Six Degrees of Kevin Bacon: Economists’ Version

Justin Lahart at the Wall Street Journal suggests a new party game for economists (or at least something to keep you awake if a conference gets dull): Six Degrees of Joe Stiglitz. He’s suggesting the econ version of the Paul Erdos number in math:

If you drew a diagram linking different mathematicians, many of the lines would cross at Erdos. That’s given rise to what’s known as an Erdos number. People who collaborated directly with Erdos have an Erdos number of one. People who collaborated with those collaborators have an Erdos number of two. And so on. Erdos, who died in 1996, has an Erdos number of zero.

Why choose Stiglitz? Well, Stiglitz himself has an Erdos number of four, which means that any economist’s Erdos number is at least as low as their Stiglitz number, plus four. But more to the point, Stiglitz has coauthored broadly and written across many fields, so is arguably the most Erdos-like figure in modern economics.

I’ll bite: My Stiglitz number is three, and I can get there at least three different ways:

Stiglitz-DiamondBlanchard-Wolfers, Stiglitz-AtkinsonLeigh-Wolfers, or Stiglitz-BlinderReis-Wolfers.

Steve Levitt is also a three: Stiglitz-SachsPoterba-Levitt, and so Steve endows both Stephen Dubner and Sudhir Venkatesh with a four.

Dan Hamermesh has written so many papers that it is hard to know, but I’m guessing he’s also a three: Stiglitz-AzariadisDrazen-Hamermesh.

Among Freakonomics bloggers, Ian Ayres has the lowest Stiglitz number at two: Stiglitz-Nalebuff-Ayres, or Stiglitz-Edlin-Ayres. (And Ayres is also a secondary reason for Levitt’s three.)

Need help calculating your Stiglitz number? Here are the rules of what constitutes a fair link. And here’s Stiglitz’s C.V.

My guess is that some of the key links (people with a Stiglitz number of one) will turn out to be: Aghion, Akerlof, Arnott, Atkinson, Azariadis, Basu, Blinder, Boadway, Boskin, Brock, Cass, Dasgupta, Diamond, Dixit, Easterly, Eaton, Edlin, Frydman, Fudenberg, Furman, Gilbert, Greenwald, Grossman, Heal, Hellman, Hosios, Jaffee, Nalebuff, Newberry, Niskanen, Orszag, Przeworski, Radner, Rothschild, Sachs, Sah, Salop, Shapiro, Shell, Solow, Stern, Stockman, Tirole, Walsh, Weiss, and Woodford — just to list the first couple of dozen among the multitude of people boasting a Stiglitz number of one.

And when we are done playing the Six Degrees of Joseph Stiglitz, let me suggest that for subsequent rounds we try Six Degrees of: Larry Summers, Gary Becker, Milton Friedman, Kenneth Arrow, Paul Samuelson, John Maynard Keynes, and Alfred Marshall.

These get much harder with earlier generations of economists, because team production of research is increasing.

Jeff S.

Yes, but what are their Tarjan numbers?


Given that you often refer to baseball (a sport I don't follow) on this blog, eg. the Red Sox thing you've got going now, I thought you'd like to know the following:

Hank Aaron (a major league baseballer of some note apparently) and Paul Erdős received honorary degrees from Emory University on the same day and took the opportunity to autograph a baseball together. Hence Hank Aaron has an Erdős number of 1 (for this "collaboration").


The paper of Sanjeev Goyal, Marco van der Leij and José Luis Moraga-Gonzalez ( studies the economists network based on the co-authors during the 1970s, 1980s and 1990s.

They show the increasing number of economists, reviews, publications... and the fact that economists are more and more connected (40% in a same network), more and more closed (9 degrees of separation) and more and more small worlds...

The network is organized around galaxies of stars, with Stiglitz and Tirole in frontline.

And in my phd thesis, I use the interaction of nobel prizes described in their autobiography to map the laureates network (an under estimation of the creative discussions). The 61 laureates form a small community and the degree of separation is not even 3.

In sum, if you are one of the connected economists (40%), you are already closed to Stiglitz. And, the closer you are to Stiglitz, the closer you are to nobel community.



This is now an Actual Internet Phenomenon, as it's been parodied in the fine webcomic, XKCD:

(You have to hover your mouse over the comic to get the 'hidden' punchline.)


Personally, I'm calling for an
(Annabel) Chong-Erdos-Bacon-Stiglitz number index...

Any takers?


Who is the leigh you are connected to?


JC: Andrew Leigh, at Australian National University. Also a blogger.


Let's see whether we find an economist whose Erdös number is lower than his Stiglitz Number.

Eric Maskin is close. His Erdös number is 2

(Unfortunately) he has published with R. Radner who has a Stiglitz number of 1 which makes Maskin's Stiglitz number a 2.

Procrastinators step forward!


Does this mean we'll soon see the advent of Erdos-Bacon-Stiglitz numbers?


Wow, my Stiglitz number is 5. Stiglitz-Sachs-Poterba-Levitt-Kessler-Gerber-me

Jane @

I once handed Stiglitz a drink at an Economics Department party at Columbia University. And if I showed my friends and family the picture someone snapped of him, me, and my daughter, they'd ask, "Who's that?" and they wouldn't be talking about me. Or my daughter.

But he did seem like a nice guy.