Teaching and learning, labor and fairness

It seems a requirement that any conversation about higher education in America must begin and end with costs and economic outcomes. Along the way, our economic analysts nod to the power of knowledge (economic research shows it improves career prospects!), or the value of an educated citizenry (it produces more innovation, which of course produces more jobs, or at least more Steve Jobs’). Their implications are both personal and national: completing a college degree will get you a job, and a better, more effective and efficient higher education system will produce more jobs for us all. Unfortunately, as many new college graduates, and our country’s unemployed are now discovering, this view of the value and power of higher education is simply not true. Learning is not improved by a more efficient labor model, and inequality and injustice in our labor market are not solved with more learning. Learning is a complicated mess of students and teachers, texts and technology, content and curricula, bodies and brains. Learning is not a labor issue.

The economic mode of analysis typically begins with a blank slate. Dylan Matthews titles his ten-part series on the economics of higher education “The Tuition is too Damn High.”  In his last post on possible solutions, he addresses the relationship between cost and quality:

But it raises a troubling point: we, at the present moment, have literally no idea how good different higher education institutions are. We don’t know anything about which are better at imparting given bodies of knowledge, which are better at getting their students paying jobs, which are better at producing voters and soldiers and other contributors to civic life, or any number of other outcomes.

Because a simple and clear relationship between cost and quality doesn’t exist, it is assumed we have no idea what quality is.  Because we can’t measure quality with the same simple metrics by which we measure money, it is assumed that we have no idea what learning is. To understand the relationship between learning and money, we apparently need to measure them each by a single number, or a weighted average of a common set of metrics.

When we do find a number, we can be sure it is a bad one. Robert Gordon in the New York Times begins his lamentation about the sorry state of education in this country

For most of American history, parents could expect that their children would, on average, be much better educated than they were. But that is no longer true. This development has serious consequences for the economy.

Gordon proceeds to offer as evidence the mediocre performance of American schoolchildren on the tests from the Program for International Student Assessment and an allusion to colleges which have “longstanding problems with quality.”

A recent study by economists (and including the President of Northwestern University, also an economist) of the differences in learning outcomes compared tenure track to non-tenure track teachers. The provocatively titled “Are Tenure Track Professors Better Teachers?” asked what the consequences to student learning of a labor model in which some full time workers have tenure and some do not (despite often having long-standing relationships with the college). They found that students who had untenured faculty in introductory courses were more likely to take a more advanced course, and performed better in these courses than their counterparts who had introductory courses taught by tenured or tenure-track faculty.

Jordan Weissman, in his summary at the Atlantic notes:

Previous studies have suggested that colleges tend to hurt their graduation rates by hiring more part-time and non-tenure faculty. But Shapiro and his team wanted to measure the impact of tenure on “genuine student learning,” a notoriously tricky task.

The assessment of learning was through grades in those second courses. Figlio, the lead author on the paper, told Dan Berrett of the Chronicle of Higher Education:

Northwestern uses a four-point scale for grade-point av­er­ages, which Mr. Figlio said is a bet­ter proxy for learn­ing than stu­dent-sat­is­fac­tion sur­veys or standardized tests. “It’s not per­fect,” he said, “but frank­ly it’s the only thing I can think of.”

Berrett closes his mis-titled piece “Adjuncts make better teachers” (mistitled because the comparison group were not part-time, temporary adjuncts but rather full time lecturers) with the following quote from the paper

“Per­haps,” they wrote, “the grow­ing prac­tice of hir­ing a combina­tion of re­search-in­ten­sive ten­ure-track fac­ul­ty members and teach­ing-in­ten­sive lec­tur­ers may be an ef­fi­cient and edu­ca­tion­al­ly pos­i­tive so­lu­tion to a re­search uni­ver­si­ty’s mul­titask­ing prob­lem.”

Here you have it. A higher education labor model which splits research and teaching entirely may be efficient and educationally positive.

These economists and economic reporters treat education as an enormous, noisy and wasteful machine. We hear talk of “inputs,” “levers” and “outputs.”  We feed money into this beast and out come college graduates, ready to grow our economy. If only we could figure out how to adjust these levers, they muse, so that we had to feed less money in, and the same amount of learning would come out. Wouldn’t that be great? What if we could shift down this “labor costs” lever and see what happen to the learning outcomes?

And here is where I must strenuously object. College is not a machine (it’s made out of people! Bowling Green is made out of people!). Learning is not a labor issue.

What do I mean by this? The labor costs lever is only weakly connected to the factors that do change student learning. While education is immensely complicated, listing the important elements of a successful educational experience is not that difficult. Ask any group of college students and teachers, they’ll tell you. Motivated and prepared students. A coherent and structured curriculum. Interesting content and activities at the right cognitive level. A class size where the student feels attended to. Engaging pedagogical style. A trained and expert teacher with resources and support.

Can we affect learning by fiddling with the labor model lever? Of course, but it would be mediated through these factors. It would be cheaper to make all college classes 400 students, that will constrain the attention that can be devoted to any one of them. It would constrain the assignments and feedback offered. What if professors’ entire job consisted of teaching one class of 20 students? That would open pedagogical doors and leave more time for teacher preparation and activity design.

But ultimately, we must admit that most labor practices alone, good or bad, have a weak correlation with learning outcomes. Adjuncts and part time labor can and do provide great learning experiences. Tenured experts can provide awful ones. Charismatic lecturers can inspire in 300 person lectures, and a disorganized and disillusioned grouch can create an awkward and useless active learning discussion.  The reason is not their labor status, but whatever effect this labor status has on the other important elements of the learning situation.

The econometric study of “Are Tenured Professors Better Teachers?” would be better titled “Are Tenured Professors (at a wealthy, selective private school with motivated, traditional-aged college students, in large introductory courses,  with lecture style pedagogy) Better Teachers (than long term lecturers at the same college, who were far better paid than most adjuncts, who had taught at the college for longer than six quarters).”

But how could these lecturers be better teachers than their tenure-track counterparts? Were they selected to be better teachers whereas the tenure track were selected based on scholarship? Were they pressured to teach well, believing that their jobs were dependent on student learning and satisfaction? If they taught better, what did they do that was better? IF whatever they are doing is improving student learning outcomes, why can’t we ask tenure track professors to do that as well? This is the question of someone who is interested in learning and not merely cost efficiency. 

The modified title isn’t quite as catchy, but it illustrates how many important variables one must control for to reach a comparison of learning outcomes based on labor categories. This isn’t merely ignoring the elephants in the classroom, this is a statistically sophisticated elbowing past elephants to reach a small anteroom of teacher characteristics and incentives. The difference between these groups may be statistically significant, but it is not meaningful for understanding student learning.

As one of the people inside this machine, I can tell these macroeconomic observers that the inner operations are neither simple gears nor complex magic but something in between. Something human. While we are all rightly confused about the connection between labor practices and human learning, this doesn’t mean that practitioners have no idea how learning works. Fostering genuine student learning isn’t just “a notoriously tricky task,” but our life’s work. This isn’t simply “imparting given bodies of knowledge” but filling the pail and lighting the fire (to bend a phrase from Yeats). And are we concerned with providing economists with nationally comparable measures of student learning outcomes? Mostly not.

So if changing labor practices are not going to change student learning outcomes, why pay attention to labor practices at all? Because not merely learning outcome efficiency, but a national code of ethics should inform our labor practices. We’ve outlawed child labor not because it is inefficient, but because it is exploitative. We should consider the same when we see that higher education is now 76% part time labor. For all those statistics that economists such as Gordon begin with the importance of education, what about the importance of health benefits and employment stability? A labor force that is 90% part time might just fight to “produce” as much student learning as the current one (76% part time) in the coarse ways that economists measure student learning. But that wouldn’t make it right.

So here’s my prescription for an improvement in our national dialogue on education. Reporters, do you care about student learning? Then report on the things that actually matter for student learning. Learn a bit about pedagogy. What is the curriculum? What is the content? Read some of the people who care about student outcomes beyond how much they make, at the National Institute of Learning Outcomes Assessment.  Consider the research and guidelines of the American Association of Colleges and Universities in their Liberal Education and America’s Promise (LEAP) Program, in which they offer a vision of learning (and its measurement) that goes beyond standardized testing and beyond the only thing that Mr. Figlio can think of.

And to my fellow higher education faculty, tenured or not: don’t take the bait. Don’t criticize the details of a new study of the impact of teacher pay or tenure on student learning. A labor model is not a learning model. The best thing that a labor model can do is take care of food, clothing, security and health, and let teachers and students focus on learning. The issues of labor in higher education are issues of fairness, not efficiency.