i. Fire works a little differently than people imagine. A great ask-science comment. See also AugustusFink-nottle’s comment in the same thread.
iii. I was very conflicted about whether to link to this because I haven’t actually spent any time looking at it myself so I don’t know if it’s any good, but according to somebody (?) who linked to it on SSC the people behind this stuff have academic backgrounds in evolutionary biology, which is something at least (whether you think this is a good thing or not will probably depend greatly on your opinion of evolutionary biologists, but I’ve definitely learned a lot more about human mating patterns, partner interaction patterns, etc. from evolutionary biologists than I have from personal experience, so I’m probably in the ‘they-sometimes-have-interesting-ideas-about-these-topics-and-those-ideas-may-not-be-terrible’-camp). I figure these guys are much more application-oriented than were some of the previous sources I’ve read on related topics, such as e.g. Kappeler et al. I add the link mostly so that if I in five years time have a stroke that obliterates most of my decision-making skills, causing me to decide that entering the dating market might be a good idea, I’ll have some idea where it might make sense to start.
“Are stereotypes accurate or inaccurate? We summarize evidence that stereotype accuracy is one of the largest and most replicable findings in social psychology. We address controversies in this literature, including the long-standing and continuing but unjustified emphasis on stereotype inaccuracy, how to define and assess stereotype accuracy, and whether stereotypic (vs. individuating) information can be used rationally in person perception. We conclude with suggestions for building theory and for future directions of stereotype (in)accuracy research.”
A few quotes from the paper:
“Demographic stereotypes are accurate. Research has consistently shown moderate to high levels of correspondence accuracy for demographic (e.g., race/ethnicity, gender) stereotypes […]. Nearly all accuracy correlations for consensual stereotypes about race/ethnicity and gender exceed .50 (compared to only 5% of social psychological findings; Richard, Bond, & Stokes-Zoota, 2003).[…] Rather than being based in cultural myths, the shared component of stereotypes is often highly accurate. This pattern cannot be easily explained by motivational or social-constructionist theories of stereotypes and probably reflects a “wisdom of crowds” effect […] personal stereotypes are also quite accurate, with correspondence accuracy for roughly half exceeding r =.50.”
“We found 34 published studies of racial-, ethnic-, and gender-stereotype accuracy. Although not every study examined discrepancy scores, when they did, a plurality or majority of all consensual stereotype judgments were accurate. […] In these 34 studies, when stereotypes were inaccurate, there was more evidence of underestimating than overestimating actual demographic group differences […] Research assessing the accuracy of miscellaneous other stereotypes (e.g., about occupations, college majors, sororities, etc.) has generally found accuracy levels comparable to those for demographic stereotypes”
“A common claim […] is that even though many stereotypes accurately capture group means, they are still not accurate because group means cannot describe every individual group member. […] If people were rational, they would use stereotypes to judge individual targets when they lack information about targets’ unique personal characteristics (i.e., individuating information), when the stereotype itself is highly diagnostic (i.e., highly informative regarding the judgment), and when available individuating information is ambiguous or incompletely useful. People’s judgments robustly conform to rational predictions. In the rare situations in which a stereotype is highly diagnostic, people rely on it (e.g., Crawford, Jussim, Madon, Cain, & Stevens, 2011). When highly diagnostic individuating information is available, people overwhelmingly rely on it (Kunda & Thagard, 1996; effect size averaging r = .70). Stereotype biases average no higher than r = .10 ( Jussim, 2012) but reach r = .25 in the absence of individuating information (Kunda & Thagard, 1996). The more diagnostic individuating information people have, the less they stereotype (Crawford et al., 2011; Krueger & Rothbart, 1988). Thus, people do not indiscriminately apply their stereotypes to all individual members of stereotyped groups.” (Funder incidentally talked about this stuff as well in his book Personality Judgment).
One thing worth mentioning in the context of stereotypes is that if you look at stuff like crime data – which sadly not many people do – and you stratify based on stuff like country of origin, then the sub-group differences you observe tend to be very large. Some of the differences you observe between subgroups are not in the order of something like 10%, which is probably the sort of difference which could easily be ignored without major consequences; some subgroup differences can easily be in the order of one or two orders of magnitude. The differences are in some contexts so large as to basically make it downright idiotic to assume there are no differences – it doesn’t make sense, it’s frankly a stupid thing to do. To give an example, in Germany the probability that a random person, about whom you know nothing, has been a suspect in a thievery case is 22% if that random person happens to be of Algerian extraction, whereas it’s only 0,27% if you’re dealing with an immigrant from China. Roughly one in 13 of those Algerians have also been involved in a case of ‘body (bodily?) harm’, which is the case for less than one in 400 of the Chinese immigrants.
v. Assessing Immigrant Integration in Sweden after the May 2013 Riots. Some data from the article:
“Today, about one-fifth of Sweden’s population has an immigrant background, defined as those who were either born abroad or born in Sweden to two immigrant parents. The foreign born comprised 15.4 percent of the Swedish population in 2012, up from 11.3 percent in 2000 and 9.2 percent in 1990 […] Of the estimated 331,975 asylum applicants registered in EU countries in 2012, 43,865 (or 13 percent) were in Sweden. […] More than half of these applications were from Syrians, Somalis, Afghanis, Serbians, and Eritreans. […] One town of about 80,000 people, Södertälje, since the mid-2000s has taken in more Iraqi refugees than the United States and Canada combined.”
“Coupled with […] macroeconomic changes, the largely humanitarian nature of immigrant arrivals since the 1970s has posed challenges of labor market integration for Sweden, as refugees often arrive with low levels of education and transferable skills […] high unemployment rates have disproportionately affected immigrant communities in Sweden. In 2009-10, Sweden had the highest gap between native and immigrant employment rates among OECD countries. Approximately 63 percent of immigrants were employed compared to 76 percent of the native-born population. This 13 percentage-point gap is significantly greater than the OECD average […] Explanations for the gap include less work experience and domestic formal qualifications such as language skills among immigrants […] Among recent immigrants, defined as those who have been in the country for less than five years, the employment rate differed from that of the native born by more than 27 percentage points. In 2011, the Swedish newspaper Dagens Nyheter reported that 35 percent of the unemployed registered at the Swedish Public Employment Service were foreign born, up from 22 percent in 2005.”
“As immigrant populations have grown, Sweden has experienced a persistent level of segregation — among the highest in Western Europe. In 2008, 60 percent of native Swedes lived in areas where the majority of the population was also Swedish, and 20 percent lived in areas that were virtually 100 percent Swedish. In contrast, 20 percent of Sweden’s foreign born lived in areas where more than 40 percent of the population was also foreign born.”
vi. Book recommendations. Or rather, author recommendations. A while back I asked ‘the people of SSC’ if they knew of any fiction authors I hadn’t read yet which were both funny and easy to read. I got a lot of good suggestions, and the roughly 20 Dick Francis novels I’ve read during the fall I’ve read as a consequence of that thread.
“On the basis of an original survey among native Christians and Muslims of Turkish and Moroccan origin in Germany, France, the Netherlands, Belgium, Austria and Sweden, this paper investigates four research questions comparing native Christians to Muslim immigrants: (1) the extent of religious fundamentalism; (2) its socio-economic determinants; (3) whether it can be distinguished from other indicators of religiosity; and (4) its relationship to hostility towards out-groups (homosexuals, Jews, the West, and Muslims). The results indicate that religious fundamentalist attitudes are much more widespread among Sunnite Muslims than among native Christians, even after controlling for the different demographic and socio-economic compositions of these groups. […] Fundamentalist believers […] show very high levels of out-group hostility, especially among Muslims.”
ix. Portal: Dinosaurs. It would have been so incredibly awesome to have had access to this kind of stuff back when I was a child. The portal includes links to articles with names like ‘Bone Wars‘ – what’s not to like? Again, awesome!
x. “you can’t determine if something is truly random from observations alone. You can only determine if something is not truly random.” (link) An important insight well expressed.
xi. Chessprogramming. If you’re interested in having a look at how chess programs work, this is a neat resource. The wiki contains lots of links with information on specific sub-topics of interest. Also chess-related: The World Championship match between Carlsen and Karjakin has started. To the extent that I’ll be following the live coverage, I’ll be following Svidler et al.’s coverage on chess24. Robin van Kampen and Eric Hansen – both 2600+ elo GMs – did quite well yesterday, in my opinion.
xii. Justified by More Than Logos Alone (Razib Khan).
“Very few are Roman Catholic because they have read Aquinas’ Five Ways. Rather, they are Roman Catholic, in order of necessity, because God aligns with their deep intuitions, basic cognitive needs in terms of cosmological coherency, and because the church serves as an avenue for socialization and repetitive ritual which binds individuals to the greater whole. People do not believe in Catholicism as often as they are born Catholics, and the Catholic religion is rather well fitted to a range of predispositions to the typical human.”
I have had a look at two sources, the Office of Refugee Resettlement’s annual reports to Congress for the financial years 2013 and 2014. I have posted some data from the reports below. In the cases where the page numbers are not included directly in the screen-caps, all page numbers given below are the page numbers of the pdf version of the documents.
