There is nothing I like more when reading scholarly work than the thoughtful interrogation of long-standing myths about a particular phenomenon. Take, for example, the article by economists Nicola Persico, Andrew Postlewaite, and Dan Silverman, “The Effect of Adolescent Experience on Labor Market Outcomes: The Case of Height” (http://ssrn.com/abstract=293122). For years economists (and frankly any casual observer) noted that taller people make more money and seem to command greater power. This was widely explained as a bias in favor of height (or as discrimination against short people). These three economists took that hypothesis and explored whether there was a stronger determinant of success as reflected in labor market returns. Surprise! It turns out if you look at height of white males over time against labor market success you discover that boys who were tall in high school have stronger labor market returns. Adult height is only relevant to the extent that it correlates with youth height.
Bank and Cheffins’ paper is of the same ilk. They take a look at a long standing puzzle: why does the United States lack the corporate pyramid structures common in other countries of the world? In some countries it is common for a successful individual or wealthy family to have a small stake in the overall operations of cascading companies, but to have a sufficient stake in the chain that they control billions of dollars worth of corporate activities with only a limited personal or family investment. These arrangements have given rise to a variety of well documented concerns, including that a small number of investors exercise enormous control at the expense of minority shareholders, that tiny elites are given disproportionate access to corporate power, and that these investors have huge sway with government. The often told example of this kind of arrangement, retold by Bank and Cheffins, is of the unmarried brothers, Oris Paxton Van Sweringen and Mantis James Van Sweringen who were able to invest less than 20 million dollars but control eight class I railroads with combined assets of over two billion dollars.
In explaining the relative absence of these kinds of pyramid structures in the United States, it has become common to speculate that their absence is due to the United States’ tax on dividends to corporate shareholders, introduced in 1935. Other countries, including Canada, allow inter-corporate dividends to pass tax free. As the standard story goes, the imposition of the U.S. tax would have made investing through cascading or pyramid companies tax inefficient, and so investors would not have been as inclined in the U.S. as in other places to use pyramid structures to enhance their corporate control. Put simply, the taxation of intercorporate dividends added to the already heavily taxed distribution of corporate profits. (Corporate profits, at least for much of United States’ history have been taxed in the corporation and again when distributed to individual shareholders. This additional tax ensured that they were also taxed when distributed between corporations.)
After devoting some time to distinguishing pyramid structures from other corporate structures, Bank and Cheffins turn to the question of why the United States has managed to avoid the proliferation of corporate pyramids. They walk through the evidence marshaled to support the standard story, which relies to varying measures, but always significantly, on the imposition of the tax on intercompany dividend distributions in the mid-1930s. Bank and Cheffins then turn to a dataset they created based on filings with the Securities and Exchange Commission between 1936 and 1938, the first years in which the dividend tax applied. To reveal at least part of their punch line – they discover that the imposition of the tax had very little role to play in explaining the pyramid puzzle.
Bank and Cheffins dismiss some alternative explanations for why the introduction of intercorporate dividends taxation did not create a disincentive for pyramid structures. They dismiss the explanation that dividends were not an important part of corporate finance. Although the importance of dividends has fluctuated they find they were a core element of corporate life in the 1930s. They also dismiss a substitution effect – that firms may have been avoiding the payment of dividends by buying back shares (this simply did not hold on the evidence). Finally, general market trends, namely an overall decline in the market, which may have explained the limited role of the tax, is dismissed on the evidence of relatively strong share performance during the period under examination.
Once a scholar has questioned a supposed correlation, and proposed a new possible correlation, it’s fun to see what explanations are offered for the observed relationship. For example in the paper by Persico et al. the authors tentatively suggest that it is cultural and social factors that led to the increased labor market outcomes for tall youth. Put simply: tall youth are more likely to build self-esteem by enrolling in athletics and clubs and that self-esteem building activity informs deeply their later ability to command top market wages.
Bank and Cheffins do not disappoint. They turn to offer some tentative alternative explanations for the rarity of pyramid structures in the United States. While their analysis is detailed and sophisticated, in summary, they find that pyramid corporate structures were simply never a major part of the U.S. corporate landscape. Indeed, the only commercial activity that was riddled with corporate pyramids in the 1930s was in the utilities sector. Bank and Cheffins suggest that this sector offered a welcome mat to pyramids because the businesses were usually small, isolated, locally owned and controlled, and financially weak. Among several other reasons carefully reviewed by Bank and Cheffins, their need for capital funding and the high risk the sector presented for investors explains why utilities became a haven for corporate pyramids. The introduction of the Public Utilities Holding Company Act might explain why this sector was substantially reorganized. Given that other industries have not suffered from some of the financial market limits that characterized the utilities sector in the 1930s, Bank and Cheffins conclude “the well-developed financial markets from which the U.S. has benefited may do considerably more to explain the rarity of corporate pyramids than do tax or other regulatory variable” (at 57).
This is an excellent and interesting article that grapples with a long-standing puzzle. Its conclusions should unquestionably inform public policy developments on corporate governance. It is well worth the read.