Hello There, Guest!  

[ODP] Crime among Dutch immigrant groups

#1
Author:
Emil O. W. Kirkegaard

Title:
Crime among Dutch immigrant groups is predictable from country-level variables

Abstract:
Data about crime suspects among Dutch immigrant groups was obtained from an older Dutch-language report. Correlations with predictor variables varied a lot. The mean correlations were: National IQ (-.55), age heaping, 1900 (-.44), Islam prevalence (.24), international general socioeconomic factor (-.45). Crime among immigrant groups is substantially predictable from country-level variables in the Netherlands.

Files:
https://osf.io/69pcd/files/

Previous thread:
http://www.openpsych.net/forum/showthread.php?tid=183

Edit 14th Jan. Updated first post with new title+abstract per request.
 Reply
#2
Quote:Previous studies have found that immigrant crime proneness and other socially important traits are strongly predictable from their or their parents' countries of origin, se e.g. [1, 2, 3, 4] and further references cited by these papers.

Should be : see e.g.,

Quote:Generally, cognitive measures correlated stronger in the smaller samples, perhaps being inflated by sampling error.

I think I understand what you mean, but you should be careful when you use this word, because some people may believe that you affirm that sampling error causes correlation to be higher, when it's the opposite.

Quote:The simple correlations of crime variables and predictors is shown in Table 1.

Table 1: Correlations of predictor variables with crime variables.

Since young men commit most crimes, the most important variables are the crime ages 12-17 and 18-24 for men, both of which show strong correlations with predictors.

3 Predictor vector correlations

etc...

Same problem as before. If you're using correlations, and not regression, the word predictor may be confusing for some people.

Quote:Another way to examine it, is to calculate the mean relative crime ratios by age and generation groups.

Is it crime ratio, or is it crime rate (as described in your table 6) ?

Your table 4 shows that the 2nd gen turkish immigrants in Denmark have much higher crime rate than 1st gen. But why so ? Even given the transferability hypothesis, which says that "immigrants tend to keep their psychological traits when they move to a new country" I'm pretty sure that it does not predict that the crime rate will increase so dramatically among 20-29 aged men. I don't think it's due to sample size either. So what happened there ?

Quote:Another way to use Table 2.13 on page 128 in the report is to note whether there are differences between first and second generation.

It must be made clear that you're referring to the 2005 Dutch report of crime.

Quote:One explanation could be that the Dutch have more success with integration of immigrants.

I think you should cite one reference or two.
 Reply
#3
Meng Hu,

This is a robust finding we see in other immigrant groups. Many immigrants come from societies where the State has only recently monopolized the use of violence. In such societies, there is still a strong belief that each adult male has the right to use violence on his own behalf and for his kin, and not just for self-defence. A man may legitimately use violence to defend his "honor" or his reputation, or simply to act pre-emptively against a perceived threat. Since the State has not pacified social relations, at least not until recently, there has been no selection to lower the capacity for personal violence, especially among young males.

In this kind of social setting, personal violence is held in check by the threat of retaliation, not only by the victim but also by his kinfolk. There is thus a kind of dynamic equilibrium that keeps violence within reasonable limits (although it continually spills over against unprotected outsiders). Moreover, this equilibrium becomes supported by various cultural prohibitions that further limit the circumstances where antisocial behavior is considered legitimate. People may fear retaliation from the spirit world, for instance. Or there may be taboos of various sorts that discourage doing harm to certain classes of people (children, holy men and women, etc.).

In short, many immigrants come from an environment where violent behavior is possible but nonetheless subject to external constraints, either the threat of retaliation or various cultural prohibitions.

The situation changes when people leave these societies and enter one where such constraints are weak or nonexistent. Violent, antisocial behavior can thus manifest itself more easily. This is less so for the first generation, who are still inculcated with cultural prohibitions that limit violence and other forms of antisocial behavior. This is more so for the next generation, who are more assimilated and correspondingly feel less restrained in their behavior. Ironically, the more assimilated the immigrant community becomes, the more it will diverge behaviorally from its host society.

This divergence will take the form of violence against soft targets. In general, the non-white on white violence we increasingly see is not motivated by hate. It's motivated by a justified perception that "whites don't fight back." Whites are soft targets.

In Western societies, the State has monopolized the use of violence, and in theory the State is supposed to keep such violence in check. It can't because the police were never designed for a situation where almost every young male is a potential threat. Moreover, any effort to contain such violence will disproportionately affect young males from certain ethnic groups. Accusations of racism will inevitably follow.
 Reply
#4
The increase from 1st to 2nd is not predicted by ST hypothesis, but it is consistent with it. Perhaps Peter's theory is right. A simpler theory is that 1st generation is more thankful to Denmark for taking them in, while 2nd grew up here and do not feel there is anything to be thankful for since they are not familiar with the generally worse environments in their origin countries.
 Reply
#5
Meng Hu,

Thanks for reviewing. Replies to your comments below.

Quote: Should be : see e.g.,

Fixed.

Quote: I think I understand what you mean, but you should be careful when you use this word, because some people may believe that you affirm that sampling error causes correlation to be higher, when it's the opposite.

That is exactly what I mean. Sampling errors go in either direction.

Quote: Same problem as before. If you're using correlations, and not regression, the word predictor may be confusing for some people.

Singular regression is identical to correlation. I don't see how it could be confusing. Whether to call one variable "predictor" ("independent variable" etc.) is only a matter of which variable is intended to predict the other. Clearly, we are not trying to predict e.g. national IQ based on crime rate in the Netherlands, so it is clear which is which.

Quote: Is it crime ratio, or is it crime rate (as described in your table 6) ?

Your table 4 shows that the 2nd gen turkish immigrants in Denmark have much higher crime rate than 1st gen. But why so ? Even given the transferability hypothesis, which says that "immigrants tend to keep their psychological traits when they move to a new country" I'm pretty sure that it does not predict that the crime rate will increase so dramatically among 20-29 aged men. I don't think it's due to sample size either. So what happened there ?

Fixed. It is a rate as the caption says. Table 6 is Dutch data, not Danish.

I don't know. It is not a new thing, it has been noted by commentators for years.

Quote: It must be made clear that you're referring to the 2005 Dutch report of crime.

Inserted "2005 Dutch".

Quote:I think you should cite one reference or two.

For what? It is a proposed explanation, not an actual claim that it is happening or that others are saying it is happening.

Draft updated to version 5.
 Reply
#6
Emil, maybe you should see if you can incorporate a discussion section regarding Peter Frost's comment. It would also be good if you can find some ways to include the research on personality differences by nations, from David P Schmitt and Robert R McCrae. If I remember correctly, there was an article (I don't remember which one) where McCrae says that some immigrant groups remain similar in behavior to people of the same racial stock who still live in their country of origin, although they differ in some aspect of their personality.

I say this, because until now, the ST hypothesis has only been applied to IQ, and crime. But I wonder if you can generalize the ST hypothesis to all kind of behaviors.
 Reply
#7
I have played around a bit with various other factors, including OCEAN traits and Hofstede's cultural dimensions.

I think the current national OCEAN estimates are not very good, and thus I've been reluctant to rely much on them.

I am not too familiar with the correlations of personality with crime. From looking at the:

Ellis, L., Beaver, K. M., & Wright, J. (2009). Handbook of crime correlates. Academic Press.

There are some OCEAN traits which are related to crime.

6.1.7 lists studies with extraversion and crime, generally positively related.

6.1.12 lists studies with neuroticism. They are mixed between sig. pos. and null findings.

Wiki mentions some studies with regards to conscientiousness, which is negatively related to crime. https://en.wikipedia.org/wiki/Conscienti...e_outcomes

This paper mentions C and A as correlates. http://www.researchgate.net/publication/...01e9000000

There is a dissertation here, but judging from my quick skimming of it, it is not very good. From Acknowledgements "First and foremost, I thank God for the perseverance to do this as well as the wisdom and knowledge that I have gained in this process.Getting this degree challenged everything I ever thought I knew, and pushed me beyond what I thought were my limits.".

https://dspace.iup.edu/bitstream/handle/...sequence=1

In any case, it found the usual low A low C predicts crime. See e.g. Table 17.

So, we should perhaps reexamine all the previous crime data to see if A and C has incremental predictive ability beyond Islam and g. We could meta-analyze all the results.
 Reply
#8
I know that the link between personality and criminality is not well established. For example, Richard Lynn said that psychopathic personality can be an important factor behind delinquent behaviors, and he said that blacks have higher "psychopathic score" than whites, but this view has been now challenged by meta-analytic reviews (although I didn't read them carefully to be fully able to make a critical comment on their works).

However, what I was thinking is that, if personality changes little among immigrants, it is predicted by ST hypothesis, and "may" be an explanation as to why the criminality rate does not change so much between 1st and 2nd generation immigrants. That said, with regard to Frost's comment, I would like to see how the 3rd generation immigrants are doing on delinquent behaviors.
 Reply
#9
There are too few 3rd gens in Denmark right now for reliable data. They are being closely followed by the stats agency. They do not have higher grades than 2nd gens, but they have somewhat higher lower rates of unemployment. But the datasets are pretty small, only a few hundred persons above age 15 in total. The group however is growing very fast, so in a few years we will know about crime rate among 3rd gens.
 Reply
#10
Any more objections to this paper? :)
 Reply
 
 
Forum Jump:

Users browsing this thread: 1 Guest(s)