The Myth of Achievement Tests, by James Heckman, John Eric Humphries, and Tim Kautz is more than a meticulous work of social science. It also is an objective, but hard-hitting, analysis of the testing that has driven school reform, and a warning about the unintended harm done by ill-conceived policies.

One of the best things about the book is its scholarly objectivity. Its problem is its excessively scholarly language. A book this important should be read by everyone, not just those who will give it the multiple close readings that its prose requires.

As I have explained, there are reasons to worry that the scholarship of Heckman et. al will be misappropriated by some school reformers. So, let me quickly dispense with my two concerns. I question the general policy recommendation on page 434 that begins, “1. Character skills should be measured and integrated into school accountability programs.” I see a powerful argument for measuring character skills for diagnostic purposes but not for accountability. But, that is their recommendation, their opinion on policy, not a problem with their impeccable research.

Secondly, I wish their wording had been less clinical. It is a jolt to repeatedly read sentences such as, “As a group, GEDs are about as smart as high school graduates but lack the character skills of graduates.”

On the surface, The Myth of Achievement Tests is about the General Education Development (GED) but it has implications about many or most of the main school improvement policy issues. For instance, it also documents the unintended negative effects of testing and other reform policies. Heckman et. al explain, “This book shows that this faith in tests deceives students and policy makers and conceals major social problems.”

“With its promise of a cheap route to high school equivalency,” Heckman et. al conclude, “the GED induces students to drop out of high school.” The GED allows society to deny the sad facts about dropout rates. If we count students who pass the GED as high school graduates, the dropout rate declined dramatically between 1968 and 2004. If we count them as dropouts, the dropout rate increased during that time.

If a GED was actually the equivalent of a high school diploma, we could justifiably pat ourselves on our collective backs. But, after controlling for male students’ skills, GED recipients barely produce better life outcomes than dropouts. Female GED graduates do somewhat better. But, by creating the illusion that a GED is comparable in real world value to a high school diploma, we set students up for making bad decisions.

We also deceive society as a whole. If we count GEDs as graduates, the black- white racial gap in school completion has virtually disappeared. If we don’t count GEDs as diplomas, the gap has barely decreased.

The origins of the GED, it could be argued, were innocent enough. The test was an easy way to award high school diplomas to World War II veterans. During their service, veterans had learned plenty in the schools of hard knocks about character skills. It probably would have been better to award degrees based on that experience. Suggesting that a test made the process of granting diplomas more valid opened the door to self-deception. And, the cost of pretending that a test added real value became greater as degrees were awarded primarily to non-veterans.

It is hard to pretend that passing a test is the equivalent of the learning gained through attending high school unless the test represents a doable but real challenge. During the 1960s, the average age of GED-takers was 29 years-old or more. This was consistent with giving a second chance for adults. When the test became a normative option for struggling teens, it became an especially attractive shortcut.

A typical high school student attends class for 1000 hours a year. The turning point came around 1970, when the percentage of GED-takers who were 19 years old or younger started to increase significantly. By 1989, few GEDs were being awarded to adults who left school for military service, and the average study time for the majority of those who passed the test was about forty hours.

Heckman et. al explain the direct harm to students of that deception. Polling data shows 40% of dropouts said that earning GED would be easier than graduating from high school. This was second to chronic absenteeism as part of the explanation of why they failed to earn a diploma.

Then, policy-makers made it even easier to be deceived by the GED option. Oregon, for instance, offered GED preparation in schools and the easier access to the tests contributed to a 40% increase in the dropout rate.

Heckman et. al then explain, “The GED program is a symptom of a deeper problem … failing to produce necessary character skills. It is possible to tackle this problem, but not simply by raising standards on achievement tests.”

Some will disagree, but I am convinced by Heckman that the GED came of age as post-World War II high schools started to “create an adolescent society with values distinct from those of the larger society and removed from the workplace.” Previously, when half of teens did not attend high school, they were “supervised and mentored by adults, often as apprentices.”

Worse, with the relaxation of school discipline during recent decades, it is more difficult to teach students the importance of cooperation and perseverance, much less showing up for work. And, that brings us to the most destructive aspect of the GED and focusing on other tests. Success in life is not primarily determined by cognitive skills, but by so-called “soft skills.”

GED-takers, when tested on Self Esteem, Locus of Control, and Self-concept, are between high school graduates and dropouts. In other words, they and others who fail to graduate need the help of adult mentors who could “scaffold” a “discipline protocol.” It would have been more difficult, perhaps, for school to help students develop the character skills they need for a happy, healthy life, but it would have been much more worthwhile.

In explaining the need for schools to tackle the socio-emotional skills that are necessary for success after school, Heckman et. al make a statement that may be the single best indictment of test-driven school reform. They remind us, “Research in neuroscience, psychology, and economics shows that character skills are more malleable in adolescence than are cognitive skills.”

As mentioned earlier, some readers (like me) may love Heckman et. al’s analysis of the science but they might be expecting a let-down as they reach their policy recommendations. After all, their emphasis on “grit” must be seen in the context of the methods of KIPP and “No Excuses” schools. But, their proposed solutions are humane and nurturing. One set of solutions draws upon the successes of apprenticeship programs. Workplace-based adolescent programs are especially promising.They recommend programs like “Tools of Mind” that teach the incremental theory of intelligence. Heckman et. al seek programs where young people are taught that their capabilities are malleable.

As far as concrete policy suggestions, none of them seem controversial to me. Heckman et. al conclude, “America is a second chance society. The GED program should be retained as a second-chance option ….” But GED graduates should not be seen as high school graduates in social statistics. Also, GED prep centers should not be allowed in high schools.

More generally, Heckman et. al say that we must “recognize that problems with the achievement test can’t be fixed by making it more cognitively challenging.” Instead of test-driven school improvement efforts, they call for programs to support families and parenting. They seek “secular programs to foster moral and character education in schools.”

In other words, the evidence, analysis, and recommendations of Heckman et. al are mostly inconsistent with the contemporary test-driven reform movement. They are very consistent with the humane policies recommended by the Broader Bolder Approach, and “the Resistance” or we teachers who oppose corporate reform. And, because these scholars are so scrupulously objective, their work can be invaluable in rolling back test-driven reform.