The Census Flap illustrates that history professors are useful to society

28 07 2010

John Ibbitson has a good article in the Globe about the Canadian census controversy. The dispute over whether participation in the 2011 Canadian census should be mandatory or voluntary has got me thinking about the place of social science in Canada and about the curriculum of university history departments.

Let me state that the Conservatives’ complaint that the current census questionnaire is morally illegitimate is especially curious, given the Conservative governments were in charge when the censuses of 1991 and 2006 were taken! If the mandatory long-form census is so terrible, why didn’t they make it voluntary in 2006? Why is the Tory 2008 party platform silent on the issue of the census?

Ibbitson calls the showdown over the census a “manufactured crisis” and suggests that it had been cooked up by brilliant political strategists. I believe in Occam’s Razor, so I would tend to regard the crisis as something caused by simple lack of education. I find that ignorance of how things like statistics and the census work is common among people who lack much in the way of formal education.

I once had a summer job cleaning and driving automobiles. One of my colleagues was a nice guy who had left school at 16. I recall a conversation in which he insisted that the ninth digit of the alphanumeric serial number on Ontario driver’s licences denoted the driver’s sexual orientation. Apparently, an even number indicates heterosexuality. I tried to remonstrate and explain that this was all nonsense, but to no avail.

Now one of the things the politicians who are hysterically opposed to the mandatory census have in common is that they did not attend or complete university. They just don’t know anything about economics, stats, data systems, management, or history or the other boring things the professors drone on about.

Consider Stockwell Day. I have no doubt that Stockwell Day is a decent person. I am told that he has great people skills and is a kind employer to his secretaries and other immediate subordinates. He is a nice guy when you meet him in the flesh. The problem is that Day is now called upon to pronounce on a complex system he doesn’t appear to understand. Some of his comments about the census are ridiculous.

Mr Day recently had this to say about the current legislation, which gives Statistics Canada a highly theoretical right to fine or imprison people who do not complete the census: “Do you think it is right that you can threaten your neighbour with jail time if she doesn’t tell you if she has mental issues or not?” he wrote. “Or who does what chores in the house? Or whether she is a Jew or not? Don’t you find that one even a little bit chilling?”

Mr Day is here alluding to the fact that the Canadian census asks people to declare their religion. Questions about religion are nothing new– they were asked when the census was taken in 1851, 1861, 1871. This question is largely a relic of the days of when we were a more religious society and when sectarianism really mattered. Today fewer people attend church, so the importance of the Catholic-Protestant split had faded. But given that some religious groups still have special privileges under the Canadian constitution, it is useful for the government, municipal planners, etc, to know how many people of each religion there are in Canada. Census data about the proportion of Catholics in the population is, in fact, used to plan school capacity in the fast growing suburbs of Toronto.  How much land should we set aside for the Catholic high school in this new subdivision?

The fact that a minister of the Crown would suggest that previous Canadian governments have collected the census for nefarious or “chilling” purposes is outrageous not only because it destroys faith in the Canadian government but because it so factually inaccurate.  There is nothing chilling about, say, the 1981 Canadian census. Its creators were not trying to turn Canada into a Benthamite panopticon or lay the groundwork for a dictatorship of either the extreme right or the extreme left. They just wanted to know what percentage of the population had a fridge.

Day`s insinuation that census data about religion and ethnicity is collected for  sinister purposes is the type of claim I would expect from perhaps one of the more radical Marxist sociologists, not a mainstream politician. Day appears to be hinting that a future dictator might use census data to persecute minorities and that we are living in some sort of Weimar Republic. This is sheer alarmism that will feed the paranoia of every nut out there. Day`s comment is about as credible as the leftwing grad student I once heard comparing the Beavers, the most junior branch of Scouting, to the Hitler Youth. Are not Beavers sometimes transported on highways that are very similar to the Autobahn of the Weimar Republic?

Of course, university-educated people who were born in Canada and who read newspapers will likely recognize Day`s absurd statement for what it is. They know that the Canadian government, which includes a vigorous privacy commissioner, is one of the most data-trustworthy governments in the world.

The danger is that people who are on the margins of our society culturally and economically, such as semi-literate native-born Canadians or refugees from dictatorships, might very well buy into Day`s alarmism and refuse to complete the 2011 census. Friends who have worked as census takers tell me that immigrants from repressive regimes are very reluctant to provide data.  In many parts of the world, the conspiracy theories about the government are sometimes true. As a minister of the Crown, Mr Day has an obligation to put their fears to rest, not to stoke them. Mr Day’s comments are extremely destructive of the levels of social trust that have hitherto characterized Canadian society. He is essentially inviting new Canadians and members of the non-voting classes in society to distrust the State and to fear their neighbour. That will have big implications for future police officers, game officers, teachers, etc., not just for the census officials who go door to door with clipboards. Day`s comments remind me of the Tea Party people in the US who are trying to delegitimize their own government in the eyes of their fellow taxpayers, which is a surefire recipe for increased tax evasion.

It should be stressed that the Canadian government uses the census to produce aggregate data about communities. I know that the word aggregate is a long one with many syllables,  I`m sorry about that, but the concept is an important one. As far as I know, it has never used the census database to identify specific individuals for the purposes of law enforcement, taxation, or military conscription. [If I am wrong about this, please correct me]. Conscription in the First World War was based on a questionnaire that was mailed to all adult males in late 1917. Many of these forms were not returned on time.  When immigrants from Eastern Europe began settling in Canada in the early 20th century, many of them were reluctant to complete the census. Unlike earlier waves of immigrants from Britain and the United States, the Tsar’s former subjects were understandably distrustful of the State. In the lead up to the censuses of 1901 and 1911, Laurier tried to defuse their concerns by explaining what the census is actually used for in a democratic country. We want to find out how much wheat the average farmer produces. We don’t want to steal your men folk. That was his message.

Another thing that needs to be explained is that individual data only becomes available to the public decades later, long after you are dead. You are allowed to look at individual data from the 1850 census for Springfield, Illinois.(see below for the entry for Abraham Lincoln`s household. The data from the 1950 census is still protected.

If you look at the data that has been published over the years by the Dominion Bureau of Statistics/Statistics Canada, you will see the numbers of members of very small ethnic and religious minorities are always rounded up to the nearest ten so that specific individuals cannot be identified. This means that if one person in Iqaluit told the census taker they are Jewish, the Community Profile will say that there are precisely 10 Jews in that census district. This suspiciously round number is designed to protect the privacy of individuals. Rounding up from 1 to 10 in each community probably results in a slight over-estimate of the number of members of very small minorities in the national population.

Even in the bad old days when Canada had explicitly racist laws such as the Chinese Head Tax on the books, the Canadian government has never used census data to track down specific individuals. To have done so would have been to destroy trust in the confidentiality of next census. The government did not use the results of the 1941 census to identify the addresses of Japanese Canadians. It used other records to round them up.

I don’t like to attack specific individuals on my blog, but it should be pointed out that Mr Day lacks a university degree. See here. When I was an undergraduate, I did coursework that involved census data. I also studied quantitative approaches to social research, how to run SPSS on a computer, the use and abuse of numbers in history, etc. Most readers of this blog probably had similar training in numbers, even if they have now half forgotten everything they learned about dependent variables and sample sizes.  I suspect that Day would have benefitted from that sort of training.  Had he had similar opportunities, he likely would have had a better understanding of the census.

Sometime professors in the social sciences and the humanities wonder whether the teaching components of their jobs are worthwhile. Some academics are plagued by self-doubt as to whether their teaching work actually benefits society in any way. After all, we aren`t teaching people how to treat cancer or to inspect aircraft engines for dangerous faults.  Stockwell Day’s absurd comments show why social science teaching is valuable to society. If all social scientists stopped teaching university students, Stockwell Day level ignorance and alarmism would be far more common in our society.

This is one of the reasons I am passionate in my belief that history and other disciplines needs to become more quantitative, not less so. I believe that attacking statistical illiteracy through education will improve society in the long run, since it will encourage people to think more rationally. Of course, narrative political history will probably always be the heart of historical studies, but we do need to also teach students more about ways of studying masses of people. Implementing solutions to environmental and other social problems requires an electorate that can understand numbers and abstract concepts, who can connect the grams of CO2 they produce each time they drive to the store with the megatonnes of CO2 our society emits each year.

Unfortunately, the quality of training in quantitative methods that liberal-arts students receive varies considerably from university to university. The census controversy has got me thinking about how I can do a better way of teaching undergraduates about quantitative as well as qualitative methods to historical research.

I am thinking of teaching a class on quantitative methods for historians. The textbook would be History By Numbers, which was written by Pat Hudson, a social historian who just retired from Cardiff University. The students would also read George Emery`s book The Facts of Life. Emery was the chair of grad studies in the history department at Western when I was working on my PhD. Students would work with census, OECD, Angus Madisson`s stuff, and other data. They would learn different ways of converting historic monetary values into present-day values. The students would also be exposed to the literature on the limitations of census data, such the great article by Bruce Curtis on the totally botched census of 1861, which produced crap data– among other problems, census day in 1861 was in the middle of the winter, which meant the ink of many census officials was frozen. The course would also talk about the social construction of quantitative data and the political history (e.g., the 1860s campaign for Rep by Pop in Upper Canada, which was informed by data from the garbage 1861 census). Students would also learn about the history of statistics methods with a focus on the Scottish Enlightenment.

Let me say that as social scientist (i.e., as someone who elected to do a PhD in a history department located in a faculty of social science rather than a history department in a faculty of humanities), I can only say that I am appalled by the sheer incompetence and stupidity that some members of the Harper cabinet have displayed on this issue. Mr Harper has an undergraduate degree in economics and then did an MA thesis in political science (see here), so one might have thought that members of his cabinet would understand the importance of gathering relatively hard quantitative data. When Harper became Prime Minister, I hope that a new era of evidence-based public policy grounded in hard social science might be upon us.  One might have thought that university graduates would understand why a mandatory census would produce better data than a purely voluntary one. The Tory proposal to make participation in the census voluntary will skew the results of the census, since the self-selected civic minded folk who actually complete the forms are unlikely to be representative of the general population. Their profile is more likely to be like that of the average voter—older and wealthier than the average adult.

Don`t get me wrong. The Canadian census in its current form may or may not be the best way of gathering data. If Stockwell Day were proposing to scrap the data on cost grounds, I would listen to him respectfully.  The United Kingdom is planning to abolish its census on the grounds that this form of data collection is obsolete in the internet age. That is an argument that Canadians ought to consider carefully.  A style of data collection that emerged out of the Scottish Enlightenment may no longer be “fit for purpose” to use a management term.

It would be reasonable for Canada`s Conservatives to attack the current census as technologically obsolete waste of money. It is, however, unreasonable for Stockwell Day to suggest that all previous Canadian governments that have engaged in a form of tyranny by making the census mandatory, especially since his government was in office at the time of the sinister 2006 census!!!

My own view is that it would be better to have no census at all in 2011 than a voluntary participation census. A voluntary census would produce unreliable data and would be expensive to run.

The Pew Center in the US has published a piece on the ongoing census debates in Canada and the UK.



6 responses

28 07 2010
Ashley Thomson

Your column illustrates John Start Mill’s dictum that while not all conservatives are stupid–all stupid people are conservative.

28 07 2010


28 07 2010

Some very good points about general innumeracy, but I wouldn’t give historians as a group much credit. I doubt that, as you suggest, most historians have much familiarity with quantitative reasoning and methods. (And I sure hope those that have are not the only readers of your blog!) Such mandatory courses make history undergraduate enrolments wither, although if you ask a lot of faculty members, they will come up with other reasons that privilege the discourse of cultural history over the objectivist conceits of statistical analysis, blahblahblah.

P.S. the Nazis did use the census to track down Jews (using IBM computing power)…

28 07 2010

I’ve used some elements from “History by Numbers” when I teach 2006 (Approaches to European History) and have students do some elementary statistical analysis from the Old Bailey Online database. I think I’ll need to do some more exact modelling of a sample analysis so that they understand that they need to say more than “a lot” did this and “only a few” were convicted of that.

I find that the continuing ed students come to the course feeling “burned” over stats since they have to take the social science stats course for their B.Ed. and most heartily dislike that course. So I spend a while trying to show them how a bit of statistical literacy will only help their own historical work and reading. It’s a bit of an uphill battle because many are self-declared math-phobics.

The entire census kerfuffle has me annoyed — it really seems to be a combination of pandering to the base and statistical illiteracy!

28 07 2010 · Tuesday & Wednesday Census Media Roundup

[…] Andrew Smith’s Blog: The Census Flap illustrates that history professors are useful to society […]

30 07 2010
Insert Real name

If the data collection methods of the Census are outmoded for present & future Canadian purposes, what can replace them and still maintain some accuracy?

We are not like some other countries, where central and regional governments have a long-established mandate to provide access to population, employment, property and taxation registries for the ongoing work of national statistical authorities (e.g. Statistics Norway has an English language brochure at that briefly describes their data sources), and administrative procedures are in place to protect people’s privacy while preserving data collection.

The earlier comment that the Conservatives’ Census changes (and it does seem like they are going to stick to them…) are “a combination of pandering to the base and statistical illiteracy” is so true, but I can’t help but think there’s a baser move afoot here (which andrewdsmith has already alluded to): destroy the popular legitimacy of such Census surveys, so that their “unfitness for purpose” is a foreordained conclusion.

I know that kind of comment sounds very tin-foil-hattish, but consider that the current conservatism in this country is very fond of “maximal” solutions: it’s never enough just to simply win with arguments against your opponents (in this case, they are those who advocate evidence-based, rather than conviction/gut-based policies), rather you must delegitimize, remove their oxygen, destroy them. (Today’s conservatives are rather like Lenin’s Bolsheviks in that respect.)

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: