Saturday, July 26, 2008

Demography and culture: French Canada's fall and Québec's isolation

Over the past few months, Statistics Canada has been releasing data sets and analysis from the 2006 Census. Statistics Canada released information on language use, immigration and citizenship, and inter-provincial migration. The textual analyses at Statistics Canada's website are good, but the interactive map created by the Canadian Press and hosted at is wonderful. Of note is the continuing fall in Francophone numbers and proportions outside of Québec and New Brunswick.

What happened? It's important to note that the belief in a French Canadian revanche des berceaux, of a nationalism-driven birth rate under British occupation that saw every family produce large numbers of children, is a legend. The disparity between Ontarian and Québec (and fertility fertility rates emerged after 1870, when Ontario moved in the direction of a demographic regime characterized by sharply falling death and infant mortality rates in tandem with falling birth rates, while Québec--more conservative, more rural--lagged behind. For a long while, this was enough to keep the Francophone proportion of Canada's population stable at roughly 30%. But then the 1960s hit.

Between 1850 and 1950, owing to a high fertility rate among French Canadian women, the proportion of Francophones in the Canadian population held at 30%. The fertility of the Francophones then dropped below the Canadian average toward the mid-1960s, contributing to a decrease in the proportion of the total Canadian population speaking French as a mother tongue -- from 29% in 1951 to 25% in 1986. Between 1926 and 1960, the fertility rate of women in Quebec moved closer to that of other Canadian women. In effect, the ratio of the fertility rate of Quebec women to other Canadian women dropped from 1.45 in 1926 to 1.30 around 1940 and 1.15 around 1950. Between 1960 and 1970, fertility declined very rapidly. The total fertility rate of women in Quebec dropped from 3.9 to 2.1; the fertility of other Canadian women declined but not so markedly, from 4.0 to 2.5. By 1974, fertility in Quebec had declined to 1.8 children/woman, and by 1986 the fertility rate had fallen still further to 1.4, while that of other Canadian women held at between 1.7 and 1.8. The fertility rate was 20% lower in Quebec in 1986 than in the other provinces. These differences in fertility between Québec and the rest of Canada have significantly affected the demographic situation. Census data have shown that the completed fertility of Francophones was 80% higher than that of Anglophones for women born at the turn of the century, but this gap narrowed rapidly over the years and disappeared for women born between 1931-36.

These differences also affected Francophone populations outside of Québec, as the same source notes, with Francophone TFRs outside of Québec falling not only below the levels of non-Francophone TFRs ("[i]n Ontario [in 1986], the total fertility rate for Francophones was 1.54, compared to 1.61 for Anglophones and 1.75 for the other groups"). The fertility of Québec--perhaps a reasonably good proxy for Francophone Canada as a whole--is not the lowest-low fertility of Italy, and it might well have only anticipated trends in English Canada. Even so, the province of Québec is going to have to deal with relatively low if not negative population growth and relatively rapid aging, further complicated by issues with immigration, as evidenced by the loud debate on the topic of how best to integrate immigrants. (The old endogamy of the pre-1960s period is most certainly gone, even if the problem of dealing with Others remains, as with other peoples and cultures.)

Survival will be an unsurmountable challenge for most Francophone minorities outside of Québec. In most of the rest of Canada, Francophones form significant minorities (and, occasionally, majorities) only in the "bilingual belt" stretching more-or-less along the Québec border. Can this be changed? It's questionable. Francophones outside of Québec once exhibited higher TFRs than Francophones inside Québec, but this situation has reversed itself. In the 1996-2001 period, there was a significant amount of Francophone out-migration to Ontario, New Brunswick, and Alberta, suggesting a certain potential for the revitalization of those communities, but again Francophones in Ontario and Alberta form only a small portion of the population, while the stable Francophone community of New Brunswick has often been a net exporter of population to Québec. Without any remarkable demographic event--a Francophone baby boom,. mass immigrations from Francophone Europe and Africa, et cetera--non-Francophones won't have any particular reason to view French as particularly relevant to their lives. Journalist Chantal Hébert was right to point out in a December 2007 article for The Toronto Star that "French still an abstraction in much of Canada": "[O]utside Quebec, Francophones make up only 4 per cent of the population. With French an abstraction in so many parts of Canada, the motivation to learn it as a second language is decreasing. Because most Anglophones learn French at school, the peak bilingualism rate for Anglophones outside Quebec occurs in the 15 to 19 age range. Over the past decade, it has slipped from 16.3 per cent to 13 per cent. The census also shows the retention rate of Anglophones who have learned the language is slipping." Even in 2001, Statistics Canada discovered that overall, "43.4% of Francophones reported that they were bilingual, compared with 9.0% of Anglophones." For those Francophones living outside of Québec, living in an overwhelmingly non-Francophone society with few if any barriers to intermarriage, these pressures create a perfect environment for language shift. It's not surprising that there was an overall decline in the numbers of Francophones outside of Québec. As Gilles Grenier put it in his paper "Linguistic and Economic Characteristics of Francophone Minorities in Canada: A Comparison of Ontario and New Brunswick" (Journal of Multilingual and Multicultural Development (18.8): 1997), in mixed environments

