Four fine posts from economics blogs

Rankings of economics blogs

Among the things economists are notorious for in academia is our intense--close to obsessive--interest in rankings. Rankings of journals, of particular papers, of individual economists, and of departments. (This assertion should be backed by some citations, but I don't want to make time to find them. Just trust me.) So I figured it was just a matter of time until someone produced a detailed ranking of economics blogs. Especially since I invited my many millions of readers to produce one over a year ago. 

But I haven't seen one. And a few minutes spent Googling suggests that there isn't one. So I guess it falls to me to begin this neglected but vitally important task.

What follows are preliminary rankings of economics blogs, computed in four different ways. Lest anybody take these rankings too seriously, after presenting them I"ll discuss eight big, big limitations on the methods and data used.

Ranking 1, using data from BlogPulse. BlogPulse is useful and fun. It claims to have identified 14.2 million--and counting--blogs.

Econ Rank Blog BlogPulse Rank
1 Marginal Revolution 88
2 Tim Worstall 119
3 Crooked Timber 146
4 Café Hayek 322
5 Angry Bear 450
6 The Becker-Posner Blog 525
7 Brad DeLong's Semi-Daily Journal 552
8 EconLog 703
9 Asymmetrical Information 879
10 Dynamist Blog 957
11 Mahalanobis 1136
12 SCSU Scholars 1426
13 Mises Economics Blog 1430
14 Agoraphilia 1532
15 Adam Smith Institute 1603
16 Eclectic Econoclast 1780
17 Knowledge Problem 1926
18 A Constrained Vision 2085
19 Truck and Barter 2106
20 Newmark's Door 2296
21 Freakonomics 2309
22 General Glut's Globblog 2606
23 Cold Spring Shops 2852
24 voluntaryXchange 3087
25 Division of Labor 3098
26 William J. Polley 3212
27 Vox Baby 3525
28 Deinonychus antirrhopus 3635
29 The Sports Economist 5516
30 New Economist 5777
31 The Liberal Order 7038
32 Peter Gordon's Blog 9454

I think it's interesting that in a universe of 14.2 million blogs, three of the top 200 are economics blogs. Is that more than we'd expect, or less? And I'll note immodestly that the top two link here, to the Door.

Ranking 2 uses information from The Truth Laid Bare. TTLB is a well-known and well-respected source of blog rankings. In last week's Washington Post article on political blogs, featuring my wife, David von Drehle cited TTLB rankings.

Econ Rank Blog TTLB--Links Rank BlogPulse Rank
1 Crooked Timber 45 3
2 Asymmetrical Information 77 8
3 Daniel W. Drezner 80 n/a
4 Brad DeLong's Semi-Daily Journal 107 6
5 Marginal Revolution 109 1
6 Dynamist Blog 153 9
7 Tim Worstall 167 2
8 Adam Smith Institute 365 10
9 Angry Bear 417 5
10 The Becker-Posner Blog 476 5
11 EconLog 685 7
12 EconoPundit 887 n/a
13 AtlanticBlog 1219 n/a
14 Prestopundit 1250 n/a
15 Knowledge Problem 1349 12
16 Deinonychus antirrhopus 1790 21
17 Coyote Blog 1799 n/a
18 BusinessPundit 1897 n/a
19 Truck and Barter 1927 14
20 Stumbling and Mumbling 2282 n/a
21 Cold Spring Shops 2311 17
22 Newmark's Door 2792 15
23 A Constrained Vision 2811 13
24 Mahalanobis 2895 8
25 Econbrowser 2906 n/a
26 Vox Baby 3026 20
27 The Sports Economist 3239 22
28 Brad Setser's Web Log 3318 n/a
29 The Idea Shop 3416 n/a
30 General Glut's Globblog 3504 22
31 Economist's View 3673 n/a
32 voluntaryXchange 3898 18
33 Eclectic Econoclast 4194 11
34 The Angry Economist 5440 n/a
35 New Economist 5777 28
36 Ben Muse 6964 n/a
37 William J. Polley 7092 24
38 Let's Fly Under the Bridge 8705 n/a
39 Heavy Lifting 8707 n/a
40 The One-Handed Economist 17967 n/a
Agoraphilia n/a
Café Hayek n/a
Division of Labor n/a
Freakonomics n/a
Mises Economics Blog n/a
Peter Gordon's Blog n/a
SCSU Scholars n/a
The Liberal Order n/a

Where "n/a" signifies "not available". 

