This post was originally drafted in October and November 2013. I have kept the post largely as is, with some updates clearly marked as such. At the time of my late October 2013 editing round, the page had between 750 and 800 likes, it now has over 1000 (see historical Facebook data for more context).

Clarification: This post concentrates on the reach of “open borders” among “Internet libertarians” but I don’t intend to indicate that this is the only target audience for the site. So far, Internet libertarians, and people who are at the edges of that category, have been a major source of likes and engagement for the site. I expect that Internet libertarians will continue to have a high, but (proportionally) declining contribution to the like counts for the site.

A few weeks ago, I played around with Facebook Graph Search with the goal of trying to figure out how close Open Borders: The Case is to exhausting its potential target market among various segments of the world population. None of the findings were particularly surprising, but I think they’re still worth recording and sharing. The numbers here relate somewhat with my discussion of open borders and the libertarian priority list (part 1 here, part 2 here, part 3 pending). They also relate with the Drake equation-style estimation of the value of open borders advocacy. However, in this post, I’ll concentrate mostly on reporting the numbers, along with a bit of background on my prior speculation and my speculation after seeing the numbers.

My prior views

Prior to looking at these numbers, I had been of the view that there’s a broad category of person called “Internet libertarian” who likes websites that make arguments from libertarian premises, and that:

• Of the category of “Internet libertarians”, a reasonable percentage (say 10%) would be in a position to like Open Borders: The Case if they came across the site and browsed it a decent bit. When I talk of “the site” I mean the site as it currently stands, without any changes being made to tailor to the audience. Note that my guess would be that about 50% of Internet libertarians would claim to be for radically freer migration in principle, but only about 20% would care about the issue sufficiently to be willing to publicly like our site.
• My original estimate of “Internet libertarians” was that it would be something in the range of 50,000 (I was going off the estimates of how many people like mainstream libertarian-leaning websites and looking at the lower end of the range).

Combining the above, my original estimate was that there would be about 5,000 “Internet libertarians” who would be willing to like the site if they came across it and read it a bit. As of the time of writing this post, the website has between 750 and 800 Facebook likes. This suggests that there would be low-hanging fruit of about 4,000 people who would be willing to like the site more or less as is, but simply haven’t yet encountered the site (or have, but not for long enough or frequently enough to consider hitting the “Like” button).

My views after looking at the data

Although the numbers don’t definitively falsify the hypothesis, a close look at the numbers did lead me to update my model. My overall conclusion (POOTA numbers alert): is that one can subdivide libertarians into roughly three clusters (note that the clustering is quasi-arbitrary, and there is in reality a continuum, but it’s easier pedagogically to frame things in terms of clusters):

• A core (perhaps 2,000 people or so) of relatively hardcore libertarians — including people who work in academia, for think tanks, or for libertarian publications, and people who maintain active libertarian-leaning blogs — who are disproportionately likely to like many libertarian pages (including Open Borders). This core of people does not necessarily include only libertarian extremists. The people here may include anarcho-capitalists or moderate minarchists. It may range from social justice-loving bleeding-heart libertarians to dogmatic natural rights believers. The key feature of this core of people would be that they spend considerable effort researching, writing, or in other ways promoting libertarian ideas. Many of them may have jobs in academia or think tanks where libertarian writing is part of their professional work. The website’s penetration with this group of people seems fairly strong. I would guess about 25% of the people in this group (total: 500) would be willing to like Open Borders, and about half of those who would be willing to like it have already encountered it and liked it (total: 250). Update: As of February 2014, I think we have covered about 70% of such potential likers.
• A much larger pool of people (about 50,000 by my estimate, based on feedback from some others) who like reading libertarian material on the Internet and occasionally comment on libertarian websites. Not all of these people need identify with the label of libertarian: many of them may consider themselves free-market liberals or free-market conservatives or perhaps “fiscally conservative, socially liberal” and some may even oppose key tenets of libertarianism. But they do engage regularly, and at least in some cases actively (for instance, by leaving comments on blogs), with libertarian content on the Internet and in real life (for instance, by attending local Students For Liberty events or going to Cato talks). My guess is that of this group, perhaps 5% or so would be willing to like the Open Borders website as is if they saw it (total: 2500), and we have probably penetrated about 10% of this market (about 250 people).Update: My guess is that we may have gone up to about 15% of the people in this group.
• An even larger pool (about 150,000 by my estimate) of people who like libertarian websites on Facebook and occasionally read libertarian blogs but aren’t really into closely following them. I expect that only about 1% (total: 1,500) or so would like the Open Borders website, and perhaps about 5% of the people in that group have liked the site already (total: 75). Update: My guess is that we are now at 7-8% of this group.

Note that the numbers I provided in this paragraph are very rough guesstimates based on the numerical intuition I acquired after compiling the numbers below. I certainly did not derive the numbers using any sort of mathematical procedure from the data.

