<p>Fwiw, while it highly probable and plausible that the SAT score DO track the family income, it is good to remember that the origin of that “data” is none other than the questionaire filled by students when taking the test. Considering that the students hardly know the REAL finances of their parents, the numbers are hardly verifiable. After all, it is the same source that want us to believe that more than 40 percent of students have a GPA above a B level! </p>
<p>Inasmuch as it might have become old news on these shores, it remains that there are people who have used better methods to track the enrollment of students at the elite colleges. One such source is </p>
<p>The data suggest, but obviously don’t prove, that a lot of Midwesterners make that same judgment, Global. Especially in states that have top-notch public flagships. Let’s compare New Jersey and Michigan, states roughly comparable in population; actually at 9.8 million Michigan is slightly larger than New Jersey’s 8.8 million. New Jersey is on average a wealthier state, in fact often ranked as the wealthiest in the nation on a per capita basis, and it’s got a ton of high-end suburbs, so you’d expect more high-SAT kids coming out of high-end schools in NJ. But Michigan’s got some pretty affluent areas, too; Mitt Romney wasn’t the only rich kid on his block when he was growing up there. Still, you’d expect more NJ kids than Michigan kids to apply to, be admitted to, and attend Ivies and other elite private schools, just from the demographics of the two states.</p>
<p>But when you compare numbers, the difference is measured in orders of magnitude. In 2010, Michigan sent 159 freshmen to Ivies; New Jersey sent 1,265. Penn was the top Ivy for Michiganders; MI sent 34. NJ sent 257. MI sent 32 to Cornell; NJ sent 319. MI sent 26 to Harvard; NJ sent 86, but that may not be a fair comparison because NJ residents clearly preferred favorite son Princeton where they sent 180, to MI’s 18. </p>
<p>OK, that’s NJ’s backyard and I think we’ve established there’s a lot of regionalism in this, so what about a neutral court? Duke? NJ 92, MI 16. Vandy? NJ 54, MI 16. Emory? NJ 85, MI 8. OK, Michiganders don’t much care for the South.</p>
<p>How about the West? Stanford? NJ 37, MI 10; ok, at least that’s a little closer. Caltech? NJ 9, MI 5; closer still. USC? NJ 46, MI 16, less than 3-to-1, but who cares about USC?</p>
<p>How about Michigan’s home turf, the Midwest? Chicago? NJ 67, MI 31; uh-oh. Northwestern? MI 81, NJ 74, that’s more like it . . . but it’s only 1 school. WUSTL? NJ 73, MI 27, but wait a minute, Michiganders tend to think of St. Louis as one of the few cities in the country that vie with Detroit for the highest murder rate, so naturally they tend not to go there. Notre Dame? MI 92, NJ 85, Northwestern-like numbers, but any true Michigander would hate to hang his hat on the state’s connection with Notre Dame; yuck!</p>
<p>So what’s going on here? Clearly NJ dwarfs Michigan in elite private school admissions. Is it just demographics? Are New Jerseyans really smarter? (And if they are, who do they live in New Jersey, for gosh sakes?) </p>
<p>No, I think it’s clear what’s going on here. Michigan has an outstanding public flagship, and most of the top-end students in the state–not all, but most–are perfectly happy to go there. Not so in NJ. Sure, some top NJ kids end up at Rutgers, but very few academically top end kids in NJ are going to speak with pride about going to Rutgers. They’re desperate for some prestigious private alternative; the more prestigious, the better, but if push comes to shove they’ll take pretty much any top 50 private school, and actually a lot of them will take Michigan over Rutgers; in 2010 Michigan drew 234 freshmen from NJ, which puts Michigan pretty high up in the pecking order of “elite” schools that NJ kids actually decide to attend once their choices are clear. And Michigan kids deciding to attend Rutgers? Zero, zilch, nil, nada. Not a single one. Terms of trade are pretty clear. </p>
<p>So it’s no surprise that 75th percentile SAT scores are 130 points higher at Michigan than at Rutgers, and 25th percentile almost 200 points higher. The high-end kids in NJ go private; the high-end kids in MI mostly go public, and the University of Michigan is strong enough to pull in a lot of high-end kids from NY and NJ to boot.</p>
<p>Proof? Well, I can’t offer drop-dead proof, but here’s a clue: Among 2011 college-bound seniors, 3,341 New Jerseyans sent SAT score reports to Cornell, the most popular Ivy in NJ, followed by Princeton with 3,227. In Michigan, 702 sent SAT score reports to Harvard, the most popular elite private school among Michigan residents, followed by Northwestern with 674. That’s a ratio of nearly 5-to-1. Roughly 7 times as many New Jerseyans as Michiganders took SAT subject tests, which would have been necessary to apply to Harvard or Princeton. Roughly 5 to 6 times as many New Jerseyans scored in the 700-800 range on the SAT Reasoning Test, but that’s not surprising given that about 30 times as many students took the SAT in New Jersey, an SAT-dominant state, as in Michigan, an ACT-dominant state. And we know that in the same year, about 4,000 kids in Michigan scored 31 or higher on the ACT, which is probably roughly comparable to the number of NJ kids who scored 700+ on both CR and M on the SAT.</p>
<p>So we have pretty similar top ends in the two states, maybe a slight edge to NJ because of demographics and more top-end schools. But most of the top-end students in NJ go private, and most of the top-end students in MI go public. It may be a sensible in choice in each case; whether the Ivies are worth it may depend on where you live, and what your other options are. For top-end Northeasterners, the answer seems clearly to be “yes.” For Midwesterners, at least in states like Michigan that have an outstanding public optiob, most people seem to say “No.” The Ivies aren’t even on their radar, because they’ve got great public options much closer to home.</p>
<p>Fair enough. There are some Michiganders, and even a few OOS, for whom Michigan State is a more sensible choice, especially given FA offers, tuition differentials, and such. I’ve actually known quite a few in that group, and I respect that choice. Some of the MSU grads I know have done terrifically well for themselves, others only moderately well but still well worth what they put into it. More power to 'em. Heck, Teresa Sullivan, the much-lauded (of late) president of UVA, is a MSU alum. And she was good enough to become Provost at Michigan. (Only kidding!)</p>
<p>But peace, brother. I think we’re on the same side here. FWIW, MSU got 40 freshmen from NJ in 2010. Rutgers got zero from Michigan. As I said earlier, the terms of trade are clear.</p>
<p>There is an enormous amount of solid data that correlates academic achievement with income<<</p>
<p>You wrote test scores and GPA. Care to offer a glimpse of the solid DATA that correlates test scores to income? And again, nothing that is based on the BS collected by the College Board on the student questionnaire amounts to data.</p>
<p>Oh Jym, Tulane allowed many more than 112 that is how many accepted to enroll. The allowed could be numbered in four digits. High applications, high admit rates, and low yield is Tulane’s defining image. </p>
<p>Yes but isn’t it reasonable to assume that higher income correlates with higher ability level, & thus higher income & higher SAT scores are both byproducts of high ability?</p>
<p>AD, what is the origin of the numbers used in this quotation? Is the income used in the 1972 NLS sample … not reported by the students themselves. </p>
<p>Again, and again, while the assumption that income and academic performance are related is PROBABLY true in general terms, using the self-reported numbers by teenagers when taking the SAT is just as valid as asking them about their future plans. </p>
<p>There are a number of academic attributes “we” simply accept as being true and decide it is not worth spending time building a verifiable pool of data. This is true for the SAT vs Income correlation. Similarly, nobody seems really interested in spending much time and resources on the impact of Early Decisions on the lower SES students. It is easier to assume that the ED favors the rich and well-connected. And then plenty assume the assumptions were the product of verifiable research. And then it is repeated. </p>
<p>Fwiw, what do you ASSUME might the correlation be of SAT scores/income in a group that would be culled from all the freshman Pell grantees at UC-Berkeley. What do you assume might be revealed if you only include zero EFC Asians?</p>
<p>A good bet is that you might to draw different conclusions, or conclude that the students might come from families that are adept at playing musical chairs with the reportable income of small businesses.</p>
<p>And fwiw, one could start an entire discussion about how a lower SAT score is INTERPRETED by admissions officers. Aren’t we led to believe that adcoms routinely place the students in their racial or SES contexts? Meaning that the disadvantaged earns some brownie points to level the playing field?</p>
<p>bclintonk, thanks so much for the work you did digging out the numbers in your post comparing college choices for MI and NJ students. It would be interesting to see how other states with strong public flagships compare.</p>
<p>@bclintonk You probably shouldn’t use sent SAT data when comparing NJ and Michigan applicant interests because of the ACT-heavy body of applicants in Michigan. I really enjoyed your post though.</p>
<p>Well, the number of SAT score reports sent does have validity for some schools—those that require SAT Subject Tests even for applicants submitting the ACT. That would include Harvard, Princeton, and a few others. It’s true that a kid in an ACT-dominant state (or anyone who submits ACT scores, for that matter) can apply to Yale, Brown, or Penn without sending any SAT score reports. But a complete application to Harvard or Princeton requires SAT score reports. The College Board data on the number of college-bound seniors sending SAT score reports to various colleges would include anyone who sent Reasoning Test scores, anyone who sent Subject test scores, and anyone who sent both. So the number of Michigan kids who sent SAT Score Reports to Harvard does represent an absolute upper bound on the number of Michigan kids who sent in complete applications to Harvard.</p>
<p>I take your point, though. Some Michigan kids might have applied to other private schools that don’t require Subject Tests at all, or don’t require them of those submitting the ACT (usually ACT with writing). Just because Harvard received the most SAT score reports from Michigan kids doesn’t necessarily mean it was the elite private school the most Michigan kids applied to. We know, for example, that Northwestern and Notre Dame are the two top private schools actually attended by the most Michigan kids in 2010, with 81 and 92 in their respective freshman classes. Neither actually requires Subject Tests, though Northwestern “recommends” them, which probably means most serious applicants are going to take them and send them. Michigan kids do send SAT score reports to Northwestern in numbers similar to Harvard; in 2011, 702 sent SAT score reports to Harvard and 674 sent SAT score reports to Northwestern, but it’s entirely possible that there was some undetermined additional number of applicants to Northwestern who submitted only ACT scores, reasoning that “recommends” means “not required” and they’d take their chances and submit only the ACT score. </p>
<p>Notre Dame doesn’t even “recommend” Subject Tests, it just says it will use them only if they enhance the application, which makes it sound much more optional. And sure enough, only 209 Michiganders send SAT sore reports to Notre Dame. No doubt many more applied with only their ACT scores; we don’t know how many more.</p>
<p>Yes, it would. Unfortunately, I have a “day job” that limits how much number-crunching I can do on this.</p>
<p>Here’s one additional little clue, though: Minnesota and Wisconsin are in many ways two peas in a pod, very similar in population, obviously in the same region, similar ethnic and religious composition. Minnesota has a somewhat higher per capita income because there are more high-tech and white collar professional jobs around the Twin Cities. Madison’s economy is in some ways comparable to the Twin Cities but Madison’s much smaller; Wisconsin’s biggest city, Milwaukee, is still primarily a lunch-bucket, blue collar town. That difference in income levels and educational background might have something to do with it, but a quick scan of the data on my spreadsheets says that Minnesota sends about 41% more kids to elite private universities than Wisconsin does. At the Ivies the gap is even bigger, with Minnesota sending about 60% more.</p>
<p>Why? Well, income and educational levels of the states’ households aside, I think one reason may be the state flagships. Ask anyone in Wisconsin which of the two has the better state flagship, and they’ll say Wisconsin, without a doubt. Many will tell you Wisconsin is one of the best universities in the world. Ask the same question of Minnesotans and some will gamely defend the Gophers, but a sizable fraction will admit, perhaps sheepishly, that Wisconsin probably has the edge. And I think that’s right. The University of Minnesota is not chopped liver and there are some areas where it matches or exceeds Wisconsin, but overall, the University of Wisconsin is who the University of Minnesota wants to be. Well, they want to be Michigan but they know they’re a long way off from that; Wisconsin is who they’re chasing for the foreseeable future. (I like to say the University of Minnesota is like the kids in Garrison Keillor’s Lake Wobegon: “above average”).</p>
<p>For that reason, a lot of the better students from Minnesota actually go to the University of Wisconsin; 738 in the 2010 freshman class, to be exact. And the 502 kids Minnesota sent to elite private universities were on top of that. To be sure, a lot of Wisconsites come to the University of Minnesota, too, but parts of western Wisconsin are in the Twin Cities metro area and much of western Wisconsin is more oriented toward the Twin Cities than to Madison and Milwaukee, and hey, they get tuition reciprocity, so why not go to the more local school? Some may come for particular programs where Minnesota is as good or better than Wisconsin, but I suspect fewer of the Wisconsinites coming to Minnesota would say they did it because Minnesota’s the better school, whereas probably most Minnesotans going to Wisconsin would say that. </p>
<p>Anyway, just a little anecdotal evidence, but worth considering.</p>
<p>Another pairwise comparison: Maryland and Virginia. They border each other, obviously, and share the DC metro area. Virginia is about 40% larger in population, both are pretty high-income but Maryland is definitely higher. Both have pretty good state flagships but Virginians tend to think more highly of theirs. (My own view is that UVA is probably not quite as good as many Virginians think it is, while UMD-College Park is probably a bit better than many Marylanders think it is, but that’s another story). If you look at the numbers from each state that attend elite private universities. Maryland just absolutely trounces Virginia: despite a much smaller population, Maryland sends about 34% more kids to elite private universities. </p>
<p>Why? Well, if a Virginia kid wants to go to the best university in the state or region, many will think that’s UVA; or at any rate, many will think it’s the best at an affordable sticker price. A smaller number will say it’s Duke, which Virginians attend in larger numbers (about twice as many Virginians at Duke, one of the few elite private universities Virginians attend in larger numbers than Marylanders). If a Maryland kid wants the best in the state or region, she’ll probably think Johns Hopkins or Georgetown, both attended by far more Marylanders than Virginians. UMD would run a distant third, or worse. Some Marylanders might say it’s Penn, which for many might be as little as an hour or two away; Maryland, despite a relatively small population, is Penn’s 7th-largest source of students. A few might even say it’s UVA; in 2010, 121 Marylanders enrolled as freshmen at UVA, compared to only 38 Virginians who enrolled at UMD-College Park.</p>
<p>And it again shows in the SAT scores at the two flagships. With Virginia retaining a lot of the top students in the state and attracting even more from other states, SAT scores are 30 points higher at the 25th percentile and 60 points higher at the 75th percentile at UVA than at UMD-College Park. So it becomes self-reinforcing; with higher stats, UVA is higher ranked and continues to attract higher-end students. With lower SAT medians, UMD is lower ranked and is shunned by more top-end Marylanders in favor of private colleges and universities, or even OOS publics- in addition to the 121 who enrolled at UVA in 2010, 104 enrolled at Michigan, 57 at UNC-Chapel Hill (which restricts OOS admissions), and a whopping 318 at Penn State. Which of course would tend to depress the 75th percentile SAT scores at UMD.</p>
<p>So this generally seems consistent with my hypothesis.</p>
<p>Countering that, though, is California which has a great public university system and still sends thousands upon thousands of top students to top private universities, near and far. Don’t know what to make of it.</p>
<p>There are plenty of possible reasons. One of them is that California is a very populous state, and that beyond the obvious public school systems, there are really (relatively) few highly desirable schools. Yes, there is Stanford and Caltech. Yes there is USC and perhaps Santa Clara. Look at the number of seats at the LACs … the Claremont schools with the possible addition of Oxy. </p>
<p>One reality is that the students who are qualified (and perhaps lucky) enough to be accepted at one of the private schools in California or at a prestigious OOS school … do seize that opportunity. And this could explain part of the attraction of an exodus for the student who DOES HAVE choices. The remaining pool (without options outside the local/regional schools) do end up atteding. </p>
<p>Despite all the noise made on behalf of the UC system, some REAL numbers are telling. While there is an overlap in admission between Stanford and Cal. there is hardly one in terms of enrollment. Simply stated, except for a handful (and that means a number that fits on the digits of a human being) all students accepted at Stanford and Cal do end up enrolling at Stanford. By choice for sure, and also because of a superior financial aid at the Farm. People do pick schools that maximize their opportunities, mininize the costs, and also pick schools that have the highest prestige. All in all, it looks like the first choices are local prestige (read Stanford or Caltech), then OOS prestigious private, and then a number of choices including public big names and LACs. After that, you really have a mixed bag of CSU and mediocre schools. Obviously, there is a last sidenote, and that is made out of the Junior Colleges a la De Anza who ARE bona fide pipelines to the big UC names to the massive number of transfers. Again, telling that third (or fourth) years students at Cal WILL sit in classes with students who were JUCO transfers. </p>
<p>In terms of exodus, it is also telling to see to which schools Stanford loses cross-admits. It is mostly composed of the Ivies and the two main tech schools. Of course, one should segregate that along geopraphical pools, but one could do worse that assuming that it does lose more than a few local kids to the Northeast.</p>
