Three Japanese researchers (Toshihiro Okubo, Tetsuji Okazaki and Eiichi Tomiura) point to this interesting research.
Clustering is just like an industrial social network in many ways. Just like we see people respond/reciprocate faster to people in their social network, it is expected firms in clusters to do faster transactions with each other. There is a lot of learning etc which flows within firms as well.
Also interesting to note how regional banks finance much of the Japanese industrial networks. The regional banks become a part of the process as well..
India also has quite a few clusters spread across the country. One can expect similar results to show there as well. Infact many local and regional banks in the 20th century came up in the 20th century just to finance these local/regional industries as well. But only few of these banks could succeed. Post multiple failures in mid-20th century, the regulator and government forced many banks to either liquidate or merge with other bigger banks. Eventually branch banking took over and banks from selected places ended up serving all national and regional industries.
It could be an interesting study as well to look at contribution of these local banks in local economies when they existed. But one needs a fair bit of data to prove such kind of linkages..