Because of my interest in governance/ownership issues I'm thinking about equities here. Share purchases are already subject to stamp duty, yet this hasn't stopped the level of trading gradually increasing and with it the costs for those on whose behalf it is undertaken. And this hasn't gone unnoticed - Warren Buffett talked a few years back about a new law of motion - as motion increases, returns decrease. And as Watson Wyatt put it in its Remapping Our Investment World report that the increase in trading "has enriched the broking community and impoverished the average pension fund".
But would, for example, whacking up stamp duty shares even further actually do any good? It would likely only make a difference if the impact was definitely felt, but felt by who? Most funds obviously delegate investment management to fund managers. But fund managers aren't going to shoulder the cost, they will just pass it on to the client. Thus it would ultimately sting pension funds and others, not the intermediaries who are doing the trading.
Another not entirely rubbish argument is that if increased taxes made trading in equities prohibitive, this might encourage investors to pile more into derivatives, specifically contracts for difference.
Of course there are good arguments against both of these points. The increased cost of trading is the point of course - it's intended to hurt in order to encourage lower levels of trading. And if trading levels fall far enough, pension funds might actually end up better off. And we're talking about the secondary market here really - as the Berle and Means quote below makes clear, this isn't affecting the allocation of capital to businesses. Similarly if you're bothered by a flight to CFDs then presumably these could be taxed too so that they don't become too relatively attractive.
But I can't help thinking that if this issue is already well-known to investment consultants, for example, that there must be better ways to address it. Why can't funds put turnover limits on their portfolios, or make other changes to mandate design? Or why not focus trustees' attention more directly on investment costs as a key part of their duties? If Watsons are right and fees have gone up 50% in five years you would think that the funds' self-interest ought to kick in, yet it doesn't seem to.
This does make me wonder sometimes about the relative importance of the principal-agent problem in the trustee-fund manager relationship (and the beneficiary-trustee one) versus the fund manager-company one. The principals in the first case have (most of the time) far more power over the agent than in the second case, yet seem to rarely exercise it effectively. And we seem to spend a lot more time focusing on the second one.
But if we got trustees, for example, thinking more about the costs they are incurring they could work with their agents to reduce them without the need to turn to tax. It seems at present that (as Gillian Tett suggested in the Prospect interview) the real problem is getting the agents to think about their own financial self-interest and act on it.