The Perfect and the Free

Michelle Obama's misleading read on school segregation trends

Last week, in a public address commemorating the 60th anniversary of the Supreme Court's Brown v. Board decision, Michelle Obama warned that "today, by some measures, our schools are as segregated as they were back when Dr. [Martin Luther] King gave his final speech."

This is a stark statement -- made even more stark by the fact that virtually no school integration occurred between the Brown decision in 1954 and Dr. King's assassination in Spring 1968.  Southern communities operating separate school districts for whites and blacks were in no hurry to merge them. And even once merged, residential segregation -- which reached its highest point at mid-century -- implied that a reliance on neighborhood schools would perpetuate segregation.  It would take a series of further court cases, beginning in 1968, to make integration happen

Can it really be true, then, that America's public schools are as segregated now as they were at a point where the word "busing" had not even entered our vocabulary?  It all boils down to the definition of the word "segregation."

Consult your favorite dictionary and the term segregation will most likely incorporate the word "separation."  By this definition, schools are segregated when students of different races attend separate schools.

By this definition, schools are in fact much less segregated today than they were in 1968.  The series of court decisions that began in 1968 dramatically reduced the degree to which white and black students attended separate schools.  In more recent years, courts have reversed those decisions in some cases.  At the same time, though, America's neighborhoods have become significantly more integrated than they once were.  These two countervailing trends have more or less canceled one another out.  This is the conclusion of a comprehensive review released by Sean Reardon and Ann Owens last fall.  Students of different races are much less likely to attend separate schools than they were in 1968.

This explains Michelle Obama's use of the term "by some measures."  She's right that there are some measures employed by school segregation researchers that establish different trends.  But they do not define segregation as the tendency for students to attend separate schools.  Instead, they define segregation as the tendency for black students to attend schools with a high concentration of nonwhite students.

Back in 1968, attending a majority-nonwhite school was exceptional.  Nationwide, white students accounted for nearly 4 out of every 5 public school attendees.  In most parts of the country, "nonwhite" was a euphemism for "black."  The fact that over three-quarters of black students attended a school where the majority of their classmates were black served as a stark indicator of efforts to keep students of different races in separate schools.

Things are very different today.  For one thing, the public school population has changed.  Only 52% of public school students are white.  For another, "nonwhite" no longer means "black."  Attending a majority-nonwhite school is no longer all that exceptional.  The National Center for Education Statistics projects that within 10 years the public school population will be majority nonwhite.  So schools could be 100% integrated -- every school a microcosm of the nation -- and every school would be majority nonwhite.  Segregation measures based on the two different concepts -- separation versus exposure to nonwhite majorities -- would yield exactly opposite conclusions.

In summary, "some measures" of segregation are highly misleading because they mistake demographic trends for actual decisions to send students of different races to different schools.  Stephan and Abigail Thernstrom pointed this out a week ago.

Now, to be fair, there is definitely resegregation going on in some parts of the country.  In districts that stop busing, there is a definite uptick in the tendency for students of different races to attend different schools.  High-poverty schools often struggle to attract and retain high-quality teachers.  But we no longer live in a world where it is possible to judge a school by the color of students' skin.  Majority-nonwhite schools are the emerging norm.  Many of these schools -- not enough, but many of them -- do an excellent job of educating children.  My own children attend a public school where over 70% of their classmates are nonwhite.  It's a great school.  We need to focus our efforts on the schools that aren't succeeding, which is not the same thing as the schools serving nonwhite students.

Have a job? Thank an immigrant.

Do immigrants take jobs away from Americans?

Imagine a world where there is a finite amount of work to be done, regardless of the size of the population.  In this world, it's pretty easy to say that any increase in population reduces the amount of work per capita.  If society hoped to keep its residents fully employed, it would do whatever it could to limit population growth.  Restricting immigration is the easy way to do it, but nations that have sought to limit their population have resorted to much more nefarious policies than that.

We don't live in a world, or more specifically a nation, where there is a finite amount of work to be done.  

