Around the world, being close to someone who has had an abortion makes people substantially more likely to support unrestricted abortion.

That's among the findings about abortion in a new BuzzFeed News/Ipsos poll of 23 nations. We looked at the marriage equality results from the survey in a post last week. In at least 10 of these countries, people who are close to someone who has had an abortion were over 50% more likely to support abortion without restriction.

The survey was conducted online over a two-week period ending May 8. We have the clearest picture of attitudes in 16 countries where the internet is widely accessible because it is possible to get an online sample that is representative of the country as a whole. These are concentrated heavily in Europe and North America. There are obviously large parts of the world that aren't represented in this survey, but unfortunately data collected online in these countries simply wouldn't give us an accurate look at the overall feel of a country because so many people lack internet access or local laws make it hard to poll on social issues.

But we included some data from seven additional countries — Mexico, Brazil, China, South Africa, India, Turkey, and Russia — because they appear to reflect broader public opinion based on comparisons with other public opinion surveys. They're marked with an asterisk as a reminder that these figures may not be truly representative.

This is what we found.

1. This is how many people support unrestricted abortion around the world.

Several countries that we surveyed in Western Europe had widespread support for allowing abortion “whenever a woman decides she wants one” — essentially unrestricted abortion. In other parts of the world, we found widespread support for allowing abortion with restrictions, like in cases of rape or when pregnancy could cause the death of the mother.

2. No country surveyed favored a total ban on abortion.

Even in countries where opposition to abortion was strongest, a majority still favored allowing abortion in at least some circumstances. Total opposition to abortion was strongest in Brazil*, but even there we only found roughly 18% of our sample saying they believe the procedure should be illegal without exception.

3. Being close to someone who has had an abortion makes people substantially more likely to support unrestricted abortion.

Pollsters have long found that people who know someone who is gay or lesbian makes them much more likely to support same-sex marriage. To test whether there was a similar effect on opinions about abortion, we asked respondents if they had ever had an abortion or if they were “close to someone who has had an abortion.” We found a substantial impact in every country we surveyed — sometimes dramatic ones.

In Russia*, for example, people close to someone who has had an abortion were nearly three times more supportive of unrestricted abortion. In Ireland, Poland and Australia, they were around twice as supportive.

4. This is how many people say they are close to someone who has had an abortion (or have had one themselves).

Russia* appears to top this list, probably a reflection of the fact that abortion was used essentially as a form of birth control under communist rule. High numbers in China* are possibly due to the country’s one-child policy, a 1979 law which prohibited families from having a second child. Mothers who became pregnant again were often forced by the government to abort their pregnancy. (The policy was relaxed in 2013 to allow for two children in most households.)

5. Women may be more willing to talk about their abortions with people close to them in some countries than others.

That last chart doesn’t simply reflect how common abortion is in each country — it also hints that in some countries, it’s much more common for women who have had abortions to talk about it.

Compare China* and and the United States, for example: Data from the United Nations shows that about the same percentage of women have had abortions in both countries, around 2% of women of childbearing age as of 2010.

6. Women may not necessarily be more willing to talk about having an abortion in countries with the strongest support for abortion rights.

Support for unrestricted abortion is highest in Sweden, France, and Great Britain, ranging from 78% to 66%. But these three countries fall solidly in the middle of the pack when we rank how common it is for people to be close to someone who has had an abortion, ranging from 38% to 31%. This puts them alongside — or even well below — several countries with far more conservative views on abortion where more people report being close to someone who has an abortion.

There may be other factors that explain this — like different ways people interpret the question — but it's notable that these measures don't line up. Mexico* is a particularly interesting example: 39% of people reported being close to someone who has had an abortion, giving it the fourth-highest figure on that question. But it appears to be the second-most anti-abortion country in our study, where more people said they favored a total ban on abortion (16%) than supported unrestricted abortion (14%).

7. Just because a country has liberal abortion laws — and abortions are frequently performed — doesn’t mean the public supports them.

Take a look at South Africa*, where only 21% of people in our survey said they support unrestricted abortion — even though South Africa has one of the most liberal abortion laws in the world. First-trimester abortion has been unrestricted in South Africa since 1996, and second-trimester abortion restrictions are seen as very liberal because they include concerns like mental or socioeconomic harm to the mother from her pregnancy.

Roughly 90,000 women in South Africa had abortions in government clinics in 2012–2013, according to the most recent years of data available from South African health officials, but experts estimate that the real total could be even higher: Around 50% of the abortions performed in South Africa are done illegally, because of backlogs in service at public hospitals.

8. There seems to be popular support in Ireland for changing its restrictive abortion law.

Ireland’s social policy is in the spotlight after it became the first country in the world to establish marriage equality by popular vote last month. While our poll does not show that Ireland is also moving rapidly to the left on abortion — only 37% of the population currently supports unrestricted abortion — there is a solid majority in favor of allowing abortion in circumstances now prohibited under Irish law.

Ireland had a total ban on abortion before two years ago, when the law was changed to allow abortion only when a mother’s life is in danger. This was the result of a court ruling following the death of an Indian woman named Savita Halappanavar who miscarried and then died in an Irish hospital after being refused an abortion.

In addition to the 37% of people who favor no restrictions on abortion, our survey found 38% of Irish voters said “abortion should be permitted in certain circumstances, such as if a woman has been raped.” Only a total of 15% said they favor the existing life exemption or support a total ban on abortion.

Attitudes toward abortion and marriage equality don’t always go hand in hand.

We took a look at this chart in our story last week on global attitudes on marriage equality, but it’s worth another glance. While Catholic countries in the Americas and Europe either have majorities supporting marriage equality or are heading in that direction — with the notable exception of Poland — support for unrestricted abortion remains very low in many of them, especially outside Western Europe. And a couple of countries that appear liberal on abortion — Turkey* and Hungary — don’t show any signs of embracing marriage equality anytime soon.

*The seven countries asterisked throughout this post have too many people without internet access for us to be sure that our online survey was truly representative of the country as a whole. We only included ones where we believe our findings appear likely to still reflect broader public opinion based on data from other sources, but we can’t be 100% sure that the data for these countries are truly representative.