Caleb Anderson: Education Fallacies

A number of core assumptions significantly impact policy design in education, and elsewhere, and it is helpful to know what these are.

Following the example of Thomas Sowell, I have labelled these assumptions fallacies. Awareness of these fallacies can help us to better deconstruct some of the assumptions that shape education policy design, and which often trap unwary educators, and others, because of their appeal to social justice. 

These ideas are not entirely new ideas, and nor are they confined to education. What should surprise us is that these assumptions persist, in spite of clear evidence that they often make things worse. These ideas include the concept that education policy changes are a zero-sum game, that changes made to benefit one group will ultimately benefit all, and that you can play with the rules of the game (i.e. that you can experiment with educational policy, almost endlessly) without doing damage.  

In my view, these fallacies lie at the very heart of why things are getting worse, and they also explain why change is often made with reckless disregard to prior experience, and simple common sense. I will now endeavour to say why I have called these assumptions fallacies.

1.  The concept of a zero-sum game assumes that shifting resources (or investment) from one part of the system to another has no significant negative effect on the efficacy of the wider system. Thus the zero-sum assumption maintains that overall outcomes will remain broadly similar when investment moves from areas of comparatively good performance, to areas of comparatively poor performance, with longer-term benefits accruing from a more equitable dispersion of opportunity. In reality, this often proves not to be the case. Systems are generally much more complex than is realised, and there are frequently unanticipated downsides when policymakers play with things. People, institutions, and ideas, add or subtract, value in myriad ways from every system. Mechanisms of added value are very difficult to identify, anticipate or replicate, and they are embedded in complex feedback loops. For example, costly initiatives to meet the needs of underachieving children are assumed to automatically yield benefits for all children, and for society at large. While such initiatives have obvious appeal, there is little evidence to support this hypothesis. Often huge investments in education, frequently enabled by disinvestment elsewhere, show no gains whatsoever for target groups, and contribute to a decline in educational standards overall. 

2. Equally fallacious is the idea that policy initiatives are always “worth a shot”, even when there is little prior evidence that they will work. The reluctance of policymakers to trial these initiatives first, to examine what has happened in other jurisdictions, to stage implementation, and to study and to tweak where necessary, tends to suggest that motivations are often more philosophically grounded than evidentially based. Initiatives to improve outcomes in underachieving cohorts are often very short on delivery, as evident in the tendency to keep adjusting, and fudging, the actual outcomes themselves to cover obvious policy failures. 

3. The third fallacy follows from the first two, that you can play with things without any significant risk of damage to the wider system. In reality, the zero-sum approach is anything but zero-sum, and invariably results in reduced motivation, innovation, creativity, and productivity within the wider system. 

By way of example, early last year the media were falling over themselves to report on "findings" that educational streaming was racist, and should be abandoned without delay.  While it might well be argued that streaming contributes to differential educational outcomes, which is why I assume it has been labelled racist, it is far from clear what the overall results of abolishing streaming would be for the wider system, and for educational achievement as a whole.  We also do not know how significant streaming is alongside a myriad of other factors influencing educational outcomes.  For example, in the United States data seems to show that absent fathers are a much greater determinant of educational underachievement (and life success) than streaming, and it seems reasonable to assume that this may be the case in New Zealand also, and this is just to name one variable.  Correlation and causality are not the same thing. Because streaming exists alongside poor educational outcomes, does not make it the cause of these outcomes, much less the primary cause.  

That the motivation for the abandonment of streaming may be more philosophical than empirical is evident in just how quickly the argument was broadened to include differential grouping of students in primary and intermediate schools.  Differential grouping allows teachers to manage multiple groups of learners concurrently, ensuring that students get material and instruction commensurate with their ability or readiness.  Differentiated instruction was one of the reasons why New Zealand was once near the top of the OECD educationally, and why teachers from other countries came to study the New Zealand education system.  Now, this, too, is considered racist, and schools are being invited, if not strongly encouraged, to dispense with differential groupings in favour of whole-class or mixed ability groupings.  I know of no multivariate, multidisciplinary and peer-reviewed research which would support such a move.  The fact that a growing number of schools are abandoning streaming and differentiated grouping, and are enjoying the experience, doesn't cut the evidential mustard.  That we can throw away a proven educational model with such abandon is a concern.

Tragically, policy failure often lags the implementation of the policy by such a large margin that those who implemented these policies never come to see, and doggedly deny, that these changes were a bad idea in the first place, often saying that the problem was that they were implemented too early or too late, too fast or too slow, or that an adjustment here or there might have made the difference. No one is accountable for the outcomes of these policies.  There seems little that educators, wise to what is really going on, can do in the face of a plethora of damaging policy initiatives, other than to brace themselves for the inevitable, and to hold on to what they know from experience works.   Until we are brave enough to face the real causes of underachievement, we are in for more of the same and the highest price will be paid by those at the bottom.

And the solution... while systems are complex, these systems are generally able to adjust themselves if they are left alone, a process called homeostasis.  Not all of the philosophical and ideological justifications in the world can compensate for a lack of clear empirical evidence. Governments can be bought, and policy design in New Zealand is often reckless, and hopelessly ideological.  Poor outcomes are inevitable when policy-making is done this way, and we seem incapable of learning from our mistakes.  Our obsession with ideological causes, in the absence of clear supporting (multivariate - and multidisciplinary) evidence, and our willingness to sacrifice the needs of higher achievers in order to equalize educational outcomes, guarantee the progressive erosion of educational standards... if you cannot lift achievement at the bottom, then lower it at the top.  The deleterious effect of this on higher achieving students, on education at large, and its ultimate effect on our economy, are considered worthy sacrifices if greater social cohesion is the end result. The fact that it makes us all materially poorer seems of little consequence.  Social cohesion remains elusive due to systemic denial of the real causes of social breakdown and dysfunction.

Caleb Anderson, a graduate history, economics, psychotherapy and theology, has been an educator for over thirty years, twenty as a school principal.