The other throne speech

Another interesting piece from the Cato Institute.

Canadian election results

So the election results are in. Of the 308 parliament seats at stake, the Conservatives won the largest number, 124, so they will form a minority government. How do the seat counts compare with the popular vote? Well the two leading parties won a somewhat larger share of the seats than their share of the vote. The Conservatives won 40.3% of the seats with 36.3% of the vote, while the Liberals won 33.4% of the seats with 30.2% of the vote. On the other hand, the NDP got just 9.4% of the seats even though they had 17.5% of the vote. The Bloc Québécois benefited the most from our first-past-the-post system, winning 16.6% of the seats with just 10.5% of the votes. And the Green Party won no seats at all, even though they got 4.5% of the vote!

I haven't looked at the breakdown of popular vote here in the province of Ontario, but here's a barchart from today's issue of the newspaper Dose. Although the title suggests that it shows popular vote, it actually shows the number of ridings in which each party was leading or elected at press time (presumably late last night). In addition to the poorly-chosen title, the graph is a disaster. Although the count for "Other" is 0, the bar looks like it has a height of 1. There's the usual unnecessary 3D effect. But one of the worst aspects is the way the party names have been placed in shaded bars underneath the bars themselves! One's eye tends to stack each data bar with its title bar, so the visual effect distorts the numbers. Sheesh!

Ethical investing

Like most people, I have investments, largely in the form of mutual funds. So that means I'm supporting and profiting (at least in the long run) from the companies whose stocks make up the funds. And which companies are those? I haven't got a clue. But what if some of them are tobacco companies? How about companies that are major polluters? Companies that don't treat their workers well (unsafe working conditions, unions not allowed, etc.)? Companies that manufacture weapons or--and this is trickier--weapons components? Obviously the list could go on and on ...

So what to do? Three choices I can think of are (1) don't invest; (2) invest to maximize profits (subject to an acceptable level of risk) and then donate some of the profits to worthy causes; and (3) invest "ethically";

Option (1) doesn't seem sensible. First of all, I don't think I have an ethical problem with investing per se. In fact it seems like a good thing to me. For example, venture capital helps support innovation. Second, I have to plan for the future (my kids' education, my retirement, etc.). Third, are there any good alternatives to investing? For the time being, let's move on to the other two options.

Option (2) is straightforward. Follow conventional financial wisdom to trade off profits and risks. As I noted above, this inevitably means that I'm to some extent complicit in the practices of the companies in which I'm investing. If I contribute some of my profits to worthy organizations, can I compensate for these harms? For example, if some of my profits come at the cost of polluting the environment, can contributions to environmental organizations make up for the damage done?

Option (3) is to invest ethically. For example, I might avoid investing in tobacco companies, companies that I judge to be egregious polluters, and weapons manufacturers. But how am I to do this? I'm not sure I have the time and energy to research the practices of various companies on an ongoing basis. An alternative is to purchase "ethical" mutual funds. A number of such funds exist, with different definitions of what types of practices are unethical. An obvious problem is that "ethical" mutual funds are likely to underperform. Naturally, if ethical considerations are ignored, there's a broader choice of companies in which to invest, and a competent fund manager ought to achieve higher profits. Compounding this, is the issue of management fees. Someone has to do the ongoing research to judge which companies have ethical practices, and this comes with a cost. I've been told that, indeed, "ethical" mutual funds generally don't perform as well as other funds, delivering a long-term absolute rate of return perhaps 2% lower than other funds.

So why not return to option (2), maximize profits, and contribute each year's profit differential to worthy causes? (This would be over and above my normal charitable contributions.) This seems like perhaps the best course of action, except that I have the nagging feeling that it's a cop-out.

So I appeal to you, gentle reader, to disabuse me of my misconceptions or point out what I've overlooked. Can anyone see a way out?

Egocasting?

I just uncovered an interesting tidbit: Gilles Duceppe's disparaging comments about statisticians were pre-planned--see the update to my earlier post. And yet:

"The sad thing about relying on blogs for information about political choices is that they are high on opinion and low on fact and preach to the choir rather than address issues factually, substantively and inclusively."

So says Michael Bugeja, a professor (of journalism, apparently) at Iowa State, as quoted today in an unperceptive Globe and Mail article titled "Wired up, plugged in, zoned out". Bugeja goes on to say that

"We need print newspapers to do that [address issues factually, substantively and inclusively] because the issues requiring factual analysis are more complex than ever and cannot be stated simply in a TV sound bite or Internet news brief."

Huh? I thought we were talking about blogs, not TV sound bites.