I had some trouble with how to deal with the images included in the post; I hope it looks okay now, at least it does on my laptop – but if it doesn’t, I’m not sure I care enough to try to figure out how to resolve the problem. Anyway, to the data!
The one above is the only figure/chart from the 2014 report, but I figured it was worth including here. It’s from page 98 of the report. It’s of some note that, despite the recent drop, 42.8% of the 2014 US arrivals worked/had worked during the year they arrived; in comparison, only 494 of Sweden’s roughly 163.000 asylum seekers who arrived during the year 2015 landed a job that year (link).
All further images/charts below are from the 2013 report.
It’s noteworthy here how different the US employment gap is to e.g. the employment gap in Denmark. In Denmark the employment rate of refugees with fugitive status who have stayed in the country for 5 years is 34%, and the employment rate of refugees with fugitive status who have stayed in the country for 15 years is 37%, compared to a native employment rate of ~74% (link). But just like in Denmark, in the US it matters a great deal where the refugees are coming from:
“Since their arrival in the U.S., 59 percent of refugees in the five-year population worked at one point. This rate was highest for refugees from Latin America (85 percent) and lowest for refugees from the Middle East (48 percent), while refugees from South/Southeast Asia (61 percent) and Africa (59 percent) were positioned in between. […] The highest disparity between male and female labor force participation rates was found for respondents from the Middle East (64.1 percent for males vs. 34.5 percent for females, a gap of 30 points). A sizeable gender gap was also found among refugees from South/Southeast Asia (24 percentage points) and Africa (18 percentage points), but there was hardly any gap among Latin American refugees (3 percentage points). Among all refugee groups, 71 percent of males were working or looking for work at the time of the 2013 survey, compared with 49 percent of females.” (p.94)
Two tables (both are from page 103 of the 2013 report):
When judged by variables such as home ownership and the proportion of people who survive on public assistance, people who have stayed longer do better (Table II-16). But if you consider table II-17, a much larger proportion of the refugees surveyed in 2013 than in 2008 are partially dependent on public assistance, and it seems that a substantially smaller proportion of the refugees living in the US in the year 2013 was totally self-reliant than was the case 5 years earlier. Fortunately the 2013 report has a bit more data on this stuff (p. 107):
The table has more information on page 108, with more details about specific public assistance programs.Table II-22 includes data on how public assistance utilization has developed over time (it’s clear that utilization rates increased substantially during the half-decade observed):
Some related comments from the report:
“Use of non-cash assistance was generally higher than cash assistance. This is probably because Medicaid, the Supplemental Nutrition Assistance Program (SNAP), and housing assistance programs, though available to cash assistance households, also are available more broadly to households without children. SNAP utilization was lowest among Latin Americans (37 percent) but much higher for the other groups, reaching 89 to 91 percent among the refugees from Africa and the Middle East. […] Housing assistance varied by refugee group — as low as 4 percent for Latin American refugees and as high as 32 percent for refugees from South/Southeast Asia in the 2013 survey. In the same period, other refugee groups averaged use of housing assistance between 19 and 31 percent.” (pp. 107-108)
The report includes some specific data on Iraqi refugees – here’s one table from that section:
The employment rate of the Iraqis increased from 29.8% in the 2009 survey to 41.3% in 2013. However the US female employment rate is still actually not much different from the female employment rates you observe when you look at European data on these topics – just 29%, up from 18.8% in 2009. As a comparison, in the year 2010 the employment rate of Iraqi females living in Denmark was 28% (n=10163) (data from p.55 of the Statistics Denmark publication Indvandrere i Danmark 2011), almost exactly the same as the employment rate of female Iraqis in the US.
Of note in the context of the US data is perhaps also the fact that despite the employment rate going up for females in the time period observed, the labour market participation rate of this group actually decreased between 2009 and 2013, as it went from 42.2% to 38.1%. So more than 3 out of 5 Iraqi female refugees living in the US are outside the labour market, and almost one in four of those that are not are unemployed. A few observations from the report:
“The survey found that the overall EPR [employment rate, US] for the 2007 to 2009 Iraqi refugee group in the 2013 survey9 was 41 percent (55 percent for males and 29 percent for females), a steady increase in the overall rate from 39 percent in the 2012 survey, 36 percent in the 2011 survey, 31 percent in the 2010 survey, and 30 percent in the 2009 survey. As a point of further reference, the EPR for the general U.S. population was 58.5 percent in 2013, about 17 percentage points higher than that of the 2007 to 2009 Iraqi refugee group (41.3 percent). The U.S. male population EPR was nine percentage points higher than the rate for Iraqi males who arrived in the U.S. in 2007 to 2009 (64 percent versus 55 percent), while the rate for the Iraqi females who arrived in the U.S. in 2007 to 2009 was 24 points higher for all U.S. women (53 percent versus 29 percent). The difference between the male and female EPRs among the same group of Iraqi refugees (26 percentage points) also was much larger than the gap between male and female EPRs in the general U.S. population (11 points) […] The overall unemployment rate for the 2007 to 2009 Iraqi refugee group was 22.9 percent in the 2013 survey, about four times higher than that of the general U.S. population (6.5 percent) in 2013” (pp. 114-115).
“Key Figures – Denmark: Among 35 year old men in 2004: 28% convicted at least once (non-trafic), 14% convicted at least twice (non-trafic), 12% prison sentence (suspended or not).”
From a course lecture note. I’ve written about the crime rates of immigrants in Denmark before (Danish link). The number you need to know from that article is this one: In 2007, 27,2% of (n=1449) male descendants of non-Western immigrants at the age of 20-29 years old got a conviction. I will emphasize that this is in that year alone; this is not an estimate of how many of the 30 year-olds got convicted while they were at the age of 20-29 – this is a snapshot, and during one year more than a fourth of these people got convicted of a crime.
You’d be tempted to say that the fraction of non-Western descendants in Denmark that commit crime while at the age of 20-29 corresponds to the fraction of Danes at the age of 35 who’ve ever been convicted. It’s not quite that bad, because the descendant numbers include traffic violations which are excluded in the other measure and traffic crimes make up a large chunk of the total – 58% of convictions of all descendants (Statistics Denmark doesn’t make it easy to separate non-Westerners from the rest) were traffic-related in 2011 (STRAFNA1). It’s noteworthy that the proportion of all crimes which are traffic-related when using this data at least seems to be significantly higher for ethnic Danes than it is for descendants; for persons of Danish ethnic origin 67% of all convictions were traffic-related (STRAFNA1). If we trust the 58% estimate above, roughly 16% of non-Westerners got a non-traffic conviction in 2008. Note that numbers vary across sources; this measure gives 117.517 traffic law convictions out of 200.091 total convictions, which corresponds to ~59% – I don’t have a good explanation for why the sources differ here. Using the numbers from StrafNA1 only gives you 102.265 traffic law convictions in total, 14575 (7%) of which were committed by immigrants or descendants (who make up 10,1% of the population).
Of course one might argue that the ‘key figures’ above include descendants and immigrants at the age of 35 as well – but I don’t think using it as an ‘ethnic Danes’ ballpark estimate is too problematic, it’s the best I’ve got anyway. So while the fraction of non-Western descendants in Denmark at the age of 20-29 who get convicted of a crime during any given year doesn’t exactly correspond to the fraction of Danes at the age of 35 who’ve ever been convicted, it probably does correspond to more than half (~57% – ~16/28).
The ‘key figures’ for 35 year olds also included a recidivism measure; half of those convicted during their first 35 years of life got at least one more conviction. Note that if you want the hypothetical proportion of repeat offenders in the descendants group at the age of 35 to be similar to the Danish total, the number of repeat offenders in the 27,2%/~16% (year by year) group would have to be very low and the number of total convicts would have to be very high. According to this article (Danish), ‘for ordinary criminals the recidivism rate is 30 % within 2 years of release’ (“For almindelige kriminelle er tilbagefaldsprocenten på 30 procent inden for to år efter løsladelsen.”). My brief look at Statistikbanken didn’t give me any numbers on recidivism rates (the menu here is blank), and I’m not sure it’s a good idea to use this estimate in calculations here because the use of the word ‘release’ likely means that the people included in this measure served time – and most convictions do not lead to jail time (..and the recidivism rate for a previous jail convict is likely different from the recidivism rate of a person who has not served jail-time). I’m lazy and it’s probably not a good estimate to use so I won’t model or do a lot of number crunching on this stuff. However it’s safe to say from the data that either a huge number of non-Western descendants will end up having been convicted of a crime, or a quite big number of them commit a huge amount of crime each. Unless you assume a high recidivism rate it’s also safe to say that the proportion of criminals grows pretty damn fast with crime rates like that (even though the growth rate falls ‘over time’). There certainly isn’t far from 16% to 28% when you add a significant amount to the first number each period and you have a lot of periods in which to add more stuff.
Update: The numbers in this recent (Danish) publication on recidivism rates seem relevant. It confirms my suspicion that the group of people who’ve been released from jail after having served their time have quite high recidivism rates (60%) compared to other groups. On average offenders with only ‘grundskole’ (1st-9th grade), the educational grouping with the by far highest average recidivism rate, had a recidivism rate of 44%. Via that link I also came across this publication from Statistics Denmark which may be of interest – there’s a lot of data here. They haven’t written the stuff in English, but they have added English translations of key concepts at the end of the publication so that it should theoretically be possible to read the tables if you’re patient.
