A Canadian – not just Quebec – scandal

Last September, former premier Jacques Parizeau issued a cri d’alarme about the high dropout rate among francophone students in Quebec – boys in particular (see page 74). To maximize impact, he published his warning neither in the respectable La Presse nor in the sophisticated Le Devoir, but in the downmarket popular tabloid Le Journal de Montréal.

Maybe his article is no more than acknowledgement of a widely known malaise that few expect to change. Or maybe, in due course, Parizeau will be seen as prophet of a cultural change in attitudes among Québécois. Six months after his cri d’alarme, interesting Quebec initiatives are underway that may have an impact (admittedly they may not). Jacques Ménard, a leading figure among Canada’s financial elite, has brought together a 27-member Groupe d’Action sur la Persévérance et la Réussite Scolaires (action group on school continuation and success). As title for their report, they chose Savoir pour pouvoir (knowledge for power), a paraphrase of a famous aphorism by the 17th-century philosopher Francis Bacon who, by the way, wrote it in neither English nor French but in Latin.1

If Quebec elites are distressed over dropout rates in their province, so too should be the elites in several other provinces. Once every five years, the census provides a snapshot of the state of educational achievement among Canadians aged 15 and over. Among those who drop out as teenagers, some return to school and obtain secondary school certification at an older age. As of the 2006 census, the lowest dropout rates are among those ages 25–34. Figure 1 on page 92 shows census data on people in that age range without high school certification by sex, by province and, in the case of Quebec, by English and French mother tongue.2 Four provinces recorded higher dropout rates than Quebec francophones: Manitoba the highest, followed by Saskatchewan, Newfoundland and Alberta.

Is the concern about school dropouts merited?

Sceptics point to the impressive scores of Canada’s provincial school systems relative to other countries participating in the Programme for International Student Assessment (PISA), a major international exercise in assessment of 15-year-old secondary school students conducted every three years by the OECD. Among 57 participating countries in the latest round, in 2006, Canada ranked third on the key measure, the “combined science” index. If we rank the ten provinces among the 57 participating countries via this index, the top ten become Finland (1), Alberta (2), Hong Kong (3), British Columbia (4), Ontario (5), Canada overall (6), Taiwan (7), Estonia (8), Japan (9) and Quebec (10).

The problem with any comparison based on averages is that we learn very little about those likely to drop out of school. Potential dropouts tend to turn up in the bottom tail of test score distributions – if they turn up at all. Their school attendance is often erratic and schools may be less than diligent in ensuring that such students actually write a PISA test.

To return to the question, concern about dropouts is merited. First, because very high dropout rates among significant minority groups – such as Aboriginal men in the Prairies (figure 2) – destroy any sense of community. And second, because failing to complete high school has dire economic and social consequences for those who drop out.

Once in the labour force, school dropouts experience much lower employment rates and earn less than students with higher education levels. In 2006, the employment rate among those whose highest education level was high school certification was 64 per cent, compared to only 38 per cent among those who had not completed high school. And while the average income reported by people whose highest educational level was high school certification was $24,200 in 2006, the comparable average among those without certification was $18,700 – a difference of nearly 30 per cent.3 The probability of someone between the ages of 25 and 64 without high school certification reporting an income below the 2005 after-tax low-income cutoff (LICO) is nearly a third.4 This is double the rate for those with a trades certificate, and nearly three times the rate for those with a university degree.

In sum, high school completion may be a low rung on the educational ladder, given that some form of postsecondary training is a requirement for good prospects of earning a “middle-class” income. But it is an important rung, as not obtaining high school certification is a good indicator of subsequently realizing abnormally low incomes.

Parizeau implicitly identifies three causes for high student dropout rates. These factors are intimately intertwined and assessing their relative importance is no easy task:

  • First is school quality, interpreted broadly to include curriculum and provincial education regulations, quality of teacher training offered by education faculties, experience of teaching staff in particular schools, school infrastructure and so on.
  • Second is the disturbing emergence of a gender gap whereby boys perform consistently less well in formal education than girls.
  • Third is the idea that relevant cultural, political and administrative elites are, for whatever historical reasons, insufficiently committed to educational success. They may be committed to educational success for their own children, but tolerant of widespread educational failures elsewhere in their communities.

To these three factors, I add a fourth: the syndrome of low Aboriginal education outcomes, especially among those who identify as North American Indian/First Nation.

School quality

The two lumbering elephants in any analysis of student performance are socioeconomic status of students’ families and in-school factors that collectively define school quality. Much of the rigorous analysis of student performance has been done in the United States. In all studies of low-performing U.S. schools, low socioeconomic status affects student outcomes in the expected direction, but in general, in-school factors cumulatively matter more.

A representative study illustrating all this is one by Eric Hanushek and Steven Rivkin of black/white student performance in Texas schools.5 Among the many in-school factors that might be modified, those that are administratively simple to change – such as lowering class size – do not help much. Hanushek and Rivkin stress three school-related factors that dramatically lower black student performance: the negative effect of a high student turnover rate (which adversely affects student performance of both those who change schools and their classmates), the negative effect on outcomes when teachers have limited teaching experience, and a negative effect from a high share of black students within a school (presumably because of low academic expectations by teachers and peers).

Socioeconomic and school quality factors are intertwined. Low-income parents lack job security and are more mobile than those with stable middle-class jobs. Teaching is more arduous in low-performing schools in low-income neighbourhoods than in middle-class schools. Teacher collective bargaining agreements typically base compensation on seniority and formal training, so that there are few incentives to make a career of teaching in “tough” schools. Not surprisingly, good teachers gravitate away from them.

“Trouble with boys”

The gender gap has become a serious educational problem.6 Among those aged 45 and over in Canada, men have lower high school dropout rates and higher rates of postsecondary training than women. Among those under age 45, these rankings have been reversed – in all provinces. Four of seven Canadians aged 25–34 without high school certification are men (figure 1). At higher educational levels, women are outperforming men by a similar ratio: four of seven Canadians aged 25–34 with university degrees are women.

Canadians under age 45 at the time of the 2006 census were born in 1961 or after. They have participated in a major social transformation in the social role of women over the last half-century, a transformation that has occurred throughout the industrial world. One feature of this transformation has been the removal of barriers to women’s education and a consequent “catching up” to men in terms of earnings. In Canada in 2006, the ratio of average after-tax female/male incomes was only 58 per cent among those aged 55–64; among those aged 25–34 it was 78 per cent. The combination of “catching up” among younger women and a larger shift toward higher educational levels among women than among men augurs well for Canadian women achieving overall parity within the next decade or so.

That female/male income gaps are narrowing, that the female 25–34 high school dropout rate has declined by two thirds relative to that for women over 45 and that many more women are obtaining university degrees are all welcome changes. But why aren’t men achieving comparable educational gains? Why has the male 25–34 dropout rate declined by less than half relative to that for men over 45? If the rate for women aged 25–34 achieving postsecondary training has risen relative to the 35–44 rate, why has the rate for men stagnated?

The Canadian labour market is evolving toward jobs requiring formal training, with a prerequisite of competence in the core academic skills of reading, writing and numeracy taught in the elementary and secondary school system. The supply of well-paying jobs requiring physical strength and manual dexterity but limited formal education is declining. These emerging gender gaps pose novel problems for the next generation. How will well-educated women fare in choice of partners? Will undereducated men be able to find partners?

Aboriginal education

Parizeau describes the status quo with respect to Quebec francophone dropout rates as “a shocking situation, a huge waste of human talent that imperils the future. I experienced a similar shock in 1962, when a census survey of Quebecers’ education showed that 54 per cent of adults over the age of 25 had not gone past Grade 6.” He places Quebec’s current political and administrative elites, plus those in education faculties, in the same purgatory as the Roman Catholic hierarchy that tolerated francophone Quebec’s low educational outcomes prior to the Quiet Revolution of the 1960s.

