Thank you, Mr. Chairman. Thanks to the committee for inviting me back. It is always a pleasure to be here.
Today's topic is a fraught one, so it is important, as always, to mention that although I am working in my capacity as research director of the C.D. Howe Institute, I am speaking for myself and not necessarily the institute or its board of directors or its members, many of whom may have quite different views on income trusts.
The trust issue is a fraud. It is a little problem that grew. It was a problem that was flagged in the report of the Technical Committee on Business Taxation, a committee that reported nine years ago, having been struck by a previous finance minister and chaired by Jack Mintz.
The issue, as everyone knows, is that investors, especially tax-exempt such as pension funds and individuals through other RRSP holdings and non-resident investors, are attracted to income trusts because of their ability to use debt, or leverage, to eliminate Canadian income tax liability at the corporate level.
The technical committee highlighted potential problems with the growth of trusts and partnerships model and went on to discuss solutions employed elsewhere. The technical committee's central recommendation was a neutral tax policy toward corporate capital structures, as would be possible through a corporate distributions tax, not very different with respect to trusts from the mechanism put forward recently by the current government.
Clearly the government's decision was ultimately the right one, if late in coming and not revamping the system as much as circumstances might warrant, but it was generally the right choice. Here is why.
The impetus to choose a particular capital structure is a market distortion, and that means costs as well as benefits. Some benefits accrue to income trust unitholders, especially non-residents and tax exempts whose investments are exposed to less corporate income tax than others. The costs, however, are more diffuse and arise from the constraints the trust model imposes on a business of capital structure.
An income trust cannot grow organically through retained earnings; it can only grow by going back to capital markets, reissuing new trust units, or by borrowing. These are legitimate options, and trusts use them. However, those options are also costly. They have constraints such as that issue in new units is diluted to existing holders. Issuing new debt is costly because it raises the business's total carrying costs by raising total risk, and that also limits payouts to all unitholders. Again, those are costs that may have offsetting benefits, but do they?
For instance, do income trusts have special governance features that make them more responsive to unitholders' interests? No. Unlike common shareholders, unitholders' rights are defined only within a trust indenture, which may be written by the trustees themselves and exist entirely outside the corporate law framework.
Do income trusts, which are generally thought to be bound to a fixed stream of distributions to unitholders, do a better job of holding managers to account for financial performance? No. The board of directors of a common-share corporation could just as easily instruct management to implement a fixed, high-dividend payout policy or to leverage the business to ensure management did not overbuild in its own interests. The trust model is unnecessary for exerting that sort of management discipline.
Do income trusts bring special characteristics to capital markets, so the overall market performs better? Now, that is interesting. Real estate investment trusts, for example, make it possible for retail investors to round out their portfolios with diversified investments in commercial real estate that would not otherwise be available to them. The trust, of course, or the business benefits from its ability to attract retail investors who would not otherwise be investing in those companies.
When it comes to ordinary business income trusts, matters are different. The risks and assets they bring to the retail marketplace are no different from those available through ordinary corporate structures. Their governance does not offer an improvement over corporations. The constraints imposed by required distributions do not improve management performance in any way that could not be generated or achieved in a common-share corporation. Yet, as I explained, the trust model imposes constraints on capital structure.
The constraints imply clear costs but do not deliver clear benefits. That is why, on balance, a neutral tax policy with respect to corporate form is the right policy. The reason is that there is nothing special about trusts that would warrant tax favouritism. As l've suggested and written elsewhere, more business tax policy changes are warranted. For example, upstream taxes paid on distributions to pensions and RRSPs should be refunded to unitholders and shareholders, so that pensioners do not end up bearing more than their fair share of corporate tax. That would be an extension of current policy, not a reversal of it.
Thank you.