Update on Corporate Political Activity

John Coates is the John F. Cogan, Jr. Professor of Law and Economics at Harvard Law School. This post relates to research by Professor Coates discussed on the Forum here, as well as a recent post on a Manhattan Institute Legal Policy Report, discussed here. Work from the Program on Corporate Governance about corporate political spending includes Corporate Political Speech: Who Decides? by Lucian Bebchuk and Robert Jackson, discussed on the Forum here. A committee of law professors co-chaired by Bebchuk and Jackson submitted a rulemaking petition to the SEC concerning corporate political spending; that petition is discussed here.

Corporate politics continue to generate controversy. Recent items of note include (1) the US Supreme Court’s decision to expand the reach of Citizens United in Western Tradition Partnership; (2) the continued increase in the number of and support for shareholder proposals calling for disclosure of corporate political activity; and (3) a recent “study” sponsored by the conservative Manhattan Institute (and described on the Forum here) purporting to find that – as the Wall Street Journal put it – “politics spending pays” – contrary to my own research, which finds that large public companies that were politically active before Citizens United experience a decline in their industry-adjusted market value after the decision. Each of these developments is discussed briefly below.

Western Tradition

In Western Tradition, without hearing or briefs, the US Supreme Court summarily reversed, by a partisan 5-4 vote, a detailed Montana Supreme Court decision upholding laws restricting corporate election activity in that state. Notably, the lower court had unanimously agreed that the state laws in question were good policy, based on Montana’s long history of out-of-state companies corrupting its elections.[1] An example noted by that court:

In 1903, [the] Anaconda [Company, then controlled by Standard Oil], closed down all its industrial and mining operations … throwing 4/5 of the labor force of Montana out of work. … Its price for sending its employees back to work was that the Governor call a special session of the Legislature to enact a measure that would allow Anaconda to avoid having to litigate in front of [two judges in] Butte, who had turned down bribe offers from Anaconda.

This and other scandals led to the enactment of the laws challenged in Western Tradition, in 1912 – long before the US Supreme Court ever found any law, federal or state, unconstitutional under the First Amendment. None of this history was addressed by the US Supreme Court, which seemed to think it self-evident – not in need of explanation – that its own peculiar views of “speech” and corporate governance should trump all countervailing concerns about corruption, democracy or federalism. This despite the fact that 75% or more of the US people think it went off the rails in Citizens United.

Corporate Shareholder Proposals

The recent spring proxy season included continued high and increasing levels of support for shareholder proposals addressing corporate political activity. As summarized here (free registration required) by ISS, and contrary to some media reports, the number of shareholder proposals increased in 2012 – in fact, political activity was the focus of more than 100 proposals, the largest issue class within the broad “environmental and social” issue category tracked by ISS. With the increase in proposals has also come large variation in the content of the proposals, and shareholders – particularly institutions – are still working through what is often a multi-year process to come to firm positions on some of the specifics in the proposals.

Nevertheless, the primary focus remains disclosure of political activity, with more than 50 proposals focused on disclosure. Of those, 13 resulted in voluntary agreements by the companies to begin disclosing their political activities, as summarized by the Center for Political Accountability here. Of those that went to a shareholder vote, average vote levels were above 30% – including increased voting support at 12 of the 18 companies where the resolutions were resubmitted from the prior year. For a second year in a row, one such proposal received majority support, this time at WellCare. In evaluating these numbers, a benchmark can be found in the corporate governance area, where proposals calling for elimination of takeover defenses obtained lower vote levels for many years before eventually climbing to the current majority vote norm. As long as the US Supreme Court continues to impede regulation of corporate political activity, and the Congress and the regulatory agencies continue to face partisan gridlock even as regards disclosure rules, it is likely that shareholder support for political disclosure proposals will continue to grow.

The Manhattan Institute Report and the Uncertainties of Politics

Recently, the Wall Street Journal singled out one of my research papers as worthy of attack. Summarizing a new “study” of corporate political activity by two political consultants published online by the Manhattan Institute, the Journal declared that the report “found methodological and other errors” in my research, including “problems … with causation and selection bias,” and concluded that “political spending pays,” increasing returns for shareholders “by 2% to 5% a year.”

The two consultants conducted no actual research of their own – indeed, they do not seem to have actually read my research, but instead cribbed a batch of canned econometric critiques already explicitly addressed in my papers. Their “critique” is shallow and self-serving, and both it and the Journal’s opinion relying on it miss their marks.

