Knowledge is a commodity

The map of ranks for cities in the selected European countries in 1990. Image courtesy of Qian Zhang.

The map of ranks for cities in the selected Euro­pean coun­tries in 1990. Image cour­tesy of Qian Zhang.

When econ­o­mists talk about pro­ducers and consumers—the people that make stuff and the people that use it—they’re usu­ally thinking about com­modi­ties like coffee, wheat, or oil. Not knowledge.

That’s because knowl­edge hasn’t really been quan­tifi­able before. But today, in this era of the “data deluge,” it is. For the first time in his­tory, we can put num­bers on what we know.

That’s what North­eastern Dis­tin­guished Pro­fessor of Physics, Com­puter Sci­ence, and Health Sci­ences Alessandro Vespig­nani and his team did in research pub­lished last week in the Journal Sci­en­tific Reports. They looked specif­i­cally at the “com­modity” of physics and how it’s been passed around the world over the last fifty years.

First they assigned geo­graph­ical loca­tions to all of the papers pub­lished in the jour­nals of the Amer­ican Phys­ical Society between 1960 and 2009, based on the affil­i­a­tion of the authors. Papers written by North­eastern researchers would be tagged “Boston, Mass.” Papers written by Notre Dame authors would be tagged “South Bend, Indiana.” The effort took a good year to com­plete, and was mostly under­taken by the article’s first author, grad­uate stu­dent Qian Zhang.

The authors used this “geolo­cal­iza­tion data” to under­stand not just the raw number of papers pub­lished in each place, but rather the number of cita­tions to those papers exchanged among loca­tions. They call these cita­tions “the cur­rency of the knowl­edge exchange.” Like dol­lars and euros, they are a proxy for knowl­edge pro­duc­tion or con­sump­tion, depending on whether the final bal­ance is pos­i­tive or negative.

While the APS jour­nals are based in the US, only 43 per­cent of the total pub­li­ca­tions came from US authors, but the trend was not static across the 50 years they looked at. In the six­ties 86 per­cent were Amer­ican, in the last ten years it’s down’ to about 37 per­cent. The team focused their analysis of all this data on the last twenty years, when the national rep­re­sen­ta­tion among the APS jour­nals was rep­re­sen­ta­tive of physics around the world.

They found that while New York City was number two in physics pro­duc­tion in the 1990s, it’s not even on the map today. Twenty years ago the east and west coasts of the US pro­duced the majority of the nation’s physics knowl­edge, but today it’s scat­tered across the country.

Whereas two decades ago Europe was more of a con­sumer of physics, it’s become a major pro­ducer. And whereas China was absent from the physics con­ver­sa­tion twenty years ago, it’s the leading con­sumer today. That’s because they are churning out papers faster than you can say “par­ticle accel­er­ator” and are citing boat­loads of pre­vious work in the process.

This gets to an impor­tant dis­tinc­tion, Nicola Perra, one of the authors, told me. Pub­lishing lots of arti­cles doesn’t make a country a top-​​ranking physics pro­ducer. That pro­duc­tion needs to be valu­able. It’s deemed such by other researchers who cite the work in their own papers. “Con­sid­ering just the number of arti­cles is not the real indi­ca­tion of posi­tion or impor­tance,” he said. “You need to con­sider the entire system.”