The popular vote in the popular imagination


Every presidential election year brings new focus on the idea of abolishing the Electoral College, especially 2000 and 2016 when the popular vote winner lost the race. If only the truly democratic popular vote of the country had been followed, the idea goes, our candidate would have won!

Applied retroactively, this critique is a mistake, ignoring how the rules make the game. If the participants were operating under different rules, their strategies would have been different. Republicans would campaign more in California, and Democrats would put greater effort into Texas, for example, while both would spend less time in Florida. Moving from 48% to 50%+1 in Florida would be less valuable than taking 48% and campaigning elsewhere.


The general assumption seems to be that a popular vote rule would favor the Democratic Party, but I don’t think that’s a guarantee. It’s based on two elections run under different rules. It assumes strategy doesn’t change when rules change—and not only for candidates. The popular vote rule would also change the calculus for voters, especially for hidden minority party members whose preferred candidate has so little chance of winning that they don’t bother to vote. Allow me to take the long way to the point. As reliably Republican as Utah is, the Democratic presidential candidate still tends to get about a quarter of the popular vote there in recent years. In two solidly Democratic states with larger, more culturally and ethnically diverse populations, California and New York, the Republican presidential candidate gets a few points over a third of the popular vote. That so many of them cast votes for candidates who are certain to lose the state hints at more voters with the same preferences but less motivation.

Not only that, but because state legislatures control Congressional district boundaries, down-ticket races tend to be more lopsided, and thus even more discouraging to minority party voters. The Cook Partisan Voting Index has New York’s 15th District, which covers Yankee Stadium and the Bronx Zoo, as the most partisan. The 15th District’s results were more assured ahead of time than anywhere, yet 4.9% of the voters there in 2016 still took the time to pull the lever for the other presidential candidate. I’m not suggesting there are a great many low-flying Republican voters there, but under the direct popular vote such as exist would have more incentive to vote.

The obvious counterpoint is the voters who would vote but realize their preferred candidates are going to win and don’t spend time running up the score. Under a direct popular vote, the incentive changes for them, too. What we don’t know is which of these groups, the discouraged local minority party voters or the assured local majority party voters, is larger. Keep in mind many people don’t vote now and still won’t vote if the rule changes. So if the popular vote rule were implemented, are we sure about who would gain more votes?


In a country as large and heterogenous as the United States, there are good reasons to think the Electoral College helps reduce social conflict. It’s also embedded in the Constitution, so removing it requires a lot of political capital, a heavy strategic concern that requires other agenda items to give way.

Perhaps it would be a more fair, democratic system. It would also be fairly unique. To use an example, the German political system is generally regarded as democratic enough, and as producing tolerably good results, and it is not even close to the simple popular vote rule. The German chancellor is elected by the Bundestag, roughly the equivalent of the US House of Representatives. Direct election of a head of government by pure popular vote is less common than you’d think. Political parties, government bodies, and other institutions generally mediate between votes and outcomes, and the popular vote rule would have to overcome their resistance as well.

None of this is to say it’s not worth doing. That’s a subjective judgment I can’t make for you. But I see the impetus for the direct popular vote as either trying to relitigate past elections, or as a better way to achieve the proponent’s goals. The latter might be a mistake. The former certainly is.

There are more podcasts in heaven and earth, Horatio

The New York Times recently ran a devastatingly bad article about podcasts. The meta reading, as is so often the case, is that the article is about the New York Times itself and its view of culture, a view it shares with other elite media outlets.

The article profiles an abortive and unintentionally funny attempt at a podcast by two freelance writers who expected their mediocre efforts would be met with instant success. They were not, of course, but I suppose they were networked with the author of this piece and somehow consented to be interviewed. I don’t fault them for rolling the dice. Hell, failing and then being profiled in the NYT probably raises their profile, so it seems like a win-win.

But this is not just a story about a failed podcast. It’s a story about a failed podcast in the New York Times. The article focuses on only the largest podcasts (by audience), consigning everybody else to forgettability. It’s true that most podcasts don’t strike gold, and even more “fail” in the sense that they don’t catch on and eventually get abandoned by their producers. But this is true in many walks of life. Most new businesses fail. Most new products flop. Most new books are ignored by almost everybody. Would there be any point to writing an article about Peak Book?

