The Percentage Sign as a Signaling Device


Jason Kuznicki

Jason Kuznicki is a research fellow at the Cato Institute and contributor of Cato Unbound. He's on twitter as JasonKuznicki. His interests include political theory and history.

Related Post Roulette

98 Responses

  1. Avatar greginak says:

    Most peoples inability to use or understand stats is an easy, and good, target. I think it would be better to have a year of stats in HS over lots of other subjects.

    I don’t think your point about NPR and liberals holds at all and it comes as just trying to slam D’s. Its a common point for liberals to note NPR gets a tiny tiny fraction of the budget so cutting it is only a stunt.

    Signaling is one of those great points that gets trotted out way to often. If often comes off as way to simply dismiss what a person says without dealing with there arguments.Report

    • Avatar Jason Kuznicki in reply to greginak says:

      I don’t think your point about NPR and liberals holds at all and it comes as just trying to slam D’s. Its a common point for liberals to note NPR gets a tiny tiny fraction of the budget so cutting it is only a stunt.

      Among the well-informed, yes. But if the average liberal voter knew that NPR got .01% of the federal budget, then we should absolutely expect a large number of people to volunteer this information to pollsters. They don’t. Because they don’t know.Report

      • Avatar ThatPirateGuy in reply to Jason Kuznicki says:

        Jason I have two points

        First the general public gets simple questions wrong. 53% percent knew how long it takes the earth to go around the sun.

        Last, politics is war. I mean that literally. It is the best form of war as way fewer people die or get injured. Voting is merely making the size of a coalitions army known. We let the bigger army win without a fight because that is the best way we know to keep the bloldster down.Report

      • Avatar tom van dyke in reply to Jason Kuznicki says:

        Numerical literacy. The Bush “tax cuts for the rich” came/come to $60-100B, 4-7% of the current deficit. Yet they have garnered far more than 4-7% of the talk time about the deficit.

        In fact, my calculations make it that “tax cuts for the rich” have made it into exactly 106% of all Democrat speeches.Report

        • Avatar NoPublic in reply to tom van dyke says:

          The Bush Tax Cuts* actually come in at closer to 12-14%. Which makes someone’s point I’m sure.

          *The “for the rich” part of them may be less depending on how you define “rich”, but that’s a going forward number should the make up change and nothing to do with Bush.Report

    • Avatar James K in reply to greginak says:

      I definitely agree with this. Most people believe its easy to tell lies with statistics but that’s not really true, but using statistics to lie to people who don’t know about statistics is easy.

      Some of the research to come out of behavioural economics also suggests that statistically-capable people are less vulnerable to a range of cognitive biases as well.Report

  2. Avatar Pat Cahalan says:

    I was wrong about the minimum wage (I guessed 4%). All the rest of ’em I was within a reasonable margin of error. Do I get a cookie?Report

  3. Avatar Ryan says:

    In other words, American society is not composed of any significant percentage of Mentats.

    People are bad at ratios, not just percentages. People can get their head around a number, but ask them to compare that number to another number and the wheels fall off.Report

    • Avatar WardSmith in reply to Ryan says:

      93.7% of statistics are made up on the spot. Or was that 97.3%?Report

      • Avatar Trumwill in reply to WardSmith says:

        They say 3 percent of the people use 5 to 6 percent of their brain
        97 percent use 3 percent and the rest goes down the drain
        I’ll never know which one I am but I’ll bet you my last dime
        99 percent think we’re 3 percent 100 percent of the time
        64 percent of all the world’s statistics are made up right there on the spot
        82.4 percent of people believe ’em whether they’re accurate statistics or not
        I don’t know what you believe but I do know there’s no doubt
        I need another double shot of something 90 proof
        I got too much to think about
        -Todd SniderReport

    • Avatar DensityDuck in reply to Ryan says:

      I recall–can’t give a cite, I’m afraid–but I remember reading that human senses can distinguish “some versus none” far better than “some versus more”. That is, if you put a feather on my hand I can tell you; but if you take the feather away and then put two feathers on my hand, I probably can’t tell you that there are twice as many feathers now.Report

      • Avatar Mike Schilling in reply to DensityDuck says:

        One vs. two, I’d expect to do OK. Three vs. ten, I doubt I’d do better than “a lot more”. And a million vs. a billion? Forget about it.Report

        • My uncle once told me that there was an indian tribe that had a number system that included zero, one, two, and many. I don’t know if that’s true or not, but I’ve found it insightful, in its own way, to the way that people approach numbers.Report

          • Avatar Chris in reply to Trumwill says:

            He was probably talking of the Pirahã. It was once thought that they had word for one, two, and many (not zero). Now it’s thought that their quantity terms aren’t even that sophisticated. They essentially have terms for small or smaller quantity, large or larger quantity, and many.

            They also don’t distinguish singular and plural grammatically.Report

  4. Avatar Pat Cahalan says:

    I think a more interesting comparison would be how often the *voting* public gets this stuff wrong.Report

  5. Avatar Katherine says:

    I actually do know the approximate percentages (ie: very small) for foreign aid and public broadcasting, but that’s because I’m a wonk and make pie charts of the budget in my spare time. I do remember strongly believing at one point that the defense budget was ~50% of US spending (rather than the ~50% of discretionary spending that it is), and was fairly shocked to see how much of the US budget qualified as ‘mandatory’.

    I actually know a lot less about the Canadian budget because, since the provinces are technically responsible for health spending, it’s much harder to get any relevant info. out of looking at the federal budget. Most of it is made up of “transfers to persons” (pensions, old age security, EI), “transfers to provinces”, and “transfers to other groups” (payments to First Nations, agricultural subsidies, etc).

    The minimum wage number I did not know – I would have pegged it at around 5% – and, proving your argument, I immediately distrusted it. So, I checked the Bureau of Labor Statistics website.

    The website’s statistics allow me to calculate that in 2010, there were 124 million American workers. Of workers paid by the hour, 1.8 million (1.45%) earned the federal minimum wage of $7.25, and 2.5 million (2%) earned wages under the minimum wage (since there’s a variety of exemptions to it). However, the BLS report notes:

    “The estimates of the numbers of minimum and subminimum wage workers presented in the accompanying tables pertain to workers paid at hourly rates; salaried and other non-hourly workers are excluded. As such, the actual number of workers with earnings at or below the prevailing Federal minimum is undoubtedly understated. Research has shown that a relatively small number and share of salaried workers and others not paid by the hour have earnings that, when translated into hourly rates, are at or below the minimum wage. However, BLS does not routinely estimate hourly earnings for non-hourly workers because of data concerns that arise in producing these estimates.”

    So the BLS stats may underestimate things, but likely not by much, since salaried jobs tend to pay higher than paid-by-the-hour ones. Still, the number (3.45%) is higher than – in fact, more than double – the 1.7% cited by Jason in the Washington Examiner. Which leads me to the question – where did he get his statistics?Report

  6. Avatar BlaiseP says:

    We ought to parse away the rhetoric of politics from the mechanics of legislatures and governing bodies. Conflating the two is pretty much nonsense. Campaigns are torchlight parades and bunting and baby kissing and strawboater hats and tour buses and confetti and cameras and songs and slogans. Actual governance is the dullest of prose and accounting, waist deep in the swamps populated by the other political (and cannibalistic) alligators and the special interest groups who blow through the halls of government like the weather.

    Emotions can never be squared to the facts. It little matters what we say, what matters is what they hear. Does it matter what the actual fraction of people are LGBT or what percentage of the federal budget goes to NPR? Not in the Campaign Stage of politics: the politicians are appealing to fears and prejudices and hopey changey stuff. We seem to like it this way, herd animals that we are. Once they’re elected, we don’t pay attention to their legislative records except through the myopic lenses of self-interest anyway, so why should the politicians make rational choices?Report

    • Avatar James K in reply to BlaiseP says:

      What you say is true, and it suggests to me that the problem goes back to the voter. After all, the role of the voter in the democratic process is to hold elected representatives to account. Politicians only engage in the pomp and circumstance because it works. Ideally any politicians that tried to demagogue would be laughed out of politics.

      Now actually getting that to happen, there’s the tricky part.Report

  7. Avatar Katherine says:

    I would like to say, Jason, that even though your minimum wage stats are wrong, I agree with your larger point – and if I had come across minimum wage stats that were closer to what I expected, and provided by a source ideologically closer to me, I would probably not have taken the time to check them out. Which is something I should work on.Report

    • Avatar Jason Kuznicki in reply to Katherine says:

      If you’d like to say that the Bureau of Labor Statistics is wrong, please take it up with them.

      They’re probably in the pay of the Koch brothers, you know.Report

      • Avatar Katherine in reply to Jason Kuznicki says:

        There’s no need to be snarky – we’re both using the Bureau of Labor Statistics, and seem to be getting different results. I was using the 2010 stats:

        My work involved a few calculations. From the second page of the report:
        -72.9 million Americans were paid at hourly rates; these represented 58.8% of all workers. 72.9/0.588 –> 124 million workers total.
        – 1.8 million hourly-paid workers earned the federal minimum wage (1.45% of all workers)
        – 2.5 million hourly-paid workers (2% of all workers) earned below the federal minimum wage due to various exceptions to it
        – there are no statistics on how many salaried workers earned the minimum wage or below, so these are probably underestimates, but not by a lot as salaried workers tend to make more than paid-by-the-hour workers

        So, the most recent figures have it at 3.45%.Report

        • Avatar Jason Kuznicki in reply to Katherine says:

          You were the one to suggest ideological bias, not me.

          But either way, we’re talking about a difference between two figures with two years’ gap between them. I’m willing to believe the bottom end of the labor market has gotten a lot worse in that time. Meanwhile, the overall point stands. Most people don’t give anything near an accurate number when asked the question.Report

          • Avatar DensityDuck in reply to Jason Kuznicki says:

            “You were the one to suggest ideological bias, not me.”

            She’s pointing out her own ideological bias, not accusing you of being influenced by it.Report

            • Avatar Katherine in reply to DensityDuck says:

              Exactly. I wasn’t saying “I consider the statistic you gave suspect because you’re right-wing and therefore not to be trusted,” I was saying “I suspected the statistic because it was lower than I, a left-winger, expected, and it came from someone with political disparate views, so I investigated it. Although I was correct, it is a bad thing that I would probably not investigate a statistic that jived with my expectation and came from a left-wing source.”Report

          • Avatar Alan Scott in reply to Jason Kuznicki says:

            The number cited in Jason’s article is 1.7%. That’s not an issue of changes in the labor market–that’s an issue of difference in methodology.

            Jason’s figures appear to reflect the percentage of total workers who were paid hourly, and whose hourly pay was at or below the minimum wage. (applying that method to Katherine’s 2010 data would result in a figure of 2.0%)

            Katherine’s figure is the percentage of hourly workers who make minimum wage or below (using Jason’s 2008 data, that would be 3.0%).

