Wasn't part of the leadup to the Industrial Revolution in Britain the enclosure of previously commonly-owned peasant lands? That seems like both the archetypal sort of top-down-imposed change to evolved institutions that James Scott types would decry (in fact I remember some left-libertarians like Kevin Carson banging on a lot about the evils of enclosure back in the day) and something that could plausibly have contributed to the Great Enrichment by enabling more intensive economic exploitation of land.
Here’s my comment with my prediction before reading on: 70% that the places that re-rolled on governance got better.
I have been told that in general some of the best things for economic growth are wars and natural disasters. (Indeed, I expect/hope to see something similar coming out of Covid.) On many levels, razing the current system and letting a new one grow allows it to get into a better equilibrium.
That said, if you do a big enough level of destruction (e.g. x-risk) everything is just dead and there’s nothing good left. You can destroy the FDA and expect a better thing to rise up, but I think just destroying the United States doesn’t get you a better United States.
My prediction here is that most of the places that got taken over and re-done got better. I’m 70% that was the outcome. Main reason why I’m wrong is if they didn’t get to do the things that the people wanted and instead that the French made them all do the same top-down things that the French wanted (e.g. losing a democracy and getting a totalitarian leader).
(One other reason why I’m wrong is that it turns out Scott wrote a fun, narratively twisty post where the answer is “Gotcha, assigning a probability here is fundamentally confused” and they got better in some ways and worse on others and it all comes down to whatever philosophical assumptions you bring to the table. And then I will feel a little silly.)
Alright, that’s my prediction. 70% that the places that re-rolled on government improved.
60% chance the difference is surprisingly small to the point of irrelevancy, on general principle that most interesting studies of politics-adjacent interventions don't turn up much.
In event there's a relevant distinction, 65% in favor of top-down innovation being better than evolved wisdom. Mostly cause I suspect Europe around that time period leaned much more towards parasitic entrenched methods than effective evolved ones.
Fiiiiiine, I'll predict it from a position of ignorance. I have some vague memories of decimal time being imposed by some French revolutionary government or other, which obviously did not catch on. I have an even vaguer feeling that that was emblematic of all such programs, but I don't know if that is a memory or a gut instinct. As such, my wild guess is a three-quarters chance that the reformed polities did better than the others - simply because (no probabilities attached to this part) they had the advantage of being welcomed within the imperial trade network and protection, despite all of the terrible policies from on high they now had to deal with.
Change of institutions is not necessarily good or bad. What matters is what the revolution changes society into and how that is done. Frequently, radical reforms involve trying to impose grand political and social schemes on people which do not work because they are coercive and unnatural. Less frequently, rapid institutional change is for the better. The heuristic that rapidly sweeping away tradition and institutions for replacement is going to end badly is a good one. However, there are counter examples which makes this heuristic not perfect.
Kind of like the tension between a competition/antifragility dynamic vs. slack/coordination. It's tough to find a meta-rule to distinguish when to use which approach. Like, do we lean into federalism to defuse political tension, letting states compete and experiment; or do we abolish the filibuster and experiment nationally because of externalities and races-to-the-bottom?
Well, I guess Napoleon was pretty good at reforming. As was pointed out, though, Communists tended to suck at it. Was the British Empire any good at it?
It seems like lots of revolutions end up being taken over by the most radical and ruthless, because radicals are more willing to start shooting other revolutionaries once the original government is overthrown - see Stalin, Robespierre, Khomeini, Mao, etc.
Taking the plunge here: I'm predicting that the states that had only their institutions (with a side of land reform) overhauled involuntarily by the French did better than the ones that didn't. The ones that felt a heavier revolutionary French hand, for whatever reason, I'm predicting did less well, but probably still (at least) about as well as those that retained all their indigenous features status quo ante. I doubt any of the unaffected states did better by a considerable margin than any affected state.
I'm cheating a bit because I studied this partly in undergrad, and I know that even the states that were defeated by but managed to retain the vast majority of their independence of the French (Prussia, Austria) adopted many reforms that the French had implemented in the Confederation of the Rhine, and certainly didn't roll them back in the lands they annexed or re-annexed after the final victory of the "counter"-revolution. I know also that the economic reforms, in particular, laid the groundwork for Germany's later economic boom in the second half of the 19th century.
My predictions: 40% effects not very large, 35% conquered better, 25% conquered worse. This includes savviness about which result Scott would write about, alas.
This is late, I've already read the article, but my prediction was "good, because I've never heard of Napoleon having skill failures in his command staff like Lysenko." I think picking competence is self-reinforcing, and I'd guessed that Napoleon would be an "A-lister hiring A-listers", as the saying goes.
Mostly ignorant prediction: conquered areas with culture closer to France’s did about the same while more culturally independent areas were worse off. 65% confidence
Another country that avoided getting conquered by Napoleon was Russia. Didn't do so well over the ensuing 105 years.
But then, Russia is huge, which can make up for subpar institutions. The UK is not, so it needed good institutions to survive the war, and it had them. (Being an island was useful too.) Disclosure: I am from the UK, which is probably why I predicted that conquered states would do less well.
Preregistering prediction: I think the reformed areas do worse for a while afterward, but then end up doing better in the longer term (like 10 years after).
Registering prediction in advance: Radically reformed patients (sorry, I see everything in terms of clinical trials now) performed slightly better in the short term (because backed by an empire), slightly worse in the medium term while dealing with unforeseen patient-specific consequences, and equally to slightly better in the long term.
Maybe this is because my politics are less libertarian than Scott's, but I struggle to understand the appeal of trying to classify "reforms" generally, or even "reforms imposed from outside," as good or bad. It seems like trying to classify "life events" as good or bad; it just depends on the individual thing, and they are so different that trying to cast judgment on them as a category doesn't reveal anything.
This reminds me of a lesson I got out of studying Japan. For those who don't know, prior to the Meiji Restoration Japan was divided into about 300 mini-states. Unlike European feudal duchies, they really were statelets: centralized organized territories with salaried bureaucracies. This system kind of knew it was ending after the Opium War destroyed the local geopolitical order. And it really knew it was going to end after 1853. But the actual war didn't erupt until 1869. What happened in those intervening years? Well, each of these statelets tried to strengthen itself in the face of these changes.
Some embraced radical westernization, adopting European style institutions and even sending secret expeditions to study or cooperating with Europeans to smuggle in trade and books. Others instead reformed along traditionalist lines, seeking to strengthen traditional institutions and reform the state into earlier, superior forms. A few others took other paths. I was fascinated by two that took opposite paths: one state emulated Great Britain and another emulated Tang China. Both, of course, adapted it to their local conditions. Which fared better?
They both hugely outperformed expectations. A westernized army and economy apparently was about even with a modernized version of the Tang militia and economic system. The states that did poorly were the ones that failed to reform. All the states that reformed in any direction did pretty well, often in proportion to how much they could reform those institutions.
This made me consider that the capacity to reform and make a program work might be more important than the actual contents of the program. And I often think this is an underappreciated factor of analysis. Libertarianism is always the correct response to an ineffective or incompetent government. If the government's an idiot, it's best the government do as little as possible. It's only when the government is competent that we can debate whether private or public options are better. For example, with healthcare, there are two questions. Firstly, all the policy we actually talk about. Secondly, is the US government actually capable of getting it done? Think of the healthcare.gov website. That was just a straight capacity failure. No one was sitting there debating the advantages or disadvantages of having a functioning website. The government just couldn't get it done. And I think we don't talk about that kind of thing enough.
I wonder how much of the positive changes came from the specific technocratic/policy reforms the French enacted vs. just destroying the existing systems and letting new systems blossom. Considering this all happened right around the start of the industrial revolution, it was probably a good time for radical change no matter the political or ideological reforms that were enacted.
To put it more bluntly it seems destruction forces progress, which will often lead to better outcomes, but I doubt many people would want to have their entire communities and livelihoods destroyed just because it'll improve economic conditions in the area fifty years later.
Registering a prediction of 30% probability that the reformed areas did better. I'm counting a case where they both did equally well as a "No" for resolution. Predicting relatively low since the title of the post contains "consequences" which has a negative connotation thus biases me towards expecting poor results in the reformed constituencies. No other factors went into this prediction.
It seems likely that there's a strong selection effect in that the reforms studied were the famous Napoleonic reforms and not some random radical reforms that went really badly and no one thinks about anymore. After all, the anti-radical-reform argument isn't that it's *impossible* for radical top-down reform to go well, just that such reforms don't *tend* to go as well as evolved institutions. It seems like picking one of the most famous reforms in history, which instituted the Napoleonic code that French law is *still based on*, is stacking the deck a little.
(My prediction was that the occupied places did better (60%), despite my general skepticism of radical reform. This was based on googling "Napoleonic code," seeing that it's famous and still in use in modern France, and concluding that it must be pretty good since cultural evolution kept it around this long.)
Prediction: comparing principalities that were reformed or not reformed has the same methodological problems as comparisons on any not-very-random sorting.
Registering my prediction in advance: I predict that the non-reformed systems did at least 5% better (on whatever metric the researchers used), and I am 60% confident in this prediction.
> So the authors ask: did the radically-reformed polities do better or worse than the left-to-their-traditions polities?
My immediate reaction was "oh, come on, what do you mean by 'better'?"
And the paper makes clear that their measure was... urbanization rates. (They also looked at GDP directly, but it sounds like their GDP data was so fuzzy and unreliable that they weren't comfortable using it as a measure.)
I wish the post had given this more attention. "Radical reform increases urbanization rates" is a) kind of unsurprising (modern ideas correlate strongly with urbanization for all kinds of reasons), b) not a refutation of anything in Seeing Like a State (the author kind of hates cities, right?), and c) not a sufficient basis for the kinds of value claims the post and the paper's authors seem inclined to make.
In other words, this is framed as an objective qualitative argument against James Scott's thesis, but it fails to convince because begs the question by reducing the problem to "legible" statistics and values. You had one job!
Would it really kill economists to stick some reference to effect size in the abstracts of these things? If this were a clinical trial, and the authors were like "hey, good news, napoleolimumab increases modernization, now FDA-approved for your unsightly problems with Prussian stagnation" they'd have the common decency to say "10 years of french reforms increases the primary endpoint of urbanization in 1900 by 9% from 41% to 50%" But instead i had to hunt through the text for it like a goddamn animal (page 23, if you care. They also have a point estimate of 36% GDP increase.
But now that they forced me to go through the paper, some not-terribly informed thoughts:
1. I have the usual concerns about econ research -- were any of these analyses pre-specified? How many different analyses were tried before they went with this one, etc.
2. If I am reading this correctly, by 1850 no changes are seen. So all the positive effect of the new institutions is from 1850-1900. Interesting.
3. Riffing on '2' -- maybe this can be spun as another example of "industrialization changes everything" or "conservatism is a better default in the absence of massive scientific/technological change." Blowing up institutions in 700 AD does you no good, because there's no innovation to take advantage of, you just get chaos. Blowing up institutions in 1800 AD helps, because it enables social shifts to take advantage of new modes of production.
4. And just for honesty: my prediction was "can't discern an obvious effect" (which in retrospect was idiotic given that if it were a null effect it never would have been published)
Prediction: 80% that the reformed states surpassed the non-reformed states, but only for a short (no more than a decade) time period before competition forced the other states to adapt.
Keep in mind that WWII era Germany and Japan went through radical reforms in the late 19th century, so I doubt they can be described as organically evolved societies.
The problem with this is a limited metric being used to derive a conclusion about a complex whole. If your measure of "better or worse" is simply aggregate economic growth this is an unsurprising result, even to many on the "traditionalist/evolved institutions" side. The trads (including me, and Burke) would argue, though, that there is a lot more to the health of a society, human happiness, and the common good than GDP.
So the state looked at a legible system (the conquered villages) and an illegible system (the unconquered villages), compared then according to its metrics (GDP, economic growth, etc.) and decided the legible system did better? Color me unimpressed. Almost every such reform looks better from the state's POV - that's the whole point of _Seeing Like a State_.
As Burke said, "The Revolution did not make France free, it made France formidable."
Bonapartism offered a sort of cautious version of the French Revolution, a hard-headed military man's version of rational modernization. It remained highly influential around the world for more than a century, such as in Bolivarism in Latin America, Kemalism in Turkey, and Nasserism in the Middle East.
Was gonna say 80% that radical (but well-intended) revolutions make matters worse short term (10 years) and 60% that they make room for necessary but impracticable in current system reforms. But then it dawned on me that "better-or-worse" is too heavily depended on the deciding criteria to form a meaningful opinion. China does just great economy-wise, but human rights and pollution are better left unsaid.
This paradigm bothers me. It seems implausible that there is some general, usually-applicable rule that entrenched institutions/traditions are either better or worse than systems imposed through radical reforms. Surely, an entrenched institution can be effective and efficient or ossified and redundant. And just as surely, a radical reform can be well-conceived or foolishly overconfident.
If that's right, then at best the paper here proves either that the French Revolution's reforms were well-devised or that the systems they replaced were weak, or both. Since the old systems were typically similar, and the reforms were all in the same direction if not literally the same, it suggests that *in this case* the reforms were an improvement. I don't see how we can possibly generalize from that result. Isn't the issue always case-specific?
My immediate thought when asked to make a prediction was "on what metric?" That basically sums up my frustration with this topic in general.
Evolved systems will, by nature, optimize, but there's no reason to think they will optimize on the metric that "we" want to optimize. In that sense, the problem with Brazilia wasn't poor optimization, it was misalignment between the metrics of the designers and the populace.
The same problem faces most modern reforms. We have broad disagreements about the different metrics Even if we totally agree about what a reform would do, different people will feel differently about it due to their different metrics. No policy is positive on all reasonable metrics, so no policy is universally agreed to be good.
That isn't to say agreeing on what the effects a reform will have is trivial. Even as a non-economist, I'm comfortable saying today's global economic system is a lot more complicated than that of pre-industrial France.
Prediction at the point of dare, before reading the rest: the reformed zones did initially better, but over 10-30 years returned to baseline, and no statistically significant difference is observed. More or less based on the idea that people are stimulated by New And Different and everyone tries harder, followed by Reversion To The Mean -- over a modest time (like 25 years) people fall into whatever rut suits their nature. (Long-term change only comes from technological or environmental changes.)
I feel like there's a good analogy with software systems. As large code bases grow, they tend to get messy: early architectural decisions are in conflict with new requirements; hacks and shortcuts never get fixed and snowball into monstrosities. So then there's always the question: do we rewrite from scratch and make it better this time?
Instead, the conventional advice is to always be refactoring opportunistically. Refactoring means you're not working from scratch: you're taking what exists and re-architecting it into something better. You might rewrite a chunk of code, but since it's a small piece, it's easier to be sure you're not missing functionality that already exists.
Although the opportunistic refactoring tends to be the superior approach, setting up incentives to make it happen is tricky. With pressure to meet deadlines and little recognition for it, refactoring often doesn't get done. When it comes to government, I think we see the same effect: the news doesn't cover small changes and politicians don't get a lot of credit for it, so there's not much "refactoring" of laws or institutions that gets done. Cruft accumulates, inefficient processes ossify, and eventually we get so fed up we want to do a "rewrite".
75% reform. I think a lot of this is based this off the most salient data point in my mind: Japan, which shouldn't necessarily generalize to all forms of foreign-imposed reform. But it's clearly a good example that such a thing is possible and often overlooked.
Two hundred years of historical changes (in France, no less) summarized in one blog post and I'm supposed to come to some kind of conclusion about our culture?
The (so-called) Glorious Revolution of 1688 made England more economically dynamic by introducing modern practices from Amsterdam such as corporations.
The Republican Revolution of 1861, made possible by Southern Democrats walking away from their Senate majority, allowed the passage of many modernizing pieces of legislation that forward-thinking pro-business politicians like Henry Clay had been talking about since sometimes as long ago as 1812.
18th Century Enlightened Despotism in Central and Eastern Europe, carried out by pragmatic conservatives, was fairly successful as well.
>So the authors ask: did the radically-reformed polities do better or worse than the left-to-their-traditions polities?
I'd guess the results would be dependent both on how long after the "conquer and reform" we look, and on a metric I don't really have a better name for than 'core v. periphery'. Conquest is inherently destructive, and the transition period is unlikely to create prosperity out of thin air whereas it can definitely do the opposite. On the other hand, I'd absolutely expect it to lead to massive gains in the medium-to-long term as the new philosophies both have greater freedom to adapt to new circumstances and simply because they're guaranteed to be replacing something that has failed in at least one key metric. I'd expect it to look something like an accelerated industrialization, which might be difficult to disentangle because if we're talking Napoleonic it'd be happening at the same time as 'conventional' industrialization. The destructive effects might be a little weaker at peripheral regions but the positive significantly less as well.
Hard to put a probability on something I'm guessing is going to vary in both space and time, but to take a swing at it: 80% of significant improvement in the 'core' regions on a timespan of up to a century. Comparatively little change on the periphery, maybe 60% chance of improvement in the same time scale with a preceding decline?
One thing to keep in mind about Edmund Burke was that, while he had a practical side, he also had a Romantic side. He was talented at romanticizing what was in the interests of the rich and powerful. Rationalists tend to be out of touch with the Romantic movement that arose from the mid-18th Century to the mid-19th Century in reaction to 18th Century Enlightenment Rationalism, although it led to some of the greatest works of art of all time.
I would hazard a guess that the radically reformed places did better, but I'm pretty unsure about this. Maybe 60%? That said, there definitely do seem like some confounding factors. Napoleon's empire didn't last forever, and when it was falling apart, maybe that affected places in it differently? Could be that the chaos hit there worse, could be that they started invading neighbors to recapture some of that early expansionary spirit. And when people were fighting against Napoleon, they'd probably treat the areas he conquered differently from the places downriver that weren't. Maybe?
My uncertainty stems from radical reform seeming dangerous, but the status quo being not ideal either, and my general ignorance of European history.
I feel like the call for predictions needs to be amended with dimensions... Like better on what axis? Wealth? Infant mortality?
I would say the reformed ones became slightly wealthier, but that may be tempered by the economic losses of being conquered and the associated destruction.
Also, I know almost nothing about Napoleonic legal codes.
I think what I got most strongly out of Seeing Like a State was the language of legibility and how that drives State reform. (J. C.) Scott selects certain schemes to increase legibility that fail catastrophically, of course, as the subtitle makes explicit, but the deeper concern I see from him is over _what gets lost_, which is illegible, and therefore never can be counted as a loss (which this study, of course, cannot account for, looking as it does at the exact same legible metrics that Napoleon was looking for). Of course, his examples are double-bounded, because they have to be _both_ examples of catastrophic failure _and_ examples which we can look back on legibly now and say 'oh yeah, shit... that was a terrible idea.' Which also means (and J.C. Scott is explicitly concerned about this) that those examples are poor examples anymore because we know so much more now and can avoid the mistakes...
In this respect, (J.C.) Scott seems similar to McIntyre (who, remember, was a diehard Marxist before he became a Thomist and Catholic–and I think it shows) and his After Virtue. You lose things, many of them perhaps permanently, with radical reform and centralization, you even lose the language in which to express and, finally, understand what has been lost and why it matters.
Samzdat's read of Seeing Like A State grapples more directly with this, as does most of Lou's writing there.
I think I disagree with Scott's initial sorting. Placing evolved institutions on one hand and complaints about vetocracy on the other strikes me as not wrong exactly, but incomplete. I think it's pretty clear that both threads have their answers to vetocracy, and I think they both recognize the problem. Vetocracy, after all, frequently prevents organic evolution of communities, societies, and institutions. Granted, I'd agree that erring on the side of evolution is more likely to result in the rise of vetocracies than centralized systems (since a vetocracy can be an organic evolution in the first place, but then prevents evolution).
I think this is borne out by the remainder of the essay. Scott points out the counter-example to the trend identified by Acemoglu as being the UK, the font of all modern organic evolution ideology and one that proves it can produce outcomes as good or better than any other system. I agree.
The difference is that the organic evolution in the UK tended (for reasons that are probably impossibly complex) to minimize or at least control the growth of vetocracy and the other malefic consequences of organic reform. The ossified German states that escaped the table-flip of the French Revolution had not just been entirely captured by the malefic side of evolution, but their ruling elites saw the maintenance of such as existential (even when their neighbors and social equals down the river and across the border did better in the long run and often in the short, too). As a result, and given the complete capture of political, economic, religious, and (frequently, but not always) cultural power in those small states and estates, the organic evolution that produced that elite class killed evolution.
And yes, of course, the clear liberalism of revolutionary French institutional reform is fairly obviously good from a standpoint of promoting human thriving (probably showing my bias here, but the massive growth of surplus and the the rapid reduction of poverty and all other forms of misery worldwide since the American/French/Industrial Revolutions seems like a track record well worth owning). Returning to theme, the UK had evolved most of this liberalism organically and didn't need a revolution to instantiate it or permit the industrial revolution to take off.
Tangent-ish: one of the things that fascinates me about Napoleon was the union of the liberal and totalitarian in one person. He was stridently in favor of liberalizing as far as possible (by the understanding of his time) all forms of ordinary day-to-day to existence, mainly commercial, social, and religious. On the other hand, he would tolerate absolutely no dissent from his subjects nor any neutral power in Europe. He had no use for civil or political liberty of any kind, and would imprison, torture, and kill any who dared to think independently in public. He could take criticism privately, though. But that was mainly from the officers he lived with on campaign, and on military matters only.
And then on the other hand, you have some places where some extremely radical reforms were imposed technocratically, and the results were terrible. Scott name-checks collective farming, but let's not forget Maoist China and the Khmer Rouge. So I disagree that the radicalness of reform has much to do with success. I think it has to do with correctness of reform. This would apply to post-fascist reconstruction of Europe and Japan.
By the same token, let's not forget the violent reaction of the European people to even the very modest (on average) counter-revolution imposed by the victors in the Napoleonic wars. That (along with the revolutionary forces unleashed by previous Ancien Regime excess) led to a century convulsed by many and serial (of admittedly varying intensity) revolutions in France, Spain, Germany, Russia, the Ottoman Empire, Hungary, Poland, and even the UK (Chartists, etc.) That is, those were relatively non-radical reforms imposed fairly gently (again, on average; every counter-revolution was different), but they were clearly wrong, and so some more heads rolled. I think that had to do with correctness more than radicalness.
Final note: the economic core of Europe. By some analyses, France was not (and is not) in it. You might have heard of the blue banana. https://en.wikipedia.org/wiki/Blue_Banana. It's a theory (that may have been concocted by a French dissident to critique his government's policy) that there's a large amount of easily accessible coal and iron in defined areas of Europe, and this was a primary driver of industrialization. The parts of France within the banana (Nord Pas-de-Calais, Picardie, and Alsace-Lorraine, primarily, and Rhone-Alpes secondarily) were/are some of the most industrialized areas of the country.
I don't know whether I agree with it, but it does seem that, after the end of the Napoleonic wars, France entered a sustained period of economic, political, and (especially) demographic decline, relative to the other nations of Europe. France's political unity and favorable position on the Atlantic coast allowed it to punch somewhat above its weight at one point, in seizing a large overseas empire during the New Imperialism, but other than that, it was fairly downhill until today, having joined the ranks of the Europoors with Italy and Spain. I'm only partially kidding about that.
What makes me skeptical of the theory is the motivated reasoning of its originator, whose purpose was to criticize excessive centralization in Paris, outside the banana. It still rings true to me, but I don't have a ton of confidence.
To the extent that the foregoing is just a re-writing of Scott's essay, I can only say I started writing this about half way through reading it.
From a position of ignorance, I would guess that the ones receiving the Napoleonic code were ~70% likely to do better. It seems like codes of laws that stuck were mostly centralised and European principalities weren't small enough to benefit from the effect of being a small community where everyone knows each other.
Radical reformers improved legible metrics like GDP per capita, lumber per acre farmed, and taxes collected. However, other metrics like quality of life, lifespan, and subjective happiness decreased. The length of time that the radical political was in effect did not last longer than the amount of time it was implemented (using Napolean's exile as 0 on the time axis).
My guess was 74% that the radical reforms would be net positive. In retrospect I think this was mostly a failure; given what I knew, I should have been able to be more confident in this.
Re the more general debate between radical reformers and traditionalism, I kind of like the analogy I made of society as a damped harmonic oscillator, which is probably the most stereotypically rationalist thing I've ever written
I think the discussion gets muddled by trying to compare such materially different places as Brasilia and Napoleonic France, especially in light of the stark differences between their stated goals. Comparing England's process of industrialization with that of the USSR and Communist china might be more fruitful, especially considering that Mao's revolution and later reforms were all (ostensibly) done to benefit a peasant class that existed under a very old-school feudal system.
But if all the successful instances of rapidly 'reforming' longstanding systems(post-war Japan, Napoleonic France, Germany etc) all basically neoliberal , then that tells us much more about neoliberalism than technocracy itself.
My take away is that the question "should my priors favor evolved systems over designed ones" is kind of silly, because a given policy's origins are completely incidental to its results.
Acemoglu measures the current status of conquered lands in 1850, 35 years after Waterloo, which was, apparently, long enough for the bad effects of the breaking things and killing people to recede.
Some of the shorter term effects of the Napoleonic Wars were dire. For example, the chaos caused by Napoleon's failed invasion of Russia in 1812 led to the breakdown of government wolf control efforts across Eastern Europe, which led to wolf packs running amok for a few decades until they could be brought under control again.
The Napoleonic Wars had the good side effect of being so bad that they discouraged Europeans from going to war with each other all that often from 1815-1913. They had the bad side effect of entrenching Reaction in control of much of Europe, discouraging British-style cautious democratic reforms until the second half of the 19th Century.
In Switzerland, most historians probably agree that the reforms imposed by Napoleon had an overall beneficial long-term effect. To some degree, elites the French drove away from power managed to return in some parts of Switzerland, but a return to the outdated political system was impossible and beneficial compromises were found. I suppose that apart from Switzerland, other areas profited in a similar way.
But I think we should be careful to extrapolate from this. One important point is, of course, that France was culturally very close to the European countries it modernized, and the French people who took an active role had intimate knowledge of these countries. I would suppose that the track record for imposed reforms („nation building“ etc.) with a much greated cultural distance is much worse. Sometimes, it can go well, but in many cases, it does not.
The other important point, which is also emphasized in the text, is that the reforms conducted under Napoleon were actually rather moderate, as far as their content was concerned. They did not aim at some far-fetched utopian goals. They may have been more radical than comparable cases in how much energy they put into removing old elites and outdated structures, but what they replaced it with looks rather moderate from today‘s perspective and probably also from a contemporary perspective.
"The evidence suggests that areas that were occupied by the French and that underwent radical institutional reform experienced more rapid urbanization and economic growth, especially after 1850. There is no evidence of a negative effect of French invasion. "
Let us see. The Franco-Prussian War, World War I, The Russian revolution, World War II, The Holocaust, the Iron Curtain. The Collapse of the Soviet Union. The butcher's bill, say on the order of 100 million.
Rather than looking simply formally at "reform or not," it's important to look at the substance of a reform. The Napoleonic Code was strongly influenced by the Justinian Code of the 6th century AD, itself a codification of older Roman Law. So while the Napoleonic Code may have been "new" to the little duchies outside of France upon which it was imposed, it was hardly radical from the perspective of Western legal tradition. Perhaps the only lesson here is that, as legal codes go, the Napoleonic Code was pretty good. This would also accord with the non-correlation of Great Britain, which already had a good legal system.
The first thread ("institutions have evolved to something like a time tested optimum and reform is hubristic, etc etc.") suggests an essentialism that plagues a lot of conservative philosophy. As if "work ethic" and "value" and "character" and economic outcomes were perfectly fixed features of physical reality and not dynamically shifting systems interacting and changing each other. There's no reason that the territory or the agents within it can't shift and mutate and render what worked however long ago obsolete. An organism can be perfectly evolved into its niche until one day a random amino acid gets swapped in a virus genome then suddenly that organism's time-tested mating ritual spreads a new deadly illness. This seems to be more or less what happened with anti-maskers and covid, the notion that something could shift so radically in the environment and suddenly make a lot of indoor commerce dangerous is not just unthinkable but *laughably* ridiculous to some people.
I think about this whenever I hear Lincoln Project conservatives like Tom Nichols smugly dismiss younger generations as entitled and whiny. As if it were like a cosmological constant that x units of Hard Work = housing/a car/stability or whatever and there are no larger forces reshaping the relationships of people to capital.
A personal pet peeve of mine is the assumption that unintended consequences and perverse incentives, the things that are supposed to render reform and central government action unworkable, are exclusive to reform. For example the very notion of "private property" seems to work very differently in theory compared to practice. In the libertarian ideal, in theory there'd be some sort of maximal respectful non-interference with the property of others barring extreme externalities, etc. In practice homeowners aggressively leverage whatever government power they can muster to shut down the construction of more housing on other lots of private property anywhere near them. In practice the idea of property in peoples' heads seems to be something like: "This property is my investment, I can and must do whatever is necessary to protect its appreciation, including restricting the supply of nearby housing through government force." Nimbyism is bipartisan but its especially ironic that many of these people consider themselves ardent capitalists.
Early on I was keen on Acemoglu as someone digging into the questions Mencius Moldbug was interested in, and even recommended his work to Moldbug. Later, around the time of Why Nations Fail, I became much more skeptical of his work (and Johnson's).
I can't find the post where I made this critique now (it might have been in the comments) but I recall them supposedly teasing out the effects of colonialism... by using malaria as a proxy for certain kinds of colonial institutions, rather than comparing Ethiopia & Thailand to neighbors or even considering the possibility that malaria's direct effect was confounding their results. If you trace back the links on that Coyne post though, I do note that the Acemoglu paper is in the tradition of Mancur Olson, who was also writing after Germany & Japan surged after WW2 (and in contrast to someone like James Scott). Greg Cochran would say that's because places with high human capital can rather easily recover from destroyed physical capital.
I predicted the non-reformed areas would do better. Looks like I was wrong, but how wrong.
I have some unformed ideas about the difference between being conquered by an external force and having a reforming faction win from within the existing society. As the review points out, post-conquering, if the new system is "good" it will work well and if not it won't. The given examples (Germany, Japan after WWII) are external-force examples and I suppose the Napoleon examples are external-force as well, though some of the same duchies were probably once under the control of Charlemagne, so maybe more of a high-water mark of French expansion than external-force per se.
Contractors seem to do well after and during military action. There are things to rebuild and replace. I don't know how to quantify "GDP due to replacing what was destroyed" versus "really new growth."
Civil wars to institute "reform" seem to go differently (even the French revolution came apart after a few decades, hence Napoleon). When a different societal faction becomes dominant, the now-subordinate faction will potentially not cease to exist after all and there may be ?a greater tendency to authoritarian dynamics to keep the new social order in place, which might stifle growth as well. The status quo ante may be closer after a civil war.
The French Revolution wiped away a lot of the remnants of feudalism, a system that had emerged 1000 years before to provide local security against raids by Vikings and other Roving Bandits by allowing Stationary Bandits to set up to defend local turf. It worked pretty well under the conditions of the late Dark Ages (e.g., terrible roads). But by the 18th Century, feudalism was obviously outdated, as the more recent Absolute Monarchs insisted.
And the feudal system of layering complex bargains on top of other complex bargains and so forth and so on for a thousand years had made getting much done legally awfully complicated. The British had a system of allowing judges to creatively misread the legal record of the common law, but more conscientious Continentals tend to be bogged down by the weight of the past. So, the streamlined Code Napoleon was welcome. But keep in mind that the Code Napoleon wasn't all that radical. It reflected the neoclassical feelings of the age, that this is what the wisest Romans would have done if they were around today.
My primary worry about the radical reform paper, which you touched on but didn't state explicitly, is that if these places had similar pre-existing institutions and economic circumstances, and France enacted the same reforms in all of them, then of course there was a consistent pattern, and this is kind of like having a sample size of 1.
Their claim that reform failures can be blamed on not being radical enough seems absurd. All the most famous examples of top-down reforms gone catastrophically wrong were pretty radical, no? Or was the problem with the Great Leap Forward that Mao wasn't thinking big enough?
I settled on the prediction that the French-invaded places were equally likely to have gotten better as worse. Once I saw the answer, part of my brain was very insistent that I was really leaning towards the reformed places doing better all along. Dammit, brain.
You need to read Scott's other book. "The Art of Not Being Governed." Seriously, with what's going on in Burma... it is important to understand why the Burmese state is having so much difficulty and the highland polities he talks about.
"So maybe the moral of the story is something like - replacing stagnation and entrenched interests with good reform is good, and with bad reform is bad. Which sounds obvious, but I do think that considerations of "is this potentially challenging a carefully evolved system of traditions?" is less important than I originally believed."
Isn't that a subset of the belief "life is complicated"? I'm being mildly facetious, but only mildly. The world when stacked end to end is full of disagreements. However, a lot of it boils down to "doing good things is good" & "doing bad things is bad", and then the rest becomes finding ways to make it easier to do good things, and harder to do bad things, and easier to reverse mistakenly doing bad things.
I know that sounds very over-simplistic, but there's a lot of muddling in the middle that makes sense, even if you have enough brainpower to chew on the big ideas.
Prediction before reading: 60% that the radically reformed polities did better.
> Second, isn't it sort of weird that Britain, the country that got least invaded by Napoleon and had some of the deepest-rooted institutions of all, was the one that really kicked off the Industrial Revolution?
My low-confidence understanding of this is that Industrial Revolution <= steam engine <= innovation spurred by having to get coal out of the ground <= deforestation leading the British to investigate costlier alternatives. So you could say the cause is material. But I've also heard that Britain emerged from the mercantilist era with the largest 'innovator / merchant' class – I believe it saw the greatest conversions of the aristocracy into an entrepreneurial class of the European nations.
This sounds a lot like survivorship bias: if the revolution lead to a huge improvement (define improvement...), then the researches give it some attention and include in the study, but how many revolutions there have been that didn't really change anything?
Doesn't the presumption that these systems "evolved" misstate how evolution works? Pre-some kind of phase change, like revolution, what you have is a system that is extremely well adapted to those who have held power, but is not particularly robust in the face of any significant change in circumstances--either from an internal or an external force. For it to be evolution, you'd have to have some kind of driver for fitness, which I would argue effectively doesn't or only minimally exists in monarchies/dictatorships. The argument that monarchical systems which have built up over time because they've existed for a long time are somehow evolved takes a horribly cramped view of what drives evolution.
As for the UK driving the Industrial Revolution, the UK had already had its revolution--in the English Civil War, and in the Glorious Revolution of 1688. So it's not surprising, if were using revolution-driven "evolution" as an indicator, that the Industrial Revolution started in the UK.
Prediction: 80% the ones taken over by Napoleon did better. Reason: feudal system is trash and so the bar is already pretty low. By destroying the existing power structures you're mostly removing assholes that are making things worse.
I may be in danger of making the hypothesis untestable, but what hypotheses exactly are we comparing? The French-occupied lands are counted as instances of top-down design, and others are considered bottom-up? The post points out that some of the French reforms actually made things more bottom-up. Was Roman law completely eliminated from the French reforms, or just the parts that seemed not to fit?
I'm also going back and forth on using GDP as the measure. What other confounding factors might influence that? What measure would I use, if I could use anything at all? Utils? How long does it take the higher growth rate to amortize the blood and treasure spent to put the French in power? Is it possible to get the top-down reforms without paying the blood and treasure? Might there be less costly ways of destabilizing bad equilibria? Yes, I think I have a bias showing.
It seems possible to me that a poorly governed city might still have advantages for growth compared to the countryside, while coming out behind a comparable well governed city. If the laws are clear, stable, and fairly administered that might matter more than their origin, unless we think top-down/bottom-up delivers advantages in clarity and honesty.
Even good ideas can be implemented badly. I think what matters is whether people can innovate and adapt. Have I learned something new, or just moved the goalpost?
My advanced guess was that there is no significant correlation. But yeah, at least someone asked a testable question, and tried to answer it instead of logic about it, the pitfall LW-infused rats and postrats tend to succumb to.
Scott: in the future, would you consider leaving a top-level comment to which everybody who wanted to leave a prediction could reply, in order to collect and contain the predictions?
My prediction was 70% of conquered territories would do better.
I admittedly have read the entire article, but I did register a prediction (to myself) that the replaced systems would do better. My (no expertise) gut reasoning is that lots of economic things work better at bigger scales, and that the benefits of having a common legal/economic system "gifted" to you by a larger power probably has positive effects, regardless of the actual merit of those systems.
>>Are you sure you want to keep reading now? You’ll never get another chance to predict this from a position of ignorance!<<
Aarggh!
I truly don't know the answer.
In my youth I would have sided with the reformers. I had lots of solutions for the ills of the world. I still have solutions, but now I'm much more cautious about imposing them.
I am going to predict that the radically-reformed polities did better… not because of any philosophic leaning but because of what little I know about Napoleon's reforms
>>The idea was that the occupying American forces couldn't care less about the entrenched power structure and vetocracy in Germany and Japan, so they rammed through whatever reforms seemed like good ideas at the time, and they were in fact mostly good ideas. On the other hand, the Soviet Union tried the same thing in East Germany and that went less well.<<
It is reasonable to presume that what you are imposing matters. In this case capitalism is better than collectivism, at least as measured by economic growth.
I think Seeing Like A State only applies to complex systems, which resolves some of the apparent contradictions here. Complex systems have many interacting components. Cities and agricultural systems, profiled in Seeing Like A State, are good examples. Top-down reform doesn't work in complex systems because they are inherently difficult to model. Breaking up oligarchies isn't in the same category. Economists can confidently recommend breaking up extractive institutions because the first order effects are so good. The second order impacts are just a rounding error. In contrast, a city is a complex system that is all about second and third order impacts. Evolved institutions can be irreplaceable in complex systems, but aren't necessarily better in other settings.
Originally 0.7 on conquered dutchies outperforming non conquered ones but flipped my prediction after misremembering the title of this post to be "The Cost Of Radical Reform".
If we think of this as analogous to searching for a global minimum when one is currently stuck at a local minimum, then the question of whether radical reform is a good idea can be rephrased as asking whether we should try a large jump or a small one. There is no fully general solution to this problem, but we do have some heuristics.
One heuristic is, "How bad is our current equilibrium? Is it more or less ok, or are people really miserable?" This lets us distinguish between the cases of a healthy neighborhood where Robert Moses wants to build a highway and a crappy principality full of oppressed peasants.
Another, similar heuristic would be, "Are the people resisting change a small elite or a large underclass?" A small elite may just be trying to hold on to their rents, but broad support suggests that the current system is actually working reasonably well and shouldn't be lightly discarded.
Another hypothesis: Bottom-up evolutionary changes work better when the social/political/economic environment is changing fairly slowly, but top-down revolutionary changes work better when the environment has changed enough to render the current systems dysfunctional. Seems like this hypothesis is consistent with the findings: Napoleonic Europe wasn't exactly the most functional environment, so maybe the places that had French soldiers marching through had the advantage of clearing the dysfunction quicker.
Another observation: Napoleonic France had a wealth of examples of Things Not To Do in the ten years preceding the Coup of 18 Brumaire. (Example #1: The French Republican Calendar, where "Brumaire" was a thing.) So maybe the Napoleonic code had the advantage of a revolutionary framework, with most of the revolutionary rookie mistakes having already been made.
60% confidence that the reformed lands do better. However, I think this moment in history was somewhat unique, as it was on the eve of the industrial revolution. Going into the industrial revolution, it would have been uniquely beneficial to have a common legal system with one of the most powerful economic actors in Europe. So, while I think it's likely the reformed countries did better, I think we have to be a little careful about abstracting too much from it.
Might be worth pointing out that Napoleon already represented a step back from the most radical excesses of the French Revolution (no "rational" ten-day weeks or temples of Reason).
<i>Europe at the time had so many tiny duchies and principalities and so on that you can actually do a decent experiment on it - for every principality Napoleon conquered and reformed, there was another one just down the river which was basically identical but managed to escape conquest.</i>
Also, the German Confederation (basically the post-Napoleonic equivalent of the Holy Roman Empire) passed various economic reforms, most notably the customs union of 1834, which would have disproportionately benefitted the smaller German states -- states which, due to their size, had also proved easier for the French to conquer. I guess you could say that this was indirectly due to Napoleon, insofar as the German Confederation wouldn't have existed without him, but still, the appropriate conclusion wouldn't be "Radical, top-down reform is inherently good," but "Free trade is better for the economy than a series of tiny, protectionist micro-states."
"So the authors ask: did the radically-reformed polities do better or worse than the left-to-their-traditions polities? ... If you're really feeling bold, post a comment with your prediction before reading further."
To predict which will do better, I'd need to know how the authors will measure "better". I predict that the areas brought into the larger state with a unified legal system and government will do better economically, partly because the Code Napoleon was a pretty good code, but mostly because it's a lot easier to do business with people who have the same laws as you, and where you know how or whether law and contracts will be enforced.
But when I've read about various people living under Napoleon's occupying forces, they generally came to hate the French. So I'd guess that those areas were less happy.
I am, however, skeptical that the experiment can actually be run as described, because I've never heard anything about there being unoccupied pockets within French-held territory. And if there were, there must have been something unusual about those pockets, such as being too remote, inaccessible, or poor to bother with. That would wreck the experiment.
Integration into the empire's economy, language & market were my considerations (over the non-immediate term).
Relatedly, on colonialism there's evidence (from another paper by Acemoglu et al., though they tell a different narrative, and maybe other papers) that, in the long term, countries never colonized do worse than those colonized. I think being conquered over land where your population is integrated into the broader system is beneficial (e.g. Russia in Siberia, or Napoleon) whereas being exploited by a handful of officials (e.g. Congo, Hispaniola) is certainly less beneficial, maybe even detrimental.
Yeah well, you addressed the key concern in your writeup. Here's another prediction for you to make: suppose you successfully spread the meme that "radical reforms" are good, and "designed systems are better than evolved ones." This is true because "there was a study."
Do you think more free markets are likely to come of that?
This makes sense. I think Scott would agree that the larger "local knowledge gap" is not between one nation state and its neighbor, but between the hyperlocal community and the nation state it belongs to.
Maybe national governments are say 95% interchangable with their neighbors, and if the French policies were already proving their worth back home (compared to their old system), they would be quite likely to do better than their neighbor's old systems too.
Did they measure any social outcomes? GDP growth doesn't give a conclusive answer on whether it was a desirable outcome, and urbanization is not inherently good. Cities are dysgenic fertility shredders, so in the long run when all of Europe has either adopted these reforms or they've become irrelevant, then earlier urbanization should have negative effects.
Was too immensed in reading to think of a number, but I was pretty certain the invaded looked better. Retroactively, I'd rate myself at around 80%. It's in fact perfectly consistent with my priors. Authority structures ossify to serve the people on top, overthrowing them makes way for progress. Obviously.
I think the problem with contrasting economic accounts with Scott is that they're concerned with different things. Meaning, not mutually contradictory. Economists study "legible" data like GDP, Scott looks at the wellbeing of the downtrodden. It can simultaneously be true that the business is booming and that the people are immiserated, and the texbook example of this is precisely the (England's) Industrial Revolution. (Which we do have legible data for. People were shorter, their lives were shorter.)
The takeaway from Scott is not to cease progress, but to watch for the little men that get trampled by it. The two views can be reconciled by not trampling the little men in the first place. To be more precise, by carefully rejecting easily-ossifiable and exploitable structures of authority so that the little men don't have to carefully weave their lives around their idiosyncracies only to get them thrown into disarray by a sudden disruptive wave of necessary but careless top-down reforms. Scott himself is an anarchist, which, obviously. (Needless to say, so am I.)
(To extend a hand to the pro-capitalist libertarians which make up like a half of the readership here, I commend you for sharing the above intuition. You just need to stop refusing to accept that capitalism does not solve this problem, in fact, it's itself easily-ossifiable and exploitable.)
In a time of rapid technological change, destruction of the old order will usually be a net good.
Soviets are a special case. In their homeland they replaced a +/- feudal system with a quasi-modern one, which had some beneficial effects. In the states they invaded, they replaced an emerging modern free-market system with a planned economy, with disastrous effects.
So, it depends if you're upgrading or downgrading the social system, I guess, and you have no way to tell if both are recent.
Perhaps big picture changes, like opening up markets, work well when externally imposed. But small scale bureaucratic changes are also necessary, and these work best when there is time and flexibility for them to evolve to fit the local cultural norms. Just look at how democracy ( a great idea) didn't work well in Arab Spring countries.
Ok I'll bite: Before reading past the "prediction pause", I predict 60-40 that Napoleon's radical reforms did worse than the traditional ones.
My reasoning? I mostly buy the Seeing Like A State-style arguments for bottom-up, illegible policies; on the other hand, much of what I've heard about Napoleonic reforms--including driving on one side of the road only--seem very good to me. But back on the first hand again, I might expect hear about the best Napoleonic reforms and not to hear about all the missteps. So, 60-40 against radical reforms.
(Also, I'm a physicist with little background in history, so my confidence is not much better than 50-50.)
Actually, despite systematic anti-Brazilian propaganda being dispensed through the world for unconfessable reasons, Brasília is widely considered "a singular artistic achievement, a prime creation of the human genius" and, accordingly, was chosen as a UNESCO World Heritage site.
I wonder to what extent this is an artifact of Napoleon's defeat at Waterloo. To explain, some Dutch history:
From the latter part of the 17th century, a group of wealthy merchant families were able to mostly reserve Dutch government offices to themselves, resulting in a ruling class of 'regents,' who served their own interests. The result was dissatisfaction among the middle class, who revolted twice during the 18th century. The second revolt was put down by Prussian intervention. At this time, The Netherlands was a client state of the English and the Prussians, with immense debts, which were paid for with extreme austerity and high regressive taxes, while an elite was extremely wealthy. In 1794/5, Napoleon managed to conquer the country, helped by revolts against the regents. This resulted in The Netherlands becoming a client state of the French. It was not really ruled by the French, but nevertheless instituted huge reforms, like forming a unitary state, adopting some democracy, adopting (progressive) income and wealth taxes, etc.
Napoleon's main interest in The Netherlands was to get money to fuel his wars, military support and a place to launch attacks on Britain from. The Dutch pushed back against this, eventually making Napoleon so angry that he forced the Dutch to accept his brother as king: Louis Bonaparte.
However, Louis turned out to have ambitions of his own: to be a good king who served the interests of the Dutch. He adopted the Dutch nationality and demanded the same from his ministers. He did his best to learn Dutch and made Dutch the official language of the court. He didn't crack down on smugglers that evaded the Napoleonic economic blockade of Britain, which helped the Dutch economy a lot & he refused to institute conscription. He also refused to write off 2/3rds of Dutch loans to the French.
All in all, he was a good king. Napoleon said of him: ″Brother, when they say of some king or other that he is good, it means that he has failed in his rule.″
Napoleon deposed Louis, but two years later he made the ill-fated decision to invade Russia, which in turn led to his later defeat at Waterloo.
If Napoleon had not been defeated and had been able to use The Netherlands as a milk cow for a long time, burdening it with debts, forcing its people to become soldiers, etc; wouldn't The Netherlands have stagnated as long as it remained under that rule? Perhaps the Dutch got very lucky by getting the best of the French reforms, but without suffering that much from Napoleon's more abusive desires.
So perhaps the lesson here is to let reformers institute their most sensible reforms and then banish them to Elba.
I think some way to integrate those kinds of predictions into the post itself would be really interesting. Some widget to make a prediction on a (pre-defined) question after the introduction, and some summary of the community predictions at the end. I am not sure if the LW Style prediction widget would suffice already, my gut feeling is that only seeing the community predictions at the end of the post would be better, but the LW style widget might already provide some interesting insight, without spoiling the rest of the post. It could obviously be gamed, but you aren't really cheating anyone but yourself, so that doesn't seem like that much of a problem.
- In 1861 Italy was unified and 1871 Germany was unified
I wonder how the paper could claim that everything good was just because of the French invasion rather than the a general revolutionary spirit or pressure to change.
As a comparison, the modern success of Japan is usually due to the fact that foreign powers colonized/influenced their neighbours, while they escaped conquest.
The deference shown to a "carefully evolved system of traditions" needs to consider whether the conditions under which the system evolved still hold. The early 1800s is around the time when a graph of GDP per capita throughout human history curves from essentially a flat line to a vertical one. Waterloo was in 1815 and the paper discusses growth post 1850, so it is likely that value of a French invasion is not the particular designed system they imposed, but the weakening of constraints on future adaptation.
I don't think it makes sense to have a fixed prior about the relative value of evolved vs. designed systems across all time and space without taking into account particular contexts.
My prediction was that areas conquered by Napoleon would have done much better. But honestly this wasn't much of a prediction, I spent most of my time trying to remember which areas Napoleon conquered. This includes most of the currently richest areas of Europe, making the answer kinda obvious.
"But then, isn't the whole point of Seeing Like A State that things which seem obviously true sometimes aren't?"
No, the point is that a top-down view misses local, tacit and embodied understandings. It only seems "obviously true" that rectangular grid squares are superior if you have a preference for rectangular grid squares. Very different things are "obviously true" from a local and embodied perspective.
"But then you would lose the right to apply Scott to most modern political debates, where there are no peasants to be found, and everything is the weird mix of extractive and altruistic typical of modern states."
Have you considered (the conflict theory perspective) that these "altruistic" policies are not altruistic? People get in a position of power and redistribute value from their opponents to their clients. Also, this is top-down imposed order. Some guys can obviously afford to pay taxes, other people are obviously too poor, lets put them in boxes according to how I evaluate their needs and surpluses.
You may find it relevant that GDP is an imperfect estimate of economic productivity. It seems easy to see why imposing a top-down system of legibility would result in a measured increase in GDP, for two reasons. You're moving value from illegible areas to legible areas, and you're requiring that that value be traded or documented somehow (whereas before value remained in the local illegible sphere).
Thoughts: It's said the Polish peasants weren't really bothered by the Partitions. They were just under new management.
The Congo and India and the Soviet Occupation Zone were set up as extractive colonies, whereas the Americans understood that a strong Japan and West Germany were the keys to containing Communism. The DDR did later become better supported, but the damage had been done.
"Daron Acemoglu is the most-cited economist of the past ten years, and I've never heard anyone say a bad word about him"
I'm not - really - an economist, and from what I know Acemoglu is far from being incompetent, but his "Why Nations Fail" was really weird to me in that it very clearly split historical periods and personalities into good (inclusive) and bad (extractive). He also seems to jump to the conclusion that democracy is always definitely better for inclusive institutions, and IDK how much I believe that.
But in general, I kind of agree with the idea about radical reform failing when it doesn't go far enough.
Regarding why Britain kick ass during the Industrial Revolution, it's worth reading Anton Howes' substack, since Britain certainly wouldn't seem to be the betting favourite before it all kicked off.
The reform principalities did somewhat better than the traditional ones, at about 65% sure. Lots of confounders and I am not sure how I would qualify 'better' - wealth is about the only objective measure.
Isn't the obvious explanation here that radical change comes with inherent problems, but that if the existing system is bad enough, it's still fairly possible to make large improvements? The French revolution swept aside feudal structures that were centuries behind their expiration date, and the Russian revolution might have worked out well if power hadn't ended up with people who were literally the worst (and even _those_ people managed to improve dramatically on industrial production).
Meanwhile, the combination of democracy, capitalism, science and markets mean that Western society seems fairly proof against that kind of ossification. In these cases, the problems with radical reforms would almost certainly outweigh the benefits.
"Does radical reform help or hurt" is a very, very different question than "would the implementation of modern legal systems including the Napoleonic Code raise GDP and urbanization vs. traditional legal systems in European polities."
For the record, my prediction as to the first question is "I have no earthly idea". I wouldn't even know how what to include in a data set to try to build a prediction off of, nor would I be able to quantify a value judgment as to whether the outcomes of reforms were "good" or "bad".
My prediction as to the second question is "yes, probably." I'm still not really willing to put a number on "probably" without knowing more specifics. Low confidence, though - maybe 60% - but part of that is admittedly because I expect to see counterintuitive results presented in this forum.
However, I think framing the question this way kind of gives away the answer. I mean, we know that "modern" legal systems are at least correlated with greater urbanization, so it would be kind of surprising if someone found that modernizing and standardizing legal codes resulted in decreased urbanization. We also have a plausible mechanism as to why and how modern, standard legal codes facilitate people living in larger cities in which they are more likely to have dealings with lots of strangers.
Also, in terms of the object question, I think you can see a clear mechanism where having standardized legal codes might facilitate trade (this even shows up now, in things like the US not importing foreign medications, and supply chain problems stemming from Brexit). So I think it's reasonable to predict that this would increase GDP.
After reading the results, I see that my interpretation of the question didn't exactly match the study, since the authors seem to be talking about destroying the institutional power of elites - which isn't exactly the same thing as implementing a standard code of laws. But that's exactly why I was reluctant to put a number on a prediction in the first place.
Finally, I think it's important to note that the choice of metrics here really matters. Why should greater rates of urbanization be counted as an indicator of successful reform? "Living in a city" isn't a terminal goal for most people, right? What was life like in these cities, compared with life in villages? What happened to life expectancy? Leisure time? Quality time with family? Number and quality of relationships? Mental health? I imagine the results would be more mixed here.
I also wonder to what extent our choice of metrics reflects a kind of survivorship bias. Development as reflected in GDP and urbanization may have produced competitive societies that were able to outcompete "traditional" societies, but as a result when we try to define "success" we're defining it based on "which societies won this competition" rather than "which societies were good for the people in them." There's a risk, in other words, that we're choosing metrics based on what we think success looks like starting with the assumption that we're the most successful and building outwards from there. I think the Johns Hopkins report (Global Health Security Index) shows the pitfalls of this kind of thinking - the metrics they chose were heavily biased in favor of the US and other "WEIRD" countries, and as a result some people were blindsided by the fact that the US had one of the worst pandemic responses in the world.
So I think in some sense when we're choosing metrics based on what we think makes a society successful we're in effect asking "will adopting [reforms that make societies more like ours] have the effect of [making societies more like ours]". And in that sense, it makes sense that the more radical the reform *in our direction*, the quicker will be the resultant move in our direction. But this tells us nothing about where future reforms will move us to.
Prediction: Frenchies did well on whatever metrics the authors select. This is based on two things: one, Napoleon as a "make the trains run on time" sort of guy, two, we say "metrics" in this context, not "measures" – I presume there's a good reason the French word won out. Or, to summarize both of those things: seeing like a state actually works, in that a strong state with a powerful executive organ can actually get things done. 80%.
If nothing else, I know from history that a lot of places (especially in Germany) did not undo the reforms after Napoleon was defeated, and other unconquered places tried to imitate them. So this suggests that what Napoleon did was valuable.
On a theoretical level, we should expect the politic system a society adopts to depend on the rate of material change within that society. If your society changes less over time, you can have more veto points, because the slow leakage of ideas through the system matches the input. If your society is constantly varying, you should see fewer veto points, because folks are constantly inventing new carriages/methods of piracy/articles of clothing/siege strategies etc., and you can't afford to wait to respond to them. And the Napoleonic conquests happen to occur right about when the Industrial Revolution hit Continental Europe. So we should expect the ancien regimes to be adapted to the slow hierarchically-constrained change discussed at https://slatestarcodex.com/2019/04/22/1960-the-year-the-singularity-was-cancelled/ and https://scholars-stage.blogspot.com/2010/08/notes-on-dynamics-of-human-civilization.html. Conversely, Napoleonic codes better match the pace of factory innovations.
Another worthwhile question is whether the states able to fend off Napolean were already healthier than those he conquered. Perhaps they already had the kind of anti-fragility Napolean introduced in conquered states, albeit likely not capitalist antifragility.
My prediction was 70% confidence that the French reforms were beneficial. Obviously getting conquered isn't great, but textbook growth theory says that good institutions are the most important driver of economic growth and I had the unfair advantage (relative to people at the time) of knowing that the Napoleonic Code is still in use two hundred years later.
Ah yes, the French Revolution - so extraordinarily innovative that they ended up being too weak to fend off a clever popular strongman who eventually made himself Emperor self-crowned in the presence of the Pope in the cathedral of Notre Dame https://en.wikipedia.org/wiki/Coronation_of_Napoleon_I Yep, that sure kicked out all those old dusty monarchical and imperial traditions! 😁
There's a reason you mention the Code Napoleon as the new legal code imported by the French into the occupied areas, and not the Code Robespierre or the Code Révolutionnaire, and it is because that after the fervour, ferment, and degradation of the Revolution into the Terror, it was easier for France and French institutions to slip back into the older, more stable, tradtional models. Yes, they did a lot of changes and reforms, but it wasn't feasible in the long run to be constantly and continually pulling down, sweeping away, and imposing hither, thither and yon.
Did the authors of this look at French colonial possessions alongside "yeah we had the French here for a while but then they left" in regards to how successful they were in the long run? (I am open to the argument that post-colonisation French possessions did better, but since most of Africa was colonised by everyone in the times of the Scramble https://en.wikipedia.org/wiki/Scramble_for_Africa , we can't really find a "this state was never colonised or occupied by outsiders" to be a control sample. Here, have a collaborative music video between Malian and French-Spanish musicians https://www.youtube.com/watch?v=J43T8rEOg-I). Small duchies and principalities such as Liechtenstein, Monaco and Luxembourg managed to survive to this day.
The point about "being invaded and occupied isn't so bad, why your economy will get a boost!" has me wincing. Germany was a developed European country at the time, the occupying powers didn't do much meddling apart from "no more Nazis" and pouring in money for reconstruction, apart from glaring at each other as former allies separated again, and it ended up divided in two halves, and East Germany was supposed to be doing less well. Do our economists do a comparison here?
Japan was a nation that was more or less forcibly Westernised, but even pre-war it was very eager to throw off the past, modernise fast, and on the model of its Western inspirations, become a colonial power itself. Oops! The US occupation did do a heck of a lot more restructuring and interfering here, but again, I think this was (economically) more a matter of throwing fuel on the fire rather than build from the ground up and do away with old feudal Japan.
And the main points I'm taking away from this is that the occupying powers eventually left. The French are not still in control of European territories, the Allies and Soviets have both left Germany, and Japan is in charge of its own government.
Economic development of an occupied territory is tricky and controversial business. One of our complaints over being colonised by Britain is that, post the Act of Union, Irish native industries were deliberately starved so as not to be competitors with British ones, and that the Irish market was treated as a dumping ground for British goods. We didn't get the same share of the Industrial Revolution, Northern Ireland with Belfast as the ship-building centre did that, and even Northern Ireland was more reliant on the linen industry which is not heavy manufacturing or high finance.
Indeed, even pre-Act of Union, this was a problem addressed by, amongst others, Dean Swift; we are still quoting his "Burn everything British but their coal". He wrote "A Proposal for the Universal Use of Irish Manufacture" in 1720, described as "a polemic on Irish affairs published anonymously by Jonathan Swift, his first such pamphlet after becoming Dean of St Patrick's. In it he attacks the English mercantilist policy which is draining Ireland of her wealth." https://celt.ucc.ie//published/E700001-024/
"It is the peculiar felicity and prudence of the people in this kingdom, that whatever commodities or productions lie under the greatest discouragements from England, those are what we are sure to be most industrious in cultivating and spreading. Agriculture, which hath been the principal care of all wise nations, and for the encouragement whereof there are so many statute laws in England, we countenance so well, that the landlords are everywhere by penal clauses absolutely prohibiting their tenants from ploughing; not satisfied to confine them within certain limitations, as it is the practice of the English; one effect of which is already seen in the prodigious dearness of corn, and the importation of it from London, as the cheaper market. And because people are the riches of a country, and that our neighbours have done, and are doing all that in them lie, to make our wool a drug to us, and a monopoly to them; therefore the politic gentlemen of Ireland have depopulated vast tracts of the best land, for the feeding of sheep.
(Footnote: Mr Lecky points out that in England, after the Revolution, the councils were directed by commercial influence. At the time there was an important woollen industry in England which, it was feared, the growing Irish woollen manufactures would injure. The English manufacturers petitioned for their total destruction, and the House of Lords, in response to the petition, represented to the King that ‘the growing manufacture of cloth in Ireland, both by the cheapness of all sorts of necessaries of life, and goodness of materials for making all manner of cloth, doth invite your subjects of England, with their families and servants, to leave their habitations to settle there, to the increase of the woollen manufacture in Ireland, which makes your loyal subjects in this kingdom very apprehensive that the further growth of it may greatly prejudice the said manufacture here.’ The Commons went further, and suggested the advisability of discouraging the industry by hindering the exportation of wool from Ireland to other countries and limiting it to England alone. The Act of 10 and 11 Will. III. c. 10, made the suggestion law and even prohibited entirely the exportation of Irish wool anywhere. Thus, as Swift puts it, ‘the politic gentlemen of Ireland have depopulated vast tracts of the best land, for the feeding of sheep.’)"
And for the "burn everything British", here is the provenance:
"I heard the late Archbishop of Tuam mention a pleasant observation of somebody's; ‘that Ireland would never be happy till a law were made for burning everything that came from England, except their people and their coals.’ Nor am I even yet for lessening the number of those exceptions.
Non tanti mitra est, non tanti judicis ostrum.
But I should rejoice to see a staylace from England be thought scandalous, and become a topic for censure at visits and tea-tables.
If the unthinking shopkeepers in this town had not been utterly destitute of common sense, they would have made some proposal to the Parliament, with a petition to the purpose I have mentioned; promising to improve the ‘cloths and stuffs of the nation into all possible degrees of fineness and colours, and engaging not to play the knave according to their custom, by exacting and imposing upon the nobility and gentry either as to the prices or the goodness.’ For I remember in London upon a general mourning, the rascally mercers and woollen-drapers, would in four-and-twenty hours raise their cloths and silks to above a double price; and if the mourning continued long, then come whining with petitions to the court, that they were ready to starve, and their fineries lay upon their hands.
I could wish our shopkeepers would immediately think on this proposal, addressing it to all persons of quality and others; but first be sure to get somebody who can write sense, to put it into form."
My guess: it won't change anything with a 50% certainty, and largely because to be in Europe in this revolutionary age means changing eventually.
Subsidiary is that it'll make things better at 20% because, if I remember properly, the Nepoleonic reforms were fairly standard stuff like 'let people work wherever they want' and 'peasants can wear whatever colors they want', and the reforms didn't touch the dangerous stuff, like family and religion (at least the Napoleonic reforms didn't, not after the terror).
30% it makes things worse largely on my general supposition that breaking apart institutions sucks and that being invaded sucks.
My prediction was that the Napoleon-reformed states gained a small but significant increase, of the order of around 10% of GDP. My reasoning was that having lots of incompatible weights and measures plus lots of fees and internal customs duties, cannot have helped trade.
How about positive effects from a bigger market? Created by unifying the conquered areas. Less obstacles to trade, lower barriers for entry, simply better capitalism...
I'll predict that the conquered regions did better, but only because I've heard that after Napoleon was defeated, merchants in the liberated regions lobbied to keep Napoleonic law, arguing it was better for trade.
"is this potentially challenging a carefully evolved system of traditions?"
Interesting wording, which actually differentiates between these cases to a reasonable extent.
Peasants, after all, have had a lot longer to evolve good farming traditions than guilds had had to evolve good industrial traditions (Napoleon being, y'know, right in the middle of the Industrial Revolution). Peasants have had a lot longer to evolve good farming traditions than *anybody* has had to do *anything*, except maybe societal fabric traditions.
(Social traditions are a lot harder to nail down when and whether there has been this kind of paradigm shift. Of course, in the case of societal fabric the state of theory on what makes for good cohesion and effectiveness is kind of *notoriously terrible*, so it's hard to actually point to many cases of people trying to use theory (Social Justice, the Cultural Revolution and psychiatry probably count, but I can't think of much else).
@SSC: comments are a mess: substantive comments drown in a sea of one-liner predictions. I suggest that when you call for predictions in a post not exclusively devoted to a call for predictions, make a comment thread specifically for predictions.
Where to start on this? Perhaps with a confession: I didn't make a prediction because as a still-practising historian it feels wrong to guess the answer to a historical question rather than look at some evidence and hypothesis. If had done I would have probably got the answer wrong though because I have a deep- seated aversion to modernist reforms, which would likely skew any other priors. This may be relevant to the critique that follows.
I would be concerned on the underlying methodology of this work, simply because I'm slightly stumped as to where you would find accurate comparable measures for GDP, urbanisation or anything else that would suit such an analysis in pre-Napoleonic Europe (it's not as if weights, measurements or even clocks and calendars were in any way standardised, so actual measurable economic measuresare expectedto be reconstructable?). These sort of statistics are only gathered by modern states not small feudal societies. So in this case they must have been invented, probably using existing data (comparability unknown) and some level of assumption. A twentieth/twenty-first-century creation of an eighteenth-century GDP for a small Rhineland duchy is not the same as a twentieth-century GDP measure for a Rhineland town, which is a serious methodological issue. Just because you want to compare something doesn't mean you can: the assumption you can do so is perhaps a sign that you are inclined to believe that states are quantifiable and manageable, which is likely to skew your conclusions. This is why professional historians tend not to produce theories like this: their reputation can survive most things, but being known for mishandling data to reach a conclusion is a serious problem akin to being a pseudo-historian.
I'd also note that the UK is not an outlier here, but rather was ahead of the curve. Whereas much of Europe had a new liberal order imposed by conquest around 1800, the English and Scots had forged theirs in the wars of the seventeenth-century, where absolutism was defeated and the commons asserted their authority over the aristocracy (you could argue there was a Dutch conquest in 1688 if you really want, but it was only really opposed in Ireland). Revolution, however glorious, had already happened for the UK, and indeed allowed the formation of the UK.
Acemoglu et al are also travelling in distinguished footsteps. Their thesis seems, purposefully or not, to justify Marx, who had already identified the importance of revolution in changing society. This might not be particularly significant as an academic discovery therefore: that the transition from feudalism to capitalism was regularly violent and beneficial is probably one of the most accepted (and possibly less original?) parts of Marx's analysis.
So in summary, this looks like a retread of Marx using the perennially dubious technique of trying to force data to tell you what you want to see, a methodological no-no in every field of thought. And if the finding is significant all it shows is that (in Marxist terms) it is beneficial to replace feudalism with capitalism, a summary that would have found few opponents in 1909 a century before it was reached. This may be unfair, and the conclusions reached seem likely enough from the priors that liberal economies do better than centralised feudal ones (for those thinking feudalism equals lack of economic control, check out the medieval legislation on prices and wages), and the rule of law is better than rule by decree. But from a histirian's perspective this looks like data mining to make a point, and this seems unacceptable methodologically. With that in mind I'd be inclined to doubt the book's conclusions are valid unless it becomes clear that there is good and reliable data being used. Especially because the conclusions can be used to justify violent revolution to achieve an aim now, whereas the revolution might be irrelevant and the changes that came about (however achieved - I think the Scandinavians got there without revolution?) are what is important.
Since I've no way to determine whether or not French reforms would improve things, I'd just wager that they increased fungibility of the state-machine. 11/9 odds that they improved the status quo
> If you're really feeling bold, post a comment with your prediction before reading further.
I am feeling bold, so 60% on the reformed systems performing at least somewhat better (expecting to be proven wrong, though, given the title and what I expect Scotts take on this to be).
70% chance that the French-reformed areas did better. I vaguely remember reading something about the influences of French-imposed legal reforms in my AP European History class.
My conclusion from this debate of technocracy vs tradition is then that in order to predict which one will be better, you have to have economic theory (or whatever theory is appropriate). If you know that markets typically work better, then it's possible to predict that pro-market reform will likely improve things.
I suspect that modern urban design theory could also have predicted that Brasilia was a bad idea.
Requested prediction: my gut says that the conquered people did better along objectively measurable dimensions in the short term (greater production in monetary terms, fewer deaths by starvation?), but worse on squishier dimensions (lower subjective well-being, possibly higher rates of alcoholism and depression).
Prediction: reforms dramatically increased urbanization, since prerevolutionary Europe had systems that discouraged urbanization. Post industrial revolution, urbanized areas exploded in GDP, so I believe GDP would have gone up proportional to urbanization, but not any higher than can be explained by urbanization. Politically I'm conservative.
Oooh: Advanced prediction. For me the same as the top one I see. No substantial difference. Gears type explanation: They are both part of European culture which is the primary economic driver.
Possible confounder would be that the states reformed by France probably are on average physically closer to England, and thus the primary point from which industrial development diffused.
I think the places that underwent radical reform should exhibit much wider variance in outcome than places which didn’t undergo change. I’m 80% confident here.
I think at least some places that underwent radical reform should have performed better than they used to: I’m 70% confident here.
I think that at least 20% of the places that underwent radical reform should have performed worse than they used to. I’m 70% confident here.
I think there’s going to be a bunch of questions around metholodgy that will difficult to know or tell how much some places did or didn’t improve. I’m 90% confident these questions will exist and 80% confident they’ll be Enough to make it very difficult to concretely say much here besides “sometimes it works well, sometimes it really doesn’t,” with a slightly higher weight towards the “it doesn’t” end of the distribution.
Isn’t it the case that it’s the least successful entrenched systems that tend to get radically revolutionised? Well functioning successful ones less so - and are more likely to be the ones imposing revolutionary change on others.
I think if that’s true, it’s going to skew how likely it is that radical revolutions result in functional improvements
I didn't have the nerve to make a prediction, but I wondered if immigration/emigration would be a good test for which places had gotten better or worse.
I remember visiting the big castle in Salzburg and hearing the guide talk about all the improvements and new walls being built and wondering why on earth they keep building walls when it already looks impossible to take. So remember my reaction when I heard how it fell: Napoleon was moving through the area with his army, without any intention of going towards the city. So the city rulers hurried to send messengers after him to surrender.
Napoleon isn't exactly your average ruler - it's the kind of guy apparently you want to be conquered by. On the other hand I finished reading Conquest and Cultures, and a common pattern is that conquered people do A LOT better than their freer neighbors. And yet, east germany did (much) worse under the soviets. Why?
Well, one explanation is that being conquered exposes you to a higher level of civilization - after all, they managed to conquer you so they must be doing something better. And also bring you in a market of technology and exchange that is usually a lot better than you previous single country level. In East Germany's case it just happened that being conquered by Russia was just objectively worse than being conquered by the West - worse market, worse tech, (much) worse ideologies. After all, the betters conquering the worse is just an average tendency, not a rule. And Germany was once the kind of country that tried to conquer the world. So them losing that battle and getting divided up like a big pie is not the same phenomenon as a lone country getting gobbled up by an empire.
When writing about modern development, economists often talk about moving peasants off the land and into more productive work, e.g., factories. It strikes me that old European systems like common land allowed a lot of very marginal peasants to hang on in the countryside, where their inconsistent labor was of course not very productive. When the commons were abolished and the system rationalized, they could no longer feed themselves and moved to the cities. So it was easier for factory owners to find workers in places where the countryside had been de-feudalized, and industry grew faster. This makes perfect sense at an Econ 101 level. On the other hand the mechanism by which this was achieved was cleansing millions of people off the land, using the threat of starvation, and forcing them to move to places where by this argument their descendants did not really benefit until 50 years later. The time scale makes this data useless for talking about modern reforms, since nobody is going to wait 50 years for good results. Heck, the radical reforms in Russia may turn out in a century to have been great. Who cares?
I used to work in factory automation. When I first started somebody explained to me something about how bottom-up evolution interacts with top-down redesign:
Him: We're going to go to [place] and do major upgrades to a 20-year-old [assembly line].
Me: They'll be happy - the new [robotics gear] is way more efficient.
Him: No, it never goes that way. We design and install this new system that is obviously better. Then the client complains that their operation is less productive.
Me: How? The [robotics gear] is way faster than what we're replacing it with. And the [operator interfaces] aren't cobbled together out of old circuit boards and TV screens.
Him: It's the operators. They get *really good* at operating the line. They know all it's quirks. They do little hacks to make things work more smoothly, or they'll figure out a way to run [different product] through the line when it's otherwise idle.
Me: So they'll get more productive on this line, but not until they've worked on it for a while?
Him: Yeah, they'll be way ahead of where they were eventually. But they'll hate every moment of it, and curse us engineers for swooping in and disrupting their system that was working great.
---
Can we reconcile the top-down planning vs. bottom-up evolution in a similar way? Big reforms change which landscape you're on (for better or worse) and accumulation of metis moves you toward its highest point?
Whenever I recommend someone read Seeing Like a State I also recommend they read The Ghost Map or some other book where top down reform worked out very well. Sometimes coordination problems are hard and some order needs to be imposed.
But with regards to Britain specifically it wasn't so traditional as all that. There was enclosure, toll pikes, and all sorts of changes. But Britain had Parliament where different interests could bargain and change things in a way that incorporated some degree of bottom up knowledge. Pure autocracies can't do that and have to rely on tradition to a greater extent.
You will find that to be an important lesson as you try to run a business. When you have to judge each proposal as “good, let’s do it” or “bad, let’s not”, you should carefully consider the devil’s advocate position to make sure you aren’t missing something, but you have to hold in check the intellectual urge to give extra weight to counterintuitive arguments just because they are cool. A good idea is a good idea.
My prediction before reading is a bimodal distribution. Radically reformed small entities will 60% have done better and 40% have done worse. So in particular, the variance of outcomes will be higher.
Ok, prediction. Napoleonic conquest marginally positive, where disrupted communities relatively more open to subsequent urbanization and industrialization. And also to nationalism, state bureuacracy, and eventually EU. 60% confidence. Null case: Weber's iron cage enforces most of this anyway, so Napoleonic conquest not very significant.
I predicted "no effect" from radical reform, as I see a lot of commenters did, but most people aren't explaining their reasons so I will: I metagamed it. The fact that you're bothering to ask the question implies it has an interesting answer, and the most interesting answer after you set up the reasons to believe both pro and con is "haha, trick question, it doesn't matter!"
Just finished Joseph Henrich's "The Weirdest People in the World". He suggests it was a cultural psychology that changed due to first the Catholic church limiting cousin marriages among other things and then after due to Protestants promoting individualism,learning to read and friendly behavior to people other than your relatives. Lots more detail in the book of course and but super well researched. We are "WEIRD" (Western,Educated, Industrious,Rich Democracies)!
Part of me wonders if the answer is democracy. You do those things which seem intuitive (popular) if they work great, if they don’t then you vote the people out and do something else. It’s not perfect (good ideas implemented poorly go away) and it’s not like an authoritarian like Napoleon can’t get it right by accident, but it is probably why democracies are so much richer than non-democracies.
The weakness of the "evolved institutions" side of things for me has ways been that the only thing institutions are evolving to maximize is longevity. There's a good reason to think a system that has lasted a long time will last a lot longer, no reason to think that system is more beneficial than the alternatives.
In a peasant context, or any other individual or small community context, we can imagine that systems will evolve towards being useful to those carrying them out. In a larger societal context though where people can exploit each other and have divergent interests the system can easily be long lasting based on its benefits to some, even as it harms most.
> Take a second to make a prediction here - a real prediction, with a probability attached.
I predict that 2/3 of the radically reformed polities did better than their left-to-their-traditions polities. I'm generally sympathetic to the Seeing Like a State argument, but my gut sense is that the polities in 19th century Europe were saddled with too much dead-weight tradition. So on balance, they needed some shaking up.
On “getting conquered by a foreign enemy is good for economic growth.” This is one of Sowell’s findings across lots of different groups far beyond Japan and Germany in Wealth, Poverty, and Politics.
But his argument was that it was isolation that kills and cultural exchange, regardless of whether it is forced, that creates prosperity.
A selection effect that's possibly worth pondering on when thinking about the French invasions: being invaded probably selects for not being powerful enough to resist invasion, which probably correlates with having a less effective economic/political system in place, which correlates with a higher probability that installing a new system causes general improvement.
Ok I'll bite and make a prediction without reading past the break...
The areas conquered by Napoleon did better in long term progress indicators on average, 70% confidence.
A) These areas would have been on average more impacted by the wars and therefore stronger postwar catch-up growth (think post WW2 growth on a smaller scale).
B) The reforms that survived the process of 10-15 years of reform followed by a reaction could be some of the more effective.
The 30% goes to the many potential confounders, me overestimating A, me underestimating the downsides of B, or just missing something entirely.
0.7 makes things better. 0.5 it makes it better in a way that also creates some meaningful costs (more inequality or social strife or something like that).
Prediction: Since Daron Acemoglu is the author I'm guessing that authoritarian imposition of pro-democratic institutions are good, and since this is the case they brought up that's probably what happened here. In the wake of Napoleon, principalities were turned into polities with on average slightly more democratic institutions than before and they were better off for it.
It's a binary question, so without considering any evidence I should start at 50%. What kind of evidence can I consider? For one thing, the Napoleonic Code is even in this post called "modern", so it has at least endured, which is a point in it's favour- it's at least not so much worse that it's been abolished everywhere. It's been so socially ingrained that I even have trouble imagining not having a family name for people, so it's definitely been successful in the sense that it's enduring.
I'd think that the reformed policies make the principalities more legible compared to the unconquered ones, which should have an impact. I also think the reforms should take energy that would otherwise be used productively elsewhere, so I'm expecting at least an initial dip, which could possibly have a sustained impact through compound returns?
On the whole I personally just don't have enough background knowledge to be very certain, so I'll stay close to my starting point. Still, I think this is moderate evidence to positive impact, so I'll move a couple of decibans towards positive.
let's say 61% probability that the radically-reformed polities do better than the left-to-their-traditions polities.
To me it seemed obviously true that Napoleon’s interventions were an improvement. That is often the case with empires because they have to standardize trade and law in order to scale. Rome had a similar impact, as did the British empire.
Before reading, the reformer did better around 80% of the time, because traditionally, monarchy are really bad for the economy. They don't want train because they allow easier revolt, they don't care about progress because the powerful don't have any risk of becoming poor, and the way to become powerful is to suck up other powerful people
Prediction: there will be a bell-curve of outcomes, centered around slight improvements. Even though the reforms were top-down, they were made by people with similar cultures who had absorbed many of the improvements.
You psyched me out! I predicted that the kingdoms were sclerotic and inefficient and so with 75% confidence Napoleon’s reforms increased measurable qualities ... eventually. But I consciously factored in my belief about why you were asking and that overrode my small knowledge of history.
As to why the Industrial Revolution started in Britain and not France, Britain had a several hundred year lead on reforming their monarchy. The Magna Carta was a different sort of reform, but it began the diffusion of power from a narrow aristocracy to wider elite class, who went on to have lots of first sons who ran family businesses and third sons who had lots of free time to seek clout by inventing stuff.
I think the same thread runs through why Communism failed, and it’s a point Nassim Taleb makes in economic contexts; you can be brilliant and see five years into the future, or just some guy if you keep your options open because in five years the answer may be obvious. Command economies - like kingdoms but also communism - are fragile to planning problems. The Soviets accidentally killed tens of millions of their own people with farming edicts and missed the boat on computers. And as long as the guy in power refused to get it, the right thing might never evolve.
Of course, a diffusion of power can also experience the fragility of planning problem, if the big players collude. Revolutionary reform may be effective at breaking up those cozy relationships, and may endure if it thoughtfully distributes power instead of consolidating it.
I think the answer is some combination of “energy is required to leave some local maxima and reach a new, higher maxima” and “most random paths away from a local maxima do not lead higher on net.” Meaning, I wonder if 19th century Western Europe was particularly ripe for this in a way that other attempts at reform were not. And that the incoming reforms had a sufficient degree of objective betterness that most of the work was done in smashing up inert, entrenched power structures and less in railroading cultural shifts.
Perhaps there is a distinction to be drawn between frameworks that enable valuable organic developments to be rewarded and reinforced (e.g., the rule of law, or, more specifically, intellectual property law) versus frameworks privileging entrenched interests. The Napoleonic Code was a quantum advance over "what the king or prince or whatever says is the law is the law," and of course the development of the common law, well before the NC, was one of the major factors in priming Britain for its era of hegemony.
OK, from the perspective of someone who has lived in France for a long time and been involved with the country for forty years.
First, Napoleon and the Revolution are not the same thing, and the Revolution itself went through various stages. It's often forgotten that the original Revolution was just the kind of high-level imposition of new ideas on France itself that Scott talks about, and was often deeply and violently resented. Some of the really wacky ideas (decimal days and years for example) never really caught on. Even the metric system only really became accepted in the last century everywhere, and, even today, if you buy half a kilo (500 gm) of apples in the market, you generally ask for "une livre", the old, pre-revolutionary measure equating to the English pound.
Second, what matters is the objective. Scott's agricultural examples, also in his book "Against the Grain", refer to a system that had evolved over a very long period of time and reached a kind of stasis. The judgement about whether they worked is essentially technical. So-called "modern" systems, alleged to give a better result technically, did not do so. Indeed, in Europe at least, so-called "permaculture" systems are now coming into fashion. But political and economic change is very different, and hard to evaluate objectively. What is true is that the French invasions gave a helpful shove to a process which was anyway inevitable - the construction of larger political and economic units based on popular sovereignty rather than inherited royal prerogative. But this was coming anyway, and in certain countries, like England, it came more slowly and peacefully, since the English ruling classes took fright at the prospect of a revolution there. (Even today, though, much of the British political system is distinctly less "modern" than the French, to the point where one historian memorably desired Britain as "the last functioning medieval state in Europe." ) The same was true, incidentally, of Japan: they modernised in their own way and at their own pace, taking what was needed from outside. Japan was already a modern state before WW2, and the "changes" much talked about as a result of the US occupation were largely cosmetic. Essentially, the Japanese carried on as before. But this was because the balance of power was different (the Japanese had a highly literate and well-educated society with lots of social capital, and hardly any USians spoke Japanese). What happens today is that foreigners descend on a society, like some in Africa, where little is written down and much of life is regulated by custom, and replace (as I've seen) traditional practices in the market with a new code of commercial law, literally translated from German, or whatever, which requires an infrastructure of literate tradespeople, inspectors, lawyers and judges to enforce it, none of which exist.
If being conquered was good, the Roman Empire falling shouldn't have led to massive decreases in standard of living.
But it did.
Likewise, if you look at Eastern Europe vs Western Europe post World War II, Western Europe did vastly better than Eastern Europe. Why? Because Eastern Europe was controlled by the USSR.
Colonized countries generally did better than non-colonized countries - unless the colonized country was developed, in which case, the developed country did better.
There are an enormous number of examples of these things.
The main differentiator seems to be the quality of the culture that won out/conquered them.
The idea that organic systems are inherently good is obviously deeply flawed. That is why Chesterton's Fence is such an important heuristic.
Well, I can safely say that I made the worst prediction. Lessons for myself: think whether you accounted for all the relevant information, are answering the right question and are not spouting complete insanity.
Prediction: net GDP growth for reformed polities is higher than for non- reformed polities up to 1900, 70% confidence. Not going to bet on the effect size, though.
The experiment isn't really testing the hypothesis as described, however. It's just testing "did the Napoleonic reforms generally improve GDP, urbanization, etc., in provinces where they were implemented." You've got a good sample size testing that hypothesis, but if you want to test the hypothesis of "Do centralized authoritarian reforms do better than evolved organic change?" you'd need to test many different examples and see what happens. This is like doing an experiment determining that penicillin treats infection and and concluding "We find no evidence for the hypothesis that eating mold is bad for you."
Registering my prediction in advance: I think on average there is no difference between the reformed and non-reformed systems.
0.6 on the side of the radical reformers
Traditional does better on average - 75%
Wasn't part of the leadup to the Industrial Revolution in Britain the enclosure of previously commonly-owned peasant lands? That seems like both the archetypal sort of top-down-imposed change to evolved institutions that James Scott types would decry (in fact I remember some left-libertarians like Kevin Carson banging on a lot about the evils of enclosure back in the day) and something that could plausibly have contributed to the Great Enrichment by enabling more intensive economic exploitation of land.
60% confidence napoleon ones do better, not much confidence I know
Advance prediction: 70% confident that the radically-reformed areas on average outperformed the control group.
Here’s my comment with my prediction before reading on: 70% that the places that re-rolled on governance got better.
I have been told that in general some of the best things for economic growth are wars and natural disasters. (Indeed, I expect/hope to see something similar coming out of Covid.) On many levels, razing the current system and letting a new one grow allows it to get into a better equilibrium.
That said, if you do a big enough level of destruction (e.g. x-risk) everything is just dead and there’s nothing good left. You can destroy the FDA and expect a better thing to rise up, but I think just destroying the United States doesn’t get you a better United States.
My prediction here is that most of the places that got taken over and re-done got better. I’m 70% that was the outcome. Main reason why I’m wrong is if they didn’t get to do the things that the people wanted and instead that the French made them all do the same top-down things that the French wanted (e.g. losing a democracy and getting a totalitarian leader).
(One other reason why I’m wrong is that it turns out Scott wrote a fun, narratively twisty post where the answer is “Gotcha, assigning a probability here is fundamentally confused” and they got better in some ways and worse on others and it all comes down to whatever philosophical assumptions you bring to the table. And then I will feel a little silly.)
Alright, that’s my prediction. 70% that the places that re-rolled on government improved.
80% on Napoleonic reform, but I think I'd heard something about this case before, so that might be tainted.
Prediction: 80% chance Napoleonic reforms improved things (I probably mostly am basing this on positive associations with the metric system)
Registered prediction:
60% chance the difference is surprisingly small to the point of irrelevancy, on general principle that most interesting studies of politics-adjacent interventions don't turn up much.
In event there's a relevant distinction, 65% in favor of top-down innovation being better than evolved wisdom. Mostly cause I suspect Europe around that time period leaned much more towards parasitic entrenched methods than effective evolved ones.
I predict the Napoleonized states probably did about the same.
But I’m always wrong.
70% confidence that radically reformed areas did better.
better: 50-60% confidence
70% reform
Fiiiiiine, I'll predict it from a position of ignorance. I have some vague memories of decimal time being imposed by some French revolutionary government or other, which obviously did not catch on. I have an even vaguer feeling that that was emblematic of all such programs, but I don't know if that is a memory or a gut instinct. As such, my wild guess is a three-quarters chance that the reformed polities did better than the others - simply because (no probabilities attached to this part) they had the advantage of being welcomed within the imperial trade network and protection, despite all of the terrible policies from on high they now had to deal with.
Change of institutions is not necessarily good or bad. What matters is what the revolution changes society into and how that is done. Frequently, radical reforms involve trying to impose grand political and social schemes on people which do not work because they are coercive and unnatural. Less frequently, rapid institutional change is for the better. The heuristic that rapidly sweeping away tradition and institutions for replacement is going to end badly is a good one. However, there are counter examples which makes this heuristic not perfect.
Kind of like the tension between a competition/antifragility dynamic vs. slack/coordination. It's tough to find a meta-rule to distinguish when to use which approach. Like, do we lean into federalism to defuse political tension, letting states compete and experiment; or do we abolish the filibuster and experiment nationally because of externalities and races-to-the-bottom?
Well, I guess Napoleon was pretty good at reforming. As was pointed out, though, Communists tended to suck at it. Was the British Empire any good at it?
It seems like lots of revolutions end up being taken over by the most radical and ruthless, because radicals are more willing to start shooting other revolutionaries once the original government is overthrown - see Stalin, Robespierre, Khomeini, Mao, etc.
Taking the plunge here: I'm predicting that the states that had only their institutions (with a side of land reform) overhauled involuntarily by the French did better than the ones that didn't. The ones that felt a heavier revolutionary French hand, for whatever reason, I'm predicting did less well, but probably still (at least) about as well as those that retained all their indigenous features status quo ante. I doubt any of the unaffected states did better by a considerable margin than any affected state.
I'm cheating a bit because I studied this partly in undergrad, and I know that even the states that were defeated by but managed to retain the vast majority of their independence of the French (Prussia, Austria) adopted many reforms that the French had implemented in the Confederation of the Rhine, and certainly didn't roll them back in the lands they annexed or re-annexed after the final victory of the "counter"-revolution. I know also that the economic reforms, in particular, laid the groundwork for Germany's later economic boom in the second half of the 19th century.
My predictions: 40% effects not very large, 35% conquered better, 25% conquered worse. This includes savviness about which result Scott would write about, alas.
This is late, I've already read the article, but my prediction was "good, because I've never heard of Napoleon having skill failures in his command staff like Lysenko." I think picking competence is self-reinforcing, and I'd guessed that Napoleon would be an "A-lister hiring A-listers", as the saying goes.
Mostly ignorant prediction: conquered areas with culture closer to France’s did about the same while more culturally independent areas were worse off. 65% confidence
Another country that avoided getting conquered by Napoleon was Russia. Didn't do so well over the ensuing 105 years.
But then, Russia is huge, which can make up for subpar institutions. The UK is not, so it needed good institutions to survive the war, and it had them. (Being an island was useful too.) Disclosure: I am from the UK, which is probably why I predicted that conquered states would do less well.
Preregistering prediction: I think the reformed areas do worse for a while afterward, but then end up doing better in the longer term (like 10 years after).
Registering prediction in advance: Radically reformed patients (sorry, I see everything in terms of clinical trials now) performed slightly better in the short term (because backed by an empire), slightly worse in the medium term while dealing with unforeseen patient-specific consequences, and equally to slightly better in the long term.
Maybe this is because my politics are less libertarian than Scott's, but I struggle to understand the appeal of trying to classify "reforms" generally, or even "reforms imposed from outside," as good or bad. It seems like trying to classify "life events" as good or bad; it just depends on the individual thing, and they are so different that trying to cast judgment on them as a category doesn't reveal anything.
This reminds me of a lesson I got out of studying Japan. For those who don't know, prior to the Meiji Restoration Japan was divided into about 300 mini-states. Unlike European feudal duchies, they really were statelets: centralized organized territories with salaried bureaucracies. This system kind of knew it was ending after the Opium War destroyed the local geopolitical order. And it really knew it was going to end after 1853. But the actual war didn't erupt until 1869. What happened in those intervening years? Well, each of these statelets tried to strengthen itself in the face of these changes.
Some embraced radical westernization, adopting European style institutions and even sending secret expeditions to study or cooperating with Europeans to smuggle in trade and books. Others instead reformed along traditionalist lines, seeking to strengthen traditional institutions and reform the state into earlier, superior forms. A few others took other paths. I was fascinated by two that took opposite paths: one state emulated Great Britain and another emulated Tang China. Both, of course, adapted it to their local conditions. Which fared better?
They both hugely outperformed expectations. A westernized army and economy apparently was about even with a modernized version of the Tang militia and economic system. The states that did poorly were the ones that failed to reform. All the states that reformed in any direction did pretty well, often in proportion to how much they could reform those institutions.
This made me consider that the capacity to reform and make a program work might be more important than the actual contents of the program. And I often think this is an underappreciated factor of analysis. Libertarianism is always the correct response to an ineffective or incompetent government. If the government's an idiot, it's best the government do as little as possible. It's only when the government is competent that we can debate whether private or public options are better. For example, with healthcare, there are two questions. Firstly, all the policy we actually talk about. Secondly, is the US government actually capable of getting it done? Think of the healthcare.gov website. That was just a straight capacity failure. No one was sitting there debating the advantages or disadvantages of having a functioning website. The government just couldn't get it done. And I think we don't talk about that kind of thing enough.
My prediction is they are probably about the same but will have some statistical differences that are noticeable but not overwhelming.
70% that radical reforms did better.
I wonder how much of the positive changes came from the specific technocratic/policy reforms the French enacted vs. just destroying the existing systems and letting new systems blossom. Considering this all happened right around the start of the industrial revolution, it was probably a good time for radical change no matter the political or ideological reforms that were enacted.
To put it more bluntly it seems destruction forces progress, which will often lead to better outcomes, but I doubt many people would want to have their entire communities and livelihoods destroyed just because it'll improve economic conditions in the area fifty years later.
Registering a prediction of 30% probability that the reformed areas did better. I'm counting a case where they both did equally well as a "No" for resolution. Predicting relatively low since the title of the post contains "consequences" which has a negative connotation thus biases me towards expecting poor results in the reformed constituencies. No other factors went into this prediction.
Napoleonic does better by a relatively small margin. - 90%
It seems likely that there's a strong selection effect in that the reforms studied were the famous Napoleonic reforms and not some random radical reforms that went really badly and no one thinks about anymore. After all, the anti-radical-reform argument isn't that it's *impossible* for radical top-down reform to go well, just that such reforms don't *tend* to go as well as evolved institutions. It seems like picking one of the most famous reforms in history, which instituted the Napoleonic code that French law is *still based on*, is stacking the deck a little.
(My prediction was that the occupied places did better (60%), despite my general skepticism of radical reform. This was based on googling "Napoleonic code," seeing that it's famous and still in use in modern France, and concluding that it must be pretty good since cultural evolution kept it around this long.)
Prediction: comparing principalities that were reformed or not reformed has the same methodological problems as comparisons on any not-very-random sorting.
"Daron Acemoglu is the most-cited economist of the past ten years, and I've never heard anyone say a bad word about him"
For some skeptical words about Acemoglu's self-confidence in his own big picture analyses, see:
https://www.unz.com/?s=Acemoglu&Action=Search&authors=steve-sailer&ptype=isteve
My prediction: the invaded countries do better, 70%
Registering my prediction in advance: I predict that the non-reformed systems did at least 5% better (on whatever metric the researchers used), and I am 60% confident in this prediction.
> So the authors ask: did the radically-reformed polities do better or worse than the left-to-their-traditions polities?
My immediate reaction was "oh, come on, what do you mean by 'better'?"
And the paper makes clear that their measure was... urbanization rates. (They also looked at GDP directly, but it sounds like their GDP data was so fuzzy and unreliable that they weren't comfortable using it as a measure.)
I wish the post had given this more attention. "Radical reform increases urbanization rates" is a) kind of unsurprising (modern ideas correlate strongly with urbanization for all kinds of reasons), b) not a refutation of anything in Seeing Like a State (the author kind of hates cities, right?), and c) not a sufficient basis for the kinds of value claims the post and the paper's authors seem inclined to make.
In other words, this is framed as an objective qualitative argument against James Scott's thesis, but it fails to convince because begs the question by reducing the problem to "legible" statistics and values. You had one job!
Would it really kill economists to stick some reference to effect size in the abstracts of these things? If this were a clinical trial, and the authors were like "hey, good news, napoleolimumab increases modernization, now FDA-approved for your unsightly problems with Prussian stagnation" they'd have the common decency to say "10 years of french reforms increases the primary endpoint of urbanization in 1900 by 9% from 41% to 50%" But instead i had to hunt through the text for it like a goddamn animal (page 23, if you care. They also have a point estimate of 36% GDP increase.
But now that they forced me to go through the paper, some not-terribly informed thoughts:
1. I have the usual concerns about econ research -- were any of these analyses pre-specified? How many different analyses were tried before they went with this one, etc.
2. If I am reading this correctly, by 1850 no changes are seen. So all the positive effect of the new institutions is from 1850-1900. Interesting.
3. Riffing on '2' -- maybe this can be spun as another example of "industrialization changes everything" or "conservatism is a better default in the absence of massive scientific/technological change." Blowing up institutions in 700 AD does you no good, because there's no innovation to take advantage of, you just get chaos. Blowing up institutions in 1800 AD helps, because it enables social shifts to take advantage of new modes of production.
4. And just for honesty: my prediction was "can't discern an obvious effect" (which in retrospect was idiotic given that if it were a null effect it never would have been published)
Prediction: 80% that the reformed states surpassed the non-reformed states, but only for a short (no more than a decade) time period before competition forced the other states to adapt.
Keep in mind that WWII era Germany and Japan went through radical reforms in the late 19th century, so I doubt they can be described as organically evolved societies.
The problem with this is a limited metric being used to derive a conclusion about a complex whole. If your measure of "better or worse" is simply aggregate economic growth this is an unsurprising result, even to many on the "traditionalist/evolved institutions" side. The trads (including me, and Burke) would argue, though, that there is a lot more to the health of a society, human happiness, and the common good than GDP.
So the state looked at a legible system (the conquered villages) and an illegible system (the unconquered villages), compared then according to its metrics (GDP, economic growth, etc.) and decided the legible system did better? Color me unimpressed. Almost every such reform looks better from the state's POV - that's the whole point of _Seeing Like a State_.
As Burke said, "The Revolution did not make France free, it made France formidable."
Bonapartism offered a sort of cautious version of the French Revolution, a hard-headed military man's version of rational modernization. It remained highly influential around the world for more than a century, such as in Bolivarism in Latin America, Kemalism in Turkey, and Nasserism in the Middle East.
Was gonna say 80% that radical (but well-intended) revolutions make matters worse short term (10 years) and 60% that they make room for necessary but impracticable in current system reforms. But then it dawned on me that "better-or-worse" is too heavily depended on the deciding criteria to form a meaningful opinion. China does just great economy-wise, but human rights and pollution are better left unsaid.
This paradigm bothers me. It seems implausible that there is some general, usually-applicable rule that entrenched institutions/traditions are either better or worse than systems imposed through radical reforms. Surely, an entrenched institution can be effective and efficient or ossified and redundant. And just as surely, a radical reform can be well-conceived or foolishly overconfident.
If that's right, then at best the paper here proves either that the French Revolution's reforms were well-devised or that the systems they replaced were weak, or both. Since the old systems were typically similar, and the reforms were all in the same direction if not literally the same, it suggests that *in this case* the reforms were an improvement. I don't see how we can possibly generalize from that result. Isn't the issue always case-specific?
My immediate thought when asked to make a prediction was "on what metric?" That basically sums up my frustration with this topic in general.
Evolved systems will, by nature, optimize, but there's no reason to think they will optimize on the metric that "we" want to optimize. In that sense, the problem with Brazilia wasn't poor optimization, it was misalignment between the metrics of the designers and the populace.
The same problem faces most modern reforms. We have broad disagreements about the different metrics Even if we totally agree about what a reform would do, different people will feel differently about it due to their different metrics. No policy is positive on all reasonable metrics, so no policy is universally agreed to be good.
That isn't to say agreeing on what the effects a reform will have is trivial. Even as a non-economist, I'm comfortable saying today's global economic system is a lot more complicated than that of pre-industrial France.
Prediction at the point of dare, before reading the rest: the reformed zones did initially better, but over 10-30 years returned to baseline, and no statistically significant difference is observed. More or less based on the idea that people are stimulated by New And Different and everyone tries harder, followed by Reversion To The Mean -- over a modest time (like 25 years) people fall into whatever rut suits their nature. (Long-term change only comes from technological or environmental changes.)
Now to read the rest...
I feel like there's a good analogy with software systems. As large code bases grow, they tend to get messy: early architectural decisions are in conflict with new requirements; hacks and shortcuts never get fixed and snowball into monstrosities. So then there's always the question: do we rewrite from scratch and make it better this time?
The conventional advice is never rewrite from scratch (https://www.joelonsoftware.com/2000/04/06/things-you-should-never-do-part-i/). Sure, you'll design it better the second time, but it's going to take a long time before you work out all the kinks and replicate all the features in the old code.
Instead, the conventional advice is to always be refactoring opportunistically. Refactoring means you're not working from scratch: you're taking what exists and re-architecting it into something better. You might rewrite a chunk of code, but since it's a small piece, it's easier to be sure you're not missing functionality that already exists.
Although the opportunistic refactoring tends to be the superior approach, setting up incentives to make it happen is tricky. With pressure to meet deadlines and little recognition for it, refactoring often doesn't get done. When it comes to government, I think we see the same effect: the news doesn't cover small changes and politicians don't get a lot of credit for it, so there's not much "refactoring" of laws or institutions that gets done. Cruft accumulates, inefficient processes ossify, and eventually we get so fed up we want to do a "rewrite".
75% reform. I think a lot of this is based this off the most salient data point in my mind: Japan, which shouldn't necessarily generalize to all forms of foreign-imposed reform. But it's clearly a good example that such a thing is possible and often overlooked.
Two hundred years of historical changes (in France, no less) summarized in one blog post and I'm supposed to come to some kind of conclusion about our culture?
The (so-called) Glorious Revolution of 1688 made England more economically dynamic by introducing modern practices from Amsterdam such as corporations.
The Republican Revolution of 1861, made possible by Southern Democrats walking away from their Senate majority, allowed the passage of many modernizing pieces of legislation that forward-thinking pro-business politicians like Henry Clay had been talking about since sometimes as long ago as 1812.
18th Century Enlightened Despotism in Central and Eastern Europe, carried out by pragmatic conservatives, was fairly successful as well.
>So the authors ask: did the radically-reformed polities do better or worse than the left-to-their-traditions polities?
I'd guess the results would be dependent both on how long after the "conquer and reform" we look, and on a metric I don't really have a better name for than 'core v. periphery'. Conquest is inherently destructive, and the transition period is unlikely to create prosperity out of thin air whereas it can definitely do the opposite. On the other hand, I'd absolutely expect it to lead to massive gains in the medium-to-long term as the new philosophies both have greater freedom to adapt to new circumstances and simply because they're guaranteed to be replacing something that has failed in at least one key metric. I'd expect it to look something like an accelerated industrialization, which might be difficult to disentangle because if we're talking Napoleonic it'd be happening at the same time as 'conventional' industrialization. The destructive effects might be a little weaker at peripheral regions but the positive significantly less as well.
Hard to put a probability on something I'm guessing is going to vary in both space and time, but to take a swing at it: 80% of significant improvement in the 'core' regions on a timespan of up to a century. Comparatively little change on the periphery, maybe 60% chance of improvement in the same time scale with a preceding decline?
One thing to keep in mind about Edmund Burke was that, while he had a practical side, he also had a Romantic side. He was talented at romanticizing what was in the interests of the rich and powerful. Rationalists tend to be out of touch with the Romantic movement that arose from the mid-18th Century to the mid-19th Century in reaction to 18th Century Enlightenment Rationalism, although it led to some of the greatest works of art of all time.
Guessing the reforms were good, mostly because countries are still described as having "Napoleonic" civil law at the present time.
I would hazard a guess that the radically reformed places did better, but I'm pretty unsure about this. Maybe 60%? That said, there definitely do seem like some confounding factors. Napoleon's empire didn't last forever, and when it was falling apart, maybe that affected places in it differently? Could be that the chaos hit there worse, could be that they started invading neighbors to recapture some of that early expansionary spirit. And when people were fighting against Napoleon, they'd probably treat the areas he conquered differently from the places downriver that weren't. Maybe?
My uncertainty stems from radical reform seeming dangerous, but the status quo being not ideal either, and my general ignorance of European history.
I feel like the call for predictions needs to be amended with dimensions... Like better on what axis? Wealth? Infant mortality?
I would say the reformed ones became slightly wealthier, but that may be tempered by the economic losses of being conquered and the associated destruction.
Also, I know almost nothing about Napoleonic legal codes.
I think what I got most strongly out of Seeing Like a State was the language of legibility and how that drives State reform. (J. C.) Scott selects certain schemes to increase legibility that fail catastrophically, of course, as the subtitle makes explicit, but the deeper concern I see from him is over _what gets lost_, which is illegible, and therefore never can be counted as a loss (which this study, of course, cannot account for, looking as it does at the exact same legible metrics that Napoleon was looking for). Of course, his examples are double-bounded, because they have to be _both_ examples of catastrophic failure _and_ examples which we can look back on legibly now and say 'oh yeah, shit... that was a terrible idea.' Which also means (and J.C. Scott is explicitly concerned about this) that those examples are poor examples anymore because we know so much more now and can avoid the mistakes...
In this respect, (J.C.) Scott seems similar to McIntyre (who, remember, was a diehard Marxist before he became a Thomist and Catholic–and I think it shows) and his After Virtue. You lose things, many of them perhaps permanently, with radical reform and centralization, you even lose the language in which to express and, finally, understand what has been lost and why it matters.
Samzdat's read of Seeing Like A State grapples more directly with this, as does most of Lou's writing there.
I think I disagree with Scott's initial sorting. Placing evolved institutions on one hand and complaints about vetocracy on the other strikes me as not wrong exactly, but incomplete. I think it's pretty clear that both threads have their answers to vetocracy, and I think they both recognize the problem. Vetocracy, after all, frequently prevents organic evolution of communities, societies, and institutions. Granted, I'd agree that erring on the side of evolution is more likely to result in the rise of vetocracies than centralized systems (since a vetocracy can be an organic evolution in the first place, but then prevents evolution).
I think this is borne out by the remainder of the essay. Scott points out the counter-example to the trend identified by Acemoglu as being the UK, the font of all modern organic evolution ideology and one that proves it can produce outcomes as good or better than any other system. I agree.
The difference is that the organic evolution in the UK tended (for reasons that are probably impossibly complex) to minimize or at least control the growth of vetocracy and the other malefic consequences of organic reform. The ossified German states that escaped the table-flip of the French Revolution had not just been entirely captured by the malefic side of evolution, but their ruling elites saw the maintenance of such as existential (even when their neighbors and social equals down the river and across the border did better in the long run and often in the short, too). As a result, and given the complete capture of political, economic, religious, and (frequently, but not always) cultural power in those small states and estates, the organic evolution that produced that elite class killed evolution.
And yes, of course, the clear liberalism of revolutionary French institutional reform is fairly obviously good from a standpoint of promoting human thriving (probably showing my bias here, but the massive growth of surplus and the the rapid reduction of poverty and all other forms of misery worldwide since the American/French/Industrial Revolutions seems like a track record well worth owning). Returning to theme, the UK had evolved most of this liberalism organically and didn't need a revolution to instantiate it or permit the industrial revolution to take off.
Tangent-ish: one of the things that fascinates me about Napoleon was the union of the liberal and totalitarian in one person. He was stridently in favor of liberalizing as far as possible (by the understanding of his time) all forms of ordinary day-to-day to existence, mainly commercial, social, and religious. On the other hand, he would tolerate absolutely no dissent from his subjects nor any neutral power in Europe. He had no use for civil or political liberty of any kind, and would imprison, torture, and kill any who dared to think independently in public. He could take criticism privately, though. But that was mainly from the officers he lived with on campaign, and on military matters only.
And then on the other hand, you have some places where some extremely radical reforms were imposed technocratically, and the results were terrible. Scott name-checks collective farming, but let's not forget Maoist China and the Khmer Rouge. So I disagree that the radicalness of reform has much to do with success. I think it has to do with correctness of reform. This would apply to post-fascist reconstruction of Europe and Japan.
By the same token, let's not forget the violent reaction of the European people to even the very modest (on average) counter-revolution imposed by the victors in the Napoleonic wars. That (along with the revolutionary forces unleashed by previous Ancien Regime excess) led to a century convulsed by many and serial (of admittedly varying intensity) revolutions in France, Spain, Germany, Russia, the Ottoman Empire, Hungary, Poland, and even the UK (Chartists, etc.) That is, those were relatively non-radical reforms imposed fairly gently (again, on average; every counter-revolution was different), but they were clearly wrong, and so some more heads rolled. I think that had to do with correctness more than radicalness.
Final note: the economic core of Europe. By some analyses, France was not (and is not) in it. You might have heard of the blue banana. https://en.wikipedia.org/wiki/Blue_Banana. It's a theory (that may have been concocted by a French dissident to critique his government's policy) that there's a large amount of easily accessible coal and iron in defined areas of Europe, and this was a primary driver of industrialization. The parts of France within the banana (Nord Pas-de-Calais, Picardie, and Alsace-Lorraine, primarily, and Rhone-Alpes secondarily) were/are some of the most industrialized areas of the country.
I don't know whether I agree with it, but it does seem that, after the end of the Napoleonic wars, France entered a sustained period of economic, political, and (especially) demographic decline, relative to the other nations of Europe. France's political unity and favorable position on the Atlantic coast allowed it to punch somewhat above its weight at one point, in seizing a large overseas empire during the New Imperialism, but other than that, it was fairly downhill until today, having joined the ranks of the Europoors with Italy and Spain. I'm only partially kidding about that.
What makes me skeptical of the theory is the motivated reasoning of its originator, whose purpose was to criticize excessive centralization in Paris, outside the banana. It still rings true to me, but I don't have a ton of confidence.
To the extent that the foregoing is just a re-writing of Scott's essay, I can only say I started writing this about half way through reading it.
From a position of ignorance, I would guess that the ones receiving the Napoleonic code were ~70% likely to do better. It seems like codes of laws that stuck were mostly centralised and European principalities weren't small enough to benefit from the effect of being a small community where everyone knows each other.
Radical reformers improved legible metrics like GDP per capita, lumber per acre farmed, and taxes collected. However, other metrics like quality of life, lifespan, and subjective happiness decreased. The length of time that the radical political was in effect did not last longer than the amount of time it was implemented (using Napolean's exile as 0 on the time axis).
65%?
My guess was 74% that the radical reforms would be net positive. In retrospect I think this was mostly a failure; given what I knew, I should have been able to be more confident in this.
Re the more general debate between radical reformers and traditionalism, I kind of like the analogy I made of society as a damped harmonic oscillator, which is probably the most stereotypically rationalist thing I've ever written
https://shakeddown.wordpress.com/2018/04/22/modeling-society-as-a-damped-harmonic-oscillator/
I think the discussion gets muddled by trying to compare such materially different places as Brasilia and Napoleonic France, especially in light of the stark differences between their stated goals. Comparing England's process of industrialization with that of the USSR and Communist china might be more fruitful, especially considering that Mao's revolution and later reforms were all (ostensibly) done to benefit a peasant class that existed under a very old-school feudal system.
But if all the successful instances of rapidly 'reforming' longstanding systems(post-war Japan, Napoleonic France, Germany etc) all basically neoliberal , then that tells us much more about neoliberalism than technocracy itself.
My take away is that the question "should my priors favor evolved systems over designed ones" is kind of silly, because a given policy's origins are completely incidental to its results.
Acemoglu measures the current status of conquered lands in 1850, 35 years after Waterloo, which was, apparently, long enough for the bad effects of the breaking things and killing people to recede.
Some of the shorter term effects of the Napoleonic Wars were dire. For example, the chaos caused by Napoleon's failed invasion of Russia in 1812 led to the breakdown of government wolf control efforts across Eastern Europe, which led to wolf packs running amok for a few decades until they could be brought under control again.
The Napoleonic Wars had the good side effect of being so bad that they discouraged Europeans from going to war with each other all that often from 1815-1913. They had the bad side effect of entrenching Reaction in control of much of Europe, discouraging British-style cautious democratic reforms until the second half of the 19th Century.
In Switzerland, most historians probably agree that the reforms imposed by Napoleon had an overall beneficial long-term effect. To some degree, elites the French drove away from power managed to return in some parts of Switzerland, but a return to the outdated political system was impossible and beneficial compromises were found. I suppose that apart from Switzerland, other areas profited in a similar way.
But I think we should be careful to extrapolate from this. One important point is, of course, that France was culturally very close to the European countries it modernized, and the French people who took an active role had intimate knowledge of these countries. I would suppose that the track record for imposed reforms („nation building“ etc.) with a much greated cultural distance is much worse. Sometimes, it can go well, but in many cases, it does not.
The other important point, which is also emphasized in the text, is that the reforms conducted under Napoleon were actually rather moderate, as far as their content was concerned. They did not aim at some far-fetched utopian goals. They may have been more radical than comparable cases in how much energy they put into removing old elites and outdated structures, but what they replaced it with looks rather moderate from today‘s perspective and probably also from a contemporary perspective.
"The evidence suggests that areas that were occupied by the French and that underwent radical institutional reform experienced more rapid urbanization and economic growth, especially after 1850. There is no evidence of a negative effect of French invasion. "
Let us see. The Franco-Prussian War, World War I, The Russian revolution, World War II, The Holocaust, the Iron Curtain. The Collapse of the Soviet Union. The butcher's bill, say on the order of 100 million.
Napoleon should have stayed home.
That is a hard pass.
Rather than looking simply formally at "reform or not," it's important to look at the substance of a reform. The Napoleonic Code was strongly influenced by the Justinian Code of the 6th century AD, itself a codification of older Roman Law. So while the Napoleonic Code may have been "new" to the little duchies outside of France upon which it was imposed, it was hardly radical from the perspective of Western legal tradition. Perhaps the only lesson here is that, as legal codes go, the Napoleonic Code was pretty good. This would also accord with the non-correlation of Great Britain, which already had a good legal system.
The first thread ("institutions have evolved to something like a time tested optimum and reform is hubristic, etc etc.") suggests an essentialism that plagues a lot of conservative philosophy. As if "work ethic" and "value" and "character" and economic outcomes were perfectly fixed features of physical reality and not dynamically shifting systems interacting and changing each other. There's no reason that the territory or the agents within it can't shift and mutate and render what worked however long ago obsolete. An organism can be perfectly evolved into its niche until one day a random amino acid gets swapped in a virus genome then suddenly that organism's time-tested mating ritual spreads a new deadly illness. This seems to be more or less what happened with anti-maskers and covid, the notion that something could shift so radically in the environment and suddenly make a lot of indoor commerce dangerous is not just unthinkable but *laughably* ridiculous to some people.
I think about this whenever I hear Lincoln Project conservatives like Tom Nichols smugly dismiss younger generations as entitled and whiny. As if it were like a cosmological constant that x units of Hard Work = housing/a car/stability or whatever and there are no larger forces reshaping the relationships of people to capital.
A personal pet peeve of mine is the assumption that unintended consequences and perverse incentives, the things that are supposed to render reform and central government action unworkable, are exclusive to reform. For example the very notion of "private property" seems to work very differently in theory compared to practice. In the libertarian ideal, in theory there'd be some sort of maximal respectful non-interference with the property of others barring extreme externalities, etc. In practice homeowners aggressively leverage whatever government power they can muster to shut down the construction of more housing on other lots of private property anywhere near them. In practice the idea of property in peoples' heads seems to be something like: "This property is my investment, I can and must do whatever is necessary to protect its appreciation, including restricting the supply of nearby housing through government force." Nimbyism is bipartisan but its especially ironic that many of these people consider themselves ardent capitalists.
Unfortunately, I cannot guess in advance because I'd already blogged about the paper. Multiple times, in fact!
https://entitledtoanopinion.wordpress.com/2011/11/14/chris-coyne-sort-of-responds-to-acemoglu/
Early on I was keen on Acemoglu as someone digging into the questions Mencius Moldbug was interested in, and even recommended his work to Moldbug. Later, around the time of Why Nations Fail, I became much more skeptical of his work (and Johnson's).
https://entitledtoanopinion.wordpress.com/2017/09/04/stephen-broadberrys-accounting-for-the-great-divergence/
I can't find the post where I made this critique now (it might have been in the comments) but I recall them supposedly teasing out the effects of colonialism... by using malaria as a proxy for certain kinds of colonial institutions, rather than comparing Ethiopia & Thailand to neighbors or even considering the possibility that malaria's direct effect was confounding their results. If you trace back the links on that Coyne post though, I do note that the Acemoglu paper is in the tradition of Mancur Olson, who was also writing after Germany & Japan surged after WW2 (and in contrast to someone like James Scott). Greg Cochran would say that's because places with high human capital can rather easily recover from destroyed physical capital.
I predicted the non-reformed areas would do better. Looks like I was wrong, but how wrong.
I have some unformed ideas about the difference between being conquered by an external force and having a reforming faction win from within the existing society. As the review points out, post-conquering, if the new system is "good" it will work well and if not it won't. The given examples (Germany, Japan after WWII) are external-force examples and I suppose the Napoleon examples are external-force as well, though some of the same duchies were probably once under the control of Charlemagne, so maybe more of a high-water mark of French expansion than external-force per se.
Contractors seem to do well after and during military action. There are things to rebuild and replace. I don't know how to quantify "GDP due to replacing what was destroyed" versus "really new growth."
Civil wars to institute "reform" seem to go differently (even the French revolution came apart after a few decades, hence Napoleon). When a different societal faction becomes dominant, the now-subordinate faction will potentially not cease to exist after all and there may be ?a greater tendency to authoritarian dynamics to keep the new social order in place, which might stifle growth as well. The status quo ante may be closer after a civil war.
Good food for thought.
People who keep mentioning Brasilia: try working for a month in Rio.
The French Revolution wiped away a lot of the remnants of feudalism, a system that had emerged 1000 years before to provide local security against raids by Vikings and other Roving Bandits by allowing Stationary Bandits to set up to defend local turf. It worked pretty well under the conditions of the late Dark Ages (e.g., terrible roads). But by the 18th Century, feudalism was obviously outdated, as the more recent Absolute Monarchs insisted.
And the feudal system of layering complex bargains on top of other complex bargains and so forth and so on for a thousand years had made getting much done legally awfully complicated. The British had a system of allowing judges to creatively misread the legal record of the common law, but more conscientious Continentals tend to be bogged down by the weight of the past. So, the streamlined Code Napoleon was welcome. But keep in mind that the Code Napoleon wasn't all that radical. It reflected the neoclassical feelings of the age, that this is what the wisest Romans would have done if they were around today.
My primary worry about the radical reform paper, which you touched on but didn't state explicitly, is that if these places had similar pre-existing institutions and economic circumstances, and France enacted the same reforms in all of them, then of course there was a consistent pattern, and this is kind of like having a sample size of 1.
Their claim that reform failures can be blamed on not being radical enough seems absurd. All the most famous examples of top-down reforms gone catastrophically wrong were pretty radical, no? Or was the problem with the Great Leap Forward that Mao wasn't thinking big enough?
I settled on the prediction that the French-invaded places were equally likely to have gotten better as worse. Once I saw the answer, part of my brain was very insistent that I was really leaning towards the reformed places doing better all along. Dammit, brain.
You need to read Scott's other book. "The Art of Not Being Governed." Seriously, with what's going on in Burma... it is important to understand why the Burmese state is having so much difficulty and the highland polities he talks about.
"So maybe the moral of the story is something like - replacing stagnation and entrenched interests with good reform is good, and with bad reform is bad. Which sounds obvious, but I do think that considerations of "is this potentially challenging a carefully evolved system of traditions?" is less important than I originally believed."
Isn't that a subset of the belief "life is complicated"? I'm being mildly facetious, but only mildly. The world when stacked end to end is full of disagreements. However, a lot of it boils down to "doing good things is good" & "doing bad things is bad", and then the rest becomes finding ways to make it easier to do good things, and harder to do bad things, and easier to reverse mistakenly doing bad things.
I know that sounds very over-simplistic, but there's a lot of muddling in the middle that makes sense, even if you have enough brainpower to chew on the big ideas.
Prediction before reading: 60% that the radically reformed polities did better.
> Second, isn't it sort of weird that Britain, the country that got least invaded by Napoleon and had some of the deepest-rooted institutions of all, was the one that really kicked off the Industrial Revolution?
My low-confidence understanding of this is that Industrial Revolution <= steam engine <= innovation spurred by having to get coal out of the ground <= deforestation leading the British to investigate costlier alternatives. So you could say the cause is material. But I've also heard that Britain emerged from the mercantilist era with the largest 'innovator / merchant' class – I believe it saw the greatest conversions of the aristocracy into an entrepreneurial class of the European nations.
75% reform, because Doylistically, you wouldn’t be posting about this paper if it didn’t contradict Scott.
This sounds a lot like survivorship bias: if the revolution lead to a huge improvement (define improvement...), then the researches give it some attention and include in the study, but how many revolutions there have been that didn't really change anything?
No significant effect/no difference made by Napoleonic laws. 0.65
Doesn't the presumption that these systems "evolved" misstate how evolution works? Pre-some kind of phase change, like revolution, what you have is a system that is extremely well adapted to those who have held power, but is not particularly robust in the face of any significant change in circumstances--either from an internal or an external force. For it to be evolution, you'd have to have some kind of driver for fitness, which I would argue effectively doesn't or only minimally exists in monarchies/dictatorships. The argument that monarchical systems which have built up over time because they've existed for a long time are somehow evolved takes a horribly cramped view of what drives evolution.
As for the UK driving the Industrial Revolution, the UK had already had its revolution--in the English Civil War, and in the Glorious Revolution of 1688. So it's not surprising, if were using revolution-driven "evolution" as an indicator, that the Industrial Revolution started in the UK.
Conquered better: 80%
Conquered worse: 15%
Inconclusive/other: 5%
Prediction: 80% the ones taken over by Napoleon did better. Reason: feudal system is trash and so the bar is already pretty low. By destroying the existing power structures you're mostly removing assholes that are making things worse.
I may be in danger of making the hypothesis untestable, but what hypotheses exactly are we comparing? The French-occupied lands are counted as instances of top-down design, and others are considered bottom-up? The post points out that some of the French reforms actually made things more bottom-up. Was Roman law completely eliminated from the French reforms, or just the parts that seemed not to fit?
I'm also going back and forth on using GDP as the measure. What other confounding factors might influence that? What measure would I use, if I could use anything at all? Utils? How long does it take the higher growth rate to amortize the blood and treasure spent to put the French in power? Is it possible to get the top-down reforms without paying the blood and treasure? Might there be less costly ways of destabilizing bad equilibria? Yes, I think I have a bias showing.
It seems possible to me that a poorly governed city might still have advantages for growth compared to the countryside, while coming out behind a comparable well governed city. If the laws are clear, stable, and fairly administered that might matter more than their origin, unless we think top-down/bottom-up delivers advantages in clarity and honesty.
Even good ideas can be implemented badly. I think what matters is whether people can innovate and adapt. Have I learned something new, or just moved the goalpost?
My advanced guess was that there is no significant correlation. But yeah, at least someone asked a testable question, and tried to answer it instead of logic about it, the pitfall LW-infused rats and postrats tend to succumb to.
Scott: in the future, would you consider leaving a top-level comment to which everybody who wanted to leave a prediction could reply, in order to collect and contain the predictions?
My prediction was 70% of conquered territories would do better.
I admittedly have read the entire article, but I did register a prediction (to myself) that the replaced systems would do better. My (no expertise) gut reasoning is that lots of economic things work better at bigger scales, and that the benefits of having a common legal/economic system "gifted" to you by a larger power probably has positive effects, regardless of the actual merit of those systems.
Preregistered prediction: Napolean-affected areas did better on average. 80% confidence.
>>Are you sure you want to keep reading now? You’ll never get another chance to predict this from a position of ignorance!<<
Aarggh!
I truly don't know the answer.
In my youth I would have sided with the reformers. I had lots of solutions for the ills of the world. I still have solutions, but now I'm much more cautious about imposing them.
I am going to predict that the radically-reformed polities did better… not because of any philosophic leaning but because of what little I know about Napoleon's reforms
>>The idea was that the occupying American forces couldn't care less about the entrenched power structure and vetocracy in Germany and Japan, so they rammed through whatever reforms seemed like good ideas at the time, and they were in fact mostly good ideas. On the other hand, the Soviet Union tried the same thing in East Germany and that went less well.<<
It is reasonable to presume that what you are imposing matters. In this case capitalism is better than collectivism, at least as measured by economic growth.
Theory: Being conquered for a short period can be helpful but only if the conqueror withdraws after breaking up the local monopolies.
I think Seeing Like A State only applies to complex systems, which resolves some of the apparent contradictions here. Complex systems have many interacting components. Cities and agricultural systems, profiled in Seeing Like A State, are good examples. Top-down reform doesn't work in complex systems because they are inherently difficult to model. Breaking up oligarchies isn't in the same category. Economists can confidently recommend breaking up extractive institutions because the first order effects are so good. The second order impacts are just a rounding error. In contrast, a city is a complex system that is all about second and third order impacts. Evolved institutions can be irreplaceable in complex systems, but aren't necessarily better in other settings.
So can we find an analogous situation if we examine the effect of imposing Roman Law?
Originally 0.7 on conquered dutchies outperforming non conquered ones but flipped my prediction after misremembering the title of this post to be "The Cost Of Radical Reform".
If we think of this as analogous to searching for a global minimum when one is currently stuck at a local minimum, then the question of whether radical reform is a good idea can be rephrased as asking whether we should try a large jump or a small one. There is no fully general solution to this problem, but we do have some heuristics.
One heuristic is, "How bad is our current equilibrium? Is it more or less ok, or are people really miserable?" This lets us distinguish between the cases of a healthy neighborhood where Robert Moses wants to build a highway and a crappy principality full of oppressed peasants.
Another, similar heuristic would be, "Are the people resisting change a small elite or a large underclass?" A small elite may just be trying to hold on to their rents, but broad support suggests that the current system is actually working reasonably well and shouldn't be lightly discarded.
I'm not very good at thinking about probabilities for my beliefs, but I'm thinking something like:
20% the radical reforms fared much better
60% the radical reforms fared slightly better
20% the traditional systems fared slightly better
Another hypothesis: Bottom-up evolutionary changes work better when the social/political/economic environment is changing fairly slowly, but top-down revolutionary changes work better when the environment has changed enough to render the current systems dysfunctional. Seems like this hypothesis is consistent with the findings: Napoleonic Europe wasn't exactly the most functional environment, so maybe the places that had French soldiers marching through had the advantage of clearing the dysfunction quicker.
Another observation: Napoleonic France had a wealth of examples of Things Not To Do in the ten years preceding the Coup of 18 Brumaire. (Example #1: The French Republican Calendar, where "Brumaire" was a thing.) So maybe the Napoleonic code had the advantage of a revolutionary framework, with most of the revolutionary rookie mistakes having already been made.
60% confidence that the reformed lands do better. However, I think this moment in history was somewhat unique, as it was on the eve of the industrial revolution. Going into the industrial revolution, it would have been uniquely beneficial to have a common legal system with one of the most powerful economic actors in Europe. So, while I think it's likely the reformed countries did better, I think we have to be a little careful about abstracting too much from it.
Feelings Bold - 60% that being conquered by Napoleon is positively correlated with GDP in the present day. Let's say, 10% better on average.
Going back to the post now.
0.5 that conquered and reformed states did significantly better.
0.3 that they did only slightly better.
0.2 that they did worse.
Might be worth pointing out that Napoleon already represented a step back from the most radical excesses of the French Revolution (no "rational" ten-day weeks or temples of Reason).
<i>Europe at the time had so many tiny duchies and principalities and so on that you can actually do a decent experiment on it - for every principality Napoleon conquered and reformed, there was another one just down the river which was basically identical but managed to escape conquest.</i>
I don't think this is accurate -- if you look at the map of Napoleon's conquests ( https://en.wikipedia.org/wiki/First_French_Empire#/media/File:French_Empire_(1812).svg ), pretty much all the tiny little duchies ended up as part of Napoleon's empire.
Also, the German Confederation (basically the post-Napoleonic equivalent of the Holy Roman Empire) passed various economic reforms, most notably the customs union of 1834, which would have disproportionately benefitted the smaller German states -- states which, due to their size, had also proved easier for the French to conquer. I guess you could say that this was indirectly due to Napoleon, insofar as the German Confederation wouldn't have existed without him, but still, the appropriate conclusion wouldn't be "Radical, top-down reform is inherently good," but "Free trade is better for the economy than a series of tiny, protectionist micro-states."
Prediction comment:
I want to hedge and say that in short-ish term unreformed were better and in long-ish term reformed win; 0.7 for either of those in separation.
"So the authors ask: did the radically-reformed polities do better or worse than the left-to-their-traditions polities? ... If you're really feeling bold, post a comment with your prediction before reading further."
To predict which will do better, I'd need to know how the authors will measure "better". I predict that the areas brought into the larger state with a unified legal system and government will do better economically, partly because the Code Napoleon was a pretty good code, but mostly because it's a lot easier to do business with people who have the same laws as you, and where you know how or whether law and contracts will be enforced.
But when I've read about various people living under Napoleon's occupying forces, they generally came to hate the French. So I'd guess that those areas were less happy.
I am, however, skeptical that the experiment can actually be run as described, because I've never heard anything about there being unoccupied pockets within French-held territory. And if there were, there must have been something unusual about those pockets, such as being too remote, inaccessible, or poor to bother with. That would wreck the experiment.
70% confidence that reform outperforms tradition.
prediction: Napoleonic reforms / modernisations improve things drastically in the longer run - 80%
I predicted 80% conquered areas do better.
Integration into the empire's economy, language & market were my considerations (over the non-immediate term).
Relatedly, on colonialism there's evidence (from another paper by Acemoglu et al., though they tell a different narrative, and maybe other papers) that, in the long term, countries never colonized do worse than those colonized. I think being conquered over land where your population is integrated into the broader system is beneficial (e.g. Russia in Siberia, or Napoleon) whereas being exploited by a handful of officials (e.g. Congo, Hispaniola) is certainly less beneficial, maybe even detrimental.
Yeah well, you addressed the key concern in your writeup. Here's another prediction for you to make: suppose you successfully spread the meme that "radical reforms" are good, and "designed systems are better than evolved ones." This is true because "there was a study."
Do you think more free markets are likely to come of that?
This makes sense. I think Scott would agree that the larger "local knowledge gap" is not between one nation state and its neighbor, but between the hyperlocal community and the nation state it belongs to.
Maybe national governments are say 95% interchangable with their neighbors, and if the French policies were already proving their worth back home (compared to their old system), they would be quite likely to do better than their neighbor's old systems too.
Prediction: reformed is only slightly better on average (60%), and has lower variance (ie both the best and worst systems are traditional) (70%)
Did they measure any social outcomes? GDP growth doesn't give a conclusive answer on whether it was a desirable outcome, and urbanization is not inherently good. Cities are dysgenic fertility shredders, so in the long run when all of Europe has either adopted these reforms or they've become irrelevant, then earlier urbanization should have negative effects.
Was too immensed in reading to think of a number, but I was pretty certain the invaded looked better. Retroactively, I'd rate myself at around 80%. It's in fact perfectly consistent with my priors. Authority structures ossify to serve the people on top, overthrowing them makes way for progress. Obviously.
I think the problem with contrasting economic accounts with Scott is that they're concerned with different things. Meaning, not mutually contradictory. Economists study "legible" data like GDP, Scott looks at the wellbeing of the downtrodden. It can simultaneously be true that the business is booming and that the people are immiserated, and the texbook example of this is precisely the (England's) Industrial Revolution. (Which we do have legible data for. People were shorter, their lives were shorter.)
The takeaway from Scott is not to cease progress, but to watch for the little men that get trampled by it. The two views can be reconciled by not trampling the little men in the first place. To be more precise, by carefully rejecting easily-ossifiable and exploitable structures of authority so that the little men don't have to carefully weave their lives around their idiosyncracies only to get them thrown into disarray by a sudden disruptive wave of necessary but careless top-down reforms. Scott himself is an anarchist, which, obviously. (Needless to say, so am I.)
(To extend a hand to the pro-capitalist libertarians which make up like a half of the readership here, I commend you for sharing the above intuition. You just need to stop refusing to accept that capitalism does not solve this problem, in fact, it's itself easily-ossifiable and exploitable.)
I predict that radically-reformed areas perform better, 70% confident.
In a time of rapid technological change, destruction of the old order will usually be a net good.
Soviets are a special case. In their homeland they replaced a +/- feudal system with a quasi-modern one, which had some beneficial effects. In the states they invaded, they replaced an emerging modern free-market system with a planned economy, with disastrous effects.
So, it depends if you're upgrading or downgrading the social system, I guess, and you have no way to tell if both are recent.
My probability was .7 that the reformed ones did better, for essentially the same reasons of legal cruft and rent-seeking.
Perhaps big picture changes, like opening up markets, work well when externally imposed. But small scale bureaucratic changes are also necessary, and these work best when there is time and flexibility for them to evolve to fit the local cultural norms. Just look at how democracy ( a great idea) didn't work well in Arab Spring countries.
Ok I'll bite: Before reading past the "prediction pause", I predict 60-40 that Napoleon's radical reforms did worse than the traditional ones.
My reasoning? I mostly buy the Seeing Like A State-style arguments for bottom-up, illegible policies; on the other hand, much of what I've heard about Napoleonic reforms--including driving on one side of the road only--seem very good to me. But back on the first hand again, I might expect hear about the best Napoleonic reforms and not to hear about all the missteps. So, 60-40 against radical reforms.
(Also, I'm a physicist with little background in history, so my confidence is not much better than 50-50.)
Actually, despite systematic anti-Brazilian propaganda being dispensed through the world for unconfessable reasons, Brasília is widely considered "a singular artistic achievement, a prime creation of the human genius" and, accordingly, was chosen as a UNESCO World Heritage site.
https://whc.unesco.org/en/list/445/ .
I wonder to what extent this is an artifact of Napoleon's defeat at Waterloo. To explain, some Dutch history:
From the latter part of the 17th century, a group of wealthy merchant families were able to mostly reserve Dutch government offices to themselves, resulting in a ruling class of 'regents,' who served their own interests. The result was dissatisfaction among the middle class, who revolted twice during the 18th century. The second revolt was put down by Prussian intervention. At this time, The Netherlands was a client state of the English and the Prussians, with immense debts, which were paid for with extreme austerity and high regressive taxes, while an elite was extremely wealthy. In 1794/5, Napoleon managed to conquer the country, helped by revolts against the regents. This resulted in The Netherlands becoming a client state of the French. It was not really ruled by the French, but nevertheless instituted huge reforms, like forming a unitary state, adopting some democracy, adopting (progressive) income and wealth taxes, etc.
Napoleon's main interest in The Netherlands was to get money to fuel his wars, military support and a place to launch attacks on Britain from. The Dutch pushed back against this, eventually making Napoleon so angry that he forced the Dutch to accept his brother as king: Louis Bonaparte.
However, Louis turned out to have ambitions of his own: to be a good king who served the interests of the Dutch. He adopted the Dutch nationality and demanded the same from his ministers. He did his best to learn Dutch and made Dutch the official language of the court. He didn't crack down on smugglers that evaded the Napoleonic economic blockade of Britain, which helped the Dutch economy a lot & he refused to institute conscription. He also refused to write off 2/3rds of Dutch loans to the French.
All in all, he was a good king. Napoleon said of him: ″Brother, when they say of some king or other that he is good, it means that he has failed in his rule.″
Napoleon deposed Louis, but two years later he made the ill-fated decision to invade Russia, which in turn led to his later defeat at Waterloo.
If Napoleon had not been defeated and had been able to use The Netherlands as a milk cow for a long time, burdening it with debts, forcing its people to become soldiers, etc; wouldn't The Netherlands have stagnated as long as it remained under that rule? Perhaps the Dutch got very lucky by getting the best of the French reforms, but without suffering that much from Napoleon's more abusive desires.
So perhaps the lesson here is to let reformers institute their most sensible reforms and then banish them to Elba.
Meta-comment: a surprising number of commenters (including me) picked *exactly* 70% in favor of reform.
I chose it because it seems right for "strong intuitions, but epistemic humility because this topic is complicated". Perhaps others thought alike.
A histogram of peoples' prediction values would be interesting.
I said "Napoleonic code is way better, 80%". I should have said 90%.
Napoleon-conquered did better, 70%
I think some way to integrate those kinds of predictions into the post itself would be really interesting. Some widget to make a prediction on a (pre-defined) question after the introduction, and some summary of the community predictions at the end. I am not sure if the LW Style prediction widget would suffice already, my gut feeling is that only seeing the community predictions at the end of the post would be better, but the LW style widget might already provide some interesting insight, without spoiling the rest of the post. It could obviously be gamed, but you aren't really cheating anyone but yourself, so that doesn't seem like that much of a problem.
I have not read the paper, but considering that:
- Napoleon fell in 1815
- 1848 was a famous year of revolutions
- In 1861 Italy was unified and 1871 Germany was unified
I wonder how the paper could claim that everything good was just because of the French invasion rather than the a general revolutionary spirit or pressure to change.
As a comparison, the modern success of Japan is usually due to the fact that foreign powers colonized/influenced their neighbours, while they escaped conquest.
The deference shown to a "carefully evolved system of traditions" needs to consider whether the conditions under which the system evolved still hold. The early 1800s is around the time when a graph of GDP per capita throughout human history curves from essentially a flat line to a vertical one. Waterloo was in 1815 and the paper discusses growth post 1850, so it is likely that value of a French invasion is not the particular designed system they imposed, but the weakening of constraints on future adaptation.
I don't think it makes sense to have a fixed prior about the relative value of evolved vs. designed systems across all time and space without taking into account particular contexts.
My prediction was that areas conquered by Napoleon would have done much better. But honestly this wasn't much of a prediction, I spent most of my time trying to remember which areas Napoleon conquered. This includes most of the currently richest areas of Europe, making the answer kinda obvious.
"But then, isn't the whole point of Seeing Like A State that things which seem obviously true sometimes aren't?"
No, the point is that a top-down view misses local, tacit and embodied understandings. It only seems "obviously true" that rectangular grid squares are superior if you have a preference for rectangular grid squares. Very different things are "obviously true" from a local and embodied perspective.
"But then you would lose the right to apply Scott to most modern political debates, where there are no peasants to be found, and everything is the weird mix of extractive and altruistic typical of modern states."
Have you considered (the conflict theory perspective) that these "altruistic" policies are not altruistic? People get in a position of power and redistribute value from their opponents to their clients. Also, this is top-down imposed order. Some guys can obviously afford to pay taxes, other people are obviously too poor, lets put them in boxes according to how I evaluate their needs and surpluses.
You may find it relevant that GDP is an imperfect estimate of economic productivity. It seems easy to see why imposing a top-down system of legibility would result in a measured increase in GDP, for two reasons. You're moving value from illegible areas to legible areas, and you're requiring that that value be traded or documented somehow (whereas before value remained in the local illegible sphere).
0.7 radical reform worked
Thoughts: It's said the Polish peasants weren't really bothered by the Partitions. They were just under new management.
The Congo and India and the Soviet Occupation Zone were set up as extractive colonies, whereas the Americans understood that a strong Japan and West Germany were the keys to containing Communism. The DDR did later become better supported, but the damage had been done.
"Daron Acemoglu is the most-cited economist of the past ten years, and I've never heard anyone say a bad word about him"
I'm not - really - an economist, and from what I know Acemoglu is far from being incompetent, but his "Why Nations Fail" was really weird to me in that it very clearly split historical periods and personalities into good (inclusive) and bad (extractive). He also seems to jump to the conclusion that democracy is always definitely better for inclusive institutions, and IDK how much I believe that.
But in general, I kind of agree with the idea about radical reform failing when it doesn't go far enough.
Prediction: most likely (~80 %) no significant difference or somewhat mixed.
Regarding why Britain kick ass during the Industrial Revolution, it's worth reading Anton Howes' substack, since Britain certainly wouldn't seem to be the betting favourite before it all kicked off.
https://antonhowes.substack.com/
The reform principalities did somewhat better than the traditional ones, at about 65% sure. Lots of confounders and I am not sure how I would qualify 'better' - wealth is about the only objective measure.
Isn't the obvious explanation here that radical change comes with inherent problems, but that if the existing system is bad enough, it's still fairly possible to make large improvements? The French revolution swept aside feudal structures that were centuries behind their expiration date, and the Russian revolution might have worked out well if power hadn't ended up with people who were literally the worst (and even _those_ people managed to improve dramatically on industrial production).
Meanwhile, the combination of democracy, capitalism, science and markets mean that Western society seems fairly proof against that kind of ossification. In these cases, the problems with radical reforms would almost certainly outweigh the benefits.
I'm boldly saying, before reading, that the duchies Napoleon conquered did better. Now back to reading.
"Does radical reform help or hurt" is a very, very different question than "would the implementation of modern legal systems including the Napoleonic Code raise GDP and urbanization vs. traditional legal systems in European polities."
For the record, my prediction as to the first question is "I have no earthly idea". I wouldn't even know how what to include in a data set to try to build a prediction off of, nor would I be able to quantify a value judgment as to whether the outcomes of reforms were "good" or "bad".
My prediction as to the second question is "yes, probably." I'm still not really willing to put a number on "probably" without knowing more specifics. Low confidence, though - maybe 60% - but part of that is admittedly because I expect to see counterintuitive results presented in this forum.
However, I think framing the question this way kind of gives away the answer. I mean, we know that "modern" legal systems are at least correlated with greater urbanization, so it would be kind of surprising if someone found that modernizing and standardizing legal codes resulted in decreased urbanization. We also have a plausible mechanism as to why and how modern, standard legal codes facilitate people living in larger cities in which they are more likely to have dealings with lots of strangers.
Also, in terms of the object question, I think you can see a clear mechanism where having standardized legal codes might facilitate trade (this even shows up now, in things like the US not importing foreign medications, and supply chain problems stemming from Brexit). So I think it's reasonable to predict that this would increase GDP.
After reading the results, I see that my interpretation of the question didn't exactly match the study, since the authors seem to be talking about destroying the institutional power of elites - which isn't exactly the same thing as implementing a standard code of laws. But that's exactly why I was reluctant to put a number on a prediction in the first place.
Finally, I think it's important to note that the choice of metrics here really matters. Why should greater rates of urbanization be counted as an indicator of successful reform? "Living in a city" isn't a terminal goal for most people, right? What was life like in these cities, compared with life in villages? What happened to life expectancy? Leisure time? Quality time with family? Number and quality of relationships? Mental health? I imagine the results would be more mixed here.
I also wonder to what extent our choice of metrics reflects a kind of survivorship bias. Development as reflected in GDP and urbanization may have produced competitive societies that were able to outcompete "traditional" societies, but as a result when we try to define "success" we're defining it based on "which societies won this competition" rather than "which societies were good for the people in them." There's a risk, in other words, that we're choosing metrics based on what we think success looks like starting with the assumption that we're the most successful and building outwards from there. I think the Johns Hopkins report (Global Health Security Index) shows the pitfalls of this kind of thinking - the metrics they chose were heavily biased in favor of the US and other "WEIRD" countries, and as a result some people were blindsided by the fact that the US had one of the worst pandemic responses in the world.
So I think in some sense when we're choosing metrics based on what we think makes a society successful we're in effect asking "will adopting [reforms that make societies more like ours] have the effect of [making societies more like ours]". And in that sense, it makes sense that the more radical the reform *in our direction*, the quicker will be the resultant move in our direction. But this tells us nothing about where future reforms will move us to.
Prediction: Frenchies did well on whatever metrics the authors select. This is based on two things: one, Napoleon as a "make the trains run on time" sort of guy, two, we say "metrics" in this context, not "measures" – I presume there's a good reason the French word won out. Or, to summarize both of those things: seeing like a state actually works, in that a strong state with a powerful executive organ can actually get things done. 80%.
The reformed places did better. (85%)
If nothing else, I know from history that a lot of places (especially in Germany) did not undo the reforms after Napoleon was defeated, and other unconquered places tried to imitate them. So this suggests that what Napoleon did was valuable.
On a theoretical level, we should expect the politic system a society adopts to depend on the rate of material change within that society. If your society changes less over time, you can have more veto points, because the slow leakage of ideas through the system matches the input. If your society is constantly varying, you should see fewer veto points, because folks are constantly inventing new carriages/methods of piracy/articles of clothing/siege strategies etc., and you can't afford to wait to respond to them. And the Napoleonic conquests happen to occur right about when the Industrial Revolution hit Continental Europe. So we should expect the ancien regimes to be adapted to the slow hierarchically-constrained change discussed at https://slatestarcodex.com/2019/04/22/1960-the-year-the-singularity-was-cancelled/ and https://scholars-stage.blogspot.com/2010/08/notes-on-dynamics-of-human-civilization.html. Conversely, Napoleonic codes better match the pace of factory innovations.
Now I'm going to go read the rest of the article.
Another worthwhile question is whether the states able to fend off Napolean were already healthier than those he conquered. Perhaps they already had the kind of anti-fragility Napolean introduced in conquered states, albeit likely not capitalist antifragility.
My prediction was 70% confidence that the French reforms were beneficial. Obviously getting conquered isn't great, but textbook growth theory says that good institutions are the most important driver of economic growth and I had the unfair advantage (relative to people at the time) of knowing that the Napoleonic Code is still in use two hundred years later.
Ah yes, the French Revolution - so extraordinarily innovative that they ended up being too weak to fend off a clever popular strongman who eventually made himself Emperor self-crowned in the presence of the Pope in the cathedral of Notre Dame https://en.wikipedia.org/wiki/Coronation_of_Napoleon_I Yep, that sure kicked out all those old dusty monarchical and imperial traditions! 😁
There's a reason you mention the Code Napoleon as the new legal code imported by the French into the occupied areas, and not the Code Robespierre or the Code Révolutionnaire, and it is because that after the fervour, ferment, and degradation of the Revolution into the Terror, it was easier for France and French institutions to slip back into the older, more stable, tradtional models. Yes, they did a lot of changes and reforms, but it wasn't feasible in the long run to be constantly and continually pulling down, sweeping away, and imposing hither, thither and yon.
Did the authors of this look at French colonial possessions alongside "yeah we had the French here for a while but then they left" in regards to how successful they were in the long run? (I am open to the argument that post-colonisation French possessions did better, but since most of Africa was colonised by everyone in the times of the Scramble https://en.wikipedia.org/wiki/Scramble_for_Africa , we can't really find a "this state was never colonised or occupied by outsiders" to be a control sample. Here, have a collaborative music video between Malian and French-Spanish musicians https://www.youtube.com/watch?v=J43T8rEOg-I). Small duchies and principalities such as Liechtenstein, Monaco and Luxembourg managed to survive to this day.
The point about "being invaded and occupied isn't so bad, why your economy will get a boost!" has me wincing. Germany was a developed European country at the time, the occupying powers didn't do much meddling apart from "no more Nazis" and pouring in money for reconstruction, apart from glaring at each other as former allies separated again, and it ended up divided in two halves, and East Germany was supposed to be doing less well. Do our economists do a comparison here?
Japan was a nation that was more or less forcibly Westernised, but even pre-war it was very eager to throw off the past, modernise fast, and on the model of its Western inspirations, become a colonial power itself. Oops! The US occupation did do a heck of a lot more restructuring and interfering here, but again, I think this was (economically) more a matter of throwing fuel on the fire rather than build from the ground up and do away with old feudal Japan.
And the main points I'm taking away from this is that the occupying powers eventually left. The French are not still in control of European territories, the Allies and Soviets have both left Germany, and Japan is in charge of its own government.
Economic development of an occupied territory is tricky and controversial business. One of our complaints over being colonised by Britain is that, post the Act of Union, Irish native industries were deliberately starved so as not to be competitors with British ones, and that the Irish market was treated as a dumping ground for British goods. We didn't get the same share of the Industrial Revolution, Northern Ireland with Belfast as the ship-building centre did that, and even Northern Ireland was more reliant on the linen industry which is not heavy manufacturing or high finance.
Indeed, even pre-Act of Union, this was a problem addressed by, amongst others, Dean Swift; we are still quoting his "Burn everything British but their coal". He wrote "A Proposal for the Universal Use of Irish Manufacture" in 1720, described as "a polemic on Irish affairs published anonymously by Jonathan Swift, his first such pamphlet after becoming Dean of St Patrick's. In it he attacks the English mercantilist policy which is draining Ireland of her wealth." https://celt.ucc.ie//published/E700001-024/
"It is the peculiar felicity and prudence of the people in this kingdom, that whatever commodities or productions lie under the greatest discouragements from England, those are what we are sure to be most industrious in cultivating and spreading. Agriculture, which hath been the principal care of all wise nations, and for the encouragement whereof there are so many statute laws in England, we countenance so well, that the landlords are everywhere by penal clauses absolutely prohibiting their tenants from ploughing; not satisfied to confine them within certain limitations, as it is the practice of the English; one effect of which is already seen in the prodigious dearness of corn, and the importation of it from London, as the cheaper market. And because people are the riches of a country, and that our neighbours have done, and are doing all that in them lie, to make our wool a drug to us, and a monopoly to them; therefore the politic gentlemen of Ireland have depopulated vast tracts of the best land, for the feeding of sheep.
(Footnote: Mr Lecky points out that in England, after the Revolution, the councils were directed by commercial influence. At the time there was an important woollen industry in England which, it was feared, the growing Irish woollen manufactures would injure. The English manufacturers petitioned for their total destruction, and the House of Lords, in response to the petition, represented to the King that ‘the growing manufacture of cloth in Ireland, both by the cheapness of all sorts of necessaries of life, and goodness of materials for making all manner of cloth, doth invite your subjects of England, with their families and servants, to leave their habitations to settle there, to the increase of the woollen manufacture in Ireland, which makes your loyal subjects in this kingdom very apprehensive that the further growth of it may greatly prejudice the said manufacture here.’ The Commons went further, and suggested the advisability of discouraging the industry by hindering the exportation of wool from Ireland to other countries and limiting it to England alone. The Act of 10 and 11 Will. III. c. 10, made the suggestion law and even prohibited entirely the exportation of Irish wool anywhere. Thus, as Swift puts it, ‘the politic gentlemen of Ireland have depopulated vast tracts of the best land, for the feeding of sheep.’)"
And for the "burn everything British", here is the provenance:
"I heard the late Archbishop of Tuam mention a pleasant observation of somebody's; ‘that Ireland would never be happy till a law were made for burning everything that came from England, except their people and their coals.’ Nor am I even yet for lessening the number of those exceptions.
Non tanti mitra est, non tanti judicis ostrum.
But I should rejoice to see a staylace from England be thought scandalous, and become a topic for censure at visits and tea-tables.
If the unthinking shopkeepers in this town had not been utterly destitute of common sense, they would have made some proposal to the Parliament, with a petition to the purpose I have mentioned; promising to improve the ‘cloths and stuffs of the nation into all possible degrees of fineness and colours, and engaging not to play the knave according to their custom, by exacting and imposing upon the nobility and gentry either as to the prices or the goodness.’ For I remember in London upon a general mourning, the rascally mercers and woollen-drapers, would in four-and-twenty hours raise their cloths and silks to above a double price; and if the mourning continued long, then come whining with petitions to the court, that they were ready to starve, and their fineries lay upon their hands.
I could wish our shopkeepers would immediately think on this proposal, addressing it to all persons of quality and others; but first be sure to get somebody who can write sense, to put it into form."
My guess: it won't change anything with a 50% certainty, and largely because to be in Europe in this revolutionary age means changing eventually.
Subsidiary is that it'll make things better at 20% because, if I remember properly, the Nepoleonic reforms were fairly standard stuff like 'let people work wherever they want' and 'peasants can wear whatever colors they want', and the reforms didn't touch the dangerous stuff, like family and religion (at least the Napoleonic reforms didn't, not after the terror).
30% it makes things worse largely on my general supposition that breaking apart institutions sucks and that being invaded sucks.
My prediction was that the Napoleon-reformed states gained a small but significant increase, of the order of around 10% of GDP. My reasoning was that having lots of incompatible weights and measures plus lots of fees and internal customs duties, cannot have helped trade.
I predict the reformed states did slightly better on average with lower variance. Although I'm dubious that this is measurable.
My prediction was 70% that traditional polities did better.
Prediction: 65% chance Napoleon made things better.
How about positive effects from a bigger market? Created by unifying the conquered areas. Less obstacles to trade, lower barriers for entry, simply better capitalism...
I'll predict that the conquered regions did better, but only because I've heard that after Napoleon was defeated, merchants in the liberated regions lobbied to keep Napoleonic law, arguing it was better for trade.
"is this potentially challenging a carefully evolved system of traditions?"
Interesting wording, which actually differentiates between these cases to a reasonable extent.
Peasants, after all, have had a lot longer to evolve good farming traditions than guilds had had to evolve good industrial traditions (Napoleon being, y'know, right in the middle of the Industrial Revolution). Peasants have had a lot longer to evolve good farming traditions than *anybody* has had to do *anything*, except maybe societal fabric traditions.
(Social traditions are a lot harder to nail down when and whether there has been this kind of paradigm shift. Of course, in the case of societal fabric the state of theory on what makes for good cohesion and effectiveness is kind of *notoriously terrible*, so it's hard to actually point to many cases of people trying to use theory (Social Justice, the Cultural Revolution and psychiatry probably count, but I can't think of much else).
65% Napoleon was a net positive.
@SSC: comments are a mess: substantive comments drown in a sea of one-liner predictions. I suggest that when you call for predictions in a post not exclusively devoted to a call for predictions, make a comment thread specifically for predictions.
Where to start on this? Perhaps with a confession: I didn't make a prediction because as a still-practising historian it feels wrong to guess the answer to a historical question rather than look at some evidence and hypothesis. If had done I would have probably got the answer wrong though because I have a deep- seated aversion to modernist reforms, which would likely skew any other priors. This may be relevant to the critique that follows.
I would be concerned on the underlying methodology of this work, simply because I'm slightly stumped as to where you would find accurate comparable measures for GDP, urbanisation or anything else that would suit such an analysis in pre-Napoleonic Europe (it's not as if weights, measurements or even clocks and calendars were in any way standardised, so actual measurable economic measuresare expectedto be reconstructable?). These sort of statistics are only gathered by modern states not small feudal societies. So in this case they must have been invented, probably using existing data (comparability unknown) and some level of assumption. A twentieth/twenty-first-century creation of an eighteenth-century GDP for a small Rhineland duchy is not the same as a twentieth-century GDP measure for a Rhineland town, which is a serious methodological issue. Just because you want to compare something doesn't mean you can: the assumption you can do so is perhaps a sign that you are inclined to believe that states are quantifiable and manageable, which is likely to skew your conclusions. This is why professional historians tend not to produce theories like this: their reputation can survive most things, but being known for mishandling data to reach a conclusion is a serious problem akin to being a pseudo-historian.
I'd also note that the UK is not an outlier here, but rather was ahead of the curve. Whereas much of Europe had a new liberal order imposed by conquest around 1800, the English and Scots had forged theirs in the wars of the seventeenth-century, where absolutism was defeated and the commons asserted their authority over the aristocracy (you could argue there was a Dutch conquest in 1688 if you really want, but it was only really opposed in Ireland). Revolution, however glorious, had already happened for the UK, and indeed allowed the formation of the UK.
Acemoglu et al are also travelling in distinguished footsteps. Their thesis seems, purposefully or not, to justify Marx, who had already identified the importance of revolution in changing society. This might not be particularly significant as an academic discovery therefore: that the transition from feudalism to capitalism was regularly violent and beneficial is probably one of the most accepted (and possibly less original?) parts of Marx's analysis.
So in summary, this looks like a retread of Marx using the perennially dubious technique of trying to force data to tell you what you want to see, a methodological no-no in every field of thought. And if the finding is significant all it shows is that (in Marxist terms) it is beneficial to replace feudalism with capitalism, a summary that would have found few opponents in 1909 a century before it was reached. This may be unfair, and the conclusions reached seem likely enough from the priors that liberal economies do better than centralised feudal ones (for those thinking feudalism equals lack of economic control, check out the medieval legislation on prices and wages), and the rule of law is better than rule by decree. But from a histirian's perspective this looks like data mining to make a point, and this seems unacceptable methodologically. With that in mind I'd be inclined to doubt the book's conclusions are valid unless it becomes clear that there is good and reliable data being used. Especially because the conclusions can be used to justify violent revolution to achieve an aim now, whereas the revolution might be irrelevant and the changes that came about (however achieved - I think the Scandinavians got there without revolution?) are what is important.
Since I've no way to determine whether or not French reforms would improve things, I'd just wager that they increased fungibility of the state-machine. 11/9 odds that they improved the status quo
> If you're really feeling bold, post a comment with your prediction before reading further.
I am feeling bold, so 60% on the reformed systems performing at least somewhat better (expecting to be proven wrong, though, given the title and what I expect Scotts take on this to be).
2/3rds that the non-conquered did better.
70% chance that the French-reformed areas did better. I vaguely remember reading something about the influences of French-imposed legal reforms in my AP European History class.
My conclusion from this debate of technocracy vs tradition is then that in order to predict which one will be better, you have to have economic theory (or whatever theory is appropriate). If you know that markets typically work better, then it's possible to predict that pro-market reform will likely improve things.
I suspect that modern urban design theory could also have predicted that Brasilia was a bad idea.
Prediction: probably depends on the reforms, but Napoleonic (being mostly good and modern) will be positive.
Call this the _The Mouse that Roared_ hypothesis.
(You could even cast Peter Sellers as Napoleon if he weren't dead)
I think traditional does better on average, with a 65% probability. (TL;DR: I am reading Seeing Like A State)
Requested prediction: my gut says that the conquered people did better along objectively measurable dimensions in the short term (greater production in monetary terms, fewer deaths by starvation?), but worse on squishier dimensions (lower subjective well-being, possibly higher rates of alcoholism and depression).
Prediction: reforms dramatically increased urbanization, since prerevolutionary Europe had systems that discouraged urbanization. Post industrial revolution, urbanized areas exploded in GDP, so I believe GDP would have gone up proportional to urbanization, but not any higher than can be explained by urbanization. Politically I'm conservative.
Oooh: Advanced prediction. For me the same as the top one I see. No substantial difference. Gears type explanation: They are both part of European culture which is the primary economic driver.
Possible confounder would be that the states reformed by France probably are on average physically closer to England, and thus the primary point from which industrial development diffused.
Pre-read predictions:
I think the places that underwent radical reform should exhibit much wider variance in outcome than places which didn’t undergo change. I’m 80% confident here.
I think at least some places that underwent radical reform should have performed better than they used to: I’m 70% confident here.
I think that at least 20% of the places that underwent radical reform should have performed worse than they used to. I’m 70% confident here.
I think there’s going to be a bunch of questions around metholodgy that will difficult to know or tell how much some places did or didn’t improve. I’m 90% confident these questions will exist and 80% confident they’ll be Enough to make it very difficult to concretely say much here besides “sometimes it works well, sometimes it really doesn’t,” with a slightly higher weight towards the “it doesn’t” end of the distribution.
Isn’t it the case that it’s the least successful entrenched systems that tend to get radically revolutionised? Well functioning successful ones less so - and are more likely to be the ones imposing revolutionary change on others.
I think if that’s true, it’s going to skew how likely it is that radical revolutions result in functional improvements
I didn't have the nerve to make a prediction, but I wondered if immigration/emigration would be a good test for which places had gotten better or worse.
I remember visiting the big castle in Salzburg and hearing the guide talk about all the improvements and new walls being built and wondering why on earth they keep building walls when it already looks impossible to take. So remember my reaction when I heard how it fell: Napoleon was moving through the area with his army, without any intention of going towards the city. So the city rulers hurried to send messengers after him to surrender.
Napoleon isn't exactly your average ruler - it's the kind of guy apparently you want to be conquered by. On the other hand I finished reading Conquest and Cultures, and a common pattern is that conquered people do A LOT better than their freer neighbors. And yet, east germany did (much) worse under the soviets. Why?
Well, one explanation is that being conquered exposes you to a higher level of civilization - after all, they managed to conquer you so they must be doing something better. And also bring you in a market of technology and exchange that is usually a lot better than you previous single country level. In East Germany's case it just happened that being conquered by Russia was just objectively worse than being conquered by the West - worse market, worse tech, (much) worse ideologies. After all, the betters conquering the worse is just an average tendency, not a rule. And Germany was once the kind of country that tried to conquer the world. So them losing that battle and getting divided up like a big pie is not the same phenomenon as a lone country getting gobbled up by an empire.
When writing about modern development, economists often talk about moving peasants off the land and into more productive work, e.g., factories. It strikes me that old European systems like common land allowed a lot of very marginal peasants to hang on in the countryside, where their inconsistent labor was of course not very productive. When the commons were abolished and the system rationalized, they could no longer feed themselves and moved to the cities. So it was easier for factory owners to find workers in places where the countryside had been de-feudalized, and industry grew faster. This makes perfect sense at an Econ 101 level. On the other hand the mechanism by which this was achieved was cleansing millions of people off the land, using the threat of starvation, and forcing them to move to places where by this argument their descendants did not really benefit until 50 years later. The time scale makes this data useless for talking about modern reforms, since nobody is going to wait 50 years for good results. Heck, the radical reforms in Russia may turn out in a century to have been great. Who cares?
65% that reformed principalities do better
I used to work in factory automation. When I first started somebody explained to me something about how bottom-up evolution interacts with top-down redesign:
Him: We're going to go to [place] and do major upgrades to a 20-year-old [assembly line].
Me: They'll be happy - the new [robotics gear] is way more efficient.
Him: No, it never goes that way. We design and install this new system that is obviously better. Then the client complains that their operation is less productive.
Me: How? The [robotics gear] is way faster than what we're replacing it with. And the [operator interfaces] aren't cobbled together out of old circuit boards and TV screens.
Him: It's the operators. They get *really good* at operating the line. They know all it's quirks. They do little hacks to make things work more smoothly, or they'll figure out a way to run [different product] through the line when it's otherwise idle.
Me: So they'll get more productive on this line, but not until they've worked on it for a while?
Him: Yeah, they'll be way ahead of where they were eventually. But they'll hate every moment of it, and curse us engineers for swooping in and disrupting their system that was working great.
---
Can we reconcile the top-down planning vs. bottom-up evolution in a similar way? Big reforms change which landscape you're on (for better or worse) and accumulation of metis moves you toward its highest point?
Whenever I recommend someone read Seeing Like a State I also recommend they read The Ghost Map or some other book where top down reform worked out very well. Sometimes coordination problems are hard and some order needs to be imposed.
But with regards to Britain specifically it wasn't so traditional as all that. There was enclosure, toll pikes, and all sorts of changes. But Britain had Parliament where different interests could bargain and change things in a way that incorporated some degree of bottom up knowledge. Pure autocracies can't do that and have to rely on tradition to a greater extent.
You will find that to be an important lesson as you try to run a business. When you have to judge each proposal as “good, let’s do it” or “bad, let’s not”, you should carefully consider the devil’s advocate position to make sure you aren’t missing something, but you have to hold in check the intellectual urge to give extra weight to counterintuitive arguments just because they are cool. A good idea is a good idea.
My prediction before reading is a bimodal distribution. Radically reformed small entities will 60% have done better and 40% have done worse. So in particular, the variance of outcomes will be higher.
Ok, prediction. Napoleonic conquest marginally positive, where disrupted communities relatively more open to subsequent urbanization and industrialization. And also to nationalism, state bureuacracy, and eventually EU. 60% confidence. Null case: Weber's iron cage enforces most of this anyway, so Napoleonic conquest not very significant.
I predicted "no effect" from radical reform, as I see a lot of commenters did, but most people aren't explaining their reasons so I will: I metagamed it. The fact that you're bothering to ask the question implies it has an interesting answer, and the most interesting answer after you set up the reasons to believe both pro and con is "haha, trick question, it doesn't matter!"
Just finished Joseph Henrich's "The Weirdest People in the World". He suggests it was a cultural psychology that changed due to first the Catholic church limiting cousin marriages among other things and then after due to Protestants promoting individualism,learning to read and friendly behavior to people other than your relatives. Lots more detail in the book of course and but super well researched. We are "WEIRD" (Western,Educated, Industrious,Rich Democracies)!
Part of me wonders if the answer is democracy. You do those things which seem intuitive (popular) if they work great, if they don’t then you vote the people out and do something else. It’s not perfect (good ideas implemented poorly go away) and it’s not like an authoritarian like Napoleon can’t get it right by accident, but it is probably why democracies are so much richer than non-democracies.
The weakness of the "evolved institutions" side of things for me has ways been that the only thing institutions are evolving to maximize is longevity. There's a good reason to think a system that has lasted a long time will last a lot longer, no reason to think that system is more beneficial than the alternatives.
In a peasant context, or any other individual or small community context, we can imagine that systems will evolve towards being useful to those carrying them out. In a larger societal context though where people can exploit each other and have divergent interests the system can easily be long lasting based on its benefits to some, even as it harms most.
60% that "traditional" ones did better
> Take a second to make a prediction here - a real prediction, with a probability attached.
I predict that 2/3 of the radically reformed polities did better than their left-to-their-traditions polities. I'm generally sympathetic to the Seeing Like a State argument, but my gut sense is that the polities in 19th century Europe were saddled with too much dead-weight tradition. So on balance, they needed some shaking up.
On “getting conquered by a foreign enemy is good for economic growth.” This is one of Sowell’s findings across lots of different groups far beyond Japan and Germany in Wealth, Poverty, and Politics.
But his argument was that it was isolation that kills and cultural exchange, regardless of whether it is forced, that creates prosperity.
A selection effect that's possibly worth pondering on when thinking about the French invasions: being invaded probably selects for not being powerful enough to resist invasion, which probably correlates with having a less effective economic/political system in place, which correlates with a higher probability that installing a new system causes general improvement.
Ok I'll bite and make a prediction without reading past the break...
The areas conquered by Napoleon did better in long term progress indicators on average, 70% confidence.
A) These areas would have been on average more impacted by the wars and therefore stronger postwar catch-up growth (think post WW2 growth on a smaller scale).
B) The reforms that survived the process of 10-15 years of reform followed by a reaction could be some of the more effective.
The 30% goes to the many potential confounders, me overestimating A, me underestimating the downsides of B, or just missing something entirely.
0.7 makes things better. 0.5 it makes it better in a way that also creates some meaningful costs (more inequality or social strife or something like that).
Prediction: Since Daron Acemoglu is the author I'm guessing that authoritarian imposition of pro-democratic institutions are good, and since this is the case they brought up that's probably what happened here. In the wake of Napoleon, principalities were turned into polities with on average slightly more democratic institutions than before and they were better off for it.
> Take a second to make a prediction here
61% in favor of radically-reformed polities
thought process:
It's a binary question, so without considering any evidence I should start at 50%. What kind of evidence can I consider? For one thing, the Napoleonic Code is even in this post called "modern", so it has at least endured, which is a point in it's favour- it's at least not so much worse that it's been abolished everywhere. It's been so socially ingrained that I even have trouble imagining not having a family name for people, so it's definitely been successful in the sense that it's enduring.
I'd think that the reformed policies make the principalities more legible compared to the unconquered ones, which should have an impact. I also think the reforms should take energy that would otherwise be used productively elsewhere, so I'm expecting at least an initial dip, which could possibly have a sustained impact through compound returns?
On the whole I personally just don't have enough background knowledge to be very certain, so I'll stay close to my starting point. Still, I think this is moderate evidence to positive impact, so I'll move a couple of decibans towards positive.
let's say 61% probability that the radically-reformed polities do better than the left-to-their-traditions polities.
To me it seemed obviously true that Napoleon’s interventions were an improvement. That is often the case with empires because they have to standardize trade and law in order to scale. Rome had a similar impact, as did the British empire.
Traditional does better at first, but then Reform catches up and surpasses. Probability 65%.
Before reading, the reformer did better around 80% of the time, because traditionally, monarchy are really bad for the economy. They don't want train because they allow easier revolt, they don't care about progress because the powerful don't have any risk of becoming poor, and the way to become powerful is to suck up other powerful people
Predicting that the reformed polities did modestly better overall, with 65% confidence.
Prediction: there will be a bell-curve of outcomes, centered around slight improvements. Even though the reforms were top-down, they were made by people with similar cultures who had absorbed many of the improvements.
You psyched me out! I predicted that the kingdoms were sclerotic and inefficient and so with 75% confidence Napoleon’s reforms increased measurable qualities ... eventually. But I consciously factored in my belief about why you were asking and that overrode my small knowledge of history.
As to why the Industrial Revolution started in Britain and not France, Britain had a several hundred year lead on reforming their monarchy. The Magna Carta was a different sort of reform, but it began the diffusion of power from a narrow aristocracy to wider elite class, who went on to have lots of first sons who ran family businesses and third sons who had lots of free time to seek clout by inventing stuff.
I think the same thread runs through why Communism failed, and it’s a point Nassim Taleb makes in economic contexts; you can be brilliant and see five years into the future, or just some guy if you keep your options open because in five years the answer may be obvious. Command economies - like kingdoms but also communism - are fragile to planning problems. The Soviets accidentally killed tens of millions of their own people with farming edicts and missed the boat on computers. And as long as the guy in power refused to get it, the right thing might never evolve.
Of course, a diffusion of power can also experience the fragility of planning problem, if the big players collude. Revolutionary reform may be effective at breaking up those cozy relationships, and may endure if it thoughtfully distributes power instead of consolidating it.
I think the answer is some combination of “energy is required to leave some local maxima and reach a new, higher maxima” and “most random paths away from a local maxima do not lead higher on net.” Meaning, I wonder if 19th century Western Europe was particularly ripe for this in a way that other attempts at reform were not. And that the incoming reforms had a sufficient degree of objective betterness that most of the work was done in smashing up inert, entrenched power structures and less in railroading cultural shifts.
In other words I am a Scott, except when I’m not.
Perhaps there is a distinction to be drawn between frameworks that enable valuable organic developments to be rewarded and reinforced (e.g., the rule of law, or, more specifically, intellectual property law) versus frameworks privileging entrenched interests. The Napoleonic Code was a quantum advance over "what the king or prince or whatever says is the law is the law," and of course the development of the common law, well before the NC, was one of the major factors in priming Britain for its era of hegemony.
OK, from the perspective of someone who has lived in France for a long time and been involved with the country for forty years.
First, Napoleon and the Revolution are not the same thing, and the Revolution itself went through various stages. It's often forgotten that the original Revolution was just the kind of high-level imposition of new ideas on France itself that Scott talks about, and was often deeply and violently resented. Some of the really wacky ideas (decimal days and years for example) never really caught on. Even the metric system only really became accepted in the last century everywhere, and, even today, if you buy half a kilo (500 gm) of apples in the market, you generally ask for "une livre", the old, pre-revolutionary measure equating to the English pound.
Second, what matters is the objective. Scott's agricultural examples, also in his book "Against the Grain", refer to a system that had evolved over a very long period of time and reached a kind of stasis. The judgement about whether they worked is essentially technical. So-called "modern" systems, alleged to give a better result technically, did not do so. Indeed, in Europe at least, so-called "permaculture" systems are now coming into fashion. But political and economic change is very different, and hard to evaluate objectively. What is true is that the French invasions gave a helpful shove to a process which was anyway inevitable - the construction of larger political and economic units based on popular sovereignty rather than inherited royal prerogative. But this was coming anyway, and in certain countries, like England, it came more slowly and peacefully, since the English ruling classes took fright at the prospect of a revolution there. (Even today, though, much of the British political system is distinctly less "modern" than the French, to the point where one historian memorably desired Britain as "the last functioning medieval state in Europe." ) The same was true, incidentally, of Japan: they modernised in their own way and at their own pace, taking what was needed from outside. Japan was already a modern state before WW2, and the "changes" much talked about as a result of the US occupation were largely cosmetic. Essentially, the Japanese carried on as before. But this was because the balance of power was different (the Japanese had a highly literate and well-educated society with lots of social capital, and hardly any USians spoke Japanese). What happens today is that foreigners descend on a society, like some in Africa, where little is written down and much of life is regulated by custom, and replace (as I've seen) traditional practices in the market with a new code of commercial law, literally translated from German, or whatever, which requires an infrastructure of literate tradespeople, inspectors, lawyers and judges to enforce it, none of which exist.
If being conquered was good, the Roman Empire falling shouldn't have led to massive decreases in standard of living.
But it did.
Likewise, if you look at Eastern Europe vs Western Europe post World War II, Western Europe did vastly better than Eastern Europe. Why? Because Eastern Europe was controlled by the USSR.
Colonized countries generally did better than non-colonized countries - unless the colonized country was developed, in which case, the developed country did better.
There are an enormous number of examples of these things.
The main differentiator seems to be the quality of the culture that won out/conquered them.
The idea that organic systems are inherently good is obviously deeply flawed. That is why Chesterton's Fence is such an important heuristic.
Chance the reforms were destructive: 99%
Well, I can safely say that I made the worst prediction. Lessons for myself: think whether you accounted for all the relevant information, are answering the right question and are not spouting complete insanity.
Prediction: net GDP growth for reformed polities is higher than for non- reformed polities up to 1900, 70% confidence. Not going to bet on the effect size, though.
The experiment isn't really testing the hypothesis as described, however. It's just testing "did the Napoleonic reforms generally improve GDP, urbanization, etc., in provinces where they were implemented." You've got a good sample size testing that hypothesis, but if you want to test the hypothesis of "Do centralized authoritarian reforms do better than evolved organic change?" you'd need to test many different examples and see what happens. This is like doing an experiment determining that penicillin treats infection and and concluding "We find no evidence for the hypothesis that eating mold is bad for you."
70% reforms turned out better