Rational Constructivism and the Metric System

~ Frederic Sautet ~ 

One of the most interesting examples of rational constructivism in history is the development of the metric system during the French Revolution. The disturbing and scary fact about that Revolution (and the Russian one as well) was the large-scale attempt to reconstruct society through the annihilation of many hitherto existing de facto rules. The replacement of the Gregorian calendar and the implementation of a new system of weights and measures are two instances of rational constructivism that the French revolutionaries attempted. The idea was to use the decimal system so as to make calculation easier for time, distances, weight, etc. The meter (one ten-millionth of the distance from the Equator to the North Pole), the gram, and the liter were defined as the standard units of weights and measures (see here and here).

An article in The Economist this week relates the failed attempt during the French Revolution to implement the decimal system for time keeping. Days in the Republican Calendar were divided into ten hours, each hour into 100 minutes, each minute into 100 seconds. It didn’t catch on and Napoleon abolished the new system in 1805. He didn’t abolish the system of weights and measures however, which eventually became the International System of Units (SI) and was adopted by every nation but three (Burma, Liberia, and the United States).

The Economist article doesn’t tell us much about the interesting side of the story. Strangely it laments that SI was not adopted for the keeping of time and for the calendar, but it doesn’t explain why it failed. Why all this happened is an important question for those who regard rules as an evolutionary process in the Hayekian sense. Many libertarians in the US hate the metric system, as it represents the pinnacle of top-down state-imposed rules. Interestingly, the metric system never caught on in the US when it did in Canada and other ex-Anglo colonies such as Australia and New Zealand (as well as the rest of the world). This probably goes to show the deep cultural divide that exists between American society and the rest of the Common Wealth, especially with regard to the role of government. The costs associated with the imposition of a change in measurement units are huge. Not only people need to grasp what a meter represents compared to a foot, but also every machine that functions with the old units need to be changed or adapted. While Napoleon imposed constructivist changes such as the adoption of the Civil Code (although it was based on the former legal rules in use), he abolished the Republican Calendar and the use of decimals to measure time. Why?

It is likely that in spite of its ease of use for navigation, the public didn’t feel comfortable with 10-hour days. People kept on using the old calendar. (Moreover, a “republican” calendar may not have been to Napoleon’s liking.) Some rules are too costly to impose and eventually fail. Even in Europe, the use of the “horsepower” concept is still common, even though one should normally use kilowatts. The generations that were born decades after the widespread use of horses came to and end still use horsepower as opposed to kilowatts. Clearly, radical top-down rule changes are bound to fail because the costs are too high – the new rules won’t stick. It reminds me of the scene in Bananas in which the hero  (Woody Allen) goes to South America and participates in a revolutionary coup. The new leader imposes de jure changes (such as using Swedish for daily conversation and wearing one’s underwear outside of one’s pants), which don’t make any sense to anyone and are too costly. As a result the revolution fails. 

And yet, the metric system was eventually adopted (or imposed) in almost all countries. More importantly, it came after several failed attempts to adopt universal system of measures that were more “rational” than what existed before (see for instance the system proposed by John Wilkins). Is the successful imposition of the metric system an instance that rational constructivism can work, as Hayek’s opponents would say? Or does it simply show that some imposed changes may work but there is no way of knowing which ones because an ex-ante cost-benefit analysis is truly impossible (as the failure of the Republican Calendar shows)?

Comments

[Rafe Champion]

Pete, I don't know about the metric system "catching on" in Australia, it was imposed by our socialist PM Whitlam, along with double digit inflation and unemployment etc etc. He was elected (narrowly) on a promise of reform with the slogan "Its Time" after about 35 years of conservative rule.

[sean]

imposed? no it was american arrogance, which btw i love, that said "we dont give a shit what is rational, we rule and suck it."

i might sound facetious, but im serious, america rules because of that.

ahahah

[Mathieu Bédard]

That was a very interesting read!

So, when people use costly evolutionary rules, can top down imposition of cheaper constructivist rules be successful given that the transition cost is low enough? Sounds fair enough to me, but the real question for me would be why was it so cheap in the first place to shift to another set of rules for seemingly very little benefits?