As to the original remark that: ‘There certainly isn’t far from 16% to 28% when you add a significant amount to the first number each period and you have a lot of periods in which to add more stuff,’ note that if we assume that the two-year descendant recidivism rate is 50% and that the traffic crime proportion estimate is correct so that ~16% of the male descendants at the age of 20-29 got a non-traffic conviction during 2008, then the proportion of descendants with a conviction after two years is 0.16 +(1-0,5)*0.16 = 24%. A 50% recidivism rate is higher than the average recidivism rate of the lowest educated group in the publication linked to above. As I said, there isn’t far from 16% to 28%.
i. I was considering covering this study in a bit more detail, but I decided against it because workplace filters probably would not like it very much – it would contain words such filters do not like (no, I’m not thinking of words like ‘sociodemographic characteristics’ or ‘multiple regression analyses’). I know a few people sometimes read my blog from work and if you’re one of them, let me just say that you should probably not read this while at work.
“The age- and sex-adjusted incidence of myocardial infarction increased from 274 cases per 100,000 person-years in 1999 to 287 cases per 100,000 person-years in 2000, and it decreased each year thereafter, to 208 cases per 100,000 person-years in 2008, representing a 24% relative decrease over the study period. […]
The proportion of patients who underwent revascularization within 30 days after myocardial infarction increased from 40.7% in 1999 to 47.2% in 2008 (P<0.001 for trend). Among patients with ST-segment elevation myocardial infarction, 49.9% underwent revascularization in 1999 as compared with 69.6% in 2008 (P<0.001 for trend). Among patients with non–ST-segment elevation myocardial infarction, 33.4% underwent revascularization in 1999 as compared with 41.3% in 2008 (P<0.001 for trend) […]
The proportion of patients with myocardial infarction who were known to have undergone troponin I testing increased from 53% in 1999 to 84% in 2004, with stable testing rates between 2004 and 2008. […]
The age- and sex-adjusted 30-day mortality after myocardial infarction decreased from 10.5% in 1999 to 7.8% in 2008 (P<0.001 for linear trend). This decrease was driven by the case fatality rate for non–ST-segment elevation myocardial infarction, which decreased from 10.0% to 7.6% (P<0.001 for trend); there was no significant change over time for ST-segment elevation myocardial infarction (P = 0.81). The multivariable adjusted odds ratio for death at 30 days after myocardial infarction was 0.76 (95% confidence interval [CI], 0.65 to 0.89) in 2008 as compared with 1999.”
Short version: Fewer people got a(n ST-segment elevation) myocardial infarction even though more people were subjected to fancy testing, more people got access to fancy treatment, and the people in the sample who got a non-ST-segment MI during the study period were less likely to die from it. But…
“observed reductions in case fatality rates could be attributable to secular trends in ascertainment of myocardial infarction and decreased severity on presentation, as well as any improvements in management of acute myocardial infarction.44 The observation that mortality after ST-segment elevation myocardial infarction (which is less influenced by the use of highly sensitive biomarkers) did not decrease over time provides support for this hypothesis.”
This could still be considered good news because if decreased severity on presentation reduces mortality it’s probably a good idea to at least have a closer look at that variable; on the other hand it’s bad news because fancy testing is expensive. Another thing:
“given the integrated medical care delivery structure in the health system that we studied and the magnitude of recent improvements in the control of risk factors within our population, our results may not be fully generalizable to other health care settings.”
Good luck finding MSM-coverage of the study including this part. I’d probably have removed the word ‘fully’. The population risk factor development during the period is a major confound.
iii. International migration: A panel data analysis of the determinants of bilateral flows by Anna Maria Mayda.
“According to the international migration model, pull and push factors have either similarsized effects (with opposite signs), when migration quotas are not binding, or they both have no (or a small) effect on emigration rates, when migration quotas are binding. It is not clear, ex ante, which one of the two scenarios characterizes actual flows. Migration policies in the majority of destination countries are very restrictive, which should imply binding constraints on the number of migrants. On the other hand, even countries with binding official immigration quotas often accept unwanted (legal) immigration.8 Restrictive immigration policies are often characterized by loopholes, that leave room for potential migrants to take advantage of economic incentives. […]
My empirical analysis also finds that inequality in the source and host economies is related to the size of emigration rates as predicted by Borjas (1987) selection model. An increase in the origin country’s relative inequality has a non-monotonic effect on the size of the emigration rate: the impact is estimated to be positive if there is positive selection, negative if there is negative selection. Among the variables affecting the costs of migration, distance between destination and origin countries appears to be the most important one: Its effect is negative, significant and steady across specifications. On the other hand, there is no evidence that cultural variables related to each country pair play a significant role. Demographics – in particular, the share of the origin country’s population who is young – shape bilateral flows as predicted by the theory. Since the effect of geography and demographics works through the supply side of the model, their impact should be even stronger when migration quotas are relaxed, which is what I find in the data. […]
Since immigrants are likely to receive support from other immigrants from the same origin country already established in the host country, they will have an incentive to choose destinations with larger communities of fellow citizens. Network effects imply that bilateral migration flows are highly correlated over time, which is what the data shows.”
iv. Via npr:
“It’s a sound you would never want to hear in real life, but this a safe way to eavesdrop. Just one warning: For the first two minutes of this video, nothing happens, nothing I could hear, anyway. Then there’s a countdown, and at 2:24 from the top … the bomb bursts; at 2:54 the blast hits.”
v. Does Thinking Really Hard Burn More Calories? Interesting piece. Unfortunately(?), “for most people, the body easily supplies what little extra glucose the brain needs for additional mental effort.”
I would be very interested in seeing a study on this including type 1 diabetics. Hard thinking for extended periods of time – like, say, a four-hour chess game or an exam – impacts my blood glucose in a very significant way; it drops like a stone if I don’t take precautions. This is despite the fact that hard thinking under such circumstances is often, as mentioned in the article, linked to stress and the release of cortisol, one of the primary functions of which is to increase blood sugar.
vi. TV from a different world:
I’ve been postponing writing a post like this about the book for ages. I didn’t really know how best to approach it. In the end I decided that I had to at least post something, and the stuff below’s where I ended up:
I guess the first thing to note is that the book is not just about ‘the fall of Rome’, even though I’ve frequently mentioned the book in that context here on the blog. Mostly it’s a book about migration patterns. The place is Europe, the time-frame covered is from the last part of the (western) Roman Empire up to the end of the first millenium. The (very) short version of the first 2-300 (?) pages is this:
(it’s the very short version!)
Keyword: Völkerwanderung (Danish: folkevandring). Before going any further, do read Razib Khan’s review of the book here (the image is from that post).
After Heather is finished talking about the migration patterns of Germanic ‘peoples’ (it’s necessary to add the ‘…’ He spends quite a bit of time talking about what the migration units most likely looked like, but I’ll not spend time on that here); Goths, Vandals, Franks etc. – I’d never even heard about half of the groups he mentions in the first part of the book – he talks about Slavic migration patterns, and after that he also spends quite a bit of time on the behaviour of the Vikings. I really liked part of the last bit of the book, the part about the early state formations in Central Europe (and beyond) and how these, according to him, were linked to immigration and development. I don’t know if Heather has read Mancur Olson, but he certainly reads as if he has and this most certainly does not make me like the book any less. As mentioned Heather’s treatment stops around the end of the millenium, though he does talk a little about later medieval migration patterns and -developments. Technically, the way I started out the paragraph could give you the wrong idea about how the book is structured so I should clarify. Heather rarely completely stops talking about group X or Y after he’s moved on to group Z, rather he always keeps coming back to stuff he’s previously covered, comparing the patterns observed; making arguments for why and how the experiences of groups X and Y were similar, which motivational factors the groups shared or didn’t share, or perhaps how the consequences of the different (?) migration strategies compared to each other. Sometimes it becomes a little repetitive, but he’s very thorough, and I liked that aspect because it also made it easier to remember the differences between Lombards, Sueves, Heruli and Sarmatians, to name but a few of the groups in question.
As I was reading the book, one thought that frequently crossed my mind was: ‘you have to start somewhere.’ The truth is that I know nothing about most of the stuff covered in the book – how much do you know about the migration patterns of Early Medieval Europe? – and I’ve only read one book now, so I still don’t know much. There’s a little overlap with other stuff I’ve read, but not much – The Classical World stops at Hadrian, and that’s a long time before the Goths really started taking to the road. Heather was a place to start, and I don’t think it’s a bad place to start. But it’s hard to evaluate the accuracy of an account when you don’t know which sources the author has excluded and basically don’t really know anything about the subject matter – you need to take some stuff on faith, and that’s harder to do when the evidence is sparse. And the evidence you need to rely upon when covering stuff during this period of European history is, it turns out, not what most historians would call optimal – as Heather puts it himself:
“The archaeological reflections of many first-millenium migratory processes […] will often be straightforwardly ambiguous in the sense that you could not be absolutely certain, just on the basis of archaeology alone, that migration had occurred.”