Why am I giving the New York Times grief? Well, it’s only a small part of the podcasting world that is produced by large outlets you can see from 8th Avenue. There are many great podcasts out there with decent-sized audiences, a comfortable enough return for their producers, and probably no hope of ever getting on the radar of the elite media. I listen to as many as I have time for. (Sadly, there’s never enough time, especially during football season.) For example, The History of Byzantium is an excellent history podcast that is essentially one man’s passion project, with donations and some ads making it worth his time to keep it going. I couldn’t count the number of episodes of The Bible Geek podcast. puts out a regular podcast I’m sure will never make the Times. These all have devoted listener bases but just aren’t big enough or official enough to attract notice in an elite media take on their field. A reader with no exposure to podcasts would have no idea of the richness of the field from reading the article, and in fact would get the opposite impression. NYT’s blinders with respect to organic, bottom-up cultural phenomena simply don’t allow them to tell the right story.

I’d love to continue and list more worthwhile podcasts, but I have to keep reading the multi-part Washington Post series about why summer is bad.

Worden on religion

On pp. 77–78 of his short book The English Civil Wars: 1640–1660, Blair Worden makes a very succinct statement about the social value of religion:

Religious tolerance went against the grain of seventeenth-century thinking. In a society without a police force or, ordinarily, a standing army, the preservation of order will seem dependent on the coherence, even the uniformity, of ideas and beliefs. The coexistence of faiths within a nation’s frontiers was generally assumed to be unattainable and undesirable.

Add to this how the church filled so much of the space now filled by the state—birth, marriage, and death records, for example, or relief for the destitute—and the more advanced economy—a very large part of the disposition of labor, property, and tax money—and disputes about the beliefs and organization of the church take on much greater importance than they possibly could in the 21st century.

Science vs. expertise

In the a recent episode of EconTalk, where Russ Roberts interviews the great Bill James, the father of sabermetrics, they discuss the distinction between science and expertise as James delineates it in a great essay called “The Note”. It begins:

Handwriting expertise, more clearly than any other field that I know of, illustrates the difference between science and expertise. Science uses general rules and principles, understood by millions of people, to work toward increasing our shared understanding of the world in which we live, a key word being “shared”. What is learned by the scientist is in no sense the property of the scientist, if it is truly science and not commerce. If in our field we were to discover, for example, that tall hitters do best against tall pitchers and short hitters do best against short pitchers, the value in this would not be to me, but to a baseball team by way of its manager. The manager would be as much the owner of this knowledge as the analyst who discovered it. Crucial to that fact—to the shared nature of its ownership—is how it is known. If I were to say that I know that tall hitters hit best against tall pitchers and you should believe me because I am an expert in this field, no one would believe me and no one should believe me. Others would believe me only if (1) I were to present evidence demonstrating that it is true, and then only if (2) others were able to study the same subject and reach the same conclusion. Studying the same subject and reaching the same conclusion or a different conclusion does not require expertise limited to a small cadre of persons; it merely requires that we apply the rules of scientific enquiry which are universally owned and widely understood. Shared principles yield shared knowledge.

Expertise, on the other hand, is the property of the expert. The only way that WE know that Document 1 and Document 2 were or were not written by the same person is that an expert says so. The only way to get a second opinion is to ask another expert. If the other expert cannot get access to the documents, you can’t get another opinion.

James is not the first to identify this division, but he puts it very well.

How should we deal with newcomers?

The Atlantic had an interesting piece recently about how gentrifying neighborhoods attract more police resources, due to calls from the newer residents, and how this disproportionately affects racial/ethnic minorities. I think it reads more productively as a criticism of policing rather than of gentrification per se, though the author doesn’t seem to, but more interestingly it evokes a widespread social issue. To begin my explanation, let me highlight these passages:

When low-income neighborhoods see an influx of higher-income residents, social dynamics and expectations change. One of those expectations has to do with the perception of safety and public order, and the role of the state in providing it. The theory goes that as demographics shift, activity that was previously considered normal becomes suspicious, and newcomers—many of whom are white—are more inclined to get law enforcement involved. Loitering, people hanging out in the street, and noise violations often get reported, especially in racially diverse neighborhoods.