Ah, yes, God forbid anyone rely on blogs and news via the Internet. We need print newspapers that publish reliable stuff. Like this:

"In the current Canadian election, more and more people are getting their news via blogs and subscriptions to Web services that align with their own beliefs, so they hear few opposing viewpoints."

Um, would it be too much to ask for some evidence that this is the case? I'm not disputing the claim that people are increasingly turning to the Internet for news and opinions. But are they thereby filtering out opposing views?

The author of the Globe and Mail article, Tralee Pearce, writes that

"The filters and search devices used to make all this information manageable, [many observers say], are isolating people into niches fashioned to their particular tastes and beliefs. Instead of going to common sources, whether newspapers or broadcast TV, to get the daily news, users are getting only the 'daily me'."

Pearce notes that American historian Christine Rosen has termed this phenomenon "egocasting". Well there's a buzzword. Never mind that it reverses the roles of the producer and the consumer of information.

But more to the point, is it true? One of the things I find quite wonderful about the Internet is that it makes it so easy to access such a broad variety of information and viewpoints. And technology for web syndication like RSS makes this much easier. But perhaps most people just stick to what they feel comfortable and familiar with.

Of course, this sort of narrow filtering could never happen with the print media. First, the major newspapers present such a broad array of viewpoints, you could never accuse them of filtering the content. Second, it's not as if most people only read newspapers and magazines that tend to agree with their own beliefs.

Propaganda and truth

Since I've been thinking a bit about propaganda, let me share a thought. For some time now I've been bothered by public health campaigns (or other education campaigns) that don't tell the truth. One example (suggested by my sister-in-law) is the much-mocked "This is your brain on drugs" tv ad. It's a striking visual metaphor: cracking an egg onto a hot frying pan. But is it truthful? It refers to "drugs" (presumably illicit), but which ones? What's the evidence that they "fry" your brain? Perhaps your brain has to already be fried before you'd consider using them. Or perhaps some of the drugs actually enhance your brain. But I don't think that truthfulness was the point of the ad. It was meant to convince, not to genuinely inform. And how do you convince people? By using the techniques perfected by Joseph Goebbels, the Nazi Minister of Propaganda. Ok, maybe that's a cheap shot, but I think propaganda is a dirty business.

As a matter of principle, I think that public health campaigns should tell people the truth. I remember with HIV/AIDS the "safe sex" slogans that eventually gave way to the more honest term "safer sex". From a pragmatic standpoint, if you tell people something, and then a little later admit that "maybe that wasn't quite true", you risk losing credibility. Sometimes you'll hear someone say something condescending like "people can't handle the truth." What's that supposed to mean?

One objection that may be raised is that you have to simplify things to communicate with the public. I don't dispute that, but you can still tell the truth. If you need to simplify something you can provide an indication that that's what you've done (and perhaps suggest where to get more details).

Another objection might center on the word "truth", which of course opens up a whole philosophical can of worms. When Jesus used the word, Pontius Pilate replied "What is truth?" But it wasn't an honest question, it was jaded cynicism. Without denying the complexity of the question (when asked honestly), I think it can be provisionally set aside by saying that "the truth" means what one honestly believes to be true. I don't deny that that sounds circular.

Advertising is an inevitable aspect of the marketplace. But the marketplace shouldn't rule supreme. I can accept that marketers will influence my choice of shampoo, but I don't think they should be involved in the content of public health campaigns!

"We don't need more statisticians."

I'm watching the Canadian leaders debate, and just heard Bloc Québécois leader Gilles Duceppe say "We don't need more statisticians." Of course I can't let that pass without comment!

First, the context. The debate was focusing on healthcare. Healthcare in Canada is governed by the Canada Health Act, through which the federal government ensures that the provinces and territories meet certain requirements, such as free and universal access to insured health care. (That description of the Act is taken from this helpful overview.)

Duceppe noted that under Liberal governments, the number of federal government employees in healthcare-related areas (presumably in Health Canada and the Public Health Agency of Canada) has increased considerably. But delivery of healthcare (including paying for nurses and doctors) is a provincial responsibility. Duceppe argued that what is needed is more nurses and doctors, not more federal employees such as ... statisticians (among other examples he mentioned).

Now, it is widely accepted that more doctors and nurses are needed, and I agree. But I am dismayed that Duceppe should so casually dismiss the value of federal health employees. The recently-established Public Health Agency of Canada seems particularly relevant given the current concerns about avian flu.

Duceppe's disparaging mention of statisticians is, of course, particularly galling to me. Sure, if you go to hospital you won't be treated by a statistician. But that doesn't mean you're nobenefitingng from the work statisticians do. For example, statisticians play a key role in the Cochrane Collaboration, which is an important part of the movement towards evidence-based healthcare. For a stimulating account of statistics in medicine, see Stephen Senn's book "Dicing with Death: Chance, Risk and Health" (which I reviewed for CMAJ).

I often find Gilles Duceppe's comments to be quite insightful, but he was way off the mark this time.

Update 14Jan2006: According to the transcript of the debate, I didn't get the quote quite right. Here's the relevant excerpt (the italics are mine):

Moderator: "In the area of social policy, let's go to health care, which, of course, many Canadians want to know a lot about, and Mr. Duceppe, this question is for you. Provinces including Quebec have increased the role of the private sector in the delivery of health care. Would you support a federal ban on private for-profit health clinics?"

Gilles Duceppe: "I mean, I don't want people to have to use their credit card to go to hospital, but the thing is, it is a provincial jurisdiction. It is a Quebec jurisdiction, and a main problem is the fact that the federal government cut in the payment transfers for health since 1994. Recently, they put more money, but not at the level it was in 1994. Just consider that in Ottawa, in the health department, there's 10,000 civil servants, and they're not managing a single hospital. Five years ago, there were 506 employees for promotion, like against cancer or smoking, thing like that. Today nowadays, we have 4,561. We don't need inspectors. We don't need statisticians, we need doctors and nurses. I'm confident the province of Quebec will be able to settle that problem and make sure you go with your health card and not your credit card."

But here's something very interesting I stumbled on. The snappy line actually dates back about 7 years! Here's an excerpt from Hansard 171 of the 36th parliament, 1st session, February 1st, 1999 (again, the italics are mine):

Mrs. Pauline Picard (Drummond, BQ): "Mr. Speaker, the intentions of the federal Minister of Health are clear. He wants to control the provinces' exercise of their constitutional rights with respect to health. Will this federal interference in health matters not mean more public servants, statisticians and inspectors, rather than more doctors, nurses and clinical staff, which is what the public really wants?"

Hon. Stéphane Dion (President of the Queen's Privy Council for Canada and Minister of Intergovernmental Affairs, Lib.): "Mr. Speaker, as always, the Government of Canada intends to respect the Constitution fully."

Some hon. members: Oh, oh.

Hon. Stéphane Dion: We intend to work with the provinces in a spirit of respect for the Constitution as it applies to the health sector, just as a certain federal minister wanted to work with the provinces in the education sector 10 years ago.

I only happened to find this because I did a Google search on duceppe statisticians. And why am I bothering to point this out? Well, it clearly shows that Duceppe's words were pre-planned. Not exactly a shocker: we all know these things are carefully scripted. But the original statement by Pauline Picard was pretty ignorant and I don't think serving it up a second time makes the Bloc Québécois look too good!

I'd also like to note Mrs. Picard's words "... rather than more doctors, nurses and clinical staff, which is what the public really wants". Well even if that is true, it sounds like we're pandering just a little, doesn't it? Allow me to suggest an analogy: suppose you walk into your doctor's office and demand a prescription for an antibiotic. Does that mean you should automatically get it? Could it be that what you want isn't necessarily what you need?

The other PR

PR can mean public relations or proportional representation. I suggest that we reserve PR for the latter, and abandon the term "public relations" in favour of the more honest descriptor propaganda. (Incidentally, I'm planning to go the Canadian War Museum tomorrow to see an exhibit called "Weapons of Mass Dissemination: The Propaganda of War". It was developed by the Wolfsonian-Florida International University and they have a virtual exhibition you can visit. Update 9Jan2006: As part of the exhibition, the Canadian War Museum also has a really good section on their website about Canadian Wartime Propaganda.)

Now, back to proportional representation, which is getting lots of media play these days. John Ibbitson recently wrote an interesting piece in the Globe & Mail titled "PR: Democracy you can really trust". (I would link to the online copy, but it's only available to subscribers.) He writes:

"The argument most often put forward by detractors of PR is that it will lead to unstable Parliaments in which larger parties are held hostage to the agendas of smaller, special-interest parties, leading to repeated political crises and frequent elections."

But ironically, Ibbitson points out, the opposite is true (at least currently in Canada):

"Moving from first-past-the-post to proportional representation would actually make the House of Commons more stable."

Based roughly on recent polls, he shows that a likely outcome under PR would be "a stable coalition of the Liberals and the NDP." And he argues that "a PR-based House would also be far more regionally representative." Finally he suggests, as I have, that PR "tends to improve voter turnout, and it more closely represents the popular will."

Today's Globe & Mail online poll asked readers

"If the Green Party, for example, received 5 per cent of the popular vote in the federal election should it also receive 5 per cent of the seats in the House of Commons?"

