pageok
pageok
pageok
Demographics of VC Readers:

The Quantcast website has an interesting profile of the demographics of VC readers here. Hat tip: Tom Smith.

I. Findings that don't surprise me.

VC readers are disproportionately male, relatively to the internet average.

VC readers are disproportionately highly educated (people with graduate degrees).

II. Findings that do surprise me.

The relatively high percentage of Hispanic readers (25-30% more than the internet average).

The comparatively low percentage of white and Asian readers (below average in both cases).

The age profile of readers is older than I would have expected (age 45-54 is the best represented demographic).

III. Data that I wish Quantcast had, but doesn't.

It would be very interesting to know the ideological breakdown of the readership. Social science research shows that most people prefer to read commentary from viewpoints they agree with, and prefer to avoid that which comes from other parts of the political spectrum. That would predict that we have a mostly libertarian and fusionist/conservative readership. On the other hand, at least judging by the comments, the VC may have a larger proportion of liberal and leftist readers than most libertarian and conservative websites do. However, judging the overall readership by the commenters is a tricky business because commenters may be unrepresentative of the readership as a whole for a wide variety of reasons. For example, people who disagree with a post may be more likely to comment on it than those who agree.

It would be helpful to have the absolute numbers for many of Quantcast's categories, rather than just percentages relative to the "average" website.

Realist Liberal:

It would be very interesting to know the ideological breakdown of the readership. Social science research shows that most people prefer to read commentary from viewpoints they agree with, and prefer to avoid that which comes from other parts of the political spectrum. That would predict that we have mostly a libertarian and fusionist/conservative readership. On the other hand, at least judging by the comments, the VC may have a larger proportion of liberal and leftist readers than most libertarian and conservative websites do.


As a moderate liberal I disagree with some of the conspirators' politics regarding role of government (but largely agree with the social side). I don't read many conservative blogs but I do read VC because I find that the quality of the postings and comments is exceptionally high, even in the times I disagree with the point of the post. So if I'm any example (and who knows if that is the case), VC may have a higher than expected readership among liberals because of the quality of the blog. Ironically, I read this blog more than any liberal blog that I can think of.
12.10.2006 6:11pm
Elliot Reed:
Do you have any idea how they get these data?
12.10.2006 6:18pm
tab (mail):
"How do you collect your data?

We collect our data through affiliations with partners, who include advertisers, publishers, ISPs and advertising networks. All the data we collect is anonymous and contains no personally-identifiable information.

Anyone can participate. Web publishers can obtain refined analysis on their site by utilizing a Quantcast measurement pixel. If you want to help make Quantcast's analysis even more accurate then please become a Quantified Publisher."

http://www.quantcast.com/
12.10.2006 6:29pm
Ty:
Mr. Reed: Thank you from the bottom of my heart for your appropriate use of 'data' as plural. That's my pet peeve.

I would be interested to know the source of the data. This seems very susceptible to selection bias.

-Ty
12.10.2006 6:30pm
AnonPerson (mail):
I actually tend to enjoy reading well-written, thoughtful, conservative blogs, because I find it interesting to read well-argued viewpoints that I may disagree with. They are new and interesting. Things that I already know about and agree with are old and boring.

As an academic, this is not at all surprising to me. Who wants to read old, well-established work, repeatedly? For me, that's what reading blogs that exactly fit my politics is like. Boring.
12.10.2006 7:11pm
RainerK:
Interested in my "ideology"? No problem: Solidly libertarian in principle. Motto: Get Govt off my back and activists too! Sometimes depends on the issue, but always in favour of liberty and true justice (that is, not the kind that just follows the laws).

Ladies and gentlemen, post it here, let the VC know where you stand!
12.10.2006 8:10pm
John (mail):
I'm not so sure you can't tell a lot from the comments. To be sure, a dissenter does have an incentive to weigh in; but that usually prompts a wide ranging discussion as those who dissent from the dissenter weigh in themselves.

On balance, I do think the comments show a libertarian preponderance. Which is why I like the VC (confirming your other point...).
12.10.2006 9:36pm
David Schraub (mail) (www):
Speaking as a liberal VC loyalist, it wouldn't surprise me if you had more liberal readers than average for a right-leaning site. Y'all write well, write smart, are respected, are respectful, and (perhaps most importantly), aren't psychotic. All that makes this site a must-read even for liberals who are interested in law.
12.10.2006 10:39pm
liberty (mail) (www):
Why not conduct your own poll, then? Easy enough to gather the data and likely as accurate as any that someone else conducts. Your poll would only include data on those who respond, but if they use affiliates, ISPs and advertisers that will skew data as well.
12.10.2006 10:50pm
DG:
The ethnicity data in quantcast is very shaky, and I wouldn't rely on it - this is a beta, after all. I've run a bunch of sites I'm very familiar with through the tool, and from what I can tell, its very accurate for "common audience", shaky on most of the other information, and completely inaccurate on ethnicity. An example - digg.com, which shows greater black and latino share than asian or white. Quite unlikely, given the subject matter, which is (mostly) seriously geeky.
12.10.2006 10:58pm
Ilya Somin:
Why not conduct your own poll, then? Easy enough to gather the data and likely as accurate as any that someone else conducts. Your poll would only include data on those who respond, but if they use affiliates, ISPs and advertisers that will skew data as well.

A good question. The problem with polls limited to those who choose to particpate is selection bias. People who choose to respond are likely to be very different from those that don't. That's why most professional pollsters use different methods to ensure random selection from the target population. Quantcast tries to achieve the same result with their panel data, though I don't know enough about the methodology to be sure if it's right.
12.10.2006 11:03pm
The Real Bill (mail):
Male, M.S., Caucasian, 36, moderate libertarian (whatever the hell that means)
12.11.2006 12:22am
M-Kel (mail):
Female, 22, M.A. Candidate this spring, hopefully to be followed by law school, and though I'm always wary of labels I'd have to go with The Real Bill in classifying myself as a moderate libertarian politically, and a libertarian constitutionally.
12.11.2006 12:40am
Sarah (mail) (www):
I wonder if the data is thrown much by the people who run script blockers, ad blockers, anonymizing proxies, and so forth. I mean, I have to assume that there's hole in the data that would best be described as technically proficient, well-educated, predominately male, predominately "white"/Asian, 18-35 years old, and at least slightly libertarian-minded. I mean, if I didn't want to comment on posts, I'd be using a proxy myself (I'm female/265/bachelor's degree/"white"/libertarian.) And most of my friends think I'm insufficiently paranoid about such things. Considering the number of users who don't even see ads anymore, let alone click on them, how reliable can this (advertiser-provided) data really be?
12.11.2006 2:13am
logicnazi (mail) (www):
A lot of these usage tracking systems (e.g. alexia) use opt in data gathering. Things like whether they pay users and the like have a huge effect on the data. Furthermore if it works like the alexia toolbar, basically convincing people to run a useless tollbar that sends back all their data to the company, it will bias the user base against younger more computer competent users.

Also what do they do with decline to states, e.g., they can't figure out/you don't say your race.

Frankly I don't trust internet usage studies at all unless I find evidence that a very high percentage of the individuals asked to participate did so.
12.11.2006 4:21am
lucia (mail) (www):
Since I'm all for empiricism, I decided to check out my site/blogs Quantcast statistics. Quantcast seems to get some things right and definitely gets some things wrong.

What does it get right for my site? It says my hobby site (the knitting fiend) visitors are overwhelmingly female. Absolutely true: after all, it's a knitting blog. It also tells me they are overwhelmingly young (25-34), which matches the impression I get from comments.

What does it get wrong? I would have thought site visits would be easier to figure than demographics, but strangely, Quantcast way undercounts unique site visitors; the blog alone gets 3 time what Quantcast says I get. (I checked another popular knitting blog the panopticon"-- and Quantcast way, way, way undercounts his visitors. They say he gets 1K visitors a month; in fact his blog gets over 2K a day.

Quantcast gives the web site operator an opportunity to increase the accuracy by putting javascript on their pages. I'm not going to do it-- I bet very few bloggers will.

I guess VC visitor with a blog can check and see if the Qantcast statistics seem to make sense. That might give us an idea whether or not we should rely on them for anything.
12.11.2006 8:50am
sbw (mail) (www):
Hah! Let's see...

Libertarian? I believe in limited government but libertarians are moral relativists, which I am not.

Conservative? I believe in the rule of law, but many conservatives fall back on "traditional values" that they cannot otherwise substantiate which makes them more absolutist and authoritarian.

Fusionist? All the vices of Conservatives and Libertarians with none of the virtues.

American Liberal? I believe government should stay out of one's personal life, but many American liberals are closet leftist authoritarians who cloak their views in moral clich├ęs. They believe in state control of the economy which they manipulate using the tyranny of the majority unless, of course, they are out of power, and then they believe in an activist judiciary. Overgrown mirror neurons make them feel everyone's pain more acutely than their own -- so that they are willing to hijack other people's money, not to solve a problem, but to cover it up for a time so they don't have to look at it now.

Progressive? I favor social reform, but much reform made in the progressive name is anti-social. Their pragmatic approach is often limited to a single pass that leaves solutions worse than the original problem. Rationalization to justify their feelings supplants deductive skills. Who tells them something is more important than what is said, and style points count. Present incarnations are found in lefty bloggers who demonize all who differ and take great glee in the rhetoric they use to do so. There is no conclusion to which they will not jump, provided it is expressed with passion.

Socialist? Historically blind, they represent the triumph of hope over experience.

Classical Liberal? They stress the importance of rationality, property rights, constitutional limitations of government, protection of civil liberties and Adam Smith's approach to the economy. In other words, libertarian with moral relativism held in check by conservatives natural law. Not a bad start, but which natural law?

So, I'm still looking...

What else should I consider?
12.11.2006 9:20am
JWR (mail):
Moderate liberal, and the first comment by "Realist Liberal" describes me almost perfectly. I, too, read this blog more than any liberal blog. It's simply the best, most intelligent, most even-handed (which is by far the most important criteria for me) legal blog out there. A must-read for lawyers (law profs, etc.) who aspire to those same qualities.
12.11.2006 9:31am
Mike BUSL07 (mail) (www):
Male, 27, (JD '07), Hawkish zionist libertarian. I know this is very corny, but I would like to see a poll on this site sort of along the lines of "which VC blogger do you most identify with?" That, I think, would also be quite revealing, (and entertaining).
12.11.2006 9:42am
Hoosier:
Wait! I'm ignorant. Can I still read your blog?
12.11.2006 10:55am
Bruce Hayden (mail) (www):
Of course, the sub-demographic that was missed with the education level is that a lot of the posters seem to be legally trained. No surprise, given the profession of most of the Volokh Conspriracy.

Despite what I view as slightly conservative leanings of the Conspirators (on average), the posters seem to be slightly to the left of center - my guess is that more posters here voted for Kerry and Gore than did for Bush (43). (Just my uneducated and unsupported guess based on comments here).

But yes, also fairly libertarian, which is why the fusion discussion could get legs here. Definately seem to have about as many liberal-libertarians here as conservative-libertarians. I found that discussion interesting, since I am conservative-libertarian, and didn't understand until then why there was any liberal allure to libertarians.
12.11.2006 12:05pm
James Dillon (mail):
On the other hand, at least judging by the comments, the VC may have a larger proportion of liberal and leftist readers than most libertarian and conservative websites do.

I think that's because the very high quality of the bloggers (and some, though far from all, of the commenters) here makes possible a level of conversation that can appeal to intellectually serious readers across the political spectrum. I'm rather to the left of any of your regular bloggers, but I enjoy reading, and occasionally debating, their points of view, and find myself in agreement with (or persuaded by) their arguments more often than I would have expected.
12.11.2006 12:21pm
tefta2 (mail):
How is a viewer's age, income, education, etc. gathered? This is unsettling.

BTW - It's amusing that other commenters look upon the VC as conservative. I'd say most posts and comments lean left.
12.11.2006 7:49pm
Public_Defender (mail):
I enjoy this page for utilitarian reasons. Most judges are now on the right side of the political spectrum. Arguments that would have persuaded Earl Warren or Thurgood Marshall won't get me very far, but libertarian arguments can sometimes persuade a judge to rule for my client. For example, I got a murderer out of prison at least five years earlier than she otherwise would have gotten out in large part by arguing contract theory.

Also, linking my criminal defendants with the interests of civil defendants can also help me win. For example, criminal defendants often face junk science from cops who think they're "experts" in everything they attended a one hour class about. And any experienced DUI lawyer can tell you about the junk science behind breathalyzer machines.

This site is mainly interesting because the Conspirators work to persuade, not just rally the base. They let very harsh criticism remain in the comment sections, which shows both confidence and humility (confidence that their ideas can withstand criticism, humility that their ideas might sometimes deserve criticism).

As to the statistics, I question their validity. I don't think I have ever accurately filled out one of the surveys. Do the web sites really believe that I'm a seventy year-old from Albania?
12.12.2006 6:39am
lucia (mail) (www):
Public_Defender: The web sites probably note your IP address is associated with an American service and discount your claim to be "from Albania". (Not that you couldn't be "from" Abania, but they sure know you aren't there now!)

As I noted, the demographics they gave for my site seem plausible. (Though I can't vouch for the racial make up) The demographics for VC seem more or less plausible. (Except that people have noted they aren't confident in the racial make up.)

The site visits numbers were way off for my site, and laughable for some other knitting sites I know. (And I based my claim of laughable on the other sites "site meter" statistics.)
12.12.2006 8:04am
Spartacus (www):
right-leaning libertarian, socially liberal but morally conservative (go figure); male, 38, white, married w/ 1 child (soon to be 2), BA/MA/JD still struggling aroung $50 per annum
12.12.2006 11:55am
tefta2 (mail):
Public_Defender -- I don't understand your reference to 70 year old Albanians, but I am a 72 year old second generation Albanian-American who alas has never been to Albania although a number of cousins have come over to visit us.

How has this been factored into the VC profile?
12.12.2006 3:16pm
A. Zarkov (mail):
"A good question. The problem with polls limited to those who choose to particpate is selection bias."

You mean response bias. Selection bias has to do with the sampling frame.
12.12.2006 4:06pm
tefta (mail):
A.Z. - Are you responding to me? If so, I didn't participate in any poll on this blog.
12.12.2006 4:37pm
Ilya Somin:
You mean response bias. Selection bias has to do with the sampling frame.

This is a minor technical point, but I think that selection bias is the right term. In a self-selected sample, it is the respondents themselves who do the selection, and thus a bias in the result is selection bias.
12.12.2006 5:35pm