How is a poor voter to know?

Naftali Bennett is an Israeli politician who was the minister of education in the years 2015-2019. With another round of elections in Israel in the offing (the 4th round in two years), Israeli voters will soon have to decide if they are impressed with Bennett’s past achievements (and with his trust-worthiness).

During his tenure as minister of education, Bennett emphasized the importance of math studies and introduced a program which, he claimed, would allocate resources toward improving math education in Israel. Recently, Bennett was quoted as saying the following regarding the newly published results of the TIMSS international math test for 2019:

These are tremendous news and a tremendous achievement for Israeli math students. Israeli students leapt from the 16th place [in the 2015 TIMSS test] to the 9th place in their math ranking. The disparities were reduced, and these are the international tests, not national tests. The proportion of excellent students is 15%, 3 times the average of the rest of the world.

Gil Gartel, a commentator on matters of education on the Sicha Mekomit website, is not impressed. In fact, he claims that each and every one of Bennett’s assertions is factually wrong.

Reading the study report reveals that Bennett’s tenure as minister of education contributed nothing to the achievements of Israeli students as far as those are reflected in the TIMSS international math test.

The those who took the test were not a sample of all Israeli students, achievements did not show a marked improvement, those who were tested did not rank in the 9th place in the world, the disparities were not reduced and (statistically speaking) not a single excellent student was added in the socio-economic disadvantaged groups.

Gartel claims the test sample was manipulated to exclude the least successful students (making the results unrepresentative), and that the countries which participated in the 2019 test were not the same ones that participated in 2015 (making the ranking comparison meaningless), and in any case were certainly not all the countries in the world (making comparisons to “the rest of the world” baseless). He also purports to show that disparities between groups were not reduced. Finally, Gartel quotes the report as flatly stating that:

In Israel, and in 15 other states, no substantial change has occurred in students’ achievements between 2015 and 2019.

Since Bennett has focused his attention on the issue of math education, it seems that deciding between these diametrically opposite views of the outcomes Bennett’s tenure would be an important item in the public discourse. And yet, only a handful of stories about the TIMSS results appear in Israeli media – most of those simply conveying Bennett’s self-congratulating comments.

It seems voters are expected to do their own homework.

9 Responses

  1. A good example, Yoram. Such sad fluffiness about impact is all too frequent.

    A political decision (irrespective whether by an executive within his/her responsibility or by a citizen jury) should always publish a consensus prediction about the various expected measurable outcomes / evidence of their policy decision (or rejection). Evidently, this requires a prior methodic judgement debate and agreement, sadly lacking or even imposed by the organisers.

    One of the purposes why I developed Prediki is to allow falsifiable policy decisions. The dilemma: political players simply shy rigorous accountabilitiy and prefer its pretense through electoralism.


  2. Hubertus,

    I don’t think that it is realistic to have public policy (or science for that matter) managed through a sequence of predictions and observations of whether previously stated predictions were met or not.


  3. Yoram,

    Not sure why you would say such a thing. Science is separable from the metaphysical and nonsensical precisely by this method: observation of evidence and falsifiable predictions.

    For policies, it a key problem that the vast majority of people are not sufficiently aware that their (own or regurgitated) ideas which sound good on paper could be flawed, actually most likely will be. That’s just how we end up implementing ill-conceived policies, even entire political systems, where whole countries decline into despair and misery over just one or two decades.

    I say that rigorous falsifiability of policies will solve this huge problem: before deciding a policy, let’s agree on falsifiable predictions about intended impact (and acceptable tradeoff) and check back. Note how this solves your story above, of Bennet working with unclear (not ex-ante defined) criteria: by tracking pre-agreed ex-ante predictions, any “poor voter” would know.

    Yes, it is an extra step, but it is well worth the small effort, it is actually indispensable, just like the shift to evidence-based medicine when Helmont suggested to cast lots to split treatment groups and observe the outcome, defined as funerals.


  4. > Not sure why you would say such a thing. Science is separable from the metaphysical and nonsensical precisely by this method: observation of evidence and falsifiable predictions.

    I think (and I believe this is the standard view among philosophers of science) that this is a very simplistic view of how knowledge is acquired or how is should be acquired.


  5. Yoram:> I think (and I believe this is the standard view among philosophers of science) that this is a very simplistic view of how knowledge is acquired or how is should be acquired.

    Presumably this is a reference to the Popper/Kuhn/Lakatos argument that took place during the 1960s. Since that time (I had the misfortune to study it as a sociology undergrad, 1969-72) it’s true to say that philosophy of science has disappeared up its own postmodern backside, along with the rest of the social sciences. John Horgan has an interesting piece on the topic:


  6. Presumably, Yoram’s lack of arguments is just that: a lack of arguments.

    @Keith> On the linked interview: astonishing that people still conflate falsification with falsifiability, not to mention the (metaphysical) reference to some religious “scientific consensus“.


  7. Hubertus,

    > lack of arguments

    Yes – these are pretty deep waters and while it is a rather interesting subject, I don’t think it is relevant enough to the matter of democracy to merit diving deeply into it in the current context.


  8. A simplistic view of democracy apparently precludes thinking about one of its key still missing ingredients..


  9. I agree with Simon that individual policies are falsifiable in the sense that they can be evaluated ex post as to whether they achieve specified outcomes (and that those who make the proposals will be judged on the basis of past predictions). But I also agree with Yoram that epistemic considerations are orthogonal to democracy (but then I don’t think the latter has anything to do with outcomes). Democratic governance means that the people (or a descriptively representative subset) rules, even if this means we all go to hell in a handcart.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: