Bobby Reddy has a post at CLS' blog, in which he reports that:
I review an array of empirical studies on U.S. dual-class firms, which embrace a range of performance measures – not just Tobin’s Q or derivatives thereof. I find that, on average, dual-class firms do show lower firm valuations than one-share, one-vote firms. However, when assessed from the perspective of accounting measures of performance, dual-class firms show greater operating performance than matched one-share, one-vote firms. Furthermore, investors in notional portfolios of inferior voting stock of dual-class firms earn at least equal, and possibly greater, returns than investors in notional portfolios of similar one-share, one-vote firms.
If dual-class firms are valued less than one-share, one-vote firms, but perform better financially and generate higher buy-and-hold stock returns than one-share, one-vote firms, it would suggest that the market is excessively discounting dual-class firms. There is no evidence that, on average, dual-class structure is used to exploit public stockholders, at least not to the extent that it outweighs the structure’s positive impact on operating performance. The market appears to discount dual-class firms on the assumption that public stockholders will be harmed, but that harm is not, on the whole, born out in practice.
Go read the whole thing.