William Watson (an economics professor from McGill) argued against the OPBO in the Post last week. He makes two main arguments. First, that any Finance Minister who 'fudged' the numbers would pay a political price from voters. Second, that the OPBO likely doesn't have any particular advantage in making macro forecasts.
It's good to have a public discussion on the role of the office and I'm glad that Bill has contributed.
On the first point, I agree that discipline from voters is the ultimate check on bad behaviour by politicians. However, I wouldn't want to rely on it as the only one. Myself, I think well-designed institutions (such as the OPBO) can also contribute to improving the incentives and ultimately behaviour of politicians.
On the second point, I agree that the OPBO likely has little advantage over others in making (or re-weighting as Bill suggests) macro forecasts. However, that is not the only thing that the OPBO does or could do. Taking macro forecasts and making fiscal projections from those forecasts seems like something that would pay low private returns--I don't see a large incentive for private institutions to put a lot of effort into it. Moreover, beyond fiscal forecasts, the OPBO is charged with pricing policy initiatives. (See, for example, this pdf released today from the CBO in the US.) I don't see non-government groups of economists having the expertise, the data access, or the incentives to do this kind of work.
Myself, I think it's always an interesting and useful question to ask whether the private sector could replace a particular public sector role, and to think carefully about why or why not. In this case, however, I think that the 'public good' aspects of providing credible, unbiased fiscal information make the laissez-faire argument difficult.