A lot still needs to be learned about what exactly determines assimilation, but it is clear that the fact that Anglophones and Francophones are more mixed together than they used to, mainly because the Canadian society has become more urbanised and communication systems more developed, leads to a more widespread use of the common majority language. If Francophones in Quebec and in New Brunswick have been able to maintain their language, it is because there is a geographical separation between them and the other communities. This does not mean that those Francophones do not use English for some of their economic activities, but at least French is still the dominant language in their own communities. One major reason of the assimilation in Ontario is that there are less and less towns and villages where the majority of the population is French.(299).

This last sentence is critical, since mass language shift from French to English is not unique in the history of North American Francophones. Starting in the late 19th century, relative economic underdevelopment propelled a tremendous migration of Francophones out from their traditional settlement areas along the St. Lawrence River and the Gulf of St. Lawrence to adjacent parts of the continent. Large and thriving communities of Franco-Americans (concentrated in New England, particularly in that region's industrial cities) and Franco-Ontarians (concentrated in northern and eastern areas adjacent to Québec) formed at this time, encouraging some to believe in the idea of a greater Québec encompassing those communities. That vision failed, as the Franco-American community was whittled away through immigration restrictions and acculturation to the Anglophone culture surrounding them. Franco-Ontarians, who with few exceptions like in Anglophone-majority communities relatively isolated from Québec, may be about to follow. And no, the Francophones who are tourists in Maine or long-term residents in Florida don't make up the same sorts of communities. What the long-term effects of a Québec isolated in its language from most of the rest of Canada and attached to a wider Francophone world and (through a traditional Ameriphilia) to the United States will be on a Canada faced with its own regional challenges, I leave to my readers to speculate.

Why is all this particularly relevant to readers of Demography Matters? Shifting population balances are central to our work here. The shrinkage of working-age populations as aging proceeds is something that we've looked at, just as we have taken a look at the effects of emigration on the long-term futures of different countries, just as we have taken a look at unbalanced sex ratios. The transformation of populations via cultural or linguistic shifts is just as surely an issue of note. How's Catalan faring in the Generalitat and the Balearics and Valencia, in the context of Spain's recent immigration boom, and if badly, could this lead to anti-immigration sentiment or legislation? Is the proportion of first-language Russophones in Latvia shrinking significantly, and if so could this lead to a cautious liberalization of immigration legislation? Will Russia try to make good on its promises to the Russophones of the former Soviet Union? Just what is going on with Spanish in the United States and what does this suggest about Hispanics? Et cetera, et cetera, et cetera.

Demography matters, and so do the underlying cultural issues that not only help define demographics but influence the ways in which we think of those shifts. In Canada, English Canadians and French Canadians have dealt with each other substantially in relation to their demographics ("Are the French have too many children?" "Are the English trying to overwhelm us with immigrants?"), managing to stifle until relatively recently serious discussion about what's actually going on. If you think that this was about Eurabia, well, yes, it is in part about that, but it is more importantly about the need to come up with examples from the past and the present of underdiscussed and often misdiscussed issues sine, after all, Demography Matters.

What happened to Flemish France, what's going on with the Tibetans, and what's happening to the ethnic Poles in the former Soviet Union? Come back here to find out.

Tuesday, July 08, 2008

Investing in Human Capital - Quantity or Quality?

A mail from a friend today alerted me to this fine op-ed in the IHT by Philip Bowring. Bowring's main point is quite neat even if of course the ultimate conclusion to re-write the practices of national accounting may be a bit far-fetched. Bowring's starting point is the economies of East Asia which, I am sure my readers will, agree have performed quite well in the 1990s and 2000s with the important intermezzo of the Asian currency crisis in 1997. In fact, one could argue that with the crisis in 1997 East Asian economies embarked on an even further solidified path of growth driven by investment and accumulation of capital and, we should never forget, foreign exchange reserves. You see, this was part of the scheme at the offset in the sense that these economies would not risk to go hat in hand to the IMF the next time crisis loomed at the doorstep. This is not about the Asian currency crisis and its aftermath per se but if you want more I have a synopsis up here.

