More public support for higher education

A story on Congressman Jim Webb's proposal to more completely fund veterans college education ran this morning on NPR's Morning Edition. More specifics on Congressman Webb's proposal is to found here on his website. Another related story ran on Marketplace Morning Report. On October 1st a new law comes into effect that changes how students loans are distributed and the rates of interest charged. For more on this click here. The impetus for this new law comes from the Campaign for America's Future .

The G.I. Bill

One way of describing what college professors do is to say that we perpetuate middle class privilege. We could also say that we help to protect the economic gains many families and the American economy overall enjoyed as a result of the G.I. Bill. For more on the G.I. Bill and its impact on American society click here.

socioeconomic standing and access to higher education

Over the last several days I have been struggling with some numbers I had seen reported. These numbers are statistics on minority enrollments and the persistence of enrollment gaps between African-Americans, Hispanic Americans, Non-Hispanic White Americans and Asian Americans. A study by Carnavale and Fry for ETS "Crossing the Great Divide: Can we Achieve Equity When Generation Y Goes to College?" notes that, although the total number of minority students (African-American and Hispanic American) enrolled in college is projected to rise as is the percentage of total enrollments, a significant gap between the share of 18-24 year olds in these groups and the number of African American and Hispanic American 18-24 year olds enrolled as students will grow. (By 2015, 550,000 fewer young Hispanic-Americans enrolled than one would expect given projected population and over 200,000 fewer young African-Americans.) This is a very funny number. The projected population totals are based sensibly on current trends in the growth of different groups (presumably it is based firstly on the number of children born between 1991 and 1997 and then with adjustments made for complicating factors such as immigration, deaths, and so on). However, the appendix to "Crossing" explains that projected enrollments are based on the assumption that current enrollment rates (obscenely lower for African-American and Hispanic Americans 18-24 year olds) will continue. (see Projection Methodology p. 55 of "Crossing"). So numbers such as those for Fall 2005 from the Census Bureau were used. These numbers show that while 40% of non-Hispanic White Americans 18-24 years old were enrolled full time or part time in four year colleges and 50% of Asian Americans only 29% of African Americans 18-24 years old were enrolled part time or full time in four year colleges and only 24% of Hispanic Americans of this age group were. Given the age range reported (18 year olds to 24 year olds) one would expect to see a fraction of 18-24 year olds who have completed their undergraduate degrees. And indeed you do see this. According to the Census Bureau the numbers were 4% for White-Americans, 3.5% for African Americans, only 1.82% for Hispanic Americans, and almost 9% for Asian Americans in October 2005. More importantly and my main reason for looking at these numbers are that regardless of ethnicity there are significant differences in enrollment rates -in either two year or four year colleges or in four year colleges- according to SES (socioeconomic standing). Haveman and Smeeding( "The Role of Higher Education in Social Mobility", The Future of Children, vol. 16, no. 2., Fall 2006, pp. 125-150) note that 90% percent of students from families with incomes in the top quartile of earnings were enrolled in college twenty months after graduation (73% in four year colleges) only 60% of students from the bottom quartile were enrolled in college twenty months after graduation (with fewer than half enrolled in four year colleges, 37% in two year colleges, and 17% enrolled in vocational/technical schools.) To assume that low enrollment rates will persist -as Carnavale and Fry implicitly do- is not very helpful. Similarly, while racial discrimination is in part the cause of lower SES, it is much more important to focus attention on social class as an obstacle to higher education and success. To focus on race seems to invite substantive versus formalist interpretations of the achievment gap. In other words, that the differences in achievment are in large part due to cultural factors. However, controlling for race, there is a strong statistically significant relationship between family income levels and enrollment in college (in either two or four year schools or in four year schools). Using data for October 2005, I find that enrollment rates were about 23% points higher for 18-24 year olds students from families with incomes in the $100,000 to $150,000 range than they were for 18-24 year olds whose families earned between $30,000-$40,000 for example. (These numbers are derived from regression I have run using Census Bureau data on fall enrollments for Fall of 2005.) I am incredibly impressed with a paper entitled "Reducing Inequality in Higher Education: Where Do We Go From Here?" by Cornell University's Ronald Ehrenberg . I am impressed first because of its focus on SES and the inequalities that are perpetuated by both the differences in access to higher education as well as differences in levels of success after a student enters colleges. I am also impressed because of its focus on an analysis of what possibly can be done -in terms of financial aid, education policy at the primary and secondary levels, and within college itself- to assist students from economically disadvantaged backgrounds . For good measure there is a thoughtful discussion of the direct and indirect financial pressures colleges and universities face and how these pressures have led them to worry far less about their role in perpetuating inequality. The paper was presented at the Conference on Economic Inequality and Higher Education: Access, Persistence, and Success, Syracuse University, September 23-24, 2005). Ehrenberg's main points are (1) that little advance has been made in enrollments of students from low income families since the 1970s, (2) that states have increasingly faced tight budgets and, as a result, have kept state spending per student at more or less constant levels since the 1970s, (3) that increased spending per student has therefore been funded largely by tuition increases, (4) that , perversely, as the private returns to college have grown and the gap in earnings between the college educated and those without college degrees has increased, the tuition increases have been more or less tolerated by middle class families. The increases are regarded as the price to be paid for protecting their position. (5) that the increase in tution at four year colleges has shifted proportionately more and more low income students into two year colleges, (6) that given the difficulty of transitioning from two year to four year colleges, especially to those schools where transfers from two year colleges are relatively rare -persistence- the ability of transfer students to complete their undergraduate degrees has fallen.

He outlines several very recent programs at flagship public universities such as UVA, the University of Texas, and UNC-Chapel Hill to increase enrollment of students from low-income families as well as very recent efforts by Princeton (1998), Harvard (2005), and Yale (2006) to increase enrollment rates and graduation rates of students from low-income families. Very selective schools by definition admit and graduate mostly students from wealthier families. Hill, Winston, and Boyd ("Affordability: Family Incomes and Net Prices at Highly Selective Private Colleges and Universities" Journal of Human Resources, 40 (fall 2005)) note that 108,721 students earned undergraduate degrees at the 28 highly selective private colleges and universities in their study. This implies that well under 10% of all the students earning bachelor's degrees in spring of 2002 received their degrees from such schools. Given the cost of attending such institutions, one can safely assume that families of relatively modest means were receiving some sort of financial aid. Assuming that all middle class and low-income families attending these schools received some kind of aid, students from families earning less than $51,000 a year represented only 17% of students graduating from such schools.

To make these very exclusive schools more inclusive is fine. However, students from wealthy families are disproportionately represented in these schools and always will be. Besides this, as noted above only a small percent of college graduates in a given year earn their degrees from such institutions. Therefore changes here cannot be expected to create more than a ripple overall. What is needed are similar programs at the colleges and universities where the majority of students pursue their education. However as Ehrenberg points out there is a danger in such institutions being more inclusive because these "lesser" colleges and universities face even more risk -both financially and politically- if they follow suit.

The financial resources of such schools are far more limited. As Hill,Winston, and Boyd (in the study of highly selective private colleges and universities already mentioned above) note there is a statistically significant inverse relationship between the net price students from low-income families are asked to pay for their schooling and the size of the endowment per student a particular school has. In the case of schools with much lower endowments per student Ehrenberg notes that they "...do not have the large flow of annual giving and the large endowments that the University of Virginia or the University of Chapel Hill have and thus it will be more difficult for them to develop the resources necessary to support programs such as Access UVA or the Carolina Covenant."

He goes on to point out that "If these flagship institutions are successful in diversifying their student bodies along racial/ethnic and socioeconomic lines, they run the risk of alienating the traditional supporters of their institutions; the upper and middle class families whose children attend high school that have been large feeders schools for them. This may translate into political pressure to reverse policies that led to the diversification, or a further loss of financial support for the public institutions." (Ehrenberg, p. 14). Other schools face the same political problem if students from the families that have typically supported these schools have a more difficult time being admitted into these schools because equally or more deserving students from poorer families have taken their place.

Therefore the current efforts that are being made to make access to higher education more equitable are likely to have less impact than one would hope for or what is needed to really change the way opportunity is distributed. A few cards in the deck are being moved around and most of the aces are still at the top. The entire deck needs to be reshuffled.

the insanity of college rankings

The very best book to read on the subject is Paul Boyer's College Rankings Exposed. It should be available right next to copies of U.S. News and World Reports annual College Rankings issue (what those at USNWR refer to as their swimsuit issue.) In a recent AP piece (Colleges Struggle to Quit the Rankings Habit) Brian Kelly, editor of USNWR, is quoted as saying that its annual rankings of colleges and universities simply provide "consumers" -hungry for hard data- with the same sort of rankings offered for cars, refrigerators, and the like. This is very telling. Are the families who enroll their students in college really consumers? There is a great deal of data on marriage and the sort of characteristics that contribute to a man becoming a good marriage partner. Weddings are, like college, a major expense. USNWR could rank cities across the U.S. according to the quality of its me. Part of these rankings could be based on what thsoe performing marriage ceremonies in other cities have heard about men in a particular city. In the case of some cities they might have direct knowledge -current or not-while in others they may have little or no information at all. We could then place a great deal of weight on these scores regardless. Families with these rankings in hand could make a much more informed decision about what city is the best one to go to if you wish to locate a husband. Einstein is quoted as saying that "everything that can be counted does not necessarily count; everything that counts cannot necessarily be counted". And this is at root the point. Just as there is not one universally valid criteria for selecting someone with whom you will share your life with there is not a universally valid criteria for selecting a college or university. While we can all agree on a fairly limited list of factors that make one big screen TV better than another or, similarly, one set of tires better than another how on Earth can it be argued that there is a relatively finite list of factors that determine waht constututes a good school? Consider also that when it comes to rankings that the most elite schools are so because they serve, as they always have, the elite. Socioeconomic status is a powerful deteminant of academic achievement. Highly ranked schools are largely populated by students from "highly ranked families"; the well-to-do and the powerful. Wealthy alumni can afford to make large contributions. Large contributions together with tuition revenues mean that such schools are able to recruit and retain faculty at relatively high salaries. It also allows them to ask faculty to carry light enough teaching loads that permits success in both teaching and research. Finally having already completed their primary and secondary education at exceedingly good schools, students come far readier for the rigors of higher education. To have completed one's undergraduate career at a highly ranked college or university is more often than not symptomatic of an exceedingly advantaged background rather than having achieved some thing spectacular through hard work, perseverence or exceptional intellectual gifts. For "lesser" institutions to climb the summit means drawing students from such backgrounds. The likelihood of their being able to attract a critical mass of students from such advantaged backgrounds to materially change their fortunes is small at best. Especially for small liberal arts colleges -modelled after top liberal arts colleges from the east coast- founded over a hundred years ago by immigrant communities throughout the U.S. the chances are small. The alternative is to encourage their faculties to develop a more national profile (and, via slick marketing to make these acheivements known to a wider national audience.) in the hopes of raising the heavily weighted peer review in thge USNWR rankings. The fact is that the top ranked schools, thriving in a very different environment, are unlikely to pay much attention. (Check out this story from NPR's Morning Edition September 15 2006, "Marketing Higher Education Gets Sophisticated" ) And as the following illustrates, deans and presidents of competing institutions are unlikely to be swayed by suchinformation no matter how impressive it might actually be. Say that you have two Deans of two colleges, A and B, who are asked by USNWR to rank one another. A can given B a rank of one (a low ranking), a five (a high ranking) or give B a rank of three. B has the same set of choices as far as A is concerned. The deans of colleges A and B may dismiss out of hand the notion of giving their competitor a rank of one. Both Dean reconginize that truly miraculous things are happening at the other college and so briefly consider giving the other college a ranking of 5. What stops either from doing so is the following. Consider college A. If its Dean gives B the ranking it deserves and B does not reciprocate but instead gives A a ranking of 3 (the rank it received in last year's USNWR) then A will drop relative to B in the rankings. Of course if B gives A a rank of 5 then, relative to B, A's position is unchanged. A family whose student is considering both colleges A and B will note that A and B are both highly ranked if the latter happens and that A has a much lower rank if the former happens. Regardless of what B does then, the worst thing that can happen to A if its Dean gives B a ranking of 3 is that its relative ranking will remain unchanged. However , if A's Dean gives college B the ranking that she feels it truly deserves and B does not return the favor the worst thing that can happen is for college A's relative position to fall 2 points. Now, a student who might have considered college A or B will see a ranking of 3 for A and of 5 for B and choose B. It follows from this that both Deans will choose to award a rank of 3 to the other college. In fact, if you look at peer rankings for liberal arts colleges in last year's USNWR -outside the top fifty schools- peer rankings do indeed cluster around three. The end result is a tremendous amount of effort and time are expended to change a situation that look at the problem strategically is unlikely to change. And the peer ranking is important. It is given a great deal of weight in USNWR's overall ranking. In frustration, colleges respond by putting more resources into encouraging students and faculty to perform specatacular feats as well as trying to tweak other numbers that would also help push the overall rankings higher. The most infamous such strategy has been the practice of sending rejection letters to students who appear to have included a given school as a back-up school. (See "Glass Floor: Colleges Reject Top Applicants,Accepting Only the Students Likely to Enroll" by Daniel Golden, WALL STREET JOURNAL, May 29, 2001.) When students either failed to make the cut at their preferred school but were also turned down by the "lesser" school or worse still, they were accepted at their preferred school but turned down by the "lesser" school it became readily apparent that admissions policy was in part being driven by the desire to lower that college's acceptance rate, and, to a degree, bolster its USNWR rankings. Passing over students with perhaps greater potential but with lower ACT/SAT scores (fn1), putting money into campaigns to boost alumni contributions and failing to accept students probably heading elsewhere anyway all have been done in an effort to boost rankings. Silly as these things are, it is quite impossible for a college to unilaterally opt out. To do so would be suicidal. That is what is so heartening about the agreement reached earlier this summer by the Annapolis group. As with all such agreements the risk of defection is high. However, the only way out of such a bad situation is through such a cooperative agreement. While I am sympathetic to the notion that families and their students need help in making the decision about where they are going to go to school. It is not an easy choice. What Vasaar College's President Catherine Bond Hill suggest as an alternative to USNWR's rankings is an online database where students and families would have access to data and could construct school rankings whose value would be based on weights that they choose (see "A Better Way to Rank Colleges" .) One of the key things one should understand to their maerrow after graduating from any institution of higher learning that is worth its salt is that there is seldom one universally valid answer to important questions. Perhaps that lesson should start with the college selection process itself. fn1: See "Paving a Path for Poor Students' College Dreams" where reporter Wendy Kaufman notes that tuition has continued to rise and financial aid and scholarship money has not even more schools seeeling to boost rankings "...are emphasizing schoalrships based on high grades and test scores, not financial need")