September 6th, 2014
from the Philadelphia Fed
Marshall (1890) noted long ago that, in industrial clusters, “[t]he mysteries of the trade become no mysteries; but are as it were in the air, and children may learn them unconsciously” (p. 271). Much progress has been made since Marshall made this famous statement about understanding exactly how agglomeration, knowledge spillovers, and innovation fit together, and what their impact on local and national economies might be.
Over the past two decades, research in this field has particularly surged. This has been unquestionably due in large part to the substantial data now available to better study these questions. It also surely has some connection to the growth of the knowledge economy, the fascination with Silicon Valley and related clusters, and similar popular themes. Particularly encouraging is the extensive set of connections being made from economic geography to adjacent fields such as labor economics, entrepreneurial finance, and business management.
Much, however, remains to be accomplished. With the handful of important exceptions reviewed previously, we still have not opened the black box of how clusters operate. Most of our empirical work has instead been comparisons over places. Better empirical guidance about the microinteractions within clusters with respect to innovation will allow us to differentiate among models and build stronger theoretical frameworks. Fortunately, the data are within reach to do so. Many employer–employee data sets are being linked to information on innovation (e.g., inventor identifiers), and these data sets often allow precise spatial coordinates of firms. This should prove a powerful lever for seeing further inside local areas, and it is likely that the line of work will increasingly draw upon network theory (e.g., Jackson, 2008; Dempwolf, 2012; and Kerr and Kominers, 2014).
We also need better insight into the long-term nature of agglomeration and innovation—the life cycles of innovative places. This is true within countries—innovation cores have shifted between Detroit, Boston, Silicon Valley, etc., and will continue to do so—but also true across countries. How does the rise of Bangalore impact Boston? Does the development of global innovation centers in rapidly emerging countries complement or substitute for those in advanced economies? The study of agglomeration and innovation is exciting because we are starting to make some progress at understanding the local and global nature of these phenomena.
Finally (although by no means closing off a comprehensive list), we need to develop better guidance about what local policymakers can do to foster agglomeration and innovation in their cities. This is a big and difficult question, and we are right to be cautious that we do not have all the answers. However, governments have spent billions of dollars on this, and many will continue to fund “be the next Silicon Valley” initiatives. This is true in advanced economies, in nations currently looking to transition from resource-dependent to a knowledge-based economy, in developing countries looking to leapfrog growth stages, and everywhere in between. Economists must continue to provide insight on these critical matters, and ideally our insights will get sharper faster.