Thursday, January 19, 2017

Immigration and Birthrate

"Let’s talk about the link between immigration and low reproduction rates"
This is a really weird article. It talks about how below replacement birth rates mean the population will decline. Which is true by definition.
Then about how some countries have lots of immigrants. Then it does nothing to link the two. So in spite of asking to talk about the link it doesn't.

I wanted to look first to see if there was a link. As the article does nothing to show there is.
I took a list of countries by their percentage of immigrants
And one of countries by their birthrate
I created this combined dataset of Country, Birthrate and Immigrant % and put it here

The correlation between birthrate and the percentage of immigrants in a country is weak.

> cor(data$FertilityRate, data$ImmigrantPer)
[1] -0.3463663
I am willing to bet you at odds that the correlation between wealth and birth rate and between wealth and % of immigrants is higher. That having money causes immigrants to come to your country and you to have less children. Not that people choose between having a child and a 25 year old Ethiopian.

So Irish Times please do talk about what is at best a weak link between immigration and low reproductive rates.


Paul Rubin said...

There are two decision analytic, if not exactly statistical, connections that come to mind. In countries with reasonably advanced economies and below-replacement-rate birth rates, there may be economic policy arguments for encouraging immigration. That doesn't mean it will happen; there may cultural impediments. Japan, for instance, has an aging population and BRR birth rates, so they probably should be trying to bring in immigrants to keep jobs filled. The numbers suggest that's not happening.

On the other end, countries with high birth rates might be discouraging immigration, thinking they already have enough mouths to feed. Perhaps more importantly, such countries may be unlikely destinations of choice for immigrants. Most of the countries in your data set with high birth rates seem to be in Africa, and not necessarily the most economically advanced countries on that continent. If I'm leaving Poland looking for work, Niger is unlikely to be my first choice.

So looking for a statistical relationship between birth rate and immigration is dicey to start. I think it might be wise to segment the sample, say by GDP or GNP. Also, birth rate might not be the most useful population variable. If the logic in fact is that aging populations need to replace workers, then birth rate adjusted for survival to, say, age 20 might be better. Sad to say, but sometimes a high birth rate exists because too many children die, leaving the parents to start over.

David Curran said...

Great points Paul.

Most countries with very high birth rates are too poor to be a big draw for people.

There are very odd countries. Places like Japan (low both) , United Arab Emirates (loads of immigrant men not many babies) and Vatican City (all immigrants 0 birthrate) for example.

My analysis of the link was pretty half assed. It just annoyed me that someone wanted to talk about a link and then did no analysis of it.