How Dual Class Share Structures Affect Innovation

In a new paper, we seek to fill a gap in research on the possible benefits of dual class share structures and how they might promote innovation. We start with a bit of history.

Shareholder democracy has been fundamental to the global rise of equity markets since the mid-19th century. The rule of one-share, one-vote sought to balance two competing interests: 1) entrepreneurial insiders wanting to maintain their control of the firm after the public offering of shares, and 2) minority shareholders concerned about the expropriation of their funds through dividend rights, private benefits of control, etc.  However, dual class share structures granting more voting rights to insiders gained popularity in the U.S. in the first two decades of the 20th century, increasing tension between the two interests.  In the 1920s, rising regulatory concerns over the expropriation of funds from minority shareholders led the New York Stock Exchange to  ban dual class shares in 1940. Four decades later, dual class firms were again allowed on the NYSE as the hostile takeover wars reached a peak.  By 1994 the NYSE, Amex, and NASDAQ had all adopted uniform rules allowing dual class shares, once again favoring insiders over minority shareholders.

While adoption of dual class share structures after the 1980s was initially slow, Google and other high-tech firms prompted a surge in 2004.  From 2005 to 2015, the proportion of initial public offerings that created dual class share structures rose from a mere 1 percent to 24 percent.  More than half of all U.S. technology, media, and telecommunications companies going through an initial public offering in 2015 chose a dual class status.  However, many recent academic studies suggest that the disproportionate insider control created in a dual class firm is associated with lower firm value, larger agency costs, and weaker oversight by the board of directors.  This has reignited the century-old debate over whether dual class shares should be banned in favor of a one-share, one-vote model.  But lacking from the research is any evidence of how a dual class structure might promote the positive benefits touted in the 1980s; namely, management’s  support of innovation.  As Lucian Bebchuk and Kobi Kastiel noted in their 2017 paper, “The Untenable Case for Perpetual Dual-Class Stock,” such evidence has been almost nonexistent.

In our paper, we seek to address this empirical scarcity by comparing innovation, as measured by the number and quality of patents,  at dual class and single class firms.  Since dual class structures are typically adopted at the time of an IPO, insiders might choose them only when innovative potential is high.  To counteract this complication, we examine innovation rather than R&D spending, because  firm insiders can control R&D spending but not whether it will result in a patent.  In addition, we use the matched sampling technique made popular in a seminal 2010 paper by Paul Gompers, Joy Ishii, and Andrew Metrick, “Extreme Governance: An Analysis of Dual-Class Firms in the United States,” which controls for factors leading to the choice of  a dual class structure at an IPO.  We augment this model by  considering patents pending at the time of the public offering and patents filed immediately after the offering.  We use data with consistent firm identifiers before and after the IPO date.  In this way, we compare dual class firms to single class firms with a similar patent pipeline as well as other factors known to influence the choice to adopt the dual class structure.

We find that a dual class structure enhances innovation as measured by patent output, quality, and creativity.  Specifically, we find a positive association between disproportionate insider control and the number of patent filings and patent citations generated by a firm’s innovations.  We also find that a dual class structure promotes more creative innovations as measured by how many industries the patents based on them affect.  While we confirm recent research indicating that a dual class structure does reduce firm valuation overall, we find that the improved innovation in dual class firms mitigates this loss.

In addition, we find that a dual class share structure improves innovation only in financially constrained firms.  Harry and Linda Deangelo argued in their 1985 paper, “Managerial Ownership of Voting Rights: A Study of Public Corporations with Dual Classes of Common Stock,” that private firms needing additional funds to continue their innovation will issue non-voting stock, giving them a hybrid status that allows them to retain innovative control while raising additional funds.  We believe that we are the first to find empirical support for this argument.

Also, dual class share structures can act as an antitakeover measure, and recent literature is mixed on whether such measures increase or reduce innovation.  Our paper broadly addresses this issue, but we specifically address this in one direct aspect by evaluating how dual class structures affect executives’ efforts to innovate. Telling evidence is provided by examining whether executives directly file their own patents. By law, any author listed on a patent filing must have spent extensive time on the innovation or the patent could be invalidated.  In other words, an executive who puts his name on a patent filing takes the substantial risk of being tied directly to the patent’s success or failure. We find a strong association between dual class share structures and a high number and quality of patents filed directly by executives.

Finally, we examine the effects of a dual class share structure on innovation after an IPO.  Much of the debate over dual class shares has centered on whether they should be “sunsetted:” eliminated after a certain period of time or a triggering event.  In support of this argument, we find that the positive benefits of a dual class share structure on innovation extend only about five years past the IPO.  We also find that executives who file their own patents produce more and better innovations only within the first five years after an IPO.

Our findings support the use of dual class shares as a way to protect innovative entrepreneurs seeking additional funds from the public markets.  However, our findings also support sunset provisions that would eliminate a dual class structure after a certain period of time.

This post comes to us from professors Lindsay Baran at Kent State University, Arno Forst at the University of Texas, Rio Grande Valley, and M. Tony Via at Kent State University. It is based on their recent paper, “Dual Class Share Structure and Innovation,” available here.