Snap Inc.'s (SNAP) initial public offering marked a turning point in investors' attitude toward voting structures that vest control in a small group of insiders.

Shares the instant messaging company sold to the public in March carries no voting rights, while company founders Evan Spiegel and Bobby Murphy own stock giving them 88.5% of the vote and company executives have the rest.

Snap's aggressiveness crystallized institutional investor opposition to such structures, and on July 31 S&P Dow Jones Indices LLC said it would no longer include companies with dual-class or multi-class shares that give insiders at least some control in its indices, including the S&P 500, S&PMidCap400, and the S&PSmallCap600.

Snap's stiff-arm to public investors is only the latest episode in a long history that Andrew Winden chronicles in his new paper "Sunrise, Sunset: An Empirical and Theoretical Assessment of Dual-Class Stock Structures." Winden, a former corporate partner at Morrison & Foerster LLP and a fellow at the Rock Center for Corporate Governance at Stanford Law School, surveys the voting arrangements that companies have used to ensure that founders continue to control a company after it goes public and - occasionally - to give minority stockholders a measure of oversight.

Some of those structures are useful models for drafters who want to balance the interests of founders and public stockholders.

The New York Stock Exchange barred dual-class listings in 1926 in the wake of opposition to Dodge Brothers Inc.'s offering of non-voting shares the previous year, a policy NYSE adhered to until 1985 with a few exceptions, including Ford Motor Co. (F) NYSE also allowed Nike Inc. (NKE) o go public in 1980 with a structure that gave founder Phil Knight the right to elect 75% of the athletic apparel company's directors and public stockholders the right to elect the other 25%. All Nike shares have an equal vote on matters other than director elections. Other companies, including Molson Coors Brewing Co., have used similar schemes, which Winden endorses.

Giving investors board seats, Winden writes, "is probably the best compromise between entrepreneurs' desire for management discretion and investors' desire for protection from poor management decisions." As he notes, Knight wrote in his recent memoir Shoe Dog that he wouldn't have been willing to take Nike public had he not been able to retain control of the board. That position suggests the disadvantages to unyielding institutional investor opposition to dual-class structures, Winden writes, which could lead to "more companies staying private longer, diminished opportunities for investment by retail investors and less publicly available information about the most vibrant engines of growth in the economy."

The American Stock Exchange and Nasdaq were more forgiving than NYSE. Amex allowed Wang Laboratories Inc. to go public in 1976 as long as it adhered to several requirements, the first of which anticipated Nike's voting structure. The class of low-vote stockholders had to have the right to elect at least a quarter of the directors; the high-vote shares could have no more than ten times the vote of low-vote shares; the voting rights of the low-vote shares couldn't be diminished by the creation of additional stock; and the high-vote shares could lose certain preferences if their number fell below a certain percentage. The Amex also strongly suggested that the low-vote class have a dividend preference. After Wang went public in 1976, at least 22 companies adopted a similar formula in their IPOs, and seven companies that were already public adopted dual-class structures.

Despite the importance of antecedents like Wang and Nike, much of the innovation in dual-class structures has come in the last 10 to 15 years, Winden writes, and has often centered on sunset provisions that provide for the end of dual-class voting. Only 36% of the companies in his study that went public before 2000 included such provisions, while 61% of those that have gone public in this century have, with a further increase since 2009.

Companies such as Fitbit Inc. (FIT) , Groupon Inc. (GRPN) and Yelp Inc. (YELP) have provided for the end of dual-class voting a specified period of time after an IPO, generally between five and 10 years. There has also been an increase in provisions that provide for the end of dual-class voting upon the death or incapacity of the company founder. Moelis & Co. broke new ground in 2014 when it provided that founder Kenneth Moelis's high-vote stock would lose that trait if he were let go because he would no longer "devote his primary business time and effort to the business and affairs of the Corporation and its subsidiaries or because he suffered an incapacity."

A founder's ability to transfer high-vote shares is also a key issue. Google Inc.'s (GOOGL) 2004 IPO "caused a sea change" on the issue, Winden writes. Previously, companies tended to allow high-vote shares to be transferred, which allows for one family to control a company for generations; Tyson Foods Inc. (TSN) and Ralph Lauren Corp. (RL) are examples. Only 13 of the 68 companies in Winden's study that went public after 2004 allowed for the free transfer of high-vote stock.

Winden favors sunset provisions based on death, incapacity and a founder's separation from a company and believes that companies should limit a founder's ability to transfer shares to a family member or favored successor. But he opposes time-based sunset provisions as an unreasonable limit on a founder's ability to implement his "idiosyncratic vision," a position that runs counter to the one taken by Harvard Law School's Lucian Bebchuk and Kobi Kastiel in their paper "The Untenable Case for Perpetual Dual-Class Stock," which appeared earlier this year in the Virginia Law Review.

Whatever their terms, dual-class structures don't exempt companies from takeovers or other forms of market discipline. Winden observes that only six of the companies included in a 1987 study of companies with dual-class structures are still listed, and only two remain dual-class companies, while the rest have been bought or gone bankrupt. And only eight of the 43 in a 1985 study of dual-class structures adopted between 1961 and 1984 are still listed dual-class companies; other have been purchased, gone bankrupt, or collapsed their dual class structures.

Don't miss these top stories on TheStreet:

Editors' pick: Originally published Aug. 18.

More from Corporate Governance

What Is a SWOT Analysis and When Are They Done?

What Is a SWOT Analysis and When Are They Done?

What Is Vertical Integration and What Are The Benefits?

What Is Vertical Integration and What Are The Benefits?

What Is an Oligarchy and What Does It Mean in 2018?

What Is an Oligarchy and What Does It Mean in 2018?

CBS Board Knew About Sexual Misconduct Claims Against Moonves, Report

CBS Board Knew About Sexual Misconduct Claims Against Moonves, Report

Wednesday Wrap-Up: You Gotta Invest

Wednesday Wrap-Up: You Gotta Invest