Now I'm pretty skeptical about polls like this that aren't based on random samples. For what it's worth, 57% of the 21075 responses were "Yes". (Confession: one of those responses was mine.)

Joe Dawson pointed me to a Canadian Press piece that raises the spectre of strategic voting:

"Stakes are high as polling suggests Martin and Conservative Leader Stephen Harper are in a statistical dead heat, leading observers to warn an ugly fight for dominance lies just ahead. And that means NDP supporters who fear a Conservative win could become crucial in the Jan. 23 vote as some consider shifting alliances to tip the balance in the Liberals' favour, says the latest Decima Research survey."

Former NDP leader Ed Broadbent recently discussed specific proposals for proportional representation in an Edmonton Journal article titled "A democratic idea whose time has come". Definitely worth reading.

Finally, Ray Deonandan pointed me to an interesting website whose mission is

"to protect and improve Canadian sovereignties and democracy through education, dialogue, and advocacy, especially using existing and emerging communications tools."

They go on to say:

"We believe that Canada should maintain a position as a responsible and involved global citizen, and we support the continued development of a socially progressive Canadian society. We remain committed to the value of democratic decision-making, the value of citizen's rights over those of corporations, the value of reciprocal solidarity with other self-determination movements and/or organizations around the world, and the value of cultural diversity."

Amen.

Names, damned names, and statistics

When I tell people that I'm a statistician, the usual response is a blank stare. Explaining that I work with statistics only makes matters worse. Those who have been exposed to statistics at university often blurt out "I had to take a statistics course -- and I hated it!" What they remember of the course is mostly that it was boring and there were a lot of formulas. Those who have had no formal exposure to statistics seem to think it might have to do with collecting and tabulating figures, like sports statistics or national economic figures. This isn't completely off the mark, but by itself it's a poor description of what statisticians do.

And about half the time, mention of "statistics" elicits the helpful response: "There are lies, damned lies, and statistics!" (Often attributed to Mark Twain, but apparently originally from Benjamin Disraeli.) Something of a variant on this is the claim that "You can prove anything with statistics!"

Clearly there are several issues at play here, including minimal public knowledge of what the field of statistics is about, poorly taught statistics courses, and prejudices about empirical reasoning. Statisticians must accept a good part of the blame for each of these. (John Nelder writes [1] that "Almost nobody knows what statisticians do, and we in turn have been remarkably ineffective in explaining to non-statisticians what we are good at.") But part of the problem is the word "statistics" and its difficult-to-pronounce-and-spell sibling "statistician".

A statistic is a function of a set of observations, for example the total, the average value, the maximum value, or what have you. Governments have always wanted to keep track of information about the state (like births and deaths, imports and exports, agricultural production, etc.), which is where the word statistic comes from.

"Statistics" means more than one statistic, but confusingly it also refers to the study of how to draw conclusions from observations. A more formal term for this is inductive inference, to be contrasted with deductive inference. Deductive inference (or simply deduction) is classical logic: when the premises are true and the argument is valid, the conclusion must be true. If all swans are white, and Tom is a swan, then Tom is white. Inductive inference (or induction) is not so simple. Suppose we observe 100 hundred swans and they are all white. We might conclude that all swans are white. But this conclusion might be incorrect. (Apparently there are black swans, by the way.) Uncertainty is inevitable: for example in political polling, the stock phrase is that the results are accurate to within plus or minus 3%, 19 times out of 20. Uncertainty is inevitable because of the variability that we find everywhere: political opinions vary, height and weight differ, some people are more susceptible to certain diseases than others (perhaps due to differences in genetics, among other things). When we try to measure something accurately several times, we get slightly different answers. This is sometimes called measurement error, or noise, but in a sense it's just another source of variability. Probability theory lets us describe variability. For example, if we toss a fair coin 4 times, the probability of getting 4 heads is one sixteenth. But statistical inference uses probability theory to deal with the inverse problem: if we toss a coin 4 times and it comes up heads each time, can we conclude that it's not a fair coin?

Given that statistics has such broad relevance, it's a shame that it has been saddled with such a poor name. If "a rose by any other name would smell as sweet", I'm hoping that statistics by another name will smell sweeter!

Bill Cleveland suggests the name data science. John Nelder suggests "statistical science" [1]. And a friend of mine suggests, tongue-in-cheek, that statisticians could be called "noise-busters".

Of the above suggestions, my preference would be "statistical science", so that a statistician would be a "statistical scientist". But maybe there's a better name out there somewhere ...

[1] Nelder J.A. From statistics to statistical science. The Statistician. Vol. 48, No. 2 (1999), 257-269.