And often, archaeological evidence is almost all you have, even though we’re dealing with stuff taking place little more than 1000 years ago. I did not know that so few historical sources exist, but that’s apparently the way it is. Sometimes the most detailed piece of evidence that you have available for analysis is a description written by the worst enemies of the group you’re interested in knowing more about; a description written by someone who most likely was never neither within 500 kilometres nor within 100 years of any of the people whom he described. Most parts of Slavic Europe was basically prehistoric until some time after the end of the first millenium, something I most certainly didn’t know.
Heather’s account is compelling, though some parts of it I find more compelling than others; like Razib I feel a bit uncertain about the proposed link between the proportion of females within the migratory unit and long-run language transmission, but then again linguistics is yet another subject I know next to nothing about so that link may be eminently plausible to people more well-versed in such matters. I also find it a bit hard to see why the link between pottery types and language should be as strong as he would like to make it, making some of the conclusions he draws less certain than he’ll have them be. I consider this to be more of a minor point though, because it’s not obvious to me either why language similarities should necessarily carry more weight than should similarities in material culture when thinking about how to model and stratify populations of the past optimally; it depends on what you want to achieve with your model. Stated another way, I don’t see why it’s all that important which languages the people implementing the Korchak material culture spoke; the cultural diffusion was significant whether the people involved were at that point Germanic speaking or Slavs. Heather includes a few genetic data in his treatment of the western Viking diasporas, but he doesn’t even mention DNA evidence when dealing with the Slavs, which I find problematic (full disclosure: The comment #3 at Razib’s post linked to above was close to making me not buy the book). You get the feeling that Heather has set out to tell a Grand Narrative, and a natural inference to make then is that this means that he’s probably also subconsciously weighing the evidence in a manner that makes his Grand Narrative more likely to be true and, vice versa, competing models less likely to be true. But people who’ll be complaining for many years to come about what he’s supposedly written in his great work will likely get a lot of things wrong, because the argument he’s trying to make is in fact not as strong as you’d probably think from just reading about it – as he puts it himself in his last concluding chapter:
“migration should generally be given only a secondary position behind social, economic, and political transformation when explaining how it was that barbarian Europe evolved into non-existence in the course of the millenium.”
Heather is not the ‘migration is everything’-strawman that will likely be knocked down many times in the years to come, he’s not denying that a lot of other factors were likely even more important than migration – he states this fact explicitly in his book! But his book also just happens to be about migration, because he thinks immigration was important too. And to someone who does not know a lot about the subject matter, he makes a strong case for that general point. Though people who know more about the period may find his arguments less convincing than I did.
The book is well written and even though I’d have liked to read more about e.g. the material culture of the various migration units, the book is probably long enough as it is (618 pages + 76 pages of maps and notes). If you want to know more about ‘why people migrate today, in the year 2012’ this may not be the best book to get (on the other hand you could also do a lot worse), but it does contain a rather neat description of the (‘a?’) theoretical framework of modern migration studies as well as several examples of how to apply the framework in question. I’ll quote a few key passages related to this point from the last part of the book below:
“Comparative studies provide two basic points of orientation when thinking about the likely causes of any observable migration flow. First, it is overwhelmingly likely that a substantial difference in levels of economic development between adjacent areas will generate a flow between the two, from the less-developed towards its richer neighbour. […] The second point is equally basic. In the vast majority of cases, the precise motivation of any individual migrant will be a complex mixture of free-will and constraint, of economic and political motives. […] Taken together, what both of these observations stress above all is that migration will almost always need to be understood against prevailing patterns of economic and political development. […] Understood properly, and this is the central message screaming out from the comparative literature, migration is not a separate and competing form of explanation to social and economic transformation, but the complementary other side of the same coin. Patterns of migration are dictated by prevailing economic and political conditions, and another dimension in fact of their evolution; they both reflects existing inequalities, and sometimes even help to equalize them, and it is only when viewed from this perspective that the real significance of migratory phenomena can begin to emerge.”
I included this quote also to point out just how wrong-headed it can be to look at immigration as an isolated phenomenon that you can just analyze separately from other important societal phenomena. This is part of why immigration is important, and it’s a point Heather repeats again and again – the fact that this variable interacts with and is dependent upon so many other important variables of interest, and that development and immigration patterns in particular are very closely connected. A lot of people implicitly know this to be true, but many people also don’t know that they know this (or don’t know that they know that they know it…).
This is the part I haven’t said yet, but this is probably all you really need to know: I recommend the book. It has a lot of good stuff and you’ll learn a lot from it even though you have little to no background knowledge.
Before I started out this post I thought it would be the last one in the series, but at the end of the day I decided to wait with the crime data until later. This part will mostly deal with public expenditures and stuff like that. Here’s a link to the previous post in the series.
*While non-Western immigrants make out 6% of the population at the age of 16-64, they make up 10% of all people in Denmark who derive their main income from government transfers (…’are provided for by the government’ is perhaps a more ‘direct’ translation. The Danish term used in the report is: ‘er på offentlig forsørgelse’). In this framework, the concept of government transfers includes various direct income transfer programs like unemployment benefits (kontanthjælp, dagpenge), and early retirement programmes (efterløn, førtidspension), as well as governmentally subsidized employment programs (ansættelse med løntilskud, fleksjob). People working for the government are not included. (p.87-88) The ‘% of X who are provided for by the government’-measure is not the ratio of people in the sample who have received the various transfers included in the measure over the course of a year, it is rather based on a sum of all the people who have over various points in time during the year been receiving these transfers. If you have a group of one hundred people and twelve of them each received a transfer for one month during that year, that would translate to 1% of that population being provided for by the government; it’s a rough measure of the amount of ‘full-time recipients’ and should be interpreted as such. For people who receive early retirement transfers from the government the overlap between the total number of recipients over the course of a year and the number of ‘full-time recipients’ is naturally much larger than it is when it comes to transfers like unemployment benefits. (pp.87,104)
*In Denmark, two of the main social assistance programs for people who are in the workforce are ‘kontanthjælp’ and ‘dagpenge’. Kontanthjælp is the basic income support system for people without any kind of supplemental job insurance, and you can only receive it when you’ve basically depleted your assets – if you have liquid assets worth more than ~$2.000 (Danish link), you do not have the right to receive this transfer. In this context, a car you might need to drive to work is considered a liquid asset. Dagpenge is a more generous job insurance scheme subsidized by the government; the transfer payments are higher and they are completely independent of personal wealth. Approximately one in 4 (24%) of all people who receive kontanthjælp are non-Western immigrants. (p.87) 7% of all non-Western immigrants at the age of 16-64 receive kontanthjælp, whereas the corresponding number for people of Danish origin is 1,5%. (p.91)
*As the employment rates of non-Western immigrants are lower than the employment rates of people of Danish origin, it makes sense that they are also more likely to be provided for by the government. 38% of non-Western immigrants are provided for by the government, whereas the corresponding numbers for people of Danish origin and Western immigrants are 24% and 16%. (p.87)
*More than half of Lebanese-, Iraqi-, and Somali immigrants are provided for by the government. And more than half of all women from Lebanon, Somalia, Jugoslavia, Iraq and Turkey are provided for by the government. (p.87)
*Middle aged immigrants in particular have much lower employment rates than people of Danish origin at the same age, and they are thus much more likely to be provided for by the government. 60% of male non-Western immigrants at the age of 50-59 and 61% of female non-Western immigrants at the age of 50-59 are provided for by the government. The corresponding numbers for males and females of Danish origin are 23% and 26%. (p.87)
*The country of origin is an important variable when considering the likelihood that an individual immigrant is provided for by the government. 20,7% of all males of Danish origin at the age of 16-64 were provided for by the government in 2010. For Western immigrants combined it was 13,9% of males at the age of 16-64 who were provided for by the government, and for non-Western immigrants combined it was 36,7% of males at the age of 16-64 who were provided for by the government. Some more detailed numbers for male Western and non-Western immigrant populations – first the Western countries: Sweden (19,3%), Germany (18,6%), Great Britain (18,0%), Iceland (16,8%), Italy (15,7%), Norway (14,9%), Poland (12,9%), USA (11,0%), Netherlands (10,1%), France (8,8%), Romania (8,0%), and Lithuania (3,3%). The corresponding numbers for non-Western countries: Lebanon (57,8%), Iraq (51,5%), Somalia (50,1%), Bosnia-Hercegovina (45,6%), Ex Yugoslavia (44,4%), Iran (44,1%), Morocco (41,7%), Sri Lanka (37,3%), Turkey (37,0%), Afghanistan (35,1%), Vietnam (31,4%), Pakistan (29,5%), Russia (20,4%), Thailand (16,5%), Philippines (14,8%), India (9,7%), China (7,8%), and Ukraine (2%). (p.94)
*The female numbers are generally higher. I shall have to make a small digression here before I deal with those numbers: When the Danish Welfare Commission (Velfærdskommissionen) analyzed the distributionary features of the the Danish welfare system considering the gender variable, they found (Danish link) that females were on average net benefactors and males on average net contributors over an entire life span – a newborn male could, given current policies at the time the report was made, expect to pay in 0,8 million kroner ($150k) more than he’d receive over his lifespan, whereas a newborn female at that time could expect to receive 2,4 million kroner ($435k) more from the government than she’d contribute in taxes ect. Danes who are interested can read chapter 3 of this report – unfortunately I do not think an English version of that report exists. It’s likely that the relative contribution rates have changed somewhat by now, but it would surprise me a lot if they are much different now, as most of the reasons for these distributional consequenses of the welfare system have not changed much.
*Either way, as mentioned above when it comes to the females the numbers are generally higher for all groups. Of the females of Danish origin at the age of 16-64, 26,3% of them were supported by the government in 2010. For female immigrants from Western countries, the corresponding number was 18,9% and for non-Western female immigrants the number was 39,1%. Below some country-specific data – first Western countries: Sweden (24,3%), Poland (24,0%), Norway (23,5%), Great Britain (21,0%), Iceland (20,8%), Germany (18,7%), Romania (15,4%), Netherlands (14,2%), USA (12,4%), France (11,6%), Lithuania (11,5%), and Italy (11,3%). Non-Western countries: Lebanon (66,2%), Somalia (55,6%) Ex Yugoslavia (54,9%), Iraq (53,6%), Turkey (51,3%), Bosnia-Herzegovina (49,9%), Morocco (49,4%), Pakistan (45,1%), Iran (42,8%), Afghanistan (41,7%), Sri Lanka (41,6%), Vietnam (39,2%), Thailand (23,0%), Russia (20,9%), India (18,6%), China (13,9%), Ukraine (12,5%), and Philippines (11,7%). (p.95)
*The report doesn’t talk about the data much, but when analyzing the numbers above there are a couple of observations worth making here. The first is that the Swedish numbers are problematic to compare with the rest of the Western countries – it is quite likely that part of the reason why the Swedish numbers are high is that many of the ‘Swedish immigrants’ Denmark receive are in reality immigrants from non-Western countries who have used Sweden as a stepping-stone to enter Denmark, because Swedish immigration laws are much more lax than are the Danish, and it is much easier to enter Denmark via Sweden than, say, via Somalia. One other thing to note here is that the non-Western countries with high dependency rates are almost exclusively countries with large muslim populations. The non-Western immigrants from Thailand, China, Russia, India, and Ukraine in fact all ‘do better’, some of them much better, than people of Danish origin – and most of these populations are perfectly comparable to the immigrant populations from Western countries.
*Calculating net contribution rates is beyond the scope of a report like this, but I thought it would be worth including a few numbers from the publications of the Danish Welfare Commission (Velfærdskommissionen, also mentioned above). The short version is this (pp.121-122):
The graphs display the calculated net contribution to the government finances of males (the first one) and females (the second one) depending on age given the policies that were in effect at that point in time. The calculations are based on the Danish DREAM model.
Green = Danish origin.
Dark blue = immigrants from ‘developed countries’ (direct translation: ‘more developed countries’).
Turquoise = descendants of immigrants from -ll-.
Red = immigrants from ‘lesser-developed countries’.
Grey = descendants of -ll-.
They calculate in the report (p.123) that when looking at the financial net contributions to the government over the lifespan of an individual the estimated net present value (…NPV) of a male immigrant from a lesser-developed country is -0,28 mio. kroner ($50k), whereas the NPV of a female immigrant from a lesser-developed country is -4,4 mio. kroner ($800k). The NPV of a new-born male descendant of an immigrant from a lesser developed country is -0,17 mio. kroner ($30.000), and the NPV of a new-born female descendant of an immigrant from a lesser-developed country is -3,13 mio. kroner ($570k). The NPVs of immigrants from more-developed countries are 3,04 mio. kroner/$553k (males) and -0,65 mio. kroner/-$118k (females). The estimates are from 2004 and they are sensitive to changes in policy, but not that sensitive.
*Off topic, but I thought I should mention it anyway: The Florida Birth Defects Registry in 1999 estimated the lifetime costs for a child with Down Syndrome to be nearly $500,000. A Danish estimate would be much higher, but note that this cost estimate is significantly lower than the cost estimate of an average female immigrant from a lesser-developed country. In the 90es it was despite this not uncommon in Denmark to see political arguments to the effect that we needed to import immigrants from the Third World in order to save the Danish welfare state from economic ruin in the long run.
*Anyway, they remark in the Welfare Commission report that:
‘The negative contributions pr. person for immigrants and descendants from lesser-developed countries have a significant effect on the total future public-sector budget-balance problem, because both these groups are growing fast. In 2003 these two groups made up 4,7 % of the population, whereas they in 2040 are expected to make up 11,8% of the population, if the present (low) level of immigration is unchanged.’
(“De negative bidrag pr. person for indvandrere og efterkommere fra mindre udviklede lande har en betydelig effekt på det samlede fremtidige finansieringsproblem for den offentlige sektor, fordi begge disse grupper vokser med betydelig hast. I 2003 udgjorde de to grupper tilsammen 4,7 pct. af befolkningen, mens de i 2040 forventes at udgøre 11,8 pct. af befolkningen, hvis den nuværende (lave) indvandring fastholdes.” – p.125)
*As mentioned before, the overlap between the number of people who are in fact full-time recipients of a given public transfer payment and the number of people who have received a certain type of transfer payment only during a short time period over the course of the year depends on the nature of the transfer. A way to measure the average duration people receive a certain type of transfer is to divide the number of calculated full-time recipients with the number of people who have at some point during the year received the transfer. Immigrants from non-Western countries who receive temporary transfers on average receive those transfers for a longer period of time than do people of Danish origin or immigrants from Western countries and this is particularly the case when it comes to kontanthjælp: Non-Western immigrants who receive kontanthjælp on average receive it for 52% of the year, whereas the corresponding number for people of Danish origin is 40% – which is again significantly higher than the number for Western immigrants, which is 31-32% (judging from the graph on page 104; no numbers are given in the text).
The third post in the series, here are the first two posts. This part will deal with education and I must admit that it’s less data-heavy than the previous two posts, in part because I felt it was necessary to spend some time explaining how the Danish education system actually works here (and in part because I feel there’s a limit as to how much time I can justify spending on posts like these). I’ll do another post on crime later on, so this is not the last post in the series. Anyway, here goes:
*In 2010, 44% of male descendants of non-Western immigrants and 61% of female descendants of non-Western immigrants in Denmark at the age of 30 had finished an education leading to a vocational/professional qualification (see below for some notes on terminology). The corresponding numbers for people of Danish origin at the age of 30 were 73% and 79%. The education level of non-Western female descendants has increased over time; in 2004 the number was 44%. (p.65)
*It was a bit harder to translate stuff from this section than the rest because the Danish education system is a bit different from that of e.g. the US, creating a few problems related to terminology. The terminology I’ve used in this section when I was in doubt follows this source. So, which educations are in fact included in the ‘education leading to a …’ (abbreviated ELVQs in the following) measure above and which are not? ELVQs include (Danish link) various technical educations (electrician, carpenter,…), further education leading to a degree (BA, MA, PhD) as well as various other educations (office education, teaching, nursing,…). A high school degree is not included in the set, nor is a grundskoleuddannelse (see below), and if you’re a college drop-out who have not obtained a degree you’re also not included in the set of people with an ELVQ. The idea is of course that if you have an ELVQ, you have finished an education that has given you some specific skills that are useful in terms of finding and retaining employment. I decided this would also be as good a place as any to add a bit more background info about the Danish education system you might need to make sense of the numbers in the report – it’s not in there, so no page references. In Denmark the lowest attainable ‘formal education level’ (i.e. disregarding drop-outs before that point) you can have is completion of the 9th grade (grundskoleuddannelse). The graduation exam is called ‘Folkeskolens afgangsprøve’. Technically it’s a little complicated as to where exactly to put high school in terms of grades, because some people finish 9th grade and then go to high school directly (I did) whereas others take 10th grade first at the same place they took 1st-9th grade before they go to high school. The coursework in Danish high schools is the same for people who went to 10th grade before going to HS and for people who didn’t, and HS classes are a mix of both types of students. I’m not completely sure if you’re required to take 10th grade before you can enroll in a vocational(/technical) education like carpentry, but I think some of them do demand that you have 10th grade before you can start, or at least that you have taken some of the specific courses (Danish, maths). Adult immigrants without an education can take a ‘basic adult education’ which is supposed to confer the same skills as a traditional grundskoleuddannelse (in a shorter amount of time) – after they have that they can move on to a vocational education or secondary education.