I think the amount of agreement between the BlogPulse rankings and the TTLB--Link rankings is reasonable, with a couple of exceptions. Both Mahalanobis and The Eclectic Economist rank a bit better according to BlogPulse. I don't know why.

Ranking 3 is also based on TTLB's data, but this ranking is based not on links but on hits.

Econ Rank Blog TTLB--Hits Rank TTLB--Links Rank
1 Brad DeLong's Semi-Daily Journal 61 107
2 Angry Bear 127 417
3 Daniel W. Drezner 133 80
4 Tim Worstall 285 167
5 Truck and Barter 454 1927
6 EconoPundit 484 887
7 Coyote Blog 488 1799
8 Cold Spring Shops 494 2311
9 BusinessPundit 503 1897
10 Econbrowser 648 2906
11 Knowledge Problem 673 1349
12 General Glut's Globblog 740 3504
13 AtlanticBlog 785 1219
14 Deinonychus antirrhopus 981 1790
15 Newmark's Door 1301 2792
16 Prestopundit 1316 1250
17 A Constrained Vision 1353 2811
18 Eclectic Econoclast 1678 4194
19 Let's Fly Under the Bridge 2143 8705
20 The Idea Shop 2250 3416
21 Heavy Lifting 2826 8707
22 Ben Muse 4544 6964
Crooked Timber n/a 45
Asymmetrical Information n/a 77
Marginal Revolution n/a 109
Dynamist Blog n/a 153
Adam Smith Institute n/a 365
The Becker-Posner Blog n/a 476
EconLog n/a 685
Stumbling and Mumbling n/a 2282
Mahalanobis n/a 2895
Vox Baby n/a 3026
The Sports Economist n/a 3239
Brad Setser's Web Log n/a 3318
Economist's View n/a 3673
voluntaryXchange n/a 3898
The Angry Economist n/a 5440
New Economist n/a 5777
William J. Polley n/a 7092
The One-Handed Economist n/a 17967

Although a lot of data is missing for the TTLB-hits rankings, the agreement between the TTLB rankings for the 22 blogs that have values for both is quite reasonable (simple correlation coefficient = .819; Spearman's rho = .798). One interesting point raised by this table is that with three mild exceptions (Dresner, Worstall, and Prestopundit), the rankings of economics blogs based on hits are higher than their rankings based on links, sometimes substantially so.

I leave explaining this as an exercise for the reader.

Finally, ranking 4, using information from Google. Google currently claims to index over 8 billion Web pages. Google assigns each of those pages a PageRank, concisely characterized as "Google's measure of the importance of [the] page".

PageRanks are integers from 0 to 10, 10 being most important. I read a while back (but can't now find the source) that 10s are very rare, on the order of only two or three dozen. In a few of minutes of looking I found only four: Google itself, the New York Times, Harvard, and Statcounter. MIcrosoft, AOL, the Washington Post, Yahoo, Yale, and Oxford are among the sites that earn only a 9. Wal-Mart and General Motors are 8s. The very well-known blogs Instapundit, Daily Kos, and Drudge Report each earn a 7.  I have absolutely no evidence for this, but I guess PageRank might be logarithmic, like the Richter Scale.

Blog Page Rank
Brad DeLong's Semi-Daily Journal 8
Adam Smith Institute 7
Angry Bear 7
Asymmetrical Information 7
BusinessPundit 7
Crooked Timber 7
Daniel W. Drezner 7
Dynamist Blog 7
EconLog 7
Economics Unbound 7
Knowledge Problem 7
Mises Economics Blog 7
The Becker-Posner Blog 7
A Constrained Vision 6
Agoraphilia 6
ArgMax Economics 6
AtlanticBlog 6
Café Hayek 6
Chicago Boyz 6
Club for Growth 6
Cold Spring Shops 6
Coyote Blog 6
Deinonychus antirrhopus 6
Division of Labor 6
EconoPundit 6
Freakonomics 6
General Glut's Globblog 6
Law and Economics Blog 6
Mahalanobis 6
Marginal Revolution 6
Newmark's Door 6
Peter Gordon's Blog 6
The Angry Economist 6
The Idea Shop 6
The Sports Economist 6
Tim Worstall 6
Truck and Barter 6
volunaryXchange 6
Vox Baby 6

The rest of the economics blogs I checked had a PageRank of 5 or lower.

