12

Background

This site is supposed to be about the following:

Specific issues with governments, policies and political processes

Real problems or questions that you’ve encountered

And not about the following:

Anything not directly related to governments, policies and political processes

Questions that are primarily opinion-based

Questions with too many possible answers or that would require an extremely long answer

However, it is impossible to entirely remove bias from how questions are asked. Likewise in answering, voting, commenting, and moderating. Every action has the possibility of some bias behind it. Add the word "Politics" into the mix, and it is inevitable that even within the accepted scope of "specific issues with policies" with lead to a discussion of opinions. We must all aim to be responsible in how we do our actions, but there is no guarantee that is the case. Not everyone reads meta, and there is no presence like academic or scientific rigor elsewhere that exists here. Only the collective actions of the community and the Stack Exchange company.

Question

Because of this reality, it is worth asking the following question "Does Politics Stack Exchange, as a whole, have a bias"?

Corollary

Furthermore, in answering that question, some sort of quantification or measurement must be taken (for the answer to the above question to elicit fact). But whoever partakes of this measurement or task is almost certainly part of this community, and thus the collective bias. Thus I ask the following corollary as well: "How is it possible to objectively quantify this bias, if we, in this community, compose part of it?"

23
  • 3
    This is an international site not sure how it can have a basis
    – Joe W
    Feb 25 at 16:50
  • 13
    Quantifying the site's bias, if it exists, is going to be difficult. I've seen several users with conservative views accuse this site of having a left-wing bias, but I've also seen users with left-wing views accusing it of having a right-wing bias.
    – F1Krazy
    Feb 25 at 17:27
  • 20
    As long as extremists of all political flavors accuse us of being biased for the other side, then we should be at least somewhere near the center.
    – Philipp Mod
    Feb 25 at 17:35
  • 4
    Also, what exactly are we saying when we say the site itself has a bias? Do you mean the moderators? But so much of the moderation comes from users. The users' bias? There are definitely members of the community with a right-wing bias, and members with a left-wing bias, and some with a strong centrist bias, others with a strong pro-India bias, and many more. Can you just average the bias? Count the number of "biased" members?
    – divibisan
    Feb 25 at 21:45
  • 3
    @JoeW Tell that to people calling for "bipartisan" actions on European issues. Just like the rest of the internet, the site leans USA. That's fine, of course, but that does mean that European Overton windows will experience it as inherently right-leaning, even if it's in the form of required vocal defense of what's considered the standard east of the Atlantic. I see "left-wing" defenses that the most right wing parties/pundits in my country would shrug about and say "well, yeah, duh...".
    – DonFusili
    Feb 26 at 10:16
  • 2
    @DonFusili People? Sure there are people on the site with a bias but that doesn't mean the site as a whole shares that bias.
    – Joe W
    Feb 26 at 15:21
  • 8
    @divibisan Yes, you can average bias. If the number of left leaning members outnumbers the number of right leaning members, on average voting will tend in that direction as well. One can earn additional upvotes by repeating the usual Democratic talking points (and throwing in a swipe at Trump for additional points), while repeating Republican talking points earns additional downvotes.
    – Sjoerd
    Feb 28 at 18:15
  • 6
    @DonFusilli Democrats are on a number of issues far-left in Europe. Among others: Third Term Abortion (that is: abortion allowed till the day before natural birth), $15/hour minimal wage (higher than any European country, even in PPP - en.wikipedia.org/wiki/… ), and Critical Race Theory.
    – Sjoerd
    Feb 28 at 18:20
  • 2
    @Sjoerd None of which are mainstream in the Democratic party. The $15 minimum wage didn't get anywhere, and late term abortion and critical race theory are mostly far-right boogeymen without practical implications. Apart from that, France eg has a PPP of 12.23 for a 35 hour work week. Adjusted to 40 hours, that's almost $14. If you add in the guaranteed 5 weeks (minimum) paid vacation (vs 0 in the US) & 11 paid public holidays (again vs 0) you get above $15. If a minimum wage of $15 is far-left compared to that, what is the current 7.25? Far-far-right?
    – tim
    Mar 2 at 8:00
  • 1
    @F1Krazy Have there been any liberals accusing left leaning, or conservatives accusing right leaning?
    – frеdsbend
    Mar 3 at 3:47
  • 1
    @Philipp "Complaints from both extremes means we're even" is interesting, but not a very good proxy for measuring bias within the center half or three quarters. For example, a "left of center" will of course be criticized by right extremes, but also likely by left extremes because the nature of extremism is that everything else is "influenced by the other side".
    – frеdsbend
    Mar 3 at 4:05
  • 6
    Also, not all users participate in all tags. For example, I usually browse new questions with a q=-[united-states]-filter to get rid of most of the US stuff I'm not interested in, and I usually abstain from voting on US specific questions and answers as well.
    – Hulk
    Mar 5 at 12:38
  • 3
    @Phillip That is a particularly strong case of the golden mean fallacy (argument to moderation fallacy), and I am disappointed (but not surprised) in the number of upvotes. Extremes from both sides expressing something is biased does not imply a tendency towards the thing being at the center or even balanced. frensbend describes one example of how this can happen. Another example is if side A expresses a thing is too biased to side B, and people on side B express the thing is biased to side A in order to gaslight the valid concerns of those of side A. This is a remarkably common phenomenon. Mar 24 at 5:42
  • 1
    By Polish standards - visible, but moderate left wing bias in net result of up/down voting clearly on social issues, somewhat as well on economic issues. Answers as such can be from any side from political spectrum which makes them sort of balanced. On touchy issues I learnt to look to heavily downvoted answer to find some valuable stuff which I would not consider as good indicator of neutrality.
    – Shadow1024
    May 26 at 20:17
  • 2
    Though if we try to quantify the bias, I'd expect that there are differences pending which type of question is being discussed. There is let's say issue of US politics (where everyone is American or anyway emotionally invested in spite of living ocean away), ME conflicts but there are less divisive questions as well. There are questions concerning ex. STV where there are no tribal believes and the discussion there seem without any visible bias.
    – Shadow1024
    May 26 at 20:28
5

