Honestly, I wasn’t planning to make the article this long. I just thought I could point out some obvious statistics many journalists seem to be oblivious of. But as I looked into publicly available statistical data, I noticed that some of their inaccurate assumptions were potentially fatal.
It all started with an article by the Guardian. It was entitled ‘Why have young people in Japan stopped having sex?’ and began with a story of a Japanese woman who used to be a professional dominatrix and now is a sex and relationship counsellor.
With the help of catchy content, the article took off immediately and got more than 70,000 shares and likes on Facebook (which is a lot). Many other journalists followed by writing their own articles – most of them critical of the original article – to get a piece of the action.
A friend of mine shared a lot of these articles. As I read through them, I started asking myself questions: Are Japanese people really sexless? Is sexlessness Japan’s unique problem? Is it related to the low birth rate?
Then I realised something: while individual stories and interviews were interesting, some of their comments on Japan’s social situation were rather dubious.
‘One reason for the lack of babies is the emergence of a new breed of Japanese men, the otaku, who love manga, anime and computers – and sometimes show little interest in sex’, writes Anita Rani from the BBC.
As represented by this sentence, the article is based on a seemingly innocent assumption that the apparent lack of interest in sex is a major cause of the low birth rate, which turned out to be wrong, according to my research. (It also talks about immigrants, without showing the connection to the article’s main issue.)
I came up with one hypothesis: many married people I know have had less education than the ones who are still unmarried. Maybe the education level has something to do with the birth rate?
I did some quick research and found out that there was a strong correlation between countries’ educational attainment and their fertility rates; it was a well-known fact. ‘This is interesting,’ I said to myself.
I started digging various publicly available statistics to see whether other things that journalists were writing were statistically sound. One question led to another and, before I knew it, I had 20 sets of statistical data ranging from literacy rates to frequency of sex.
Here are some of the things I found out:
1: Sexless Means a Low Fertility Rate
This association repeatedly comes up. In the aforementioned Guardian article, Haworth writes: ‘increasing numbers (of Japanese people) can’t be bothered with sex. For their government, “celibacy syndrome” is part of a looming national catastrophe. Japan already has one of the world’s lowest birth rates’, assuming that the lack of interest in sex is highly related to the country’s low birth rate.
But the truth is this: frequency of sex doesn’t correlate to the fertility rate.
The x axis shows how many times people have sex per year on average, and the y axis shows countries’ fertility rates.
As you can see, the plots are completely scattered and there is no apparent trend. The correlation between the frequency of sex and the fertility rate is virtually non-existent. (0.01 out of 1.00. If you want to know more about this number, look here: Pearson Product-Moment Correlation)
People in countries like Greece, Croatia, and Bulgaria have sex many times a year, but their fertility rate is as low as Japan’s. In fact, the fertility rate varies widely between the countries where people have sex 80-120 times a year. Most of the countries are within this range.
It may seem counter-intuitive but it is completely possible. For example, you can have sex as many times as you want without having a baby if you practise correctly protected sex. On the other hand, you can still have a baby even if you have sex only once in your lifetime. People have control over having children regardless of how frequently they have sex.
This also tells us the existence of other factors that really affects birth rates. But a lot of people (including professional researchers) have already discussed this and that’s not the main theme of this article.
Are you suspicious about this number? I also calculated the correlation using different statistics by Durex from a different year: the percentage of respondents who have sex weekly. This time, the number was -0.2 out of -1.0, which means a very weak reverse correlation. The more regularly people have sex, the slightly lower the birth rate.
Sexlessness and fertility rates are two separate things. Many articles I’ve read confused these two and, as a consequence, drew inaccurate assumptions about Japanese social issues.
This is a very important fact because, if the link between sexlessness and the low fertility rate is non-existent, many ‘arguments’ based on Japan’s lack of interest in sex are invalid as the premise is false.
I actually thought that frequency of sex correlated to birth rates before I did this research. So the result was somewhat unexpected to me too.
The causes of low fatality rate is complex. Saying that the low libido is the cause may be be catchy and gets attention, but it’s overly simplistic and untruthful.
2: A Low Birth Rate is Japan’s Unique Problem
To be fair, many journalists understand this one; the low birth rate is far from a Japan-only problem. In an article entitled ‘No, Japanese People Haven’t Given Up on Sex’, Joshua Keating from Slate writes: ‘Japan is a leading indicator of a trend rather than an outlier. Birth rates are falling almost everywhere in the world except for sub-Saharan Africa.’
However, many people continue to imply that Japan’s low birth rate is something exceptional. Justin McCurry from the Guardian, in his article about Japanese women seeking instant sexual partners though an agency, opens the article by writing that ‘in 2003 Japan’s birth rate hit a record low of 1.29 – the average number of times a woman gives birth during her lifetime – one of the lowest rates in the world.’ He seems to be oblivious to the fact that a low fertility rate is a strong trend amongst highly educated nations.
This figure shows the strong reverse correlation (-0.8 out of -1.0) between countries’ IQ scores and their fertility rates. The x axis represents the IQ scores and the y axis represents the fertility rate. As you can see, if you go to the right side of the figure (high IQ score), the fertility rate gets lower. This is a typical linear correlation.
Mind you, I know that the IQ is a highly problematic concept. So let’s say it about educational attainment.
Japan’s birth rate is not particularly low considering that the higher the educational attainment, the lower the birth rate.
Japan is merely following this global trend. If there are true outliers, they are countries like Singapore or Cuba.
IQ is highly correlated to various educational indexes. I also used educational indexes, percentage of secondary education enrolment and literacy rate, and calculated the correlation between them and the fertility rates and got similar results.
While this correlation alone doesn’t necessarily imply that the educational attainment is the direct cause of the low birth rate, you can assume there’s some relationship (whether direct, indirect, or through another variable like different lifestyles) between them. The correlation is very strong.
3: Long Working Hours Cause Sexlessness.
Some people blame Japan’s infamous working hours for sexlessness. William Pesek from Bloomberg says, ‘. . . the exclusion of more and more Japanese from the lifetime employment system that’s long been the cornerstone of Japan Inc., forcing many to work additional jobs. If you leave for work at 6 a.m. and get home close to midnight, including weekends, where is there time for dating?’
He doesn’t explicitly say that long hours cause sexlessness: he says ‘dating’, which may or may not involve sex. So it might not be fair to quote him here. To make my point, I quote someone from Facebook: ‘the problem I had with the Guardian article is that the lack of sex here was positioned as a racial/cultural problem. It’s really a problem of overwork.’
But long work hours almost don’t correlate at all to sexual infrequency.
The x axis shows the average annual hours actually worked per worker, and the y axis shows how many times people have sex a year.
The distribution is quite flat, except for Japan where people have much less sex than any other country. Spain and Greece are also far away from the other countries as people apparently have much more sex. Otherwise, there’s only a weak positive correlation (0.18 out of 1) between working hours and frequency of sex: the longer you work, the more you have sex. But this correlation is very weak and not definitive at all.
Longer work hours don’t imply less sex, if anything they might imply more sex.
(*Edit: A few people pointed out that this comparison is highly questionable due to many factors. I have to admit that I was quite hasty on this one. Check out the excellent comment by Christopher Magor in the comment section below.)
4: Slow Economic Growth Causes Sexlessness
William Pesek is very clear on this point: ‘the root of Japan’s supposed sex drought isn’t culture, but economics,’ he writes. He seems to be quite convinced by this. A few paragraphs later, he also writes that ‘the real issue is that many avoid traditional, committed relationships out of doubts about the future that are based on economics rather than culture.’
He also says that ‘if low libido were strictly societal, why do the Czech Republic, Poland, Singapore, South Korea, Spain and Taiwan have fertility rates as low as Japan’s?’ But we already know the answer: frequency of sex has very little to do with fertility rates. So it’s totally possible that other countries have lower fertility rates while they have more sex.
Anyway, the economy (economic growth, to be more precise) doesn’t correlate much to sexual frequency.
The x axis is the real growth rate and the y axis is the frequency of sex per year. The correlation is -0.185 out of -1.000, which is a very, very weak reverse correlation: the more the economic growth, the slightly less likely the sex.
The assumption that slow economic growth is the cause of sexlessness is very likely to be false.
Of course you can still argue that Japanese people are particularly sensitive to economic mood. But in this case, you would need to provide a convincing reason why Japanese people can be so ‘different’.
5: Women’s Social Advancement is the Cause of Sexlessness
Japan’s gender gap is notoriously bad. But some people think that women are relatively stronger than before and that makes men scared of them, which causes low sexual drive.
Roland Kelts from the Guardian presents one of the ‘theories’, ‘that Japanese women have become stronger socially and economically at the very same time that Japanese men have become more mole-ish and fully absorbed in virtual worlds, satiated by the very technological wizardry their forebears foisted upon them, and even preferring it to reality.’
However, the gender gap doesn’t correlate to the frequency of sex.
The x axis represents The Global Gender Gap Index by the World Economic Forum (the higher the better). The y axis represents the frequency of sex.
There’s almost no correlation between the gender gap and the frequency of sex. The number is 0.13 out of 1, which means there’s only a very weak correlation: the lower the gender gap, the more the sex.
Gender gap is not a good indicator of sexlessness.
This relationship is not obvious at all. Some people think that men are intimidated by strong, independent women. Other people think that modern women are freer to express their sexuality. Both can be true, or false. But the link to the frequency of sex is not apparent in these data.
I know this article is lengthy. So if you just skip to this part, no worries: you can just read these bullet points to get the main idea.
- Sexlessness has nothing to do with low birth rates. They should be discussed separately.
- The low fertility rate is not Japan’s unique problem. It’s a very strong trend among the highly educated nations.
- Sexlessness has nothing to do with a low marriage rate either. They should also be discussed separately.
- Many people mix sexlessness and fertility rates all together and try to come up with something. They are two possibly independent variables.
- Working long hours doesn’t make you sexless.
- Slow economic growth doesn’t make you sexless.
- Narrowing the gender gap won’t make people have more sex.
The statements above are statistically not very accurate in favour of simplicity. They should be something like, ‘based on the correlation, it is likely that…’ Before you say something about correlation and causation, read the main article and the footnote.
I’m also aware of the possibility that my conclusions are false as my research is far from thorough. Read the footnote to learn more about this.
So, What Was the Real Problem?
If sexlessness doesn’t relate to any real social problem, what is really wrong with having less sex? I mean, as long as people are happy they should be OK, right?
Interestingly, the frequency of sex doesn’t correlate with a country’s suicide rate, which can be a good indicator of happiness. And, ironically, Durex’s own study shows that the frequency only has a very weak correlation (0.27 out of 1) with sexual satisfaction. Quantity doesn’t mean quality. (But, for the record, Japanese people are also extremely low – 15% – on sexual satisfaction.)
Sexual choice is really a personal matter. Talking about it as if it has a real impact on other social issues is potentially misleading.
I tried to keep the main article short. I didn’t exactly succeed, but there is just so much to talk about.
This article’s weakness
Most of the statistics in this article heavily rely on Durex’s sexual survey, and one can still question the validity of such a survey. They conducted the survey online in most countries. Also, all the numbers are obviously self-reported.
Also, I am aware that in order to measure the possible impact of sexlessness on Japanese society in particular, you have to examine the historical sexlessness within Japan. Unfortunately, I wasn’t able to get any statistics on this. It is quite possible that frequency of sex has unique meanings in some countries.
If you know other good statistics available on this subject, please let me know. I think we need to validate the results with other statistics.
About the article by Abigail Haworth
Actually, I quite enjoyed reading the Guardian article by Abigail Haworth. She tells an interesting story about the sex counsellor and her clients. She also interviewed quite a few people and I think their stories represent quite well the feeling of certain young Japanese people.
The way journalists (at least the ones I examined here) write stories is to come up with catchy stories, write them and connect them with seemingly related numbers. The problem is that when they try to connect their stories to social issues, they sometimes make logical leaps, and some people buy their ‘arguments’.
Wait, correlation doesn’t imply causation and causation can exist without correlation!
…you may scream. Well, I think it’s easy to understand that correlation doesn’t imply causation, right? (But we still make this mistake because it can be confusing.)
But when it comes to the fact that the absence of correlation doesn’t imply the absence of causation, it’s much harder to explain. I should be aware of this because most of this article is related to this.
Let’s think about sexlessness and fertility rates. I said that there’s virtually no correlation between the frequency of sex and the birth rate. However, this doesn’t mean that there’s not causation: there obviously is! You have to have sex in order to make a baby. In other words, sex causes babies, albeit not always.
You can also think that if the frequency of sex is 0, the fertility rate will also be 0 (assuming of course there’s no sperm donation).
But the data show that there’s no correlation. What does it mean?
When correlations are absent despite the existence of causation, it possibly means that there are other external factors that neutralise the results, which is exactly the case in this instance: birth control. In this case, it is highly likely that the high sex frequency doesn’t produce more children because people use condoms and possibly abortion when that fails. (Theoretically, you can also think of other factors like epidemics.)
How about non-linear correlation?
The numbers that indicates the correlation is called ‘he Pearson product-moment correlation coefficient’, also known as Pearson’s r. One of the pitfalls of using it is that it can only capture linear correlation. That’s why you always have to make plot charts to see if there’s any apparent non-linear correlation, which I did.
It’s still possible that meaningful correlation exits
In this article, I threw in every country with available data. But there’s still a possibility that we can find meaningful correlation by filtering out countries by other variables. I can think of a few possibly relevant variables: religion, region, race, language family, immigration rate, GDP per capita etc. Those elements seem to be relevant to sex and marriage.
It’s also possible that a particular variable is meaningful despite the global trend, if there are unique factors that are only present in Japan
For example, if there’s a very unique cultural element in Japan, it will be possible that a particular variable still correlates to some other variables, only in Japan.
For example, suppose that, for some reason, Japanese people don’t use condoms (which, by the way, would be very ironic since Japan produces extremely high quality condoms). If this happened, more sex would obviously mean more babies.
When something unique like this exists, global trends don’t mean much in that particular case.
Sexlessness is quite independent from other variables
I tried to find correlations between frequency of sex and other variables (IQ, birth rates, education index, secondary school enrolment, suicide rates, life expectancy, GDP, percentage of people over 65, working hours, gender gap, marriage rate, divorce rates, sex satisfaction, and economic growth.) But the only apparent correlation I found is the literacy rate (and the education index, to some extent). So, highly educated people are somewhat likely to have sex more frequently. If this is true, Japan’s sexlessness is a complete outlier.
Well, maybe it’s just culture after all? Some of the articles I talked about deny this simple assumption, with varying reasoning which I refuted. In that case, the real question will be this: how do we know if this is something we call ‘culture’? And where did this ‘culture’ really come from?
Abigail Haworth (20/Oct/2013). Why have young people in Japan stopped having sex? The Guardian. http://www.theguardian.com/world/2013/oct/20/young-people-japan-stopped-having-sex
Roland Kelts (27/Dec/2011). Japan leads the way in sexless love. The Guardian http://www.theguardian.com/commentisfree/2011/dec/27/japan-men-sexless-love
Justin McCurry (4/Apr/2005). Japan’s virgin wives turn to sex volunteers. The Guardian. http://www.theguardian.com/world/2005/apr/04/japan.justinmccurry
Anita Rani (24/Oct/2013). The Japanese men who prefer virtual girlfriends to sex. The BBC. http://www.bbc.co.uk/news/magazine-24614830
Joshua Keating (23/Oct/2013). No, Japanese People Haven’t Given Up on Sex. Slate. http://www.slate.com/blogs/the_world_/2013/10/23/are_japanese_people_really_having_less_sex_than_anyone_else.html
William Pesek (26/Oct/2013). The Lust Beneath Japan’s Sex Drought. Bloomberg. http://www.bloomberg.com/news/2013-10-25/the-lust-beneath-japan-s-sex-drought-.html
Central Intelligence Agency (2012). GDP (OFFICIAL EXCHANGE RATE). https://www.cia.gov/library/publications/the-world-factbook/fields/2195.html
Central Intelligence Agency (2013). AGE STRUCTURE. https://www.cia.gov/library/publications/the-world-factbook/fields/2010.html
Central Intelligence Agency (2013). The World Fact Book. https://www.cia.gov/library/publications/the-world-factbook/rankorder/2127rank.html
Durex (2005). 2005 Global Sex Survey. http://www.durex.com/en-jp/sexualwellbeingsurvey/documents/gss2005result.pdf
Durex (2006). Sexual Wellbeing Global Survey. http://www.durex.com/en-NZ/SexualWellbeingSurvey/pages/default.aspx
Human Development Reports (2009). Human development index 2007 and its components. http://hdr.undp.org/en/media/HDR_2009_Tables_rev.xls
OECD (2005). StatExtracts. http://stats.oecd.org/Index.aspx?DatasetCode=ANHRS
Richard Lynn (2006). Intelligence and the Wealth and Poverty of Nations. http://www.rlynn.co.uk/index.php?page=intelligence-and-the-wealth
The World Bank (2012). GDP growth (annual %). http://data.worldbank.org/indicator/NY.GDP.MKTP.KD.ZG
The World Bank (2012). GDP per capita (current US$). http://data.worldbank.org/indicator/NY.GDP.PCAP.CD
The World Bank (2012). Life expectancy at birth, total (years). http://data.worldbank.org/indicator/SP.DYN.LE00.IN
The World Bank (2012). School enrollment, secondary (% net) http://data.worldbank.org/indicator/SE.SEC.NENR
Unesco (1997). Literacy, culture and communication. http://www.unesco.org/education/information/wer/WEBtables/Ind2web.xls
UNSD (2008). Marriages and crude marriage rates, by urban/rural residence: 2002-2006. http://unstats.un.org/UNSD/Demographic/sconcerns/mar/mar2.htm
World Economic Forum (2013). The Global Gender Gap Report 2013. http://www3.weforum.org/docs/WEF_GenderGap_Report_2013.pdf
World Health Organization (2011). Suicide rates per 100,000 by country, year and sex. http://www.who.int/mental_health/prevention/suicide_rates/en/
If you've read this far, you should probably subscribe to my blog and follow me on twitter or Facebook.