*A Danish ELVQ perhaps needless to say significantly increases employment opportunities. For 30-39 year old male non-Western immigrants who had only a grundskoleuddannelse/basic adult education, the employment rate was 58% in 2010 (females: 45%, p.79). For those with a vocational education, the employment rate was 76% (females: 78%). For those with a medium-cycle higher education (‘mellemlang videregående uddannelse’), the employment rate was 82% (females: 84%). For those with a long cycle higher education (MA or equivalent/higher), the employment rate was 79% (females: 77%). (p.65 unless otherwise specified)
*When you look at the descendants of non-Western immigrants at the age of 30 years, 41% of males and 25% of females have only a grundskoleuddannelse. The corresponding numbers for males and females of Danish origin are 18% and 13%. 22% of male- and 30% of female descendants of non-Western immigrants have a vocational education at the age of 30; the corresponding numbers for people of Danish origin are 40% and 30%. When it comes to medium-cycle higher education, the numbers for non-Western descendants are 6% and 15%; the corresponding numbers of people of Danish origin are 10% and 24%. 10% of male descendants and 8% of female descendants of non-Western immigrants at the age of 30 have a long cycle higher education; 13% of males of Danish origin and 15% of females of Danish origin at that age have one. As mentioned above there’s generally a pronounced gender difference when it comes to the education of non-Western descendants, as 61% of female descendants and 44% of male descendants at the age of 30 have a ELVQ. (p.67)
*I’ll add a couple of cautious remarks here regarding how to interpret the numbers above, cautious remarks which are not included in the report (so no page references): a) There’s probably a significant power issue here when considering forecasting based on these numbers, because the number of non-Western descendants in this age group (30-years-old) is quite low – n=558 (males) and n=559 (females). b) In terms of forecasting, heterogeneity might also be an issue. It matters if you’re looking at descendants born before or after 1983-84, because the composition of new immigrants changed at that point (in the medium run, so did the composition of immigrants in Denmark as a whole). I already talked a bit about related matters in the comment section here. Non-Westerns who came before, say, 1980 mostly came here to work; on the other hand the number of non-Westerns with fugitive status or family reunification status increased dramatically after 1983 due to policy changes implemented at that point. Another dimension along which heterogeneity is relevant is the change in the country profile of descendants, change which is not only driven by a change in the immigration patterns but also related to fertility differences across subpopulations; the total fertility rate of Somali immigrants is almost twice that of Turkish immigrants (86% higher, p.26) and these differences aren’t new. It should perhaps be made clear here that even if the change in the composition of non-Western descendants in the past might have had adverse effects on some human capital measures (SES of parents, IQ…) of the descendant group ‘as a whole’, it’s far from certain that this will lead to lower educational outcomes of the group in the future – for example, political commitment to improve educational outcomes of these groups might more than make up for the other effects. From 2004 to 2011 the educational outcomes of non-Western descendants improved, but there were only 72 non-Western descendants altogether in 2004 so it’s hard to draw strong conclusions from this as we once again run into the power issue.
*One way to try to draw inferences about the future educational profiles is to look at the educational profile of descendants currently aged 20-30 years old and compare them with the historical educational profiles of the 1980-generation (the current 30-year-olds). This is done below, the first graph contains data for the current 20-30 year-olds, the second contains data for the current 30-year-olds, green = females, blue = males – the lower ones are for non-Westerns, the graphs show how big a percentage of the group had obtained an ELVQ at any given age between 20 and 30. For example, 40% of non-Western males have an ELVQ at the age of 28 (and this was also the case for the 1980-generation):
*Part of the reason why I’ve focused mostly on descendants is that it is very hard to figure out the education levels of (first-generation) immigrants, because the data the authors made use of includes only educations which are completed at Danish educational institutions. In other words, both an Italian nuclear physicist educated in Rome and a poor Sudanese woman without a primary school education will have an ‘unknown’ education level (uoplyst) in these data sets, making it harder to pinpoint just exactly what is going on. A big majority of immigrants do not have a Danish education – 77% of Western and 69% of non-Western immigrants do not have a Danish education. (p.80) However, it seems relatively clear that at least when dealing with non-Western immigrants, an ‘unknown’ education level probably most often translates to a ‘low education level’ – the employment rate of non-Western female immigrants with an unknown education level is just 33% (p80).
Thanks for the feedback.
And just a remark in case you were in doubt (most people probably weren’t, but just in case) – yes, I know very well that it doesn’t make all that much sense to report population estimates on a population of millions of people 40 years into the future down to almost fractions of a person without even including error bars (like the 6.139.618 population estimate for 2050. 618 you say? Not 617?). But the report doesn’t include error bars and I don’t feel comfortable rounding these numbers – so I decided from the start to just report the numbers they give and work with those; there are all sorts of problems related to doing anything else. So anyway, here’s some more stuff from the report:
*According to Statistics Denmark’s latest model estimates, the number of non-Western immigrants in the population will grow with 39% from 2011 to 2050, so that there will be 358.000 non-Western immigrants in Denmark. Today the number is 258.000. The corresponding increase in the number of Western immigrants is estimated at 47%. (p.48)
*The number of descendants of Western immigrants is expected to increase significantly during the period, so that by 2050 the number will be 4,4 times higher than it is today. The number of descendants of non-Western immigrants is likewise expected to increase over time, by a factor of 2,2. Despite these differences in growth rates, the number of non-Western descendants is still expected to turn out to be a little more than 3 times as high as the number of Western descendants by 2050. (p.48) This is because the current number of descendants of non-Western immigrants living in Denmark is much higher (115.597) than the current number of descendants of Western immigrants (18.016) living in Denmark. (p.49)
*The ratio of the Danish population categorized as people ‘of Danish origin’ is expected to decrease over time from 89,9% in 2011 to 84,7% in 2050. (p.48)
*The total Danish population (people of Danish origin, immigrants and descendants combined) is expected to grow by 578.990 people from 2011 to 2050, from 5.560.628 people in 2011 to 6.139.618 people in 2050. The subset of Western immigrants living in Denmark is expected to increase by 79.876 over that time period, from 170.758 to 250 634. The number of descendants of Western immigrants is expected to grow by 61.477, from 18.016 to 79.493. The part of the total Danish population growth from 2011 to 2050 which can be explained by Western immigrants and their descendants is thus equal to 141.353, which is roughly one-fourth of the total estimated population growth (24,4%) (p.49). The subset of non-Western immigrants living in Denmark is expected to increase by 99.413 from 2011 to 2050, from 258.146 to 357.559. The number of descendants of non-Western immigrants is expected to grow by 135.221, from 115 597 to 250 818.
The part of the total Danish population growth from 2011 to 2050 which can be explained by non-Western immigrants and their descendants is thus equal to 234.634, or 40,5% of the total estimated population growth. The part of the population growth over the period explained by people of Danish origin is 203.003, or 35,1% of the total population growth – despite the fact that this group makes up ~85-90% of the population over the entire time period in question. (all numbers from Tabel 1.18, p.49. They didn’t actually report these specific growth component percentages in the report, but it doesn’t take much work to calculate them from the data provided and I thought they’d be interesting to have a look at.)
*When looking at age groups, a few developments are noteworthy. In 2050, people of Danish origin are expected to make out 80,7 % of people at the ages of 40-64, vs. 91,4% today. The employment level of this age group is relatively high, compared to other age groups, which is part of what makes this development interesting – non-Western immigrants in general have much lower levels of employment than do people of Danish origin; more on that stuff below. Another factor perhaps worth noting is that the percentage of immigrants from non-Western countries above 64 years old is expected to increase from 1,2% today to 7,9% in 2050 (the expected growth of Western immigrants in that age group is much smaller – from 2,5% to 3,5%). (p.49)
*The employment rate [beskæftigelsesfrekvens] of males of Danish origin was 75,1% in 2010. The employment rate of females of Danish origin was 73,0% in 2010. The employment rate of male Western immigrants was 62,8% and the employment rate of female Western immigrants was 57,4%. The employment rate of non-Western male immigrants was 53,9% in 2010. The employment rate of non-Western female immigrants was 44,6% in 2010. (p.54)
*The employment rate of non-Western male descendants was 55% in 2010, and the employment rate of female non-Western descendants was 56%. (p.51)
*The employment rate differences between people of Danish origin and non-Western immigrants are particularly pronounced in the age group of 50-59 year olds: Whereas the employment rate of that age group was 79% for females of Danish origin, the corresponding number for non-Western female immigrants was 38%. (p.51)
*The employment rate difference between males of Danish origin and male immigrants of non-Western origin was 21 percentage points in 2010, whereas the employment rate difference between females of Danish origin and female immigrants of non-Western origin was 28 percentage points in 2010.(p.51)
*In 1996 the difference in the employment rates of males of Danish origin and those of male immigrants of non-Western origin was 40 percentage points. The corresponding difference in the employment rates of females of Danish origin and those of female immigrants of non-Western origin was 44 percentage points in 1996. 1996 was two years before the first election where immigration policy became a major factor (though in terms of formation of the government, it did not decide the election – that didn’t happen until 2001).
*The previous 2008-report from Statistics Denmark contained a nice illustration of how the employment rate differences between non-Western immigrants and Danes vary with age and I decided to include it in this post – you can find it at page 65. The numbers are from 2007. Dark-blue = males, light-blue = females, the y-axis is the employment rate difference between people of Danish origin and non-Western immigrants measured in percentage points, the x-axis is age:
So, to take an example, the employment rate of non-Western female immigrants at the age of 40 was approximately 35 percentage points lower than the employment rate of females of Danish origin at the age of 40 in 2007.
