Institutional investors hold the overwhelming majority of shares of US publicly traded companies and routinely vote on management and shareholder proposals (contributing to some concerns about asset management concentration; see Franzoni 2019). They are able, in principle, to collectively control most publicly traded firms. These investors increasingly recognise that they need to look beyond the financial performance of the companies they invest in and ask how the company contributes to society more broadly. The CEO of BlackRock, Larry Fink, wrote in his 2019 annual letter to CEOs that “society is increasingly looking to companies, both public and private, to address pressing social and economic issues”. Engagement through proxy voting is an important way in which institutional investors can prod companies to be more socially responsible. It is therefore natural to ask what objectives institutional investors pursue when they exercise the votes attached to the shares they hold. What are the preferences of institutional investors?
Using a by now standard approach in political science, we elicit institutional investors’ preference, or ideology, based on how they vote. The W-NOMINATE method we apply maps investors’ ideology onto a left/right dimension based on how they vote. We find that most public pension funds are left-leaning and support a more social and environmentally friendly orientation of the firm. A few mutual fund families are more ‘money conscious’ investors, and lie on the right. The largest asset managers and the proxy adviser ISS (based on its voting recommendations) are placed in the centre. We also find that institutional investors’ ideology differs along a second dimension, which reflects differences in their governance stance, with management-disciplinarian investors, the proxy adviser Glass Lewis among them, at the top pitted against more management-friendly investors at the bottom.
In a new study (Bolton et al. 2019), we apply the standard spatial model from political science to elicit institutional shareholders’ ideology based on how they vote. This methodology does not take an a priori stand on the proposal attributes investors care about and is silent about their motivations. The preferences of investors are revealed entirely from their voting patterns. The main finding is that institutional investors’ ideal points map onto a line, where the far-left investors are best described as socially responsible investors (those that vote most consistently in favour of pro-social and pro-environment shareholder proposals), and the far-right investors’ votes can be described as money-conscious investors (those who oppose proposals that could financially cost shareholders). In other words, the issue that most separates institutional investors is the degree to which they weigh social responsibility. Our main finding echoes the political science findings developed from US Congress roll call voting, where legislators are aligned along a liberal-conservative spectrum (Poole and Rosenthal 1985, 2007). We show the ideological distribution of investors in Figure 1. The top panel shows the results of a one-dimensional scaling, and the bottom panel a two-dimensional scaling. Mutual funds appear as orange dots, public pension funds as blue.
Figure 1 Ideal points: One- and two-dimensional W-NOMINATE
The ideology of most pension funds is to the left, while that of the largest mutual funds is to the centre-right. The funds voting in line with the proxy adviser ISS recommendations are squarely in the centre. The other main proxy advisor, Glass Lewis, together with the main index-funds Vanguard and BlackRock, is centre-right. Along the governance dimension, Glass Lewis and pension funds tend to be more management-disciplinarian than most mutual funds. The more socially responsible large funds, such as Nuveen, PIMCO, DFA, and Grantham, Mayo, are more management-friendly, while BlackRock, Vanguard, and GAMCO are both more profit-oriented and more management-disciplinarian. Including director elections in the analysis moves investors in the management-disciplinarian direction, confirming that negative votes on directors are one of the main forms in which institutional shareholders express their dissent with management and board decisions. Whether these ideological differences reflect the ideology of their client bases we cannot say. It is not even clear that clients are aware that the funds they invest in have systematic ideological biases.
We also find that governance is a second dimension separating investors, with investors differing on how tight a discipline should be imposed on management. It sees Glass Lewis and a few public pension funds taking a tough stand on management on one side, and most of the large mutual fund families displaying a friendlier attitude on the other. Our results differ somewhat from the proxy voting literature in that we do not find that large institutions follow the proxy advisers closely.
Different types of proposals
Our methodology further generates the predicted investor voting blocs on different categories of proposals, allowing us to study how votes are affected by proposal and sponsor characteristics. We find that most proposal votes set the left funds against those that are centre or right. The exception are governance proposals, where the centre tends to side with the left, and against management, which is located near the right end of the first dimension. Along the second dimension, shareholders are divided in the middle, with the exceptions of the ‘Say on Pay’ votes, where in a significant fraction of cases the centre voted with Glass Lewis and the management-disciplinarians, and also the ‘Social’ proposals, where on the contrary Glass Lewis and the management-disciplinarians are in the same camp against the centre and the management-friendly funds. Figure 2 reports an example of predicted coalitions (top panel) and deviations (bottom panel) for the 2012 Citibank Say on Pay vote, when 55% of the shareholders voted against the compensation package. Subsequent to the vote, the CEO, Vikram Pandit, resigned.
Figure 2 Voting blocs and deviations in the Citibank ‘Say on Pay’ vote on 17 April 2012
The characteristics of the firms that the investors are voting on also matter. Higher past returns and governance variables such as board size and the fraction of independent directors result in smaller and more extreme left camps, and, all else equal, higher management support; poison pills and unequal voting rights have the opposite effect, suggesting that firms with stronger minority shareholder rights tend to be firms with a broader shareholder support of management. On the second dimension, the presence of a golden parachute, a poison pill, a classified board, a higher fraction of independent directors, a smaller board, and higher institutional ownership are associated with a larger support for the management-disciplinarian approach.
Tradeoffs between the first and second dimensions
Finally, for each proposal, we measure how investors trade off first- and second-dimension considerations. We find that most of the proposals are either purely first-dimension issues or a mix of the two dimensions, but with a greater weight on the first. Interestingly, the few proposals that give more weight to the second dimension tend to be in the Say on Pay category, although a good number of director proposals also have a strong second dimension.
The interpretation of the dimensions we found is, of course, open to discussion – as much as is the meaning of liberal and conservative in politics. The sorting on Say on Pay may reflect different beliefs about how much executive compensation contributes to shareholder returns. Alternatively, there could be agreement about what compensation maximises shareholder returns, but the left may be more open to lowering shareholder returns in ways that promote environmental and other social objectives.
Bolton, P, T Li, E Ravina, and H Rosenthal (2019), “Investor Ideology,” Journal of Financial Economics, forthcoming.
Franzoni, F (2019), “The effects of concentration in the asset management industry on stock prices”, VoxEU.org, 3 June.
Poole, K, and H Rosenthal (1985), “A spatial model for legislative roll call analysis,” American Journal of Political Science, 29, 357-384.
Poole, K, and H Rosenthal (2007), Ideology and Congress (2nd edition of Congress: A political economic history of roll call voting, 1997). Transaction Press.