Dynamic Patterns of Industry Convergence : Evidence from a Large Amount of Unstructured Data

The rapid changes in the market environment as well as technological innovations drive firms to devise differentiated products or services that functionally overlap with other products or services. Such overlapping in markets with converged products or services eventually drives industries to converge. However, despite the significant impact of industry convergence on the economy, our understanding of the phenomenon is limited because previous studies largely come from the scientific/technological perspective. Accordingly, in this paper, we investigate the trends and patterns of market-side industry convergence for entire sectors of the U.S. economy. To do so, we conduct a co-occurrence-based analysis of text mining for a large volume of unstructured data — newspaper articles from 1989 to 2012 — and suggest an industry convergence (IC) index based on normalized pointwise mutual information (PMI). We find that industry convergence has been increasing over time, but it does not take place throughout all industries. Some industries show strong convergence intensity while others do not. We also identify industry groups that are converging over time as well as those that are diverging, which creates heterogeneous patterns of industry convergence. These findings suggest that significant transformation is under way in the economy, but the trends and patterns are quite heterogeneous, depending on the industry. In addition, this study provides a method for anticipating the future direction of industry convergence