Daniel sent us this one, and I want to read the key finding straight off the top because it stopped me cold. A Jerusalem Post survey, thirteen hundred Israelis, and seventy percent of them believe the ceasefire reflects a US concession to Iran. Two thirds oppose it. That alone is striking. But the detail that really grabbed me is this: the survey was conducted at two points in time, close together, and between those two points there was a forty percent shift in sentiment. Not a gradual drift. Within a short window. So the question Daniel is raising is: what does that tell us about trust, not just in Israel but everywhere? How does Israel compare to other democracies? Is this kind of volatility a sign of something structural, or are we watching a political cycle play out? And what's the internet and social media doing to all of it?
That forty percent swing is the number I keep coming back to. Because forty percent is not noise. That is not margin of error. That is people who held one view, and then, in a matter of days or weeks, held the opposite view. Or at minimum, people who were undecided tipping hard in one direction. Either interpretation is significant.
The sample is not tiny. Thirteen hundred respondents, a thousand Jews, two hundred and twenty-eight Arabs. That's a methodologically serious survey. This isn't a Twitter poll.
Right, and the two-point design is actually really important here. A lot of opinion surveys give you a snapshot. This one gives you a before and after, which is much harder to dismiss. You can argue about what caused the shift, but you cannot argue the shift didn't happen.
By the way, today's script is powered by Claude Sonnet four point six.
Feels appropriate that an AI is helping us process a story about information environments and who people trust.
It does have a certain poetry to it. To understand why a forty percent swing is both shocking and, in a way, not surprising at all, we need to look at where Israel sits right now—and that means going back to its political foundations.
Israel spent roughly the first three decades of its existence as a Labor-dominated state. Ben-Gurion's party, Mapai, essentially ran the country from independence in 1948 until 1977. That's nearly thirty years of a single political bloc controlling government, the Histadrut labor federation, the major newspapers, and the cultural institutions. Trust in those institutions was partly organic and partly structural, because the institutions and the ruling party were the same thing.
Then Menachem Begin wins in seventy-seven, which is still called the Mahapach, the upheaval, and the whole frame shifts.
And what's interesting is that the right's rise wasn't just about security policy. A huge part of it was Mizrahi Jews, Sephardic Jews from Arab countries, who felt excluded and condescended to by the Ashkenazi Labor establishment. So distrust of institutions was already baked into the coalition that brought the right to power. It wasn't incidental.
Which means Israel's political right came to power on a wave of institutional skepticism and has, over time, become the institution. That's a tension worth sitting with.
It really is. And now you have a situation where the governing coalition includes parties that are themselves deeply suspicious of the judiciary, the press, the military brass, the intelligence services. The skeptics are running the place, and a significant portion of the public is skeptical of them in turn.
Trust in Israeli institutions is not declining from some pristine baseline. It's been contested terrain for a long time. The question is whether what we're seeing now is a new phase or an acceleration.
That question applies globally. The OECD data puts Israeli trust in government at around forty-five percent, which is mid-range. Above France and Italy, well below Switzerland and Finland. So Israel is not an outlier in the numbers, but the volatility, that forty percent swing, that's what sets it apart.
The number is average. The turbulence underneath it is not.
That gap between the average number and the turbulence underneath it is actually the most interesting thing about the Israeli case when you put it next to the comparative data. Because if you look at the Edelman Trust Barometer from this year, the headline finding is that seventy percent of respondents globally say they are hesitant to trust people who hold different values. That's not a fringe sentiment. That's a majority disposition toward distrust as a default posture.
Which reframes the question a little. We tend to ask why trust is declining. But maybe the prior question is whether high institutional trust was ever the natural state, or whether it was a specific historical condition that required specific circumstances to maintain.
That's exactly the right frame. The postwar period in Western democracies, roughly 1945 to the mid-seventies, was anomalous. You had high growth, low inequality, relatively consolidated media, and governments that had just won a war or built a welfare state. Those are not normal conditions. They're exceptional ones.
