"Opening Up Innovation: Strategy, Organization and Technology".Discuss

Paper to be presented at the Summer Conference 2010

“Opening Up Innovation:
Strategy, Organization and Technology”
at
Imperial College London Business School, June 16 – 18, 2010

THE FALL AND RISE OF THE LOCAL BREW
Peter Swann

Nottingham University Business School

Abstract:
In the presence of strong economies of scale, a fall in transport costs can lead to a more pronounced geographical concentration of production. This is very apparent in the growing concentration of breweries in England from 1900-1970. The number of breweries in England fell sharply between 1900 and 1970, and so also did the range
of locations in which they were located. In 1900, many small villages could boast a brewery of their own, but by 1970, the vast majority of breweries were located in towns and cities. However, as the number of discerning

consumers has grown, with a pronounced taste for horizontal product variety, the tendency towards geographical concentration has declined in the last forty years. The consumer group CAMRA, arguably the most successful consumer group ever formed, has been instrumental in educating English beer drinkers to appreciate horizontal product differentiation in beers and ales. Since the foundation of CAMRA in 1971, we have seen many new micro-breweries enter the industry, and a rapidly increasing geographical dispersion of these micro-breweries.

JEL – codes: R30, L66, O18

The Fall and Rise of the Local Brew

Process Innovation, Horizontal Product Innovation and the
Geographic Dispersion of Breweries in England, 1900-2004
Paper written for possible presentation at
The DRUID Conference, London 2010
Version 1
19th February 2010

Abstract

In the presence of strong economies of scale, a fall in transport costs can lead to a more pronounced geographical concentration of production. This is very apparent in the growing concentration of breweries in England from 1900-1970. The number of breweries in England fell sharply between 1900 and 1970, and so also did the range of locations in which they were located. In 1900, many small villages could boast a brewery of their own, but by 1970, the vast majority of breweries were located in towns and cities. However, as the number of discerning consumers has grown, with a pronounced taste for horizontal product variety, the tendency towards geographical concentration has declined in the last forty years. The consumer group CAMRA, arguably the most successful consumer group ever formed, has been instrumental in educating English beer drinkers to appreciate horizontal product differentiation in beers and ales. Since the foundation of CAMRA in 1971, we have seen many new micro-breweries enter the industry, and a rapidly increasing geographical dispersion of these micro-breweries.

1. Introduction

The objective of this paper is to explore the relationships between horizontal product differentiation and demand for variety, on the one hand, and the geographical dispersion of activity on the other. This is done with reference the history of the English 1 brewing industry between 1900 and 2004.

A fundamental result, explored in a series of publications by Krugman (1980, 1991, 1992), has shown that in an industry producing a transportable and standardized commodity where production is subject to increasing returns, a decline in transport costs will tend to lead to greater industrial and geographical concentration. This happens because large producers in a major city enjoy a sufficient cost advantage over smaller producers in neighbouring towns and villages that that they (the large producers) can undercut the smaller producers in their local market. As a result, production becomes ever more concentrated amongst a few large scale producers who then transport their commodity for sale throughout over an ever greater region.

The history of the brewing industry in England from the start of the 20th century until the early 1970s is a striking example of this phenomenon. According to data from the Brewery History Society (2005), the number of distinct breweries in England fell from 1,324 in 1900 to 141 in 1975. Not only did the number of breweries decline rapidly, but so also did the range of locations in which they were located. In 1900, many very small villages could boast a brewery of their own. By 1970-1975, the vast majority of breweries were located in towns and cities, with only a few exceptions. And indeed, 141 breweries in 1975 does not give a true picture of the extent of concentration, as a large share of the industry was controlled by the ‘big six’ brewers.

Why did this happen? Two essential technological factors which enabled this concentration were these: (a) technical change in brewing which made the product more easily transportable; and (b) a reduction of transport costs over the last 100 years. One particular development which was particularly influential was the development of ‘keg’ beer, or ‘brewery conditioned’ beer which was sterilised in the brewery and had CO2 added, making it easier to transport over large distances.

Two essential strategic influences on concentration were: (a) the successive history of mergers and acquisitions, accompanied by the closure of small breweries and the relocation of production to new, large scale breweries; and (b) the substantial investments made by the big brewers in developing national brands of beer which could be sold over the whole country. These brands would be highly standardised products with little regional variation.

Some suggested that these developments were in the consumer interest because they would lead to cheaper beer for the consumer (e.g. Vaizey, 1960). Moreover, some have commented on the very variable quality of beer before these developments. Others regretted the loss in regional and local variety,2 which was part of the pleasure

1 We have not included data on breweries in Wales or Scotland as the historical records in Brewery History Society (2005) are not so detailed.

2 The importance of regional product variety to regional pride and identity was memorably summarised by C. de Gaulle: “How can you govern a country that has 246 different varieties of cheese?” (here quoted from Jay, 1996, p.112)

1
of beer drinking, and also regretted the sterilisation of beer, which they considered made beer a rather bland drink. Out of this discontent arose CAMRA (The Campaign for Real Ale), formed in 1971. The founders were concerned about the decline of variety and the blandness of modern beer and sought to create a pressure group which would persuade the brewers to change their strategies. CAMRA has been described as the most successful consumer pressure group ever formed.

It is arguable that CAMRA were unable to stop the progress of consolidation and standardization amongst the largest brewers. But they were very successful in educating an increasing number of beer drinkers about the finer points of ‘real ale’ and the pleasures of regional diversity. Since CAMRA appeared on the scene, a large number of small new entrants have appeared who boosted the number of distinct breweries to 480 by 2004 – roughly the same figure as in 1939. And a most noticeable increase is seen in the geographical dispersion of these new entrants.

This paper illustrates and explores two different episodes in the history of the English brewing industry: the first from 1900-1970 and the second from 1971 onwards. The first episode is of industrial and geographical concentration, product standardization and transportation. These trends continue in some parts of the industry beyond 1970, but during this second episode we also see an interesting counterpoint: a story of increasingly educated and discerning consumers with a taste for variety, the entry of new producers of real ale, increased regional and local variety and a geographical dispersion of small breweries.

The rest of the paper is organised as follows. Section 2 briefly illustrates two very simple variants of the basic Krugman model. In the first, falling transport costs leads to industrial and geographical concentration, when there is no horizontal product differentiation and no taste for variety. In the second, where there is horizontal product differentiation and a marked taste for variety, this process of concentration only happens if there are strong economies of scope. Sections 3 and 4 examine the first historical episode (1900-1970). Section 3 describes the main technological and strategic factors behind increasing concentration, and Section 4 illustrates this with data from the Brewery History Society (2005). Sections 5 and 6 examine the second historical episode (1971-2004). Section 5 describes the activities of CAMRA and the technological trends that made the widespread entry of micro-breweries possible. Section 6 illustrates the effects on entry, total brewery numbers and geographical dispersion – again using data from the Brewery History Society (2005). Section 7 concludes by trying to reconcile the different features of these two historical episodes.
2. Two Models of Declining Transport Costs and Geographical Dispersion

As noted in the introduction, a fundamental result, explored in a series of publications by Krugman (1980, 1991, 1992), has shown that in an industry producing a transportable and standardized commodity where production is subject to increasing returns, a decline in transport costs will tend to lead to greater industrial concentration. This happens because large producers enjoy a sufficient cost advantage over smaller producers in their region. As a result, production becomes ever more concentrated amongst a few large scale producers who then transport their commodity over an ever greater area.
2
As the spirit of this paper is essentially empirical, we don’t offer a formal statement of the Krugman model, just the simplest possible graphical representation. This is undoubtedly a highly simplistic account of what has happened in this industry. However, by presenting two variants of this basic model, we can provide an (incomplete) explanation of some of the differences in the two historical episodes under consideration. In the first variant, falling transport costs leads to industrial and geographical concentration, when there is no horizontal product differentiation and no taste for variety. In the second variant, where there is horizontal product differentiation and a marked taste for variety, this process of concentration only happens if there are strong economies of scope.

Variant 1

The first variant is illustrated in Figure 1. This illustrates the geographical distribution of population (the diamonds) and breweries (the circles) in a hypothetical small country. There are four regions (NW, NE, SE and SW) with boundaries defined by the black lines. There are four breweries, each located in a different region. In part (a) of the Figure 1, transport costs are relatively high. To make it as simple as possible, let us assume that transport costs within a region are negligible while transportation across regional boundaries is expensive.3 In that case, it is only viable for each of the regional breweries to operate within its own region; it is unable to compete outside its region.

Then part (b) of Figure 1 shows the situation after a reduction in transport costs which make it viable for any of the regional breweries to operate in any of the four regions. So it is now as if the regional boundaries have been removed. Again, to make the exposition as simple as possible, assume that transport costs anywhere within the country are now negligible. If beer is an undifferentiated commodity, and the products produced by the four regional breweries are identical, then competitiveness is defined by cost alone. As it stands, the brewery in the SE is the largest and as there are economies of scale, this SE brewery will have the lowest average costs.

One possible outcome (and this has been very common in the brewing industry) is that the largest and most profitable brewery acquires the smaller breweries, closes them and expands production at its main site (in the SE region). The outcome is shown in part (c) where a single large brewery now supplies all of the country.

This model is a gross simplification, of course, but it captures an essential idea. Declining transport costs can undermine regional or local monopolies and make it viable and profitable for a large brewer to pursue a three pronged-strategy: mergers and acquisitions with smaller breweries; consolidation of production at one site to exploit economies of scale; and the marketing of national brands which are transported around the country.
3 This non-linearity in the transport cost function is a bit implausible, but it greatly simplifies the exposition. The essence of the result does not depend on it, however.
3
Figure 1
Declining Transport Costs, Without Horizontal Product Differentiation
(With Economies of Scale)

NW NE

SW SE

(a) High Transport Costs (b) Low Transport Costs (c) Low Transport Costs
Before Concentration After Concentration
4

Variant 2

The second variant of this model is illustrated in Figure 2. This has the same starting point as in Figure 1 – except that the beers produced by each of the regional breweries are distinctively different (and are marked with different shades of grey in Figure 2). Part (a) shows the case of relatively high transport costs where, once again, each regional brewer supplies all its beer within its home region and none outside the region. Part (b) then shows the situation with relatively low transport costs where, as before, we make the simplifying assumption that transport costs are negligible within the country, so any brewer can supply any location at the same cost. And Part (c) shows an outcome that may or may not happen: this is the case where one brewery located in the SE region produces all four types of beer and supplies these to the whole country.

The key difference between Figure 1(b) and Figure 2(b) is the role of product differentiation. In Figure 1(b) the question was this: can a single plant located (say) in the SE region supply the whole country with a single type of beer at lower cost than the four existing breweries? There the answer depended on economies of scale. If economies of scale are strong enough the answer is, yes, and the outcome is shown in Figure 1(c).

In Figure 2(b) we should ask a subtly different question: can a single plant located (say) in the SE region supply the whole country with all four different types of beer at lower cost than the four existing breweries? Here the answer depends on economies of scope. If economies of scope are strong enough, then the answer is, yes, and the outcome would be as shown in Figure 2(c). If economies of scope are not strong enough, then the answer is, no, and the outcome would be as shown in Figure 2(b).

Or a further possibility is that the pattern in Figure 2(b) is replaced not by 2(c) but by 1(c). That is, after the decline in transport costs, the four breweries merge, but instead of trying to produce all four regional beers, the new enlarged brewery just produces one. Is that a viable outcome? It will depend on the strength of demand for variety. If the demand for variety is strong enough, then 1(c) would not be an equilibrium, because entrants would find it profitable to buy up the empty regional breweries (or build new capacity) and reintroduce the regional beer. But if demand for variety is not strong, then 1(c) could be an equilibrium here.

We shall argue below that during the first historical period under consideration, there was not an especially strong demand for horizontal product differentiation and economies of scale were strong; so as transport costs decline, the outcome was a tendency towards industrial and geographical concentration as shown in Figure 1(c). But during the second historical period under consideration, demand for horizontal product differentiation was growing and economies of scope are limited; so as transport costs decline, the outcome is a tendency towards geographical dispersion as shown in Figure 2(b).

5
Figure 2
Declining Transport Costs With Horizontal Product Differentiation
(With and Without Economies of Scope)

NW NE

SW SE
(a) High Transport Costs (b) Low Transport Costs (c) Low Transport Costs
but 1o Concentration and Concentration

We can summarise the possible outcomes of these Krugman type models as follows:

Economies of Scope

Strong Weak
Strength of Demand Strong 2(c) 2(b)
for Variety Weak 1(c) 1(c)

The outcome depends in an obvious way on the strength of demand for variety and the strength of economies of scope. We shall see in the following sections that this essential insight goes some way towards explaining the changing trends in the industry before and after 1970.
3. Historical Episode 1: Background

In this section, we list five factors in particular that are generally agreed to have shaped the trend towards geographic and industrial concentration in the English brewing industry over the period 1900-1970. In a short paper, this is of necessity a very brief and superficial account. For further details, reference is made to the sources cited here.

3.1 Economies of scale

There is quite widespread agreement that economies of scale are important throughout brewing. These economies of scale are to be found in production, purchasing, distribution, advertising and marketing. This view is supported by Gourvish and Wilson (1994, pp. 505-508), Hawkins and Pass (1979, pp. 60-78), Hornsey (2003, ch. 7, 9) and Vaizey (1960, pp. 83-87 and pp. 88-100).4

In his magnificent History of Beer and Brewing, Hornsey (2003) has described in great detail the origins of these economies of scale. Hornsey (2003, Chapter 7) describes developments in the science of brewing starting in the late 17th Century. He notes particularly important developments in the late 19th century, and describes the period 1860-1880 as the ‘Golden Years’ of brewing science. Hornsey argues that the growth of this scientific approach to brewing and the feasibility of constructing very large breweries played a key role in the emergence of economies of scale in brewing.

3.2 Declining Transport Costs

Large breweries are only viable if the output from the brewery can be transported cheaply over a wide area. That was difficult in an era when long distance transport depended on canals and horse-drawn vehicles. But it became much more realistic in the era of railways and even more so in the age of motor vehicles.
4 But not all share this view. By contrast, Protz (1979) argues that the much-vaunted economies of scale never arose and cites the higher prices charged for national beer brands than for local brands. A middle view would be that these economies of scale do indeed exist, but did not result in lower prices for consumers – partly because of the market power of the big six brewers and some x-inefficiency in those companies.

7
Vaizey (1960, ch. 7) notes that the cost of transporting raw materials to a brewery are not very significant. The really important source of transport costs is the retail delivery of beer. Gourvish and Wilson (1994, pp. 149-151) describe how the growth of the railways played a huge role in developing Burton-on-Trent as one of the main centres of the brewing industry in England: “Perhaps no other town and industry in Victorian Britain demonstrated better the benefit of the railways” (Gourvish and Wilson, 1994, p. 151).

Hornsey notes that before 1900, brewery amalgamations mainly involved small local firms within a few miles of each other and it was rare for any local monopolies to emerge. Thereafter, much more ambitious mergers took place: “with improved transport connections, such moves were now quite feasible” (Hornsey, 2004, p. 592). Hornsey (2003, pp. 549-556) also notes that up to 1939, brewers used a mix of road, rail and horse-draw carriages to transport their beers. After the end of petrol rationing in the 1950s, this situation changed rapidly and road transport soon dominated all aspects of retail delivery.

3.3 Greater Transportability of Beers

Even with economies of scale and declining transport costs, large breweries are only viable if the beer will travel well. It is fair to say that for a long time, traditional beers did not travel very well. The advent brewery-conditioned beers from the 1930s onwards made such transportation much more realistic.

Brewery-conditioning was a 20th Century innovation. Scientific developments in filtration, carbonisation, refrigeration and pasteurisation had revolutionised bottled beer manufacture (Hornsey, 2003, p. 595). The beer was biologically ‘dead’, which means that it was chemically stable. This means that it could be stored for much longer periods and that it would travel much better than ‘live’ beers. These developments were then applied to production of ‘keg’ beer – biologically ‘dead’ beer in steel barrels. The first experiments were by Watneys in the 1930s but the main impetus towards ‘keg’ beers started in 1946 – see section 3.5 below (Hornsey, 2003, pp. 670 ff.)

3.4 Consumer Tastes

Merger activity in the brewing industry and the consequent growth in industrial and geographical concentration lead to a steady decline in the number of brands (Hall, 1979, 1980). While not everyone was happy with that trend (see Section 5), many consumers were content with a limited range of national brands. Following customer reaction to mediocre and unpredictable beers in the 1940s (Hawkins and Pass, 1979, p. 54), consumer demand for quality and reliability was stronger than the demand for variety per se.5 Using the classification of consumer types described in Swann (2009, Chapter 15) it is arguable that many beer drinkers during this period behaved like ‘Galbraith’ consumers – that is, consumers whose tastes and buying behaviour are receptive to advertising and marketing efforts by the large brewers, and who are attracted to the ‘big name’ national brands (such as Watneys Red Barrel and Double
5 Advertisements for Ind Coope’s beer Long Life, which was very popular in the 1960s, stressed that, “it never varies!”

8
Diamond). And this consumer characteristic in turn reinforces the process of concentration: regional brewers could not compete in supplying ‘keg’ bitters and lagers because they could no afford the huge advertising campaigns required to create ‘big name’ national brands.

3.5 ‘History Matters’

We noted above that reaction to mediocre draft beer during and after World War 2 had helped to build the momentum behind national ‘keg’ brands. Far from real ale at its best, as we see it today, these beers were cloudy, full of sediment and unpleasant to drink. One historical ‘accident’ is said to have played an important role in helping to promote ‘brewery conditioned’ beer. Hornsey (2003, p. 671) describes how the brewer J.W. Green (later Flower’s Breweries) of Luton, Bedfordshire, started experimenting with ‘keg’ beer when locally stationed US airmen were very disenchanted with drinking murky draft beer. This ‘keg’ (later known as Flower’s ‘keg’) soon gained a fervent and widespread following, and by around 1960, two ‘big name’ national brands of ‘keg’ (Double Diamond and Watneys Red Barrel) had become “vogue drinks” (Hornsey, 2003, p. 672).

* * * * *

The result of these five factors was a steady growth of industrial and geographical concentration. The large numbers of mergers and acquisitions has been well documented – see Gourvish and Wilson (1994), Hall (1979, 1980), Hawkins and Pass (1979), Hornsey (2003), Mark (1985) and Millns (1998). Following merger or acquisition, it was common to find that the acquirer or new company would close all but one of the breweries and concentrate all production at one site (or a very small number of sites).
4. Historical Episode 1: Evidence on Brewery 1umbers and Dispersion

Figure 3 illustrates the trends in industrial and geographical concentration of breweries in England over the period 1990-1975. There are four snapshots in Figure 3, relating to the years 1900, 1925, 1950 and 1975. Figure 3 uses the Ordnance Survey National Grid to break the map of England into a grid of ‘locations’, each 1Km by 1Km. In these figures, there are four different marks:

Mark on Map Number of Breweries at
Location (1Km*1Km)
small circle 1
medium circle 2-4
medium square 5-8
large square 9 +

The map has also super-imposed county boundaries to aid interpretation. The declining numbers of breweries and growing geographical concentration over the period is readily apparent.
9

Figure 3
Geographical Dispersion of Breweries, 1900-1975

1900 1925

Source: Author’s calculations using data from Brewery History Society (2005)
10

Figure 3
Continued

1950 1975

Source: Author’s calculations using data from Brewery History Society (2005)
11

Figure 4
Active Breweries in England, 1900-2004

1500
Series pre-1950 is a Formation of CAMRA
in England 10 year moving average
1000
Active Breweries
500 Total
of Established
Number 1971 Onwards

Established
0 Before 1971

1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000

Source: Author’s calculations using data from Brewery History Society (2005)
Figure 5
Breweries and Locations in England, 1904-2004

1600 2.00
1400 1.75
Locations 1200 Breweries 1.50 Location

(LH Scale)

Numbers of Breweries and 1000 1.25 Number of Breweries per
Breweries
per Location
800 (RH Scale)
1.00
Locations
(LH Scale)
600 0.75
400 0.50

200 0.25
0 0.00
1904 1914 1924 1934 1944 1954 1964 1974 1984 1994 2004

Source: Author’s calculations using data from Brewery History Society (2005) and various Gazetteers
12
Figure 4 gives and even more stark indication of the fall in brewery numbers from 1900. Between 1990 and 1975, indeed, the number of breweries fell by nearly a factor of 10. The very different developments from 1971 onwards will be discussed in the next two sections, but it is worth examining the continuing downward movement in the number of ‘pre-1971 breweries’ in Figure 4. This falls from 167 in 1970 to 55 in 2004 (see also Figure 7 which is a detail from Figure 4).

Figure 5 gives another indication of the trend in brewery concentration. This places the trend in the number of breweries alongside the number of ‘locations’ (as described above) in which there is at least one active brewery. The number of active locations falls rapidly too (a factor of 5 or more) but not as rapidly as the number of breweries. The implication, also shown in Figure 5, is a steady decline in the average number of breweries per 1Km*1Km ‘location’.
5. Historical Episode 2: Background

Now we turn to the second historical episode – from 1971 onwards. We list three factors that are generally agreed to explain the resurgence of small breweries from 1971 onwards. As in Section 3, this is of necessity a very brief and superficial account, and for further details, reference is made to the sources cited here.

5.1 CAMRA

The most important element in this episode was the history of CAMRA, and the initiatives to which it led. CAMRA (The Campaign for Real Ale) was founded in 1971 as a consumer pressure group in reaction to what had happened in the English brewing industry from 1900-1970. The founders were concerned about the decline of variety and the blandness of modern beer. Some in CAMRA also objected to the highly concentrated industrial structure of the industry, over and above its implications for product variety and quality. Millns (1992) notes that the ideas behind CAMRA were not entirely new: for example, in the 1960s, an organisation called the Society for the Preservation of Beers from the Wood (SPBW) was formed with some similar objectives. But CAMRA was successful while its predecessors were not.

Protz and Millns (1992, p. 7) summarise the early objectives of CAMRA very succinctly:

“CAMRA created the climate for and awareness of the possibility of choice, and helped to hold out the promise that choice might be fulfilled. But no brewer would have responded …. if there had not been evidence of continued and increasing public demand. That demand was stimulated and publicised by CAMRA in its early stages of existence as a single-issue campaign dedicated to preserving real ale.”

The founders of CAMRA were reacting to the displacement of ‘real ale’ by ‘keg’ beer. ‘Real ale’ is matured by secondary fermentation in the container (cask) from which it is dispensed. Real ale is therefore a living product and is not pasteurised or filtered and no extraneous carbon dioxide is added. ‘keg beer’ is brewed using the same primary fermentation process as real ale, but after that it is filtered and/or
13
pasteurised. No further conditioning can therefore take place in the container (‘keg’) from which it is served, and therefore it is known as ‘brewery-conditioned’ beer. Carbon dioxide (and sometimes nitrogen) is added artificially.

The current membership of CAMRA is approximately 95,000 (CAMRA, n.d.), but its influence is much greater than that. Many people who are not members of CAMRA have learnt to become more discerning and more active consumers (as defined by Bianchi, 1998) who appreciate the qualities of real ale and appreciate regional and local variety.

We said before that, in terms of the consumer categories described in Swann (2009, ch. 15), the drinker of ‘keg’ beer is a ‘Galbraith’ consumer. By contrast, the ‘real ale’ drinker is a ‘Bourdieu’ consumer who marks his/her distinction by his/her consumption patterns. The ‘real ale’ drinker would demonstrate distinction by shunning the ‘big name’ keg beers. But the ‘real ale’ drinker is also a ‘Douglas’ consumer, whose consumption is influenced by loyalty to peers. Accordingly, the ‘real ale’ drinker will favour the ‘small name’ and local brews. In short, the demand is for variety and tradition rather than homogeneity and branding.

5.2 Micro-breweries

As the objectives of CAMRA are very different from the objectives of the large brewers, the success of the CAMRA initiative also depended on the availability of technologies for micro-breweries. Without that, there could not have been the large scale entry of small regionally dispersed micro-breweries dedicated to making real ale.

It is not clear that micro-breweries have had the equivalent technological change to the micro-computer. As such, the growth of micro-breweries in the 1970s cannot be attributed to technology push. It is essentially driven by a change in demand, building (in the UK) on the CAMRA initiative. Nonetheless, the technologies for micro-brewery are available and, judging from the number of books on ‘how to set up your own micro-brewery’,6 entry into this market is affordable even if survival is far from guaranteed.

5.3 Economies of Scope

Why do the big brewers not react to the influx of small entrants by product proliferation, as in other industries? That is, why do they not produce a wide variety of beers to match those produced by the micro-breweries? The main answer is that economies of scope in brewing are of limited importance – even if economies of scale are very important.

It is generally agreed that economies of scope in production are not very important. This is the view taken by Hawkins and Pass (1979, p. 116), Hornsey (2003, pp. 627-628) and Vaizey (1960, p. 84) in their histories of brewing. Some more recent statistical studies have come to a similar conclusion (Tremlay and Tremblay, 1996,

6 For example: Bruning (2009), Calagione (2005) and Daniels (2006).

14
2005; Tsiritakis, 1990). And, this is also the view of Berghoff et al (n.d.) who argue from their knowledge of the brewing process:

“large batch size often hinders the major brewers’ success in the craft brewed market because the production of small batches, necessary to produce the product variety and freshness demanded by craft-beer drinkers, does not fit profitably into scale production operations.”

One exception, which could lead to economies of scope in production, is where the large brewer chooses to mimic the ‘late configuration centre’ concept used in car plants. This is a way of deriving economies of scope from underlying economies of scale, by generating a large number of slightly different models by adding (or not adding) additional features at the end of the process. The drinker might prefer not to know that this process is sometimes used by the brewers (Hornsey, 2003, p. 628):

“For logistic reasons, the mega-brewers would have liked to have achieved an even greater reduction in the number of available brands but, by making use of a number of versatile post-fermentation techniques, such as the addition of isomerised hop extracts, and the use of colouring agents, they could manufacture a variety of end-products from a single brewing run. This practice, together with the use of highly imaginative packaging, enabled a number of low-volume beer brands to survive.”

It is argued that such economies of scope as there are in brewing are to be found in advertising, marketing and distribution. Cook (1997) in particular argues that economies of scope in distribution are very important. But even then, it is clear that most of the large brewers prefer to maintain a clear brand focus and for that reason limit the variety of beers they produce (Hall, 1979, 1980)

Craft brewers (the US term) and real ale brewers (UK term) obviously cannot achieve the same economies of scale in production as the big brewers. But since small batch production is the appropriate scale for real ales (craft beers), then small brewers don’t really suffer a competitive advantage vis a vis the large brewers, because production of small batches doesn’t really fit the business models (and large scale plant) of the large brewers.
6. Historical Episode 2: Evidence on Brewery 1umbers and Dispersion

Figure 6 is derived in a similar way to Figure 3. It shows the trends in entry of new breweries in England over the period 1970-2004 subdivided into the periods 1970-74, 1980-84, 1990-94 and 2000-04. The marks in Figure 6 are as below:

Mark on Map Number of breweries at
Location (1Km*1Km)
small circle 1 Incumbent
medium circle 2-4 Incumbents
medium square 5-8 Incumbents
large square 9 + Incumbents
triangle 1+ Entrant

Figure 6
Geographical Dispersion of 1ew Entrants, 1970-2004

1970-1974 1980-1984

Source: Author’s calculations using data from Brewery History Society (2005)
16

Figure 6
Continued

1990-1994 2000-2004

Source: Author’s calculations using data from Brewery History Society (2005)
17

Figure 7
Active Breweries in England, 1971-2004

200 600
Established Total Number Established from 1971 Onwards
Before 1971 (RH Scale) Total Number of Active Breweries
(LH Scale)
150 450
Established
1971 Onwards
100 (RH Scale) 300
50 150
NumberofPre1971BreweriesStillActive
0 0
1970 1975 1980 1985 1990 1995 2000

Source: Author’s calculations using data from Brewery History Society (2005)
Figure 8
Breweries Active Sometime During 1971-2004 by Type of Location
35%
30%
25%
20%
15%
10%
5%
0%
Small Village Medium Village Large Village/ Medium Town Large Town/ Med/Large City
Under 850 851-3,700 Small Town 16,201-60,000 Small City Over 187,001
3,701-16,200 60,001-187,000

Established
1971 or Later

Established

Before 1971
Source: Author’s Calculations using data from Brewery History Society (2005) and Office of National Statistics (n.d.)
18
The striking features of entry during periods 1980-84, 1990-94 and 2000-2004 are both the extent of entry and its geographical dispersion. The triangle marks occupy many new and unoccupied locations. Indeed, referring back to Figure 5, it is clear that the grow in numbers of breweries over this second period is matched by an equivalent growth in the number of occupied locations, indicating that these new breweries were occupying new space.

Figure 7, which is a detail from Figure 4 shows the cumulative entry of new breweries (formed since 1971) net of exits. This grows rapidly, albeit with a marked cyclical pattern.7 Referring back to Figure 4, it is clear that the total number of breweries in England in the early 2000s were back to level not seen since about 1940.

Figure 8 provides an interesting comparison of the sorts of locations in which the pre-CAMRA breweries and the post-CAMRA breweries are located. The horizontal axis of Figure 8 is calibrated in such a way as to ensure that the distribution of post-CAMRA breweries is distributed uniformly across those six size categories. Thus, one sixth of the post-CAMRA breweries are located in small villages, one sixth in medium villages, one sixth in large villages or small towns, one sixth in medium towns, one sixth in large towns or small cities, and the final sixth in medium or large cities. The distribution of pre-CAMRA breweries, by contrast, is heavily skewed towards towns and cities.

In short, an essential characteristic of the new entry has been its greater regional dispersion, with many of the micro-breweries located in villages or even on isolated farms. What lies behind this dispersion? There are several possible explanations and without further evidence it is hard to discriminate between them. One appealing explanation, however, is that the regional and local origins of a beer have become an important characteristic of the product in the eyes of the consumer – or, “place is what we drink” (Swann, 2006). The best-informed consumers may know the production heritage of each brand of beer, but the less well informed may judge production heritage by geographical origins. A beer from Burton-on-Trent is surely the product of one of the giant breweries. By contrast, a beer from a farm on the Lizard Peninsula in South Cornwall or Wasdale Head in the heart of the Lake District, must surely be the product of a micro-producer of real ale.
7. Conclusion

This paper has compared two episodes in the history of the brewing industry in England. During the first, 1900-1970, we see a rapid decline in the number of breweries and a much greater industrial and geographical concentration. This is consistent with Krugman’s model of declining transport costs geographical dispersion, where economies of scale are very important and there is little horizontal product differentiation and little demand for variety.

During the second, we see a rapid increase in the number of micro-breweries, many located in out-of-the-way places, producing a wide variety of real ales – even if they
7 The downturn from 1985-90, and the downturn around 2000 coincide with periods of sharply falling GDP growth in the UK.

19
only account for a relatively small share of the total beer market. This is consistent with a variant of the Krugman model in which economies of scope are weak when compared to economies of scale, and there is both a strong demand for and a strong supply of product variety.

The main features of the first episode seem to be:

⦁ economies of scale
⦁ declining transport costs
⦁ greater transportability of beers
⦁ demand for reliability rather than variety

⦁ ‘Galbraith’ consumers
⦁ growing industrial and geographical concentration
⦁ declining numbers of breweries in fewer locations

The main features of the second episode seem to be:

⦁ some more discerning and more active consumers
⦁ ‘Bourdieu’ consumers
⦁ appreciation of local variety

⦁ establishment of micro-breweries
⦁ limited economies of scope for big brewers
⦁ wide geographical dispersion
⦁ geographical location may take on a new meaning

Most of all, the contrast between these two episodes illustrates how the implications of declining transport costs on geographical dispersion of activity can vary enormously, with the outcome depending critically on the following:

⦁ economies of scale
⦁ economies of scope
⦁ horizontal product differentiation
⦁ demand for diversity

⦁ demand for reliability and predictability

In the first episode, we observe rapidly increasing geographical concentration because

(a) and (e) are the most important factors. In the second, we observe rapidly increasing geographical dispersion because (c) and (d) seem most important and because (b) is not important.
References

Berghoff S., C. Hackett, L. Hun, P. Liebwald and J. Reddinger (n.d.) Tapping into the Craft-Beer Industry, Guerneville, CA: Stumptown Brewery: www.stumptown.com/ articles/mgmtbeer.html

Bianchi M. (1998) The Active Consumer: 0ovelty and Surprise in Consumer Choice, London: Routledge
20
Bruning T. (2009) The Micro-brewers’ Handbook, 2nd Edition, Norwich: Paragraph
Publishing

Brewery History Society (2005) A Century of British Brewers 1890-2004, Longfield, Kent: Brewery History Society

Calagione S. (2005) Brewing Up a Business: Adventures in Entrepreneurship from the Founder of Dogfish Head Craft Brewery, Hoboken, NJ: John Wiley & Sons

CAMRA (n.d.) About Us, http://www.camra.org.uk/page.aspx?o=about

Cook G.C. (1997) “A comparative analysis of vertical integration in the UK brewing and petrol industries”, Journal of Economic Studies, 24, 152 – 166

Daniels R. (2006) Brewers Association’s Guide to Starting Your Own Brewery, Boulder, CO: Brewers Publications

Gourvish T.R. and R.G. Wilson (1994) The British Brewing Industry 1830-1980, Cambridge: Cambridge University Press

Hall G.C. (1980) “Mergers and the Product Mix: A Case Study of the Brewing Industry”, in K. Cowling, P. Stoneman, J. Cubbin, J. Cable, G. Hall, S. Domberger and P. Dutton, Mergers and Economic Performance, Cambridge: Cambridge University Press, pp. 214-237

Hall G.C. (1979) “The effects of mergers on the product–mix offered by the brewing industry”, Applied Economics, 11, 21–34

Hawkins, K. H. and Pass, C. L. (1979) The Brewing Industry, A Study in Industrial Organisation and Public Policy, London: Heinemann Educational

Hornsey I.S. (2003) A History of Beer and Brewing, Cambridge: The Royal Society of Chemistry

Jay A. (1996) The Oxford Dictionary of Political Quotations, Oxford: Oxford University Press

Krugman P.R. (1980) “Scale Economies, Product Differentiation, and the Pattern of Trade”, American Economic Review, 70, 950-959

Krugman P.R. (1991) “Increasing Returns and Economic Geography”, Journal of Political Economy, 99, 483-499

Krugman P.R. (1992) Geography and Trade, Cambridge, Mass: MIT Press

Mark J. (1985) “Changes in the British Brewing Industry in the 20th Century”, in D.J. Oddy and D.S. Miller (eds.) Diet and Health in Modern Britain, London: Croom Helm
21
Millns T. (1992) “CAMRA: The History”, in R. Protz and T. Millns (eds.) Called to the Bar: An Account of the First 21 Years of the Campaign for Real Ale, St Albans: CAMRA Ltd

Millns T. (1998) “The British Brewing Industry, 1945-1995”, in R. G. Wilson and T. R. Gourvish (eds.) The Dynamics of the International Brewing Industry since 1800, London: Routledge

Office of National Statistics (n.d.) 0eighbourhood Statistics, http://neighbourhood. statistics.gov.uk/dissemination/

Protz R. (1978) Pulling a Fast One, What the Brewers Have Done to Your Beer, London: Pluto Press

Protz R and T. Millns (1992) “Foreword”, in R. Protz and T. Millns (eds.) Called to the Bar: An Account of the First 21 Years of the Campaign for Real Ale, St Albans: CAMRA Ltd

Swann G.M.P (2006) “Place is What We Think With: Spatial History, Intellectual Capital and Competitive Distinction”, in G. Vertova (ed.) The Changing Economic Geography of Globalisation, London: Routledge

Swann G.M.P. (2009) The Economics of Innovation: An Introduction, Cheltenham: Edward Elgar Publishing

Tremblay C.H. and V.J. Tremblay (1996) “Firm Success, National Status, and Product Line Diversification: An Empirical Examination”, Review of Industrial Organization, 11, 771-789

Tremblay V.J. and C.H. Tremblay (2005) The U.S. Brewing Industry: Data and Economic Analysis, Cambridge, Mass: MIT Press

Tsiritakis M., C. Campbell and A. Merikas (1990) “The cost of scale inefficiency in the Canadian brewing industry: a multi product cost function approach”, The Journal of Applied Business Research, 7, 16-24

Vaizey J. (1960) The Brewing Industry 1886-1951: An Economic Study, London:
Pitman
22