How have relationships changed in the last ten years?

I have written previously on sexual behavior in 2021 and how many people had sex in 2021, based on the General Social Survey. We have seen a slow climb upward where fewer young people, men and women both, are having sex.

In 2021 15% of all men did not have sex in 2021, while 22% of all women did not have sex. The numbers were much higher for the youngest demographic (18-25); 42% of young women did not have sex while 23% of young men did not have sex.

I also looked briefly at sex within relationships. Depending on the demographic, only between 5% and 15% of sex occured outside of relationships. Sexual activity is closely linked with current relationship status. Most sex occurs in relationships and the people having sex the most frequently are men and women in relationships.

What I didn’t look much at was relationship status. It’s possible that the pandemic, as well as the general trend toward having less sex, is also reflected in relationship status.

I will focus mainly on three variables from the 2021 GSS: age, sex and the “Posslqy” variable which measures having a “steady partner.”

The Posslqy variable and what it measures

To avoid ambiguity on what the Posslqy variable measured in the GSS, here is the question that was asked and the four potential responses:

Which of these statements applies to you?

  1. I am married and living in the same household as my husband or wife
  2. I have a steady partner, and we live in the same household
  3. I have a husband or wife or steady partner, but we don’t live in the same household
  4. I don’t have a steady partner

This is a bi-annual variable that the GSS began collecting in 2012. This means we have data for the years 2012, 2014, 2016, 2018 and 2021. This gives us a ten year window to observe a trend, multiple replications over that time and a period before the COVID-19 pandemic.

In this article I use the term “single” to refer to people who responded “I don’t have a steady partner.”

What Posslqy does not measure

This variable can tell us who is married, both married and cohabiting and who is in a steady partnership. As such, it is a good variable for serious relationships. However, by being conservative in its measurement of “steady” partnerships it will also be an underestimate of total relationships.

Many people are in sexual and romantic relationships they would not list as “serious.” Friends-with-benefits (FWBs), “situationships” and casual daters all fall into this category.

The relationship is also a snapshot, so it will underestimate relationships over time. For example, it will exclude people who are recently broken up as well as people who will enter into relationships in the near future.

This means we can’t assume that anyone who responded as being single has been perpetually single, romantically uninvolved or sexually inactive. It will be an underestimate of romantic involvement, but we can’t know by how much.

Relationship stability, 2012-2021

Here is a table of the raw data:

GSS singles 2012 through 2021
GSS, Singles 2012-2021

If you have read my article on how to know if your sample size is sufficient, you will recall that around 1000 people is the gold standard for nationally representative surveys. At this sample size we can generalize to the entire US population with a margin of error of around 2%.

If you look at the bottom row, you will see that the percentage of singles in the US has remained highly stable pre and post pandemic. For the last ten years, American singledom has hovered around 30%.

Graph of Singles, GSS Data 2012 2021
GSS, Singles 2012-2021

Corroborating the GSS data

As a large, random sample with high compliance the GSS is as close to a truly representative sample as we get in data collection.

However, it’s always good to corroborate. We can check other nationally representative samples and compare them with the GSS dataset. A lot of research does not use random nationally representative samples, but rather convenience samples. We will typically find random representative samples conducted by polling firms, or by the US Census.

The Pew Research Center conducted a large, nationally representative survey of single Americans in 2019. Here are their results:

Pew Research Center, 2019

The Pew report tells us that 31% of Americans were single, which is consistent with the data for 2014-2021.

A second source we can look to is the US Census Bureau. The US census does not give us a measurement of singlehood, instead limiting itself to married and cohabiting couples. This will underestimate the number of people in relationships (clearly not all people in relationships are married or living together), but it will give us a figure to adjust upward from.

According to the US Census Bureau, in 2021 58.4% of adults were living with a spouse or unmarried partner. We therefore can expect the total number of relationships to be above 58.4%.

In the GSS 2021 data, 50% of adults were living with a spouse and 13% were living with a steady partner. We only see a 5% deviation across the US census and the GSS.

GSS breakdown by age

Here are the figures for 2012-2021:

GSS, Singles by Age 2012 2021
GSS, Single by Age, 2012-2021

Singlehood has increased for the age group 34-41 by 8%. Meanwhile, singlehood has also decreased for the 18-25 age group by 17%. There is only a 1% total difference between 2012 and 2021 in total singles for age groups combined.


Relationships are not randomly distributed. Knowing the total number of Americans in relationships is not the entire picture. Age is one of the variables that we should look at. The GSS collected data up to 88 years of age. Below I have divided GSS age data into three groups, stopping at age 41.

Within the 18-25 age group, 43% are single. For 26-33, 37.4% are single. In the 34-41 age group 22.9% are single. In total, between ages 18-41, 34.5% are single.

The younger someone is the less likely they are to be in a relationship. Only in the 30+ range do we begin to see more people in relationships than the national average, at 77% versus 70%.


In the 18-25 age group, 52.6% were single. For 26-33, 36.2% were single. 34-41 gives us 21.1% single. In total, 34% single.

In 2018, before the COVID-19 pandemic, more of the youngest age group were single than they are now. Age groups 26-33 remained quite stable at 37.4% and 36.2%, as did ages 34-41 at 22.9% and 21.1%.

