Bear with me. I'm going to discuss the work of the Auditor General without any reference to a politically damaging scandal. Yes, it will be boring. But it's important. And I promise to talk about the gun registry just to give everybody something to disagree about. The role of the auditor general is simple: to make sure tax dollars are used as Parliament intends. Mostly, that means figuring out who spent what money where. But the AG is also required to report to Parliament if "satisfactory procedures have not been established to measure and report the effectiveness of programs." Note the AG doesn't do the measuring and reporting herself. She simply checks to see if it's being done. Yes, I know. Boring. But this is important. Governments don't create programs in order to keep bureaucrats busy. They do it to achieve specified goals. We have to check to see if those goals are being met. When they are, the program works and should be kept or expanded; when they are not, the program doesn't work and it should be modified or scrapped. Anyone object so far? No. Of course not. Everyone says public policies should be based on evidence, not hunches, politics, or ideologies. And if you don't check the evidence, you don't have evidence-based policies. Which brings me to the long-gun registry. We've talked about it for years and years. Most people have an opinion. Most people are sure their opinion is supported by the facts. And most people are sure those who disagree with them are ignorant, deluded, or dishonest. And what are the facts everyone is sure support their side? Incredibly, on the day MPs narrowly voted to keep the registry alive, various media outlets -- this newspaper included -- argued about the annual operating cost of the registry. No fact is more basic. No fact is simpler to measure. No fact is more necessary in judging whether the registry should be maintained. And yet, after years of debate, it still wasn't nailed down and commonly agreed on. But the truly amazing thing is that as poorly informed as the registry debate is, it is actually better-informed than many. After all, the RCMP did conduct an internal evaluation of the Canadian Firearms Program -- of which the long-gun registry is a part -- and this evaluation is a public document. True, the government suppressed the evaluation for many months. True, it is horribly written and confusing. True, it's not easy to find on the Internet. But it exists. And it provides some reasonably objective information on which one can base one's views. If one is so inclined. That's often not the case with important matters of public policy. Sometimes internal program evaluations aren't done. Or the terms are narrowly circumscribed. Or the results are fudged. Or the evaluation is quietly slipped into the desk drawer of the deputy minister and never seen again. No one likes to be told the programs they oversee are useless, after all, and it's even less pleasant to have the bad news publicized. Even a public rebuke by the Auditor General may not be enough to change this. In 2001, for example, the AG revealed that the federal government didn't really know how much money it spends on its illicit drug policies, it didn't really know how big the drug problem is, it hadn't set measurable goals, and it wasn't conducting the research that would be needed to determine if it was meeting its goals, if it had any. Pretty horrible stuff. So what happened? Not much. The situation today isn't a whole lot different: The government continues to spend enormous amounts of money without really knowing whether that money is accomplishing what it's supposed to. Things are done a little differently in the United States. The GAO -- Government Accountability Office -- is the almost-90-year-old investigative arm of Congress. It's fully independent. It has a staff of 3,100 and a budget of $538 million (U.S.). Its work is rigorous and universally respected. All of which sounds like Canada's AG. But there's one big difference. The GAO not only tracks money. It evaluates programs to determine if they do what they are supposed to do. The GAO's reports are crisp, clear, and comprehensive. And they're all available at a single website (gao.gov) where anyone can type a search word and instantly get everything the GAO has ever said about the subject. In Ottawa, a group of mainly retired civil servants has proposed the creation of an "Evaluator General" to do what the Auditor General does not (see their website at evaluatorgeneral.ca). I don't know whether that's necessary, or whether it would be better to expand the AG's office into a Canadian GAO. I also don't know how much this would improve the quality of discussions about public policy. Human biases are tenacious and not even the finest research will penetrate the skulls of partisans and zealots -- as American policy debates demonstrate with depressing frequency. But not all of us are partisans and zealots, fortunately. Many of us are rational. We would like public debates to be better informed. And we would like "evidence-based policy" to be something more than a slogan. For there to be any chance of that happening, we need something like a Canadian GAO.