Compared to What?: An argument for a conservative approach
Recently, I watched an interesting youtube video of economist Thomas Sowell giving a talk about public schools. In response to a question about what to do over the idea that standardized testing may not be objective and may be biased, Sowell said this:
Compared to what? … That’s the question economists always ask: “Compared to what?”… Nothing is easier than to prove that something human has imperfections. I’m amazed at how many people devote themselves to that task.
This very simple statement – it almost seems like common sense! – is quite hard for many to grasp. Whether the issue is standardized testing (whether to do away with it), or any other percieved injustice of society, many people’s reaction is to point out the flaw and use this as prima facie evidence that the system giving rise to it must be fised, abolished, or reconstructed.
Sowell’s point – one I share – is that pointing out a flaw in a thing is not the only step necessary toward arguing against the thing. The next ste – one not often taken up – is to argue that a concrete proposal for a solution will be better than and have fewer flaws than the system trying to be replaced.
As far as standardized testing goes, it is often said that the tests produce biased results, “teaching to the test” or some other undesired result. This by itself is often taken as an argument against standardized testing. But: compared to what? If we abolished standardized testing, will we not have equal or worse problems of having no adequate measure of student ability, school performance, and academic progress?
The way I see it, removing standardized testing has just as many downfalls as keeping it and the question is whether the downsides (“teaching to the test,” ranking students unidimensionally) are worth the benefits (accountability, measuring the progress of students, teachers, and schools). The problem is that, like many others, the issue tends to be seen less as a “trade off” of ill for good effects, but as an all or nothing issue: either we find the perfect way or we abolish the imperfect one.
As a PhD student, I have the good fortune to talk often with radicals, and one recently told me that they were “anti-institution.” They pointed out that institutions often stifle creativity (by locking themselves into pre-established patterns), perpetuate injustices in the name of procedures, and are generally ranked hierarchically.
Here is another example of the “all or nothing” view, rather than seeing the issue as one of “trade offs.” Where I, a believier that insitutions allow for optimal functionality in society, am fully prepared to admit that institutions sometimes do the things mentioned by my friend, the radical, I believe it can be argued that institution’s benefits (stability, efficiency of production, continuity) are well worth these downsides.
Now, none of this is to say that asking “Comapred to what?” should be used to justify opposition to change or advocacy of leaving things alone per se. I would be completely glad to see any undue bias eliminated from standardized testing so that the results are better and fairer. I would be happy also to see rules put in place that made it difficult for institutions to be corrupt. But what always has to be recognized when contemplating these things is that the cure cannot be any worse than the disesase and our excitement for a cure must not lead us to discount this idea. More directly, we can’t let our zeal for stamping out injustice allow us to remake things in a way that may end up worse than what we are trying to erase.
In the case of standardized testing, we would have to make sure that we can put some method of measurement in place that gives us all the current benefits of standardized testing without the argued downsides. (Authentic project-based assessments are often argued, but these are liable to be every bit more biased than standardized testing, as they are scored very subjectively.) In the case of designing rules to curb institutional corruption, it must always be remembered the regulation intending to curb corruption often promote and hide corruption more efficiently than no rules (because it gives companies incentives to collude with or bribe the regulators, and regulators can provide good smokescreens).
Anyhow, you get the idea. So, before we decide to undo instutitions or undo standardized testing (or any other protocol or structure where injustice may result), we should ask Dr. Sowell’s question. If we are tempted to say: “this system doesn’t work,” we should follow with “compared to what?” If we can’t answer the second part, then maybe we shouldn’t be so quick to proclaim that the first part is any sort of argument for abolition.