Sunday, April 02, 2006

US Fertility I

by Edward Hugh

This post needs to be read in association with the agenda outlined in my last post. Basically I have been arguing on this blog that the US derived Total fertility Rate is a composite statistic being derived from three separate fertility regimes. Commenter SM Sterling points out that, of course, any taxonomic system is to some extent arbitrary, and that you could multiply this number considerably, it is simply that I am not convinced what positive advantage would be achieved by doing so).

Conventionally, for example over at the Population Reference Bureau, these regimes are defined ethnically:

One explanation for the higher U.S. fertility is that many European countries have racially homogeneous populations compared with the United States. In the United States, fertility rates differ among the nation's varied racial and ethnic population groups. In 1998, the U.S. TFR of 2.1 children per woman was made up of several different rates: non-Hispanic white, 1.8; black, 2.2; American Indian, 2.1; Asian and Pacific Islander, 1.9; and Hispanic, 2.9.

This is certainly part of the picture, and the united states is certainly more diverse than many EU societies. But is it as simple as this. I tried to have a first shot at this problem in this post. Basically I think, as I argue that that the impact of the most recent wave of immigration on current US fertility has been huge, but immigration is not the only pertinent factor here.

Basically, as I say there are three regimes:

1) A regime which follows the 'European Pattern' of steadily rising educational levels, increasing life expectancy and declining fertility. I say 'European Pattern' since of course this trend is being follwed increasingly in Asia, and, as many have noted, in Western Turkey and Iran (to name but some). There is nothing especially 'ethnic' about all this, it is simply a product of one path of economic development.

2) A regime which is based on high rates of school dropout, unstable partnerships, and frequent adolencent pregnancy. Since this regime, in one form or another, can be found in the UK and Ireland it has been called an 'anglo' regime, but, of course the Irish are Celts, so again ethnic labelling fails us. The parts of the Afro American population also appear to follow this profile.

3) The hispanic regime. This is a relatively high-fertility regime (around 2.8 TFR and falling), but appears to combine many of the features of relatively higher fertility to be found in traditional societies with some of the teenage pregnancy, school dropout unstable partnership patterns which characterise the second regime. As such the rate of fertility decline to be expected here is hard to foresee.

Fertility readings for the first group may be deceptive, becuase almost certainly widespread birth postponenment has been taking place over recent decades in this group, and this, as we know, produces early statistical declines in the TFRs which later reverse as the 'missing births' occur at the higher ages. What will be very interesting will be to follow the final parities for subsequent cohorts in this group.

Also to be watched is the evolution of the third regime. It is typically a 'third world' regime, and as such may be subject to rapid downward movements in TFRs, movements which can be produced by substantial upward movements in first birth ages, just as we can see in Southern Europe or in the Asian Tigers.


On the first regime, this paper by this intriguing paper by US researchers Robert Drago & Amy Varner “Fertility and Work in the United States: A Policy Perspective” is a useful background to the pre-latest wave immigration evolution of US fertility.


Anonymous said...

Regional and religious differences in US fertility are more important than ethnic ones.

For example, there are 15 states where non-Hispanic white women have TFR's of 1.8 or lower.

And there are 35 states where they have TFR's of 1.8 or higher, usually over 2.

Conversely, Puerto Rico has a TFR well below the national average -- 1.75, as of this year.

The gap between religious and non-religious is particularly wide, nearly 2 children per woman.

This is very important in the US, where practicing believers are a growing majority, rather than a small minority as in most of Europe.

Admin said...

"practicing believers are a growing majority"

I find it hard to believe that practicing believers are a growing majority, as opposed to believers. I realise of course that the US is very different from Europe (and almost anywhere else), and I understand that the population of 'practising believers' is growing, but to you really have 150 million plus people in some sort of institution worshiping every weekend?

If this is the case it would be really mind blowing.

Anonymous said...

One of the Barna Group's questions is "Have you attended a religious service, not including a special event such as a wedding or funeral, in the past seven days?" In 2005, 45% of U.S. adults surveyed answered yes. This number has been rising since 2000, when it hit a low of 40%, but is down from 1991, when it topped out at 49%.