*Back to the 2011 report: From 1996 to 2008 the employment rate of non-Western immigrants increased significantly; the male employment rate increased from 40% to 63% and the female employment rate increased from 26% to 50%. Here are two graphs from the report (p.53), click on them to view them in a higher resolution – the first one is on male data, the second is on female data:
“Indv., vestlige lande” = Immigrants from Western countries
“Indv., ikke-vestlige lande” = Immigrants from non-Western countries
“Dansk oprindelse” = Danish origin
“Eftk., vestlige lande” = Descendants, Western countries
“Eftk., ikke-vestlige lande” = Descendants, non-Western countries.
In both cases, the y-axis is the employment rate.
*Country of origin is a very important variable – not all Western countries are the same, nor are all non-Western countries the same. The employment rate of immigrants from the Netherlands is the same as that of people of Danish origin – 74%. The Polish immigrants have an employment rate of 66%, and so do the British. These employment numbers are much higher than those of the immigrants from the US, where the employment rate is just 49%. (p.57) However, ‘many young Western immigrants come to Denmark to study, they’re often only here for a short while and return home after they’ve finished their coursework here’ (paraphrasing some of the relevant remarks on p.76). The authors don’t go into any details about the US immigrants in the report, but I think it’s safe to say that they are more likely to be university students than are immigrants from, say, Poland – the lower employment rates probably shouldn’t be all that surprising. Employment rates on their own don’t care about differences in labor force participation rates.
*Non-Western immigrants generally have lower employment rates, and it’s also among these countries of origin that we find the subpopulations with the lowest employment numbers. The bottom three are Iraq (36% employed), Lebanon (35% employed) and Somalia (31% employed). Less than one in four of female Lebanese immigrants in Denmark are employed. But worth noticing here is also that some of the non-Western countries do quite well: 67% of Ukrainians are employed, and so are 63% of the immigrants from Thailand. (p.57)
A table from the report (p.57), click on it to view it full size:
As I know a lot of terms might cause problems I decided to add an explanation. It was either that, translate everything and make my own table or report some more of the numbers in the text – I decided you should have the data but I didn’t want to spend a lot of time reconstructing that table. You can probably figure out a lot of the stuff I’ve translated below on your own, but in my experience it’s very nice to not have to be the least bit in doubt when reading tables like these. If you have questions, ask:
Title: Employment rates of 16-64 year olds. 2010.
“Antal personer”: Number of people. (antal = number)
“Beskæftigelsesfrekvens”: Employment rate.
“Mænd” = Males.
“Kvinder” = Females.
“I alt” = Total/combined.
“Indvandrere, vestlige lande”: Immigrants, Western countries
Nederlandene = The Netherlands
Storbritannien = Great Britain
Polen = Poland
Rumænien = Romania
Sverige = Sweden
Tyskland = Germany
Litauen = Lithuania
Norge = Norway
Island = Iceland
Italien = Italy
Frankrig = France
“Indvandrere, ikke-vestlige lande” = Immigrants, non-Western countries
Kina = China
Tyrkiet = Turkey
Rusland = Russia
Indien = India
Jugoslavien = Jugoslavia
Marokko = Morocco
Filippinerne = The Philippines
Irak = Iraq
Libanon = Lebanon
The specific 30 countries were chosen because those were the 30 countries of origin with the highest amounts of 16-64 year olds.
The employment rate is somewhat dependent on how long people have lived here, so the authors also decided to split up the data using that variable. Again, click to view it full size:
From page 60. Additional explanation:
Title (roughly): ‘Employment rates of 16-64-year old immigrants – distributed based on the amount of time spent in Denmark. 2010.’
‘Opholdstid’ = Time spent in Denmark.
‘Under 3 år’ = Less than 3 years.
‘3-6 år’ = 3-6 years. [I think you get the picture…]
‘Over 15 år’ = More than 15 years.
*Do note when interpreting the employment numbers of e.g. Filipino women that people who are employed as au-pairs are not counted as employed. (p.59)
*’Even when taking into account differences in the amounts of time spent in the country, there are still big differences. Immigrants from Iraq, Lebanon and Somalia who have been in Denmark for at least a decade have employment rates between 30% and 41%. Immigrants from the Philippines, China and Thailand who’ve been in Denmark for at least a decade have employment rates between 67% and 75%.’ (p.59)
I’ll post at least one more post on this subject. I will probably add the posts together into one single post when I’m done.
The central Danish statistical office, Statistics Denmark, has just published a report with a lot of data on Danish immigrants, Immigrants in Denmark, 2011. I thought some of the non-Danes reading along might appreciate a post in English on this subject.
At the site, they’ve given no indications that they’re planning to translate this, so I don’t think an English version of this material is coming up anytime soon. My translation of the stuff is better than what you’d get from google translate, but do remember that I’m not exactly a professional translator. I’ve decided to page-source every bit of data for this reason, so that it’s easier to go have a look for yourself if you’re in doubt. It was most convenient for me to page-source the pdf version pages, not the official page numbers at the top of each page in the report. Don’t think of the statements below as direct quotations from the report – I’ve frequently had to reformulate the expressions used in the report. If something’s unclear, please ask away. Anyway, let’s start:
*10,1 % of the Danish population are immigrants or descendants of immigrants. (p.13)
*Immigrants make up 7,7% and descendants make up 2,4%. (p.13) [A small note here: The report only explicitly mentions the 10,1% and the 7,7%, not the 2,4% – but I think it’s safe to assume that this is a simple subtraction problem and that it makes good sense to post that number as well just for completeness.]
*60,2% of all immigrants are from non-Western countries. (p.13)
*66% of all immigrants and descendants are from non-Western countries. (p.25)
*The number of non-Western immigrants has almost sextupled since 1980. (p.14)
*From 1980 to 2011, the number of non-Western descendants has increased from 7.653 to 115.597. (p.15)
*The number of descendants of Western immigrants grew by 70% from 1980 to 2011. (p.15)
*The immigrants living in Denmark come from more than 200 countries. (p.15)
*The distribution is not uniform. Immigrants from the top 12 countries (in terms of the number of immigrants living in Denmark) make up 50% of all immigrants. (p.15)
*Turkey is at the top of the list with 32 479 immigrants living in Denmark. (p.15)
*5 out of the top 12 countries are Western countries (Germany, Poland, Norway, Sweden, GB). 7 are Non-western countries (Turkey, Iraq, Bosnia and Herzegovina, Iran, Lebanon, Pakistan, ex-Jugoslavia). (p.16)
*There’s significant variation in the age distribution of immigrants from different countries. When looking at the top twelve, 20% of the Western immigrants in that group are 60 years old or older, whereas only 10% of the non-Western immigrants in the top-twelve are 60 years old or older. (p.16)
*As to the Poles, they’re an interesting case because they’re quite different from the rest of the Western immigrants. They’re the third largest immigrant population (26 580) in Denmark – the number of Polish born people living in Denmark is higher than the number of immigrants from Sweden and Great Britain combined – and more than half of the Poles (53%) are between 20 and 40. 68% of the Polish immigrants are between 20 and 49 years old. 10 % of them are 60 years or older. (p.16)
*When looking at the descendant populations living in Denmark, 11 out of the top 12 countries are non-Western countries. More than one in five (21%) of all descendants living in Denmark are descendants of Turkish immigrants. Lebanon and Pakistan are next on the list, with 9% and 7% respectively. (p.17)
*Most descendants are quite young. 41% of them are below the age of 10, and only 10% have reached the age of 30.
