Hacker News new | comments | ask | show | jobs | submit login
AsK HN: Why can't the US change to the metric system?
44 points by robomartin on May 12, 2013 | hide | past | web | favorite | 67 comments
Every 3.14 years I have this question. I know that changing all highway signs would cost a lot of money and would take a long time. Well, had we started fifty years ago we would have been done by now.

When designing mechanical components or circuit boards (or looking at other's designs) one recurring thought is that our fractional unit system might actually cause inefficiencies and increased cost of goods.

You can look at an American design product and identify it as such by taking a few measurements. Mechanical designers will think in fractional terms and specify "nice round numbers".

For example, a feature might be set to 1/16 of an inch, when, perhaps, a smaller thickness or length would have worked just fine. In this case 1/16 is 0.0625 in, which happens to be 1.5875 mm. Now, a designer working in SI units will probably specify 1.5 mm for the same feature. This doesn't seem like much, but you are talking about 5.5% more material. That's not a trivial amount if you are making a bunch of widgets.

We ship liquids in gallon containers. A gallon is 3.8 liters. If we worked with liters, might we buy and ship three liters instead? Perhaps generating less waste?

This is a very simple example to illustrate one of the mechanism that might be at play here.

Don't know. Just a though in between coding sessions on a fine Saturday evening (Los Angeles).

When I was in 4th grade (in 1979), growing up in the US, we learned about the metric system. My teacher said that we were learning it, because the US would soon be switching over to it.

I came home and told my mother what my teacher had said. And my mother responded, "Yes, that's what my teacher told me in 4th grade, also."

Fast forward more than 30 years, and the US is no closer, despite other countries (such as England and Canada) having pulled off smooth transitions. Aside from a few signs on a highway in Massachusetts (of course) that I saw 20 years ago, marked in km as well as miles, and 2-liter bottles of soda, the metric system has completely and utterly failed in the US.

I now live in Israel, where the metric system is standard. My children cannot believe that I once used a system that wasn't so simple and standardized. But because they're not learning the metric system, but just using it, they never learned the beauty of its design, with powers of 10, and uniform prefixes. They just know how many cm tall they are, how much is in a 1 kg container of ice cream, and how many km it is to their grandparents' house.

It seems obvious to me that the metric system can and should be uniform, including in the US. But there is so much inertia, and so much history of opposing it, and so much technology tied up with the existing English system, that I don't realistically see it happening anytime soon, even if it could and should happen.

Hello fellow oleh amerika'i.

Funny thing is, there's one metric unit I truly don't like: temperature. Fahrenheit-scale temperature is just better for describing the immense ranges of cold and heat you'll see inside the average year in a very temperate place like the United States or large parts of Europe. Only over here in a warm climate like Israel does Celsius make sense, where zero degrees is actually one of the coldest temperatures of winter and forty degrees the hottest temperature of summer.

For distances, I don't really consider Imperial or Metric that much more useful. They're just slightly different.

Never in my life, which ranged from living in tropical jungles to snowy European cities, have I ever felt the desire for differences in measuring temperature.

Comparing F and C is one of the most obvious cases for metric, as I see it. 0 is freezing, 100 is boiling. It couldn't be clearer.

The weather never actually hits "boiling".

Ah, but what it it when it is -40?

P. S.: This is a trick question.

The one very good thing about Celsius scale is conversion of energy amount of heating up water to calories. Those as such are still not so useful but the conversion to J is still something you can do roughly in your head. (Of course, conversions between °C and °F are simple for anyone who started their learning of C from K&R).

Actually, temperature is the only thing which regularly makes me long for the metric system. I used to think celcius was ridiculous until I realized how useful it was as a tool to intuitively understand large temperatures. Just divide by 100! When I see 2500 F, that number is essentially meaningless to me, but when I see 1400 C, I immediately think "oh, that's like 14 times (freezing to) boiling, that's hot!".

> I immediately think "oh, that's like 14 times (freezing to) boiling, that's hot!"

You can only say that X is K times as hot as Y if you're using an absolute temperature scale.

Here's the ratio of the two absolute temperatures in Python:

   >>> (1400 - -273.15) / (100 - -273.15)
So 1400 C is only about 4.5 times as hot as 100 C.

IMHO, the metric temperature scale should have absolute zero at 0 degrees and the freezing point of water at 100 degrees. But due to historical accidents, it...doesn't.

Quick question for thermodynamics nerds: Is the mean kinetic energy of molecules in a sample proportional to temperature? I bet it's actually proportional to temperature squared, or at least some power of temperature.

Why is having to use negative temperatures a problem?

It isn't since in the fahrenheit scale temperatures drop below zero as well. It's just that we rate a lot of stuff on a scale of 0 to 100, so with the fahrenheit system it just makes sense to know that 0 is really cold and 100 is really hot.

Celsius is more intuitive for cooking.

I wouldn't say the English have pulled off a smooth transition. We still universally use miles for distance, frequently use feet and inches (almost universally for giving the height of a person; people usually weight themselves in imperial units) and drink pints of beer since bars can't even legally sell draught beer in metric units. Despite EU pressure, we're almost as stubborn as the US.

The smoothness of the transition in Canada is somewhat debatable. We use this weird mix of Imperial and metric now - people and food are (generally) measured in Imperial (I couldn't tell you my weight in kilograms, I can only remember my height in centimetres because it's just a bit short of 2m, and signs in the supermarket meat department are all $/lb (though the meat labels are in $/kg)), while most other things are measured in metric. Whether someone is more comfortable with Celsius or Fahrenheit is a generational thing, but most Canadians can think in inches and feet at least as easily as centimetres and metres, ounces and pounds more easily than grams and kilos, but (at least among the younger generations) mililitres and litres more easily than pints and gallons. The worst one is where we fill our cars in litres, measure the distance we travel in kilometres, then calculate the fuel efficiency in miles per gallon.

Actually, my elementary school math textbooks were American, awkwardly pasted over to use metric for the measurement unit, so I can keep the Imperial liquid measurements straight by just remembering that there are "2 half litres (pints) to the litre (quart)" and "4 litres (quarts) to the 4 litre bag (gallon)" (because milk comes in bags here)

There are many extremely valid arguments for why the US should change to the metric system, but I don't think that "maybe we'll round down and be more efficient" is one of them.

> We ship liquids in gallon containers. A gallon is 3.8 liters. If we worked with liters, might we buy and ship three liters instead? Perhaps generating less waste?

Might we buy and ship 4 liters instead? That's closer to 3.8. Anyway, if we rounded down to 3, what about those people who really need 3.8 liters? They'll buy two jugs, which is double the packaging, and if they don't need the extra 2.2 liters, it's a waste of the liquid.

Also, many things are sold by the pound. But a pound is ~0.45 kilograms. Surely by your reasoning people would round up to 0.5 kilograms, and thus be less efficient.

I see your point and raise you 0.35mm.

In working on a design I specified 1/4 shafts. Why? Well, from the commonly-available range of sizes this is what met the specs.

Then I went to price manufacturing. No issues here in the 'States. My Chinese CM comes back to me and says I can save money if I go with a 6mm shaft rather than 6.35mm. The 6mm shaft is more common, weighs less and costs less per unit length. And so the entire design slowly changed to metric units and we saved both money and weight.

Of course, this is a loose and probably really flawed hypothesis with nothing but personal data points and random observations for support.

Your hypothesis is that all common measurements in imperial units round down when expressed in metric units, and that somehow by switching and rounding, "efficiency" will be gained, and therefore metric units are superior.

Please think about this for a moment. Really think about what you are saying. It's ridiculous.

Er, I'm not him, but I suspect the reason 6mm shafts were cheaper is because that's what the rest of the planet orders (not anything to do with rounding or whatever)... Rare or custom specs almost always increase the price.

robomartin is the OP -- read his original post and also the comment that I was replying to. It doesn't matter that in some circumstances rounding imperial units to the nearest-but-lower metric equivalent results in cheaper parts (perhaps due to better availability). What I'm saying is that it's absurd to use that as an argument for switching from imperial units to metric.

(Don't get me wrong, imperial units are completely backwards and horrible, and I would love for the US to switch. It's just that this particular argument for doing so is stupid.)

Well, it's only absurd because my data points are easy to attack. I'm with you on that. It would be very interesting if a study looked at various efficiency metrics to compare the systems.

Like I said, I know my argument is weak and calling this idea nonsense at this stage is probably not without merit. At the same time I've seen this effect more times than I can remember in my career, which leads me to believe there might be a number of cases where it may very well be factually correct to assume that imperial is costing us money and creates more waste.

Rounding will happen eventually, but it's a slow change. Initially I'd expect everything to be in the old units but labelled metric.

As an example, here in New Zealand most bars sell beer in 500ml glasses, but the one I go to sells 568 ml glasses. The same happened when I was in the UK and the milk had to be sold in metric quantities.

You don't need to force people to change their thinking, but if you start labelling things in metric it will change over the generations.

Yeah rounding efficiencies would probably just wind up being mostly a wash as some things round down and some round up or amount to a very small difference over all.

There are better arguments that can be made for changing over to the metric system.

The US is on the metric system[0]. The people, however, are not. The children use the United States customary units, because their parents use it. The parents use it, because their parents used it. The companies use it, because the people use it.

I also happen to like the orderly ISO 216 paper sizes, but US people don't use that either.

[0] http://en.wikipedia.org/wiki/Mendenhall_Order

Never mind the paper sizes, the really crazy US measurement is the paper density: https://en.wikipedia.org/wiki/Paper_density

tl;dr - most of the world uses grammage (g/m^2, ISO 536). The US uses pounds in a ream of paper in its uncut basis size. The basis size varies with the type of paper, and so does the ream size. It's a measurement that can only make sense to people who never take paper out of the box.

I wonder if the US will ever move to international paper sizes? I would have thought this would actually be an advantage to (American) product manufacturers making printers and photocopiers. Would it not simplify product offerings if A4 was used instead of US Letter? Plus you get the advantage that A4 scales up perfectly to A3 and scales down perfectly to A5.

Perhaps the biggest thing non-metric people miss is easy conversions. A litre of water weighs one kilogram and occupies a 10cm cube. This makes it really easy to move amongst volume, weight and dimensions.

It is already the case that many goods sold in the US are labelled in both imperial and metric. The order can just be switched on labels. In theory the roads could be left using miles as has happened in the UK.

Clive Cussler used to have a preface in his books saying this: "Please forgive the inconvenience of converting measurements from what most Americans are used to. But in 1991 the United States finally became the last nation on earth to convert to the metric system."

He has long since given up.

>Perhaps the biggest thing non-metric people miss is easy conversions. A litre of water weighs one kilogram and occupies a 10cm cube. This makes it really easy to move amongst volume, weight and dimensions.

Indeed. And while that's very useful for scientists and engineers, it's not really useful for the average person. Most people don't do conversions often enough to justify switching their measuring system to metric, and if one doesn't care about doing conversions, the advantages of the metric system become much less clear. And the American government isn't going to switch the country to metric unless it provides clear advantages to most Americans.

I blew a friends mind when he mentioned wanting to have a rainwater tank that contained some number of gallons. I immediately said how much it would weigh and what the dimensions of the tank would be, and consequently how it would be too large for the area he intended.

It drives me nuts trying to follow recipes written by Americans as the measurement units are entirely random between weight and volumes. Sometimes it makes sense due to packing - eg flour is more accurate by weight than volume. But often the worst one is used.

Why not switch to a system that's based on actual physical constants instead of properties of a certain impure water mix ?

How about we start measuring distances relative to c, mass in GeV/c2, or maybe in multiples of electron masses, and so on and so forth. Sooner or later (probably much later, granted) we'll do this, so why not get ahead of the game ?

The meter is defined in relation to c: https://en.wikipedia.org/wiki/Meter#Speed_of_light

Well, yes, but read that page. I mean a vaguely reasonable fraction, not 1/299792458 per second (plus if you're going to be pedantic, second still isn't defined in terms of c).

IMO conversions between weight, volume and dimensions are very useful for the average person, too. At least when the average person cooks food, builds something, etc. Though obviously this one-to-one conversion of weight/volume only works so great if you're dealing with water or other substances with a a similar density. But even so, also knowing that granite has density that is 2.7 times that of water helps e.g. in estimating how much a rock weighs.

I'm in a country that is as metric as it can get, and I'm still infuriated enough about package sizes that are sometimes multiples of imperial units, even if expressed in metric units.

"Mix 210 g of couscous to 2.4 dl of boiling water". Absurd. Well, recently they changed it to a more usable "Mix 2 dl of couscous to 2.5 dl of boiling water." That is a lot better. I usually don't want to use a kitchen scale and want to get away with a measuring cup.

Gallons, inches and feet etc are fairly easy to handle. I've grown up in a completely metric country and yet I have fairly good intuition of most "imperial" units, mostly due spending time on the internet. Changing those would be pretty low on the list of things to change in the US. Here are some things I'd change first.

Date formatting. This is such a major pita when you see something like 02-03-04 and you have no clue what date that's supposed represent. At least with customary units, the units are usually specified but no such luxury with dates.

Paper sizes. This is annoying because converting between Letter and A4 is not trivial. You might end up with cropped or scaled pages, or with a printer waiting you to load Letter sheets to tray. That latter is very annoying when a) the printer is in other side of the building b) how to cancel the job is not obvious c) the "error" blocks the whole printer while you try to sort it out.

(Prime) Fractional sizes. While inches are fairly intuitive to me, and something like 1/4" is still good, I have no clue how much something like 7/32" is.

MPG for fuel consumption. This is arguably just a bad way to evaluate the economy of cars. Just swap to GPM (gallons per mile) if you want to keep your customary units.

Confusing list prices. The way sales tax and tipping works in the US is weird. Life is just simpler when you can just pay what the price is. It's also bit difficult to compare US prices to rest of the world prices.

Clothes sizing. Well this one applies to the whole world, but could we please sort this out someday? Of course it's bit funny to see something like "US: L, FR: S", but the funniness vanes out quickly when you try to find clothes that actually fit you.

In conclusion it is not the units that are the problem, but the way they are used.

If wanting to fit with stereotypes i think "US: L, FR: S" was meant to be the other way round.

MPG may not be a great measure but people generally have an easier time understanding 'bigger is better' so i think it works better than GPM.

It's worse than that for clothes sizes. Manufacturer A's large is not necessarily the same as manufacturer B's large.

Because Congress and Ronald Reagan: http://lamar.colostate.edu/~hillger/laws/usmb.html#disbandin...

    “The efforts of the Metric Board were largely
    ignored by the American public, and, in 1981,
    the Board reported to Congress that it lacked
    the clear Congressional mandate necessary to
    bring about national conversion. Due to this
    apparent ineffectiveness, and in an effort
    [by President Reagan] to reduce Federal
    spending, the Metric Board was dis-
    established in the fall of 1982.”

If there's anything that will cause the US to change to metric, it's the inefficiencies in manufacturing. Nothing we make to a non-metric spec can be exported. We have to duplicate production lines and tooling to build a metric version of anything. That is inefficient.

Nassim Taleb argues that the metric system is too abstract for people to relate to as where measurements like an inch (the size of your thumb) are much more easy to relate to. And as someone who studies usability I have to say that he's on to something.

To this day I still don't remember the exact number of feet in a mile. I know between 4-5 thousand, but the exact number is just absurd.

exactly 5280. Its not absurd. It has always been that number.

The mile has not even been the same size, it changed in 1959 https://en.wikipedia.org/wiki/Mile and it still is not at sea, where the nautical mile is longer. The Roman mile was 5000 feet.

I stand corrected.

not really. the mile was changed from 5000 to 5280 feet in order to make it an even number of furlongs (a furlong is 40 rods and a rod is 5.5 yard, the closest thing to 5 meters). you're in good company though, as there're also the geographical mile, the british nautical mile, the sea mile, the russian mile, the scots mile, the irish mile, the arabic mile and finally the original 5000 feet roman mile from where everything started, at least it was consistent over the entire roman empire. almost forgot: there's also the nautical mile but I'll spare that as it's really the only one whose definition is an angular measure applied to the earth surface instead of an arbitrarily decided unit.

I suspect it is mostly whatever you grew up with first. (BTW my thumb is 5cm or longer depending on where you measure.) If metric was difficult or less relatable then wouldn't people have noticed in the 190 or so countries that use it pervasively?

An inch is more or less the width of a thumb, not the length. It always surprises me how few people know that (is there a conspiracy in the USA, where parents teach their children about the inch at exactly the time their thumb is an inch long? When do they learn about the foot, then?)

That should be obvious from http://en.wikipedia.org/wiki/Ancient_Roman_weights_and_measu..., which states "The Roman foot was sub-divided either like the Greek pous into 16 digiti or fingers; or into 12 unciae or inches". So, Romans thought a thumb to be longer than a finger by 4:3.

I think it also makes sense when looking at the way one would measure stuff. How would you go about measuring something in thumb lengths with some accuracy? With widths, you can grab it between thumb and need finger of one hand, place thumb and index finger of the other hand next to it, move first pair to the other side, etc.

> An inch is more or less the width of a thumb

My thumb width is 7/8th of an inch, and I'm 1m94cm tall.

> is there a conspiracy in the USA

Although I live in the US, I am an immigrant who grew up with the metric system and consequently not part of said conspiracy.

What if you have big thumbs? In any case, you can always find connections: a cm is about the width of a fingernail, for example.

The most useful one in metric is the width of the open hand, from thumb tip to pinkie tip is about 20cm. It is surprisingly constant from person to person. I measure everything on the order of a meter with that, whether a piece of furniture will fit through a door, the size of rug I need, etc... I'm sure there are more "body measures" like this but this is the most useful to me.

You just use other things to relate to - one meter is a (somewhat larger than normal) step, your pinky is probably about one centimeter width and so on. There are as many things around you having simple sizes in the metric system as in the imperial system.

Well in my growing up 70s-80s in primary/high school, We got metric courses a few time and some of the books covered it (did you know that a meter is roughly three feet, the distance from the floor to a doorknob? I learnt that in school.)

In the early 80s there was another push many California road signs had dual postings of miles/kilometers to destinations. Cars also had dual gauges on their speedometers, large soda bottles were now in liters instead of ounces.

Then... it stopped again.

Wikipedia fills in some of those blanks http://en.wikipedia.org/wiki/Metrication_in_the_United_State...

I myself would like to see it change over, it would be a pain for a year or so, but then everything is standard with the rest of the world.

Yeah, I remember the same deal in the 80's. In Middle School out teachers made a point to tell us that the US had switched to the Metric system, and taught us all about metric units, etc. And, as you say, road signs often had both units, car speedometers had both, etc.

And then it just stopped. No notice, no noise, no nothing. Everybody just stopped talking about it, the road-signs went back to all Imperial measurements, car speedometers dropped km/h readouts. As far as I can tell, the only real legacy of metrification in the US is that anybody who does much tinkering with mechanical "stuff" that might involve bits that are imported, needs a toolbox loaded with both Metric and SAE sized tools. To be honest, I always half figured the whole Metric movement was arranged by Craftsman, Mac and Snap-On in order to double the volume of tools they sold.

For the same reason that we're all still on 12 months, 30ish days, 24h, 60m, 60s, 1000ms time.

When you put it like that, it's a bit weird. Is there an imperial fraction of a second that somehow never caught on?

This is because our system of time is based the earth circling the Sun and works best for things to slice up a circle, 360 degrees, base 6.

While we're at it. Can we have the whole world switch to UTC time and stop dealing with timezones and daylight savings?

I agree that DST is just madness but timezones are less insane in theory, they would make sense if they were determined by the longitude normalizing time so that sunrise and sunset happen at roughly similar times. Currently they're pretty screwy sure but the idea actually has some merit.

One thing to consider though is that switching everyone over to UTC you'd still have to consider local business hours. However, instead of just looking at their time zone and figuring 9-5 or standard hours you'd have to have a listing of each business's UTC hours. It might fracture more than timezones do currently.

Here's my argument for why DST is a good idea: http://messymatters.com/dst/

Everyone I have talked with about DST has had opinion, that the time to be chosen as base time used around year is the summer time. I would also change daylight at summer from 5am to 8pm, but i would also rather have in winter daylight at 16pm than 8am

If it were universally applied at the same time of the year it would be a more reasonable thing but because it's applied at different times (or not at all) depending on where you are in the span of a few weeks you have 2 times zones with 3 different offsets relative to each other.

That won't solve anything. Let's say that you leave London and 10:00 UTC, and land in another city at 15:00 UTC. Do you go to dinner? Breakfast? Can you schedule the business meeting at 17:30 UTC or this is just after midnight?

Just get an extra clock and keep it in UTC.

I worked in UTC for almost 3 years because it was company policy to have just one timezone for operations. Granted I was in Europe so the impact was minimal on my life, but not having DST and always working in UTC saved a great deal of problems when dealing with clients around the world on a real-time basis. The only organisation I ever had a problem with was the BBC because some of their people would regularly struggle to know which way was up.

Yes, that'd be lovely. No more talking about lbs and such. (I'm British but we were not taught imperial measures.)

But they'd still be saying things like

* "Where's Chechnya?" (if under 15 in 1995 you have an excuse)

* "I love being in England", when they're standing in "Glassgau" (Glasgow=Glazgo) - try it. Brave!

* "different than", "gotten", "color"...

Change to imperial units instead, just think of that extra 95ml of beer in every pint glass!

1 Cubic Foot = 7.48051948 Gallons [Fluid, US]

1 Cubic Foot = 6.42851159 Gallons [Dry, US]

1 Cubic Foot = 6.22883545 Gallons [UK]


I ask myself often why we still use inches in the EU: for pipe diameters (we still use NPS instead of DIN), trouser sizes, TV/LCD diagonal sizes (that seems to be changing)...

Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact