Here's one from Daniel that lands right in the middle of a contradiction that's been getting sharper by the month. New OECD adult skills data just dropped — the PIAAC survey — and Israel's literacy and numeracy scores sit below the OECD average. Not slightly below. And yet this is the same country with more Nobel laureates per capita than anywhere else on earth, and a tech sector that produces something like half of all exports. Those two facts don't sit together easily.
They really don't. And the timing makes it worse — the Taub Center just put out their latest reports, and the primary school metrics are deteriorating, not improving. Teacher shortages, declining instructional hours, compensation that keeps falling further behind. So the disconnect between the elite output and the average output isn't closing — it's widening.
Which is the puzzle Daniel's asking us to actually make sense of. Because on the surface, you'd look at these numbers and conclude the system is failing. But then you look at the startups, the cyber units, the Nobel roster, and you think — something is clearly working. The question is: working for whom, and for how long?
And that's what we're going to trace today — not whether the system is good or bad in some blanket sense, but how it actually functions. Because once you look at the structure, the paradox starts to resolve itself. It's not one system producing contradictory outcomes. It's multiple systems operating under one label, and the averages are hiding everything interesting.
Let's get into it. What does the data actually show?
The headline numbers first. Israel's PIAAC literacy score sits around two hundred fifty-five — that's below the OECD average of two sixty-seven, and it puts Israel in the bottom third of participating countries. Numeracy is a similar story. We're talking about adults who, on average, struggle with tasks like evaluating the credibility of a news article or calculating a percentage discount.
Yet Israel's high-tech sector is eighteen percent of GDP. Half of exports. That's not a rounding error — that's the engine of the economy. So either the PIAAC data is wrong, or something stranger is happening.
The data isn't wrong. What's happening is that Israel doesn't have one education system — it has four, and they produce radically different outcomes. The secular Jewish stream, the religious Zionist stream, the Arab and Druze stream, and the ultra-Orthodox Charedi stream. They're funded differently, they teach different curricula, and their students show up in the same national average as though they're all sitting in the same classroom.
Which means the average is a fiction. It's like averaging the heights of basketball players and jockeys and concluding the country has a height problem.
The secular stream performs near the OECD average — not stellar, but within striking distance. The Charedi stream, which is now about sixteen percent of all students and growing fast, teaches almost no secular subjects after a certain age. No math, no science, no English. So when those students take the PIAAC assessment as adults, they cluster at literacy level one or below. That drags the national average down dramatically.
The paradox starts to dissolve once you stop treating Israel as a single data point. The Nobel laureates and the Unit 8200 cyber recruits aren't emerging from the same educational experience as the adults who can't parse a bus schedule. They're coming from different streams entirely.
That brings us to the core question Daniel's really asking. Is this a system failure, or is it system design? Because from one angle, you could say the system is working exactly as intended — it produces a small, highly capable elite while maintaining political compromises that fund religious schools without demanding secular accountability.
The military-tech filter completes the picture. Unit 8200 and similar programs don't just recruit from the existing pool — they effectively run a parallel education system. They screen for raw cognitive ability, then provide three years of intensive technical training that can rival a computer science degree. That pipeline catches talent that the primary system might have left underdeveloped.
Which is how you get a country with below-average literacy scores and world-beating cybersecurity output. The military isn't fixing the schools — it's routing around them.
Here's where it gets uncomfortable. The military filter works — brilliantly, in some ways — but it's a filter, not a fix. It selects maybe a thousand to two thousand recruits per year for the elite technical units. Israel's school system has about a hundred sixty thousand students per grade cohort. So we're talking about the top one to two percent getting a world-class technical education, while the rest depend on whatever their primary school managed to deliver.
What their primary school can deliver is increasingly constrained by who's willing to teach. The Taub Center numbers on teacher pay are genuinely stark. Starting salary for an Israeli teacher is around twenty-five thousand dollars a year. The OECD average is closer to thirty-eight thousand. And it gets worse with seniority — the gap widens, not narrows. A teacher with fifteen years of experience in Israel earns about thirty percent less than their OECD counterpart, adjusted for GDP per capita.
Which means the talent pool for teaching is shrinking, especially in STEM subjects. If you're a physics graduate in Israel, you can join a startup at a starting salary of maybe sixty, seventy thousand dollars, or you can teach high school physics for twenty-five. That's not a choice — it's a filter that pushes talent away from classrooms. The Taub Center has been warning about this for years, and the effect shows up in PISA scores with a lag of about five to ten years.
The primary school decline we're seeing now reflects teacher compensation decisions made a decade ago. And the decisions we're making now will show up in the data around twenty thirty to twenty thirty-five. It's a slow-motion problem, which is exactly the kind societies are worst at fixing.
Then there's the Charedi dimension, which is the most politically sensitive part of this whole picture. Charedi schools receive state funding — substantial funding — but they teach minimal secular subjects. By eighth grade, many Charedi boys are studying almost exclusively religious texts. No algebra, no biology, no English composition. When they reach adulthood and take the PIAAC assessment, they score at literacy level one or below. That's not a bug in the measurement — it's an accurate reflection of what the curriculum prioritizes.
This isn't some accidental oversight. It's the result of explicit political compromises that go back decades. The Charedi parties have made school autonomy a non-negotiable condition of coalition participation. So the state funds schools that don't teach the subjects the state then tests adults on. The low average score isn't a failure of the system — it's a feature of the deal that keeps the system politically viable.
And when you strip out the Charedi cohort from the national average, the secular Jewish stream looks very different — close to OECD norms, sometimes above. The Arab and Druze stream sits somewhere in the middle, with its own distinct challenges around resource allocation and language barriers. But the point is, you can't understand any of these numbers without knowing which stream you're looking at.
The military produces better STEM outcomes than the school system for a pretty straightforward reason — it's not trying to educate everyone. It's selecting for aptitude, then concentrating resources on a tiny cohort with intense, practical training. A school system can't operate that way. It has to serve the entire population, including populations whose educational priorities are fundamentally different from the state's economic needs.
That tension is only going to intensify. The Charedi stream is growing at about four percent per year, compared to roughly one and a half percent for the secular Jewish population. At current rates, Charedi students will be about a quarter of all primary enrollment by twenty thirty-five. The national average scores are going to keep declining, not because anything is getting worse in the secular schools, but because the composition of the average is changing.
Which brings us back to Daniel's question about whether these facts actually make sense together. They do — once you see the system as four parallel tracks with different goals, different funding, and different accountability structures. The paradox was never in the outcomes. It was in treating the average as meaningful.
Here's where the averages stop being an academic puzzle and start being a forecast. The bifurcation we just described has a trajectory, and the numbers are moving in one direction. The Charedi population grows at about four percent a year. The secular Jewish population, roughly one and a half. By twenty thirty, Charedi students will be around twenty percent of primary enrollment. By twenty thirty-five, closer to twenty-five percent. That means the absolute number of students in the secular stream shrinks.
Which means the pool that Unit 8200 and the tech sector draw from shrinks with it. The military filter can maintain its per-capita selection rate — the top one or two percent — but one percent of a smaller number is a smaller number. The pipeline doesn't need to get worse at screening to produce fewer elite graduates. It just needs the denominator to contract.
That's already happening. The PISA 2022 data — that's the most recent full assessment — showed Israel's math score at four fifty-eight, which is fourteen points below the 2018 result and below the OECD average of four seventy-two. Reading dropped ten points to four seventy-four, just under the OECD average of four seventy-six. These aren't gentle declines. They're the steepest drops Israel has recorded.
Fourteen points in one cycle. That's the cohort that hits the workforce around twenty thirty to twenty thirty-five. The military filter can't compensate for a shrinking base of adequately prepared students, because it's not a remedial program. It assumes a certain floor.
Unit 8200's training doesn't start from zero. They're screening for kids who already have strong quantitative reasoning, some programming exposure, English proficiency. If the primary system produces fewer students who meet that bar, the unit either lowers standards or accepts fewer recruits. Either way, the output shrinks.
This connects to something Daniel's prompt gestures at that most commentary misses. The Nobel Prize thing. Israel's Nobel per capita rate is extraordinary — about twelve per ten million people. But those prizes reflect research done twenty to thirty years ago. The current crop of Nobel-caliber scientists were educated in the nineteen eighties and nineties, before the Charedi demographic shift accelerated, before teacher salaries fell off the cliff, and when the secular stream was proportionally much larger.
