On Japan's immigrant population and their crime rates

Japan's big drops in crime since 2002 are largely attributable to its aging population. Excess crime by specific foreign residents are largely attributable to their much younger ages.

Much has been said about Japan's plummeting birth rate and aging population, to the point readers of current events in Japan seem to have become numb to it. People have been made well aware of how the country's demographics will affect its debt and and social security payments in the future, but those issues still seem far away and abstract for most people.

What many people don't realize is how much Japan's aging population affects the country right now, and in ways you might not expect. One such way is Japan's crime rate, which, despite a slight uptick post-pandemic (more on that in a future post), has fallen dramatically since its peak in 2002:

(Source)

Even less explored is the way attempts to fight these demographics through immigration could turn the clock back in Japan not just on its economic malaise, but on crime rates too. 

The purpose of this post is not to warn against immigration; I think the following data and charts will show that accepting immigrant labor is a worthwhile trade off. Once accounting for their younger age, Japan's foreign residents are on net much better behaved than many locals believe they are. But a trade off it remains, and people should understand what these trade offs are when discussing immigration and crime.

Japan's crime peak can be attributed to its second baby boom. Its decline in crime can be attributed to the aging of that cohort.

It is well documented that crime rates are highest among youth, especially young men. As US crime data compiled by Pinkerton shows, under the age-crime curve, people are most likely to commit crimes starting in their late teens. This tendency continues through their 20s but begins to drop off steeply as they approach 30 and continues to drop though the remainder of their lives.

(Source)

While Japan has always had low crime rates relative to other advanced nations, they are not immune from this trend. Consider Japan's last demographic gasp of the 20th century, the "second baby boom" which occurred between 1971-74.

(Source)

If the age-crime curve applied to Japan, we would expect to see a big spike in crimes as that cohort reached its twenties. And that is in fact exactly what happened. Reported crimes began to climb sharply around 1995, when that cohort was aged about 21-24. Reported crimes in Japan peaked in 2002, when the cohort was aged about 28-31, and fell steadily for the following 20 years:

(Source)

Even if one could argue that people in their 30s, 40s and 50s still commit crimes, few would argue that many senior citizens do. As Japan's working age population goes down as a percentage of the total population, so too do crime rates. In fact, the correlation of those two things is 0.981, which is extremely high. Below are the results of my own analysis:

While the drop in since 2002 has been dramatic, it's important to stress that even at its peak, Japan was very safe by international standards. It's hard for international observers to even fathom how safe Japan is today. Compare murder rates by country:

So it is important to understand that although the crime rate fell a lot in the past years, even if Japan returned to its crime rates in the early 2000s, it would still be one of the safest countries in the world by a wide margin.

Does immigration make Japan less safe?

Lately there has been a surge of claims on social media that immigrants are making the Japanese unsafe. These claims contributed the far-right anti-immigration party Sanseito picking up 14 seats in the Diet's upper house. How warranted are these fears? Have immigrants increased crime in Japan?

Before going any further with this, it's important to give the short answer- No. As NHK reported:

In the space of about 30 years, the foreign population has nearly tripled, from about 1.3 million to 3.7 million. Meanwhile, the number of people arrested has been on a downward trend, from 14,786 in 2005 to 9,726 in 2023. [An expert] points out, "Even if we look at the trends over the past 30 years or so, even though the number of foreigners has been increasing, the number of criminal offenses committed by foreigners has actually decreased."

It is possible to steel man anti-immigrations arguments though- these figures are for all immigrants, regardless of nationality. Some could (and do) argue that crime rates could vary a lot by the nationality of the immigrant, and that if Japan allows many more such immigrants from those countries, the country would become more dangerous. Their higher rates are held up as evidence residents of those nationalities are incompatible with Japanese culture.

Recently Japan's refugee Kurd population have attracted a lot of attention in the right-wing press due to sexual assaults and other crimes committed by a handful of young men among their numbers. But the total Kurd refugee population in Japan is only about 4500 and grouped together in just a few communities, making a robust quantitative analysis difficult, and due to the political blowback from these incidents it is unlikely Japan will accept any more such refugees anytime soon.

