Each industry has a unique cost structure which basically means that every industry incurs a different set of costs. The textile industry is extremely labor-intensive so the set of costs for a textile factory would include a high labor component whereas a beer factory might have a high cost of capital component due to the types of machines used. So given a certain cost structure, is there such thing as an optimal factory size? Of course there is! And there are many different ways economists can find this "optimum" meaning most efficient but the easiest way was discovered by George Stigler in 1986 in his paper "Barriers to Entry, Economies of Scale, and Firm Size." His theory stated that if a factory is efficient, in the long run all other factories (with identical cost structures) should approach the size of that efficient factory. This means that if an industry's factories were broken down by size (as seen below), a reduction in the number of factories producing at the inefficient levels (normally very small and very large amounts) would occur over time.
This is what happened in the beer industry between 1959 and 1979.
Elzinga (1986)
You can see that beer factories which chose to produce under or at 2 million barrels decreased over time. To take an example, the smallest factories which produced 0-25 thousand barrels a year went from having 11 such factories in 1959 to just 2 in 1979. Stigler would say that a factory producing 0-25 thousand barrels a year is inefficient. On the other end of the spectrum, a factory producing over 4 million barrels a year is extremely efficient since the number of factories at that capacity increased from 2 in 1959 to 20 in 1979.
Did the inefficient factories close up shop within the 20 years between 1959 and 1979? Most likely not. The smaller factories might have realized that they were not taking advantage of economies of scale (spreading large fixed costs over increasing amounts of quantity) and increased their capacity, adding to the numbers of bigger breweries. This indicates that before the mid-1980's, economies of scale played a major role in the production of beer and that firms realized the maximum benefit of economies of scale the larger they were.
The number of breweries decreased from 465 in 1947 to 80 in 1983; an 82% decrease. This trend started after Prohibition when the drinking environment had been radically changed. Prior to Prohibition, establishments were allowed to make and sell their own alcohol and therefore, there existed many small brewpubs and breweries that owned bars, 1566 in fact (nearly the number there are today!). But after Prohibition, alcohol manufacturers were prohibited from owning bars and/or retail outlets. Therefore, the small breweries and brewpubs that had existed prior to Prohibition exited the market due to dramatically reduced profits that had previously come from their retail operations. The larger breweries took advantage of a market in which most rivals had exited.
Today there are 1,599 breweries in the US; nearly 20 times the number of breweries in 1983 and 3.5 times of that which existed at the peak in 1947. One might simply say that Americans have increased their preference for beer but an economist might choose to explain it in a different light.
First, government lifted the ban on breweries having retail operations, creating more opportunities for smaller breweries, as seen prior to Prohibition. Most economists will agree, less regulation creates more opportunity for innovation, profit maximization, and reduces barriers to enter a market.
Secondly, a new technology was created in the late 80's. That innovation was the unheard of concept of competing on the characteristics of the beer rather than the price. Into the 80's, beer producers were still competing on price and general taste but coming into the 90's, microbreweries began to pop up with boutique beers with interesting flavors that hadn't been seen for many decades. The price charged was not comparable to those of the large breweries but this didn't matter - it was an entirely different product. Today, microbreweries only control 5-7% of the beer market, which if you live in Oregon is hard to imagine that statistic being so small.
So does the emergence of microbreweries debunk the conclusion that breweries are more efficient the larger they are? Definitely not! Economies of scale refers to cost savings that occur due to increased production, ie less inputs making more outputs! Since the cost structure for craft breweries is very different from that of large non-craft breweries, both types could very well be operating at minimum cost and be benefiting from economies of scale.