Edward Hugh said...

"In 2005, 45% of U.S. adults surveyed answered yes."

If I wanted to be a stickler and a pedant I'd jump in and point out that 45% isn't a majority, and indeed, as you point out, the percentage is declining and not rising (at least over the longer time horizon), but since I am neither I will just say 'wow'. These numbers are huge.

What we are about to see is one enormous social experiment. I am fascinated, riveted even. I really want to see what gets to happen next.

Without wishing to detract from the magnitude of this phenomenon some other thoughts do come to mind on interpreting this data (as in election opinion data, or do you beat your wife surveys). Do 'practising believers' always tell the truth? Or, let me put this, ahem, another way, do 'lapsed practising believers' give responses which have any higher level of veracity than those offered by outright atheists? Are there any ways to decide?

Or again, are there reasons for thinking that agnostics are less (or more) truthful when it comes to answering such questions.

Anonymous said...


This is a very real question, and a hard one to answer. Anecdotally, my own congregation once tried to measure it, and found that for every two people who worship with us at least three out of four weekends, there are three who attend worship once or twice a month.

What fraction of this latter group of people believe they usually attend, and so would answer to a pollster that the had attended when they in fact had not? I have no idea. At the same time, these are not exactly secular people either.

Personally, I suspect the pollster's concept of a 'practicing believer' is a less-than-useful concept when trying to measure the actual religious practice of most of the United States. Robert Wuthnow (Center for the Study of Religion at Princeton University) has (by methods I know not) broken things down as one-fourth of the populace is devoutly religious (by some standard), one fourth profoundly secular, and the remaining half has some degree of interest in religion. As his measurement criteria is vague at best, it is difficult to substantiate, but it seems to mesh with my anecdotal experience.

I have seen very little work worth recommending on what effect this "degree of interest" translates into in terms of the actual practice of this largest group of Americans. It is not an un-interesting question, but I think a very hard one to study.

Now, to somehow drag this post kicking and screaming back to something more closely cogent to demographics, what will the effects of large-scale Latino migration into the United States on its religious praxis be?

Anecdotally, a Mexican co-worker (who is legal, but knows many who aren't) tells me that people who were not religious in Mexico often become religious in the U.S. (Living in fear of discovery and deportation may well do that.) At the same time, he says, the U.S. Catholic church is so different from Mexican Catholicism, that even devout migrants rarely attend mass, and the most common form of religion among Mexican migrants is a sort of folk Catholicism practiced in the home.

Admin said...

"Now, to somehow drag this post kicking and screaming back to something more closely cogent to demographics"

Well thanks for the sentiment, but funnily enough Sterling has now convinced me that all of this does have something to do with your fertility profile. Quite what, and how, may well of course be another question.

"what will the effects of large-scale Latino migration into the United States on its religious praxis be?"

Yes, and of course on your whole social system.

I think it may be important to differentiate between flows of indigenous and non-indigenous populations, and I am not seeing much discussion of this difference anywhere yet.

I know Sterling does not appreciate annecdotal evidence but still, it is the best I have in this case, since here in Sapin most of the huge migratory flux from LA into Spain is now of indigenous peoples (and remember nearly as many illegals are entering Spain at the moment as are entering the entire US, although the distribution is more diverse) the evidence seems to be that the new migrants are more evangelist than catholic. Certainly alongside the locutorios there spring up makeshift meeting places in the 'barrios', and the established churches remain near empty. Whether insecurity and distance from loved ones has any explanatory role here I really don't know.

First generation migrants are unusual in many ways. The important area to study will be the children, and their children.

"tells me that people who were not religious in Mexico often become religious in the U.S."

Well curiously this correlates with a research finding - I'll post the paper in another moment - from Gray Swicegood which suggests that Mexicans also have more children in the US than they do in Mexico, but this data is hard to interpret.

Admin said...

Sorry, it's Stirling. Another lapsus :).