<p>Considering the fiscal problems and the California version of diversity, it does not seem to hard to understand that the people who have options do tend to exercise them.</p>
<p>I can’t really speak to the situation in California, but I have always been skeptical that cross-admit data tells us anything interesting. I can speak to my own experience, now many years old but I think still valid. I just always wanted to go to. Michigan, my in-state flagship. My HS GC tried to encourage me to think about private schools. Not that he knew anything about it; he went to a public university that was not nearly as good as Michigan, and I had never heard of anyone from my hometown going to an elite private school. But he knew I had credentials that would have made me competitive, and in those days it was much easier to get into even the very best private schools than it is today. So I looked at a few and really thought about it, and decided I really preferred Michigan. I never applied to the private schools. So I never would have showed up in any cross-admit data. I “voted with my applications,” if you will. If I had wanted to go to Harvard, or Yale, or Chicago, I would have applied to that school and to Michigan, in case I didn’t get into my top choice. But since Michigan was my top choice and it was easier to get into Michigan, there was absolutely no point to applying to the schools that were more selective. Who knows if I’d have been admitted if I had applied. But my suspicion has always been that there are a lot of people like me, plausible candidates for a more selective school and a somewhat less selective school who, for whatever reason, prefer the somewhat less selective school. I suspect that happens with state universities all the time. And like me, if they truly prefer e slews selective school, they won’t bother to apply to the more selective school. They’ll apply to their top choice and an even less selective alternative, in case they don’t get into their top choice.</p>
<p>If that’s the case, then there’s going to be a profound selection bias in the cross-admit pool. Cross-admits between Harvard and Michigan will be mostly people who prefer Harvard and are applying to Michigan as Plan B. If they’re admitted to both, they’ll choose Harvard almost every time. People who prefer Michigan to Harvard will mostly not apply to Harvard, so they can’t possibly end up in the cross-admit pool. They’ll be like me: they’ll apply to Michigan, and because it’s not absolutely certain they’ll be admitted there, they’ll apply to Michigan State as a back-up. People who prefer Michigan State to Michigan won’t bother to apply to Michigan, so the cross-admit pool between Michigan and Michigan State will be biased by the presence of large numbers of people who prefer Michigan and an absence of people who prefer Michigan State. And so on down the line.</p>
<p>If that’s the case, then the “revealed preferences” of cross-admits aren’t telling us much more than which is the more selective school. Which, although I haven’t studied it,seems pretty consistent with the cross-admit data I’ve seen. The only time it could tell us anything interesting, then, is when there are two more-or-less equally selective schools doing battle.</p>
<p>Another case in point: my D1 didn’tapply to any Ivies because she just really didn’t like any of them. She much preferred LACs; to her, all the Ivies felt like big, impersonal research universities, which of course they are. She did like Brown more than the others, but she liked Brown, she concluded, because it felt more like an LAC than the others. But it didn’t feel like a real LAC, it felt like a research university masquerading as an LAC by exhibiting some LAC-like characteristics. So she applied ED to the LAC that she most favored, and was accepted. End of story. Could she have been accepted at Brown? Who knows? Her stats and other credentials certainly would have made her competitive, but we rated it a “reach.” had she not been admitted ED to her LAC, she might have applied to Brown. Or maybe not, because she really didn’t like it as well as some other LACs on her list, and it was harder to get into, so what would have been the point? I’m sure there are some cross-admits to Brown and to her LAC, and I have no doubt Brown totally clobbers her LAC in the cross-admit datat, but that’s because most of the people who apply to Bith schools really prefer Brown and are adding her LAC, which while very selective is not as selective as Brown, as a backup. Those who prefer the LAC are much less likely to apply to Brown, because,as my D reasoned, what would be the point?</p>
<p>This is not what I expected (since Michigan is more prestigious than Rutgers). So what’s going on? One possibility is that Rutgers’ proximity trumps Michigan’s prestige</p>
<p>(Another is that the data is unreliable. I’m skeptical about the number of ties that show up in the parchment.com results. There may be other explanations having to do with costs or demographics.)</p>