Making it count

One of the most frustrating aspects of being involved in a social justice movement is coming to grips with the sheer scope of the problem. Social inequalities are grounded, more often than not, in centuries of history and the evolutionary detritus of human cognition. We can point to a handful of successes like the American civil rights movement, but those were foughts that people literally bled and died for, and resulted in a system that almost immediately adapted to restore as much of the racist status quo as was legally permissible. The fact is that the fight for equality is gigantic, and it’s easy to feel as though one person can’t do much to move the massive edifice the dictates the roles of various groups in power dynamics.

Indeed, even if one wasn’t so overwhelmed by the sheer scope of the problem, it’s hard to conceive of what actionable solutions are available. The whole Occupy movement was heavily criticized for even trying to get together and spell out all of the problems. When solutions were offered (and they were offered), their very existence was denied or ignored because it fit into the more easily-digested narrative that we live in a world where people cannot solve big, diffuse problems. Certainly those who are sincerely interested in, say, seeing the end of racism can see few avenues toward true progress: the problem is inside people’s heads. How can we fix the ever-warping landscape of human psychology aside from waiting for the ‘racists’ to die off and hope that the next generation does a better job?

While I agree the task is daunting, there may be one specific lever we can exploit:

The quest for a state that “looks like America” is understandable, but the reality of lived life is more complex. And not just in racial terms (e.g., the division in politics between the white suburbs of Maryland vs. Virginia on either side of D.C.). But keeping race in mind, one consistent finding in social science is that Americans actually tend to overestimate the number of minorities. Iowa is actually more typical than we think, despite the fact that it is not typical. In the year 2000 the General Social Survey asked respondents to estimate the number of various groups in the USA. The finding of a tendency to overestimate minorities, and underestimate non-Hispanic whites, was confirmed. But, I decided to break this down by demographic. The results are below in a table.

The first row are real counts from the 2000 Census. All the following rows are average estimates of a set of respondents in the year 2000.

A table from the survey

The way to read the table is to look across the rows of respondent groups to get a summary of what the average person estimates the proportional size of the other groups to be. The top row gives the actual proportions from the 2000 census*.

So the thing that should immediately jump out at you is just how wildly inaccurate the estimates are. If you look at the ‘Total Sample’ row, the average American believes that ze lives in a country that is 59% white, 31% black, 25% Hispanic, with the remaining 36% evenly divided between Asians and Jews. Those math wizards among you will probably have noticed the most glaring part of the error – that’s 151% of the actual population. Some of this, to be fair, could be double-counting of people who identify as more than one race, but none of the minority groups were counted anywhere close to accurately.

The thing that jumped out for me, simply because it went quite strongly against my a priori guess, is that there doesn’t really seem to be much of a relationship between race and the accuracy of estimates. I thought perhaps that members of minority groups would have a better picture of what percentage of the population was represented by each group, but that doesn’t appear to be the case. If anything, it looks as though black and ‘Hispanic’** people tend to overestimate their own numbers, an effect not seen in white respondents. One potential explanation is that segregation within minority communities leaves people with a skewed idea of how many people who look ‘like them’ there are in the country at large, but I have no ready explanation for why white Americans would underestimate their numbers. It’s also worth noting that young Americans were the ones who got it the most wrong, and yet their attitudes are the ones that are, demographically speaking, closest to ‘right’ when it comes to eradicating racism.

Whatever the explanation of the findings, one thing is clear: the United States lacks basic demographic knowledge of its racial makeup. Is it so hard to imagine therefore why ‘the welfare state’ is conflated with black and brown folks (and the image of Mitt Romney’s 47%) when the country believes itself to be more than twice as black and brown as it actually is? Would we see the cities and suburbs differently if we understood just how racially divergent those two different types of communities are? What lessons can we learn from the fact that everyone seems to think that America is less white than it is – does that relate somehow to the finding that white Americans feel that anti-white persecution is more prevalent than anti-black persecution?

Here is one of those rare opportunities to do something that is easy, obviously fact-based, and can trigger some critical analysis of race dynamics: ask people to guess how many people of colour there are in the country. This is a question that clearly needs consciousness-raising, and regardless of where the conversation goes next, having more facts can’t be a bad thing.

Like this article? Follow me on Twitter!

*It is undoubtedly the case that these numbers have changed away from white statistical domination in the 12 years since the 2000 census. You can see more recent figures here.

**A problematic term, since Hispanic etymologically means ‘from Spain’, which is a gross misrepresentation of the descendants of those Central and South American countries whose families are indigenous to the region rather than descended from Spanish colonizers.


  1. arno says

    My guess for the discrepancies in estimated and actual numbers would be that the few non-whites a white person sees are more memorable than the other whites. A similar effect shows up when people from a male-dominated academic discipline look at e.g. a speaker list with slightly better balanced numbers of male and female speakers – they overestimate the number of females on the list.

    On the other hand, I would assume that members of a minority overstimate the numbers of that minority simply by sampling bias.

  2. says

    Maybe it is also the fact that the distribution of people is largely uneven? This is just a guess based on my personal experience having lived in the biggest big city in America AND in ‘Mayberry'(I drove to work on Andy Griffith Parkway, past Barney Fife’s police car).

    The places where people are packed most tightly are urban and contain more non-whites. The places that are people are more loosely distributed contain more white people, but there are more of those places than anything else in America which is part of why there are more white people overall.

  3. maudell says

    Interesting. I would be curious to see the count in Canada (particularly people’s estimates of First Nations). I suspect it would be quite high (with the amount of times I hear people believing we are being overloaded with lucky tax-free moochers refusing to assimilate… but then it may be my own cognitive bias)

  4. scorinth says

    It’s a bit weird in situations like at my university. If these numbers are correct, it seems like minorities are somewhat over-represented. But then, I know that students from abroad are not an insignificant portion of the student body here.
    Now it makes me wonder about the proportion of the students that are minorities, the proportion that are not permanent citizens, and how those two groups overlap. (It’d be a lot easier to notice a visitor from Africa than a visitor from Western Europe.)
    Anyway, I’ll be sure to start quizzing my friends and colleagues.

  5. carlie says

    I have no ready explanation for why white Americans would underestimate their numbers.

    As arno says, I think it’s because it stands out in a white person’s mind when they see someone of another race, whereas all the ones like them blend into the background (not seeing the forest for the trees, or the water if you’re a fish, whatever).

  6. says

    My sister-in-law, whose father is from Mexico, and mother is from a midwest First Nation has told me that the term Latino/Latina is preferable to Hispanic. I’m not sure how that helps, but I will defer to the right of people(s) to refer to themselves with terms of their own choosing.

  7. jamessweet says

    It’s also worth noting that young Americans were the ones who got it the most wrong, and yet their attitudes are the ones that are, demographically speaking, closest to ‘right’ when it comes to eradicating racism.

    I’m not sure how worth noting this really is, given that every demographics’ guesses were pretty tightly clustered. I don’t know what the error bars are, but it sure doesn’t look like a statistically significant deviation. Across all demographics, people seem to have a pretty equivalently skewed idea.

    I wonder how I would have done, if I’d been asked to guess before reading this… I would have nailed the number of Jews, since that is a factoid that I just happen to know off the top of my head. But the others, not so sure…

  8. No Light says

    151% eh? There’s that American exceptionalism in action, every person is being half another person too!

    The situation is likely similar here in the UK. We have 7-8% non-white citizens.

    My late father-in-law used to get practically apoplectic about the number of “them”, who he was convinced were “taking over” the country. This was fuelled by television programmes like Border Patrol, and irresponsible tabloid “journalism”.

    What he actually saw with his own eyes could not convince him. In our village of 3500 people there are 11 people who are not “white British”. Yep, so few that I can count them, I know where they (or their parents) live and work, and I know their names. Hardly a takeover, not quite the invasion the press are reporting. A massive 0.3%.

    But a 78 year old man, raised in an age when the media were a trustworthy source of information, could not believe the two young women who were telling him that the news just could not be trusted. He’d get visibly upset, and say “They wouldn’t be allowed to say stuff or print stuff if it wasn’t true”. He was the same with media scare stories about violent crimewaves led by rampaging teens. Despite the fact that our community is about 50% old age pensioners, he’d imagine that gun-toting (black) kids were hiding behind every bush.

    The only children of colour are one four year old girl and her big sister who’s six. They live with their parents behind their Chinese takeaway, and the worst thing they do is have the odd slappy quarrel over the proper pronunciation of local dialect words, when they should be doing their maths homework. No muggings, hold-ups or burglaries. Didn’t stop Dad from being scared to leave the house sometimes.

    Come the revolution, the first backs against the wall should be the scumbags of the world who make their money by terrifying vulnerable people.

  9. tychabrahe says

    Another thing to consider is that in an effort to reach out across the spectrum, there has been an emphasis more and more on cross-cultural advertising. If you search a stock image site like Shutterstock for groups of doctors or business people, you get mixed race/mixed sex groups.

    If your business happens to employ 90% whites and 70% men, because that’s pretty typical of the people available to hire in your particular technology field, you don’t dare reflect that in your company documents. I work in Chicago in a small software company. Fourteen of our fifteen employees are White. (Our CIO is of Mexican descent.) We do have a large percentage (1/3) that are of Central European descent, Polish and Croatian. Our outsourced publications feature mixed race groups, just like the ones above. (Our in-house publications usually feature a White man and a Southeast Asian woman, our Marketing director and his Burmese wife.)

    The Friendly Atheist recently blogged about a campaign of atheist billboards in the Pacific Northwest. The billboards featured Whites exclusively. Several people commented on it. People from that region pointed out that over 88% of Oregonians are White.

  10. octopod says

    For what it’s worth, if you assume that people are counting all “Hispanic” and “Jews” (sic) as white, the average total percentage is 108.75%, which is a hell of a lot more sensible than 151.7%.

    Don’t know whether this is a reasonable assumption but it sure reduces the excess.

  11. Sri says

    I think part of why everyone underestimates the number of white people is due to the highly prevalent meme of the shrinking majority. Conservatives try to stir a lot of panic with about areas where whites aren’t a majority anymore or even a plurality.

    There was a to do over the article earlier this year stating that for the first time (since they started recording it) the number of white babies born did not account for a simple majority of births in the United States. With all those narratives floating around, I’m not surprised that everybody lowers the number of white people (before I looked at the chart I was guessing closer to 60% than 70%).

    My guess is that young people have increasingly bought into the idea of the shrinking majority since it’s been yelled about for most of their lives but that they don’t think it’s important. I know that I and my friends laughed at people in an uproar about the baby article because we thought it was stupid to get so worked up about not being a majority anymore.

  12. Kalirren says

    I’m curious to what extent young peoples’ overestimation of minority abundance in the population at large comes from an actual demographic change in their peer demographic. Asians, and to a lesser degree Hispanics, are by and large newer immigrants to the United States than Whites and Blacks are. They/We have been here for fewer generations, and are underrepresented in the older segments of the population.

    I know I myself would have nailed the Black figure at 12.5, but I’d have put the Hispanic figure at around 17-20 %. Then again, that impression comes from having lived in Los Angeles, where that’s probably justifiable.

    So maybe the overestimation of minorities comes from living in a de facto segregated America, because when we think of seeing X people we think of seeing them in their X neighborhoods, where X people are common, and don’t think of just how common or rare those neighborhoods themselves are. The underestimation of whites is probably obtained by subtraction…

  13. smrnda says

    I agree that it’s lots of white panic over no longer being a ‘majority’ and the idea that this isn’t the America they grew up in. This might be why baby boomers and elderly folks who benefited from a social safety net want to shred it for future generations, the future generation are ‘the wrong kind of people’ and some older white people consider the less racist younger generation of white kids to be race traitors they don’t care about.

    Something in my own life. I came of age on the south side of Chicago and was used to being the only white (and Jewish at that) person in a given place pretty frequently. Then, when I went to college, the town I was in (to me) seemed to have absolutely no Black people whatsoever. Of course, kids from suburbs thought it had a lot of Black people, but we were coming from radically different baselines.

    A more general question – with so many Americans so poorly informed about facts they could have found out in a ten second internet search, how is democracy even possible? Also, is there some instinct to pull numbers out of one’s ass rather than actually doing the research and looking them up? If someone asked me a question like this, I’d be checking data, not doing guesswork.

  14. Sri says

    I would assume this was a study or survey done in person or over the phone (with some sense of immediacy at any rate) – so that they could check people’s assumptions, not their Google-fu.

    If researchers want to know how people view the world around them, they want the answer that the person has been subconsciously working from, not the answer the person can look up.

    My main confusion is how people managed to toss out these guesses without realizing that they added up to way more than 100%. It would be interesting to see how the survey was presented.

  15. left0ver1under says

    Whatever the explanation of the findings, one thing is clear: the United States lacks basic demographic knowledge of its racial makeup.

    Their demographic ignorance matches their geographic ignorance. The linked study was from 2006, so some of those “students” are now adults.

  16. IB says

    Back in the 60s and 70s, ‘respectable’ newspapers like the Telegraph happily ran headlines like ‘Asian Invasion’ inresponse to the plight of those Ugandan Asians expelled by Idi Amin (and they were British Citizens from the get go).
    If anything, the British press has gotten better over time, mainly through peer pressure and changing terminology (its all about the muslims now), and shifting the immigrant blame game to newer arrivals who are safe targets (East Europeans, Somalis etc).
    And people are still willing to believe the bollucks that gets printed.

  17. jesse says

    I’ll throw in with the simplest set of explanations:

    1. sampling error. I live in NYC, and it’s a majority-minority place, and even if you ask folks who know that NYC isn’t like the rest of the country, they’ll go with what they see. It’s a bit like the Copernican principle: you assume the place you live in isn’t that unusual.

    2. “Shrinking majority” meme

    3. The increased visibility of minorities and women skews people’s estimates. Someone above mentioned that this happens to people in male-dominated professions.

    3. Basic ignorance

    I mean, I don’t think any of these are particularly magical forces, and they actually make perfect sense.

    At a more minor level, there’s a whole thing you could get into about self-identification and the like, and how that affects estimates, though the interesting thing to me about the table is that the wrong estimates are all roughly the same — that is, the factor by which people are wrong doesn’t seem to vary much: note that every group asked came up with roughly similar percentages of various races.

    That would say to me there’s an interesting study to be done on how people do mental math/ cognition in relation to questions like this.

    It is worth saying, though, that even a country that is ~70% “white” has at least three majority-minority states: Hawaii, New Mexico, and lately California and Texas. That’s going to have an effect on electoral politics, if nothing else. So it isn’t like the issue of a shrinking white majority is silly or anything, given the way the US political system is set up and a small portion of the electorate can swing elections and districts.

    (And y’know, there’s a weird kind of karmic justice in the fact that the reason for setting up the Electoral College was in part to make sure some people could keep owning slaves. Well, if loads of minority voters move to Nevada, Arizona, Virginia and Texas, that same system is going to bite the descendants of those slave owners).

    And for the record, my rough-and-ready estimates when asked were always ~10-12% black, ~15% Latino, ~5% Asian, ~2% Jewish and the rest white and a smattering of others with a fudge factor for how people self-ID. But I think the Latino number and Asian number were off b/c I was using the 2010 census when the Latino population outstripped the black population for the first time, and I grant that I had actually looked it up often enough (it comes up in my work sometimes).

  18. F says

    Another bit of red.

    Of course it might seem like “they” were “taking over”, especially during the period when “we” were finally kicked out of “their” countries. (The incredible shrinking British Empire.)

  19. F says

    Plus, people suck at statistics; asking them to intuit a statistic, you are likely doomed. But the influences on their inaccuracy can be telling. Although any noticeable effects are probably best examined statistically as well. Having people examine their own failed guesses is a great idea on a personal level, though.

  20. Sivi says

    Coming late to this, but having talked over the numbers with some people, I really want to know what the results would be if participants were constrained to 100% totals, rather than (I assume) being asked to just give percentages of population in separate questions.

    I feel like the numbers wouldn’t be quite as off, and might better reflect underlying differents of demographic perception by various groups.

    Also, it seems weird that they include Jews and Asians as questions, but don’t have a category for Asians, and have Jews sorted among the religious groups rather than the ethnic groups.

  21. tim rowledge, Ersatz Haderach says

    ‘hispanic’ could be even more problematical given the history of indigenous population crash and ‘replacement’ by mostly African slaves during the early years of the European colonizations. It was very variable by exact area but Brazil had a huge influx of slaves, making their genetic heritage strongly African. Other regions might have have been negligibly affected and thus have a very different heritage.

    Yet another example of the inane attempts that people make to categorise other people in order to self-justify abusing them. Any bets on it stopping any time soon?

Leave a Reply

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