Testing mandates flunk cost-benefit analysis

By Valerie Strauss

This was written by Peter Smagorinsky, Distinguished Research Professor of English Education at The University of Georgia.

By Peter Smagorinsky

According to Wikipedia, cost–benefit analysis “is a systematic process for calculating and comparing benefits and costs of a project, decision or government policy (hereafter, ‘project’). CBA has two purposes:

1.To determine if it is a sound investment/decision (justification/feasibility),

2.To provide a basis for comparing projects. It involves comparing the total expected cost of each option against the total expected benefits, to see whether the benefits outweigh the costs, and by how much.”

I believe that it would be prudent to apply this process to the current accountability movement now being administered in public education, primarily in the form of testing mandates such as No Child Left Behindand Race To The Top. Although I am not an economist — I’m an old high school English teacher now engaged in teacher education at the university level — I believe that I understand the issues at stake as well as anyone currently employed in the U.S. Department of Education.

First, let’s consider the costs. In Texas, taxpayers will pay about $93 million this year to administer standardized tests to Texas students . . . or nearly ten times the cost of just nine years earlier. The Georgia state Department of Education pays McGraw-Hill about $11 million a year to produce the CRCT, and more than $5.4 million to NCS Pearson for the EOCT; and as listed here, these are just two of the many tests administered in my home state. The annual cost of standardized testingin the United States has been estimated at somewhere between $20 billion and $50 billion.

Some defenders of standardized testing maintain that critics exaggerate the costs in order to overstate the case against the accountability movement. Upon further consideration, I must wonder exactly what is computed to determine the costs of test administration and scoring: Teacher salaries during test preparation? The cost of #2 pencils? Operating a school building on Saturdays to accommodate testing? And so on. With that caution, I’ll accept only the lowest estimate, a mere $20 billion price tag, and proceed from that assumption. Now, what are we getting for our $20 billion?

* Some of the lowest teacher morale survey responses ever recorded.

* Massive, pervasive cheating on tests by administrators, and by the teachers bullied into enforcing their pressure for high scores.

* Six-figure bonuses for school superintendents whose scores meet a minimum standard, no matter what it takes to achieve them.

* Federal penalties for schools caught cheating, denying them essential operating budgets.

* Students who see their role models on the faculty and in the administration behave in unethical ways in order to artificially raise test scores by cheating.

* Curriculum and instruction that focus on fragmented knowledge bitsand avoid time-consuming, process-oriented, insight-driven teaching and learning.

* Immense profits for textbook publishers who have entered the competition for designing and administering the tests and writing the curriculum materials that are aligned with the tested content.

* An institutionalized assumption that poverty is not a factor in student achievement, in spite of considerable evidence to the contrary.

Based on this cost-benefit analysis, I conclude that the accountability is making education very profitable for a limited number of publishing companies who, it turns out, have invested heavily in political connections (e.g., McGraw-Hill and the Bush family). It also is quite lucrative for school superintendents who don’t get caught cheating, although it can cost them their jobs if they do get caught. Apparently, their own cost-benefit analysis of cheating typically leads to high-risk behaviors with high-stakes consequences.

Other groups pay for the accountability, and not just through the taxes they pay to raise that $20 billion annually to support the testing apparatus. Teachers increasingly dislike their jobs and consider their work environments to be hostile and depressing. According to Richard Ingersoll, 40 percent of new teachers nationwide bolt the profession within five years because of the terrible working conditions; and a new report by the Education Trust identifies the “culture” of school—the work conditions — to be the top priority in a satisfying teaching career, particularly in high-poverty schools. The primary motive for entering a teaching career and staying in it, then, has been sacrificed to the accountability movement.

Students as well do not benefit from this approach. Rather, their education is reduced to an endless series of assessments of their ability to fill in bubble sheets, at the expense of more extended thinking such as writing or composing other sorts of texts that they find useful and meaningful.

I can only conclude that the $20 billion annually spent on testing as a means of educational accountability is a poor expenditure of our tax dollars. I would now like to propose a different means of assessment that I believe has greater validity as an educational measurement, and will produce a more satisfying teaching environment that is more likely to keep the best teachers in the profession. Unfortunately, it has the downside of failing to enrich superintendents and publishing companies, but I am willing to live with that unfortunate consequence.

But first, let’s keep that $20 billion budget available. Here’s how I would reinvest it, beginning with some suggestions for addressing the needs of children more than the needs of publishing companies and other wealthy entities cashing in on the new accountability mandates:

* Provide a good nursing staff, particularly in impoverished areas, so that kids who live in poverty can undertake their studies with a reasonable degree of health and balance. Sick, dizzy, aching, itching, wounded, and distracted children with limited access to health care or guidance in navigating the health system would benefit from the immediate care of a qualified health professional.

* Expand free and reduced-price meals for children from homes where fresh food is not available, and work to improve the healthiness of the food offerings under these services. Kids who haven’t eaten are very difficult to teach effectively.

* Staff school libraries with knowledgeable, helpful media specialistswho can direct students to books that benefit their reading and educational development.

In general, invest in school infrastructures so that they are in good operating order, rather than falling apart at the seams.

Note that I am not calling for increased teacher salaries, although that would sure be nice. I am proceeding according to the assumption that for many teachers, good work conditions matter more than a high salary. So I am starting there, assuming that my $20 billion can only go so far.

My last suggestion concerns assessment. To be blunt: Standardized tests are a really stupid way to measure learning. Hardly anyone involved with education finds them to be valid; they are mostly believed to be worthwhile expenditures of time and money by people who have never taught. People like Arne Duncan and Bill & Melinda Gates.

I recommend instead that assessment proceed more authentically. Linda Darling-Hammond, Jacqueline Ancess, and Beverly Falk described this approach in the mid-1990s, and their ideas still resonate today — perhaps now more than ever. In their view, schools could institute such assessments as comprehensive, interdisciplinary projects through which students embody what they have learned during their studies. A boy might build a set of cabinets, for example, incorporating mathematics, physics, chemistry, kinesthetics, drawing, writing, speaking, and other knowledge and skills in order to design, build, polish, and then explain the project and what it involves. This proposal for project-based learning and assessment is taken up in some schools, such as Simon Hauger’s “Sustainability Workshop” in Philadelphia, where kids build solar charging stations, full-sized electric vehicles, and other machines to demonstrate their knowledge. Hauger, I should note, accomplishes these remarkable feats with kids from West Philly, not from The Main Line.

Now, some might wonder, how can this plan work, when it relies on such complicated projects and means of assessment? What about the elegant simplicity of a nice, firm test score? Doesn’t a quantitative test score tell more about a kid’s mathematical knowledge than his ability to measure a cabinet door so that it fits the frame? And what about the broader community? How will they know how this kid stacks up against another cabinet builder from Milwaukee? What if their cabinets both work equally well, albeit for different purposes—perhaps one to store DVDs, the other to display china? How will we know who won the Race To The Top under this approach?

Here’s where some of my $20 billion budget comes in handy. As part of a broader effort to increase internet capacity, some amount — let’s ballpark it at $2 billion annually, 10% of my budget, although I could be off by a few billion dollars — could be dedicated to expanding each school’s server space, or perhaps link each to a national database, so that each student’s work could be displayed. What would you rather be able to do: (1) learn that Freddi got a 79 on the CRT (and see if you can figure out what these scores even mean) or (2) go online and see a web demonstration of the new wardrobe that Freddi has designed, cut, sewn, and tailored along with a verbal account of the process she went through and the academic knowledge that she incorporated into the project?

So, there you have it, one person’s view of a cost-effective way to invest $20 billion in the necessity of educational assessment. It’s a bit more complicated that what we’ve presently got, just like learning and life in general. It puts money into classrooms and school infrastructures, instead of in the bank accounts of book publishers and the politicians they influence with contributions. It requires more work of the taxpayer in seeing and understanding educational outcomes, but the products are multidimensional and real, rather than paper-thin and abstract. And that new technological infrastructure could probably serve a few additional beneficial purposes for school districts beyond the immediate and designated purpose of publishing assessment results. I’m thinking here of one of my neighboring counties, where the computers available to teachers still use Windows 3.0 for their operating system.

I offer this proposal entirely for free, unlike the situation in states like Colorado where 35% of their federal education money is paid to consultants. You are free to take it or leave it. But whatever you do, you can’t say that it cost you too much.