Parizeau’s critique of the Quebec elite’s tolerating low education attainment is, in my opinion, largely correct. This statement needs to be qualified: first, such critiques are by their nature hard to document to the satisfaction of sceptics, and second, even if elites engage actively, it is uncertain whether more money and different educational policies at the provincial level can improve outcomes in the trenches – in the thousands of classrooms in the hundreds of schools within a provincial school system. Having stated these qualifications, I make a parallel cri d’alarme with respect to Aboriginal education.

Two preliminaries. The census employs several definitions of the Aboriginal population. One is based on ancestry. The second – and most frequently used – is self-identification. Individuals can self-identify as belonging to one of three Aboriginal groups:

  • North American Indian or First Nation (Mohawk, Ojibway, Cree and so on);
  • Métis (descendants of communities formed from the intermarriage of Indians and French fur traders);
  • Arctic Inuit.

A person self-identifying as Aboriginal does not necessarily have Aboriginal ancestry. A third definition is based on indicating that one is a “registered Indian” under provisions of the Indian Act, a statute dating from the late 19th century. The great majority of those identifying as Indian/First Nation are also registered Indians. Only registered Indians have the right to live on reserves. The census defines the “Aboriginal identity population” as all those who either self-identify as Aboriginal or indicate that they are registered Indians.

A second preliminary is to distinguish between registered Indian students attending on-reserve schools and Aboriginal students attending off-reserve schools along with non-Aboriginal students. Jurisdiction over on-reserve schools lies formally with Ottawa; in practice, control lies with the relevant band councils. With a few exceptions, jurisdiction over off-reserve schools is provincial. About a third of students living on-reserve – a higher ratio at the secondary than primary level – attend off-reserve schools. According to the census, three quarters of Aboriginals now live off-reserve and their children attend provincially run schools. This fraction includes all Métis and slightly over half of those who identify as Indian/ First Nation. On the basis of these ratios, on-reserve schools run by band councils are responsible for educating about one Aboriginal child in six; provincial governments are responsible for the other five.

In making the case that Aboriginal education – in particular education among those who identify as Indian/First Nation – is in crisis, I start with the evidence from figure 2, which compares incomplete high school rates for Aboriginal people aged 25–34 with the non-Aboriginal population. Aboriginals are divided into those identifying as Indian/First Nation and Métis, and the comparison is carried out for Canada and six selected provinces, representing more than 90 per cent of Canada’s Aboriginal population.

To summarize, the national Métis rate is twice that for non-Aboriginals; the national rate for Indians/ First Nations is nearly four times higher. Sizable gender gaps exist among all three populations, in all provinces. Aboriginal education outcomes are better than average in Ontario and British Columbia, close to the national average in Quebec and much worse in the Prairies, home to nearly half the Aboriginal population. In Manitoba, well over half of Indian/First Nation men aged 25–34 lack high school certification.

Among older Aboriginals, the great majority live in rural areas, where formal education skills are in less demand. But the image of Aboriginals living in remote rural communities now applies to a minority. In 2006, for the first time, the census recorded a small majority (53 per cent) of Aboriginals living in urban areas, and nearly a third living in cities with more than 100,000 people. Métis are the more urban of the Aboriginal subgroups (at 69 per cent), but a large minority (45 per cent) of Indians/First Nations also live in urban areas.

With a few honourable exceptions, there are no voices in the Assembly of First Nations, in Parliament or in provincial legislatures sounding cris d’alarme over Aboriginal education. Ottawa has recently set in place a high-profile Commission for Truth and Reconciliation with respect to the historical wrongs of the residential school system. Where is the equivalent high-profile commission dealing with the wrongs surrounding the present Aboriginal school system?

Policy implications

The first matter is to collect, publish and discuss the evidence. Admittedly, data may be manipulated to satisfy prejudices of the analyst, and a “good education” entails dimensions not readily measurable via statistics. Neither criticism invalidates the need to use evidence as benchmark in evaluating the status quo and setting goals.

Some school districts in Britain and the United States are experimenting in programs to address boys’ educational performance. Programs include devotion of more time to physical activity during the school day in early years, use of single-sex classes in some subjects and employment of more male teachers to serve as role models. Here is an area for more research.

With respect to the crisis in Aboriginal education, probably the key reform for on-reserve schools is to professionalize school administration by creating Indian/First Nations education boards with budgetary authority and jurisdiction over a sufficient number of schools (12 or more) to realize scale economies. This is an on-reserve equivalent to reforms undertaken among non-Aboriginal rural schools many decades ago.

Off-reserve, Aboriginal education outcomes are better, but they do not justify complacency. Here and there across Canada, a few school districts have, for many years, been aggressively tackling Aboriginal educational underachievement and have made admirable progress. Their strategies should be scaled up. In a study of 366 British Columbia schools in 43 districts that colleagues and I recently undertook, the few successful districts outperformed the others in several identifiable ways. School administrators and teachers had consistently pursued Aboriginal education success as a long-term priority; they had more adequately engaged Aboriginal leaders and the broader Aboriginal community; they made more consistent use of objective data on Aboriginal student performance; and they had a reputation for following through on policy implementation.7

Jacques Parizeau has not become president of a sovereign Quebec. If his cri d’alarme succeeds, he may have achieved something more important.


1   Groupe d’Action sur la Persévérance et la Réussite Scolaires, Savoir pour pouvoir: Entreprendre un chantier national pour la perseverance scolaire (Ménard Report) (Montreal: Author, 2009).

2 A few qualifications need to be made. Census data are not administrative data from education ministries. They are subject to potential bias – presumably upward bias – due to being self-reported. Also, readers need to bear in mind the problems of credentialism. A high school graduate from a relatively low-performing province may have learned much less than a graduate from a province with higher PISA scores. This is a potential problem in Quebec, where high school terminates at Grade 11, not 12 as in other provinces. Another complication arises in making interprovincial comparisons. Ages 20–34 are the prime ages for mobility. Those who migrate typically have higher educational levels than those who stay in their province of origin. Hence, net out-migration will increase a province’s dropout rate, as recorded here; net in-migration will lower it. PEI results are identical to those for Quebec francophones, and are particularly subject to uncertainty due to small numbers.

3 This measure summarizes self-reported after-tax income in 2005 from all sources, not only market earnings. The averages are calculated over all those with any reported income. They are not adjusted for relevant factors such as age and experience.

4 The after-tax low-income cutoff (LICO) used as threshold is $14,562, the 2005 value for one person living in a census metropolitan area with population between 100,000 and 500,000. This statistic should not be interpreted as a poverty rate because it does not adjust for family composition.

5 Eric Hanushek and Steven Rivkin, School Quality and the Black-White Achievement Gap, Working Paper 12651 (Cambridge, MA: National Bureau of Economic Research, 2006).

6 The Trouble with Boys: A Surprising Report Card on Our Sons, Their Problems at School, and What Parents and Educators Must Do is the title of an informal introduction by Peg Tyre (New York: Crown Publishing, 2008) to the educational research and policy debates surrounding the growing gender gap across western Europe and North America. Her conclusion is that the organization of primary schools penalizes boys.

7 For a comprehensive survey of Aboriginal education, with emphasis on U.S. evidence, see William Demmert and Peggy McCardle, “Improving the Education of Native American Children,” Journal of American Indian Education, Vol. 45, No. 3 (2006). Michael Mendelson, in Improving Education on Reserve: A First Nations Education Authority Act (Ottawa: Caledon Institute of Social Policy, 2008; retrieved May 3, 2009, from www.caledoninst.org/Publications/PDF/684ENG.pdf), has written about professionalizing on-reserve schooling. Along with two colleagues, I have written on the lessons to be drawn from the strategies pursued by relatively successful school districts in teaching off-reserve Aboriginal students in British Columbia (John Richards, Jennifer Hove and Kemi Afolabi, Understanding the Aboriginal/non-Aboriginal Gap in Student Performance: Lessons from British Columbia, Commentary No. 276. .