According to Japan's police agency, most foreigners who commit crimes in Japan belong to one of five nationalities, the only nationalities they give individual statistics for: Vietnamese, Chinese, Thai, Fillipino and Indonesian. Much of the national news regarding "foreign crime" regards residents of these nationalities.

It may surprise international observers that people from these countries are considered troublemakers in Japan. In other countries, East Asian and Southeast Asian immigrants are often considered "model minorities" with very low crime rates (Indeed, Vietnam and China also have low murder rates that while a bit higher than Japan are still stellar by global standards).

However, when the immigrant population approaches 4 million in a country of 127 million people, raw numbers of crimes can start to look large if they have been separated from their broader context. Below is the type of report the police typically give for foreign crime:

The drawback of data in this format is that it is not given in reference to population sizes. At first glance crimes committed by some groups appear to have exploded over the last ten years, but these numbers do not account for the increases in the population since then. When we adjust these numbers to be proportionate to the growing populations of these groups, we can get a better picture of how often foreigners (per 100,000) are arrested for crimes relative to Japanese nationals (per 100,000):

Readers should note a few things:

1. Arrest rates for all foreign residents who do not belong to the above five groups is lower than the national average. As an example, crime rates for North Americans and Europeans are far below Japanese national averages.

2. Of the groups listed, Indonesian and Filipino residents also commit fewer crimes than the Japanese national average. Chinese residents are just 8% above the average, which seems pretty negligible.

However, it is true that Vietnamese and Thai residents are arrested at much higher rates.

Some of this is likely due to their economic conditions. For example many foreigners on Japan's notorious "technical intern" visas are Vietnamese. They must take out high loans to finance their move to Japan, but are stuck working with the company that sponsored the visa with no option to switch employers. If they have a problem with that employer  and leave (which is common due to "black companies" that exploit and mistreat foreign labor), they have no way of legally making money, and may resort to crime to repay their debt and finance their return home.

However, there is another reason that should be considered- Japan's foreign residents are much younger than Japanese nationals. 

Recall that Japan's crime reports peaked in 2002 when the working age population was about 68% of the population, and fell by about 75% by the time the working age population was down to 58%:

The corollary is likely also true- were Japan to suddenly have more young people, the crime rate would likely rise again. And that is what happens as Japan receives more foreign residents. According to population statistics the largest cohort of foreign residents is 25-29 with a median age of about 28-29, whereas the median age for Japanese nationals is about 48-49. 

To use the metric of working age population above, the working age populations for the above groups are as follows:

A drawback of these statistics is that the 15-64 age range is so broad that it doesn't capture the proportions of immigrants in their 20s in each nationality, making these groups seem more similar to on another in age than they really are. But the contrast to the native population is clear- these are all dramatically higher rates than Japan had even at the very peak of their own crime rates. Given this and what has been demonstrated about the age-crime curve, it is predictable that crime rates could be higher among these younger groups. If anything, these groups crime rates are probably a good deal lower than what we would expect of Japanese nationals of the same age.

The end result of this is that immigration does not appear to take Japan into a future of unprecedented crime. Rather, it turns back the clock. Even if Japan vastly expanded immigration from the current 3% of the population, the net effect would likely only assist in taking the country back to the crime rates of its relatively recent (and still very safe) past. 

If we look at these nationalities' current arrest rates and compare them to Japan's arrest rates from previous years, we can make the following comparisons:

In fact at their current crime rates, boosting immigration from the Phillipines and Indonesia would make Japan safer than it has ever been, despite their young average ages.

Is immigration worth it?

Many people in Japan appear to worry that accepting more immigrants will change their country for the worse. While the cultural changes that could occur are more subjective and debatable, the threat to public safety may look less alarming if it is put in the perspective of changing Japan's crime rates back to what they were in, say, 2018. 

A final argument by anti-immigration forces is that "even one crime by a foreigner is too many", regardless of how they are framed in terms of rates per 100,000 residents.

But of course, the same would be true of young new Japanese nationals if the country ever succeeded in increasing birth rates; eventually, those children will reach their twenties and commit many crimes they wouldn't have if they had never been born.

Would any of the same people argue natalism efforts should stop, because even one crime committed by a young Japanese is one too many? I'm guessing not, because that sounds ridiculous. But anyone who dismisses the latter as silly but still considers the former an unacceptable threat lays their prejudices bare.