I'll be quite open about the fact that I'm as guilty as anyone on this front, and spend quite a bit of time thinking about the best way to argue a position (or position an argument!) rather than looking for the evidence whether that position is a valid one or not. In fact, if you've read Jonathan Haidt's The Righteous Mind you'll know that there's a strong case for saying that evolution has led our minds to be better at defending positions we want to take, rather than the cool, detached we think we use to get us to particular positions in the first place.
In this battle, stats become important weapons to be deployed in order to advance, or prevent reform.
For example, here's Baroness Wilcox in the Lords recently:
We welcome the close engagement of institutional shareholders and their willingness to use their voting powers. We want this to be sustained and shall continue to monitor disclosure levels. Evidence suggests that more institutional investors are disclosing their voting records and that up to three-quarters of those investors are now disclosing their votes. We will consider further action if the number of investors volunteering to disclose their voting records does not continue to increase.
Vince Cable took the same line in the Commons and it sounds good doesn't it? With three quarters of institutional invetsors disclosing their votes there is no need for Government to blunder in and upset a welcome, voluntary trend to disclosure.
There's only one problem.The stat quoted does not reflect the true position. I know because one of the few issues where I have repeatedly gone at looked at the actual evidence is disclosure of shareholder voting records. Based on my own research I think there are approximately 50 asset managers that disclose some level of voting data, and there are, by comparison, over 170 asset manager signatories of the Stewardship Code. So at best I would say a third - of asset managers - disclose some data.
But actually that oversells the position. Because of those 50 odd disclosers only about half disclose a full record. Some only report oppose votes and abstentions (I won't repeat why this isn't enough here) and some only put headline stats up, which are basically useless. The number of asset managers disclosing a full record is less than 30. And this is after a decade of pressure for transparency.
Some will counter that most of the big managers disclose. True, but they don't all disclose a full record, and they don't all disclose in a timely fashion. Off the top of my head I can think of two very well-known asset managers for whom we still don't have votes on James Murdoch's re-election as chair of BSkyB in November. In one case the most recent data available is from last summer.
The reality is that the stat - quoted by the Government as evidence of why they don't need to make voting disclosure mandatory - is based on investment industry research. This, not surprisingly, is looking for the best possible result so will prefer to quote the number of managers who disclose anything, rather than those who disclose a full record. It is also drawn from respondents to a survey, rather than looking at all asset managers.
I don't mind the industry trying to put a gloss on disclosure levels to try and prevent mandatory disclosure - proponents of reform can generate their own stats to push the argument in the other direction. But let's be clear both that this is not untainted 'evidence' and that it is being deployed to defend a predetermined position (essentially the same as the industry's) that intervention is not required.