[I used to comment on this fact back when I did political discussions, because it is often overlooked or simply ignored in discussions about what might be termed the demographic potential of descendant populations. We have no idea how many children descendants will end up having, and it makes no sense to try to draw strong conclusions out of sample from the data sets that are available now. Please have this in mind when we get to the forecasts later on. Putting the above numbers in context, the average age of women having their first child in Denmark was 29,1 years in 2010 (Statistikbanken, FOD11). I also urge people to remember here that the growth rate of population segment X in a population doesn’t just depend on the total fertility rate differential, but also on age of birth differentials. If women from population segment X get children at the age of 30 and women from population segment Y get children at the age of 20, population segment Y will grow faster than population segment X, even if every single woman in the two population segments have the same amount of children. This remark is relevant because non-Western immigrants as a rule get children at a lower age than ethnic Danes. Females of Danish origin get on average 0,21 children during the period of their lives where they are 20-24 years old. For all non-Western female immigrants, the corresponding average number is 0,35. For Lebanese women, the number is 0,72. (pp. 27-28)]
*Western descendants are much older than non-Western descendants, on average. [worsening the data problems mentioned above. Especially if you mix up the Westerns and non-Westerns – does it make sense to extrapolate birth rates of Turkish descendants in 2015 from the historical birth rates of descendants of Norwegian women?] One third of the descendants of Western immigrants are above the age of 30, whereas only 6% of the descendants of non-Western immigrants are that old. (p.18)
*Descendants from Turkey, Pakistan, Jugoslavia or Morocco make up 77% of all 30+ year old descendants from non-Western countries. (p.18)
*The total fertility rate of Somali immigrants in Denmark is 3,937. (p.26)
*In the period 2006-2010, there were an average of 64.056 living births pr. year. In the same period, there were an average of 5.860 (9,1%) children born every year of non-Western immigrants and an average of 2.310 (3,6%) children every year born of Western immigrants. The average annual number of children of descendants over the time period was just 961. (p.26)
*The report has some stats on family patterns and the degree of observed endogamy. When it comes to male immigrants from Western countries who are classified as being in a relationship, in 59% of the cases the partner is of Danish origin and in 37% of the cases the partner is an immigrant from a Western country. When it comes to the female immigrants from a Western country, 63% of the partners are of Danish origin and in one-third of the cases it’s a Western immigrant. The pattern is different when it comes to immigrants from non-Western countries. For male immigrants from non-Western countries, 13% have partners of Danish origin and 80% have partners from a non-Western country. For female immigrants from non-Western countries, 28% have partners of Danish origin and 68% have partners of non-Western origin. Interestingly, when it comes to descendants Western immigrants are more likely to have a partner of Danish origin than are first generation immigrants (83% and 85% for males and females respectively), whereas this pattern is actually reversed for females from non-Western countries, where descendants are less likely to have a Danish partner than are first generation immigrants (19% of females who are descendants of immigrants from non-Western countries with a partner have a partner of Danish origin, whereas the corresponding number for the first generation non-Western female immigrants is 28%.) 3 out of 5 non-Western descendants who are in a relationship are in a relationship with a non-Western immigrant and 18% of them have a partner who’s also a descendant of immigrants from a non-Western country. (all numbers above from Tabel 1.9, p.32)
*When it comes to the non-Western females who find Danish male partners, few of these women come from the major immigrant countries. Of the 19.981 female non-Western immigrants with a partner of Danish origin, females from Thailand, Philippines, Russia, China, Brazil and Ukraine make up 11.644 of them – 58%. (p.33)
*Females from Thailand and Philippines alone make up 39% of the non-Western females who have partners of Danish origin. (p.34)
*When it comes to females from Turkey, Pakistan and Iraq, only 2% of them have a partner of Danish origin. (p.34)
*97% of female Turkish immigrants with a partner have a partner of Turkish origin. 94% of Pakistani females in a relationship have a partner of Pakistani origin. (p.35)
*88% of Turkish descendants in a relationship have a partner of Turkish origin. (p.37)
*Today the country from which Denmark receives the largest number of immigrants is Poland. Denmark received 3850 Polish immigrants in 2010. (p.38)
*(not direct citation but paraphrasing…)’Immigrants from Western countries like USA, Spain and Italy rarely come to Denmark to live here permanently and a large share of them leave Denmark again.’ – ‘This is not the case for non-Western immigrants.’ (p.40) Some data: 77% of the Poles who came to Denmark in 2002 had left the country by January 1st, 2011. 88% of the immigrants from the US who came in 2002 had left Denmark by 2011. On the other hand, only 9 percent of Iraqis who came in 2002 had left by 2011. 24% of the Turks who arrived in 2002 had left by 2011. (all numbers from table, p.39) [the 9% number is interesting also because during that time period, Denmark actually had various policies (Danish links) in place where Iraqis who decided to leave Denmark could get a one-time cash prize for doing so.]
This post dealt with roughly the first 40 pages of the report. The report has 153 pages. So there’s a lot of stuff to cover – there’s also data on education, crime, employment, ect. I might write another post or two on this subject if people liked this one.
Major related hint: If you’d like me to write another post on this, tell me, either by using the rating system or by commenting. If I don’t get positive feedback, I probably won’t do any more work on this – it adds a not insignificant time component to not being able to just quote directly from the report because the stuff needs to be translated as well.
Plamus linked to it in the comments section and I’ve seen it linked elsewhere as well, it’s an interesting paper.
Here’s the abstract:
“A recent line of research demonstrates that cognitive skills—IQ scores, math skills, and the like — have only a modest influence on individual wages, but are strongly correlated with national outcomes. Is this largely due to human capital spillovers? This paper argues that the answer is yes. It presents four different channels through which intelligence may matter more for nations than for individuals: 1. Intelligence is associated with patience and hence higher savings rates; 2. Intelligence causes cooperation; 3. Higher group intelligence opens the door to using fragile, high-value production technologies, and 4. Intelligence is associated with supporting market-oriented policies. Abundant evidence from across the ADB region demonstrating that environmental improvements can raise cognitive skills is reviewed.”
I don’t buy 4 at all unless/before much more work is done in that field. Now it mostly just reads ‘I read Caplan’s book and people I know talk about it so I should probably mention it in my study’ to me. The other parts I don’t have strong opinions about. Below’s some stuff from the study and my remarks. Here’s Figure 1 from the paper, you have log-GDP pr. capita up the y-axis:
The ‘PRC’ in the corner is China, and there are plenty of reasons (the name of the most significant one is Mao) why you’d think it makes good sense that they haven’t managed as well as the theory suggests. The IQ-effect is huge: “Jones and Schneider […] found that across countries […]: 15 IQ points is associated with a 150 percent increase in productivity.” If you think simply in terms of labour input, this finding would suggest that in a country with an average IQ of 115, 2 average workers can be expected to add the same value to a product as (‘do the work of’) 5 workers living in a country with an average IQ of 100. Yet the private returns related to that productivity difference is very small; in the paper they mention an estimated wage differential of just 13 percent.
There’s a lot of stuff in the paper, I’ll just go through a few interesting bits I found. Here’s some stuff on environmental factors and their influence on IQ:
“there is a vast public health literature on environmental correlates of intelligence, and many of these papers study nations in Asia. A study of excessive fluoride in Indian drinking water found a 13 IQ point-difference between children “residing in two [separate] village areas of India with similar educational and socioeconomic conditions” (Trivedi et al. 2007, 178). If even half of this relationship is genuinely causal, and if intelligence has some of the technological and political spillover effects discussed below, then public health matters are of first-order concern for economic development.”
The impact of just two environmental factors of that size could in theory reduce the mean intelligence of a population with Mensa-level average IQ to that of current-day Japan. These effect sizes are huge.
“Arsenic and fluoride exposures are also associated with low IQ in the People’s Republic of China’s (PRC) Shanxi province (Wang et al. 2007, 664), even when comparing “groups [who] lived in rural areas with similar geographic and cultural conditions and a comparable level of socioeconomic development.” High arsenic exposure was associated with a 10-point IQ gap, and high fluoride exposure with a 4-point gap. In both cases, the “normal” group had an IQ of 105, 5 points above the US mean.
In the Visayas region of the Philippines, Solon et al. (2008) found evidence that lead levels reduced the IQ of children. In their study, one microgram of lead per liter of blood was associated with a 2.5 point reduction in the verbal IQ of older children, and a 3.3 point reduction in the IQ of young children. In their sample of children, the levels of lead in the blood averaged 7.1 micrograms per liter, so lead exposure could be costing the average child in this sample 15 IQ points even under conservative estimates.”
The role of nutrition is mentioned in the paper, but they don’t go much into the specifics. I’m pretty sure that’s one of the main things holding India back on the IQ-scale of Figure 1.
I think both point V and VI are only/mainly there because of the agenda of the authors and I hate that kind of thing. V is almost pure speculation using an already (with respect to which conclusions can be drawn from the findings) speculative voter preferences model from the US to talk about East Asia. Smarter people will be more likely to support free market policies if they think they’ll gain from it and they get a say in the matter, which depends mainly on how the local government decides to split up the cake. Show me a group of American professors of theoretical physics pushing for more free market policies in education (fewer gov. subsidies). No, that’s not the relevant margin, but to take an extreme example in the opposite direction, in a standard median voter model you could have an IQ increase of 30 points of the 4 top deciles having no effect on policies whatsoever, if the intelligence of the median voter is unchanged. Yeah, you might argue the IQ effects are to be had on the other side of the distribution, but model symmetry means that you could make the same argument and apply the change to the 4 lowest IQ deciles. Conceptually they probably just take up this subject to encourage further research, but I’m one of those people thinking that Caplan is drawing way too strong conclusions from his findings already, and using IQ proxies to speculate about effects in countries looking nothing like the US, having wastly different political systems – well, that’s just not very smart. Point VI is of the same kind – it smells of ‘we want to push this idea, how can we include it in the paper’-motivation. It mentions one way to increase a country’s IQ – immigration. From the paper:
“Even if scientists and public health officials quickly reach their limits in raising a person’s IQ—again, not a foregone conclusion — we still have a reliable tool for raising a nation’s IQ. Encourage immigration by individuals with higher average intelligence. Many countries implicitly do this by permitting high-skilled immigrants to enter and work legally.”
Nowhere in the paper is it mentioned that this is most likely a zero-sum game. One country’s gain is another country’s loss. And the ‘many countries implicitly do this…’ part is correct but only half of the story, as many countries, especially Western countries, also implicitly do the opposite – import massive amounts of low-IQ immigrants (and also implicitly form/maintain policies which encourage these people to have a lot of children, lowering national IQ and future human capital even further).
“In 2008, employment rates were 82% and 77% for native Danish men and women respectively, compared to 63% and 50% for non-Western immigrant men and women (Statistics Denmark, 2009).”
Here’s the link.
This is not a new idea, at least not to me, but I think this factor, when evaluating the relative merits of the US’ vs European immigration policies and results, must be somewhat overlooked in general if it’s the first time Razib thought about it (and thus – had not encountered the idea elsewhere before):