In 2013, the city of San Francisco launched Open311, a mobile app that allows residents to easily report public disorder like loitering, dirty sidewalks, or vandalism by snapping a photo and sending their location. The app can feel altruistic; residents, for example, are able to report the whereabouts of homeless people who seem to be in need of assistance. But some worry that the dispatches can result in unnecessary citations or harassment. And while broken-windows policing remains controversial, a 2015 poll suggested that it’s still largely accepted by the general public, so when people see something, they’re likely to say something. After the app launched, 311 calls increased throughout the city, and one study showed that gentrifying neighborhoods saw a disproportionate spike.

Butler, who recently wrote the book Chokehold: Policing Black Men, believes that this is a result of newcomers refusing to assimilate to longstanding neighborhood norms. [emphasis added] “Culturally, I think the way that a lot of African American and Latino people experience gentrification is as a form of colonization,” he said. “The gentrifiers are not wanting to share—they’re wanting to take over.” One of the tools they can use to take over public spaces, he argues, is law enforcement.

This is interesting, and it reminds me of the neighborhood where I lived in Alexandria, Virginia, only in reverse. There was a lot of very recent immigration, largely Salvadoran, as there has been all over the DC metro area. Granted, I was a newcomer to the neighborhood, but of similar SES and background to the longer-term residents. Being from the US-Mexico border, this immigration doesn’t bother me, but I noticed it bothered a lot of other people because the Salvadorans, to use the language of the article, ‘refused to assimilate to longstanding neighborhood norms’ regarding noise, cleanliness of public spaces, junk cars parked on the street, etc. Indeed, this is a common occurrence all across the country, and for the last decade or more has been a simmering issue. This cultural friction is one of the reasons for the surprising result of the 2016 presidential election. The Atlantic’s political wing tends to regard people who are upset by immigration as, well, deplorable. No doubt some are motivated by racism, which is indeed deplorable, but cultural frictions exist that aren’t motivated by racism. (For the record, in my ideal world immigration rules will be greatly liberalized relative to current rules.)

So which is it? Are newcomers wrong for “refusing” to assimilate to longstanding area norms, or are older residents wrong for not accepting the changes? There’s no simple answer, and I don’t mean to suggest the writer of the article or anybody else thinks there is. I don’t think we’ve had much public dialogue on the issue, and I’d like to, especially as so much governance happens outside of formal systems, and better public understanding of good rules and norms would help us resolve more social issues with nuance and wisdom before they became political issues resolved with neither.

JFK Assassination – Oak Cliff locations map

I’ve seen maps of the area between Lee Oswald’s boarding house and the site of Officer J.D. Tippit’s murder trying to figure out how the timeline works, but there is more in Oak Cliff that’s interesting to the puzzle. This map is not by any means exhaustive, but here are six locations:

The first odd thing to notice is the two Oswald residences, one where Marina Oswald lived with the children, and which Lee Oswald visited on weekends, and the boarding house where Lee lived during the week. The Oswald family had resided in the house earlier but had moved to Irving by the time Lee was at the boarding house during the work week, visiting Irving on weekends. This is odd. It’s said this was due to the Oswalds’ marital problems, but it would have been very expensive for a minimum-wage worker whose wife did not work; Oswald’s $1.25/hr minimum wage job at the Texas School Book Depository in November 1963 would be worth $10.01/hr in November 2017. Where Oswald came up with money is a question that comes up frequently when looking at his life.

The second odd thing to notice is where Jack Ruby’s apartment is in relation the Tippit shooting site and Oswald’s residence. I’m not the first to notice this, of course. It might be pure coincidence, as it’s entirely plausible Oswald was Tippit’s killer, was in panic mode, and was fleeing with little regard to direction, or to a bus stop not so close to his residences. The housekeeper said she saw him waiting for the bus at Zang and Beckley, at the end of the block, which he obviously did not take. There has also been a lot of confusion about which direction Tippit’s killer was walking when Tippit stopped him. I don’t know enough to have a firm opinion on whether the killer was walking east or west on 10th Street, or if the killer and Oswald were the same person. But it’s strange either way. Whoever the killer was, he was oddly close to Jack Ruby’s apartment when the murder occurred.

I’ve also put the site of Officer Tippit’s mysterious phone call on the map. The location so near to the Texas Theatre, where Oswald was arrested, is not really noteworthy; that’s the main drag in the area, so plenty of places people would stop in on were in those few blocks. But the phone call is strange. At this point the president has already been assassinated, and every law enforcement officer in the Metroplex is on high alert. Rather than communicating by his police radio, Tippit tries (and apparently fails) to get in touch with somebody by phone only a few minutes before he is murdered. We’ll probably never know what was so urgent and unrelated to his official task at the moment that he had to make this phone call about.