Philip Bowring starts off with the simple point that although East Asian economies in general have been doing very well on the investment front with respect to physical capital (and according to many thequality of human capital) they have not been able to halt the demographic transition which to a considerable degree has eroded the quantity foundation of these economies' human capital. In this way and as if guided by an invisible force, these economies have also transcended into lowest low (TFR < 1.5) and in fact some economies are hovering around the "single child per women" mark. This makes Mr. Bowring wonder whether in fact the economic experience of these economies are so stellar as we are lead to believe.

South Korea, Hong Kong, Taiwan and Singapore have over 40 years averaged roughly the highest consistent economic growth rates in the world. All but Korea have steadily accumulated massive surplus savings and foreign reserves. But change the national accounting principles behind these rosy numbers and a different picture emerges, one that the societies concerned have barely begun to grapple with. In one vital respect these countries (soon to be joined by China) collectively may have the worst record of investment in the future since homo sapiens evolved: Investment in the next generation.

They have the lowest fertility rates in the world. Hong Kong is at the bottom (excluding births to non-residents), with around 1.0 births per woman of child-bearing age (the replacement rate is 2.1). Taiwan, Singapore and Macau come in at 1.1, while South Korea, at 1.2. is on par with Europe's lowest, Belarus. None of these economies has had replacement-rate fertility levels since the late 1980s.


Imagine if these four economies had invested less in infrastructure and reserves and more in people. They would now most likely have much lower foreign-exchange reserves, but they would not be facing a situation in which their work forces - unless replaced by immigrants - will decline dramatically within 20 years as the population over 65 continues to grow. The payback for years of what may well have been the misallocation of resources is not far in the future.


Almost all developed and many developing countries now face demographic challenges from a reverse in the high fertility rates of the past. There are no easy solutions. But by changing national accounting principles to make child-rearing a priority, attitudes toward it might change too.

Bowring's piece poses a lot of intertwined questions which I cannot dwell by here. One would be whether the cultural edifice of Neo-Confucianism is particularly 'adept' in fostering lowest-low fertility (or high household saving rates perhaps?). In this note I shall neatly bypass this question. However one crucial question which emerges from Bowring's piece is whether economists are adequately defining investment in human capital? Now, Bowring's suggestion to change national accounting practices does not, at a first glance, make sense. It would hardly be possible to make a credible short term/high frequency measure of investment and accumulation (say, quarterly) in quality (e.g. education) and quantity (e.g. some form of fertility gauge) of human capital. Yet national accounting is not the only tool economist use to describe the growth of economies. Consequently, when looking at growth in the longer term economists refer to the collective sub-discipline of growth theory or more specifically neo-classical growth theory.

Within the context of neo-classical growth theory the contributions are multiple and I won't even try to provide an overview. Mankiw, Phelps and Romer's Growth of Nations is an excellent starting point for the intermediate scholar.

Without cutting further corners my main gripe with the growth theory framework is that economists traditionally (and this is especially the case in the context of the whole growth theory framework) tend to separate quantity and quality of human capital in a quite unsatisfactory manner. You see, according to growth theory it is a well known dictum that more people leads to lower growth in the illusive steady state. The point here would be that we are getting more people for the same capital and productivity level and thus more people to share the same sized pie. Yet, human capital does matter. As such, Mankiw et al. famously showed that investment in (i.e. accumulation of human capital) is highly conductive of economic growth. This would then mean that investment in education be considered a highly fruitful policy; I hardly think anyone will disagree here. This point also forms a strong link to Bowring's perspective of East Asia since, apart from choosing export orientation over import substitution, these economies were praised for their high saving rates not least in the context of a strong focus on education and especially high enrollment in basic education. At the time they were thus bathing in the spotlight on the expense of their developing economy siblings in Latin America.

The two points made above on human capital basically shore up at the following main point in the context of growth theory.

Few people of high quality (education) is the way to go!