In Israel's case, you add to that an existential founding project where everyone was more or less rowing in the same direction, or at least the dominant narrative said they were.
And then that narrative fractures. The Mizrahi experience we talked about earlier is one fracture line. The religious-secular divide is another. The Arab minority, who are twenty percent of the population and were twenty-two percent of that survey sample, have their own entirely different relationship with Israeli institutions, which is its own enormous subject.
The sample actually mirrors the demographic split pretty closely. A thousand Jews, two hundred and twenty-eight Arabs, out of thirteen hundred total. That's roughly seventeen percent Arab representation, which is close to the actual proportion.
Which matters for reading the seventy percent figure. Because if that number holds across the Jewish and Arab subgroups, even roughly, that's a cross-communal finding. If it's driven entirely by one group, that tells a different story. The Jerusalem Post piece doesn't break it down that granularly, but it's worth flagging.
It's the question I'd ask the researchers first. Because the ceasefire framing, the idea that it represents a US concession to Iran, lands differently depending on where you sit politically and communally.
And this is where the conflicting narratives piece gets complicated. The government's framing of the ceasefire is one thing. The opposition's framing is another. The media's framing is a third. And the public's interpretation of all three of those is shaped by which sources they trust, which by now is itself a deeply partisan variable.
Pew has US trust in national news organizations at fifty-six percent overall, down twenty points since 2016. But when you split it by party, Republicans are at forty-four percent and Democrats are at sixty-nine. That's a twenty-five point gap on whether the news is even credible.
If you're seeing that kind of gap in the US, you can be fairly confident something structurally similar is happening in Israel, where the media landscape is at least as polarized and the political stakes feel even more immediate to people.
The forty percent swing in the survey could be a genuine shift in underlying opinion. Or it could be a measurement artifact where the two time points happened to catch different activation levels of different partisan groups. Both are plausible.
The two-point design makes the genuine shift interpretation more defensible, but it doesn't rule out the other reading. What I'd want to know is whether the swing was uniform across demographic groups or concentrated in one segment. A forty percent aggregate shift that's actually a ninety percent shift among one group and flat everywhere else tells you something very specific about which community is reacting to what.
Reacting to what is the key phrase. Because the timing of the survey matters enormously. If the second measurement came right after a specific news event, a statement from Washington, a development on the ground, you'd expect a spike.
Trust is event-sensitive in a way that people underestimate. The OECD data shows Israel at forty-five and a half percent trust in government, and that number is an average across time and context. But trust in government after a military success versus trust in government after a hostage deal that the public perceives as disadvantageous are not the same number, even if the underlying institutional reality hasn't changed at all.
The institution hasn't changed. The news cycle has.
Which gets to whether this is cyclical or structural. And I think the honest answer is it's both, operating on different timescales. The short-term volatility, the forty percent swing, that looks cyclical. Responsive to events, capable of reversing again. But the long-term trend line across the OECD, across Pew, across Edelman, that looks structural. Slow, persistent erosion that doesn't fully recover between cycles.
France at forty-four percent trust in government. Italy at forty-two. Those are not countries in acute crisis right now. Those are baseline numbers reflecting decades of accumulated skepticism about whether the state delivers what it promises.
Switzerland at eighty-two percent, Finland at seventy-five. The outliers on the high end are small, high-functioning states with strong transparency norms and relatively low corruption perception scores. The lesson there isn't just that good governance builds trust. It's that trust and governance quality are mutually reinforcing over long periods, and they're mutually undermining over long periods too. Once you lose the cycle, it's genuinely hard to restart.
Which is a slightly uncomfortable thing to sit with if you're thinking about Israel's trajectory. Not because forty-five percent is catastrophic, but because the direction of travel matters as much as the current position.
Israel has some specific structural pressures that most OECD comparators don't. The judicial overhaul debate, the Haredi exemption from military service, the ongoing conflict, the hostage situation. Each of those is independently capable of moving trust numbers, and they're all happening simultaneously.
Stacked crises tend to produce stacked skepticism. People can absorb one or two institutional failures. When the failures are concurrent and visible, the cognitive load pushes toward a general conclusion rather than a specific one.
The general conclusion being: I don't know what's true, and I'm not sure who to ask.
Which is exactly what that seventy percent finding is measuring, whether people know it or not. It's not just an opinion about the ceasefire. It's a statement about information confidence—how people perceive the reliability of what they're hearing.
Right, and that's the thing about information confidence as a concept. It's not binary. It's not that people think the government is lying. It's that they've lost the ability to calibrate. They don't know how much to discount what they're told, so they discount more than they probably should, which is a rational response to an unreliable signal.
The internet made that problem structurally worse in a specific way. Not because it introduced misinformation, which has always existed, but because it removed the bottleneck. Pre-internet, you had to get your misinformation through a limited number of channels that at least had reputational skin in the game. Now the distribution cost is zero and the verification cost is entirely on the consumer.
The consumer is not equipped for that. There's a body of research on what's sometimes called the liar's dividend, the idea that in an environment where fabricated content is plausible, even true information becomes harder to trust because people can always claim the real thing is fake. The existence of convincing forgeries devalues authentic currency.
Which means you don't need to actually deceive people to erode trust. You just need to make the information environment noisy enough that people opt out of trying to figure out what's true. Epistemic exhaustion is the goal, not conversion.
That framing fits what we're seeing in the data. The Edelman finding about seventy percent being hesitant to trust people with different values isn't really about values. It's about information. If I think your information sources are unreliable, I extend that skepticism to you. And if you think the same about me, we've both made a rational local decision that produces a globally irrational outcome.
The US case is almost a controlled experiment in this. Pew has federal government trust at seventeen percent overall, which is the lowest it's been since they started tracking in 1958. But the partisan split is the interesting part. Democrats at nine percent trust in the federal government, Republicans at twenty-six.
Which means both parties are operating in a low-trust environment, just at different floor levels. And notice that the number didn't collapse uniformly. It collapsed asymmetrically, tracking which party controls the executive branch.
Trust in institutions has become a proxy variable for partisan affiliation. You don't evaluate the institution on its merits. You evaluate it based on whether your team is running it.
That's been true for a while, but social media accelerated the mechanism. The algorithmic incentive is engagement, and the content that drives engagement is content that confirms what you already believe and makes you feel that the other side is not just wrong but dangerous. That's not a bug in the system. It's the business model.
The effect on institutional trust is direct. If your feed is curated to show you the worst version of every government action and the most alarming interpretation of every policy decision, your trust number is going to be lower than someone consuming the same events through a different filter. Not because one of you is right and the other is wrong, but because you're literally looking at different information objects.
The European data is interesting here because the trust collapse in Europe has followed a slightly different trajectory than the US. France at forty-four percent and Italy at forty-two are low, but they've been low for a while. What's changed in the last decade isn't the floor so much as the volatility. You see sharper swings, more event-sensitivity, more rapid reversals.
Like what we're seeing in Israel.
And the driver in Europe has been a combination of the 2008 financial crisis, which hit trust in economic institutions first and then spread to political institutions, and then the migration crisis from around 2015 onward, which created a very specific wedge between what official sources were saying and what people were observing in their immediate environment.
The credibility gap between official narrative and lived experience is probably the most corrosive specific mechanism. Because it's not abstract. It's not about some policy decision in a capital city. It's about whether what you see out your window matches what you're being told.
In Israel that gap is particularly acute because the stakes are not abstract. The hostage situation is not a policy debate. There are families who have direct, personal, visceral stakes in whether the government is telling them the truth about negotiations, about timelines, about what was traded and what was promised.
The seventy percent who believe the ceasefire reflects a US concession to Iran, that's not just a geopolitical opinion. For a significant portion of those people, it's a judgment about whether their government was honest with them about something that directly affected their family.
Which is why the forty percent swing is both meaningful and hard to interpret. It could reflect new information changing minds. It could reflect the event-sensitivity we talked about. But it could also reflect the fact that in a high-stakes, emotionally charged environment, trust is less like a considered judgment and more like a mood state. It shifts because the emotional context shifts.
Is that cyclical or structural, though? Because mood states cycle. But if the underlying emotional charge stays elevated for long enough, the baseline shifts permanently.
That's the crux of the cyclical versus structural question. And I think the evidence leans structural at this point, for a specific reason. In previous decades, trust tended to recover after crises. Watergate hit US trust hard, but it bounced back in the eighties. The recessions of the early nineties hit trust, but it recovered. What's different since roughly 2008, and accelerating through the social media era, is that the recoveries are getting smaller. Each cycle ends at a lower floor than the previous one.
A ratchet mechanism. You can lose trust quickly and recover some of it, but you never fully recover, and the next shock starts from a lower baseline.
Once you're below a certain threshold, the dynamic changes. At high trust levels, institutions get the benefit of the doubt. At low trust levels, the burden of proof reverses. Every action is interpreted through a skeptical lens by default, which means even good-faith actions get read as self-serving. That's very hard to escape.
Is Israel at that threshold?
I think it's close. Forty-five percent sounds like it's above the halfway mark, but that's an average. If you segment by political and communal affiliation, there are almost certainly subgroups well below thirty percent. And those subgroups are not passive. They're politically activated, they're on social media, and they're producing the content that the more moderate middle is consuming.
The tail wagging the dog, information-wise.
Which is the pattern in the US too. The seventeen percent federal trust figure is an aggregate, but the people driving the discourse on institutional trust, on both sides, are not in the middle of that distribution. They're at the extremes, and the platforms reward them for it.
The question of whether Israel is exceptional or just an extreme version of a global pattern, I think the answer is both, in different respects. The structural forces, the social media dynamics, the post-crisis ratchet effect, those are global. But Israel is running the experiment at higher temperature. Higher emotional stakes, more acute security context, more recent and visible institutional failures, and a media environment that is every bit as polarized as the American one.
Which matters because the information ecosystem is denser. Everyone knows someone who knows someone with a direct stake in the decisions being made. In a country of nine million people, the hostage crisis is not distant news. It's three degrees of separation at most.
Proximity collapses the abstraction. You can't maintain a detached analytical posture about institutional trust when the institution's decisions are affecting people you actually know.
That's probably the most important thing to understand about why Israeli trust numbers are so volatile relative to the OECD average. It's not that Israelis are uniquely cynical or uniquely credulous. It's that the signal-to-noise problem is operating on top of a much higher emotional baseline. Every piece of information carries more weight because the consequences are more immediate.
Which loops back to the alternative media question. If the mainstream sources feel unreliable, and the emotional stakes are high, people don't stop seeking information. They seek it harder, through channels that feel more honest or more aligned with what they already believe. And those channels are often less constrained by editorial standards precisely because they're operating outside the institutional framework.
The paradox being that the search for more reliable information often leads to less reliable information, but information that feels more reliable because it confirms your priors and speaks in a register that feels authentic rather than managed.
Authentic is doing a lot of work in the current information environment. Authenticity has become a stylistic marker rather than an epistemic one. Something can feel raw and unfiltered and still be completely wrong.
Institutions, almost by definition, communicate in managed registers. Press releases, official statements, carefully worded briefings. Which means that as the authentic register becomes more valued, institutional communication becomes more suspect, regardless of its accuracy.
You couldn't design a better trap for governments that are actually trying to communicate honestly—it’s a system that almost guarantees confusion or distrust.
If you’re someone who reads that Jerusalem Post survey and thinks, yes, that maps to something I’ve been feeling, and you want to engage more honestly with information rather than just less, where do you even start?
The first thing I'd say, and this sounds almost too simple, is to separate the question of what happened from the question of why it happened. Most institutional communication conflates those two things, and so does most alternative media, just in the opposite direction. The raw event and the interpretation of the event are different objects. Training yourself to hold them apart is useful.
The ceasefire is a good example. The fact of the agreement is one thing. The question of what it reflects about US-Iran dynamics, or Israeli negotiating leverage, is a different thing. Seventy percent of Israelis have a view on the second question, but how many of them could tell you precisely what the agreement actually contains?
That's the gap where a lot of trust erosion lives. People form strong opinions about implications before they've fully processed the underlying facts. And the platforms are optimized to serve the implication directly, because the implication is more emotionally engaging than the document.
Which suggests a practical discipline. Before you share something, or before you update your view on something, ask whether you've actually read the primary source. Not the take on the primary source. The thing itself.
If the primary source is a government statement, which might itself be strategically worded, you triangulate. Not just across partisan outlets, but across jurisdictions. What is the Israeli press reporting versus the international wire services versus regional Arabic-language sources? They're not all right, but the overlap between them is more reliable than any single one.
The triangulation point matters because most people think they're getting diverse information when they're actually getting diverse tones on the same underlying narrative. You can read five different outlets and think you're getting a range of views, but if they're all drawing from the same press briefing, you're not.
On the transparency side, the institutional obligation is real. Governments that communicate only in crisis mode, only when they have to, have already lost the ambient credibility that makes their crisis communications believable. The Israeli government's problem right now is partly that it has spent years treating transparency as a liability rather than an asset.
The cost of that is exactly what the survey shows. When you need people to trust you about something that matters, the trust account is already overdrawn.
The Edelman data actually bears on this. The organizations that maintain the highest trust scores are the ones with consistent, proactive communication, not necessarily the ones with the best outcomes. Process transparency earns more trust than outcome transparency, counterintuitively.
Because people can evaluate process. They can watch how a decision gets made and form a view about whether it was honest, even if they don't fully understand the technical details. Outcome transparency is harder because outcomes are complicated and the gap between intention and result is easy to spin.
For listeners who want to engage civically with all of this rather than just consuming it, the most useful frame is probably: where does your local information ecosystem have gaps? Not who do you distrust, but what kinds of information are you not seeing at all? The blind spots are usually more important than the biases.
Show up for the boring stuff. Local council votes. The trust crisis is partly a function of the fact that civic attention is highly concentrated on dramatic national events and almost absent from the institutional machinery that actually determines daily life. Which creates an accountability vacuum that gets filled by exactly the kind of actors who benefit from low trust.
Right, and that accountability vacuum is self-reinforcing. When people disengage from the boring machinery, the machinery gets worse, which confirms the suspicion that engagement is pointless, which deepens the disengagement. That loop is running in every democracy we've looked at today.
Where does it go from here? Because the technology question isn't settled. We've talked about what social media has done to trust over the past fifteen years, but the next fifteen are going to look different again. AI-generated content, synthetic media, personalized information feeds that are even more precisely tuned to your priors. The mechanisms we've been describing are about to get considerably more powerful.
Which is the open question I keep coming back to. There are two plausible futures and I don't know which one we're heading toward. One is that the synthetic content problem becomes so severe that people retreat toward credentialed institutions simply because they're the only sources with verifiable identity and accountability structures. Trust recovers not because institutions earned it back, but because the alternative became unnavigable.
The other future?
The other is that synthetic content is good enough, and cheap enough, that the distinction between credentialed and uncredentialed sources collapses entirely. At that point the concept of institutional trust becomes almost meaningless because there's no stable referent. You can't trust or distrust an institution you can no longer reliably identify.
Either way, the Israeli case is going to be one of the more instructive data points. High emotional stakes, dense information ecosystem, already running at the edge of the structural decline we described. If trust can be rebuilt anywhere under those conditions, it's probably worth studying how. And if it can't be, that's also worth knowing early.
That forty percent swing in a short window is either a warning or an opportunity, depending on what produced it. We still don't fully know which.
That's probably the right note to leave it on. Big thanks to Hilbert Flumingtop for producing, as always. And a quick word to Modal, who keep our infrastructure running without us having to think about it, which is exactly how infrastructure should work. If you've found this episode useful, leaving us a review helps other people find the show. This has been My Weird Prompts. We'll see you next time.
