This week I’ve been reading Crossing the Finish Line: Completing College at America's Public Universities, Princeton University Press's hot new book about public university graduation rates, and here is my first impression: We should all take pity on the poor economists who mine databases for a living, because theirs is truly a difficult and thankless job.
First, there’s the problem of the raw data, which never come in the desired form. So step one is to sanitize the numbers, reconciling differences among multiple sources, correcting for errors and omissions, imputing numbers to fill in gaps and missing pieces, and making reasoned guesses about ambiguities, all the while keeping meticulous records about procedures and methodologies, and knowing full well that every footnote at the bottom of every table is grist for the mill for future generations of Ph.D. students anxious to pick apart what you’ve done.
Next comes an even harder step, which is deciding what questions to ask, in this case all bearing on why some students graduate from college while others don’t. How difficult it must be to wander around in a multidimensional space of a near-endless number of variables, associations, correlations, causes and effects, and then try to focus and distill the questions to the handful that the data actually could answer. Not only difficult but frustrating. Perhaps students who don’t graduate from college tend to have mothers who abused drugs while pregnant, or were spanked as infants, or merely parked in front of the TV. Sorry, the data don’t say anything about these possibilities.
Step three is to crunch the numbers to extract their hidden messages, and this must be high-anxiety time for the researchers. Because now, several years down the road, after all the averages are computed, standard deviations plotted, correlation coefficients and linear regressions charted; after the results are standardized and normalized, probabilities calculated, error bars bracketed on the data points; after all this mind-numbing, exhausting, tedious and painstaking research has been completed, there is always the possibility that the data won’t tell you anything that you couldn't have guessed at the outset.That after all this work, you’ve ended up mostly confirming the obvious.
And even if the data do reveal some surprises, one still isn’t out of the woods. After all, this is human society we’re talking about, not Higgs bosons or quarks, and so while the research may be dispassionate and rational, that isn’t necessarily true of the human beings who read about it. And to the extent the messages extracted from the data conflict with their politics, passions, prejudices and predilections, readers will likely conjure all kinds of reasons to reject them. (Remember the furor over The Bell Curve?)
For example, consider the authors’ finding (Chapter 5) that college students are significantly more likely to graduate if they enroll in the hardest schools that will admit them. This is fact, drawn from the data, not conjecture, and it mostly parallels the findings of a prior study by other scholars. But it is a socially touchy issue (which the authors handle with great delicacy) because good students from disadvantaged backgrounds often are unable to attend the colleges that would otherwise be best for them. Already, charges of “elitism” are cropping up because the authors observed that “undermatching” is harmful to a student’s future, and that highly selective universities do a better job at challenging smart students than do less selective schools. Like I said, let us all pity the poor economists.
What else are we learning from Crossing the Finish Line? Clearly, some of the findings are what one might have anticipated: that students who attended rigorous high schools graduate from college more often than those who didn’t (Figure 5.1); that top-ranked flagship public universities have higher graduation rates than their less-selective public peers (Figure 3.1a); that students from affluent families graduate more often than students from modest backgrounds (Figure 2.4a); and that high school GPAs are good predictors of graduation rates (Figure 6.1).
But even if some of these findings are not exactly shockers, just having them backed by hard data can be an important tool in rebutting unfounded opinions and avoiding pointless and unproductive debate.
I was pleased, for example, to have the authors comment on public university tuition pricing (page 190): “An obvious but important first proposition, demonstrated again in this study, is that lower net prices matter much more to students from low- and moderate-income backgrounds than to others… As economists have argued for a number of years, it is far better from this point of view [for universities] to charge a higher price and use some of the revenue to discount the price more heavily to needy students through need-based grants.” This is music to my ears, since the concept of differential tuition pricing was the basis for the Miami University tuition plan, while I was president there, and the underlying concept behind my proposals in Saving Alma Mater.
But the real value of Crossing the Finish Line will come later, after we’ve stopped skimming the book's surface and burrowed down into the 131 figures and 121 tables and 71 pages of endnotes in order to extract the second-order effects. This will obviously take some time, but that is most likely where we’ll learn how to dispel common misconceptions, reject public policies based on faulty reasoning, expose legislative assumptions that aren’t warranted, and do way with ineffective government programs. (Do you think Pell Grants have boosted enrollments for low-income students? Think again. The data say no. )
Books of this sort always suffer from TMI. They’re ponderous, hard to read, and filled with qualifications and technicalities. (“Coefficients indicate the predicted difference in graduation probability (from a probit model) between the reference group and the listed group (low-SES students), holding all control variables at their respective means.”) Unfortunately, that’s just the way good research is done and there’s no way around it.
To their credit, though, the authors (William G. Bowen, Matthew M. Chingos & Michael S. McPherson) have done their best to make the forest visible through the trees. They’ve pushed the details into appendices (some of which are on-line), and they’ve provided a useful readers’ guide to their core findings. The chapters are more-or-less self-contained, each with a helpful summary, so readers won’t have to wade through material they’re not interested in before getting to the good stuff. And one particularly helpful section is an excellent historical overview by Eugene M. Tobin: The Modern Evolution of America’s Flagship Universities. To me, that essay alone is reason enough to buy the book. Crossing the Finish Line is an important resource that scholars and policy-makers will be talking about for a long time.