That's the lagging indicator problem. We're looking at Nobel Prizes awarded now and feeling reassured about the education system. But the system that produced those laureates doesn't exist anymore. The composition of the student population has changed, the teacher talent pool has thinned, and the political deals that fund schools without secular requirements have deepened. The Nobel count is telling us about the past, not the present, and definitely not the future.
What does this mean for a parent on the ground right now, especially one who can't afford the twelve hundred dollars a year per child that the average Israeli family now spends on private tutoring? Because that number is real — Israel has one of the highest private tutoring expenditure rates in the entire OECD.
It's a tough spot. The system works reasonably well for motivated families in the secular stream who can supplement with enrichment — extra math, English tutoring, coding camps. But for families without those resources, the gap is widening. And it's not just about getting into Unit 8200. It's about basic preparedness for a workforce that increasingly demands quantitative literacy. A child who falls behind in math in fourth grade, in a school with underpaid teachers and large classes, doesn't just miss out on a tech career. They miss out on a huge range of middle-class professions.
The tutoring expenditure number tells you that parents already know this. They're not waiting for the system to fix itself. They're routing around it individually. But that only works if you have twelve hundred dollars per kid per year. Below that line, you're dependent on whatever the school can deliver.
Which is where the trajectory really bites. If teacher salaries stay thirty percent below OECD peers, the quality of instruction keeps declining, and the families who can afford to supplement keep doing so, you get a system that sorts children by parental income more efficiently than it educates them. The military filter at age eighteen is too late to undo the gaps that open up in third grade.
The paradox resolves into something bleaker than it first appeared. The system isn't contradictory — it's coherent. It produces elite outcomes for a shrinking slice of the population while average outcomes decline, and the Nobel Prizes give everyone a comforting story to tell while the numbers underneath keep deteriorating.
What does a parent actually do with this? Daniel's asking the practical question — if you don't have the tutoring budget, if you're watching these numbers and feeling stuck, what moves can you actually make next week?
The single highest-leverage thing, and the data on this is remarkably consistent, is early literacy. PIAAC research shows that literacy gaps compound — a student who struggles with reading in third grade is about four times more likely to drop out later. It's not just about books. It's about whether a child can independently access every other subject.
Because if you can't read fluently by fourth grade, you're not learning history or science from a textbook. You're falling behind in everything, simultaneously.
And the good news is, there are free interventions that actually work. The National Library of Israel runs digital literacy programs — bilingual, accessible online, no cost. Most municipal libraries have reading clubs and summer programs that cost nothing. The Ministry of Education's "Sifriyat Pijama" program distributes free books to young children. These aren't substitutes for good schooling, but they can close a meaningful portion of the gap.
The first move is: get your kid reading early and reading a lot, and use the free infrastructure that already exists. It's not glamorous, but the effect size is real.
Second thing, and this one is less obvious. The IDF screening tests — the Kodkod and the psychometric batteries — they're trainable. The IDF itself publishes free practice materials on its recruitment website. Some NGOs, like the "Atidim" program, offer free test preparation for students from peripheral communities. Starting preparation in ninth or tenth grade, rather than scrambling in twelfth, makes a significant difference.
This matters because the military pipeline isn't just for future cyber warriors. Strong Kodkod scores open doors to a much wider range of technical roles — intelligence, logistics, communications — that translate directly into civilian careers. Even if Unit 8200 isn't the goal, the screening system is a gateway.
And here's the thing parents often miss: these tests don't measure what your school taught you. They measure raw reasoning, pattern recognition, quantitative intuition. A kid from a weak school who's been practicing these skills independently can outperform a kid from a strong school who's never seen the format. The filter is imperfect, but it's more meritocratic than the school system it routes around.
The second move is: treat the IDF screening as a target you can prepare for, starting in ninth grade, using free materials. It's a backdoor into the elite pipeline that doesn't depend on your zip code or your school's budget.
Third — and this shifts from individual to collective — local advocacy actually works in Israel's system. The Taub Center's core recommendation is a twenty percent increase in teacher salaries to attract STEM graduates back into classrooms. That's a national policy fight. But municipalities have some autonomy. They can supplement teacher pay from local budgets. They can fund classroom assistants. They can invest in school infrastructure.
Parents can push for this. School board meetings, municipal council sessions, parent committees — these are levers that exist. Most parents don't use them because they feel like the system is too big to move. But at the municipal level, a few dozen organized parents can shift budget priorities.
The numbers bear this out. Municipalities that supplement teacher pay — places like Herzliya, Ra'anana, parts of Tel Aviv — have lower teacher turnover and better student outcomes, even controlling for socioeconomic factors. It's not magic. It's just paying people enough to stay.
Move three: show up locally. Push your municipality to supplement teacher salaries. It's less satisfying than a national reform, but it's actually achievable, and the effects compound.
Fourth, and this is the one that requires the most legwork from parents, but it's worth knowing about. Not all secular schools are the same. The Mofet network — it's a network of science-oriented schools within the secular stream — produces PISA-equivalent scores fifty to eighty points above the national average, even with similar student demographics. That's not a small edge. That's the difference between below-OECD and well-above.
The Ministry of Education publishes school-level performance data. It's not always easy to find, but it exists. Parents can look up how specific schools perform on the Meitzav tests — those are the national assessments — and compare options within their district. Most parents don't know this data is public.
They really don't. And there are other models emerging. Some secular schools are experimenting with project-based learning, partnerships with tech companies, integrated coding curricula starting in elementary school. The "Branco Weiss" network, the "Dror" schools — these are alternatives within the public system. They're not everywhere, but where they exist, they're worth investigating.
The fourth move is: research school-specific performance data, look for Mofet or similar programs in your area, and treat school choice within the public system as something you can actually exercise with information rather than rumor.
None of these four things fixes the systemic problem. The demographic trajectory, the teacher pay crisis, the political compromises — those require national-level decisions that parents can't make alone. But these moves can meaningfully change the trajectory for an individual child, and they don't require a private tutoring budget.
Even with those individual strategies, the systemic question remains. And this is where the forward look gets uncomfortable. The military-tech sector isn't waiting for the education ministry to fix things. Google's Campus for Startups, Microsoft's CyberSpark in Beersheba, the IDF's own expanded training pipelines — these are private and military workarounds that essentially say, we'll train the talent ourselves because the schools won't deliver it.
Which works, for now, for a few thousand students a year. But a few thousand isn't a workforce. Israel's tech sector employs something like three hundred fifty thousand people. You can't backfill that through corporate training programs when the primary pipeline is shrinking. At some point, the workarounds stop scaling.
That's the open question Daniel's prompt really forces. Can this system reform itself before the demographic math makes the problem irreversible? The Charedi population growth isn't slowing. The teacher salary gap isn't closing. The political incentives that created the four-stream system haven't changed. Nothing in the trajectory suggests a correction is coming.
Which means the country that built its entire economic identity on human capital — we don't have oil, we don't have vast natural resources, we have brains — may face a talent shortage within ten to fifteen years. Not because Israelis got less talented. Because the system stopped developing the talent it has.
You can already see the early responses. The tech sector is increasingly lobbying for expanded work visas, importing engineers from Eastern Europe and India. That's a rational short-term fix, but it's an admission that the domestic pipeline isn't sufficient. A country that once exported brainpower is starting to import it.
The real paradox isn't the one we started with. It's not that Israel produces Nobel winners despite weak schools. It's that the system works brilliantly for a shrinking few while quietly failing the growing many, and the Nobel trophies give everyone a reason to look away from the numbers that are catching up.
The numbers don't care about reputation. They just compound.
Now: Hilbert's daily fun fact.
Hilbert: The word "halo" entered astronomical vocabulary in the early Renaissance, borrowed from the Greek "halōs" meaning threshing floor — a circular disk where oxen trampled grain. The optical ring around the sun or moon was named for this agricultural shape, not the other way around. In French Guiana, the phenomenon is known locally as "couronne de pluie" — rain crown — because its appearance reliably precedes tropical rainfall within twelve hours.
A threshing floor in the sky. That's actually quite beautiful.
That one almost made sense.
This has been My Weird Prompts. Thanks to our producer Hilbert Flumingtop. If you enjoyed this episode, leave us a review wherever you listen — it helps more people find the show. We'll be back next week.
