Gay population must be higher than thought
by Matt | December 28th, 2006 |In my work during the past 6, almost 7, years on LGBT issues and activism, I’ve run across numerous numbers purporting to be the “true” or “most accurate” percentage of the U.S. population which is either gay or lesbian.
Those numbers have ranged from 2% all the way up to an unremarkably unbelievable 30%.
The number I’ve heard most often and the number I trust is anywhere between 4% and 6%.
I believe that if you talk to any reputable scholar, LGBT activist, “traditional family values” activist or just any body familiar with this stuff, they’d agree that, most likely, the 4%-6% number is probably the most reliable (for the record, I put absolutely no stock or faith in the 10% number pushed by Alfred Kinsey - I don’t believe it is reliable at all).
But… I have to question that now, since I ran across this tiny segment in an article from 365gay.com News on Edwards announcement to run for Mr. Prez:
The LGBT vote has emerged as a sizable, discrete voting bloc of 4 to 5 percent of the vote in national Congressional and Presidential elections, according to Voter News Service data. The figures show that in large metropolitan areas that percentage is much higher. Openly gay voters account for 9 percent of the vote in large cities and 7 percent of the vote in medium-size cities.
All of us who are even the slightest bit interested or involved in politics knows that voter participation is NEVER as high as the total population of any demographic. Most times, voter participation NEVER comes ANYWHERE CLOSE to the total number of any demographic. For that matter, you can even say that most times voter participation almost never reaches close to just the number of registered voters, which doesn’t include every person who may be considered a part of that demographic.
So… I’m just kind of venturing out on a limb here, but… If the LGBT “voting bloc” is thought to be at about 4%-5% nationally, then the total number of LGBT people must be MUCH, MUCH HIGHER.
According to the Federal Elections Commission, percentages of voter turn out (out of the total voting-age population) were as follows for the following most recent, major election years:
- 2004: 55.3%
- 2000: 51.3%
- 1996: 49.1%
- 1992: 55.1%
EDIT (12/29/06 11:30am): Note: The voter turn out is the percentage of the total VOTING-AGE POPULATION (all those American citizens aged 18 and over); it is not the percentage of just those persons registered to vote (i.e. voting population).
The average of those four percentages is 52.7%. Now… just for fun (and I don’t think I’d count this as entirely academically or scientifically accurate, but it gives us some idea), let us apply this voter-turn out percentage to the percentage of LGBT voters in recent elections.
4% (LGBT voting bloc) / 52.7% (total voter turn out - out of total voting-age pop.) = 7.6%
According to this - and remember it is no where near 100% reliable - the possible total number of LGBT people in the United States equals approximately 7.6% of the total voting-age population of the United States.
I would say, however, that the voter turn out among LGBT people is most likely higher, percentage wise, than the national average, due in large part to the heightened sense of politics and its many times harmful and negative effects upon this particular community. Most often, you’d find that among many minority populations, voter turn-out is higher than the national average. Therefore, this is why my little math isn’t 100% accurate; in the math-work, I use the national, voter turn out average.
So… What is the main message here? You will never find that 100% of any given demographic or group turns out to vote. NEVER. It just will not happen. The “LGBT voting bloc” is estimated at 4%-5%. Because 100% turn out never happens, the total percentage of the U.S. population which LGBT people represent MUST be higher than what is usually purported to be the “most reliable” figure of 4%-6%. There is no doubt about this, at least in my mind.
Thoughts? Come on, Ryan… you’ve been good a “ripping me a new one” here lately…
Technorati Tags: politics, voter turnout, gay, demographics
13 Responses to “Gay population must be higher than thought”
I have not taken the time to explore your logic on voter turnout/LGBT population percentage, but I find your end result and beginning comments interesting. I had always thought that LGBT people did represent 10 percent of the population… i will have to look into that more.
By Brandy on Dec 28, 2006
I think the logic is flawed, as you admit… I don’t think there’s a way to accurately measure the LGBT voting bloc. I’m not sure how VNS does it, but you could definitely ask Prysby, I think he used to work for them. It’s pretty unlikely that you’d get accurate survey data, because many gay people will either not admit it on the phone to a stranger, or would be turned off by the question and hang up (invalidating the rest of their responses). So if anything, I’d say that number is probably too low.
But yeah, unless there was a way to do it that you could guarantee more accurate results AND compare this to the size of national LGBT populations, I don’t think you can estimate it effectively.
As for the 10% guess… umm, yeah right. I’d say 6% is even too high. I have nothing to base that on, and maybe they’re all in the closet, but - I dunno.
I don’t think I’ve necessarily been “ripping you a new one” lately, I’ve just been reading more of your entries and commenting because most other bloggers are MIA right now; holidays or something, it’s leaving me with a dearth of news to read during the day. So much so that I even read the article you linked to to see what they’re saying about Edwards… he looks gay, he’s a Dem, and he claims to be from NC - sounds like a campaign for you to work on!
By Ryan on Dec 28, 2006
Your math is flawed.
You err when you assume that since about 50% of people vote, you can multiple the gay voter percentage turnout to extrapolate the gay population.
Here’s an analogy for your process:
1) There are 100 candies in a store, four are pink (4% gay).
2) A shopper buys 50 candies (voter turnout).
3) When she gets home, the shopper counts her candies. She observes that of her 50 candies, two are pink — just as we’d expect: there were four pink candies in total, the shopper bought half the total candies and ended up with half of the pink candies. Correctly, she notes that the two pink candies are 4% of her 50 candies.
4) Now the error: The shopper thinks that since she only bought half the candies, she can double the percentage of pink candies and come up with a percentage of the total candy population. She does so and comes up with 8%. Cleary, we know that’s wrong because the total number of pink candies was only 4 out of 100 to begin with.
So, as far as voter turnout as a measure of the gay population, the best one can do is take it at face value (well, I suppose one could specualte that gays or more or less likely to vote, but my impression is that it is probably a reasonably close approximation of the gay population.)
By Roch101 on Dec 28, 2006
I admit my math is flawed… my point was this: The most commonly accepted total of the gay population is 4%-6%… the voting bloc is 4%… You will NEVER find that 100% or even just a super-majority of any demographic ever turns out to vote… therefore… It is possible that the gay population could be much higher than the most commonly accepted 4% to 6%.
Again… I admit my math is flawed, but this is something to think about.
By Matt on Dec 29, 2006
No, Matt, it is still incorrect to use that extrapolation with voter turnout as your starting point. It is possible, certainly, that the gay population is higher than 4%, but voter data doesn’t point to that.
You are saying that 2 out of 50 may really be 4 out of 50, but if you are going to extrapolate by including the non-voting gay population, you have to bring along the rest of the non-voting population too. You cannot double the subset without doubling the total. So rather than the flawed “2 out of 50 may actually be 4 out of 50,” you can only speculate correctly that 2 out of 50 may actually be 4 out of 100.
If you answer only 50 questions on a 100-question test and get 2 correct (4%), it is illogical to say that if you had answered all of the questions you would have gotten 8 correct (8%). If you got 2 out of 50 questions correct and had answered all 100 of the questions, you likely would have gotten 4 out of 100 questions correct (still 4%).
What you are trying to support is that 2 out of 50 may really be 8 out of 100 and there is no evidence that the non-voting population includes twice the number of gays than does the voting population. If the gay population is equally dispersed among the voting and non-voting population, then 4% is 4% whether your sample is 50%, 10% or 100%.
By Roch101 on Dec 29, 2006
Roch I think you still misunderstand… Read the post and look at the Federal Elections Commission data again… The voter turnout (approx. 50%) is NOT based of “voting population,” but rather VOTING-AGE POPULATION. Big Difference there… although the math is still flawed with that, it isn’t as flawed as you say.
