Does exporting raise productivity? Dean Yang et al. use export demand shocks from the Asian financial crisis to instrument for increased exporting and find a big effect:

Between 1995 and 1998, the Japanese, Thai, and Korean currencies depreciated in real terms against the US dollar by 31%, 32%, and 43%, respectively. At the other extreme, the British pound and the US dollar experienced real appreciations against the yuan, by 14% and 7%. Because the exchange rate changes varied so widely, two observationally equivalent firms faced very different export demand shocks if one happened to export its goods to Korea and the other exported to the UK…

For each firm, we construct an exchange rate shock measure specific to that firm, which is the average exchange rate change of a firm’s export partners weighted by the firm’s export destinations in 1995 (prior to the Asian financial crisis). We focus on changes in exports driven by these exchange rate shocks.

Using this approach, we ask whether and how instrumented changes in exports affect measures of firm performance. We find that increases in exports are associated with improvements in total factor productivity, as well as improvements in other measures of firm performance such as total sales and return on assets. Our estimates indicate that a 10% increase in exports causes productivity improvements of 11% to 13%, nearly one-eighth of the mean productivity improvement from 1995 to 2000 in our sample.