The Singular Beverage Experience

dr_stringfellowBy The Metric Maven

GAO Report Edition

There are two chestnuts that are perennially trotted out to “show” that the US is on its way to metric. One is the 2 liter bottle of soda, the other is the fact that liquor is sold in metric quantities. The 1978 GAO Report chapter on beverages indicates a quick change took place in the 1970s, and then the metrication of beverages was frozen in place at near zero kelvin. What happened has become mythologized in the last few decades. The GAO Report is interesting as it comprehensively looks at all the beverage sectors. It is also an example of how tone-deaf Americans are when dealing with metric—something that could be remedied in five minutes.

The report begins by explaining that:

Wines and distilled spirits are converting their products to metric sizes for marketing reasons. Both are regulated by the Department of Treasury’s Bureau of Alcohol, Tobacco and Firearms; however, the producers requested the change. A considerable portion of their products are now being sold in metric sizes.

The Report notes that soft drink manufacturers have introduced products with metric volumes in many areas of the US. However, the soft drink industry “did not plan an overall metric conversion in the near future.” Milk has labels with metric equivalents, but had no plan to change-over to metric sizes.

The beer industry sells all its products in customary sizes and did not plan to convert to metric sizes. Some brewers showed metric equivalents on their labels.

This statement was made decades before the recent craft brewing movement began. Their current marketing seems to be based upon unit proliferation.

The Report first discusses the wine industry. They decided to convert their entire product line to metric. Amazingly:

…it was the Wine Institute, a trade association representing California wine producers, that petitioned the Bureau to convert to metric sizes and reduce the number of permissible sizes.

The organization wanted to reduce the number of permissible sizes from 16 customary to 7 metric sizes. A table is then included that shows only 9 of the 16 customary sizes were in general use.

The Wine Institute made the request because many imported wines were being sold in bottles containing up to several ounces less than the 4/5 quart (the fifth 25.6 ounces), the most common size used by domestic  products. The bottles used for imported wines appeared to contain the same amount of contents as those used for domestic wines. The industry believed the practice was deceptive to consumers and gave foreign producers an unfair competitive advantage.

At a hearing held on the Wine Institute request, it was brought out that imported wines should not be required to use customary-size bottles because the National Bureau of Standards study had recommended that the United States switch to the metric system over a 10-year period. Subsequently, the Bureau denied the request because it considered it inappropriate to require foreign wine producers to use customary-size bottles for sales in the United States.

It appears that the wine producers first wanted to force foreign manufacturers to produce wines in Ye Olde English, but because alcohol is one of the few government regulated industries in the US. The NBS had the authority to tell the wine producers they had to switch to metric. It appears as clear as a summer day, that given their choice, the wine manufacturers would have forced foreign manufacturers to produce a set of bottles in Olde English for the US and metric for the rest of the world.

As it was:

The Wine Institute selected the 750-milliliter (25.4 ounces) size as the primary size because it was very close to the 4/5 quart (25.6 ounces) which comprised about 48 percent of the industry’s sales. The 750 milliliter was also used in other countries. Four other metric sizes–the 3 liter, 1.5 liter, 375 milliliter and 187 milliliter–were selected by the Wine Institute because they were multiples or submultiples of the 750 milliliter and thus would enable consumers to make price comparisons between sizes. Selection of the 375 and the 187 milliliters also permitted continuation of sizes similar to those consumers and the industry were familiar with.

Miraculously, the NBS held a public hearing on the subject, and all the persons who spoke were either for the conversion or didn’t oppose it. The NBS took written comments. 40 comments were submitted, and only three were opposed to the metric switch-over.

In June of 1974, the NBS approved seven metric sizes. Six had been requested by the Wine Institute plus the 100 milliliter which “had been requested by foreign wine producers, importers, and airlines to permit importation of sherries sold in one-person servings. The conversion period would be 4 years beginning in January of 1975 and ending on December 31, 1978. Four years was chosen to allow for existing glass molds to wear-out. Foreign wine producers would now have to conform to the US metric size requirements.

The conversion process was about complete in October of 1978, when the GAO Report was issued. The amount of time given allowed for the orderly transition that took place. The largest snag was not technical but legal. Tax law was written using Olde English units, and the producers had to convert the values for tax purposes. In this case the producers put the conversion to good use by introducing reforms that might not otherwise have taken place:

… One wine company official told us that about $12,000 annually in storage costs will be saved because of the new shape, a 375-millilier bottle, will be used to replace three, 4/5-pint bottles that the company previously used. Changes in other bottles could also result in some savings to the industry. These changes could have been made without converting to metric, but the metric conversion was viewed as providing the opportunity to make the changes.

Pat Naughtin would most certainly have been pleased.

This is a good moment to look at the tone-deaf use of metric units by the authors of the Report. The first is the archaic Olde English notion that one should switch from milliliters to liters. Rather than switch to liters, one should use Naughtin’s second and third laws to make the numbers all integers and not switch metric prefixes:

100 mL, 187 mL, 375 mL ,750 mL, 1000 mL, 3000 mL

Indeed, one can use the 750 mL size as a standard price touchstone, but only if the store carries that particular size for the wine you desire. A go-to way to describe the price is in dollars (or cents) per 100 mL.  If the 100 mL item is shown as $2.00 per 100 mL and the 750 mL bottle has $1.45 per 100 mL, then it is very easy to compare, just like metric fuel efficiency. Each item should have a price per 100 mL for comparison.

The GAO next indicates that consumers will receive limited benefits. They immediately argue:

The 100 milliliter will be more difficult to make comparisons with than the other sizes; however little use will be made of this size, and consumers will feel little impact.


… Also, sizes such as 4 and 5 liters cannot be easily compared in volume to other metric sizes, such as the 750 milliliter and the 1.5 and 3 liter, thereby defeating another of the original aims of metrication.

No, just demand that shelf pricing include the price per 100 mL for each size.

The Department of the Treasury was criticized for waiting until two years after the metric wine conversion to issue the chart below:


Once again, we see the metric sizes treated like Olde English sizes with liters and milliliters, rather than just one unit. The cultural inertia to continue with factors of two makes metric seem unwieldy. This practice continues on their more graphically designed chart:


The distilled spirits industry also converted to metric and decided to reduce the number of sizes. The GAO seems shocked that they decided to do so using metric: “The size reductions could have been achieved without metrication.” Again a table is offered for the new sizes:


The metric sizes are more rational at 50 mL, 200 mL, 500 mL, 750 mL, 1000 mL and 1750 mL. The chart immediately changes the 1000 mL and 1750 mL to liters, which introduces a cognitive discontinuity.

The conversion is also being carried out under regulations prescribed by the Bureau. However, it was the Distilled Spirits Council, a trade association which represents about 95 percent of the distilled spirits industry, that petitioned the Bureau for permission to convert to metric sizes and reduce the number of permissible sizes. Among the reasons given by the Council for wanting to convert were to (1) reduce production costs, (2) permit marketing and distribution efficiencies, (3) provide better service to the public, and (4) promote exports.(26-18) and (26-19)

What precipitated the change?:

The legislation by the Congress directing NBS to study metrication also stimulated the industry to consider converting to metric.

The agreement by the European Economic Community to use 17 metric sizes for trade among the member nations was also a factor. There were arguments about what sizes to adopt, but not the fact that the industry would become metric.

When the conversion was announced in 1976, the Bureau Director stated:

..the change to metric would (1) reduce significantly the number of bottle sizes, (2) provide enough separation between sizes to deter possible consumer deception, and (3) make calculations easier because of round numbers. [italics mine] (26-20)

It is good to see that at least one person saw the advantage of integers.

The cost of conversion did not seem to be much of a deterrent: “One distilled spirits official described the conversion as posing no difficult problems for his company. He said the company is constantly making changes in its operations. New bottles are introduced, new labels developed, and production lines are adjusted to handle bottles of different sizes with varying contents. He viewed the change to metric as just another change; one not much different than his company faces on a day-to-day basis.

Another official of a distilled spirits producer told us that his company estimated it would cost $1.5 million to make the conversion. ….. This amount, however was not considered substantial in that it amounted to less than 0.5 percent of the company’s annual sales.”

One company didn’t seem to see the costs as attributable to metrication as they had timed the transition to coincide with a time when they were replacing worn-out glass molds with new ones. The GAO spends more time asserting that metric will make price comparisons harder, and still does not mention price per 100 mL as an option. Once again, milliliters and liters are implemented by the Treasury:


A more graphically designed poster again switches from milliliters to liters, and argues that it is simplicity:


The choice of 750 mL is based on one-fifth of a gallon. This has produced an incompatibility with Europe and affects our international trade. The operators of Montanya Distilleries in Crested Butte, Colorado make rum, and wanted to expand their sales into Europe. The problem is that US bottles are 750 mL and European bottles are standardized at 700 mL. The distillery could not obtain 700 mL bottles in the US so they had two choices. The first was to import 700 mL bottles from Europe, fill them with rum, and then export them back to Europe.  This was definitely cost prohibitive. The second option was to obtain permission to send the rum over to Europe in bulk and have it bottled there. The distillery chose the second option, which employs Europeans and their glassware industry.

One can argue that 750 mL makes the most sense as it is half-way between 500 mL and 1000 mL. But it is the only case of doubling in the entire chart.  It would have made more sense to go with 700 mL as it fits well with price comparisons using price per 100 mL and would increase our compatibility with other countries. Our government never seems to find a way to ever rationally study and act upon problems like this one, and US industry and commerce suffers as a consequence—but we preserved a metric version of our Ye Olde English volume!

The soft drink industry is examined next by the GAO, and here there is a bit of a surprise:

Use of metric size containers for soft drinks began in April 1975 when  a major soft drink company introduced the liter size in one of its marketing areas. The company wanted to use the new bottle which, while shorter, would cost less since it contained less glass and would perform better on production lines. The new bottle would also permit a 20-percent savings in space for bottlers because (1) more cartons could be placed in the same amount of space, (2) increased payloads would be possible for trucks, and (3) more storage capability would result in warehouses. Also, customers preferred the new bottle. (26-30)

The company viewed using the liter as an opportunity to be the first in the soft drink industry with the metric system.

Yes, the liter was first introduced in the US and not the two liter bottle. “… the company also began use of 1/2-liter and the 2-liter sizes. Both refillable and nonrefillable metric-size bottles were used.” (26-31)

But, like beer, there was a demarcation:

All the metric soft drink changes that we are aware of involved soft drinks sold in bottles. No soft drinks were being sold in metric-size cans. Soft drink industry officials told us that no conversions involved cans because the costs to convert can production facilities would be too high, about $1 million for each can production line. Also, concern was expressed on the impact changing can sizes would have on vending machines.

It is clear that machines wear-out and small aluminum cans of soda have been introduced over the years. It is very difficult to believe that in the last three decades an opportunity did not arise to make soda cans metric. The desire to decrease the amount of material in each can is monumental, would not going from a 355 mL can to a 350 mL can make sense? In recent years aluminum “bottles” of soda have made their way onto US shelves. Could they not have been changed over to metric from 251 mL to 250 mL?


The soft drink industry also saw voluntary metric conversion as meaning optional metric conversion. They assume that government would impose how they would do metric rather than just saying you have to go metric, you figure out how you would like to do that.

30lThe GAO praised the soft drink industry for using “1/2 liter, liter, and 2 liter–[which] are multiples of one another.” This in their view made it easy for consumers to compare prices.” If the soft drink industry had actually been compelled to convert then what is said next would not still exist:

Soft drinks are sold in 15 customary size containers–6-1/2, 7, 8, 10, 12, 16, 24, 26, 28, 30, 32, 36, 48, and 64 ounces. Many of these sizes do not lend to easy price comparisons with other sizes. Also many soft drinks are sold in cartons having 2, 4, 6, 8, or other quantity containers. Because of this, the soft drink cartons consumers purchase come in sizes such as 39, 42, 96, and 128 ounces–quantities which do not lend to easy price comparisons.

If there were a complete conversion to metric sizes, this could help facilitate consumer price comparisons, particularly if rational sizes like 1/2 liter, liter, and 2 liter are used. For example, the price of a carton of eight 1/2-liter bottles (4 liters) could easily be compared with the price of four 1-liter bottles or two 2-liter bottles. The same advantage could be achieved, however, if soft drinks were sold in customary sizes that were multiples of one another.

The GAO just had to take a jab at metric. Without a metric switch-over what opportunity might present itself to change the custom(-ary) units to a more rational series? Interestingly The Onion has it’s own take on the situation.

We next move on to milk:

The sizes in which milk may be sold is strictly regulated by various state laws. At present, the 1/2 pint, pint, quart, 1/2 gallon and gallon are the most commonly used sizes. They are all multiples of one another which makes price comparisons for consumers simple.

Really? compare the half-pint to the half-gallon and get back to me on that. The paragraph goes on:

Other sizes permitted include the gill, an amount equal to 1/4 pint, and 3 quarts. The 3-quart size is permitted in about 20 states but is not widely used.

This points out yet again that weights and measures have been left to the states, and the Federal government has been asleep at the switch since at least 1797. This means:

The States have not approved the sale of milk in metric sizes. … It appears that if conversion occurs, the most frequently used metric sizes would be as follows.(26-35)


Ok, let me help the GAO out here and show them how to use Naughtin’s Laws:


They point out that “Each of the above metric sizes is about 6 percent larger than the customary size it would replace.”

Beer producers vigorously opposed any metrication efforts that would make its container sizes obsolete. Again:

The sizes in which beer is sold is regulated by the States and many different sizes are authorized. Under the Federal Alcohol Administration Act, the States are provided the authority to regulate beer sizes. The sizes most widely used are 7, 12, 16, and 32 ounces. Revisions of laws in many States would be necessary if the industry converted to metric sizes

Strangely, it was cost effective for the soft-drink industry to change their bottles, but the replacement of 12 ounce bottles of beer would be prohibitively expensive?

Several industry officials believed it was inevitable that the industry would convert to metric. But, they believed conversions would not occur for many years unless required by the government. …. Industry officials believed that the United States should continue a voluntary policy of converting to the metric system. (26-36)

The GAO Report has an interesting history of the mixed-bag that is the beverage industry. The wine and distilled spirits world became metric, the soft drink industry added metric to the mix, and the milk and beer industries flat-out rejected metric. Why the consensus for metric occurred in the wine and liquor industries and how government appeared to have made it happen is still mysterious, the beer brewers were completely against it and had their way. What the Report contains is a description of the sudden change that occurred in a small segment of the US economy in the 1970s. It is a fossil that describes the living-fossil that is the United States of America when it comes to weights and measures.

If you liked this essay and wish to support the work of The Metric Maven, please visit his Patreon Page

Related essay: Taking The Metric Fifth


The Metric Maven has published a book titled The Dimensions of The Cosmos. It examines the basic quantities of the world from yocto to Yotta with a mixture of scientific anecdotes and may be purchased here.


The Right Way, The Wrong Way and The American Way

Battlebots Trophy -- A nut with a 5 inch diameter hole with 4 threads per inch

The Battlebots Trophy — A nut with a 5 inch diameter hole and 4 threads per inch

By The Metric Maven

GAO Report Edition

Fasteners are ubiquitous. In physical design, they are everywhere. The image to the left is the current Battlebots Trophy, a giant nut. As of 2016, this nut celebrates US obsolescence with its archaic 5 inch diameter hole and four threads per inch.  The 1978 GAO report has an interesting chapter on fasteners, what happened, and what didn’t. The first sentence of the fastener chapter reads:

The U.S. fastener industry which was originally opposed to metrication, began conversion efforts in 1970 in order to maintain its markets.

The industry found that in the 1960s their major customers were moving toward the metric system. One would think the US fastener manufacturers would have been in favor of metric by 1978, but the story is more complicated than expected and perhaps too American. The Report has a nice description of a fastener:

A fastener is anything which holds two things together. Nuts, bolts, screws, rivets, cotter pins, and nails are a few examples. (See following page.) Of these, the United States produces approximately two million different types. Fasteners can hold together a vast number of items. For example, a telephone is held together with about 70 fasteners. Jumbo jets contain millions; and for one model, fasteners costs represent about 10 percent of the plane’s total cost. In short, much of the nearly $2 trillions U.S. economy is held together by the $2 billion fastener industry.

The report notes that a considerable increase in the use of metric fasteners is taking place in the US. The domestic fastener industry was also under pressure from imported Olde English fasteners. At the end of the 1960s, no US engineering standard for metric fasteners existed, but an international standard did. US industry representatives claimed that the international standard had too many sizes and thread types. The values of these sizes did not follow a logical pattern it was alleged.  If the US fastener industry was going to become metric, it was argued that the US should create a new fastener system that was:

….as perfect as possible. Also, the industry did not want to give a competitive advantage to foreign producers of metric fasteners. It was felt that the foreign producers would gain an advantage if the U.S. industry merely accepted the existing international  standard for metric fasteners in its entirety.

Yes, we in the US were going to produce a “more perfect fastener” or perhaps even a perfect fastener, and in January of 1971 the report “A Study To Develop An Optimum Metric Fastener System” was released by the Industrial Fasteners Institute. The study was presented to the ten largest corporations in the US as well as the National Bureau of Standards (NBS), and technical bodies in Canada. The selected group was unanimous in its view that a detailed study should be undertaken. The GAO Report states:

The Committee’s ultimate objective was to design a metric fastener system which would be so attractive technically and economically that it would become the single internationally accepted system of threaded fasteners. (7-5)

An unshakable US faith in technical Darwinism, coupled with the belief the US would create the fittest fastener meme propelled this new study. The Special Committee published its results in 1973. It recommended a fastener system with 25 sizes and a single thread type. The first metric fastener standard based on these recommendations was released in 1974. Before the standard was completed, the new system was encountering international resistance. Britain and German standards representatives released a paper called “Why Should the International Standards Organization System for Metric Fastener Threads be Changed?” It argued that the costs and confusion were unwarranted, “the technical advantages were minimal, and the system could hardly be called “optimum.” There were complaints of protectionism and everyone having to start all over again. (7-6).

The discussions continued from 1973 to 1977 as the ISO negotiated with its US members. The US representatives finally backed off from the proposed changes to the international standard. The US standard became essentially the same as the preferred series of the ISO standard.

There was controversy about the strength grade of fasteners in the 6 to 18 millimeter range. Europeans used an international strength grade of 8.8. It has a strength capacity of 116 000 pounds per square inch. The comparable US SAE was grade 5, which has a strength of 120 000 pounds per square inch. This is about a three percent difference. It was recommended the next higher grade 9.8 be used. This fastener has a strength of about 130 500 PSI.

The Europeans went along with the proposed change, but only the US automotive industry adopted the higher grade. US farm equipment, Canadian and European manufacturers decided to use 8.8 for their threaded fasteners. The unavailability of fasteners that met the US requirement caused concern that an 8.8 fastener could be interchanged for a 9.8 version during a repair. If 9.8 was not available, it would be necessary to use 10.9, which requires an alloy steel.


The report next focused on the head sizes for the fasteners:

A major problem arose during the attempt to reach agreement on the hexagon head size for three fasteners. This was probably the most hotly debated and difficult issue considered during the 1977 ISO meetings. The schedule below shows the head sizes wanted by the United States, those used in Europe, and those agreed to at the meetings.


The Optimum Metric Fastener System study had shown that the head size for a number of fasteners was unnecessarily large. International standard sizes were widely used in Europe, but the European representatives had in 1975 agreed to reduce the head size 1 millimeter on each of the three sizes. The U.S. representatives agreed to the compromise sizes in the earlier meetings, but in 1977 returned to the demand for a smaller head for the 10-millimeter fastener size.

The Europeans would not approve an inclusion of a 15 mm head and the US would not compromise. The official standard became 10, 12 and 14 millimeter diameter fasteners with 16, 18 and 21 millimeter heads respectively. The US would use these and the 15 mm head. It became possible that several head sizes might be used for these three fastener sizes. The Report noted:

Head sizes (like strength grades) are an example of an international standard which is formally agreed to on paper but not uniformly adhered to in practice. (7-8)

The European view was that the benefits of the changes to the new system did not justify the expenses involved. The fastener standard is voluntary, and the US could do whatever it wanted. This impasse could leave US fastener manufacturers holding the bag. The GAO report states:

An official of one company told me he had stuck his neck out and stocked six metric sizes in 24 lengths. The stock included the 6.3-millimeter fastener which was one of the U.S.-proposed sizes that did not gain international acceptance. This size was being used by a major automobile manufacturer in its 1977 and 1978 models. However, the automobile manufacturer has dropped it for future models.

It was noted that maintaining Ye Olde English and metric fasteners in the US could cause considerable difficulty:

It is virtually impossible to visually identify some sizes of customary-threaded fasteners from similar-size metric fasteners. It is possible to mismatch 36 combinations of customary- and metric-threaded fasteners. The result could be either stripping during assembly or full assembly with 25- to 60-percent loss in load capacity. Thus, the accidental mismatch of fasteners could result in fastener failures.

This is a very good argument for a quick metric switch-over, with an M-day, and no “transition period,” rather than waiting for the magic of the Metric Philosophers technical Darwinism to accomplish this task over an undefined period.

On April 25, 2014 (2014-04-25), Joe Greenslade of the Industrial Fasteners Institute gave a presentation titled “Metric Fastener Standards Transition”  His view is that one metric fastener system should be used throughout the world. Mr. Greenslade calls the US attempt to create an “Optimum Metric Fastener System” a “misguided move!” He claims that from 1975 to 2013 there has been a slow but gradually accelerating adoption of metric designs.

Greenslade identifies three different fastener systems, ISO (International), DIN (German) and ANSI/ASME/ASME/SAE (US). He sees the US Optimum Metric Fastener System (OMFS) as a misguided philosophy of “since we must change we will do it better than you Europeans do.”  The OMFS attempted to eliminate fine threads, this “simplification” was rejected. The US introduction the M6.3 X 1 fastener simply because we wanted a metric version of a 1/4-20 Ye Olde English fastener, rather than using a standard M6 x 1.0 was rejected. The introduction of a new thread gauge was not accepted. The US wanted to replace the hex head with a new spline head, but that was also rebuffed. The changing of hex sizes (head sizes) by 1 millimeter on M10, M12, M14, and M22 is still causing confusion to this day. The US has finally withdrawn its proposed “optimum” metric standard.

The two metric standards that remain are ISO and DIN. DIN is very, very close to the ISO standard. They are 99.99% interchangeable, and 90% identical. The German DIN standard is to be replaced with the ISO standard. When US customers now ask for ISO they are often told “we do not stock any ISO — only DIN.” Greenslade indicates that a search for dual DIN/ISO designations on existing drawings and parts lists should be undertaken, and in these instances they should be edited so that only the ISO number is used. He also suggests that all new product design drawing designations be only ISO. Greenslade offers numerous examples of this existing redundancy “out in the wild.” The long term objective for the USA should be to use ISO and forget the past.


baskin-robbinsThe US introduction of an “optimum” standard in the 1970s has the fingerprints of American hubris all over it. Rather than finally bring some order to the chaos that is side by side Ye Olde English fasteners and metric, by eliminating the “custom”-ary versions, and using ISO metric exclusively, we instead opted to show everyone “how to do it better.” History has not judged us favorably, and the exercise in imposing a US metric “standard” on the world continues to cause discord and confusion to this day. As has been said many times “the great thing about US standards is there are so many to choose from.”

If you liked this essay and wish to support the work of The Metric Maven, please visit his Patreon Page

Related essays:

Without Metric Threads We’re Screwed

A Hole in The Screw Head

The Metric Maven has published a book titled The Dimensions of The Cosmos. It examines the basic quantities of the world from yocto to Yotta with a mixture of scientific anecdotes and may be purchased here.