The MI report (which was discussed on the Forum here) is not in truth a “study.” It contains no new data or theory, and is neither detailed nor balanced. It purports to survey prior research on corporate political activity, but its presentation is highly selective. It emphasizes studies that, it claims, support its bottom line – that hiring consultants like the authors is worthwhile because it pays off for companies – and it fails to note findings that do not support this view. For example:

  • The MI report cites a highly regarded survey by Harvard Government Professor Stephen Ansolabehere [2], but not for one of Ansolabehere’s core conclusions, that numerous prior studies of corporate politics were more consistent with the view that such activity represented a form of executive consumption – a type of perk – than that such activity helps shareholders.
  • The MI report lists in its bibliography a published article by economist Jin-Hyuk Kim [3], which finds that weak shareholder rights correlate positively with corporate political activity, but they nowhere mention the Kim article in the text of their discussion.
  • The MI report nowhere cites or addresses theoretical work on political activity by Harvard Law Professor Lucian Bebchuk and Columbia Law Professor Robert Jackson, who write that “the interests of directors and executives may significantly diverge from those of shareholders with respect to political speech decisions.” [4]

Worse, the MI report contains outright and easily refuted factual misstatements about my research. For example:

  • The MI report claims that the measure of value used in my paper (available here) – Tobin’s Q – is not used in prior research as the “primary variable of interest.”
    • In fact, Tobin’s Q has been used for over 20 years in peer-reviewed finance journals. For a sampling of papers using Tobin’s Q, see articles by Andrei Shleifer (Harvard), R. Glenn Hubbard (Columbia), and Xavier Giroud (MIT).
  • The MI report claims my results “largely disappear” when I include controls for size and industry.
    • In fact, the negative relationship between corporate politics and value increases when controls are included (see Table 6 of my 2012 paper).
  • The MI report says that “the actual effect” I observe is so small it “could very well be zero.”
    • In fact, my results suggest that one can be 95% confident that the average effect of corporate politics on shareholder results lies between negative $13 billion and $195 billion.
    • The average estimate – the proper one to focus on if one wants to focus on single number – is $104 billion.
    • Is $104 billion – or even $13 billion – nearly zero?
  • The MI report claims that my results may be caused by “reverse causation.”
    • But nowhere does it discuss or address the test – set out in Table 6 and discussed at length in my paper – that is designed to rule out reverse causation.
    • In that table, I examine the effects of political activity prior to Citizens United on shareholder value after Citizens United – and unless a time machine has been invented recently, the latter cannot have caused the former.
    • Though I do not purport to have conclusively proven that politics causes lower shareholder value, the MI report nowhere provides any reason to think that my findings can be explained in any other way.

Readers should be aware that the Manhattan Institute is not, as its name might suggest, an otherwise neutral think tank. In fact, it proudly bills itself as a “market-oriented” and “conservative” think tank, and it has been funded by the David H. Koch Charitable Foundation, the Scaife Foundation, and politically active companies such as Exxon, as documented here. The authors of the MI report are self-described as the chairman and an employee of Sonecon, an economic consultancy, a private firm that advises politically active trade groups such as the US Chamber of Commerce and businesses such as AT&T.

For a clear example of how political activity can harm shareholders, you need only review AT&T’s decision last year to attempt to purchase Deutsche Telekom’s T-Mobile subsidiary. Despite AT&T’s spending more than $5 million on lobbyists and public relations, that purchase failed because (among other things) the “political climate” proved unfavorable. The result? AT&T wasted a year, spent millions in transaction costs, and paid a $3 billion reverse break fee to Deutsche Telekom to compensate it for the lost bid. Who bore those losses? AT&T’s shareholders. Any business strategy entails risks, but boards, shareholders, and managers should think hard about how to factor political risks into the financial analyses on which major business adventures are based. Unlike many risks traditionally factored into such analyses, political risks are not easily modeled, and they are not sufficiently similar over time or across industries to permit even experts to make confident predictions on outcomes.


[1] Despite their uniform policy views, the Montana court split – the dissent felt bound by the US Supreme Court’s deeply unpopular decision in Citizens United. The Montana dissent’s deference to precedent is in marked contrast to the disrespect by the US Supreme Court, in Citizens United, to two of its own precedents.
(go back)

[2] Ansolabehere, Stephen, John M. de Figueiredo, and James M. Snyder Jr. “Why Is There So Little Money in U.S. Politics?” The Journal of Economic Perspectives 17, no. 1 (winter 2003): 105–30.
(go back)

[3] Kim, Jin-Hyuk. “Corporate Lobbying Revisited.” Business and Politics 10, no. 2 (2008).
(go back)

[4] Bebchuk, Lucian, and Robert J. Jackson, Corporate Political Speech: Who Decides?, subsequently published in 124 Harvard Law Review 83-117 (November 2010).
(go back)

Both comments and trackbacks are currently closed.


  1. Eric F. Van de Velde
    Posted Tuesday, July 3, 2012 at 6:35 pm | Permalink

    If political spending is bad for companies, is it possible it is good for top executives? The shareholder money they spend gives them connections and personal influence with the power elite of the country. This cannot be bad for one’s personal career. Who cares about the fact that shareholders pay for their social climbing?

  2. John Coates
    Posted Sunday, July 8, 2012 at 7:53 am | Permalink

    Eric — Yes, I think your suggestion is likely true.

    Two other findings in my paper – available here: http://ssrn.com/abstract=1973771:

    1) More than one in ten CEOs of S&P 500 companies in 2000 had been nominated, elected or appointed to high federal office by 2010.

    2) Those CEOs’ companies were significantly more likely to be politically active than peer companies.

2 Trackbacks

  1. […] Coates (disclosure: my Corporations prof) responds forcefully to this Manhattan Institute report on the impact of corporate political activity for shareholder […]

  2. […] serve shareholder interest; the Manhattan Institute questioned his methodology in June and he fired back last week at the HLS Forum on Corporate Governance. Disclosure of corporate political involvement […]