I think the question was about good ones and bad ones.
I mentioned the unsolicited proposal program. In those days, the purchasing agent was DSS, the Department of Supply and Services. You could make a proposal to DSS and they would shop it around the departments, asking, “Does somebody want to buy one of these things?” Sometimes they would and sometimes they wouldn't. We built our first ground station that way. We then dominated the world market.
The IRAP has been a granting program over the years, and it has been by far the best one. It's getting ruined now by the accountants. It's still on, but it's a mere shadow of its former self. The beauty of that program through the years that NRC ran it was that the judgments were made by scientists.
With other types of granting programs, there's so much bureaucracy to try to make sure the government doesn't get cheated that they're not efficient. A small company can't use them. They're too expensive. You can fool an accountant any day of the week, but you can't fool a good scientist, and it was the fact that the judgments were made by competent scientists.
I think the principle of contracting out research is a very important one. That develops the knowledge base close to the wealth generation. Government is a wealth-consuming organization. The universities, through their students, and industry are wealth generators.
To the whole idea of procuring R and D, I worked on a study in 1984. It was led by Doug Wright, who at the time was president of the University of Waterloo. We concluded in that study that there were only two roles of science in government. One was to maintain the regulatory knowledge base necessary in a department—and I'm thinking of bad mussels in P.E.I. and stuff like that with Health Canada—to maintain the department's competence as a smart buyer. Doing a lot of academic R and D in government really didn't have a place. Of course that is now collecting dust, and has been doing so for some time.