You may remember the results of a survey circulated by IPSOS a few days ahead of Pinkdot, making some interesting claims about the local attitudes towards homosexuality and inclusion.Now, I reached out to IPSOS for the details about the methodology but there really isn't much more to it than has been reported. The survey was carried out using online tools on a group of 500 respondents, reflecting the demographic makeup of the society (race, age etc.).
And this is where problems begin.
First of all, a 500 group is still small, even if could be statistically sufficient for the purpose. The responses should be representative up to +/- 5% for the entire population - but this means, effectively, a 10% wide margin of error. It is quite large even if the fundamental requirements of the survey have been fulfilled (e.g. that the respondents were correctly selected at random).One of the many problems with public opinion surveys is that they are only accurate as long as they were made without errors. This is why, for example, most public pollsters in America made catastrophic blunders in surveys ahead of elections in 2016 and 2020.Given the small sample, IPSOS' survey could also be affected but we won't know unless we poll the entire population and compare results.Secondly, even the results published already suggest that there may be something wrong with it.Remember how activists claimed the attitudes are changing and the society is ready for a repeal of 377A? And yet, according to IPSOS only 1/5 or 20% of Singaporeans are opposed to it, compared to 44% who support it. The rest just don't really care much.In other words - 80% of people in Singapore don't mind 377A, with over half of them supporting it.However, if you take a closer look, you will see that only 81% of Singaporeans identify as heterosexual. A whopping 12% reportedly identify as homosexual and the rest would prefer not to say - which we may assume is a group dominated by non-heterosexuals.
So, let's get this straight - as many as 19% of Singaporeans identify as non-hetero but only 20% are opposed to 377A? Does that mean that virtually no heterosexuals are in this group?I think the assumption that support for a repeal of 377A would be near 100% in the LGBT community, which means only 1 percentage point of those opposing 377A are heterosexual.In other words, either the survey is wrong or LGBT community in Singapore has almost no "allies".
But let's return to the basic figures - 12% of people identifying as gay would place Singapore among the absolute global top. Really?For reference, the latest surveys in Australia put the local gay/lesbian population at around 3.5%. In the UK the figures provided by the government show 3.1% identify as LGB. In the US the figures go as high as 7%. In the gay-friendly Netherlands the government estimates put it at 4 to 6%.But 12%? In Singapore? Really? Someone really saw that result and thought "I see nothing wrong about that"?Mind you, this is 12% reportedly across the ENTIRE population and all age groups, not only the young.We can observe that earlier generations show lower rates of LGB identification, which means that the Singaporean youth could be 1/5 or more gay.Seriously?Singapore would immediately become the gay capital of the world.
What really bothers me is that someone looked at it and at no point considered something might be off.I mean, these figures don't make sense.
This takes me to my final point - I think the sphere of public surveys on sensitive topics should be highly regulated. Because it's easy for private sector companies to both abuse their position and peddle misinformation (regardless of whether it is intentional or not) that the media greedy for eye-grabbing headlines are eager to circulate even further.I don't think the government should have a monopoly on surveying the public, but private companies should fall under strict rules for minimum samples and methodologies employed, on topics that may have significant influence on public opinion - and/or may mislead them about the true sentiments of the society.Throwing together an online survey for a few hundred people and not looking twice at results which look questionable at best does more harm than good.