Again… my whole point, minus all the rambling is that 4% of the voting bloc is LGBT. Groups of people NEVER have a voter turnout of a super-majority or 100%, therefore the LGBT population MUST be higher than most commonly thought.
By Matt on Dec 29, 2006
I must not be doing a very good job of explaining this. Yes, 50% of the voting-age population votes. Of that, 4% are gay. That means that 96% of those who turn out to vote are heterosexual.
Now, as you observe, groups of people never have a voter turnout of 100%. This holds true for the heterosexual population too. You suggest that because a group never has a 100% voter turnout and, because only 50% of the voting age population votes, you can double the percentage of a group who does vote to extrapolate their percentage among the voting-age population. So, the 4% of gays who vote becomes 8% if we include those who don’t vote. By the same measure, the 96% of heterosexuals who vote becomes 192% if we include those who don’t vote. See?
The only way you can say that the total gay population is a greater percentage of the total population that it is of the voter turnout is if their population is greater among those who don’t vote.
Here’s the math required for your assumption to work:
Total voting population = 50
of which:
——————————–
Voting gay population = 2 (4%)
Voting hetero population = 48 (96%)
Total non-voting population = 50
of which:
————————-
Non-voting gay population = 6 (12%)
Non-voting hetero population = 44 (88%)
————————
Addind voting and non-voting populations, we get:
—————————
Total: 100 (100%)
Total gay: 8 (8%)
Total hetero: 92 (92%)
This gives you the total 8 out of 100 for the total gay population (8%), but it does so only by reducing the hetero population from 96% among the voting population to 88% of the non-voting population. There is no evidence that the percentage of heterosexuals is lower among non-voters than it is among voters.
By Roch101 on Dec 29, 2006
Roch… this post wasn’t mean to be an honest-to-God, this is the truth and these are accurate numbers type thing. Thanks for all the math work, though.
No matter the numbers, no matter the percentages or what math is right or what math is wrong, my original point stands… The total LGBT population must be larger than what is commonly accepted (4%-6%), because voter demographics never turnout at 100% or close to 100% rates. That was the point I was trying to get to and I used very simple, admittedly flawed and, as you point out, not reliable math, to get there… but it proved my point in a very simple way that many people could understand.
By Matt on Dec 29, 2006
Here, try this:
The total population is 100. 50 of them vote.
Of the 50 who vote, 2 are LGBT (4%).
Of the 50 non-voters, what is the minimum number of them that must be LGBT for the percentage of LGBT among the total population (50 voters + 50 non-voters) to be greater than 4%?
a. 1
b. 2
c. 3
d. 4
By Roch101 on Dec 29, 2006
Time’s up. Please mark your answer and pass your paper to the front.
By Roch101 on Dec 30, 2006
C. 3
Although I’ve never been good at math.
Again… yes… the math is flawed… but could you, simply in the name of making my soul feel better, at least back me up on the original point I was trying to make?… The total LGBT population must be larger than what is commonly accepted (4%-6%), because voter demographics never turnout at 100% or close to 100% rates.
lol
By Matt on Dec 30, 2006
Correct!–and thus you illustrate the inescapable conclusion, one that completely refutes your original point that the total LGBT population must be larger because voting groups never turn out at a 100% rate.
Because, for the percentage of LGBT in the total population to be larger than the percentage of those who vote, their number must be greater among those who don’t vote — as you just illustrated by correctly picking “c.”
The corollary to that is that the heterosexual population among non-voters must be lower than it is among voters (47 out of 50 compared to 48 out of 50).
Now, it is entirely possible that this may be true, that the LGBT population is greater among non-voters than it is among voters, but we don’t have any evidence in this discussion that that is so. In fact, you speculated the opposite above, Matt, that because of a “heightened sense of politics,” LGBT may actually be more inclined to vote than other populations.
So, without some additional evidence, and taking exit polls at face value, this voter data doesn’t, at all, support the notion that the LGBT population is above 4 to 6 percent. It may well be, but this data can’t be extrapolated to make that case.
By Roch101 on Dec 30, 2006