second statistic, the Post presents the number of unarmed black men shot and killed as a percentage of the male population under review.

This change in framework has the effect of swelling the size of the apparent disparity being noted. It would have made more sense to present the disparity in this way:

Although black men make up only 12 percent of the U.S. male population, they account for 40 percent of the unarmed men shot to death by police this year, The Post's database shows.

In that new formulation, a large disparity still exists, and needs to be explained. But it's half as large as the disparity the Post seemed to be reporting.

We humans make statistical errors all the time. As you may have noted over the years, this is especially true of those of us who are major American journalists.

Analytical errors aren't sins. But concerning the important topic on which the Post has gathered so much information, statistical and analytical errors have tended to run in just one direction in recent years, at least within the mainstream press and the emerging liberal/progressive journalistic world.

These errors have been quite frequent. In this case, the error appeared in paragraph 5 of a very important front-page report.

This particular statistical blooper came from the Washington Post, a part of the mainstream press corps. It didn't come from one of the organs of the emerging liberal/progressive press corps—from the new Salon, let's say, or from one of the liberal/progressive hosts on MSNBC.

That said, the Post's statistical blooper appeared quite early in a high-profile report. From there, it jumped to MSNBC, where Joy Reid was guest hosting Hardball last week.

Unmistakably, Reid is quite "smart;" we'd also say she's less ideological than quite a few cable hosts. For those reasons, we were surprised when she repeated the Post's statistical blooper, bungling it a bit further.

We were disappointed to see Reid do that as we watched Hardball last Wednesday night. Here's MSNBC's official account of what she said:
REID (12/30/15): If you look at the statistics in the year that is just about to end, police officers have killed nearly a thousand citizens. African-Americans make up just 6 percent of the U.S. population. They account for 40 percent of the unarmed men shot to death by police this year.
As we watched the show in real time, we didn't notice that Reid had misstated the first of the two statistics. Given the fact that Reid is smarter, and more responsible, than the average cable news bear, we were surprised that she had run with the bungled comparison at all. That said, there was a lot of this sort of journalistic bungling during the past year.

Over the course of next several weeks, we'll be offering a review of the past journalistic year. On a journalistic basis, we'll be dubbing 2015 "The Year of the Liberal."

We don't mean that as a compliment.

In the realm of journalism, was 2015 really The Year of the Liberal? In our view, 2015 was the year in which an unfortunate fact became unmistakably clear—the emerging liberal/progressive journalistic world has largely adopted the journalistic culture of talk show pseudo-conservatism.

In the past year, it became unmistakably clear. The values invented by Rush and Sean have largely become the values and norms of the emerging liberal/progressive journalistic world.

In what ways has the liberal world adopted that noxious culture? At our liberal sites and orgs, we invent false claims and bogus statistics. We invent, embellish and disappear basic facts.

Our errors, which are plentiful, tend to run in preferred directions. We have our privileged tribal scripts. Relentlessly, we bend the facts in those preferred directions.

In this particular case, the statistical blooper came from the Washington Post, an organ of the mainstream press. But when it comes to race and gender, mainstream orgs like the Washington Post and the New York Times tend to share the liberal/progressive view of the world. This only heightens the liberal tendency to invent, disappear and rearrange facts in service to preferred narratives, a tendency which is now a convention within our emerging world.

The Post made a relatively minor statistical error in that report—but it made the error remarkably early, in paragraph 5 of a major report which must have been thoroughly edited. We don't mean to suggest that the Post actually knew it had made a mistake. But within the emerging liberal/progressive world, thumbs are now constantly placed on the scale in this familiar fashion.

To tell the truth, we liberals aren't super-sharp. But our rather frequent errors all tend to run the same way.

In our view, 2015 was the year when it became clear that the liberal/progressive world is playing by The Rush and Sean Rules. Our lizard brains will rush to say that this simply isn't the case. But that's what the other team's lizard brains have always said about criticisms of Rush and Sean, and now about criticisms of Candidate Trump.

Over the course of the next several weeks, we'll run through some of the basic factors which made last year The Year of the Liberal. In the past year, it became unmistakably clear:

By and large, the nation's two tribes are now playing by one set of rules.

This afternoon: Also found last week at the Washington Post

Still coming: Who are Trump voters?

36 comments:

  1. Should all races and sexes be shot at the same rate? That's the unstated assumption, when one points out that black males were disproportionately killed by police. As a data point, it's fine to point out this disparity. But, for reaching conclusions, one needs to look at the behavior of the different groups.

    E.g., black men commit murder and other violent crimes at a far higher rate than Asian women. That difference in behavior surely accounts for at least some of the difference in the rate at which each group is shot by policemen.

    ReplyDelete
    Replies
    1. As an actuary with some statistics training, you should understand that one measure of whether bias (in the sense of a systematic influence on data) exists or not is a comparison against chance. It doesn't mean there is an assumption about what should or does occur in reality.

      Do you think black men commit suicide, for example, at a higher rate than Asian women? They don't. Isn't suicide a violent crime? You need to be much more careful with your generalizations.

      How, for example, would you compare suicide rates for two groups when men are statistically more likely to commit suicde-by-cop and women are more likely to take pills? Surely the difference in the rate at which each group is shot by policemen will differ even when the motivation does not. As a data point, does the method of committing suicide matter more than the intent?

      You don't seem to look beyond the numbers any further than is required to confirm your own assumptions.

      Delete
    2. "You don't seem to look beyond the numbers any further than is required to confirm your own assumptions."

      Our David has never been more pithily described.

      Delete
    3. "You need to be much more careful with your generalizations."

      That's like suggesting vegetarianism to Jeffrey Dahmer.

      Delete
  2. Statistic Bob overlooked while focusing on one "blooper" he concedes was "accurate."

    Although black women make up only 12 percent of the U.S. female population, they account for 66.66667% percent of the unarmed women shot to death by police this year, The Post's database shows.

    ReplyDelete
    Replies
    1. African Americans make up 12% of the US population, not 6%. Putting two different statistical comparisons into the same sentence results in confusion. It doesn't matter whether you are talking about women or men. Doing that is bad writing.

      Delete
    2. Overlooking problems facing women doesn't matter whether it involves pay disparities, police shootings, or rape.

      Delete
    3. "The Post made a relatively minor statistical error in that report—but it made the error remarkably early, in paragraph 5 of a major report which must have been thoroughly edited. We don't mean to suggest that the Post actually knew it had made a mistake."

      So sez Sr. Somerby using the word "error" twice and "mistake" once to describe something he earlier conceded was "accurate."

      Delete
    4. It wasn't a minor error. If a reporter on a major paper doesn't know what percentage of Americans are black, something is majorly wrong.

      Women commit a tiny proportion of violent crimes. What problem involving them do you imagine is being overlooked in this report?

      You are only here to criticize Somerby. Why don't you take your troll ways to some other blog?

      Delete
    5. Anon @ 10:45: You speak of "bad writing." You are guilty of bad thinking. Yes, blacks make up 12% of the population, however, the 6% figure, as stated by Bob indicates the male %, and the female %, thereby adding up to 12%.

      Delete
    6. @ 11:00 AM says:

      "It wasn't a minor error. If a reporter on a major paper doesn't know what percentage of Americans are black, something is majorly wrong.
      ----

      You are only here to criticize Somerby"

      Idiot. The person who said it was a "minor" error WAS Somerby. After he said it was accurate.

      Delete
    7. Here is the statement from the Washington Post:

      "African-Americans make up just 6 percent of the U.S. population. They account for 40 percent of the unarmed men shot to death by police this year."

      They state that the % of African Americans in the US population is 6% not 12%. That is incorrect. The word males is omitted. That word is needed to make their sentence accurate. There are not exactly equal numbers of males and females in any demographic group, so that 6% is rounded off.

      According to the CDC:

      "In 2013, the population of African Americans including those of more than one race was estimated at 45 million, making up 15.2% of the total U.S. population.

      Those who identified only as African American made up 13.2% of the U.S. population-over 41.7 million people."

      Delete
    8. Well, you managed to avoid the fact you trolled Bob while throwing the T bomb at others.

      And, that was not the statement from the Washington Post.

      Idiot.

      Delete
    9. If you are comparing men only, then black men make up 12% of the male population. You cannot switch from comparing all people (6%) to just men (40%) in the same sentence without being misleading. Reid's sentence shows what happens when you do that.

      Your only interest here is to call people names.

      Delete
  3. I thought the Post did a bigger service reporting on the speaking fees of the Clinton clan.

    ReplyDelete
    Replies
    1. They neglected to mention that Ben Carson is still giving speeches for money despite being in the midst of a campaign. What's up with that?

      Delete
  4. The statistics collectec by the Washington Post don't indicate whether those men running were running away from or toward police. Further, they neglect the duty of police officers to protect bystanders and innocent civilians from violent actions by someone who is armed or behaving dangerously. When someone runs away from police, ignoring police commands, the officer has a duty to protect others from that person's actions, already demonstrated to be non-compliant. Will the runner take a hostage, hijack a car, hurt others? The officer cannot know but must nevertheless take action. It is always assumed that the shooting was to prevent the runner from getting away. It is more likely to prevent the runner from endangering other people by stopping him and taking him into custody. People do not have the right to run away from police and it is not an innocent or victimless action. It endangers public safety. That is why police sometimes shoot in that situation. It is their judgment call whether the person running poses a danger to others or not. If there is the possibility, they shoot, even when the runner is going away from the cop.

    ReplyDelete
  5. "Over the course of the next several weeks, we'll run through some of the basic factors which made last year The Year of the Liberal."

    We presume Mr. Somerby thinks this will be more beneficial than asking his readers to donate money to help him finish his book.

    ReplyDelete
    Replies
    1. At one time, this sort of thing mainly came from the right. Today, our side is doing it too! Increasingly, our own lazy, uncaring “intellectual leaders” are part of the problem too.

      Delete
    2. When Bob started this site in 1998, there were virtually no liberal organs.

      Delete
  6. Has it really been thirty-two years since Bob did the Richmond Comedy Club with Bill Maher?

    ReplyDelete
    Replies
    1. Endless forms of entertainment, distraction and piffle.

      Delete
  7. The Washington Post made a comparison which is analytically quite flawed and that at a glance suggests the odds of being shot by cops are more than 6 times as high for black men than whites.

    That gets noted. The person noting it also tells you that the real comparison is still horrifying: it's more than 3 times as likely.

    The person also notes than some other media are running with the analytically flawed comparison -- even sometimes botching it further.

    The Big Problem for many commenters here: the person who pointed out the statistical problem.

    ReplyDelete
    Replies
    1. The Big Problem for Nona Nym:

      "The Washington Post made a comparison which is analytically quite flawed and that at a glance suggests the odds of being shot by cops are more than 6 times as high for black men than whites.

      That gets noted. The person noting it also tells you that the real comparison is still horrifying: it's more than 3 times as likely."

      Actually the number of unarmed black men getting killed by police is 38.5% of the total of unarmed people getting killed. So it is 6 times more likely using the figure Nona, the Post, and Bob all use. Which makes the real misleaders Bob Somerby and friend Nona Nym.

      Delete
    2. 38.5% divided by 12-13% is not 6 times more likely. It is 3 times more likely than their representation in the population. The 12% is the correct figure because black men should be compared to white men, not to all people (male and female).

      Delete
    3. Black males and females make up about 13 percent of the population. Blake males, about 6.5 percent.

      Divide 38.5 by 6.5. Pretty close to 6 times.

      Delete
    4. 5:52 you are repeating a blunder. the 38.5 percentage is of men, blacks make up 12 percent of men

      Delete
  8. Hi Bob, a fellow hoops player from the mid-60's Mid Peninsula League here, and a long-time reader of your very interesting blog. With regard to your analysis of the "tribal style" of infotainment on MSNBC, Fox, et al., I wonder if you would find this editorial comment in today's New York Times relevant. Written by a think tanker, but I believe he's on to something. He attempts to address the conundrum that while Congress has an approval rating of about 13% nationally, incumbents nevertheless are returned to office at about the 94% level. Further, most incumbents have found they can disspense with "constituent services." They're a waste of time, as far as reelection is concerned. All that matters is a check list of positions on hot button issues that define the "ideology" of the party - taking these positions assures the incumbent of the votes of those in his/her district or state. There are fewer and fewer "cross-over" voters who stray from their party (or its ideology) in a general election. This is true even for the so-called "Trump Democrats," the blue collar males who make up a lot of his base. They always vote along conservative lines which follow the Republican ideology. Liberals and conservatives have become, in this sense, increasingly "reactionary." Name the issue, you know the position. The writer makes the point that in this sense Trump is following a trend already in motion; he's just very good at capitalizing on the increasingly polarized and simplistic method of appealing to the masses. His website, e.g., has five "positions." Nothing complicated, all very obvious. He knows you don't win elections these days by going into great detail about some "program" you're considering. Just toss the chum in the water and the voters will rise to the surface. Anyway, I found the analysis very insightful. It fills in part of the mystery of the oligarchy - how do Congress members get away with pursuing corporate interests so blatantly and still continue in the sinecure of eternal electability? Keep up the good work. http://www.nytimes.com/2016/01/04/opinion/campaign-stops/trump-did-not-break-politics.html?ribbon-ad-idx=4&rref=opinion&module=Ribbon&version=context&region=Header&action=click&contentCollection=Opinion&pgtype=article

    ReplyDelete
    Replies
    1. Bob has been throwing the same chum in his blog waters for 18 + years and his readership keeps rising to the surface. To tell the truth, they aren't super-sharp.

      Delete
    2. @1:20, if you want Somerby to be aware of this, please send him an email. He does not read these comments and he will not respond to you here. When you write such comments here, you are off topic and seem to be trying to hijack today's topic.

      Delete
    3. I guess you're the Thread Police. Most blogs have a few such officers. But Bob actually did introduce the topic I wrote about in my comment, so it's off the topic of police brutality and whether there's a demographic bias, but not off this: "In our view, 2015 was the year when it became clear that the liberal/progressive world is playing by The Rush and Sean Rules. Our lizard brains will rush to say that this simply isn't the case. But that's what the other team's lizard brains have always said about criticisms of Rush and Sean, and now about criticisms of Candidate Trump." The editorial from the Times plays into that.

      Delete
    4. I thought I was doing you a favor by telling you your communication was misdirected. Enjoy talking to yourself.

      Delete
    5. Wow, someone's got their tighty-whitey's all the way up there (anon at 4:06).

      He addressed a topic in Bob's post. Relax. If you have no interest in Harry's observations, take your own implied advice and move on.

      Delete