The clear winner in this ranking is Brad DeLong. A PageRank of 8 is quite impressive. 8s are earned by such hugely popular and important sites as Fark, Boing Boing, and Metafilter, and as noted, an 8 is higher than earned by Instapundit and Drudge. DeLong's is also the one of only two economics blogs listed in the top 100 blogs at Technorati (currently #84 among "13.8 million blogs"; the other is Crooked Timber at #60. Thanks to Henry Farrell for the correction.)

(This despite the fact that Professor DeLong once left a comment here that commanded me, "Don't be an idiot". Good advice generally, but not in that instance.)

There are twelve economics blogs with a PageRank of 7, and twenty-six, including yours truly, with a PageRank of 6.

I looked at a few other possible sources for data--Blogstreet, Bloogz, Popdex, and Technorati--but for one reason or another, I found them generally unhelpful.

Here are some of the serious limitations of these rankings.

1. We first need to decide 1) what a blog is and 2) what an economics blog is. Is a blog an economics blog based on its author(s) or its content or both?  Threshold percentages must be defined: what should they be? There clearly are no single answers to these questions.

Purely for convenience, I started with the list of 93 sources that Professor William R. Parke consults for his very useful Economics Roundtable.

2. But I didn't want to take the time to check all 93. I ended up checking 47 of the 93 (more than I initially intended). But the 47 I checked were not randomly selected; those chosen reflect my knowledge and preferences. I looked at blogs I think are best; blogs I read; and some others I was curious about. For the same reasons, and with the same non-randomness, I investigated an additional 19 blogs. In this set of 66 blogs, non-U.S. bloggers and left-wing bloggers are probably underrepresented. (But some of the sources I omitted wouldn't qualify as an "economics blog" under any reasonable defintion. Slate, for example.)

3. What time period should be used for the ranking? Life of the blog? Five years? Three months? Again, there's no "right" answer. Blogpulse uses 30 days. I'm not clear what time period TTLB and Google use. (I have no definitive information, but it seems as though Google penalizes younger blogs somewhat.)

4. Rankings based on "hits" and "links" are both problematic. Some hits are from people using search engines poorly. Some hit recorders capture just hits to the main URL and don't capture hits to individual pages or to individual posts. Some links might be from people who looked at the blog five years ago, once, and haven't updated their links. Links may well favor older blogs, because people didn't have as much to look at in the early days of the Web.

5. Except for the Google ranking, a lot of data is missing in the sources for the other rankings. I don't know why.

6. Just like student rankings via SAT scores and rankings of colleges and universities via U. S. News & World Report, "small" differences in ranks are statistically unreliable and substantively unimportant. The problem is that I see no way of knowing what "small" is. (The Google PageRank ameliorates this problem.)

7. With some effort all of these rankings could be gamed. Google would be hardest, but there's an industry devoted to advising people how to inflate their PageRank. For all I know, some of the blogs in this little study may already made such an effort.

8. I did this very fast, with minimal double-checking of data entry.

A note: I compiled the rankings Friday night and Saturday (7/22 and 7/23). By the time you read this, there could well be changes to the source data.

I don't intend to spend much, if any, additional time on this. But I'd be happy to email an Excel spreadsheet with my data, including URLs for all the listed blogs, to anyone who asks.

UPDATE 1: The Emirates Economist ranks 40th in ranking number 2 and 21st in ranking number 3.

UPDATE 2: The Google PageRank of voluntaryXchange corrected from 5 to 6. Thanks, Dave!

UPDATE 3: SCSU Scholars places 13th in ranking #2 and 14th in ranking #3. It can be located in TTLB's data by omitting the space: SCSUScholars. Thanks, King!