Markets do not come close to fully pricing the costs of war, so how can they serve as the sole barometer of whether a war was a good idea? As for the timeline heuristic, I am not even sure how to engage with it, because it is completely arbitrary. Either it can be trivially gamed by choosing favorable cutoff points, or it creates a false sense of precision that is really just luck.
More broadly, I think you are modeling foreign policy like a speculative trade: if markets rally quickly, then the bet was justified. But high-risk gambles do not become responsible merely because they might pay off. Invading Iran carried meaningful risks:
- turning a country of roughly 100 million people with weapons-grade uranium into a failed state
- knocking out enough oil supply to trigger a global recession and stagflation
- pushing parts of the regime to sprint for a nuclear weapon
- dragging the United States into a multidecade quagmire with boots on the ground to secure the Strait of Hormuz.
I reject any framework that ignores tail risk. Prudent judgment has to account for catastrophic downside scenarios and ask whether imposing those risks on billions of unwilling people was ever justified.
And to make this concrete, the people in charge of this war seem to have been operating on a version of underpants-gnome theory for foreign policy:
1. Kill a couple of regime leaders.
2. ???
3. Iranian society overthrows the regime and replaces it with a government congenial to U.S. and Israeli interests.
No one in the administration with real power seems to have seriously considered what would happen if the regime did not promptly collapse, or how it would respond when faced with an existential threat. Closing the Strait of Hormuz was the obvious response from iran and yet the administration did not treat it as a realistic possibility.
Everyone, including the Iranians, understands what happens to the global economy if the strait is not reopened within weeks. That is why Iran is attacking its neighbors’ energy infrastructure. So what is Trump supposed to do now: launch a massive ground invasion of a country with 100 million people, continue aimless aerial bombing, or pull a TACO? These are not 11-dimensional chess masters. They seem to think war is a video game where you win by killing the final boss. But the boss is dead, the game did not end, and now they are spinning their wheels, looking for an exit while trying to save face.
the war in afghanistan achieved all its original goals. it is only considered a failure due to the inexplicable decision to pursue regime change as an afterthought
Constantine converted in 312. The fall of the Western empire wasn’t until 476 and consisted of an invasion of barbarians after a big civil war. And the West started to take off again by the year 1000. Plus, we have a control in the Eastern empire, which maintained a high Christian civilization until the 1400s.
I agree that trying to evaluate the long-run impact of the French Revolution, into the 20th and 21st century, is too doubtful to be useful. But there are other examples where a historical event's effects, on a timescale of decades, seem to be rather clearer. Probably the most obvious is the Bolsheviks' seizure of power in Russia, which gave the most extreme & authoritarian faction of Russian socialists power over one of the greatest countries in Europe, & thus also allowed them to split off & empower similar extremists in the rest of the worldwide socialist movement & eventually to gain for the communist form of government control, not only of Russia, but of much of Europe & Asia. Some form of socialism would surely have gained power in some places no matter what, but it seems likely that, if the Bolsheviks had not taken over Russia, then the Leninist top-down party-state model of socialism, which tended to be more forceful & brutal than the democratic socialism of e.g. the early British Labour Party or the German Social Democrats, would not have spread anywhere near so widely. However, the Russian revolution is an 'example that proves the rule', since it had such effects by giving national & international power not just to the then-generally-disempowered socialist ideology, but to a particular, small & generally unpopular form thereof, which would otherwise have been doomed to obscurity.
Though this is beside the point, I should note also that the idea "that Christian superstition ruined ancient civilization" is quite wrong: the Neoplatonism popular among the educated pagans of the middle to late Roman Empire was hardly much less superstitious than Christianity (Celsus, the most sophisticated anti-Christian writer of the 2nd century, argued i.a. that Jesus's miracles were done by sorcery rather than divine grace & that it did not matter whether the supreme god was identified with Zeus or with the God of the Jews), & in any case this issue had little to do with the political instability & military weakening which ultimately allowed the invasions of the Germanic barbarians.
If they look good after 1 year but bad after 2 you'll still say it was a good idea? This seems like an irrational over correction against the extreme long term view of effects. It's not hard to trace the history of major events only 2 or 5 years later. Yes, existence is full of ripples that are constantly interacting and making direct causation difficult to attribute, especially after hundreds of years, which I agree is ridiculous, but when you take a major action that ripple dominates all others for the forseeable future. Stories aren't all so hard to tell.
For a citizen to only consider the economic costs and benefits of war is to be a machine, not a human. Though war is sometimes necessary (and I would argue it is NOT in the case of Iran at this time), it is always, always, always a tragedy. And I wish our best and brightest would consult their souls (and consider the souls of others) as readily as they consult the stock market.
When I consult my soul, it tells me to go around the world murdering theocrats and socialist authoritarians. Considering costs and benefits places some limits.
Your conclusion that the stock market correction, after liberation day, proves that protectionism is stupid is flawed. The tariffs were never about protecting USA industry, despite what was claimed. The way the tariff rates were calculated was ridiculous, and investors were rightly appalled at the gross incompetence shown.
Commentary about the "One Big Beautiful Bill". Also, Trump has stated numerous times that so much revenue will come from tariffs that all income tax could be abolished
I like this. Your short-termism sounds plausible, as does your metric. At the same time, my kind wife (who dislikes your essay) points out that you entirely ignore morality. My take is that short-term consequentialism is a useful metric, but so is morality, because actually we have no f-ing idea what the long-term effects of a major international action might be, and morality at least provides guidance, aside from whether it happens to correctly predict the future.
Hanania correctly points to the difficulty of evaluating "success" and the advantages of objective external measures. But he seemingly ignores their limitations. It doesn't seem too mysterious to suppose (for eample) that brutality begets brutality.
I agree with this, quite strongly in fact, with just one caveat which is economic policy which almost always _should_ to be judged on more complex and often delayed second order effects, largely irrespective of the immediate effect.
Eg one could increase the minimum wage and this could have the immediate effect of raising wages at the bottom of the distribution with minimal other apparent costs such as loss of jobs, but two years later we can clearly see evidence that this actually cost lots of jobs and lowered total income at the bottom of the distribution.
There is a difference between “intervention is good” and “arguing for intervention is a good use of time”. Even if war would be slightly beneficial, advanced market commitments for new antibiotics, regulating factory farming, deregulating ports, etc. are clearly overwhelmingly beneficial. Questions of war and peace have always been very salient to elites, and this is probably unchangeable, but they are probably too salient.
It is probably useful to “raise the sanity waterline” on the topic, or to write popular posts about it, to create a pipeline for politics and foreign policy junkies to economics, science policy, etc. Scott Alexander once wrote about his strategy here - writing popular culture war posts in the hope that 1% of those readers would read his charity posts. Good to be deliberate here!
I agree that it is still too early and the war may turn out somewhat fine even if there is no regime change, but it is already clear to me that Trump made a mistake... by going in too early without adequate military resources. The effective closure of the Strait of Hormuz is a strategic debacle that will have after effects, primarily on Gulf monarchies like Kuwait and the UAE. So is the recent Iranian stipulation that it will allow the tankers to go through the Strait but only if they pay for oil in yuan. This effectively compels the administration to take control of the Strait by force, prolonging the war and the economic pain. It also puts additional stress on European economies, which, together with Trump's bullying European countries into helping us in the war (about which they were never consulted), pushes Europe even closer to China.
I want to make a comment about French Revolution. I did not know that Zhou Enlai misread the question; in retrospect, his answer makes more sense. But the French Revolution was enormously influential beyond Napoleon, because Napoleonic wars had largely awakened nationalism in German lands, the Habsburg empire, and Poland. The Revolution also created the first modern nation-state (the US was still largely a confederation), forcing other principalities and empires of Europe to become nation-states as well. You can trace a direct effect from Napoleonic wars to unification of Germany and Italy, as well as the peoples who inhabited the Austrian empire. At some point, of course, you have to draw the line... but some world events are orders of magnitude more influential than others.
> Was $36 trillion in lost value worth a twenty-five-percentage point increase in regime change, and a weaker Iran regardless?
I don't think this is a correct analytical frame. Part of the reason that the % increase in regime change risk was what it was, is that the administration's plans were not clear to the trading public. Suppose that on the day of the attack, Trump & co had a truly atrocious, moronic plan; the traders would not know this, and would consider it more likely than 0% that they had a good plan that had a real chance of achieving its goals. However, the plan had already been locked in, good or bad - the information was secret to traders, not Trump or God.
Of course, it's difficult to get a truly accurate assessment of a plan's likely efficacy, though historians do eventually converge, that can take quite a long time. But I do think in this case it seems fairly clear that Trump & co did not have a plausible plan to accomplish regime change that failed; they had no plan that could realistically accomplish regime change and therefore failed.
I came here to say precisely this. In an event like a war with Iran, the US military has a ton of top secret information that no one else on Earth has access to. A smart American general could come up with a pretty reasonable probability of whether he'll be able to achieve Iranian regime change ahead of time, and can communicate that to the President.
When the traders see the US president has launched a huge attack on Iran on Feb 28, it looks much more likely to them that the US military had a good plan to achieve regime change they told the President about than it looked on Feb 27. So they make smart trades trying to profit off that increased probability. When it's 16 days later and the war with Iran looks like a clusterfuck, a smart trader would guess that probably actually Trump just wanted to attack Iran, and the US military didn't have an especially good plan to present.
It's hard to know for sure; maybe there was a good plan but Iran had better countermeasures. But merely going "Prediction markets went up 25%" is not sufficient.
"It’s looking the same as before now, but that should not have much influence on how we see the original decision to pull the trigger."
We get to evaluate more than the decision to pull the trigger but how well the operation was planned and executed, and not just at a military tactical level.
Imagine that the regime change odds went to 99pct, then returned to the original level. This is the market saying invasion should have almost certainly achieved regime change and yet somehow they screwed it up. So I dont think this immediate response automatically goes to the plus side of the ledger. I can accept your short term framing, but we also get to evaluate the execution quality, which is information revealed over a moderate period of time.
I mostly agree with this article but one objection I have to this approach in evaluating policy is that while it may show that an action is overall positive, it isn't useful for evaluating said action relative to alternatives.
Venezuela isn't a convincing example here but that's what I'll use. Had the US lifted sanctions on and committed to warming relations with Venezuela absent the removal of Maduro or any reform from Venezuela's government, we'd see improvement in Venezuela's economy. For this reason, the raid should be evaluated relative to such a policy rather than in isolation (In reality I think the raid probably has quite a bit of value added since the stock rally looks so great relative to anything I can see in the last 5 years in which we've seen sanctions eased and reimposed). Also, if the market had priced in an extensive military op in Venezuela one could argue that some of the rally could come from the market seeing that the chance of an extended or deep cutting intervention was unlikely (Again probably not applicable for Venezuela since we'd expect to see more movement downwards in the lead up to the op as assets were moved).
Not a bad article overall, but you do make one very obvious analytical error; the S&P 500 index value refers to the aggregated market cap of the companies it is composed of, not to the dividends they pay out. It those dividends which are counted as profits in GDP. The S&P price to earnings ratio for the index is currently 27.55 so the actual losses in come are 27.55 times lower than you calculated .
Asset prices can only be realised as income through capital gains on the sale and the income gained depends entirely on the time that they are sold, how many people are selling at the same time etc. Whilst more serious for it to happen to the whole index a 2% fall in value is a commonplace and insignificant fluctuation for any given share.
The degree to which any given investor is effected by events like this depends entirely on the extent to which their portfolio is neutral to the market vs simply following it, for example my own investments are still up since the war started.
Markets do not come close to fully pricing the costs of war, so how can they serve as the sole barometer of whether a war was a good idea? As for the timeline heuristic, I am not even sure how to engage with it, because it is completely arbitrary. Either it can be trivially gamed by choosing favorable cutoff points, or it creates a false sense of precision that is really just luck.
More broadly, I think you are modeling foreign policy like a speculative trade: if markets rally quickly, then the bet was justified. But high-risk gambles do not become responsible merely because they might pay off. Invading Iran carried meaningful risks:
- turning a country of roughly 100 million people with weapons-grade uranium into a failed state
- knocking out enough oil supply to trigger a global recession and stagflation
- pushing parts of the regime to sprint for a nuclear weapon
- dragging the United States into a multidecade quagmire with boots on the ground to secure the Strait of Hormuz.
I reject any framework that ignores tail risk. Prudent judgment has to account for catastrophic downside scenarios and ask whether imposing those risks on billions of unwilling people was ever justified.
And to make this concrete, the people in charge of this war seem to have been operating on a version of underpants-gnome theory for foreign policy:
1. Kill a couple of regime leaders.
2. ???
3. Iranian society overthrows the regime and replaces it with a government congenial to U.S. and Israeli interests.
No one in the administration with real power seems to have seriously considered what would happen if the regime did not promptly collapse, or how it would respond when faced with an existential threat. Closing the Strait of Hormuz was the obvious response from iran and yet the administration did not treat it as a realistic possibility.
Everyone, including the Iranians, understands what happens to the global economy if the strait is not reopened within weeks. That is why Iran is attacking its neighbors’ energy infrastructure. So what is Trump supposed to do now: launch a massive ground invasion of a country with 100 million people, continue aimless aerial bombing, or pull a TACO? These are not 11-dimensional chess masters. They seem to think war is a video game where you win by killing the final boss. But the boss is dead, the game did not end, and now they are spinning their wheels, looking for an exit while trying to save face.
But this metric the war in Afghanistan would have been a resounding success.
the war in afghanistan achieved all its original goals. it is only considered a failure due to the inexplicable decision to pursue regime change as an afterthought
Constantine converted in 312. The fall of the Western empire wasn’t until 476 and consisted of an invasion of barbarians after a big civil war. And the West started to take off again by the year 1000. Plus, we have a control in the Eastern empire, which maintained a high Christian civilization until the 1400s.
I agree that trying to evaluate the long-run impact of the French Revolution, into the 20th and 21st century, is too doubtful to be useful. But there are other examples where a historical event's effects, on a timescale of decades, seem to be rather clearer. Probably the most obvious is the Bolsheviks' seizure of power in Russia, which gave the most extreme & authoritarian faction of Russian socialists power over one of the greatest countries in Europe, & thus also allowed them to split off & empower similar extremists in the rest of the worldwide socialist movement & eventually to gain for the communist form of government control, not only of Russia, but of much of Europe & Asia. Some form of socialism would surely have gained power in some places no matter what, but it seems likely that, if the Bolsheviks had not taken over Russia, then the Leninist top-down party-state model of socialism, which tended to be more forceful & brutal than the democratic socialism of e.g. the early British Labour Party or the German Social Democrats, would not have spread anywhere near so widely. However, the Russian revolution is an 'example that proves the rule', since it had such effects by giving national & international power not just to the then-generally-disempowered socialist ideology, but to a particular, small & generally unpopular form thereof, which would otherwise have been doomed to obscurity.
Though this is beside the point, I should note also that the idea "that Christian superstition ruined ancient civilization" is quite wrong: the Neoplatonism popular among the educated pagans of the middle to late Roman Empire was hardly much less superstitious than Christianity (Celsus, the most sophisticated anti-Christian writer of the 2nd century, argued i.a. that Jesus's miracles were done by sorcery rather than divine grace & that it did not matter whether the supreme god was identified with Zeus or with the God of the Jews), & in any case this issue had little to do with the political instability & military weakening which ultimately allowed the invasions of the Germanic barbarians.
If they look good after 1 year but bad after 2 you'll still say it was a good idea? This seems like an irrational over correction against the extreme long term view of effects. It's not hard to trace the history of major events only 2 or 5 years later. Yes, existence is full of ripples that are constantly interacting and making direct causation difficult to attribute, especially after hundreds of years, which I agree is ridiculous, but when you take a major action that ripple dominates all others for the forseeable future. Stories aren't all so hard to tell.
For a citizen to only consider the economic costs and benefits of war is to be a machine, not a human. Though war is sometimes necessary (and I would argue it is NOT in the case of Iran at this time), it is always, always, always a tragedy. And I wish our best and brightest would consult their souls (and consider the souls of others) as readily as they consult the stock market.
When I consult my soul, it tells me to go around the world murdering theocrats and socialist authoritarians. Considering costs and benefits places some limits.
Your conclusion that the stock market correction, after liberation day, proves that protectionism is stupid is flawed. The tariffs were never about protecting USA industry, despite what was claimed. The way the tariff rates were calculated was ridiculous, and investors were rightly appalled at the gross incompetence shown.
what were they about
The tariffs are all about raising revenue; supposedly to offset tax cuts to the very wealthy
why do you believe this? im not convinced theyre really "about" anything other than the president's affection for them
Commentary about the "One Big Beautiful Bill". Also, Trump has stated numerous times that so much revenue will come from tariffs that all income tax could be abolished
he has also said many times that they would bring back manufacturing and end the trade deficit.
targeted tariffs possibly could achieve those aims, slapping tariffs on everything imported has the opposite effect
I like this. Your short-termism sounds plausible, as does your metric. At the same time, my kind wife (who dislikes your essay) points out that you entirely ignore morality. My take is that short-term consequentialism is a useful metric, but so is morality, because actually we have no f-ing idea what the long-term effects of a major international action might be, and morality at least provides guidance, aside from whether it happens to correctly predict the future.
I am with your kind wife.
Hanania correctly points to the difficulty of evaluating "success" and the advantages of objective external measures. But he seemingly ignores their limitations. It doesn't seem too mysterious to suppose (for eample) that brutality begets brutality.
I agree with this, quite strongly in fact, with just one caveat which is economic policy which almost always _should_ to be judged on more complex and often delayed second order effects, largely irrespective of the immediate effect.
Eg one could increase the minimum wage and this could have the immediate effect of raising wages at the bottom of the distribution with minimal other apparent costs such as loss of jobs, but two years later we can clearly see evidence that this actually cost lots of jobs and lowered total income at the bottom of the distribution.
There is a difference between “intervention is good” and “arguing for intervention is a good use of time”. Even if war would be slightly beneficial, advanced market commitments for new antibiotics, regulating factory farming, deregulating ports, etc. are clearly overwhelmingly beneficial. Questions of war and peace have always been very salient to elites, and this is probably unchangeable, but they are probably too salient.
It is probably useful to “raise the sanity waterline” on the topic, or to write popular posts about it, to create a pipeline for politics and foreign policy junkies to economics, science policy, etc. Scott Alexander once wrote about his strategy here - writing popular culture war posts in the hope that 1% of those readers would read his charity posts. Good to be deliberate here!
I agree that it is still too early and the war may turn out somewhat fine even if there is no regime change, but it is already clear to me that Trump made a mistake... by going in too early without adequate military resources. The effective closure of the Strait of Hormuz is a strategic debacle that will have after effects, primarily on Gulf monarchies like Kuwait and the UAE. So is the recent Iranian stipulation that it will allow the tankers to go through the Strait but only if they pay for oil in yuan. This effectively compels the administration to take control of the Strait by force, prolonging the war and the economic pain. It also puts additional stress on European economies, which, together with Trump's bullying European countries into helping us in the war (about which they were never consulted), pushes Europe even closer to China.
I want to make a comment about French Revolution. I did not know that Zhou Enlai misread the question; in retrospect, his answer makes more sense. But the French Revolution was enormously influential beyond Napoleon, because Napoleonic wars had largely awakened nationalism in German lands, the Habsburg empire, and Poland. The Revolution also created the first modern nation-state (the US was still largely a confederation), forcing other principalities and empires of Europe to become nation-states as well. You can trace a direct effect from Napoleonic wars to unification of Germany and Italy, as well as the peoples who inhabited the Austrian empire. At some point, of course, you have to draw the line... but some world events are orders of magnitude more influential than others.
> Was $36 trillion in lost value worth a twenty-five-percentage point increase in regime change, and a weaker Iran regardless?
I don't think this is a correct analytical frame. Part of the reason that the % increase in regime change risk was what it was, is that the administration's plans were not clear to the trading public. Suppose that on the day of the attack, Trump & co had a truly atrocious, moronic plan; the traders would not know this, and would consider it more likely than 0% that they had a good plan that had a real chance of achieving its goals. However, the plan had already been locked in, good or bad - the information was secret to traders, not Trump or God.
Of course, it's difficult to get a truly accurate assessment of a plan's likely efficacy, though historians do eventually converge, that can take quite a long time. But I do think in this case it seems fairly clear that Trump & co did not have a plausible plan to accomplish regime change that failed; they had no plan that could realistically accomplish regime change and therefore failed.
I came here to say precisely this. In an event like a war with Iran, the US military has a ton of top secret information that no one else on Earth has access to. A smart American general could come up with a pretty reasonable probability of whether he'll be able to achieve Iranian regime change ahead of time, and can communicate that to the President.
When the traders see the US president has launched a huge attack on Iran on Feb 28, it looks much more likely to them that the US military had a good plan to achieve regime change they told the President about than it looked on Feb 27. So they make smart trades trying to profit off that increased probability. When it's 16 days later and the war with Iran looks like a clusterfuck, a smart trader would guess that probably actually Trump just wanted to attack Iran, and the US military didn't have an especially good plan to present.
It's hard to know for sure; maybe there was a good plan but Iran had better countermeasures. But merely going "Prediction markets went up 25%" is not sufficient.
"It’s looking the same as before now, but that should not have much influence on how we see the original decision to pull the trigger."
We get to evaluate more than the decision to pull the trigger but how well the operation was planned and executed, and not just at a military tactical level.
Imagine that the regime change odds went to 99pct, then returned to the original level. This is the market saying invasion should have almost certainly achieved regime change and yet somehow they screwed it up. So I dont think this immediate response automatically goes to the plus side of the ledger. I can accept your short term framing, but we also get to evaluate the execution quality, which is information revealed over a moderate period of time.
I mostly agree with this article but one objection I have to this approach in evaluating policy is that while it may show that an action is overall positive, it isn't useful for evaluating said action relative to alternatives.
Venezuela isn't a convincing example here but that's what I'll use. Had the US lifted sanctions on and committed to warming relations with Venezuela absent the removal of Maduro or any reform from Venezuela's government, we'd see improvement in Venezuela's economy. For this reason, the raid should be evaluated relative to such a policy rather than in isolation (In reality I think the raid probably has quite a bit of value added since the stock rally looks so great relative to anything I can see in the last 5 years in which we've seen sanctions eased and reimposed). Also, if the market had priced in an extensive military op in Venezuela one could argue that some of the rally could come from the market seeing that the chance of an extended or deep cutting intervention was unlikely (Again probably not applicable for Venezuela since we'd expect to see more movement downwards in the lead up to the op as assets were moved).
I strongly agree with 'Decide what you care about first.', however I missed that you did that in the case of Venezuela.
What do you care about, and what are the short term market proxies?
Not a bad article overall, but you do make one very obvious analytical error; the S&P 500 index value refers to the aggregated market cap of the companies it is composed of, not to the dividends they pay out. It those dividends which are counted as profits in GDP. The S&P price to earnings ratio for the index is currently 27.55 so the actual losses in come are 27.55 times lower than you calculated .
Asset prices can only be realised as income through capital gains on the sale and the income gained depends entirely on the time that they are sold, how many people are selling at the same time etc. Whilst more serious for it to happen to the whole index a 2% fall in value is a commonplace and insignificant fluctuation for any given share.
The degree to which any given investor is effected by events like this depends entirely on the extent to which their portfolio is neutral to the market vs simply following it, for example my own investments are still up since the war started.