            At question is what percentage of non-hourly paid workers make at or below the minimum wage. Per Katherine’s BLS quote, that’s a non-zero number, and thus the 1.7/2.0 number underestimates the total number of workers who are being paid less than minimum wage, but it’s unclear by how much.Report

            • Avatar Katherine in reply to Alan Scott says:

              Katherine’s figure is the percentage of hourly workers who make minimum wage or below (using Jason’s 2008 data, that would be 3.0%).

              No – that number is given in the report at 6% (on the third page). I did calculations to find the total number of workers (rather than the number of hourly workers) precisely to avoid that methodological error. 3.45% is the percent of total workers who make at or below minimum wage.Report

        • Avatar Kolohe in reply to Katherine says:

          “1.8 million hourly-paid workers earned the federal minimum wage (1.45% of all workers)
          – 2.5 million hourly-paid workers (2% of all workers) earned below the federal minimum wage due to various exceptions to it”

          Well, that’s one thing I would have got very wrong. I would have said, if asked, that not only the number of people at minimum wage are greater than those below min wage, but also that the former number would be a full order of magnitude than the latter.

          I knew the min wage exception was mostly carved out of waitstaff and farmworkers, but I had no idea it was that big (and exceeded those at the minimum)Report

          • Avatar Katherine in reply to Kolohe says:

            Farmworkers may not even be included in that statistic to a great degree – the stats only cover pay-per-hour, so people who are paid based on, for example, the pounds of fruit they bring in are not included.Report

  8. Avatar Katherine says:

    By the way – does anyone know if there’s a way to insert Excel figures into posts? Because this is about the 20th time (not a statistically rigorous figure!) I’ve wanted to post some of my charts.Report

  9. Avatar DensityDuck says:

    I’m reminded of the bit in Stephenson’s “Zodiac” about pH levels. It was something like…

    “They’re putting stuff into the soil that’s pH 13. That’s more than twice the legal limit.”
    Shit, man!”
    That’s one reason why I liked Gomez. He hadn’t yet gotten jaded about toxic dumping; he still got excited about stuff like that.
    Except that I was still lying to him, in a way. pH works by powers of ten; pH 13 isn’t just twice the legal limit, it’s ten thousand times the limit. But you can’t just drop a number like that on people. They’re think you’re making it up. It’s too far beyond their daily experience; they have no idea how badly chemical-emissions laws get violated on a daily basis.Report

  10. Avatar DensityDuck says:

    It’s been my experience that pointing out actual percentages doesn’t help much, though, because–since the argument is about perceptions rather than facts–the person on the other end just finds something else to get upset about.

    “Drug companies make too much money, man, if we taxed all their profits then all their drugs would be cheap!”
    “Actually they only make 5% profit, and only in a really good year, so taxing all their profits wouldn’t really do a lot.”
    “Well yeah but they spend too much on advertising, man, if we banned drug advertising then all their drugs would be cheap!”Report

  11. Avatar Mike Schilling says:

    By the way, Slow Tuesday Night is a terrific name for a blog. As are Land of the Great Horses, Nine Hundred Grandmothers, Primary Education of the Camiroi, …

    In fact, I can’t think of a Lafferty story that wouldn’t make a terrific name.Report

  12. Avatar dexter says:

    Mr. Duck, How much of big pharma’s gross goes to salaries, bonuses and golden parachutes? I am not trying to be snarky, just looking for clues.Report

    • Avatar Simon K in reply to dexter says:

      Most of it is salaries, since drug firms basically develop IP. Actually manufacturing drugs is not that expensive. The reason drugs are so expensive is that the discovery pipeline is enourmously long (20 years is not unusual) regulatory compliance is enormously complex and expensive (mostly for good reasons) and the process of winnowing out compounds is essentiall still one of trial and error, so every drug that actually makes is to market has to pay the costs of hundreds that failed in the process. You may have heard pharma executives talk about how their pipelines are empty – this is in a sense their own fault, but it means they’re really struggling to find the successful drugs to pau for all the unsuccessful ones right now.Report

      • Avatar BlaiseP in reply to Simon K says:

        If Pfizer is any guide to the matter, more is spent on advertising than research.

        Developing a drug isn’t as expensive as all that and I have never seen a drug go 20 years in trial stages, especially if it’s just a matter of isomerizing an existing drug going off patent (e.g. omeprazole / esomeprazole aka Zegerid / Prilosec / Nexium ) : once you’ve got the lab set up, getting it through the protocols and production ramp up is a fairly constant burn and is amortized fairly quickly. That’s the big deal these days, not new drugs, but recycling existing molecules and getting them back under patent.Report

        • Avatar DensityDuck in reply to BlaiseP says:

          “Developing a drug isn’t as expensive as all that”

          Maybe not on whatever planet you’re from.

          A friend of mine’s been working on “developing” a drug that already exists and is used for the purpose it’s in development for and the FDA-mandated process has already cost over forty million dollars with two go-backs for further study.Report

          • Avatar BlaiseP in reply to DensityDuck says:

            Are you saying it took 40 million dollars to get omeprazole converted to esomeprazole? Get real.Report

            • Avatar Jason Kuznicki in reply to BlaiseP says:

              He’s saying it takes about that much to convince the government that it’s safe.Report

              • Avatar BlaiseP in reply to Jason Kuznicki says:

                No it doesn’t. It costs less than 10 million, often as little as 1 million to get a drug to Stage II trials. To get it through Stage III costs about 4 million and much of that can be done through an NIH grant.Report

              • Avatar BlaiseP in reply to Jason Kuznicki says:

                At any rate, the point I’m making resolves to the tricksy legality of reformulating old drugs to get them back under patent. Esomeprazole is the mirror image of its ancestor drug, omeprazole. Technically this called a stereoisomer, more precisely its enantiomer, and it’s where lots of drug companies are focusing their efforts.

                A whole host of drugs have gone down this route to re-profitability. Eszopiclone is just one-half of its ancestor drug and the EMA turned down its application for new active drug status.Report

            • Avatar DensityDuck in reply to BlaiseP says:

              “Are you saying it took 40 million dollars to get omeprazole converted to esomeprazole? Get real.”

              No, but it’s taken that much to get misoprostol approved for labor induction. Which, as I said, it’s already being used for, despite a black-box warning not to do it (black-box being the FDA’s way of saying “THIS WILL FUCKING KILL YOU DEAD RIGHT AWAY”.)Report

              • Avatar BlaiseP in reply to DensityDuck says:

                Misoprostol is a nightmare drug. The only reason it’s still on the shelf is because NSAIDs are, and it’s being replaced with eszopiclone for the NSAID-concomitant gastric ulcer.

                So what’s Searle doing? Back when it was just Cyctotec, it was contraindicated for pregnant women. Now it’s being shilled as an abortifacient. Don’t kid yerself about this induction phenomenon, oxytocin is what you’d use if you want to keep the cervix intact. Expect a massive bleed after every Cyctotec induction.Report

              • Avatar DensityDuck in reply to BlaiseP says:

                Oh look, someone read Wikipedia.

                So far the studies show that when you actually formulate the drug properly–that is, you don’t just crush antacid tablets and jam them up the hoo-hah–it works great. No bleeding, and in fact the rate of C-sections is lower than other drugs.Report

              • Avatar BlaiseP in reply to DensityDuck says:

                No, Duckie. I did the Pharmacia/Pfizer integration in 2003. I do know about eszopiclone and it’s in the same market space as misoprostol. Your condescending is duly noted.Report

              • Avatar DensityDuck in reply to BlaiseP says:

                So, no response to the “proper formulation eliminates adverse events”, then?

                And you’ve got a nerve ragging on me for about condescenscion.Report

              • Avatar BlaiseP in reply to BlaiseP says:

                I tried to make one point: drug formulation has become a mug’s game of molecule manipulation. You’ve made another, touting the hugely expensive process of putting Cyctotec through its paces as an abortifacient.

                Cyctotec has already been through its clinical trials and has been on the market for a long time. It’s been surpassed by other gastric ulcer drugs, notably first by omeprazole and then by eszopiclone. Mainly it survives coupled with NSAIDs as with diclofenac in Arthrotec. Now it’s being re-presented as an abortifacient, a mendacious attempt to capitalize on a major downside to the original drug’s contraindications. Tell me WTF could possibly cost 40 million dollars about that process.

                I’ll tell you what’s going on, misoprostol has turned into a dandy black market abortifacient drug in the third world, especially in Brazil. So why not lobby the government to put it on as a scheduled drug for exactly that purpose? Adverse events indeed.Report

              • Avatar Jason Kuznicki in reply to BlaiseP says:

                DensityDuck —

                Your life will grow easier once you realize that there is nothing in the universe that BlaiseP hasn’t done. Did drug company mergers, worked in refugee camps on two different continents, ran a coffee outfit in Guatemala, took down a burglar in his kitchen, grew up in Africa, lived in France, studied Torah, speaks Japanese, expert in artificial intelligence, versed in aviation, spent time on a dairy farm.

                Just roll with it. The guy can write.Report

              • Avatar Burt Likko in reply to BlaiseP says:

                I heard that BlaiseP also knows how to kill a man seventeen different ways with a teaspoon, is an expert in Polynesian macrame techniques, and collects antique pottery from the pre-Roman Balkans when he’s not teaching inner-city schoolchildren how to create organic vegetable gardens in old soup cans. Dude’s amazing.Report

              • Avatar BlaiseP in reply to BlaiseP says:

                There’s a downside to everything, Jason. I’m alone in a hotel room with a cat and haven’t lived in a house for six years, making up for lost time, after putting three kids through school, putting together money for my retirement. I have a lot of true stories to tell, and everything in my life fits into the back of my truck, two computers and four terabyte drives. I miss my books most of all.

                Want to live my life? You’re fucking welcome to try. It’s killed better men than me. Don’t begrudge me my stories, they’re all that’s really left to me.Report

              • Avatar BlaiseP in reply to BlaiseP says:

                Seconds out, Burt. Bite my ankles, an it please thee. There’s more scar tissue than your feeble jaw can bite through.Report

              • Avatar Burt Likko in reply to BlaiseP says:

                I was just trying to have a little light-hearted fun. No offense intended, BlaiseP.Report

              • Avatar Jason Kuznicki in reply to BlaiseP says:

                I find your biography incredible. It’s a double-edged word, as you know. Beyond that, I haven’t got much to say. Outside of a very few fields, no one has had a life like yours.

                You’re well-read, thoughtful, and make reasonable points on any subject you address. Not like I can or should complain about any of that.Report

              • Avatar BlaiseP in reply to BlaiseP says:

                Yeah (sadly) you couldn’t make me up if you tried. Truth is stranger than fiction.

                Shameless blog-whoring: since I haven’t been set up here with a blog, I’m still writing back at my old haunts. Perhaps this will give you some insight into who I really am.Report

              • Avatar tom van dyke in reply to BlaiseP says:

                BlaiseP, I have a friend of the same vintage, demeanor, politics, and Zelig-like experience, in the same straits. If you weren’t you, I’d swear you were him.

                I continue to disagree with the content of his assertions, but took a chance on his admirable, restless intellect and bought him a beer when we ended up in the same town after literally years of sparring on the internet.

                I found him a goodfella, a man of bright cheer and little quarrelsomeness face-to-face. His internet interlocutors would be so surprised!

                I say this to you by way of equal respect and appreciation altho we have not met, for I think I get you. And to others—that this “BlaiseP” is probably a far more sociable companion in real life than in fora like these, where it’s the restlessness of his intellect and not his personal disposition that is at the fore.

                Cheers, mate. Find yrself in LA, dinner &c. is on me. But you’ll have to tell me your real name—which I assure I would not divulge, lest you’d have to kill me.Report

              • Avatar Chris in reply to BlaiseP says:

                I love BlaiseP’s Dos Equis commercials. 😉Report

              • Avatar DensityDuck in reply to BlaiseP says:

                “Tell me WTF could possibly cost 40 million dollars about that process.”

                That is an extremely good question which the FDA has yet to provide a reasonable answer to. That’s not stopping them demanding more patients, more studies, more time, and more money be spent. Apparently they’d rather the drug be used off-label.

                Abortion is not part of the study (or the intended approved usage) in any way at all, and it’s kind of interesting that you’ve gone there, seeing as how I said nothing about it. I guess that everything involving labor that isn’t Natural The Way God Intended is a Baby Murder Conspiracy to some people.Report

              • Avatar BlaiseP in reply to BlaiseP says:

                I’ll take your word on this 40 million business, Duck, for you have no reason to lie about it. FWIW I don’t have a viewpoint on abortion. The debate is properly confined to people who have a uterus.

                This much I’ll say about misoprostol, it’s a well-understood and long since superseded drug which has acquired a new lease on life in the Third World, where access to abortion is denied or unavailable. Nobody in Big Drugco wants to advertise it as an abortifacient, not after the Plan B fiasco, so it’s billed as an induction drug. Well, yes, so it is, and it works just as well inducing labor in the first trimester as it does late in the third.Report

              • Avatar Fred Chopin in reply to DensityDuck says:

                Forgive me Mr. Duck, er, Mr. Density Duck-I’ve been meaning to ask you this, are you perhaps related to Disco Duck? Wife, husband, cousin, partner, whatever.

                Your distinct avian verbiage is almost enough to be an identical twin for the Disc Duck–love Barry White’s duets with the Duck Man.

                There’s talent in them thar genes and don’t let nobody tell ya differently!

                Heard a chorus of ducks once sing the Hallelujah Chorus from Handel’s Messiah and it was a truly blessed event!

                Just curiousReport

              • Avatar Fred Chopin in reply to BlaiseP says:

                And Mr. BlaiseP, you’re a wonder to behold! The ominsience of your curious mind has you wandering the four corners of the earth, seeking mystical connections to the to our blessed Universe and to our untouchable Godhead.

                A very deep, thanks. And please, never hesitate to rip me (and my sometimes blarney) to shreds. As if I needed to tell you that.

                I’d love to send you Bach’s 48 from the WTC. I think Bach has unique ability to ALWAYS know note comes next–the whole entire architectual possibilities he sees in all it’s tonality. Not easy, especially when you have Fugues 4, 5, even SIX voices complimenting each other. And even sometimes throwing in an orchestra!

                The Art of the Fugue–you’ll never be the same after you hear what God has to say about this matter! Best–CReport

        • Avatar Simon K in reply to BlaiseP says:

          I’m not talking about trials. That takes around 5 years. I’m talking about the complete process from the initial drug discovery through to large scale manufacturing. The full process is in fact extremely expensive. Big pharma has been trying to avoid it, which is why they’re currently in some trouble, and spending a lot on trying to cheat the system one way or another.Report

    • Avatar Jason Kuznicki in reply to dexter says:

      Mr. Duck, How much of big pharma’s gross goes to salaries, bonuses and golden parachutes? I am not trying to be snarky, just looking for clues.

      Tagged as an example of the method. Well done.Report

      • Avatar dexter in reply to Jason Kuznicki says:

        Mr. Kuznicki, “Tagged as an example of the method. Well done.” Is that a compliment or a slam? I am not being snarky, just obtuse.Report

        • Avatar Jason Kuznicki in reply to dexter says:

          I am praising you. Straightforwardly and sincerely. I asked people to be suspicious of percentages; by obvious implication, proportions are under the scope as well.

          You asked about an assumption regarding proportions. I was pleased. Honestly.Report

  13. Avatar Mike Schilling says:

    And then there’s percentages vs. percentage points, which are misused all the time, e.g. if 3% of the population is gay and 12% black, it doesn’t mean that there are 9% more blacks than gays.Report

    • Avatar Simon K in reply to Mike Schilling says:

      Yup. And the inability to tell the difference between annualized rates and actual annual rates – in early 2009 NPR kept telling me the economy had contracted by 6% last quarter where in fact it had contracted by 1.5%, 6% being the annualized rate. The actual annual contraction was, of course, much less.Report

  14. Avatar David Cheatham says:

    This isn’t a problem with percentages, it’s a problem with the woefully uninformed, thanks to the media. But people seem to be entirely happy to be very badly informed. Every time I talk about the US budget, I tend to get totally nonsensical replies.

    For example, how much money would the government save if we eliminated social security? Just, tomorrow, removed it all.

    The correct answer is actually ‘none at all’, unless you’re operating in some strange universe where we would keep collecting social security taxes without paying it out. (In fact, it would end up increasing the debt as we stopped having the current surplus to borrow from.)

    It’s the same with Medicare. People aren’t going to keep paying insurance premiums unless it actually pays out. And it’s pretty clearly a ‘tax increase’ if you force them to keep purchasing imaginary insurance.

    I’m of the opinion that almost all discussion about the US budget is total uninformed nonsense.

    For example, the Republican’s latest nonsense about ’63 billion’…those cuts were in imaginary money in the first place, money that got allocated each year but not actually spent. If I give you ten dollars to go buy lunch, and you spend $4 at Wendys and give me back $6 each day…I can’t claim I saved $5 if I reduce your lunch budget to $5. But that’s what the Republican did. It’s all well and good to cut that, but won’t do a damn thing about the deficit.Report

  15. Avatar David Cheatham says:

    And as for the questions, the amount of ‘gay’ people isn’t a good question, as that’s under rather a lot of debate. While only 3% of American are _out_ as gay, the actual percentage of homosexuality is much harder to calculate. The standard estimate is 10%, although that varies both by gender, and by what you mean by ‘gay’. And it pretends no one is bi, which is stupid.

    I suspect that it’s probably somewhere in the middle, call it 7% of humans who would only be happy with a partner of the same gender, and some unknowable amount of who would be happy with either. Unknowable because a large amount of them will either never figure it (Unhappiness with straight relationships seems to be the big clue for people that they are gay. If people aren’t frustrated with their lack of attraction for the opposite gender, they’ll never even bother looking at the same one.), or will never step outside heterosexuality even if they do become away, as straight relationships are much easier in current society.

    In fact, I think logically that the number of bisexuals should outnumber homosexuals. This is because I assume that ‘attraction to men’ and ‘attraction to women’ are, in fact, _separate_ switches inside people, which would logically mean the number of, say, men, who have their ‘attraction to men’ switch randomly flipped on (aka, bi) should outnumber the men who have that switch on and the ‘attraction to women’ switch also randomly flipped off (aka, gay). If we assume that each switch has a 1/4 chance of being flipped from ‘default’, we get 18.75% bi, plus 6.25% gay.

    Along with 6.25% asexual who have both switches off, and the rest straight. Not that there’s any reason that both switches, or the same switches in different genders, should have the same chance of flipping.

    The point is, if they are independent of each other, we should be seeing a lot more of ‘half flipped switches’ people, aka, bisexuals. Which would argue against them being independent, except that, like I said, I suspect the amount of bisexuality in the population is _epicly_ underreported, and will continue to be so until same-sex relationships are utterly unremarkable.Report

    • Avatar Katherine in reply to David Cheatham says:

      While only 3% of American are _out_ as gay, the actual percentage of homosexuality is much harder to calculate. The standard estimate is 10%, although that varies both by gender, and by what you mean by ‘gay’.

      I’ve heard the 10% stat everywhere, but what can we base it on, if not on self-reporting?

      I feel the same way about claims that “75% of rapes [or any similarly large percentage] go unreported”, which tends to be cited a lot. If they’re unreported, where do we get the number from?Report

      • Avatar Simon K in reply to Katherine says:

        The 10% number comes from Alfred Kinsey “Sexual Behavior in the Human Male”. It applied specifically to American men, and was based on his interview data. That data contained a number of biases that have since been discussed at enormous length. The long and short is that it was probably an over-estimate.

        I think the “75% percent of rapes go unreported” (or whatever the number actually is) is similarly from interview or survey data and comparing that with actual reports to the police. I’m not sure exactly where the number comes from.Report

        • Avatar Robert Cheeks in reply to Simon K says:

          Weren’t Kinsey and his associates a bunch of wankers?Report

          • Avatar BlaiseP in reply to Robert Cheeks says:

            Kinsey did some dodgy research work, that much is for sure. He was dealing with some serious child molesters and didn’t turn them in. I dunno, maybe Kinsey felt the science was more important than exposing the crimes involved. We still use the asphyxiation data compiled by the Nazis for oxygen consumption in divers and pilots, astronauts and the like. Not a pretty picture.Report

          • Avatar Simon K in reply to Robert Cheeks says:

            Literally? I’m pretty sure he said as much.Report

        • Avatar David Cheatham in reply to Simon K says:

          I suspect 10% is wrong for ‘homosexual’ in the sense of ‘men not attracted to women at all’.

          However, I’m pretty certain that’s not what Kinsey’s said. His definition of ‘homosexual’ was ‘a man who is more attracted to other men than women’, aka, a lot of bisexuals were included, so 10% seems somewhat too low in my head. I haven’t ever read Kinsey, though, because it _is_ crap.

          Seriously, at this point we should probably accept that sexuality is a lot like ‘race’…there might be some biological distinction somewhere way deep down, and a lot of testing can reveal you’re 25% this, 50% this, etc. But at least 75% of the stuff we _think_ is objective about sexuality is just society making up groups and people self-identifying. (Which is where all the bisexuals have vanished to.)

          However, annoyingly, at some reason it became political important that sexuality be ‘genetic’. (By which misinformed people actually mean ‘hormonal’, as genetics doesn’t even really determine your gender…men have like one extra gene to switch on testosterone, and that’s it, the hormones do the rest.)

          Because if people can’t ‘choose’, that makes it like ‘racism’. So suggesting ‘A lot of sexuality is a social construct’ doesn’t go over well, even when that’s not an attempt to say ‘And thus should not be allowed’.

          In my universe, people shouldn’t be discriminated against because they belong to any group, or people think they do, be it homosexuals, blacks, women, or Baptists, and it really doesn’t matter how much of membership in that group is some objective measure and how much of that group is just a term we invented to poorly categorizes things that exist in spectrum. (And, as any religious scholar will tell you, religious groups exist in a spectrum. And once in a while, so does gender.)Report

          • Avatar Simon K in reply to David Cheatham says:

            If you’re going just from the reported factoids when you say Kinsey’s report is crap, you should read it. There’s a lot more to it and the two bits most people know about, and it was a significant step forward to publish anything in a era when talking about breastfeeding was considered obscene.Report

            • Avatar David Cheatham in reply to Simon K says:

              Yes, but there’s been a lot of better studies, done in better, freer eras. I wasn’t trying to criticize Kinsey, I suspect he did the best he could.

              But if I ever read it, it would be as ‘the groundbreaking paper that actually allowed people to talk about these things’, then, you know, ‘actual true facts’. (And as I’m not studying ‘the history of sociology’, I am unlikely to do that.)Report

    • I think logically that the number of bisexuals should outnumber homosexuals.

      As I understand your gender preference taxonomy, if I’m a male, and my “attraction to men” switch is in the “on” position, then I’m either gay or bi — depending on whether my “attraction to women” switch is “on” or “off.” If both are off, I’m asexual, if men are off and women are on, then I’m straight, if men are on and women are off, I’m gay. If both are on, I’m bisexual. I’m not sure I buy into the theory although I see the intellectual attraction to it as it explains durable bisexuality.

      But in reality, these are dimmer switches, not toggles.

      Most of the time, a dimmer switch is pushed to nearly all the way on or nearly all the way off, but sometimes it falls somewhere in the middle (for reasons which are unimportant here). If we quantify it, and the switch intuitively goes from 0 to 10, your “attraction to men” rating might be a 9 and your “attraction to women” rating might be a 3 — with the result that you would want about 75% of your sexual partners to be men and 25% to be women (assuming you had your way with your romantic adventures, of course; real life is rarely so neat and clean as mathematically-expressed taxonomies). Unless you’ve got someone who is at 10 out of 10 on one axis and 0 out of 10 on the other, everyone is going to be bisexual, whether or not they self-identify that way.

      In such a taxonomy, is the “9-3 guy” gay, straight, or bi? Obviously, we can’t just ask him, because his answer would necessarily be an oversimplification of his reality into one of the four socially conventional categories. He might self-identify as “mostly straight,” for whatever reason (social pressure, personal identity, religious background, who knows) even though the majority of his attraction and likely of his behavior is toward and with other men.

      If pressed for specificity and honesty, he might say “Mostly I like men better than women, but I do like women and I’ve had a few girlfriends, mainly boyfriends, but some girlfriends too.” We can call that “mostly homosexual” but now it’s a numbers game — if 75% same-sex partners is “bisexual” when do we get into “straight with some history of experimenting”? Is that “9-1”? Do you have to be “10-0” to be “straight” or “gay”?Report

      • Avatar Jason Kuznicki in reply to Burt Likko says:

        Not only are these dimmer switches rather than toggles, but for each dimension there is more than one switch.

        “Orientation” contains many things.

        It’s the gender(s) I’ve had sex with.

        It’s the gender(s) I’d like to have sex with.

        It’s the gender(s) I think about having sex with, even if I’d rather not have sex with them, and even if, after having sex with them, I don’t like it.

        It’s the gender(s) I hope to be oriented toward, even if I’m not currently oriented toward them.

        It’s how I present myself to the world, even if it contradicts some or all of the above.

        Seriously, if you’re going to do a 1-10 scale, or even Kinsey’s 0-6, where do you place someone like Ted Haggard? He’s clearly not entirely straight on all dimensions — though that’s what he aspires to be, and it could very well be what he currently is. What’s in his mind? I’d rather not know, but it might not be what he presents to the world.

        Attempts to press this subject into a single dimension are dangerous. We miss a lot of nuance that way.Report

        • Avatar Freddy "Chopsticks" Chopin in reply to Jason Kuznicki says:

          Jason, I ask this sincerely and without bias–if a cure was found that could change one’s sexual preference–the change would have to be homosexual to heterosexual–do you believe this would this have any appeal to the homosexual community? Forget the science and all the genetic hurdles necessary to accomplish such a change, I’m just very curious whether such an idea would any interest or traction within that segment of our society.

          If someone had the choice of being either gay or straight would there be a significant amount of gays crossing over for the simple reason that it’s not worth the deep emotional pain and abuse that accompanies being gay.

          I think the pain of self-acceptance is probably the deepest, most difficult pain of all to endure.

          I sincerely hope this is not taken as transparent blatant bigotry. There is no animus or offense intended in my comments.Report

          • Avatar Jason Kuznicki in reply to Freddy "Chopsticks" Chopin says:

            Jason, I ask this sincerely and without bias–if a cure was found that could change one’s sexual preference–the change would have to be homosexual to heterosexual–do you believe this would this have any appeal to the homosexual community?

            Some might want it. I would avoid it like poison. At this point, your “cure” would ruin my otherwise very happy life. Thanks but no thanks. The last thing I’d want is to be heterosexual and no longer romantically attracted to my partner. It would crush him, crush is family, hurt our daughter’s life, and really just ruin everything.

            If someone had the choice of being either gay or straight would there be a significant amount of gays crossing over for the simple reason that it’s not worth the deep emotional pain and abuse that accompanies being gay.

            I don’t think there’s really that much “deep emotional pain” to being gay. I find even less abuse. I’m pretty happy being gay.

            I think the pain of self-acceptance is probably the deepest, most difficult pain of all to endure.

            I don’t agree. I had a pretty unhappy adolescence, but I gather that most straight people usually do too. It gets better all around, gay and straight alike.

            I sincerely hope this is not taken as transparent blatant bigotry. There is no animus or offense intended in my comments.

            No animus. I just think you’re operating on some mistake factual assumptions.Report

            • Avatar Freddy "Chopsticks" Chopin in reply to Jason Kuznicki says:

              Thanks so much for the wonderful reply, Jason–a true mensch you are sir!

              I know you could have handed me my head on a silver platter had you wanted to. In some respects my question about gender choice is similar to asking a blind person if they wanted to have their vision restored or deaf person whether they want their hearing (back).

              It can’t be generally assumed that the answer would always be a resounding “yes” because there are far too many factors, deep factors that are involved.

              Again, many many thanks. Hey, have you seen the new movie, “Mate”? Looks quite interesting. It would be a great for you to see–could you be, perhaps, talked into writing a review for the League? Bribed, perhaps?!

              Thanks again.Report

      • Avatar Mike Schilling in reply to Burt Likko says:

        Is Ted Haggard complicated? He’s gay, and he thinks gayness is pure evil, so he fights it, but when he gives in he make it as evil as possible.Report

      • Avatar David Cheatham in reply to Burt Likko says:

        Well, yeah, the ‘switches’ were an over-simplification, obviously. I was just making the point that for some reason people seem to think there’s a ‘men’ or ‘women’ switch, whereas, in reality, it is _at least_ two switches. (Along with other, much less toggled switches, like attraction to animals or whatever.)

        And if it are two switches that default to one position, and get flipped randomly, then almost _automatically_ there should be more people that have one switch flipped (and hence are bisexual or asexual) than people who have both major switches flipped and are strictly homosexual. Unless the switch flipping is from the same source.

        As for the fact it’s not a ‘switch’, yeah, that poses a strange statistical problem…there’s really no way to measure it, even in some magical world where everyone is utterly uninhibited and completely honest.

        And it’s added to by the fact that almost all ‘coming out’ stories seem to have, as a constant, being unsatisfied in their ‘standard relationships’…which raises the question as to how _anyone_ realizes they’re bisexual, because they _aren’t_ unhappy. The only possibility is having someone of the same gender happen to ask them out, and them not automatically rejecting it.

        And if they do manage to realize they’re bi, why even _bother_ with dating the same gender…it’s harder to find people, the social protocols are still somewhat weird, and society is not entirely happy with it.

        It also raises the question of what we’re _trying_ to measure, and why? Are we suddenly running a dating service? Do we need to figure out how many marriage forms to print with ‘bride and groom’ spaces, vs ‘bride and bride’ and ‘groom and groom’? (Obviously that has a simpler solution, but you know what I mean.) What, exactly, is the point of all this?

        At some point, this is sorta like measuring ‘race’….we’re not measuring any actual thing, but what people identify as. (Which is partially based on what others identify them as.) With race, sure, we could do actual genetic sampling, or melanin levels, and with sexuality we could probably do the same, or test hormones or run MRIs while showing pictures…but, why do we really care?

        Anyway, the point is that this is a _really_ bad question to see ‘How well do people do at guessing percentages?’, because the ‘real’ _percentage_ is just a guess anyway.Report

  16. Avatar Chris says:

    I don’t disagree with your point about the general… suboptimality of using percentages. Our brains just aren’t very good at dealing with percentages in pretty much any context, at least not on a conscious level (our neurons are doing calculations of proportions all the time, and quite well – we could learn a thing or two from them). It’s not by accident that much of the early heuristics and biases research by Tversky and Kahneman involved ratios and percentages (though mostly as probabilities): our interpretations of proportions is largely determined by our evaluations, instead of the other way around.

    On the other hand, I’m not sure what the alternative is. Absolute numbers? That’s even worse! While .01% of the federal budget sounds really small, that’s still in the millions of dollars, and millions of dollars sounds like a lot to most people regardless of the context. The problem is that in order to think rationally about almost every political issue, numbers are important, and in the vast majority of cases, absolute numbers just won’t tell you enough to make any sort of informed decision. Whether it’s money or time or demographics, everything is relative.

    Also, like someone else (I’m too lazy to look back and find out who), I’m not sure this is really a case of signaling. That’s not to say that percentages are never used for signaling, but I think the way we think about percentages in politics, or elsewhere, really reflects our representation of the information. It’s just that are representations don’t map onto reality very well.Report

    • Avatar DensityDuck in reply to Chris says:

      I think that people can understand percentages and proportions just fine.

      I also think–as does Jason–that percentages are far more often used as signaling than as anything of numerical meaning.

      The issue is that our society has been so trained to look at numbers and figures and decimal points as indicators of Scientific Truth that we think we need to put numbers on anything to make it real; that, and people dishonestly put numbers on their statements to falsely give them an apparent factual backing. After all, figures don’t lie, right?Report

      • Avatar Chris in reply to DensityDuck says:

        Well, there’s a fair amount of data showing that people have trouble with percentages, ratios, and probabilities. This goes both for thinking about them from a mathematical standpoint and from an evaluative one (that’s the heuristics and biases stuff).
        It’s not really a matter of “I think that…”

        From my own personal experience teaching statistics to undergrads, I can say that it is proportions and probabilities that give them the most trouble.

        On signalling, as I said before, it’s not that signalling isn’t an issue in the use of numbers, but it’s much more than that, since these numberse are likely to “signal” how people actually represent the information.Report