Could it be that now that the school were State-ran it was much easier to impose such rule change?

And was the acceptance of the metric system really so widespread and fast? How fast did the metric system become accepted in commerce? How long did it take for the metric system to spread to other countries?

[Bogdan Enache]

The "constructivist measures" were enacted by the Convention, as it says in the title of the original decree in the photo. The Revolution spread over 6 years and during this time there were no less than 5 or 6 constitutional regimes, not all as bad as the Jacobite Convention. (Among the most baffling measures took during this time, was the creation of Temples of Reason worshiping a Goddes of Reason, a strange thing to do when you proclaim an atheist society. The emblem of France, Marianne, dates from this period and it's not unrelated to this phenomenon). The new calendar, like many other republican symbols that immitate the Roman Antiquity which have become embedded in the fabric of our society, might have stayed with us - who knows? - if Napoleon wouldn't have been so keen to reconcile with the Pope and thus be crowned with the title of emperor in something resembling Charlemagne's succession and be recognised as such by the people. The metric system became widely acceped because it was taught to young people in the public school system created in this era, which was focused primarely on engineering, not on classical languages and religious history as the Church-run system before, and it spread throughout Europe at the pace of Napoleon's conquests - very fast that is.

[Tristan]

In the UK we are a hybrid - at school we're taught SI units for science (which is a blessing - imperial units are a pain) and in general, but loose goods area still advertised in pounds (although the law now states kg prices must be indicated too). Speed limits and distances are still mph/miles and people still use feet/inches and stone/lbs for many measurements.
Unlike the continent, we still buy beer in pints (good luck changing that one...) - I note they're Imperial pints which are bigger than US pints - just to add complexity.

There is a difference between generations is interesting, my parent's generation is all imperial measurements, mine is a mixture - the difference? The schools now teach metric and not imperial.

[Daniel Klein]

Nice post.

The Imperial system obviously sucks. I still don't know what a pint is.

[Roger Koppl]

I figure one crucial fact explains why the metric system caught on, whereas the revolutionary calendar did not. The calendar was rather uniform across space. People all over European civilization used the same calendar. (I assume there were some differences in 1791, especially as you go east, but I am sure they were small potatoes.) The units of the metric system, however, were not uniform across space. I think they varied from town to town. Having a uniform system of weights and measures lowered transaction costs so much that lots and lots of folk found it well worth the trouble to learn the new system. The Wikipedia article on the metric system seems to support this interpretation:
http://en.wikipedia.org/wiki/Metric_system

Something similar happened with German and Italian, which could be imposed by the state in part because they replaced a variety of local dialects

[Pietro M.]

I see no knowledge problem in moving to the metric system, only sunk costs that make a transition to a superior one economically unprofitable. I mean, SI units are better, but not enough better to bother about rejecting pints and pounds.

There should be little doubts that with sufficient coercion many things in the long run can be done: changing measure units, changing religions (Albania was christian and then forced to convert to Islam), changing languages (Tibet will be chinese within a couple of decades).

What coercion cannot do is organize a complex system, like the market, substitute a complexity-reducing institution like the price system, and align incentives when they are not.

So, I would ask, what does imperial units do that SI units can't? Nothing. So, it's only a matter of investing in getting used to the new units, something that with a lot of coercion, centralized controls and a couple of generations can be performed.

By the same token, if the government wants, it can turn everybody into a Macintosh user, into a non-QWERTY typer or into a Cadillac driver.

The basic error of constructivism is that it wants to do without complexity and ignorance reducing institutions and is at risk of imposing unlimited transaction and enforcement costs if incentives are not aligned.

In the case of metric units, I see none of these problems. The impossibility of performing costs-benefits analyses is an ontologically different problem: it is not about if it can be done, but if it should.

So, what stopped people from thinking in terms of 10 hours? Either the absence of real advantages (whereas SI units are superior), or because coercion didn't suffice. The two factors are partial substitutes, in that no coercion may be necessary when things are useful, and usefulness is not necessary when things are imposed by coercion (economic policy is an example).

Metaphorical summary: Government can create paper money systems, it can't create moneyless economies.

[Sam]

"The Imperial system obviously sucks. I still don't know what a pint is."

Aren't miles and pounds imperial measures? The USA is the only country to have a quaint and confusing combination or imperial and metric systems.

[Greg Ransom]

In his _The Constitution of Liberty_ Hayek supports "rational constructivism" in the decree of such things as uniform weights and measures by national governments.

The "principle of charity" requires that we at list take at least a half a step in the direction of thinking about how Hayek would see nothing incoherent with this position, even in light of his writings on "rational constructivism".

Frederick wrote:

"is the successful imposition of the metric system an instance that rational constructivism can work, as Hayek’s opponents would say?"

[Greg Ransom]

In the early 1980s the U.S. Federal government would only allow the purchase and use of IBM PC's by its agencies and contractors.

My DOE employed Dad has always insisted this is why the PC became the government and business standard in personal computing.

Pietro

"if the government wants, it can turn everybody into a Macintosh user"

[Bogdan Enache]

I think the adoption of the metric system is more of a Karl Polanyi story than a transaction costs story, at least in the first phase. The metric system allowed the new national states to impose uniform regulation and taxation over previously very different jurisdictions.

With regard to language, things are more complicated. The question was which literrary language should be thought in public schools and used in offcial administration. In Germany, that was based on Luther's translation of the Bible, but in Italy there were many dialects with a rich literrary tradition, like Napoletan and Venetan (Casanova wrote in this one), but the national state settled on Toscan (the language of Dante, Petrarca and Machiavelli), because the nucleous from which the new national state was situated in this area.

The calendar was the same, until the 16th century, when the Catholic Chuch adopted the Gregorian caldendar instdead of the Julian calendar, which is off by about 13 days presently. When the governments in some the Orthodox countries switched to the Gregorian caldendar, some of the Orthodox Churches, like the Church of Russia didn't. So there are basically two calendars, one civil and one religious. The shift form the Julian to the Gregorian calendar was as much of a practic matter as a signal of cultural change.

[Barkley Rosser]

Bogdan,

The modern official German language is closest to what is spoken in Hanover, Hanoverdeutsch. I do not know if it is that much closer to the Luther Bible language than others, but it might well be.

The nucleus of the Italian state was not Toscana, but Piemonte, with Savoy rulers becoming the royal family of the united Italy.

[Bogdan Enache]

Yes, that's true, The Kingdom of Piedmont-Sardinia started the Italian national state, but the first capital of Italy was in Florence, the capital of Tuscany and the intellectuals in this region had a great role in the struggle for nationhood, as it were.

With regard to Luther, I was emphasising his role in creating the literary German language (spelling, grammer rules, ponctuation etc); I personally knew that the purest German is spoken in the city of Rostock in Meklenburg, on the Baltic sea, I don't know.

[Charles N. Steele]

Frederic, I entirely disagree (respectfully, I hasten to add) with your analysis. The problem at the heart of rational construction is the knowledge problem, and there's no particular knowledge problem in design of a measuring system. It's more a matter of getting people to coordinate on a common standard, and entirely different from trying to organize an advanced economy according to a central plan.

I find it extremely difficult to believe that the costs are so great to adopting the metric system; every researcher in every scientific laboratory in the United States uses it, and then goes home and uses the English system. What are the costs, really, that I need two sets of tools -- one English and one metric -- depending on which of my bikes I'm working on?

I think the primary obstacle to switching systems is one of subjective preference. Every argument I've heard in favor of the English system has this at its root (it's certainly *my* argument). With respect to the Republican Calendar, it was just too much for people to swallow.

In conclusion, none of this is relevant for the question of whether rational construction can organize society. That's a question that hinges on how the knowledge problem can be solved.

(P.S. You gents ought to link to our blog.)

[Charles N. Steele]

Several commenters have raised the example of language. Language is an interesting case -- it's largely similar to a measurement system. One can design a perfectly good "artificial" language; there are a number of them.

But adopting a new language involves considerable startup costs; hence even if an artificial language offers benefits of simpler, more logical grammar, or more precision by care in defining words, etc. there remains a significant barrier against adopting it. Still, one can construct a workable language in a way one cannot centrally plan an economy.

But language also faces a knowledge problem; how will new concepts, ideas, knowledge be brought into the language? In English, new terms are simply developed by users with their specific knowledge of time & place. Another approach is that of French, in which an official body attempts to govern what words and constructions are and are not French -- a top-down approach. I suspect this is one reason (among several) why English, rather than French, is the international language.

One of the more successful artificial languages, Esperanto, does allow for the creation of new words by users, as needed.

[Tyler Watts]

Fred,

Interesting post... first thing to come to mind for me is the adoption and development of the US dollar. The dollar, of course, was a Spanish unit inherited from evolved practice- a combination of England's bullionism and Spain's rich silver mines in Mexico and Peru. But the Spaniards divided the unit into eigths; it was Thomas Jefferson's idea to go with a decimalized dollar. I'm wondering if this was a matter of convenience or decree. Interestingly, Robert Morris proposed a new monetary unit divided into something like 1,480 parts, so as to have a common denominator with every money then in use in America.

[Barkley Rosser]

Bogdan,

Did a bit of checking on German. Luther did create a language synthesized out of various German dialects so his Bible could be widely read. Goethe and Schiller reported further developed it, and modern standard German has come out of what they did more or less. According to Google, this standard German most closely resembles the middle German dialects (which would include Hanover). Rostock is in the zone where Lower German is spoken.

I noticed that digging around, many locations claim to have their German language being "pure," including Vienna (whose language is very different from that in Rostock), although the more authoritative sources say it is more like Middle German (a sub-species of High German).

Another item where there is strong path dependence is alphabets, which may be easier to change from above than languages. However, even with them it can be quite difficult. This is discussed in a very interesting paper published in JEBO 2008 by Young Bak Choi, "Path dependence in the Korean alphabet," which reports on how it took several centuries and some major political upheavals over several centuries for the much more "efficient" Hangul alphabet to finally be fully adopted in Korea.

[Jim Hlavac]

It's funny to hear the system of avoirdupoir - what we have here in the US -- called "imperial" when it was the Emperor Napoleon who imposed, imperially, a new measurement standard. And to hear metric called "Scientific," when I'd guess that the vast majority of the world's useful inventions, scientific, industrial, commercial, in the past 200 years or so, were done in inches and feet, pounds and ounces, right here in the US.
The measuring system of America is oh, 800, even 1000 years old. Nothing imperial about it, but the science it pushed brought us around the world and to the moon, and rescued more than a few metric nations from their dictatorial clutches. It's even apocryphal as to whether it's based on the body measurements of King John (a man by the way, very Obama-like in wanting his way, who wound up with his own "Massachusetts," that being Runnymede, after which not one potential heir to the throne was ever named John again, and free liberties began to see the light of day.)
Meanwhile, it is the US which allows both systems to be used, freely, such as on products, but the rest of the world does not allow the use of the system in place for 300,000,000 in the world's largest economy. The EU specifically outlaws a system of measurement, only ours, by the way, in a poke a stick in your eye, America mentality. Talk about your discrimination.
Further, the entire country is based on the mile, the foot, the acre. Every piece of land in the nation is divided like this. Every street sign in the country gives miles to the next distance. Every house, every car, all avoirdupois. I could go on, like every boulevard in Phoenix AZ is exactly one mile apart, and each block is 1/10th of a mile, -- what, are we to move the roads? Or change the signs so that 1 mile is now 1.67 kilometers? Or maybe we should just leave it alone, and all who visit should do what they can to deal with our American Culture -- it's what we want that counts, not what is convenient for the world.
To now change it would be such a mind boggling expense for no reason other than to "join the rest of the world" -- most of whose people would seem more than content to abandon their metric wonderlands and join the inch, foot and yard folks here.
No, it's academic as to whether the US should "adopt" metric. What we have is best of both worlds. Where necessary we use metric, where we have it all done in the "old" system we'll keep it.
Even more, has anyone every done a consideration or a study on the thought-boost and stronger math skills required for avoirdupois over metric? Could that be yet another reason that we are so far ahead of the rest of the world?
No, let the world be metric, and let us stay as we are, with both when necessary.
I also find it odd that, anecdotally in my life, many metric lovers are also lovers (though not all obviously) of parliamentary systems, and of the civil law -- another from the top down idea that America does not use, in favor of the Common Law. Odd, that 190 some odd countries are civil law, and we are alone on that too.
Again, we provide this example of what created the richest and freest society, and all I keep hearing about is how they do it over there -- wherever there is -- the very place we all escaped from to not have to deal with.
On metric, I'll never give an inch, or a yard. No, to it.

