Tuesday, June 02, 2009
Click the tab below the body of this post to read previous entries in the series about how previous generations were more depraved. One way to look at how civilized we are is to see how we behave in situations where our conduct can mean the difference between life and death for those around us -- for example, when we drive our car. Traffic deaths, of course, reflect properties of the car as much as the people involved, but teasing the two apart turns out to be pretty simple in this case.
In my brief review of Daniel Gardner's book The Science of Fear, I gave a few examples of how media coverage of some threat was outta-whack with the underlying risk, namely homicide and rape. Gardner spends a few pages talking about the epidemic of "road rage" that was allegedly sweeping across the country not too long ago, so why don't we have a look at what the data really say about when road rage may have been greater than usual.
First, here's a quick view of media coverage of "road rage," which begins in 1996:
What about actual traffic deaths, though? The data come from the National Safety Council, as recorded across several versions of the Statistical Abstract of the United States -- which, btw, is much cooler than the General Social Survey or the World Values Survey if you want to waste some time crunching numbers. I tracked the data back as far as they exist in the Stat Ab, and they include four ways of measuring traffic death rates. Here are the graphs:
The first is the most instructive -- it measures the number of deaths compared to the number of vehicles on the road and how long they're on the road. The other three measure deaths compared to some population size -- of vehicles, of drivers, or of the whole country -- but even if population size is constant, we expect more deaths if people drive a lot more. The distinction isn't so crucial here since the graphs look the same, but I'll refer to the first one because what it measures is more informative.
From 1950 to 2008, there's a simple exponential decline in traffic deaths (r^2 = 0.97). As roads are made safer, as all sorts of car parts are made to boost safety, and as people become more familiar with traffic, we expect deaths to decline, and that's just what we see. I looked at earlier versions of the Stat Ab, and there is a similar exponential decline in railroad-related deaths from 1920 to 1959 -- again, probably due to improving the technology of railroads, trains, and so on.
However, aside from the steady decline that we expect from safer machinery, there is a clear bulge away from the trend during the years 1961 - 1973. Although I haven't researched it, it seems impossible for roads to have went to shit during that time but not during the other times, or that cars made then were even less safe than the ones made before or after. The obvious answer is that people were just more reckless and/or hostile toward their fellow man in that period.
There's no other huge departure from the trend, so if any time period has been characterized by "road rage," it was The Sixties (which lasted until 1973 or '74). Consider the age group whose brains are most hijacked by hormones, and maybe by drugs and alcohol too -- say, 15 to 21 year-olds. The oldest members of this group who were driving during the road rage peak were born in 1940, while the youngest ones were born in 1958. Hmmm, born from 1940 to 1958 -- Baby Boomers. (Those born after 1957 - 8, and before 1964, are not cultural Boomers.) And this doesn't seem to be an effect of lots more teenagers on the road than at other times -- there have been echo booms afterward and yet no big swings away from the trend.
When the media and everyone hooked in to the media began talking about road rage 10 to 15 years ago, there was nothing new in the traffic death story -- indeed, the rate was continuing its decades-long decline. If you just want to know what is going on right now, the media may not be so bad at giving you that info. But this serves as yet another lesson to not believe anything they say, or imply, about trends unless there is a clear graph backing them up (or, in a pinch, a handful of data-points sprinkled throughout the prose).
I located, collected, analyzed, and wrote up all of the relevant data -- stretching back nearly 60 years -- in less than one day, and only using the internet and Excel. This shows us again that journalists are either too clueless, too lazy, or too stupid to figure anything out.
Thursday, January 08, 2009
After declining pretty steadily from 1991 to 2005, in 2006 teen birth rates showed a slight uptick. Rather than swallow what the mass media and doomsaying blogosphere infers, read the report for yourself -- what you want to know is contained in the first 5 to 10 pages. Since most people worry about the long-term trend, and where things are going, I've taken data from the report's tables and put them into easy to understand time-series graphs, broken down by race and ethnicity. I'll then address a few of the larger issues.
All birth rates are live births per 1000 women in a given group. I'll only look at births to 15 - 17 year-olds because mothers younger than that are even rarer, and people freak out less about mothers at or above the age of majority. The 18 - 19 graphs look similar, and you can create them yourselves using the NCHS' report and Excel. Update: see the end of the post for the one 18 - 19 year-old graph that is different, which shows birth rates among Hispanic 18 - 19 year-olds increasing since 2000. [End update]
Most of the recent increase is due to 18 - 19 year-old births, so that's another reason not to care about an increase in "teen pregnancy" -- 18 and 19 year-olds are adults. Moreover, there is an increase across all age groups, especially 20 - 24. So, there's nothing special about teens of any age -- the 15 - 17 year-olds increased a bit, while the 18 and 19 year-olds appear to really be part of a larger group of 18 to 24 year-olds. (Nature doesn't adhere to our numbering system, where there's a bright line between 19 and 20.) Births are just up overall, and the closer we get to the female fecundity peak in the early 20s, the stronger the signal is.
First, the NCHS has data going back to 1970, although it is not broken down by race. Still, here is that graph:
There is a downward trend throughout, with a steady oscillation around that trend. So, the rate will probably continue to decline into the following decades, and we shouldn't be fooled by a temporary increase. For the near future, it looks like the rate will remain pretty flat for about 5 years, then start to increase again, with a decrease again after that, all on a downward trend.
Next, the data with race broken down begins in 1980 and includes White, Black, Asian / Pacific Islander, and American Indian / Alaskan Native. The graph with Hispanic ethnicity is further down. Here are the birth rates for Blacks and AIANs, and for Whites and Asians:
Since the graphs look so similar to the all-race graph, we can assume that from 1970 - '79, the birth rate declined across all races. Again, we see a downward trend with a steady oscillation around it. In no case is the 2006 uptick dramatic, and it looks just like it does in the all-races graph. So again, the rates will probably remain mostly flat for the next 5 years, increase, then decrease, following the overall downward trend.
Finally, the data on Hispanic and non-Hispanic groups starts only in 1990. Here are the trends:
The patterns are the same as in the other graphs, so it wouldn't be too risky to assume that, if we had the actual data, they'd look like they do in the first graph back to 1970. There are clear race and ethnic differences -- the Black line is always above the White line -- but the downward trend and presence of oscillations does not have to do with race or ethnicity. Whatever causes them is at a societal level. The only interesting difference in these Hispanic graphs is that non-Hispanic Blacks and Hispanics have switched in the rank order: we now have more of a Hispanic problem than a Black problem.
That should give people a clearer idea of what is going on. It's silly to forecast based on two data-points -- last year and this year -- when there is plenty of other data to help out. The reason people do this is because it allows them to indulge their desire to imagine the end of the world. The slightest aberration or reversal of a reassuring trend is greeted with exhilaration by the declinist junkies -- "So there's hope that the world is doomed after all!"
Furthermore, these data are all on birth rates, not pregnancy rates. We won't know what the teen pregnancy rate was until we have abortion data, and the NCHS says that won't be for awhile. (The most recent data are from 2004.) For all we know, pregnancy rates are the same or lower than before, but more of those who do get pregnant may opt to keep the kid.
I suspect something like that is true based on adolescent sex behavior data that I've already written about. All measures of sluttiness among teenage girls -- having sex at all, having 4+ partners, not using a condom, etc. -- have declined from 1991 to 2007. Indeed, there is no change in any of the measures from 2005 to 2007. See here for the data from the YRBS. This suggets that the 2006 uptick in birth rates was not due to greater rates of bad behavior among teenagers, but to a greater aversion to abortion among today's young people.
A statement by the National Campaign to Prevent Teen and Unplanned Pregnancy hopes for the worst about these YRBS data:
For teens in school, YRBS data for 2005 and 2007 reflected small increases in sexual activity and decreases in contraceptive use that were not statistically significant, but the changes are likely large enough to account for the 3% increase in the teen birth rate (and likely a similar increase in the teen pregnancy rate) between 2005 and 2006.
The part that I've boldfaced is mostly a lie. If you look at the YRBS data in the link above, there is an apparent but non-significant increase of 1.0% in the percent who have had sex, from 2005 to 2007. However, there was also an apparent increase of 1.1% from 2001 to 2003, and of 1.5% from 1997 to 1999 -- greater than the most recent apparent increase, yet which resulted in no uptick in teen births. As for percent who are currently sexually active, there is an apparent but non-significant increase of 1.1% from 2005 to 2007 -- but there were similar increases from 2001 to 2003 (up 0.9%) and 1997 to 1999 (up 1.5%). The same is true for percent who use birth control pills. Only the percent who used a condom the last time they had sex shows an apparent and non-significant decrease that isn't matched by similar apparent decreases in previous years. Again, though, the change is not significant.
Here's what the CEO of the National Campaign told USA Today:
Sarah Brown, CEO of the National Campaign to Prevent Teen and Unplanned Pregnancy, says she is less inclined to believe abortion is driving higher teen birth rates and suggests that increases in high-profile unmarried births in Hollywood, movies and even politics is a significant factor for impressionable teens.
You have to admire her candor and pity her desperation: she admits that the purported causes came years after the effect. So, rather than Hollywood influencing a social trend, the already existing social trend began to reach even into Hollywood. After that happens, there may be positive feedback, but let's be clear about who started it. In general, Hollywood doesn't want to influence any social trend -- they want to figure out what the existing trends are and pander to them to get rich. They're greedy capitalists, not mad scientists.
When I am elected dictator, my first campaign to improve the people's mental health will be to censor anyone who, in a mainstream forum, argues that 1) Hollywood is responsible for degrading our morals, 2) poverty causes crime, 3) "it's the parents' fault," or 4) these kids these days don't know what good music is. Ah, to read the newspaper without feeling the urge to strangle 5 of the 6 people quoted -- that oughtta make everyone feel a bit more sane.
Note: here's the graph for non-Hispanic Black and Hispanic 18 - 19 year-olds:
So as with the 15 - 17 year-old case, Hispanics have overtaken Blacks, and also that their rate has been increasing since 2000. This is the only worrisome data -- and yet another reason to slam our borders shut to anyone who didn't graduate college.
Monday, June 30, 2008
Remember that better time when college coeds frolicked on the quad lawn, safe from the eyes of older males, who were drawn instead to the allure of a mature woman? Indeed, doesn't it seem like nowadays, in our Girls Gone Wild culture, we shove females into the sexual spotlight at ever younger ages? That's what you'd conclude from the 50,000 alarmist results that a Google search for "+sexualizing +young" returns, in particular the recent panic over 15 year-old Miley Cyrus posing semi-topless for Vanity Fair. The cropped picture to the left is of Elizabeth Ann Roberts, who was 16 when she was photographed nude as Playboy Playmate of the Month -- of January 1958.
On an intuitive level, though, we know that the culture must be more hostile than before to sexualizing young females -- there would be no hysteria if it were acceptable. Plus, suburban housewives and city-dwelling cougars have never hogged so much of our attention. Still, let's turn to three datasets that show the trend is, if anything, toward sexualizing increasingly older females in popular culture. We will look at data across the decades on beauty pageant winners, girls featured in nude magazines, and hardcore porn actresses.
First, take the winners of the Miss America beauty pageant, a competition determined mostly by how closely the contestant fits the ideal look of the time. A writer for the website Seduction Labs has already done an extensive analysis, so I took the age data from his work. Here is how Miss America's age has changed over the decades:
It sure looks like Miss America is getting older -- the ones from before 1940 are quite young -- and this is true: Kendall's tau for the correlation between year and age is +0.50 (p = 3 x 10^(-10), two-tailed). Admittedly, estimating the youth-obsession of each year with only one data-point -- the winner from that year -- is less desirable than averaging all contestants' ages for that year, but the data are hard enough to come by that this is the best we can do.
Next, consider the Playboy Playmates of the Month, averaged for a given year. While the 1950s had fewer data, each year still had at least 7 data-points. Using 12 data-points to estimate each year should make us more confident in the results, shown here:
Again, the average Playboy Playmate is getting older: Kendall's tau for the correlation between year and age is +0.44 (p = 3 x 10^(-6), two-tailed). The trend is clearly not linear, though, since there was a decrease in age at least from the mid-1950s, when the data begin, throughout the 1960s.
In response to a criticism brought up in the comments to the post showing that the popularity of blonds is recent, I've also calculated Kendall's tau based on the raw month-by-month data-points, rather than yearly averages: it is +0.18 (p = 1 x 10^(-10), two-tailed). As I mentioned to the commenter, I think it's more instructive to look at the year's average since the Playboy people likely have a target girl in mind for the year's subscription, based on the perceived demand. That is, the Playmates within a given year are comparable to the Miss America contestants for a given year -- they are chosen to fill out a year's run, and Miss April could just as well have been Miss December. Still, even by this perhaps overly stringent standard, the trend is positive and significant.
Finally, we look at actresses in hardcore porn movies. Collecting a representative sample of active females in a given year would be incredibly arduous, so instead I took famous actresses and determined how old they were when they made their first movie, and entered this as a data-point for the year in which they started making movies.
The lists I used are the AVN Hall of Fame, the XRCO Hall of Fame (which barely added anyone else), and a list of female porn stars by decade drawn up by the porn geeks at Wikipedia. I required each year to have at least 5 data-points; if there were too few, I merged that year's data with an adjacent year (whichever had fewer data-points than the other choice), so that the data-sparse year is excluded and the beefed-up year is included. This mostly affects the 1970s and early 1980s. Here are the ages of first-time porn stars by year of their first movie:
There is no increase or decrease over time: Kendall's tau for the correlation between year and age is nowhere near significance. There are several apparent upward and downward trends, though. This might be the only example of the 1980s and early 1990s showing greater progress by the declinists' standards. I recently analyzed a large, representative sample of porn stars and found that their average age is 23, for what it's worth. Again, that's what we really want to see: the age of the typical actress for a given year.
Maybe girls enter at earlier ages in recent times but don't reach their peak in popularity until they are in their early 20s. Another drawback of looking at age at first movie is that it ignores the recent popularity of "MILF" actresses -- maybe it's just that the variance in age is increasing. Admittedly, these pornstar data are not ideal.
Finally, we examine the popularity of beauty pageants specifically for teenage contestants. While I don't have datasets to analyze, such as the annual TV ratings, there is enough information on them to get a rough picture. First, there is Miss Teen USA, the adolescent version of Miss Universe. It was created in 1983, reached its peak for ratings in 1988, and has declined in popularity afterward, to the point where it may not even be televised anymore. And second, there is Miss Teenage America, which was created in 1962 and was last televised in 1977. Judging by its corporate sponsorship and celebrity hosts, it must have been somewhat popular. There are other beauty pageants for teenagers, but they are not even televised, and so do not count as evidence of an obsession with youth. Rather, we see a shift away from throwing young girls into the purely sexual spotlight.
Since there are no huge long-term swings up and down in these data, as opposed to the cases of sluttiness and violence, all generations can say that they've improved over previous generations, or at least done no worse. If any generation is to be accused of sexualizing younger girls in popular culture, though, it is surely the older ones. It is true that the current culture does not value women over 30, but that has never been the case -- just the opposite.
As with sluttiness, part of the declinists' misperception may be due to fashion trends, such as even prepubescent girls wearing adult-inspired clothing. That's hardly evidence of their being sexualized, though -- no guy is actually looking at them as a sex object, and dressing like an adult doesn't make you behave like one sexually. While it may be a bizarre fashion trend -- though more bizarre than when pre-pubescents started wearing two-piece bathing suits? -- it doesn't reflect a sexualization of the young.
What's causing this trend toward older sex symbols? Oh, I don't know, but I'm sure we'll get a bunch of half-baked ideas in the comments, so I'll get the goofball ball rolling. Women are having their first kid later, if at all, so there's a wider age range of females who haven't ruined their figure by giving birth. Still, according to the analysis of Miss America winners at Seduction Labs, there are other trends: starting around 1960, winners became taller, less buxom, and less hourglass in shape, in addition to older. In short, the feminine ideal in popular culture has been worn down by the march of the masculine minxes. It's a mistake to blame this on the women's movement of the 1970s, though, since most of these trends began in the early-to-mid-1960s. Radical feminists were just jumping on the bandwagon and trying to steal credit for it.
Though it's harder to measure, the manliness of these sex symbols' faces has surely increased -- go back and look at some of the Playboy Playmates from the late 1950s through the late 1960s. They look like girls, not butch transvestites (NSFW, obviously). I see this as a form of cultural decline, of course, but the declinists who decry our obsession with youth could not be more wrong.
Monday, June 23, 2008
Previously we found that your generation was sluttier, so we turn now to another great threat to civilization -- violence (between individuals). As before, our concern is with whether violent crime rates are increasing or decreasing, and not so much with the absolute level: it is easier to screw up civilization than it is to improve on it, so a decline can quickly snowball, while it may take much longer to restore things to their previous levels.
There are very good and very clear data on violent crime, so this post will be much more direct than the one on sluttiness. Let's begin with homicide. The Bureau of Justice Statistics, part of the US Department of Justice, has taken homicide data from the CDC's National Center for Health Statistics and put it into a straightforward graph. I see five trends in the graph: an increase from 1900 to the mid-1930s, a decrease from the mid-1930s to about 1960, an increase from 1960 to the late 1970s, a fairly steady high level (with oscillations) throughout the 1980s, and a decrease from 1992 to the present.
To be generous to older generations, let's say that much of this homicide is committed by 15 year-olds. That means that the cohort born in 1945 is responsible for the increase that began in 1960. I figure you have to be about 73 years old in order to decry how violent the younger generations have been -- certainly the Boomers and Gen X-ers cannot complain, while Generation Y should be thankful they've lived through such peaceful times.
The homicide data also caution against viewing the past with rosy spectacles -- there was nothing peaceful at all about the first third of the 20th century. Declinists who long for better times in the past seem to latch onto a fleeting period of rest and prosperity. That's fine, as far as worshipping one period over another goes. However, we should not think that we can easily maintain that level, whether through individual choice or institutional incentives, as oscillations and limit cycles appear to be the rule rather than the exception. We should aim instead to have a somewhat low level of Bad Things, with low-amplitude fluctuations, and not let the mere existence of waxing and waning cause us hysteria.
What about the intersection of sex and violence -- how have forcible rape rates changed over time? Again we turn to BJS data, although they do not go back nearly as far as homicide data, the earliest year being 1960. After retrieving data from this page, looking at the entire United States, forcible rape rate, from 1960 to 2006, I put them into a simple graph:
There are only two trends here: an increase from 1963 to 1992, and a decrease afterward. In fact, the two trends look pretty linear on first glance. The slope of the increasing trend is about +1.11, and the slope of the decreasing trend is about -0.85, confirming the hunch that the decline of civilization snowballs more quickly than its restoration proceeds. As with homicide, Boomers and Gen X-ers cannot complain about rape epidemics in recent generations. This is particularly true for the Boomers and Gen X-ers who manufactured and continue to prop up the myth of the campus rape crisis.
The BJS also has an index of "violent crime" that includes murder and non-negligent manslaughter, forcible rape, robbery, and aggravated assault. I used the same search function as for the rape rate graph, and the trends for this general violent crime rate look the same as the rape rate trends, so I won't include the graph. In brief, there's an increase from 1962 to 1991, and a decrease afterward.
As in the case of sluttiness, using popular culture as a means of taking civilization's pulse is highly unreliable. Before, we saw that slutty behavior has been decreasing even as perceived slutty appearances have been increasing. Here, we see that violent crime has been decreasing even as video games, movies, and TV shows have become increasingly violent. To pick just one example, gangsta rap was invisible during the 1980s and only became popular when Dr. Dre's album The Chronic came out in 1992, drawing ever larger audiences throughout the 1990s -- at the very time when violent crime was falling.
I don't believe that trends in real behavior and in popular culture are causally related in an inverse way either -- just that they are independent of each other. Cycles of fashion in the cultural realm are self-contained, and oscillations and limit cycles in real behavior are also self-contained, at least to a first approximation. I've read posts at Cognitive Daily that exposure to violent video games (and perhaps TV shows?) desensitizes people to violence within controlled, experimental laboratory settings, and that is an interesting finding. However, in examining the world outside of the lab, violent media cannot hope to account for even a trivial share of the variance across time in violent behavior.
Tuesday, June 17, 2008
I am sick of hearing Baby Boomers and Gen X-ers complain about a perceived cultural decline among the younger generations. For a variety of measures, things started to go bad already by the 1950s, became obscene during the 1960s and '70s, and plateaued some time during the 1980s. Since roughly 1990, however, things have gotten steadily better. This series will catalog such a trend for measures typically given in support of the declinist hypothesis: we begin with sexual behavior, and will eventually cover violent crime, divorce, narcissism, the arts, and whatever other examples I come across or that readers suggest in the comments. The hope is that the series will prevent the real-world picture from disappearing down the Memory Hole, as every generation thinks that patterns among its usurpers spell doom, regardless of what the data show.
Importantly, I am more interested in the slope or derivative of an indicator at some point in time, and less so in the value of the indicator at that point. The reason is simple: those who claim that our culture is declining, decaying, rotting, dying, and devolving are making an argument about whether some indicator is increasing or decreasing over time. What the declinists are really saying is that there are forces that cause promiscuity, say, to increase or to decrease. Therefore, even if some Bad Thing was lower in 1958 than in 2008, it may have been in a state of worsening then (increasing), and in a state of improving now (decreasing), so the underlying corrosive forces must have been stronger then and weaker now. It is the strength of these unseen "causes of decline" that I'm interested in.
Sluttiness is perhaps the most frequently given example of how far kids these days have fallen -- fallen, that is, from the zenith of innocence embodied by fucking your gf in the back of your car at Make-out Point (or the drive-in theater), round-robin pairing off during the sexual revolution, and the barely-covers-you costumes of the disco era and its spillover into the nightclub scene of the 1980s. Although there are not national probability samples (as opposed to convenience samples) going back decades for the entire diversity of perversions, indicators of sexual misbehavior correlate with each other, so we may need to rely on a proxy indicator if data are lacking for another.
The most straightforward indicator of sluttiness is simply the percentage of people who have had a "high" number of partners for their age. Since the declinists target the younger generations, let's look at the percentage of high schoolers who have had 4 or more sexual partners. Here are the data from the representative National Youth Risk Behavior Survey. From 1991 to 2007, this percentage has decreased. This is the strongest argument against the declinist hypothesis.
I could not find a good national probability sample that included a straightforward measure of sluttiness before 1991, but we can look at some proxies. The percentage of adolescents who have ever had sex is one: if you haven't had sex ever, you can't have had multiple partners, and earlier age of first intercourse is correlated with having more partners (that is not a tautology). The YRBS data above show that this indicator too has been decreasing from 1991 to 2007. Before then, we turn to a different dataset, although it is also national and representative: the National Survey of Family Growth. According to the CDC's summary:
Proportions were calculated for adolescent women in each year of age from 15 through 19 who reported having had premarital sexual intercourse by March 1 in 1970, 1975, 1980, 1985, and 1988. For all ages combined for each of these periods, the proportion of adolescent women who reported having had premarital sexual intercourse increased steadily (from 28.6% in 1970 to 51.5% in 1988 (Table 1)).
The 1988 figure of 51.5% is nearly the same as the 1991 figure of 50.8% from the YRBS data (see here, where the data are broken down by male vs. female). Thus, at least as far back as 1970 (and probably earlier), the fraction of teenagers who had had sex was already increasing, it peaked around 1990, and has been decreasing ever since.
We can also look at the spread of sexually transmitted diseases that are very common and have been around long enough for there to be decades of relevant data. First we look at gonorrhea. This table of gonorrhea rates by year shows that it increased from 1941 to 1946, decreased until 1957, increased until 1975, and decreased until 1997, leveling off thereafter. The main trends that emerge are a 20-year period of increase from the late 1950s until the mid-1970s, and a 20-year period of decrease from the mid-1970s to the mid-1990s.
A widespread campaign to treat gonorrhea began when the rate started to decrease, so some of the decreasing trend may be due to better medicine, but combined with the data on number of partners and virginity, some of it must also be due to lower promiscuity. In any case, the data do suggest an increasing trend in promiscuity starting in the late 1950s and lasting at least until the mid-1970s.
Next we look at type 2 herpes. Its prevalence has been decreasing since some time in the late 1990s, especially among adolescents (free full text here, popular journalism write-up here). It had been increasing at least from the mid-1970s to the mid-1990s. Because herpes is not treatable like gonorrhea is, it must be that more responsible sexual behavior has curbed its spread, again in particular among adolescents.
Taken together, these various indicators -- what percentage of teenagers have had "many" partners, what percentage has ever had sex, and what percentage has a common STD -- all argue for a period of cultural decline starting in the 1960s, perhaps as early as the late 1950s, which lasted until about 1990. Since then, however, our culture has been in a state of progress regarding teenage sluttiness. Thus, if any age cohort gets to brag about improving sexual mores, it is those born about 1975 or after.
Finally, note that the average female's appearance tells us nothing about the actual level or rate of increase/decrease in sluttiness. Because this is what most older people use to support the declinist hypothesis -- "young girls didn't used to wear thongs or jeans that low-cut when I was a boy!" -- it's worth emphasizing. Note also that more salacious dance practices among youngsters don't tell us anything real either, something I pointed out with a field study on my personal blog. Girls these days may give you a standing lapdance on the dancefloor, but -- although the male receiver may wish otherwise -- this doesn't mean she is going to fuck you. One plausible reason for the disconnect between appearance and reality is that appearances are largely driven by fashion, which changes for its own sake, rather than reflect underlying changes in preferences or behavior.
While oral sex is not worth looking at as a measure of sluttiness compared to intercourse-related indicators, it's worth mentioning that there is no "oral sex epidemic," as Oprah phrased it in a typically anti-male way. (The guys would refer to it as the "efflorescence of oral sex.") Nor is oral sex being substituted for intercourse, another worry in the mind of the declinists. Read the free pdf of the study here, or if you're lazy, a Newsweek editorial summarizing it. As is usual in these cases, the only thing that is epidemic here is a fear of an epidemic.