The Aromantic Spectrum in the Ace Community Survey

by René Mellema & Tristan Miller

Recently, there has been a lot of discussion about the interaction between the asexual and aromantic communities.  Following our review of extant aro community surveys, we contribute to the discussion by reviewing the information we have about aromantic and aromantic-spectrum respondents to the 2017 Ace Community Survey.

Main Findings

Although our survey recruits from asexual communities rather than aromantic communities, we are able to gather information about people on the aromantic spectrum, including some allosexual people. Some of our important findings are as follows:

  1. Compared to aro community surveys, our survey finds a larger number of gray-(a)romantic, demiromantic, lithromantic, and quoiromantic people as compared to aromantic people.
  2. Among our respondents, people on the aromantic spectrum were more likely to be non-binary.  Allosexual people (regardless of romantic orientation) were more likely to be men.
  3. People on the aromantic spectrum, especially those who were also on the asexual spectrum, were less likely to have had romantic relationships.  Allosexual people on the aromantic spectrum were more likely to have non-romantic significant relationships, and to be polyamorous.
  4. Both the asexual or aromantic spectrums are negatively correlated with sexual activity.
  5. On average, allosexual people have a higher sex drive, and higher frequency of sexual fantasies, masturbation, and porn consumption, regardless of romantic orientation.

Further details are below.

The Respondents

First, we would like to see how many people who took the survey were on the aromantic spectrum. For the purposes of this analysis, we define “aromantic spectrum” or “aro-spec” to include people who are aromantic, demiromantic, gray-(a)romantic, lithromantic, or quoiromantic.  We use “alloromantic” to refer to people who are not aro-spec. “Asexual spectrum” and “ace-spec” refer to people who are asexual, gray-asexual, demisexual, or questioning, and “allosexual” refers to people who are not ace-spec.

This gives the following counts for the number of aro-spec respondents.

Table shows a cross-tabulation of the ace and aro spectrums

As we can see, around half of the respondents are aromantic spectrum, and around half are alloromantic.  Among aro-spec respondents, only 1.9% are allosexual, and the rest are ace-spec. Note that our survey primarily targets asexual spectrum communities for recruitment, and so these ratios do not reflect the general population.

Next, we look at the specific romantic orientation labels used by people on the aromantic spectrum.  

Table shows percentage of ace-spec and allosexual respondents with various aro-spec identities

*Percentages show the number people in the row and column, as a percentage of all the people in the same column.  For instance, the first row and column shows that 58.97% of people who are ace-spec and aro-spec are aromantic. Percentages do not add to 100% because respondents could select multiple romantic orientations.

We compare our results to the Arospec Identities Stigmatization survey, whose respondents were 67.6% aromantic, 8.2% demiromantic, 7.4% greyromantic 1.8% lithromantic, 4.3% quoiromantic, 2.6% aroflux, 2.6% aro-spec, and 5.3% other.  These numbers aren’t quite comparable to our survey because the Arospec Identities Stigmatization survey restricted people to a single response, but it appears that our survey has fewer aromantic people and more demiromantic, gray-(a)romantic, lithromantic, and quoiromantic people.  This likely reflects differences between the ace and aro communities from which the surveys respectively recruited from.


From this point on, we define the following four groups:

  1. Aro-spec and ace-spec (AroAce)
  2. Aro-spec and allosexual (AroAllo)
  3. Alloromantic and ace-spec (AlloAce)
  4. Alloromantic and allosexual (AlloAllo)

To determine the statistical significance between groups a 95% confidence interval was estimated with the following formula: 1.96 * √(P(100-P)/N) where P is the percentage and N is the number of respondents. Note that because of the small sample size of the AroAllo group (N=98), many comparisons with this group are not statistically significant at the 95% confidence level.

Each of these groups has slightly different gender ratios.  The aro-spec groups have higher proportions of nonbinary people, and the allosexual groups have higher proportions of men.

Table shows the gender ratios of the four groups


Next, we want to measure differences in how people experience relationships. Note that “aromantic” is sometimes defined as not wanting romantic relationships; however, this is not always the definition, not all aro-spec people are aromantic, and not all relationships are romantic.  Our survey asks about “significant relationships”, which we partially define as follows:

“Significant relationships” refers to close relationships other than family or close friends – typical examples could include marriage, domestic partnerships, queerplatonic relationships, partner, boyfriend, girlfriend, etc. Significant relationships need not necessarily be sexual or even romantic.

With this in mind, the next table shows the number of people in each group who are single, married, or in a significant relationship.  We find that AroAces are most likely to be single, closely followed by AroAllos and AlloAces.

Table shows current relationship status of four groups

The difference between AroAces and the other groups becomes much larger when we ask if people have ever had a significant relationship.

Table shows percentages of each of the four groups that have ever had a significant relationship

Next, we asked people if they had ever had romantic significant relationships.  AroAces were least likely to have had a romantic relationship, followed by AroAllos, AlloAces.

Table shows percentage of each group who have had significant romantic relationships

Next, we asked about non-romantic significant relationships.  Using a 95% confidence interval AroAllos were statistically significantly more likely to have non-romantic relationships, followed by AlloAllos.

Table shows percentage of each group who have had significant non-romantic relationships

Finally, we asked whether people were polyamorous.  AroAllos were more likely to be polyamorous than any other group, and this was statistically significant using 95% confidence intervals.

Table shows how each of the four groups relate to polyamory

Sexual Behaviours and Attitudes

We asked respondents whether they had ever engaged in consensual sex, and it is clear that both the aromantic and asexual spectrums have a significant negative correlation.  AlloAllos were most likely to have had consensual sex, followed by AroAllos, then AlloAces, and finally AroAces.

Table shows how many people in each of the four groups have had consensual sex

Then we asked about the sex-repulsed/indifferent/favorable trichotomy, which we refer to as “sex disposition”.  Allosexual groups were far more likely to be sex-favorable and less likely to be indifferent or repulsed. Romantic orientation also appears to play a small role, with AroAllos being less likely to be sex-favorable than AlloAllos, and AroAces being more likely to be sex-repulsed than AlloAces.  These differences are all significant at the p<0.05 level, using a 95% confidence interval.

Table shows the sexual dispositions of the four groups

We asked people to rate their sex drive on a scale from 1 to 4, defining sex drive as “the drive to engage in some kind of sexual stimulation, whether through partnered sex or solo stimulation (e.g. masturbation)”.  There is a large and significant difference between ace-spec groups and allosexual groups. Although there appears to be a difference between the AroAllo and AlloAllo groups, it is not statistically significant.

Bar chart shows sex drive of the four groups

We also asked about the frequency of masturbation, sexual fantasies, and porn/erotica consumption.  In general, we find that allosexual people are more likely to have higher frequencies of each than people on the asexual spectrum.  When comparing AroAllos and AlloAllos, it appears that AroAllos have a slightly larger average frequency; however, the differences are not significant at the p<0.05 level.

Bar chart shows rates of masturbation of the four groups

Bar chart shows rates of sexual fantasies of the four groupsBar chart shows rates of porn and erotica consumption of the four groups

In this article, we’ve looked at the respondents to our survey who were on the aromantic spectrum, and compared them to alloromantic respondents on several measures.  While we caution that our results may depend on recruitment methods, we hope that they serve as a starting point for understanding group differences.

Leave a Reply