Note also that this should not be taken to construe that the site’s long-term appeal among “libertarians” is limited to the ~4,500 or so people that my above guesstimates suggest would like it if they saw it today. As the site gets redesigned to have more visual appeal and as the ideas become more mainstream within libertarian circles, the number of people who like it may go up. But that’s not low-hanging fruit — it requires site redesign effort and effort in making open borders more mainstream within loosely libertarian circles.

Data based on Facebook Graph Search

A few limitations of graph search worth keeping in mind:

• Due to privacy settings, some people may not appear in the lists, even though they do affect the counts. Thus, the manual counting for the “Fewer than 100 people” cases is in most cases an underestimate.
• Facebook does not provide exact counts, instead generally saying “fewer than 100 people” or “more than 100 people” if the number of people is more than (about?) 15.

Below is a partial list of libertarian Facebook pages that had more than 100 common likers with Open Borders as of October 26, 2013.

Page	Number of likes as of October 26, 2013	Link to Facebook Graph Search query for intersection
Economic Freedom	297,158	intersection
Independent Institute	246,138	intersection
LearnLiberty	245,830	intersection
Cato Institute	186,186	intersection
Reason Magazine	109,485	intersection
Students for Liberty	98,135	intersection
Mises Institute	87,911	intersection
Libertarianism.org	72,953	intersection
EconLib	41,341	intersection
Institute for Humane Studies	31,774	intersection
Skeptical Libertarian	21,935	intersection
Mercatus Center	18,405	intersection
Bleeding-Heart Libertarians blog	5,191	intersection

Here is a list of websites I checked that had less than 100 likes in common with Open Borders: The Case as of October 26, 2013. For those where the intersection is now estimated as over 100, I have indicated that parenthetically. For others, a guesstimate of the new value is provided.

(now about 60)

Page	Number of likes as of October 26, 2013	Size of intersection as of October 26, 2013, along with link to Facebook Graph Search query (parenthetical value is estimate as of February 8, 2014)
Punk Rock Libertarians	58,998	35 (now more than 100)
Libertarian Troll	8,208	17 (now about 30)
Center for a Stateless Society	6,059	50 (now more than 100)
Cafe Hayek	5,018	53 (now more than 100)
Liberty Fund	1,319	40 (now about 50)
Future of Freedom Foundation	485	39

Multiple intersections (main information October 26, 2013, updated information February 8, 2014):

• The intersection with the three most liked libertarian pages in the list was fewer than 100 people (it is now more than 100 people).
• On the other hand, the combined intersection with the Cato Institute, Reason Magazine, and the Independent Institute was more than 100 people. All three of these are relatively famous libertarian pages.
• The combined intersection with EconLib and Bleeding Heart Libertarians was fewer than 100 people, with 40 visible search results (now, about 56 visible search results).

I noticed from the above that the intersections with much smaller and more hardcore libertarian pages are not that much smaller than the intersections with relatively more widely liked pages. This suggests that our likers have a significant fraction of “movement libertarians” — people who have strong commitment to libertarian ideas and for whom it may even be part of their professional work. To test this hypothesis, on November 3, I considered intersections with people who had ever worked at a number of libertarian places. Note that many people don’t fill out their employer information on Facebook, so these numbers are all almost certainly underestimates.

Libertarian institution or publication	Number of people who like Open Borders and have ever worked for this institution or publication (link to Facebook Graph Search query) (numbers as of November 3, 2013)
Cato Institute	6 people
Reason Magazine	1 person
The Independent Institute	1 person (but this is an underestimate relative to what I know)
Students for Liberty	3 people
Institute for Humane Studies	3 people
Ludwig von Mises Institute	1 person

Although these numbers do not seem to work out to very high values, anecdotal evidence of the relatively high degree of commitment to libertarianism of people who like the site seems to bolster the hypothesis.

Penetration rates for blog subcultures

Libertarians are not a homogeneous group. There are different subcultures within libertarianism, and each subculture has a slightly different typical mix of blogs followed.

Currently, Open Borders has been mentioned in a number of libertarian blogs, but to widely varying degrees. Here’s my sense of things, based on data collected at the external coverage page (Update February 2014: I expect each of these to have moved up a couple percentage points, but not significantly, since Open Borders did not receive much additional coverage in the blogs listed below)

Libertarian (or libertarian-leaning) blog	Extent to which it’s mentioned Open Borders	Estimate of penetration: (number of regular or semi-regular readers who have liked Open Borders)/(number of regular or semi-regular readers who would like Open Borders if they visited the site), expressed as a percentage
EconLog	1 post about the site, 4 links to site posts with substantive discussion, about 5 other links to site posts, plus 2 posts about EconLog blogger Bryan Caplan’s guest posts on Open Borders	80-90%
Volokh Conspiracy	1 post linking to a page on the site, 1 post about a guest post on the site by the author Ilya Somin	30-40%
Cafe Hayek	2 posts mentioning the site in link roundups.	20-30%
Bleeding Heart Libertarians	1 post mentioning a substantive post	20-30%
Marginal Revolution	2 brief mentions of the site, 1 brief mention of a post on the site	5-10% (update: now 10-15%)