Random coincidences happen. Nothing I’ve written here may really mean anything. But I like maps, and these locations are more interesting on a map than as textual abstractions.

How to blind yourself and others to good research

Via Twitter, this gem about ideologically-driven research:

I want to highlight two parts. First, the opening act:

The criteria of valid knowledge associated with quantitative research are ones that turn women, whey they are the focus of research, into objects. This means that women are again subjected to exploitation, in that knowledge and experience are extracted from them with nothing in return, even when the research is conducted by women (Mies 1993).

Because of the many things people have in common with other people, we can find out some kinds of information about people generally by doing research involving specific people. And because people are part of the natural world, many things about them can be described with numbers. Those numbers cannot capture everything interesting, important, or even relevant about human beings, but like carbon dioxide parts per million, sometimes putting numbers to the various features of the natural world can tell us interesting and useful things. If you want to call this treating women (and people generally) as objects, fine, but you thereby imply a complete separation between humans and the rest of the natural world. Humans are special, but it does not make sense as a blanket rule in social and physical science.

If you think of quantitative data gathering as particular women being treated as objects and thereby out of bounds, what is your alternative proposal for medical research? There are certain kinds of research that are unethical, of course, but beyond those cases the overwhelming majority of medical research works this way, and women (as well as men) have certainly benefited from medical research. Does this passage imply compensated participation in research is acceptable, i.e. not exploitative? Or at least informed consent? These are standard practices. What if research participants recognize the potential benefit to people other than themselves and participate for altruistic motives, even if they don’t know the specific purpose of a particular study and aren’t compensated? Are we saying only compensated participation is acceptable?

But this is all a warm-up for the really astounding main event:

The emphasis on controlling variables further exacerbates this last problem, and indeed the very idea of control is viewed as a masculine approach.

This is really egregious. Is the objection to trying to determine causal and probabilistic relationships among variables by the technique of holding some other factors constant? That seems absurd to write out, something nobody could posit even in jest. The only explanation that makes sense is that the mysterious, unnamed academics who with the passive voice view “the very idea of control” “as a masculine approach” simply have no understanding of the processes under discussion, and think because some forms of control are objectionable the word itself must entail them in all other contexts. (Perhaps some self-controldiscipline in studying statistical methods is in order.) If this is the case, we are dealing with an entire academic subfield premised on a complete misunderstanding of what words and numbers are and do. This is a group within the academy that literally and completely has no idea what the others are talking about, but commands pontificating about it anyway. Not only is using any resources for this purpose wasteful, it actually takes us farther away from reality.

Making sure women get a fair shake is a noble (and uncontroversial) goal. Tying the improvement of the female condition to dubious, even mystical research practices does not help the goal. Fortunately, almost all scholars, including the scholarly women I know, don’t follow this approach.

As directed by the JFK Records Act of 1992, many more files relevant to the investigation of Kennedy’s assassination were released today, following a batch in July, although the current president reversed course from his earlier statement and held some back. I don’t expect too many shocking revelations to come out of this latest batch, but you never know.

For those who are interested, .pdf versions are available here.

Henry George on trade:

Free trade consists simply in letting people buy and sell as they want to buy and sell. It is protection that requires force, for it consists in preventing people from doing what they want to do. Protective tariffs are as much applications of force as are blockading squadrons, and their object is the same—to prevent trade. The difference between the two is that blockading squadrons are a means whereby nations seek to prevent their enemies from trading; protective tariffs are a means whereby nations attempt to prevent their own people from trading. What protection teaches us, is to do to ourselves in time of peace what enemies seek to do to us in time of war.

Sadly, this insight keeps being relevant.

A short deductive inference

P1. Bumper stickers for political campaigns that have ended are evidence for the public choice idea that political participation is expressive rather than instrumental. What could be more about you and less about concrete objectives than a call to action for an event that’s already happened?

P2. Bernie Sanders bumper stickers survive all out of proportion to how much support he had.

C1. Supporting Sanders is expressive at an above-average level.
C2. Sanders supporters are lazy and haven’t gotten around to taking off the stickers. (heh)