First off, the amount of work to be done in most industries depends directly on the size of the population.  The education and health sectors now account for nearly one-quarter of the jobs in the American economy, and demand for education and health rises directly with population.  Construction jobs (6% of the workforce) depend in large part on how many new houses we need to build.  Retail trade jobs (11% of the workforce) depend on the number of customers available.  Transportation jobs (5% of the workforce) depend on how many people and things need transporting.  Even much of the manufacturing industry (10% of the workforce, down from 16% two decades ago) ties directly to population size.  There are more American jobs in food and beverage manufacturing than there are in the motor vehicle, aircraft, ship building, furniture, apparel, steel, farm machinery, computer and peripheral, electrical equipment, and household appliance manufacturing industries combined.

Your standard economics-101 view of the labor market, where the demand curve slopes down and thus any infusion of new workers lowers wages, doesn't tell the whole story.  Immigrants increase both the supply of labor and the demand for labor.  Think about it this way.  Would your job be more or less secure if the population of your community suddenly plunged by 13%?  That's the difference between having and lacking immigrants in the typical American community.

The second key point to consider is that when immigrants don't take jobs, there is no guarantee that the job will be made available to a native instead.  In the manufacturing industry and elsewhere, employers have moved millions of jobs abroad in the past few decades.  Had it not been for immigration, they would have moved even more.  Adding an extra 1,000 immigrants to the economy keeps 46 manufacturing jobs here in the United States.

The third reason immigrants don't take jobs from Americans is that they are disproportionately likely to create their own.  And when immigrants start their own business, they often end up creating more jobs for others.

The key statistic to understanding how immigration benefits native employment is to consider the impact of immigration on native population.  If immigrants took American jobs, we'd see a pattern whereby an inflow of immigrants to a community led natives to leave in response.  In fact, we see the opposite pattern.  For every 1,000 immigrants who enter a county, the native population of that county increases by about 270.  If you're looking to see where opportunities are in the United States today, find a place where immigrants are.

Much ado about zero

Tomorrow, the school board governing North Carolina's largest school district will consider a proposal to prohibit teachers from awarding the grade of zero to students who fail to turn in work.  Some advocates hope to convince the board that the minimum grade on any assignment -- even for a student who fails to turn in the assignment -- should be fifty.

The argument against zero goes something like this.  Students receive one and either become discouraged or rationally compute that it is mathematically impossible to recover from the grade to pass the course and thus give up, increasing their risk of dropping out and other adverse outcomes.

The pro-zero side points out the perverse nature of handing out half-credit for no work. Students might rationally conclude that there isn't much point in doing the work.  So they might pass more courses, but won't necessarily learn any more, and some -- those motivated by the fear of a zero in the first place -- might learn less. 

What's the "right" thing to do?  That depends, as it turns out, on what you think schools are for in the first place.  

The root of the problem, as pointed out by independent consultant Ken O'Connor, is the separation of grading standards from learning.  There are good intentions behind the separation.  If you were educated in American schools, you most likely have heard some version of the phrase "if you hand in your homework, it is basically impossible to fail this course."  That's a reassuring statement for students who might otherwise be intimidated by the subject matter.  But it leads directly to the debate over zeros: when it is impossible to fail except by neglecting to hand in your work, the only students failing will be those who neglect to hand in work.

Effort-based grading leads to another problem: students can pass courses without mastering the subject matter.  The consequences of this practice in the K-12 system can be seen in the higher education system.  The majority of North Carolina high school graduates entering the state's two-year community colleges need to retake high-school level coursework in at least one subject before they can proceed to courses that earn college credit.

The fundamental question underlying the entire debate is as follows: under what circumstances should students fail?  One can think of this as a multiple choice question:

a) Students should fail when they do not demonstrate mastery of course subject matter.  This is the old-school definition of failure.

b) Students should fail when they do not turn in assignments.  This is a kinder, gentler policy, forgiving those who try hard but do not master the subject matter.  In a world where persistence matters more than knowledge, this policy makes more sense.

c) Students should never (or perhaps just rarely) fail.  This policy makes the most sense if you believe that staying in school matters the most, regardless of what you may or may not learn while there.

So, the question of when students should fail in turn boils down to a basic question about the true function of education.  Are we endowing students with knowledge that will make them happier and more productive?  Are we reinforcing "non-cognitive" skills -- showing up, putting in a good faith effort?  Or are we primarily interested in handing out diplomas, based on the notion that having the sheepskin makes the greatest difference?

The cognitive skill vs. non-cognitive skill vs. sheepskin debate is surprisingly unsettled in the academic literature.  So don't be surprised to see debates about zeros continue on for years to come.