Youth still predicted a lower likelihood of being in a relationship before the COVID-19 pandemic. The national relationship average is still weighted by the oldest demographics. However, we see an 8% reduction in singlehood comparing 2018 with 2021. The youth are in more serious relationships now than they were before the pandemic.


18-25, 60.8% single. 26-33, 32.8% single. 34-41, 14.8% single. The total, 33.6% single.

In 2012 we saw even more singles than in the recent year. 17% more of the 18-25 age group were single in 2012.

GSS Single by Age, 2012-2021
GSS, Single by Age, 2012-2021

Keeping in mind that the national average for singlehood was 30%, by 34-41 people should be less likely to be single on average. This group did not go higher than 23% and was only at 14.8% single at its lowest. This has consistently been peak relationship time, when adults are most likely to be in a relationship before age 40.

Meanwhile the youngest age group has consistently remained the most likely to be single. We see a majority of young adults between 18-25 reporting singleness annually up until the past year.

We don’t see the largest changes between 2018 and 2021, or before and after the COVID-19 pandemic and associated lockdowns. The largest change was fewer singles in the 18-25 age group, down from 52.6% to 43%.

GSS, sex and relationship status

As we’re looking at a predominately heterosexual sample, we should expect the sex ratio for relationships to be close to 1:1. For every man in a relationship there is a corresponding woman. Very few people report being in poly- relationships of any kind.

However, we see that women are consistently more likely to report not having a steady partner up until 2021.

GSS, Single by Year, Sex
GSS, Single by Year, Sex
GSS, Single by Year, Sex
GSS, Single by Year, Sex

GSS 2021, relationship status by sex and age

This is likely driven by age effects. Let’s look at 2021, by age and sex:

GSS 2021, Singles by Sex and Age
GSS 2021, Singles by Sex and Age

Here we see that men are consistently more likely to report being single. However, the gap closes as men get older. We can corroborate the GSS data with the data from the Pew 2019 survey as well:

There are likely two things happening here. First, men die earlier. This leaves more single women in the highest age bracket. Second, women date up in age-gap relationships. We usually end up with a close 1:1 ratio across the entire population, but the rate of singlehood is not evenly distributed.

That most men and women will not remain single in their 30s and up is also explained by the increasingly late age of marriage. According to the recent US census, the average age of marriage for men is 30 and for women is 28.

We can see this in the GSS data as well, by looking at the percentage of men and women who reported being in a steady partnership and/or married, but not living together:

GSS, Steady Relationship, Not Living Together
GSS 2021, In A Steady Relationship But Not Living Together
GSS 2021, In A Steady Relationship But Not Living Together

By age 26, the gap between men and women for non cohabiting or unmarried relationships closes. We see less disparity in men and women from 26 onward who report cohabiting in a steady relationship:

GSS 2012, relationship status by sex and age

Let’s compare 2021 with 2012 for age and sex:

In 2012, both young men and women aged 18-25 were more likely to be single. However, women were almost twice as likely to report being single in 2012 as in 2021 (33% vs 62%). Young men in this age bracket only differed by 3%, with more being single in 2012.

Men and women were also less likely to report being single in the 26-33 age group, by 6% and 8% respectively. 

At 34-41, men were less likely to report being single at 10% versus 25%. Women remained fairly stable at 20% vs 21%.


The overall total of singles has remained very stable. However, we see some variation across age and sex demographics within that stable trend. The youngest men and women are the most likely to be single. Young men in particular are more likely to be single than young women and young men constitute the largest single demographic.

Singledom for young men in the 18-25 and 26-33 age groups has remained fairly stable over the last ten years. Singledom for women between 18-25 has halved, while singledom for men 34-41 has more than doubled.

Why have young men consistently been more likely to be single? I would caution against looking for a single cause. This trend predates COVID-19 and it predates the rise of dating apps. Young men are less likely to want to commit, so choice may be part of the picture. Young men are also more likely to have low incomes, low education and to live at home. This makes young men less desirable mates. At the youngest range of 18-25, young men may even be less physically imposing or less masculine, making them less physically desirable. The data cannot tell you why, so you are welcome to speculate. Please share your ideas and let me know what you think may be the cause in the comments.


  1. Singledom has remained very stable from 2012 onward and despite the pandemic.
  2. The highest rate of single men is in the youngest cohort of men, aged 18-25; the highest rate of single women is in the oldest cohort of women.
  3. Young singlehood may be explained by age-gap relationships, later age of marriage and earlier mortality for men.
  4. Singlehood for men aged 18-25 has remained stable since 2012, but is almost twice as low for women in 2021. 
  5. Nationally representative samples across the GSS, Pew Research and the US Census Bureau give us very similar numbers both for 2021 and previous years. This replication is a strong indicator of the reliability of the datasets.
  1. This could be due to computer games and the internet increasing social isolation and blunting men’s social skills. Girls generally don’t spend as much time online as boys. Additionally, men often don’t have fathers or other male relatives in their surroundings, which leads to a lack of masculine character development and a deficiency in the confidence necessary for mating effort. Menelaos Apostolou’s research suggests that poor flirting skills and low mating effort are the main predictors of involuntary bachelorhood, which supports my opinion.

Leave a Reply

Your email address will not be published. Required fields are marked *

You May Also Like

What is an incel?

Correlation does not equal causation, but when a variable X is causal of a change in variable Y…