[Bogdan Enache]

Dear professor Rosser,

I think you might be right about Hanover. The city is close to the center of Germany and used to be (is) a major fair city. According to linguists, because of the commercial activity in this kind of cities, people speaking various dialects of the same language interact, so the language gets "polished", i.e. regional words, regional accents and so on are not used, the result being a lingua franca that everyone can understand. The linguists then call it a "pure form" and attempt to "model" it for teaching and formalization purposes. The same thing happened in Romania where the city of Brasov, in the middle of the Carpathian arch is the equivalent of Hanover, although withing the Romanian language or the Daco-Roman dialect doesn't very significant subdialects.

Regarding alphabet change, well, until the middle of the 19th century, the Romanian language used to be written with Cyrillic characters in the South-East part, Wallachia and Moldova (when the Soviets took Western Moldova in 1940, they switched back Romanian to Cyrillic characters and called it Moldovan language). The alphabetical transition took about two decades during which a mixed alphabet was invented and newspapers and official documents were often written with both alphabets (half a page with Latin characters and the other half with Cyrillic characters), while spelling was an experimental art, sometimes based on the etymological principle and sometimes based on the phonetic principle.

[Bogdan Enache]

...when the Soviets took Eastern Moldova, I meant

[Craig]

Very interesting post.

I was in elementary school when it was assumed that the U.S., too, would go on the metric system. Several years later -- that talk ended.

I suspect that, someday, we'll look back on the push for nationalized health care as a similar silly political fad that just never caught on here.

[BillChapman]

A fascinating discussion. Back to language, if I may ...

"But adopting a new language involves considerable startup costs; hence even if an artificial language offers benefits of simpler, more logical grammar, or more precision by care in defining words, etc. there remains a significant barrier against adopting it. Still, one can construct a workable language in a way one cannot centrally plan an economy."

Esperanto is a useful example. Esperanto hasn't yet gained the recognition it deserves. However, all things considered, it has actually done amazingly well. In just over 120 years, it has managed to grow from a drawing-board project with just one speaker in one country to a complete and living natural language with around 2,000,000 speakers in over 120 countries and a rich literature and cosmopolitan culture, with little or no official backing and even bouts of persecution.

The movement for Esperanto has always been a grass-roots one. Its initiator Dr Zamenhof had no riches and no political power. The language has continued to be used because ordinary people find it useful.

[Sam]

I'll point out that the US has adopted the metric system--back in 1866! What hasn't happened is outlawing "customary" measurement units. Federal law requires metric and customary in consumer packaging (since 1992). The federal government prefers to use metric in purchasing.

Despite mandatory metrification worldwide, there's some areas where metric isn't used, and its probably due to the US, e.g. altitude of airplanes is in feet, and distance is in nautical miles, and airspeed is in knots (English is the language used to communicate between pilots and air traffic control).

[Andy]

The French revolutionaries also tried to redefine a week into ten days from the standard 7 day week, a concept dating back to creation. Also, our 12 hour days and 60 minute hours date to the Babylonian 6 based counting system in the days of Daniel, Nebuchadnezer, Cyrus, Darius and those characters.

[Bill Stepp]

Greg,

You might be right about the Feds, IBM and the pc, but they had some help from Apple Computer (as it was then known), because it decided, over Steve Jobs' strenuous objections, not to go after the business market, except for the small graphic design segment.

[C. Walls]

A slight problem here. When you talk about the way in which the metric system failed you persistently fail yourself to talk about England which, may I remind you, was the country that founded your country.
We never adopted that system although there are attempts to make us change. But there is a more interesting aspect to this whole issue which is the Corsican Bandit's (which is how we used to know him) own comments himself about the metric system. He actually said about the system that it would never work.
Comments please?

[Shox Shoes]

A troupe of little vagrants of the world, leave your footprints in my words.