However, it is pretty clear I think that this is not true. At least, there seems to be quite strong evidence that not only quality matters but also quantity. Moreover, if everybody goes for the low quantity high quality option it will create strong negative externalities. In fact, the quantity/quality nexus of human capital is much more complicated with respect to feedback loops than the current standing theory assumes. Basically, the whole model seems to break down once we get into the situation of lowest-low fertility. Especially in this regard would be the concept of steady state which becomes very difficult to sustain in an empirical context in the sense that it becomes a proverbial moving target.

It is within this somewhat wonkish framework that Bowring's point suddenly makes sense. What he suggests is thus that in stead of focusing solely on the quality of human capital we should also look at the quantity. What he is in fact latching on to is the very finely tuned linkage between the two. The point is simply here that when economies emphasise education (quality of human capital) and especially of women as a natural part of the development process they would be wise in also focusing on quantity measures since, as other economists so famously showed, there is a tradeoff between quantity and quality in child rearing. Add to this the tempo (birth postponement) effect of fertility which comes from pushing the whole life course framework of women as they enter the labour market and you end up with a bad policy if taken too far. Moreover, evidence suggests that once the decision is made to go aggressively for quality without making sure that quantity is kept stable it is very difficult (if not impossible) to bend the stick back again.

This is what many Asian countries now confront I feel.

It's Both Actually ...

In many ways, this should not be too difficult to understand but often the debate gets completely loop sided because people focus on population growth rather than population structure. Moreover, once population structure is incorporated into the growth theory framework the goings get quite tough in terms of mathematical model complexity. Yet, as the main conclusions of the neo-classical framework persist the critique remains the same.

Interestingly, this is also why Bowring's comparison of human capital investment (the quantity) with investment in infrastructure is only as good as it goes. Roads can be built within a reasonable time frame (well, if you have the workers that is) but human capital is another matter. Basically, the decisions you take today will only matter in some 20-30 years time and the decisions you took decades ago are only making their presence felt now. This is what path dependency is all about really and also what policy makers in countries such as India, Turkey, Brazil, Morocco (I think it is too late for China) etc. would be wise to consider; both for their own good and the rest of us . The lesson here is then that while no-one can disagree in putting aside investments for the quality of human capital as well as working to get women integrated into the labour force one would be wise to also consider the quantitative perspective.

German and Japanese Populations Continue to Decline in 2007

Hello Everyone,

This is really a post to apologise to all our readers for a rather lengthy absence. This blog is far from dead, and we have simply been busy with other matters, possibly for a lot longer than we realised. Anyway Claus is now back up with a new post, and while I am here I would just mention the fact that....


Germany lost about 97,000 inhabitants in 2007. According to recent data from the Federal Statistics Office, births were up by 1.8% and deaths increased slightly by 0.7%, whereas marriages were down by 1.3%.

There were in fact 685,000 live births registered, that was 12,000 births or 1.8% more than in 2006. Thus the number of births thus increased for the first time since 1997. In 1996 and 1997, the number of births rose slightly, but this was an exception to an otherwise downward trend since 1991.

The number of deaths was down in 2006, but then increased again slightly in 2007 - from 822,000 to 827,000 (+6,000 or +0.7%). The number of deaths declined continuously from 1994 to 2001 but started an upward trend in 2002, 2003 and 2005.
In 2007, the number of deaths thus exceeded births by about 142,000. In the previous year, the birth deficit was 7,000 more.

According to provisional results, Germany had about 82,218,000 inhabitants on 31 December 2007. That was 97,000 inhabitants or 0.1% less than at the end of 2006 (82,315,000).


The Japanese government reported at the end of May that the number of children (all individuals under the age of 15 years old) had fallen to approximately 17,400,000. According to Japanese government estimates, if this trend continues, Japan is on course to lose somewhere between 26% and 31% of its total population.

There were 17,250,000 children aged 14 or under in Japan as of April 1, down by 130,000 from a year earlier, according to an annual survey by the Internal Affairs Ministry which was released to coincide with the May 5 Children's Day national holiday. Japan's population has been shrinking since 2005.

The number of Japanese children in the age group dropped by 0.7 percent from a year earlier. Children are now at a record low, constituting only 13.5 percent of the total population, down from 13.6 percent in April 2006.

Update 14 July 2008

Japan's population did in fact increase slightly in 2007, making for the first gain in three years, according to the latest estimates from the Ministry of Internal Affairs. The increase of 12,707 - which took Japan's total population to 127,066,178 as of March 31 2007 occurred entirely as a result of Japanese returning home from overseas and non-Japanese obtaining Japanese citizenship.