Exploring the Drug Cultures of Hundreds of College Campuses Through Their Crime Data

There are 13.4 million full-time college students in America, studying and partying on more than 7,000 college campuses. Of these, 20 percent have used an illegal drug in the last 30 days, and 40 percent have consumed alcohol to excess in the last two weeks.

But every college has its own unique drug scene, and each student has his or her particular drug habits, ranging from "straight edge" to "pharmacy on legs." So to really understand what the landscape of drug use on American college campuses looks like, one has to closely analyze a map. Actually, one has to delve into government and academic data and then create and compare some maps. With only weeks remaining until fresh crime stats are released detailing last year's campus crime activity, now seems the perfect time to do it.

Why don't I begin where some articles on this subject might finish, by revealing the "druggiest" college in America? That is to say, the academic institution that had the most drug arrests on its campus per 1,000 students. Well, it's the State University of New York at Oneonta. In case you haven't heard of it, it's a four-year liberal arts college in, as its name suggests, Oneonta, New York. You might be wondering how exactly SUNY Oneonta earned the not-very-illustrious accolade of being the druggiest college of the 7,000 the nation has within its borders. If you are, I'm glad, because that's where our map-making journey begins.

The Office of Postsecondary Education maintains a Campus Safety and Security Statistics database to which crime statistics are submitted annually via web-based data collection by all postsecondary institutions that receive Title IV funding (which is almost all of them). The most recent data is from 2011, with 2012's coming soon. So, if a college records an alcohol or drug incident either on or off its campus, it will end up recorded in the database, alongside all of the other incidents from all of the other colleges across the country.

I decided to focus on one particular category: on-campus drug and alcohol arrests. That, by the OPE's own glossary definition, means any drug- or alcohol-related arrest, citation or summons that takes place in "...any building or [on any] property owned or controlled by an institution within the same reasonably contiguous geographic area" and, importantly, it includes residence halls, which is obviously where a lot of drug taking and binge drinking happens.

I managed to download the data set before the shutdown of the Federal Government blacked out the OPE's database, then I narrowed down the list of 7,000 colleges to only those with more than 5,000 enrolled students. Otherwise, when you calculate the per capita rate of drug arrests you get some odd and probably not very fair results for the very small ones. This brought the total number down to a more manageable list of a thousand colleges, including all of the ones you've heard of, plus many more no one has heard of except the people who study or work at them.

The above rankings are for drug arrests per 1,000 students on each college campus in 2011. And in the top spot with 13.61, unsurprisingly (because I already mentioned it), sits SUNY Oneonta. University of Colorado at Boulder is in the No. 2 spot with only slightly fewer: 13.57. It was this university that The Daily Beast named "America's Druggiest" in December 2011, although later inspection of the figures by Colorado Police revealed that there had been a mistake: 801 arrests had been added to the OPE database when there had actually been only 351 that year (the rankings here use the updated figure). A CU spokesperson responded to the story by saying "...it doesn't much matter whether you score No. 1 or No. 10. Nobody wants to show up in a survey like that in the Top 20."

Or the Top 50, I imagine. And I would guess that no college would want more than one of its campuses showing up on the list, like the University of Wisconsin (positions 5, 12, 13, 29, 30 and 40), or for that matter SUNY, which is in the top spot, plus positions 3, 26, 27, 44 and 50. But we shouldn't get ahead of ourselves. After all, the above map and table are only based on a single year's crime stats and, as the OPE website is very careful to point out, "...valid comparisons of campus statistics are possible only with study and analysis of the conditions affecting each institution." In other words, turning up on the Top 50 list isn't necessarily the singular fault of a college's on-campus drug culture. Clearly there are other factors at work, not least the overall prevalence of drug abuse by young people who reside in each college's home state. So let's look at that on another map.

The map and tables above show the per capita rate of on-campus college drug arrests for each state. Wisconsin and Colorado are in the Top 10, no doubt thanks specifically to UW's and CU's high rate of drug arrests. The map and tables below, however, show drug use by young people statewide.

Using data from the same two-year period, we can compare the two maps to get an idea of whether a college has a serious drug problem because the state it resides in has a general problem with young people taking drugs. It seems that for University of Colorado at Boulder, that may indeed be the case, as Colorado ranks 4th for illicit drug use by 18-25-year-olds. The University of Wisconsin doesn't get off as easily, however, because Wisconsin ranks much lower for statewide drug abuse. I'll come back to UW later.

Altogether, four of the Top 10 states for college drug arrests appear on the Top 10 list for statewide drug abuse: Colorado, Montana, New Hampshire and Vermont. Perhaps the most notable difference between the two maps is North Dakota and South Dakota. The state stats for these are in the fifth lowest bracket but the first for college drug arrests. That's a big difference. So why might that be? Well, it seems to be thanks to four institutions in particular: University of North Dakota, North Dakota State University, South Dakota State University and University of South Dakota. I'm guessing these pretty much include most college students for both states and that they each have consistently high per capita rates of on-campus drug arrests for 2009, 2010 and 2011. Whatever the cause for those incidents was, it appears to be unrelated to the states' overall rate of drug use by young people of the same age.

Here are both maps next to one another to make comparing them easier.

So, overall, what number of college students could be said to have a significant substance abuse problem? Well, a few years ago, The National Center on Addiction and Substance Abuse calculated that 23 percent of college students meet the medical criteria for substance abuse or dependence. That's about triple the proportion in the general population, which is partly to be expected (because young people do more drugs than older people), but it's still certainly cause for some concern. The other major category within the data set I downloaded, and the cause of 600,000 student injuries, 700,000 assaults and 100,000 sexual assaults and date rapes every year, is alcohol abuse. Let's compare on-campus alcohol arrests by colleges and states.

I said I'd come back to the University of Wisconsin; this time it's in the No. 1 spot for on-campus alcohol arrests and also in positions 4, 10, 17, 19, 23, 30, 36, 39 and 41. In fact, it's in very nearly 10% of the Top 50 positions, which seems significant. SUNY, on the other hand, which appeared in six of the Top 50 positions for drug arrests, doesn't appear once for alcohol. This is where some real differences begin to emerge between colleges and their states' rates of alcohol abuse among young people help us highlight them.

North Dakota, South Dakota, New Hampshire and UC Boulder's state of Colorado again appear in the Top 10 (having done so previously for drug use), while several states that have so far gone unmentioned remain in the Bottom 10, like California and Florida. Florida and Texas don't feature much at all throughout the results of this analysis, in fact, whether for campus drug or alcohol arrests, which came as some surprise to me, having previously analyzed drug seizure data for all U.S. states and finding them both right at the top of the list.

Minnesota and New Hampshire appear in the Top 10 for campus alcohol arrests and statewide binge drinking. Interestingly, North Dakota and South Dakota do too, whereas for the drug data both states appeared in the Top 10 for campus arrests but the Bottom 10 for statewide drug abuse. This time, for alcohol arrests, it seems North Dakota and South Dakota colleges get a bit of a pass: They can blame their states' high prevalence of binge drinking for their high number of booze-related campus crimes. Having said that, they might still be concerned about appearing in both Top 10 lists, regardless of the reasons.

I mentioned above some of the fallout binge drinking on college campuses causes each year. The most serious consequence, of course, is death. Estimates vary on how many college students a year die from alcohol-related incidents. Former board member of Mothers Against Drunk Driving Ralph Hingson came up with an estimate of 1,700 college student deaths annually. However, this number was later criticized by the federal General Accountability Office. USA Today conducted its own analysis of college student deaths in the United States over a five-year period beginning Jan. 1, 2000, and arrived at 36 per year, a much lower but still harrowing number.

In the course of researching this article, I stumbled across a record of student deaths by alcohol- related causes that is maintained by the parents of a young female college student who herself died in an accident exacerbated by drinking. Their site is called Compelled To Act, and it lists nearly 400 student deaths since 2004 (roughly 40 a year, which is in line with USA Today's estimate).

I decided to map all the data from CompelledToAct.com and for the first time geographically plot the locations of the colleges to which the deceased students belonged when they died.

It only takes a brief glance at the map to see how many more of the students who died were male. This at first seems surprising given recent news that college binge drinking is more likely among female students than males (with 64 percent of females saying they'd exceeded the weekly guidelines at least once, compared to 60 percent of males). But when the circumstances of the students' deaths are considered, the high proportion of male students who died in part due to alcohol makes more sense, not least because men are more likely to engage in high risk physical situations like fighting and drunk driving (four out of five people who drink and drive are male).

College students who binge drink, apart from being at a higher risk of suicidal thoughts, depression and anxiety, are also far more likely to use the only kind of drug I've so far not mentioned: prescription stimulants, the most popular of which is Adderall.

Adderall is often called the "study drug" because of how many college students take it (without a legitimate prescription) to help with their studies.

Typically it's prescribed to treat patients with ADHD and narcolepsy, but college students, especially those who feel they're under a lot of pressure, use it to stay up all night and improve their focus. In fact, it's so widely used by college students that they are, as a group, twice as likely to abuse Adderall than people who are of the same age and not in college. What's more, 89.5 percent of college students who use Adderall non-medically also report past-month binge drinking, and more than half are heavy alcohol users.

The graph on the left shows just how much more alcohol non-medical users of Adderall drink than those who don't take it at all.

It's been known for a while that more and more college students are taking Adderall (which hardly seems surprising when you learn that the U.S. produces 88 percent of the world's legal amphetamine), but hard data on exactly who is taking Adderall and on which college campuses around the country are harder to establish.

I was mostly drawing a blank while trying to find good geographical data on Adderall use by college students, but then I remembered that Google had successfully been able to predict flu epidemics by analyzing flu-related searches made by people around the country. I thought I'd try to use the same tool to have a look at the subject of Adderall.

It's called Google Correlate and it works in two ways. One is by uploading a data set that contains either a time series with values or geographical attributes and values, and then letting Google find search trends that match the patterns in the data as closely as possible. For example, if you do this with a data set whose values peak in the winter but dip in the summer, Google Correlate's output might find that the best search terms that match that data series are "mittens" or "indoor heaters," because these are the things that people searched for most during the peak times in the data series. And the map the tool produces might go further by showing the most interest for those terms in states that are especially cold in the winter (and therefore most in need of heaters and mittens).

The other way the tool works is by inputting a search term, letting it find the search volume over time for it, stripping out the term, and then reanalyzing what new terms best match that data series. If none of this makes sense, but you're interested in learning more, check out the tool's FAQ. I've only tried to describe roughly how it works because of what is about to follow: the first use of Google Correlate to clearly show that searches about Adderall are highest in May and December of each year, when college students are desperately cramming for their summer and fall final exams.

The only other person I could find who has noticed a similar thing is a reporter for OKNews named Jaclyn Cosgrove, who found what looked like some kind of trend similar to the one above by searching for "Adderall" using Google Trends (not Correlate) and looking at the resulting graph over a 12-month period. Her blog post went almost completely unnoticed at the time, perhaps because the screenshots she included of her finding didn't seem to show a truly compelling and undeniable pattern.

To the left is what she found. The line does look fairly erratic.

But using Google Correlate and searching for "Adderall" produces the graph above and a seemingly impossible-to-deny correlation between searches for "Adderall" and "taking Adderall" and the two periods in the academic calendar when students are most likely to use Adderall to study.

But I wanted to make sure. After all, maybe concerned parents are actually the ones responsible for all those searches because around the times when they fear their kids are taking Adderall at college (whether they are or not), they use Google to research the drug. Seems unlikely, but it's possible. So I used Google Correlate again, but this time I inputted search terms that are much, much more likely to come from students on the brink of using Adderall to study harder.

I chose the terms "snort Adderall" and "30 mg Adderall" because some research I did on Drugs-Forum.com, a place where students discuss how to best non-medically use Adderall and what the side effects are likely to be, suggested that these are two searches a lot of students would be likely to make. 30mg is the most common starting dose for a person with no tolerance for the drug and snorting it, well, take a look.

"SWIM" is an acronym that stands for Someone Who Isn't Me - a way for the forum users to avoid incriminating themselves.

I also checked Google Trends (which works in the opposite way to Google Correlate) to see if I could find a clear representation of the trend over the last several years.

Not only does the May/December pattern hold for the last four years, the graph also shows a year-on-year increase in the volume of searches for the word "Adderall."

There are, in fact, two other studies I could find that back up the theory that Adderall use spikes around students' finals. One involved testing wastewater that came from colleges to see if it contained traces of the drug (the hypothesis being that they'd find more around May and December, which they did). The second one was published about six months ago and consisted of an analysis of 213,633 public tweets that mentioned Adderall during a six-month period between November 2011 and May 2012. By analyzing the geographical data and textual content of the tweets, they discovered that mentions of Adderall spike around final exam periods and were most common in the northeastern and southern United States. Specifically, Vermont had the highest per capita rate per 100,000 students of all states (66.4), followed by Massachusetts (54.6) and Alabama (52.2).

The next question, then, was whether Google Trends could do more than confirm the same phenomenon by matching the top three states for Adderall search activity to the top three identified by the Twitter study.

I filtered Google Trend's results to the exact same time period used in the Twitter analysis, and it showed me the Top 10 states for "Adderall" search volume. Two of the same top three states showed up: Alabama and Massachusetts. The third, Vermont, didn't appear in the Top 10 for that period but was in the Top 10 throughout 2009.

Combining the above insights with the two previous studies seems to practically close the case on the question of whether or not college students use Adderall most just before their final exams. The answer is yes and especially in northeastern and southern states.

Stressed students desperately cramming for their exams got me thinking about something I hadn't seen at all along my journey through U.S. college drug culture: any references to eight of the most renowned colleges in America, those that are part of the Ivy League.

The rankings above are based on the per capita drug and alcohol arrests that took place on Ivy League campuses between 2009 and 2011. What's obvious is that there weren't very many of them. Dartmouth College sits comfortably in first position, with more than five times more alcohol arrests than Yale University. This result matches a known problem Dartmouth has with drinking on its campus, attributed mostly to the fact that fraternities dominate the social scene there. About two-thirds of undergraduates at Dartmouth join a fraternity or sorority, which is nearly double the rate of any other Ivy League school. And Greeks are known to be more likely to abuse alcohol.

I started out with the conclusion that The State University of New York at Oneonta was the druggiest college in America. But that was just in 2011, and it didn't include data on alcohol, which as we've seen is more abused than drugs and goes hand-in-hand with the abuse of prescription medication. With that in mind, I thought I'd try to come up with a better representation of which colleges are really the most drug and booze heavy. First I checked which colleges appeared on both the Top 50 lists for most drug and alcohol arrests over the last three years. Twenty did. Then I combined the drug and booze per capita rates for these 20 into a single score and ranked them accordingly. The result is the bank of boxes below; each one's size relative to its score, with each one also showing the actual score, the college name and the institution's brand color and logo. At a glance you can see which really rank as the druggiest in the nation. And yes, that is the University of Wisconsin taking up five of the 20 boxes.

I told you I'd come back to that place.


  1. http://nces.ed.gov/fastfacts/display.asp?id=372
  2. http://usatoday30.usatoday.com/news/nation/2007-03-15-college-drug-use_N.htm
  3. http://compelledtoact.com/Tragic_listing/Main_listing_victims.htm
  4. http://www.samhsa.gov/data/NSDUH/2k11State/NSDUHsaeMapsAlts2011.htm
  5. http://www.thedailybeast.com/galleries/2010/12/11/druggiest-colleges.html
  6. http://blogs.westword.com/latestword/2011/12/cu_boulder_druggiest_campus_420.php
  7. http://usatoday30.usatoday.com/news/nation/2007-03-15-college-drug-use_N.htm
  8. http://www.iowastatedaily.com/news/article_82ad5eda-3f14-11e2-8788-001a4bcf887a.html
  9. http://www.samhsa.gov/samhsanewsletter/Volume_17_Number_3/Adderall.aspx
  10. http://news.psu.edu/story/286174/2013/09/05/research/probing-question-how-serious-binge-drinking-problem-college
  11. http://www.cdc.gov/vitalsigns/drinkinganddriving/
  12. http://www.usnews.com/news/articles/2013/05/17/study-college-women-binge-drink-more-often-than-men
  13. http://www2.potsdam.edu/hansondj/YouthIssues/1140106101.htm