SE can be extremely Marxist in some cases. Like getting great many upvotes for claiming labor markets only exist in the fetid minds of the right wing. That's an answer from the same user who says there is no bias in SE.

I'm sure one can find more examples of one kind of weirdness or another. But like I said in some comments, it's going to be extremely hard to quantify the bias across all questions and answers unless one is willing to conducts some studies that aren't cheap to conduct.

Things about SE that were easy enough to study like early answers getting more votes than later ones were studied in academic papers as well.

If you're curious how such content-bias studies are conducted when there is enough motivation; you can look at how some such studies were conducted regarding Wikipedia's contents. Mind you, any such study involves either some kind of automated process that ranks contents by some keywords or uses human judges to score a small sample of articles/pages. Neither method is exactly foolproof. An example:

In Is Wikipedia Biased? (2012), the authors examined a sample of 28,382 articles related to U.S. politics as of January 2011, measuring their degree of bias on a "slant index" based on a method developed by Matthew Gentzkow and Jesse Shapiro in 2010, to measure bias in newspaper media. This slant index purports to measure an ideological lean toward either Democratic or Republican based on key phrases within the text such as "war in Iraq", "civil rights", "trade deficit", "economic growth", "illegal immigration" and "border security". Each phrase is assigned a slant index based on how often it is used by Democratic vs. Republican members of U.S. Congress and this lean rating is assigned to a Wikipedia contribution that includes the same key phrase. The authors concluded that older articles from the early years of Wikipedia leaned Democratic, whereas those created more recently held more balance. They suggest that articles did not change their bias significantly due to revision, but rather that over time newer articles containing opposite points of view were responsible for centering the average overall.

3
  • 17 upvotes and 5 downvotes counts as a site wide basis?
    – Joe W
    Apr 10 at 1:19
  • @JoeW: who said that?
    – Fizz
    Apr 10 at 2:05
  • You pointed to an answer getting many upvotes when it only has a total of 17 upvotes and also has 5 downvotes. Though looks like with the new attention it has gotten 3 more upvotes for a total of 20.
    – Joe W
    Apr 10 at 5:11
4

There's an old saw I heard once that's worth considering:

Philosophy is the art of making a story that fits the facts; rhetoric is the art of making the facts fit a story.

I think this is doubly true when we turn to politics and society, but unfortunately it often takes some discernment to see the difference. The point, though, is that one is always telling a story, and telling a story invariably invokes a point of view. We cannot escape a trace of bias.

Musings aside, there's a difference in argumentation style between conservatives and liberals that might give the impression of systemic bias. Conservatism runs heavy on what Habermas called the normative reasoning style: arguments founded in ideas that are pre-given as true and immutable. Conservatism usually aligns itself with the status quo and sees itself as arguing from a moral high-ground, and so answers and comments from the conservative side of the spectrum tend to be on the short, straight, simple side. Few people belabor the status quo, because the status quo is (from their perspective) both self-evident and intrinsically good. Liberals, by contrast, usually feel like underdogs, because they tend to argue against the status quo in favor of some (to their eyes) positive change. This desire to 'argue for' leads them towards (Habermas') communicative reasoning style, which aims to convince and reach consensus. It's more deliberate, more complex, more geared towards nuance and fine distinctions, and all-in-all a much wordier style of expression. I imagine that all other things being equal, liberals are likely to use ten words for every word a conservative writes, if only because liberals feel the need to argue and defend things where conservatives will merely assert things as true.

And of course, this doesn't consider those out there just aiming to be 'influencers' (Habermas' dramaturgical style), or the trolls whose only goal is to score points (teleological reasoning style) regardless of which side of the spectrum they come from...

Written communication — even social media communication like this — disadvantages the normative reasoning style, because norms have most power when they are collective group expressions (pragmatically difficult to achieve on a site like this). Some conservative posters manage it, and other conservatives enter into the communicative style, so it works out. Any appearance of overall bias is just that, I think: an appearance, not a truth. The left may have more words by volume, but that doesn't necessarily translate to better answers or more cogent points.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .