TY - JOUR
AU1 - Higgins, David M
AU2 - Varian, Brian D
AB - Abstract Before 1932, Britain’s essentially free-trade policy left barely any scope for reciprocating the preferential tariffs that the Dominions applied to Britain’s exports. Thus, Britain attempted to reciprocate by means of a “soft” trade policy aimed at increasing Britain’s imports from the empire through wide-reaching publicity coordinated by the Empire Marketing Board (EMB). This article, the first econometric assessment of the EMB, argues that there was not a differential increase in the volume of those imports advertised by the EMB. Principal arguments for this failure are that British consumers were frequently unaware of the geographic origin of many commodities and that they tended to identify company brand more than country of origin. 1. Introduction The Empire Marketing Board (EMB) was a relatively transient feature of the interwar British economy, operating from just 1926–33. It was conceived as a solution to an impasse in the political economy of the British empire. In the 1920s, the tariff-autonomous Dominions and colonies extended preferential tariffs to imports from Britain (MacDougall and Hutt 1954, p. 237). Yet, apart from a few deviations, Britain clung to its longstanding policy of free trade, which precluded any reciprocation in the form of preferential tariffs. The EMB was devised as a vehicle for Britain to reciprocate, by means of a non-tariff preference, the tariff preferences that its exports enjoyed in Dominion markets (Drummond 1974, p. 29). The EMB was, effectively, a government department created to encourage British consumers to purchase more empire produce via an advertising campaign in which posters featured prominently.1 This “soft” trade policy of appealing to the consumer as a means of raising Britain’s imports from the empire was ambitious indeed. An early study concluded that efforts to create non-tariff preferences for empire goods were “pathetic” and that the EMB was no more than a matter of expediency: “[it] satisfied various desires … anything in a political environment which precluded general protective tariffs and preferences” (Drummond 1974, p. 67). The EMB operated what was, as best as can be determined, the first state-sponsored campaign advertising the symbiotic relationship between Britain and its empire. Its activities represented a form of economic nationalism directly related to the controversy of whether “openness and free exchange or support and shelter better assist a nation’s development” (Trentmann 2008, p. vii). Despite the briefness of its existence, the EMB offers a rare opportunity to assess whether publicity can effect some increase in the volume of imports from advertised sources. This article provides the first econometric assessment of whether the EMB actually achieved its objective of increasing Britain’s “empire imports,” i.e., Britain’s imports from the empire. The econometric analysis in this paper relies upon a sample of the EMB’s posters from 1927, the first full year in which the EMB existed. The econometric analysis is supplemented by qualitative evidence on consumption patterns in Britain. This article’s analysis of the EMB has similarities with other studies that have examined the extent to which country-of-origin or “Made in” campaigns determined domestic consumers’ willingness to buy domestic products. Such non-tariff barriers were deployed in India during the 1930s, the United States in the 1930s and 1980s, and Australia and New Zealand in 1986 and 1988, respectively. Often, these campaigns were motivated by concerns about global depressions, growing import penetration, and their effects on economic growth and unemployment, though “Buy Indian” was also mobilized to foster ethnic trading links and to promote economic independence (Ettenson et al. 1988; Fischer and Byron 1997; Frank 1999; Fenwick and Wright 2000; Thackeray 2017; Thackeray 2019). This article, however, differs from the above studies insofar as it assesses the extent to which the EMB increased British sales of products from the many parts of the British empire. This article proceeds as follows. The next section provides an overview of trade policy within the British empire during the interwar era. Section 3 details the origins and operation of the EMB. Section 4 tests, econometrically, whether specific posters, each of which advertised a particular Dominion and commodity, actually increased the volume of Britain’s imports of the advertised commodity from the advertised Dominion.2 Section 5 offers archival explanations for the econometric results. This section re-appraises the magnitude of the EMB’s publicity expenditure. Additionally, by means of three case studies, this section demonstrates the weaknesses in legislation designed to enhance the visibility of country- and commodity-specific produce (New Zealand butter and Ceylon tea), and, in the case of Australian beef, it argues that the EMB’s promotion of this product conflicted with a change in consumer tastes. Section 6 presents conclusions. 2. “Hard” trade policy in interwar Britain The literature on interwar Britain’s trade policy toward the empire has concentrated almost exclusively on “hard” trade policy—the tariffs and quotas that Britain imposed upon imports. Among Britain’s hard trade policies toward the empire, the Ottawa agreements of 1932 were of singular and defining significance. These bilateral agreements between Britain and seven of its tariff-autonomous Dominions, viz.: Australia, Canada, Newfoundland, New Zealand, South Africa, India, and Southern Rhodesia, entailed a variety of reciprocal preferences (McGuire 1939, pp. 256–78; Glickman 1947, pp. 443–6; Russell 1947, pp. 32–4; Drummond 1974, pp. 219–99).3 In general, under the Ottawa agreements, the Dominions increased the pre-existing margins of preference accorded to many commodities imported from Britain.4 Additionally, Britain reciprocated the widened margins of preference mainly by agreeing to a permanent exemption of imports (from signatory trade partners) from the duties assessable under the Import Duties Act of 1932; the application of these duties to empire imports had been postponed pending the outcome of the Ottawa Conference later that year.5 The exemptions to the Import Duties Act, enshrined on a permanent basis in the Ottawa agreements, conferred a sizeable margin of preference upon Britain’s imports from the seven signatory trade partners. MacDougall and Hutt (1954 p. 237) estimated that the average margin of preference applied to Britain’s empire imports increased from 2–3 percentage points in 1929 to 10–12 percentage points in 1937.6 Studies have found that the Ottawa agreements enhanced trade within the British empire as a whole (Eichengreen and Irwin 1995, pp. 15–16; Jacks and Novy 2020, p. 131) and, specifically, between Britain and its empire (Gowa and Hicks 2013, p. 453; de Bromhead et al. 2019a).7 Of particular note is the estimate that the tariffs and quotas—quotas were a new and influential phenomenon in British trade policy in the 1930s—concluded at the Ottawa Conference accounted for three-fifths of the increase in the empire share of Britain’s imports, which rose from 27 percent in 1930 to 39 percent in 1935 (de Bromhead et al. 2019a, p. 347). Given the consequentiality of the Ottawa agreements, there is the temptation to interpret Britain’s trade policy toward the empire in binary terms: before and after Ottawa. While the literature emphasizes the substantial preferences that Britain extended to empire imports in 1932, it would be mistaken to think that Britain’s trade policy toward the empire was entirely static before then. Between the First World War and 1932, Britain’s hard trade policy began to embrace a small degree of preference for certain empire imports. Britain had long imposed tariffs on several agricultural commodities, such as tea, sugar, and tobacco, for revenue purposes. The Finance Act of 1919 implemented a one-sixth reduction of the applicable duty to empire imports of these commodities (McGuire 1939, p. 258). The Act did raise the empire share of certain commodities, notably sugar (McGuire 1939, p. 259). However, these were isolated cases. The produce of temperate agriculture remained almost completely outside the reach of the British tariff system, which, therefore, offered limited scope for the application of imperial preference. Not all of Britain’s tariffs in the 1920s were levied for the purpose of revenue. Several manufacturing industries came under the aegis of tariffs through a variety of legislation, including the Finance Act of 1915, which imposed the “McKenna duties” on motorcars and a few other manufactured goods, as well as the Safeguarding of Industries Acts of 1921 and 1925.8 The Finance Act of 1919 reduced the McKenna duties by one-third for empire imports (McGuire 1939, p. 258). Subsequent protective legislation enacted in the 1920s called, at least notionally, for a preferential reduction of the duties imposed upon empire imports. These preferences were symbolic gestures on the part of Britain, since hardly any manufactures were imported from the empire. Motorcars were the only manufactured article of a non-negligible value imported from the empire (Canada). The preferential reduction of the McKenna duty on motorcars from 33.3 percent to 22.2 percent ad valorem in 1919 induced American motorcar manufacturers to establish operations in Canada in order to gain preferential access to the British market (McGuire 1939, p. 268).9 Britain’s awkward integration of manufacturing protection and imperial preference, though of hardly any economic consequence, was nonetheless an attempt to remedy the unidirectional system of preference that prevailed in the British empire during the 1920s. In the two decades before the First World War, Canada (1897), New Zealand (1903), South Africa (1903), and Australia (1907) each implemented policies of imperial preference that applied either to Britain exclusively or to the British empire as a whole.10 These unreciprocated preferences loomed large within the political economy of the British empire. As Rooth (1993, p. 73) has documented, competition between the Dominions and other primary-producing countries was intensifying in the British market during the mid-1920s. The unidirectional character of imperial preference was becoming tenuous. Unless and until Britain abandoned its adherence to free trade, there was no meaningful “hard” trade policy modus for reciprocating the Dominions’ preferences. Hence, in the mid-1920s, Britain resorted to “soft” trade policy by establishing the EMB. 3. The origins and operation of the Empire Marketing Board The EMB originated from the Imperial Economic Conference of 1923. This conference established the Imperial Economic Committee (IEC), to investigate and advise on economic and commercial issues referred to it by any of the constituent governments. The terms of reference governing the IEC stated that it should … consider the possibility of improving the methods of preparing for market and marketing within the UK the food products of the overseas parts of the Empire with a view to increasing the consumption of such products in the UK in preference to imports from foreign11 countries, and to promote the interests of both producers and consumers (IEC, First Report, 1925, p. 4). These objectives, fundamental to the success of a voluntary preference scheme, were to be achieved in three ways. First, a sustained national publicity campaign was needed to induce consumers to buy empire produce. Such publicity was to be directed to a “broad educational effort to induce empire buying … [of] selected commodities” (IEC, First Report, 1925, p. 13). Second, legislative changes governing country-of-origin labelling would enable consumers to recognize the origins of foodstuff imports, thereby facilitating the exercise of voluntary and latent preferences for empire produce. Finally, the IEC envisaged appropriate research to increase the empire’s ability to supply regular quantities of high-quality foodstuffs. These objectives were interrelated. For example, origin marking might backfire if empire produce was perceived to be inferior to foreign produce. Similarly, any disposition that consumers had toward the empire would dissipate if supplies were erratic (IEC, First Report, 1925, pp. 6–9 and 13; IEC, Eight Report, 1928, pp. 14–16). The above policies reflected a growing realization that the economic fortunes of Britain and its overseas empire were symbiotic. Britain was unquestionably the principal market for empire produce. The IEC estimated that, in 1923, 45.7 percent of Britain’s total imports of foodstuffs and drink originated from the empire. This average conceals the fact that, for certain foodstuffs such as lamb/mutton and cheese, the empire’s share was much higher, at 65.4 and 85.6 percent, respectively. In turn, empire countries were important markets for Britain; in 1923, the empire absorbed 44.9 percent of Britain’s manufactured exports (IEC, Eighth Report, 1928, pp. 28–30). However, the IEC was a purely advisory body. To implement its proposals, the IEC recommended the formation of an executive commission, subsequently the EMB, which was to be funded by the British government. The history of the EMB is well documented. Its formation represented a political compromise in which non-tariff preference was used to promote imperial economic development (Constantine 1986a, pp. 197–8). Constitutionally, the EMB differed from the IEC. The latter was an imperial body empowered by the Imperial Economic Conference, constituted of members appointed by all empire governments and reportable to them. The EMB, in contrast, was appointed by the Secretary of State for Dominion Affairs and funded by a grant, nominally £1 million per annum, which had to be approved by the British government. This amount was roughly equivalent to the value of preferences on Britain’s empire imports that had been considered (but ultimately rejected), following the general election of 1923 and Cabinet debates on the merit of tariff reform. Nonetheless, there was some overlap between these organizations. For example, the EMB was responsible for implementing the recommendations of the IEC (IEC, Eighth Report, 1928, pp. 11–12). In broad terms, the EMB’s activities can be categorized as follows: making grants for research into marketing and production; the initiation and development of economic intelligence; encouraging greater production of agricultural goods and better preparation of those goods for market;12 and, central to this article, launching publicity and education campaigns. The EMB appears to have interpreted its publicity remit in general terms: The Board’s publicity campaign, though wide in range, is limited to the preparation of a background against which more specific efforts to press the sale of Empire produce, whether undertaken by the Board itself, or by other agencies, can be thrown into relief (EMB, Work of the Board, 1927, p. 31). Table 1 shows that, at least in the early years of its operation, spending on publicity was the most important component of the EMB’s total expenditure. However, following the onset of the general depression in 1929 and a review of the EMB, the proportion of funds devoted to publicity declined significantly. Increasingly, more emphasis was placed on improving the supply-side response of empire producers. For example, grants were made to the Departments of Agriculture in India and Mauritius to conduct research on tea and sugar cane, respectively (EMB, Work of the Board, 1930, p. 166). Particular importance was placed on low-temperature research, and representatives of the Australian Commonwealth Council for Scientific and Industrial Research participated in a scientific survey of the condition of frozen meat as it was transported from the Antipodes to London (EMB, Work of the Board, 1927, p. 26; EMB, Work of the Board, 1930, p. 167; EMB, Work of the Board, 1932, p. 772). Table 1 Expenditure by the EMB on publicity, posters, and research, 1926–32 . (1) . (2) . (3) . (4) . (5) . (6) . (7) . Total expenditurea (£) Publicity expenditureb (£) Poster expenditurec (£) Research expenditure (£) Share of publicity in total expenditure (percent) Share of posters in publicity expenditure (percent) Share of research in total expenditure (percent) 1926 120,551 89,843 23,597 30,708 74.5 26.3 25.5 1927 359,405 238,126 103,885 121,279 66.3 43.6 33.7 1928 510,033 278,414 92,714 231,619 54.6 33.3 45.4 1929 599,099 222,361 64,873 376,738 37.1 29.2 62.8 1930 598,017 199,411 60,420 398,606 33.3 30.3 66.6 1931d 277,306 80,997 34,916 196,309 29.2 43.1 70.8 1932e 359,440 62,820 23,820 296,620 17.5 37.9 82.5 1926–32 2,823851 1,171,972 404,225 1,651,879 41.5 34.5 58.5 . (1) . (2) . (3) . (4) . (5) . (6) . (7) . Total expenditurea (£) Publicity expenditureb (£) Poster expenditurec (£) Research expenditure (£) Share of publicity in total expenditure (percent) Share of posters in publicity expenditure (percent) Share of research in total expenditure (percent) 1926 120,551 89,843 23,597 30,708 74.5 26.3 25.5 1927 359,405 238,126 103,885 121,279 66.3 43.6 33.7 1928 510,033 278,414 92,714 231,619 54.6 33.3 45.4 1929 599,099 222,361 64,873 376,738 37.1 29.2 62.8 1930 598,017 199,411 60,420 398,606 33.3 30.3 66.6 1931d 277,306 80,997 34,916 196,309 29.2 43.1 70.8 1932e 359,440 62,820 23,820 296,620 17.5 37.9 82.5 1926–32 2,823851 1,171,972 404,225 1,651,879 41.5 34.5 58.5 Notes: a) Total expenditure refers to the sum of cols. 2 and 4. b) Publicity expenditure includes exhibitions, shopping weeks, posters, press advertising, cinema advertising, and lectures. c) Poster expenditure includes the construction, erection, and maintenance of frames; site rentals; artists’ fees; and the printing, posting, and distribution of posters. d) Data for 1931 were brought to account on 31 March 1932. e) Data for 1932 are based on official estimates. Sources: EMB, Work of the Board (1932, pp. 6–10); House of Commons, Select Committee on Estimates (1932, p. 170, Appendix 1). Open in new tab Table 1 Expenditure by the EMB on publicity, posters, and research, 1926–32 . (1) . (2) . (3) . (4) . (5) . (6) . (7) . Total expenditurea (£) Publicity expenditureb (£) Poster expenditurec (£) Research expenditure (£) Share of publicity in total expenditure (percent) Share of posters in publicity expenditure (percent) Share of research in total expenditure (percent) 1926 120,551 89,843 23,597 30,708 74.5 26.3 25.5 1927 359,405 238,126 103,885 121,279 66.3 43.6 33.7 1928 510,033 278,414 92,714 231,619 54.6 33.3 45.4 1929 599,099 222,361 64,873 376,738 37.1 29.2 62.8 1930 598,017 199,411 60,420 398,606 33.3 30.3 66.6 1931d 277,306 80,997 34,916 196,309 29.2 43.1 70.8 1932e 359,440 62,820 23,820 296,620 17.5 37.9 82.5 1926–32 2,823851 1,171,972 404,225 1,651,879 41.5 34.5 58.5 . (1) . (2) . (3) . (4) . (5) . (6) . (7) . Total expenditurea (£) Publicity expenditureb (£) Poster expenditurec (£) Research expenditure (£) Share of publicity in total expenditure (percent) Share of posters in publicity expenditure (percent) Share of research in total expenditure (percent) 1926 120,551 89,843 23,597 30,708 74.5 26.3 25.5 1927 359,405 238,126 103,885 121,279 66.3 43.6 33.7 1928 510,033 278,414 92,714 231,619 54.6 33.3 45.4 1929 599,099 222,361 64,873 376,738 37.1 29.2 62.8 1930 598,017 199,411 60,420 398,606 33.3 30.3 66.6 1931d 277,306 80,997 34,916 196,309 29.2 43.1 70.8 1932e 359,440 62,820 23,820 296,620 17.5 37.9 82.5 1926–32 2,823851 1,171,972 404,225 1,651,879 41.5 34.5 58.5 Notes: a) Total expenditure refers to the sum of cols. 2 and 4. b) Publicity expenditure includes exhibitions, shopping weeks, posters, press advertising, cinema advertising, and lectures. c) Poster expenditure includes the construction, erection, and maintenance of frames; site rentals; artists’ fees; and the printing, posting, and distribution of posters. d) Data for 1931 were brought to account on 31 March 1932. e) Data for 1932 are based on official estimates. Sources: EMB, Work of the Board (1932, pp. 6–10); House of Commons, Select Committee on Estimates (1932, p. 170, Appendix 1). Open in new tab Table 1 also indicates that posters were a key element of the EMB’s publicity, accounting for, on average, 35 percent of publicity expenditure between 1926 and 1932. The EMB commissioned some of the best-known poster artists to produce “frame posters” that were displayed at dedicated sites in municipalities with populations greater than 100,000 (EMB, Work of the Board, 1927, p. 9). This activity was complemented by exhibitions and fairs, cinema and newspaper advertisements, and a program of lectures. The EMB reported that 22,000 schools and other organizations were on its standing list for publicity and communications, and that 11,000 copies of “contract posters” had been issued to factories where orders were being executed for the overseas empire (EMB, Work of the Board, 1930, p. 16). However, the cost of these factory contract posters was subsumed within the EMB’s total expenditure on posters, which was substantially directed toward consumers; in any event, EMB reports do not indicate which, or how many, factories directly benefitted from its poster campaign. Staff were employed to encourage local authorities to increase their purchase of empire produce. A prominent feature of the EMB’s publicity campaign were empire shopping weeks and the leasing of empire shops dedicated to the sale of empire produce. Shopping weeks were launched throughout Britain, and, for periods of between three and six months, specially commissioned shops, replete with empire imagery and the EMB’s posters, operated in Birmingham, Blackpool, Glasgow, and Liverpool (EMB, Work of the Board, various years).13 In addition to more common foodstuffs such as New Zealand butter and Ceylon tea, the EMB promoted a range of other empire produce including sago (from British Borneo), pineapples (Malaya), timber (Canada), oranges (South Africa), as well as cloves, brandy, rum, and tobacco from Zanzibar, Cyprus, Jamaica, and Nyasaland, respectively (Constantine 1986b, p. 5; plates 41, 43, 51, 55, 56A). However, to date, there has been no systematic attempt to assess the effectiveness of the EMB’s publicity activities. The contemporary evidence is limited and fragmentary. A Select Committee reported that the Board’s non-specific publicity campaign made it “impossible to obtain any direct evidence on which to base a reasoned opinion of the measure of success attending the Board’s efforts, or of the relative value of the various methods of propaganda adopted” (House of Commons, Committee on Estimates, 1928, p. 197).14 Investigations by the EMB into domestic food preferences revealed pronounced regional differences: “What London likes may prove to be what Lancashire refuses” (EMB, Butter, 1930, p. 7). Consequently, exhortations to buy empire produce were unlikely to obtain approval throughout the country. For example, Danish butter predominated in Manchester, Yorkshire, and Scotland, whereas New Zealand butter was much preferred in the South. Similarly, comparing 1909–13 with 1928, the EMB contrasted the sharp growth of Britain’s cheese imports from New Zealand with declining Canadian supplies (EMB, Cheese, 1929, p. 7; EMB, Butter, 1930, pp. 15 and 19). One purported success was an increase in the sale of Australian and New Zealand butter in Lancashire in 1931, but even Stephen Tallents, Secretary of the EMB, admitted that this increase was due to a fall in the price of empire butter, as well as the deployment of canvassers by the EMB (House of Commons, Committee on Estimates, 1932, p. 29, QQ.295–6). Questioned in Parliament whether the EMB had actually realized its aim of raising the consumption of empire produce, Tallents responded, “On the whole we have the impression that this feeling for buying Empire goods, both Home and Overseas goods, is more in the air than when we started our work” (House of Commons, Committee on Estimates, 1928, p. 120, Q.1311). The scale of the task confronting the EMB should not be underestimated. Constantine noted that not only was Britain at the center of the British empire but it was also the focus of international trade in produce because of its traditionally heavy reliance on imports (Constantine 1986b, p. 3). The predominance of foreign supplies in certain British foodstuff imports is revealed in table 2, which indicates that foreign imports of fruit and beef/veal substantially exceeded empire imports, whereas the opposite applied for cheese. For butter and lamb/mutton, the importance of foreign and empire supplies was almost equal. Finally, within Britain, pronounced variations in wealth undermined the general appeal of the EMB’s publicity campaign; its influence was largely restricted to middle-class consumers (Constantine 1986a, p. 223; Higgins 2018, pp. 100–1; Thackeray 2017, p. 405). Table 2 British imports of selected foodstuffs from empire and foreign sources, 1924 . Empire imports (thousand tonnes) . Foreign imports (thousand tonnes) . Total imports (thousand tonnes) . Empire share of imports (percent) . Foreign share of imports (percent) . Butter 2,360 2,933 5,293 44.6 55.4 Cheese 2,512 377 2,889 87.0 13.0 Fruit 2,724 5,883 8,607 31.6 68.4 Beef and veal 164 586 750 21.9 78.1 Mutton and lamb 151 119 270 56.0 44.0 . Empire imports (thousand tonnes) . Foreign imports (thousand tonnes) . Total imports (thousand tonnes) . Empire share of imports (percent) . Foreign share of imports (percent) . Butter 2,360 2,933 5,293 44.6 55.4 Cheese 2,512 377 2,889 87.0 13.0 Fruit 2,724 5,883 8,607 31.6 68.4 Beef and veal 164 586 750 21.9 78.1 Mutton and lamb 151 119 270 56.0 44.0 Note: Foreign refers to sources outside of the British empire. Source: IEC, First Report (1925, p. 29, table 2). Open in new tab Table 2 British imports of selected foodstuffs from empire and foreign sources, 1924 . Empire imports (thousand tonnes) . Foreign imports (thousand tonnes) . Total imports (thousand tonnes) . Empire share of imports (percent) . Foreign share of imports (percent) . Butter 2,360 2,933 5,293 44.6 55.4 Cheese 2,512 377 2,889 87.0 13.0 Fruit 2,724 5,883 8,607 31.6 68.4 Beef and veal 164 586 750 21.9 78.1 Mutton and lamb 151 119 270 56.0 44.0 . Empire imports (thousand tonnes) . Foreign imports (thousand tonnes) . Total imports (thousand tonnes) . Empire share of imports (percent) . Foreign share of imports (percent) . Butter 2,360 2,933 5,293 44.6 55.4 Cheese 2,512 377 2,889 87.0 13.0 Fruit 2,724 5,883 8,607 31.6 68.4 Beef and veal 164 586 750 21.9 78.1 Mutton and lamb 151 119 270 56.0 44.0 Note: Foreign refers to sources outside of the British empire. Source: IEC, First Report (1925, p. 29, table 2). Open in new tab Nonetheless, some scholars have viewed the impact of the EMB more positively. It has been claimed that the formation of the EMB was not a cynical political ploy and that its poster and advertising campaigns were important in reducing British consumers’ bias against imperial products (Moore 2016, pp. 267–76). Schwarzkopf (2012, p. 172) argued that the EMB was “one of the most important movements in the management of public opinion and mass communication.” Freeman (1977, p. 14) asserted that the EMB’s publicity schemes were “a success, the campaigns having effected a significantly increased demand for Empire goods.” This more favorable reassessment of the EMB originates from studies demonstrating a coalescence of cultural and social factors in the 1920s that created a fertile environment for the EMB’s publicity. The growth of “imperial” and “ethical” consumer movements—for example, the British Women’s Patriotic League—reflected the importance of the consumer (usually the housewife) as a positive force promoting imperial unity (Trentmann 2008, pp. 228–40). This background enhanced the appeal of the EMB’s publicity campaign by reducing psychological distance between Britain and its empire. The EMB’s posters created a Dominion discourse in which the Dominions were represented as an extension of Britain and “Britishness” (Barnes 2014; Barnes and Higgins 2020). Complementing these forces was a range of economic-industrial organizations, for example, the Empire Development Union (1922), the Empire Industries Association (1924), and the Empire Economic Union (1929), whose members were concerned about the collapse of the pre-1914 trading environment and sought to make the British empire more self-sufficient (Constantine 1986a, pp. 193–5). Did the EMB achieve its fundamental objective of increasing the sale of empire produce in the British market? The failure of the existing literature to address this question directly is partly explained by the absence of quantitative data on consumers’ attitudes to the EMB, as well as the EMB’s failure to conduct systematic market research (Kothari 2014, p. 45). The EMB’s claims of “success” were based on impressionistic and unsolicited testimonials from consumer bodies and trade organizations (Constantine 1986b, p. 16). This article seeks to determine the success of the EMB by using novel data on the content and timing of the EMB’s posters to test whether those posters did affect the volume of Britain’s empire imports. 4. Assessing the economic effect of the Empire Marketing Board This section aims to determine whether, in fact, the EMB was effective in increasing Britain’s empire imports. The analysis in this section makes use of novel data on the timing of the release of the EMB’s “frame posters,” which were its most iconic mode of publicity, and which offer an identification strategy for assessing the effect (if any) of the EMB. In particular, the analysis makes use of data on those early frame posters that made explicit reference to both a particular commodity and Dominion of origin. Consistent with the trade literature, this section uses the terminology of commodities and varieties. A variety is a bilaterally defined commodity, e.g., New Zealand butter, whereas a commodity is undefined by place of origin, e.g., butter. Did the release of the EMB’s variety-specific posters increase the volume of Britain’s imports of the advertised variety? This section, representing the first econometric assessment of the effectiveness of the EMB, addresses this question in two ways: a conventional regression analysis with a single binary variable for posters and an event study. Before proceeding to these two subsections, a discussion is given to three crucial variables used in the econometric analysis: posters, import volumes, and prices. 4.1. Data As already discussed, posters were a large component of the EMB’s activity, accounting for more than one-third of total publicity expenditure during its existence. The typical poster was the frame poster. Measuring 60 × 40 inches, frame posters were displayed for several weeks in purposely constructed frames in, initially, 700 locations throughout Britain (The National Archives, CO 760/26).15 The EMB began displaying these frame posters in January/February 1927.16 Of the frame posters displayed in the first full year of the EMB’s operation, 1927, only a small number of posters made reference to a specific Dominion variety. The remaining frame posters depicted either an empire region sans commodity, commodity sans region, or a more generic evocation of empire, such as the fanciful Empire Christmas poster of December 1927. There is the possibility that these other frame posters, alongside the many other publicity initiatives of the EMB, exerted a positive effect on Britain’s empire imports. Nonetheless, the analysis focuses on just the variety-specific posters of the EMB; if indeed the EMB enhanced Britain’s consumption of empire products, then such an effect is most likely to be observed from the most defined forms of publicity pursued by the EMB, such as variety-specific posters. The analysis is further restricted to those seven variety-specific posters for which matching trade data are available: New Zealand (dairy), Irish Free State (dairy), Canada (wheat), Mauritius (sugar), India (rice), Ceylon (tea), and Australia (cattle).17 For the purpose of matching these posters to corresponding variety-specific data on import volumes and prices, it is necessary to identify a representative commodity for several of these posters, since their pictorial content pertains to a slightly broader category of commodities. Butter is taken as representative of dairy, grain as representative of wheat, and frozen beef—chilled beef was not imported from Australia on more than an experimental scale during this period—as representative of cattle. Cheese is taken as an additional representative commodity for the New Zealand dairy poster.18 The decision to focus on only those variety-specific posters released in the year 1927 is informed by two concerns. First, the timing of the release of the 1927 posters can be identified with sufficient precision. Although the EMB coded its posters sequentially, there is no comprehensive record of when every individual poster was displayed. Yet, the minutes of the Publicity Committee and Poster Sub-Committee are comparatively detailed during the early period of the EMB’s existence and yield evidence of when, during 1927, each of the seven variety-specific posters was released. Particularly informative is “Posters for Display in the Board’s Special Frames: Approximate Programme for 1927” in the minutes of the Publicity Committee, which states the months when the posters were scheduled to be displayed (The National Archives, CO 760/22). The second, equally important concern motivating the decision to focus on only the 1927 variety-specific posters is the potential for diminishing returns (in terms of raising empire imports) of the EMB’s poster campaign. Even if the initial posters of the EMB were successful in shifting consumer expenditure from foreign to empire imports, additional posters might have produced a lesser, if any, effect on Britain’s empire imports. Here, it is worth mentioning that some varieties featured in multiple frame posters throughout the brief existence of the EMB. For example, Canadian wheat farming was featured in a frame poster displayed in January/February 1927, while a Canadian grain elevator was featured in another frame poster displayed sometime in early 1928. In testing for an effect of the EMB’s variety-specific posters on the import volume of the advertised variety, a concern would be that the EMB’s selection of which varieties to advertise (and when) was endogenously determined. The EMB never adopted a formal procedure for selecting which Dominions, commodities, or (Dominion-commodity) varieties to advertise. However, the minutes of the Publicity Committee reveal that certain guiding principles were followed in devising the program of frame posters. In a document titled “Note on Subjects suggested for Posters,” the Publicity Committee indicated that the first tranche of posters, which was to extend beyond 1927, would focus exclusively on foodstuffs, rather than the non-foodstuff produce of the Empire, such as Canadian timber (The National Archives, CO 760/22). Within foodstuffs, the following procedure was proposed: “The posters might be displayed in order of seniority of the Dominions or Colonies whose product is represented” (The National Archives, CO 760/22). It appears that the EMB adhered to this procedure, with Canada, the first Dominion (1867), being featured in the first set of variety-specific posters in January/February 1927. While it cannot be proven definitively that the selection and timing of advertised varieties was exogenously determined, there is no indication in the archives that the ordering of the varieties was influenced by import volumes or import shares. Rather, it seems that the EMB proceeded from the senior Dominions to non-self-governing colonies, featuring a leading foodstuff produced in each. Table 3 Sample composition Poster . Corresponding commodity . Non-advertised variety/varieties . Advertised variety/varieties . Quarter of poster release . Dairy Butter Argentina Australia Denmark Finland Netherlands Sweden New Zealand Irish Free State 1927Q1 (NZ) 1927Q4 (Ireland) Dairy Cheese Australia Canada Italy Netherlands New Zealand 1927Q1 Wheat Grain Argentina Australia United States Canada 1927Q1 Sugar Sugar British West Indies Cuba Peru St. Domingo Mauritius 1927Q1 Rice Rice United States India 1927Q2 Tea Tea China India Java and Sumatra Ceylon 1927Q2 Cattle Frozen beef Argentina New Zealand United States Uruguay Australia 1927Q4 Poster . Corresponding commodity . Non-advertised variety/varieties . Advertised variety/varieties . Quarter of poster release . Dairy Butter Argentina Australia Denmark Finland Netherlands Sweden New Zealand Irish Free State 1927Q1 (NZ) 1927Q4 (Ireland) Dairy Cheese Australia Canada Italy Netherlands New Zealand 1927Q1 Wheat Grain Argentina Australia United States Canada 1927Q1 Sugar Sugar British West Indies Cuba Peru St. Domingo Mauritius 1927Q1 Rice Rice United States India 1927Q2 Tea Tea China India Java and Sumatra Ceylon 1927Q2 Cattle Frozen beef Argentina New Zealand United States Uruguay Australia 1927Q4 Note: The fifth column states the quarter during which the poster was released. It should be emphasized that, in the baseline specification of the model (see table 4, col. 2), the Poster variable takes a value of 1 beginning in the quarter after the quarter during which the poster was released. Source: The National Archives, CO 760/22. Open in new tab Table 3 Sample composition Poster . Corresponding commodity . Non-advertised variety/varieties . Advertised variety/varieties . Quarter of poster release . Dairy Butter Argentina Australia Denmark Finland Netherlands Sweden New Zealand Irish Free State 1927Q1 (NZ) 1927Q4 (Ireland) Dairy Cheese Australia Canada Italy Netherlands New Zealand 1927Q1 Wheat Grain Argentina Australia United States Canada 1927Q1 Sugar Sugar British West Indies Cuba Peru St. Domingo Mauritius 1927Q1 Rice Rice United States India 1927Q2 Tea Tea China India Java and Sumatra Ceylon 1927Q2 Cattle Frozen beef Argentina New Zealand United States Uruguay Australia 1927Q4 Poster . Corresponding commodity . Non-advertised variety/varieties . Advertised variety/varieties . Quarter of poster release . Dairy Butter Argentina Australia Denmark Finland Netherlands Sweden New Zealand Irish Free State 1927Q1 (NZ) 1927Q4 (Ireland) Dairy Cheese Australia Canada Italy Netherlands New Zealand 1927Q1 Wheat Grain Argentina Australia United States Canada 1927Q1 Sugar Sugar British West Indies Cuba Peru St. Domingo Mauritius 1927Q1 Rice Rice United States India 1927Q2 Tea Tea China India Java and Sumatra Ceylon 1927Q2 Cattle Frozen beef Argentina New Zealand United States Uruguay Australia 1927Q4 Note: The fifth column states the quarter during which the poster was released. It should be emphasized that, in the baseline specification of the model (see table 4, col. 2), the Poster variable takes a value of 1 beginning in the quarter after the quarter during which the poster was released. Source: The National Archives, CO 760/22. Open in new tab Data on variety-specific import volumes and prices were obtained from the monthly reports of the Trade of the United Kingdom. Prices are taken to be average bilateral unit values, calculated by dividing the bilateral import value by the bilateral import volume. For the purpose of the panel data analysis, monthly data are aggregated to quarterly data. Quarterly-frequency data are preferable for two reasons. First, if a poster did raise the volume of a variety, this effect is unlikely to be observed in the month when, or even the month after, the poster was displayed. For a considerable period of time, increased demand might be satisfied through the depletion of inventories, while shipments were in transit. It is worth recalling the immense distances characterizing trade within the British empire, particularly Anglo-Australasian trade. The use of quarterly-frequency data allows for a delay between the display of the poster (and increased consumer expenditure) and the arrival of the commodities in a British port, without the need to introduce many (monthly) lags into the panel data analysis. Second, for some of the varieties included in the sample—the contents of the sample are discussed shortly—the value and volume of imports in particular months (though not quarters) is either trivially low or zero. Hence, it would be impossible to calculate the price variable consistently, for certain varieties, necessitating the exclusion of these varieties from the sample. The use of quarterly-frequency data therefore permits a much richer sample. The varieties included in the sample are stated in table 3. The sample consists of 33 varieties that fall within seven commodities: butter, cheese, grain, sugar, rice, tea, and frozen beef. There are 21 countries of origin represented in the sample. For each commodity, the sample includes all varieties, including the advertised empire variety, non-advertised empire varieties, and non-empire varieties, that meet both of the following criteria. First, the variety must be consistently bilaterally disaggregated in Britain’s monthly trade statistics for the entire interval from 1923Q2-1929Q4. The Trade of the United Kingdom reports the values and quantities of commodity imports on a partly bilaterally disaggregated basis. The values and quantities of commodity imports are not always bilaterally disaggregated for all sources, as the values and quantities of commodity imports from minor sources are included in a residual category of “other countries.” If, for any quarter from 1923Q2–1929Q4, the imports of a variety are relegated to the residual category of “other countries,” in which case the value and quantity of imports of the variety are unknown, then it is necessary that the variety be excluded from the sample. The second criterion is that the import volume is non-zero in every quarter, although not necessarily in every month. The interval of the analysis spans 27 periods from 1923Q2–1929Q4, making for a sample of 891 observations. Extending the interval earlier would introduce a discontinuity in the series resulting from the inclusion of Ireland in the trade statistics of the United Kingdom before 1923Q2 and would also encroach upon the erratic price fluctuations of the early 1920s.19 The interval is ended in 1929Q4, before the onset of the Great Depression and before such pathbreaking policies as the sterling devaluation of 1931, the Import Duties Act of 1932, and the Ottawa agreements of 1932. For the period from 1923Q2–1929Q4, figure 1 depicts the quarterly shares, according to value, of the following two sub-samples in the total sample: empire varieties (15), irrespective of whether they featured in a poster in 1927; and only those empire varieties (8) that were advertised in a poster in 1927.20 Figure 1 Open in new tabDownload slide Shares of empire varieties in sample (by value), 1923Q2–1929Q4. Source: Board of Trade, Accounts relating to Trade and Navigation of the United Kingdom (various years). Notes: For the varieties included in each series, see table 3. 4.2. Main analysis To assess the EMB’s effectiveness in raising Britain’s empire imports, the following panel regression equation is estimated, pooling all of the 33 varieties in the sample: $$\begin{equation} \mathit{\ln}\left({Imports}_{c,v,t}\right)={\beta}_0+{\beta}_1\mathit{\ln}\left({Price}_{c,v,t}\right)+{\beta}_2\left({Poster}_{c,v,t}\right)+{\gamma}_{c,t}+{\delta}_{c,v,q}+{\varepsilon}_{c,v,t} \end{equation}$$[1] The subscript c stands for the commodity (7 commodities), v for the variety (33 varieties), t for the period (27 quarterly periods), and q for the quarter (4 quarters). The distinction between quarterly periods and quarters should be emphasized: whereas there are 27 quarterly periods (1923Q2–1929Q4), there are only four quarters (Q1, Q2, Q3, and Q4). Equation 1 includes a commodity-period interaction fixed effect (γc,t) that controls for, inter alia, non-variety-specific demand shifts. Additionally, a variety-quarter interaction fixed effect (δc,v,q) controls for the desynchronized seasonality of Britain’s bilateral imports of agricultural commodities. For example, it would be expected that Britain’s bilateral imports of grain from Canada and from Argentina would peak in different quarters, due to the harvest season differing between hemispheres. Such variety-specific variation would not be captured by γc,t. The Price variable controls for price-induced substitution between different varieties of the commodity. The (binary) explanatory variable of interest, Poster, takes a value of 1 beginning in the quarter after the release of the variety-specific poster and continuing through the end of the interval, i.e., 1929Q4. For the 25 varieties (out of 33) in the sample that were non-advertised, this variable takes a value of 0 in every period. Columns 1–4 of table 4 present the results estimated using OLS. Column 1 reports the results excluding the fixed effects indicated in Equation 1. Column 2, which represents the preferred specification, reports the results including the fixed effects indicated in Equation 1. Whereas the assumption implicit in columns 1 and 2 is that any potential effect of the posters was enduring, it is possible that these posters galvanized consumers to shift their expenditure to advertised imports on a temporary basis. Therefore, column 3 presents an alternative specification (including the fixed effects) in which Poster takes a value of 1 for only the four periods following the display of the poster, rather than for the remaining duration of the interval. Column 4 represents another approach (including the fixed effects) to addressing this possibility, whereby the interval is truncated in 1928Q2, or shortly after the release of all of the posters (from 1927) in the sample. Reassuringly, in columns 1–4, the coefficient of Price is consistently negative and statistically significant. However, the coefficient of Poster is nowhere close to statistical significance, and its sign is inconsistent. Table 4 Britain’s variety import volumes, 1923Q2–1929Q4 . (1) . (2) . (3) . (4) . (5) . (6) . (7) . (8) . Price −0.55*** (0.18) −2.06*** (0.71) −2.05*** (0.70) −2.05*** (0.87) −0.21* (0.11) −2.21*** (0.37) −2.18*** (0.43) −1.88*** (0.40) Poster 0.01 (0.10) 0.03 (0.14) −0.01 (0.12) 0.37 (0.27) 0.05 (0.11) 0.05 (0.07) Poster (four periods) 0.00 (0.09) 0.03 (0.07) Constant 13.48*** (0.26) 13.97*** (0.51) 13.98*** (0.50) 14.35*** (0.50) 14.14*** (0.19) 13.86*** (0.26) 13.89*** (0.28) 14.39*** (0.24) Commodity-period fixed effects NO YES YES YES NO YES YES YES Variety-quarter fixed effects NO YES YES YES NO YES YES YES Interval 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 N 891 891 891 693 891 891 891 693 R2 0.04 0.89 0.89 0.88 Pseudo-R2 0.10 0.94 0.94 0.94 Estimation method OLS OLS OLS OLS PPML PPML PPML PPML . (1) . (2) . (3) . (4) . (5) . (6) . (7) . (8) . Price −0.55*** (0.18) −2.06*** (0.71) −2.05*** (0.70) −2.05*** (0.87) −0.21* (0.11) −2.21*** (0.37) −2.18*** (0.43) −1.88*** (0.40) Poster 0.01 (0.10) 0.03 (0.14) −0.01 (0.12) 0.37 (0.27) 0.05 (0.11) 0.05 (0.07) Poster (four periods) 0.00 (0.09) 0.03 (0.07) Constant 13.48*** (0.26) 13.97*** (0.51) 13.98*** (0.50) 14.35*** (0.50) 14.14*** (0.19) 13.86*** (0.26) 13.89*** (0.28) 14.39*** (0.24) Commodity-period fixed effects NO YES YES YES NO YES YES YES Variety-quarter fixed effects NO YES YES YES NO YES YES YES Interval 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 N 891 891 891 693 891 891 891 693 R2 0.04 0.89 0.89 0.88 Pseudo-R2 0.10 0.94 0.94 0.94 Estimation method OLS OLS OLS OLS PPML PPML PPML PPML Notes: * indicates statistical significance at the 10% level and *** at the 1% level. Robust standard errors have been clustered at the variety level and are reported in parentheses. The dependent variable is the volume of Britain’s variety imports. A variety is a bilaterally defined commodity. Volumes are quantities multiplied by constant 1923Q2 prices. The dependent variable is expressed in natural logarithms in cols. 1–4 and in levels in cols. 5–8. Price is expressed in natural logarithms in all columns. Poster is a binary variable taking a value of 1 beginning in the period after the variety-specific poster was displayed, and continuing for the duration of the interval. Poster (four periods) is a binary variable taking a value of 1 for only four consecutive periods, beginning in the period after the variety-specific poster was displayed. Open in new tab Table 4 Britain’s variety import volumes, 1923Q2–1929Q4 . (1) . (2) . (3) . (4) . (5) . (6) . (7) . (8) . Price −0.55*** (0.18) −2.06*** (0.71) −2.05*** (0.70) −2.05*** (0.87) −0.21* (0.11) −2.21*** (0.37) −2.18*** (0.43) −1.88*** (0.40) Poster 0.01 (0.10) 0.03 (0.14) −0.01 (0.12) 0.37 (0.27) 0.05 (0.11) 0.05 (0.07) Poster (four periods) 0.00 (0.09) 0.03 (0.07) Constant 13.48*** (0.26) 13.97*** (0.51) 13.98*** (0.50) 14.35*** (0.50) 14.14*** (0.19) 13.86*** (0.26) 13.89*** (0.28) 14.39*** (0.24) Commodity-period fixed effects NO YES YES YES NO YES YES YES Variety-quarter fixed effects NO YES YES YES NO YES YES YES Interval 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 N 891 891 891 693 891 891 891 693 R2 0.04 0.89 0.89 0.88 Pseudo-R2 0.10 0.94 0.94 0.94 Estimation method OLS OLS OLS OLS PPML PPML PPML PPML . (1) . (2) . (3) . (4) . (5) . (6) . (7) . (8) . Price −0.55*** (0.18) −2.06*** (0.71) −2.05*** (0.70) −2.05*** (0.87) −0.21* (0.11) −2.21*** (0.37) −2.18*** (0.43) −1.88*** (0.40) Poster 0.01 (0.10) 0.03 (0.14) −0.01 (0.12) 0.37 (0.27) 0.05 (0.11) 0.05 (0.07) Poster (four periods) 0.00 (0.09) 0.03 (0.07) Constant 13.48*** (0.26) 13.97*** (0.51) 13.98*** (0.50) 14.35*** (0.50) 14.14*** (0.19) 13.86*** (0.26) 13.89*** (0.28) 14.39*** (0.24) Commodity-period fixed effects NO YES YES YES NO YES YES YES Variety-quarter fixed effects NO YES YES YES NO YES YES YES Interval 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 N 891 891 891 693 891 891 891 693 R2 0.04 0.89 0.89 0.88 Pseudo-R2 0.10 0.94 0.94 0.94 Estimation method OLS OLS OLS OLS PPML PPML PPML PPML Notes: * indicates statistical significance at the 10% level and *** at the 1% level. Robust standard errors have been clustered at the variety level and are reported in parentheses. The dependent variable is the volume of Britain’s variety imports. A variety is a bilaterally defined commodity. Volumes are quantities multiplied by constant 1923Q2 prices. The dependent variable is expressed in natural logarithms in cols. 1–4 and in levels in cols. 5–8. Price is expressed in natural logarithms in all columns. Poster is a binary variable taking a value of 1 beginning in the period after the variety-specific poster was displayed, and continuing for the duration of the interval. Poster (four periods) is a binary variable taking a value of 1 for only four consecutive periods, beginning in the period after the variety-specific poster was displayed. Open in new tab Columns 5–8 replicate columns 1–4, but using a Poisson Pseudo-Maximum Likelihood (PPML) estimator, with the dependent variable, Imports, expressed in levels. Still, the coefficient of Poster is statistically insignificant. Alternative specifications of Equation 1 replace the dependent variable, Import (volume), with import values and import prices. The specifications are presented in Appendix table 1 (import values) and Appendix table 2 (import prices). These appendix tables replicate the specifications in table 4, apart from the change of dependent variable and the dropping of price as an explanatory variable. Altogether, there is no econometric evidence that the EMB’s variety-specific posters, which were its most defined posters, raised the volume, value, or price of Britain’s empire imports. 4.3. Event study As an alternative approach to assessing the effectiveness of the EMB’s advertising campaign, this paper conducts an event study taking the following form: $$\begin{equation} \mathit{\ln}\left({Imports}_{c,v,t}\right)={\beta}_0+{\beta}_1\left({Price}_{c,v,t}\right)+\sum_{i=-n}^n{\alpha}_i\times 1\left(t=i\right)\times{Poster}_{c,v}+{\gamma}_{c,t}+{\delta}_{c,v,q}+{\varepsilon}_{c,v,t} \end{equation}$$[2] Equation 2 is identical to Equation 1 apart from the definition of the Poster variable. In Equation 2, Poster takes a value of 1 for all advertised varieties—and in all periods. Thus, unlike in Equation 1, the value of Poster is time-invariant in Equation 2. In this event study, the “event” is the release of the poster. However, the event is non-synchronous, since, as already discussed, the event occurs in different quarters of 1927 for different commodities (see table 3). In the equation, i represents a (quarterly) period relative to the commodity-specific event. For the period during which the event occurs (i.e., the poster is released), i = 1. Therefore, the reference period (i = 0) is the period preceding the release of the poster. Poster is interacted with a binary variable taking a value of 1 in period i. For the above equation, n is set equal to 9. If n was set equal to any number greater than 9, then those events occurring in 1927Q4 would be excluded from the estimation of the coefficients (αi) for higher values of i, since the sample ends in 1929Q4. Setting n equal to 9 involves the estimation of 18 coefficients (αi) for the Poster interaction variable. To provide an example, consider the commodity tea. A poster advertising Ceylon tea was released in 1927Q2. For 1927Q2, i = 1; 1927Q3, i = 2; … 1929Q2, i = 9. The reference period is 1927Q1. For 1926Q4, i = −1; 1926Q3, i = −2; … 1924Q4, i = −9. Poster takes a value of 1 for Ceylon tea in all periods, both before and after the event. Meanwhile, Poster takes a value of 0 for non-Ceylon varieties of tea in all periods. Figure 2 depicts the estimated coefficients for those 18 variables that interact Poster with a binary variable taking a value of 1 in period i; it is worth reiterating that i is the period relative to the commodity-specific event. The figure also presents the 95% confidence interval for the coefficients. (The coefficients and standard errors are reported in Appendix table 3.) All of the pre-event coefficients are statistically insignificant, suggesting that, with respect to trends, advertised varieties were indistinct from non-advertised varieties. Moreover, all of the post-event coefficients are statistically insignificant, corroborating the earlier conclusion that the advertising of the EMB did not make an iota of difference. Figure 2 Open in new tabDownload slide Coefficients and 95% confidence intervals for estimated values of αi. Note: See equation 2 for αi. The event, which is the release of the poster, corresponds to i = 1. 5. Re-evaluating the failure of the Empire Marketing Board In this section, three major explanations are advanced for the failure of the EMB. First, the Board’s publicity expenditure was low when compared with that of the principal export control boards in Australia and New Zealand. Second, and more significantly, irrespective of its publicity expenditure, the EMB’s campaign could not overcome the problems of proving provenance when consumers asked for the produce of a specific country. This latter argument is based on two case studies: butter and tea. Finally, the EMB failed to recognize a pronounced shift in consumer attitudes toward frozen Australian beef. Contemporaries recognized that the publicity campaign envisaged by the EMB was “novel and difficult”; it focused on “the advertisement of an idea rather than a commodity” (IEC, Eighth Report, 1928, p. 493). Subsequent research has confirmed this statement by referring to a contemporary report that claimed that New Zealand producers thought the EMB’s posters were “ineffective as sellers of produce and of little practical value” (Barnes 2012, pp. 177–8).21 This latter literature also emphasizes the constraints within which the Board operated. For example, even at its peak in 1928, the Board’s total publicity expenditure was infinitesimal compared with total national advertising in that year: £0.28 million versus £57 million (Constantine 1986a, p. 221). Moreover, Treasury skepticism about the Board’s activities, and opposition by the Committee on National Expenditure during the economic crisis of 1931, almost resulted in the abolition of the EMB; it was saved, but the result was a significant reduction in its budget and a switch of focus to the “Buy British” campaign of 1931 (Atkins 2003, p. 15). In any event, publicity for empire produce was declining before 1931. Between 1926 and 1930, the proportion of funds expended by the Board on publicity declined from 74.5 to 33.3 percent, while that for research increased commensurately (see table 1). The explanation for the relative decline in publicity expenditure is pertinent. The EMB became concerned about devoting too many resources to publicity until scientific research improved the quantity and, more importantly, the quality of empire produce (House of Commons, Committee on Estimates, 1932, p. 743). The EMB recognized that if its publicity campaign to persuade consumers to buy more empire produce was to succeed, it was imperative that the standard of empire produce was improved (Constantine 1986b, p. 3). In fact, considering the entire period from 1926–33, total expenditure on research comfortably exceeded that on publicity (see table 1). However, it should not be assumed that expenditures on publicity and research were mutually exclusive. Research was necessary to improve the overall quality of empire produce so that it could better compete with generally higher-quality foreign supplies. Otherwise, consumers were being encouraged to purchase relatively inferior products (Atkins 2003, 4.2/9–4.2/10). In any event, the publicity expenditure intensity of the variety-specific marketing campaigns launched by the Australian Dairy Produce Control Board (ADPCB), the New Zealand Dairy Produce Control Board (NZDPCB), and the New Zealand Meat Producers Board (NZMPB) considerably exceeded that of the EMB, when calculated on a standardized basis.22 Between 1928 and 1931, the publicity expenditure intensity of the ADPCB, NZDPCB, and NZMPB exceeded that of the EMB by a factor of 1.7, 2.0, and 1.2, respectively (5.3, 6.0, and 3.8, if only the EMB’s poster expenditure is considered). Not only was the publicity expenditure intensity of the EMB lower, but it was also less focused. Included in the EMB’s publicity expenditure were such posters as “The River Mersey” poster, displayed in 1928, which makes no explicit reference to the empire or a foodstuff (Constantine 1986b, plate 19). Consequently, only a portion of the EMB’s publicity expenditure had even the potential to influence the purchasing decisions of British consumers. Moreover, following opposition by the National Farmers Union, the EMB’s publicity expenditure was required to advertise domestically produced foodstuffs, which clearly undermined the extent to which the EMB was able to promote successfully the produce of the overseas empire (Atkins 2003, pp. 4.2/4–4.2/5). The second explanation for the failure of the EMB focuses on the extent to which consumers were able to identify produce originating from the empire and specific Dominions. Exhortations to purchase “New Zealand” butter, “Ceylon” tea, or “Australian” beef require that these products are marked with an indication of their origin. The IEC recognized the importance of such marking and was favorably disposed to the introduction of legislation mandating this practice (IEC, First Report, 1925, pp. 6–10). But, the academic literature on such legislation—embodied in the Merchandise Marks Act, 1926, discussed shortly—has either overlooked the potential significance of this Act or tended to overstate its importance. In the latter category, it has been claimed that “The branding of home and colonial goods … would benefit consumers and producers alike,” and that by enabling the stamping of a good’s country of origin, the Act enabled consumers to distinguish sub-standard articles (Atkins 2003, p. 4.2/10; Trentmann 2008, p. 239). Compulsory origin marking was introduced by the Merchandise Marks Act, 1926. This Act stipulated that producers could apply for a Marking Order to ensure that specific foodstuffs were marked with their country of origin. This legislation was not comprehensive, as it did not apply to all produce: a standing committee had to be satisfied that an Order was in the public interest before giving approval.23 However, the 1926 Act did not specify a precise form of indication. Retailers were permitted to apply either the legend “Empire” if a foodstuff was produced in an overseas Dominion, or they could indicate a specific country.24 Applications for Marking Orders were successful for butter and beef, but unsuccessful for tea; no such applications were made for cheese, grain, rice, or sugar. Most of the butter imported into Britain was in bulk, in boxes, or in casks, indicating its origin. New Zealand butter, for example, was exported in consignments weighing 56 pounds. The butter was then supplied to retailers who assembled packets of butter (pats) for their customers according to the weight demanded. At this stage, only grocers knew the origin of the butter—which they were not required to indicate. Consequently, grocers were crucial in determining the kind of butter that was purchased (Ministry of Agriculture and Fisheries, hereafter MAF, Marketing of Dairy, 1932, pp. 24, 69–71, and 84). In 1930, the leading empire butter exporters, supported by the British Dairy Farmers Association and the British Empire Producers’ Organization, sought to secure a Marking Order under the 1926 Act, in order to ensure that retail butter was marked with an indication of origin. The applicants claimed that: At the present time there is no regulation requiring the marking of butter when exposed for sale either wholesale or retail with an indication of origin, and consequently, the public is not encouraged to express a preference for Home or Empire produce and is afforded no satisfactory guarantee even if such preference is expressed. The applicants are of opinion that the imposition of a mark of origin will result in a largely increased demand for Home and Empire produce (Application for a Butter Marking Order: Summary of Applicants’ Case. The National Archives, MAF 36/151). In support of their case, the applicants argued that it was impossible for ordinary consumers to distinguish butter according to its country of origin. Moreover, irrespective of its origin, imported butter was often described as “Choice” or “Best Dairy,” which was detrimental to domestic and empire producers. To prevent ambiguity, it was suggested that, at the point of retail sale, imported butter be marked by means of printing on, or affixing printed labels to each package, an indication of origin (Merchandise Marks Act, Report of the Standing Committee on Butter, 1931, pp. 5–7). Grocers’ and retailers’ organizations were adamantly opposed to the imposition of a butter Marking Order. These bodies claimed that consumers’ best interests were served when butter was sold according to price, not country of origin. Grocers were struggling to observe other regulations, such as Weights and Measures, and, because they sold many different types of butter, the requirement that each package should be marked increased the risk of prosecution (Giles to Johns, 27 October, 1930. The National Archives, MAF 36/152). Other arguments were that the scale of butter misrepresentation was exaggerated, that the proposed Marking Order would disadvantage domestic and empire butters by advertising the produce of European countries, especially Denmark, which was supplied in a consistently high quality, and that the Order would harm the British blended butter industry (Merchandise Marks Act, Report of the Standing Committee on Butter, 1931, p. 8). The Standing Committee was persuaded by the applicants and recommended that, in the retail trade, butter should be clearly marked with an indication of origin. However, the Committee did not prescribe a form of marking which differed from the 1926 Act. Consequently, grocers and other retailers remained at liberty to apply the legend “Empire” or a definite indication of origin (Merchandise Marks Act, Report of the Standing Committee on Butter, 1931, p. 13). But, use of the former indicia undermined the poster campaign of the EMB, which sought to promote New Zealand and Irish Free State butter; it also threatened to nullify the marketing efforts of the ADPCB and the NZDPCB, which sought to differentiate their butter from other empire and foreign butters. Indeed, in 1934, one year after the abolition of the EMB, the New Zealand Dairy Commission reported that, “customers who ask for New Zealand butter are told that ‘Empire’ means the same thing … [although] the publicity given by the Empire Marketing Board had helped New Zealand by encouraging consumers to buy Empire goods … New Zealand had recently tended to lose by reason of the merger of its identity … we are of opinion that it would be advisable to … [have] the marking regulations amended to require the name of the country of origin to be stated” (New Zealand Dairy Industry Commission, Dairy Industry Commission, 1934, p. 35).25 Similar problems occurred with tea imports. In 1929, London-based agents for the Indian Tea Association, the South Indian Association, and the Ceylon Association sought a Marking Order to ensure that the sale of each packet of tea was accompanied by an indication of origin. This indication could be “Foreign” when a tea consisted solely of foreign tea, “British Empire” when the tea was sourced entirely from within the empire, or “Blended British Empire and Foreign” when the tea was a blend of ingredients from within and outside the empire (Board of Trade, Committee respecting Tea, 1929, p. 4). Opponents to the scheme, including both the Tea Buyers Association and the Federation of Grocers’ Associations of the United Kingdom, concurred that, in the absence of a mark of origin, a consumer would be unable to differentiate between empire and foreign tea. Nonetheless, they argued that the characteristics of the tea trade made it impossible to impose indications of origin. For example, most of the tea consumed in Britain was blended and sold with company brands or trademarks, such as Brooke Bond’s “Red Label.” These brands possessed a valuable goodwill, and to vary them with the legends “Empire” or “Empire and Foreign” would lead consumers to believe that the quality of the tea was being varied. Dealers blended teas to obtain the required standards of quality, taste, and price, with little concern about the country in which the tea originated. Thus, it was a “common practice” to blend Java or Sumatra tea with Indian or Ceylon tea to “secure a particular quality, or to keep down the price.” The proportion of Java and Sumatra tea in the cheapest blends could exceed 30 percent. Conversely, Chinese tea was used in the most expensive blends (Board of Trade, Committee respecting Tea, 1929, pp. 5–6). In view of these arguments and the fact that it was impossible even for experts to detect the presence of foreign tea in a blend, the Committee rejected the application for a Marking Order (Board of Trade, Committee respecting Tea, 1929, p. 8). This decision was “an overwhelming victory to the distributors of tea, a victory that brought joy to the heart of those growing tea outside the British Empire” (The Grocer and Oil Trade Review, 20 April 1929, p. 80).26 Further weaknesses in the EMB’s campaign were evident in its promotion of frozen Australian beef. Contemporaries recognized that this type of beef was generally unacceptable to the British market. The substantial difference in prices for domestic, imported chilled beef, and imported frozen beef testifies to the discrimination that domestic consumers exercised in their choice of beef. There was agreement that fresh beef ranked highest in price, followed by chilled beef and then frozen beef. Some authorities ranked frozen beef from South America higher than that from Australia because the former benefitted from better quality cattle and their beef was “dressed” in a more marketable fashion (IEC, Second Report, 1925, p. 11; MAF, Beef, Mutton and Lamb, 1925, pp. 14 and 19; MAF, Cattle and Beef, 1929, pp. 25 and 34). Unsurprisingly, therefore, between 1920 and 1928, chilled beef as a percentage of total British beef imports (chilled and frozen) increased from 11 to 83 percent (MAF, Cattle and Beef, 1929, p. 13). In volume terms, comparing 1924 with 1909–13, British imports of all beef products from foreign countries increased from 83.1 to 86.7 percent, while those from the empire declined from 16.9 to 13.3 percent (IEC, Second Report, 1925, p. 6, table III). By value, Australia accounted for 54.6 percent of Britain’s beef imports from the empire in 1924, but this represented just 5.0 percent of Britain’s total beef imports, the bulk of which were obtained from South America (IEC, Second Report, 1925, p. 8, table V). By the late 1920s, the trade in frozen beef had practically disappeared, and Australia was “in the unflattering position of having to sell its frozen beef primarily to the armed services, institutions, and manufacturers of meat products in Britain because the retail trade did not want it” (Henzell 2007, p. 131).27 6. Conclusions To date, much of the economic history literature on marketing and branding has adopted a microeconomic perspective directed toward individual firms and industries (Church 2000). This article has enlarged the marketing focus by considering a central government initiative to promote the produce supplied by many firms in multiple agricultural industries between the mid-1920s and the early 1930s. The EMB’s objective was to increase Britain’s empire imports at a time when there existed significant competition from foreign countries, such as Denmark (butter) and Argentina and the River Plate (beef). Exacerbating matters, not only did British consumers have well-established preferences for these products, but there was growing European competition within the British market for branded, graded, and marketed foodstuffs, such as butter from Finland, Latvia, and the Netherlands (MAF, Marketing of Dairy, 1932, pp. 48–61). As this article has found, the EMB was unable to overcome such challenges and achieve its aim of shifting consumer expenditure toward imports from the empire. Contributing to the EMB’s economic failure was its low publicity expenditure intensity relative to that of the commodity-specific marketing campaigns of the Antipodean control boards. It is surprising, therefore, that in discussions leading to the abolition of the EMB in 1933, the Secretary of State for the Colonies claimed that he did not believe that the publicity campaigns launched by the Dominions would “ever succeed in building up a goodwill similar to that built up by the Marketing branch of the [Empire Marketing] Board” (The National Archives, CO 323/1248/19). However, as Constantine (1993, p. 365) has argued, Canadian officials mistrusted the EMB and believed spending Canadian money to promote its own produce was the only measure compatible with national integrity. Another explanation for the failure of the EMB was weakness in the Merchandise Marks Act, 1926. This Act sought to address unfair competition; it was predicated on the view that consumers were being misled into purchasing foreign produce in the belief that it was either British or empire produce. However, because retailers had considerable leeway in the choice of indication they applied to produce, there was no guarantee that consumers requesting New Zealand butter actually received this product. Although a Marking Order for beef was introduced in 1933, it conflicted with a significant change in consumer preferences away from frozen Australian beef. Moreover, such Orders did not apply to the other commodities included in the econometric analysis: cheese, grain, rice, sugar, and tea. In short, the soft trade policy of publicity that Britain pursued via the EMB was ineffective. The British market had long since been supplied by both empire and, to a substantial extent, foreign producers of foodstuffs. Non-tariff preferences taking the form of advertising were inadequate to raise the volume of imports from the empire in the 1920s. An efficacious policy of imperial preference would require tariff preferences—and that came in the following decade. Acknowledgements We wish to thank Felicity Barnes, Stephen Constantine, Jason Lennard, David Thackeray, John Wildman, the Editor (Chris Meissner), and the three anonymous referees for their helpful comments. We are also grateful for the comments received at the Economic History Society Annual Conference at Queen’s University Belfast (2019), as well as seminars at Newcastle University and the Max Planck Institute for European Legal History. All errors are our own. Footnotes 1 Constitutionally, the EMB was only an advisory committee under the Secretary of State for the Dominions, but as he (Leo Amery) was an ex officio chairman, his authority gave the EMB executive power, and it drew upon its own annual budget (Constantine 1986b, p. 3). 2 Although there were constitutional differences between Dominions and colonies, this article uses the former term, since the sample in section 4 mostly consists of imports from Dominions. 3 Some additional bilateral agreements were concluded between Dominions. 4 The margin of preference is the difference between the (lower) duty on imports from the preferred country, in this case Britain, and the (higher) duty on imports from other countries. 5 The Import Duties Act of 1932, which brought to an end Britain’s longstanding policy of free trade, was preceded by the transient Abnormal Importations Act of 1931. This earlier Act permitted the imposition of temporary duties of up to 100% ad valorem on goods imported in abnormally great quantities (McGuire 1939, p. 241). 6 The share of Britain’s imports from the empire receiving preference increased immensely, from 7% in 1929 to 60–61% in 1937. 7 Although, it should not be assumed that every bilateral trade flow within the British empire conformed to this general finding. As Jacks (2014) has argued, the Ottawa agreements raised neither the empire share of Canada’s exports nor the empire share of Canada’s imports, as the preferential trade policy was insufficient to overcome the advantage of distance realized by the United States in its trade with Canada. 8 See Varian (2019, p. 706), for a full list of pre-1932 protective legislation, which covered 9% of Britain’s manufacturing sector in 1930. 9 Yet, it should be observed how relatively minimal were Britain’s imports of motorcars (from all countries), which comprised less than 4% of domestic consumption in 1930 (Board of Trade, Fourth Census of Production, 1934, p. 342). 10 India did not adopt a policy of imperial preference until 1927 (MacDougall and Hutt 1954, p. 237). 11 Throughout this article, “foreign” denotes those areas outside the British empire. 12 The Imperial Economic Committee recognized that improvements in quality and the preparation of foodstuffs for market were equally important. The latter includes, for example, better packaging, grading, and standardization (IEC, First Report, 1925, p. 13). 13 Much of the Board’s publicity overlapped with the initiatives of the Women’s Patriotic League (Trentmann 2008, pp. 228–40). 14 See also House of Commons, Committee on Estimates (1932, pp. 25–6, Q.253; p. 28, Q.281). 15 The minutes of the 1 December 1926 meeting of the Poster Sub-Committee of the Publicity Committee indicate that the frames were due to be delivered by the end of December to the following locations: London (200), Manchester (40), Birmingham (35), Liverpool (30), Glasgow (30), “… and the remaining 365 frames between 38 other large towns on a basis of four frames to each 100,000 of the population.” According to Constantine (1986b, p. 11), the number of frames reached 1,000 by December 1927. 16 The frames accommodated sets of five posters. The first, third, and fifth posters in the frame were 60 × 40 inches. The second and fourth posters in the frame were 25 × 40 inches and often contained only text (Constantine 1986b, p. 6) 17 Britain’s monthly trade statistics include a bilateral disaggregation for only certain commodities. Even then, the bilateral disaggregation is only partial due to the category of “other countries.” An example of a poster excluded from the analysis due to a lack of corresponding trade data is the poster advertising South African oranges, since there is no bilateral disaggregation of Britain’s monthly orange imports. 18 Since bilateral cheese imports from the Irish Free State are not disaggregated in Britain’s monthly trade statistics, it is impossible to consider cheese as a representative commodity for the Irish Free State dairy poster. 19 For the price fluctuations of the early 1920s, see Aldcroft (1970, p. 32) and de Bromhead et al. (2019b, p. 131). 20 Although there were seven variety-specific posters, the sample includes eight advertised varieties, because there are two representative commodities for the New Zealand dairy poster: butter and cheese. 21 See also Constantine (1986a, pp. 223–4) and Constantine (1986b, p. 17). 22 To calculate publicity expenditure intensity, publicity expenditure is standardized by the value of imports that were publicized. Since the remit of the EMB extended to all empire goods, the EMB’s publicity expenditure is standardized by the total value of Britain’s empire imports. The publicity expenditures of the produce control boards are standardized by the value of Britain’s imports of the corresponding import varieties. For example, the publicity expenditure of the NZDPCB is standardized by the value of Britain’s imports of butter and cheese from New Zealand, in order to calculate the publicity expenditure intensity of the NZDPCB. The data sources for publicity expenditure are Empire Marketing Board: EMB, Work of the Board (1932); Australia Dairy Board: Commonwealth of Australia, Dairy Produce Control Board (various years); New Zealand Dairy Board: New Zealand Dairy Produce Control Board, Annual Reports (various years); and New Zealand Meat Board: New Zealand Meat Producers Board, Annual Reports (various years). The data sources for import values are: value of empire imports: Schlote (1952, pp. 123 and 163); and value of variety imports: Board of Trade, Trade of the United Kingdom (various years). 23 16 & 17 GEO.5. Merchandise Marks Act, 1926, s.1 (3). 24 Merchandise Marks Act, 1926, s.10 (1). 25 As far as can be determined, no further changes were made governing the marking of butter imports into Britain. 26 Moreover, Ceylon tea had to compete with the rapid growth in production of tea from Java and Sumatra. Between 1910 and 1931, the tea crop in Ceylon, Java, and Sumatra increased by 39%, 400%, and 400%, respectively (The Grocer and Oil Trade Review, 21 November 1931, p. 73). 27 See also Duncan (1996, p. 85) and Perren (2006, p. 137). References Aldcroft , D.H. ( 1970 ). The Inter-War Economy: Britain, 1919–1939 . London : B. T. Batsford . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Atkins , P. J. ( 2003 ). Food and the Empire Marketing Board in Britain, 1926–1933. Eighth Symposium of the International Commission for Research into European Food History, (mimeo) . Barnes , F. ( 2012 ). New Zealand’s London: A Colony and its Metropolis . Auckland : Auckland University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Barnes , F. ( 2014 ). Bringing another empire alive? The Empire Marketing Board and the construction of Dominion identity, 1926–1933 . Journal of Imperial and Commonwealth History 42 , pp. 61 – 85 . Google Scholar Crossref Search ADS WorldCat Barnes , F. and Higgins , D.M. ( 2020 ). Brand image, cultural association and marketing: “New Zealand” butter and lamb exports to Britain, c. 1920–1938 . Business History 62 , pp. 70 – 97 . Google Scholar Crossref Search ADS WorldCat de Bromhead , A. , Fernihough , A., Lampe , M. and O’rourke , K.H. ( 2019a ). When Britain turned inward: the impact of interwar British protection . American Economic Review 109 , pp. 325 – 352 . Google Scholar Crossref Search ADS WorldCat de Bromhead , A. , Fernihough , A., Lampe , M. and O’rourke , K.H. ( 2019b ). The anatomy of a trade collapse: the UK, 1929–1933 . European Review of Economic History 23 , pp. 123 – 144 . Google Scholar Crossref Search ADS WorldCat Church , R. ( 2000 ). Advertising consumer goods in nineteenth century Britain . Economic History Review 53 , pp. 621 – 645 . Google Scholar Crossref Search ADS WorldCat Constantine , S. ( 1986a ). Bringing the empire alive: the Empire Marketing Board and imperial propaganda, 1926–33. In J.M. MacKenzie (ed), Imperialism and Popular Culture , 193–231 . Manchester : Manchester University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Constantine , S. ( 1986b ). Buy and Build: the Advertising Posters of the Empire Marketing Board . London : HMSO . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Constantine , S. ( 1993 ). Anglo-Canadian relations, the Empire Marketing Board and Canadian national autonomy between the wars . Journal of Imperial and Commonwealth History 21 , pp. 357 – 384 . Google Scholar Crossref Search ADS WorldCat Drummond , I.M. ( 1974 ). Imperial Economic Policy, 1917–1939 . London : Allen and Unwin . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Duncan , R. ( 1996 ). The demand for frozen beef in the United Kingdom, 1880–1940 . Journal of Agricultural Economics 12 , pp. 82 – 88 . Google Scholar Crossref Search ADS WorldCat Eichengreen , B. and Irwin , D.A. ( 1995 ). Trade blocs, currency blocs and the reorientation of world trade in the 1930s . Journal of International Economics 38 , pp. 1 – 24 . Google Scholar Crossref Search ADS WorldCat Ettenson , R. , Gaeth , G. and Wagner , J. ( 1988 ). Evaluating the effect of country of origin and the “Made in the USA” campaign: a conjoint approach . Journal of Retailing 64 , pp. 85 – 100 . Google Scholar OpenURL Placeholder Text WorldCat Fenwick , G.D. and Wright , C.I. ( 2000 ). Effect of a buy-national campaign on member firm performance . Journal of Business Research 47 , pp. 135 – 145 . Google Scholar Crossref Search ADS WorldCat Fischer , W.C. and Byron , B. ( 1997 ). Buy Australian made . Journal of Consumer Policy 20 , pp. 89 – 97 . Google Scholar Crossref Search ADS WorldCat Frank , D. ( 1999 ). Buy American: The Untold Story of Economic Nationalism . Boston : Beacon Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Freeman , J. ( 1977 ). The publicity of the Empire Marketing Board, 1926–1933 . Journal of Advertising History 1 , pp. 12 – 14 . Google Scholar OpenURL Placeholder Text WorldCat Glickman , D.L. ( 1947 ). The British imperial preference system . Quarterly Journal of Economics 61 , pp. 439 – 470 . Google Scholar Crossref Search ADS WorldCat Gowa , J. and Hicks , R. ( 2013 ). Politics, institutions, and trade: lessons of the interwar era . International Organization 67 , pp. 439 – 467 . Google Scholar Crossref Search ADS WorldCat Henzell , T. ( 2007 ). Australian Agriculture: Its History and Challenges . Victoria : Csiro . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Higgins , D.M. ( 2018 ). Brands, Geographical Origin and the Global Economy . Cambridge : Cambridge University Press . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Jacks , D.S. ( 2014 ). Defying gravity: the Imperial Economic Conference and the reorientation of Canadian trade . Explorations in Economic History 53 , pp. 19 – 39 . Google Scholar Crossref Search ADS WorldCat Jacks , D.S. and Novy , D. ( 2020 ). Trade blocs and trade wars during the interwar period . Asian Economic Policy Review 15 , pp. 119 – 136 . Google Scholar Crossref Search ADS WorldCat Kothari , U. ( 2014 ). Trade, consumption and development alliances: the historical legacy of the Empire Marketing Board poster campaign . Third World Quarterly 35 , pp. 43 – 64 . Google Scholar Crossref Search ADS WorldCat MacDougall , D. and Hutt , R. ( 1954 ). Imperial preference: a quantitative analysis . Economic Journal 64 , pp. 233 – 257 . Google Scholar Crossref Search ADS WorldCat McGuire , E.B. ( 1939 ). The British Tariff System . London : Methuen . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Moore , J.A. ( 2016 ). Selling empire: a historical perspective on selling foreign products in domestic markets . Journal of Historical Research in Marketing 8 , pp. 263 – 283 . Google Scholar Crossref Search ADS WorldCat Perren , R. ( 2006 ). Taste, Trade, Technology: The Development of the International Meat Industry since 1840 . Aldershot : Ashgate . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Rooth , T. ( 1993 ). British Protectionism and the International Economy: Overseas Commercial Policy in the 1930s . Cambridge : Cambridge University Press . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Russell , R.S. ( 1947 ). Imperial Preference: Its Development and Effects . London : Empire Economic Union . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Schlote , W. ( 1952 ). British Overseas Trade From 1700 to the 1930’s . W. P. Henderson and W. H. Chaloner, trans . Oxford : Basil Blackwell . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Schwarzkopf , S. ( 2012 ). Markets, consumers, and the state: the uses of market research in government and the public sector in Britain, 1925–1955. In H. Berghoff, P. Scranton and U. Spiekermann (eds), The Rise of Marketing and Market Research, 171–92 . Basingstoke : Palgrave Macmillan . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Thackeray , D. ( 2017 ). Buying for Britain, China, or India? Patriotic trade, ethnicity, and market in the 1930s British empire/Commonwealth . Journal of Global History 12 , pp. 386 – 409 . Google Scholar Crossref Search ADS WorldCat Thackeray , D. ( 2019 ). Forging a British World of Trade . Oxford : Oxford University Press . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Trentmann , F. ( 2008 ). Free Trade Nation . Oxford : Oxford University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Varian , B.D. ( 2019 ). The growth of manufacturing protection in 1920s Britain . Scottish Journal of Political Economy 66 , pp. 703 – 711 . Google Scholar Crossref Search ADS WorldCat Periodical The Grocer and Oil Trade Review. London. Official publications Board of Trade. (various years). Accounts relating to Trade and Navigation of the United Kingdom. London: HMSO. Board of Trade. (1934). Final Report on the Fourth Census of Production (1930), vol. II. London: HMSO. Board of Trade. Merchandise Marks Act, 1926. (1929). Report of the Standing Committee respecting Tea. London: HMSO. Commonwealth of Australia. (various years). Annual Reports of the Dairy Produce Control Board. Melbourne. Empire Marketing Board. (1929). The Demand for Cheese in London. London: Empire Marketing Board. Empire Marketing Board. (1930). The Demand for Empire Butter. London: Empire Marketing Board. Empire Marketing Board. (1930). Empire Marketing Board May 1929 - May 1930, Paper No. 28. London: Empire Marketing Board. Empire Marketing Board. (1927). Note on the Work of the Board & Statement of Research Grants Approved by the Secretary of State from July, 1926, to May, 1927. London: HMSO. Empire Marketing Board. (1930). Note on the Work and Finance of the Board and Statement of Research and other Grants Approved by the Secretary of State for Dominion Affairs from July, 1926, to March 31st, 1930. London: HMSO. Empire Marketing Board. (1932). Note on the Work and Finance of the Board and Statement of Research and other Grants Approved by the Secretary of State for Dominion Affairs from July, 1926, to March 31st, 1932. London: HMSO. House of Commons. (1928). First and Second Reports from the Select Committee on Estimates … the Proceedings of the Committee. London: HMSO. House of Commons. (1932). First and Second Reports from the Select Committee on Estimates … the Proceedings of the Committee. London: HMSO. Imperial Economic Committee. (1925). Report of the Imperial Economic Committee on Marketing and Preparing for Market of Foodstuffs Produced in the Overseas Parts of the Empire: First Report – General. London: HMSO. Imperial Economic Committee. (1925). Report of the Imperial Economic Committee on Marketing and Preparing for Market of Foodstuffs Produced in the Overseas Parts of the Empire: Second Report – Meat. London: HMSO. Imperial Economic Committee. (1928). Reports of the Imperial Economic Committee: Eighth Report. Functions and Work of the Imperial Economic Committee. London: HMSO. Merchandise Marks Act, 1926. (1931). Report of the Standing Committee on Butter. London: HMSO. Ministry of Agriculture and Fisheries. (1925). Report on the Trade in Refrigerated Beef, Mutton, and Lamb. London: HMSO. Ministry of Agriculture and Fisheries. (1929). Report on the Marketing of Cattle and Beef in England and Wales. London: HMSO. Ministry of Agriculture and Fisheries. (1932). Report on the Marketing of Dairy Produce in England and Wales, Part II Butter and Cream. London: HMSO. New Zealand Dairy Industry Commission. (1934). New Zealand Dairy Industry Commission. Wellington. New Zealand Dairy Produce Control Board. (various years). Annual Reports and Accounts. Wellington. New Zealand Meat Producers Board. (various years). Annual Reports. Wellington. Appendix Table 1 Britain’s variety import values, 1923Q2–1929Q4 . (1) . (2) . (3) . (4) . (5) . (6) . (7) . (8) . Poster −0.01 (0.10) −0.00 (0.15) −0.04 (0.13) 0.39 (0.28) 0.02 (0.10) 0.03 (0.06) Poster (four periods) −0.00 (0.11) 0.04 (0.06) Constant 13.18*** (0.23) 15.26*** (0.21) 15.26*** (0.21) 15.52*** (0.11) 14.08*** (0.21) 15.28*** (0.21) 15.28*** (0.19) 15.48*** (0.10) Commodity-period fixed effects NO YES YES YES NO YES YES YES Variety-quarter fixed effects NO YES YES YES NO YES YES YES Interval 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 N 891 891 891 693 891 891 891 693 R2 0.01 0.88 0.88 0.88 Pseudo-R2 0.01 0.95 0.95 0.96 Estimation method OLS OLS OLS OLS PPML PPML PPML PPML . (1) . (2) . (3) . (4) . (5) . (6) . (7) . (8) . Poster −0.01 (0.10) −0.00 (0.15) −0.04 (0.13) 0.39 (0.28) 0.02 (0.10) 0.03 (0.06) Poster (four periods) −0.00 (0.11) 0.04 (0.06) Constant 13.18*** (0.23) 15.26*** (0.21) 15.26*** (0.21) 15.52*** (0.11) 14.08*** (0.21) 15.28*** (0.21) 15.28*** (0.19) 15.48*** (0.10) Commodity-period fixed effects NO YES YES YES NO YES YES YES Variety-quarter fixed effects NO YES YES YES NO YES YES YES Interval 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 N 891 891 891 693 891 891 891 693 R2 0.01 0.88 0.88 0.88 Pseudo-R2 0.01 0.95 0.95 0.96 Estimation method OLS OLS OLS OLS PPML PPML PPML PPML Notes: *** indicates statistical significance at the 1% level. Robust standard errors have been clustered at the variety level and are reported in parentheses. The dependent variable is the value of Britain’s variety imports. A variety is a bilaterally defined commodity. The dependent variable is expressed in natural logarithms in cols. 1–4 and in levels in cols. 5–8. Poster is a binary variable taking a value of 1 beginning in the period after the variety-specific poster was displayed, and continuing for the duration of the interval. Poster (four periods) is a binary variable taking a value of 1 for only four consecutive periods, beginning in the period after the variety-specific poster was displayed. Open in new tab Appendix Table 1 Britain’s variety import values, 1923Q2–1929Q4 . (1) . (2) . (3) . (4) . (5) . (6) . (7) . (8) . Poster −0.01 (0.10) −0.00 (0.15) −0.04 (0.13) 0.39 (0.28) 0.02 (0.10) 0.03 (0.06) Poster (four periods) −0.00 (0.11) 0.04 (0.06) Constant 13.18*** (0.23) 15.26*** (0.21) 15.26*** (0.21) 15.52*** (0.11) 14.08*** (0.21) 15.28*** (0.21) 15.28*** (0.19) 15.48*** (0.10) Commodity-period fixed effects NO YES YES YES NO YES YES YES Variety-quarter fixed effects NO YES YES YES NO YES YES YES Interval 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 N 891 891 891 693 891 891 891 693 R2 0.01 0.88 0.88 0.88 Pseudo-R2 0.01 0.95 0.95 0.96 Estimation method OLS OLS OLS OLS PPML PPML PPML PPML . (1) . (2) . (3) . (4) . (5) . (6) . (7) . (8) . Poster −0.01 (0.10) −0.00 (0.15) −0.04 (0.13) 0.39 (0.28) 0.02 (0.10) 0.03 (0.06) Poster (four periods) −0.00 (0.11) 0.04 (0.06) Constant 13.18*** (0.23) 15.26*** (0.21) 15.26*** (0.21) 15.52*** (0.11) 14.08*** (0.21) 15.28*** (0.21) 15.28*** (0.19) 15.48*** (0.10) Commodity-period fixed effects NO YES YES YES NO YES YES YES Variety-quarter fixed effects NO YES YES YES NO YES YES YES Interval 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 N 891 891 891 693 891 891 891 693 R2 0.01 0.88 0.88 0.88 Pseudo-R2 0.01 0.95 0.95 0.96 Estimation method OLS OLS OLS OLS PPML PPML PPML PPML Notes: *** indicates statistical significance at the 1% level. Robust standard errors have been clustered at the variety level and are reported in parentheses. The dependent variable is the value of Britain’s variety imports. A variety is a bilaterally defined commodity. The dependent variable is expressed in natural logarithms in cols. 1–4 and in levels in cols. 5–8. Poster is a binary variable taking a value of 1 beginning in the period after the variety-specific poster was displayed, and continuing for the duration of the interval. Poster (four periods) is a binary variable taking a value of 1 for only four consecutive periods, beginning in the period after the variety-specific poster was displayed. Open in new tab Appendix Table 2 Britain’s variety import prices, 1923Q2–1929Q4 . (1) . (2) . (3) . (4) . (5) . (6) . (7) . (8) . Poster −0.05 (0.03) 0.03 (0.03) 0.03 (0.03) −0.12 (0.35) 0.01 (0.02) 0.01 (0.02) Poster (four periods) 0.00 (0.03) −0.01 (0.01) Constant 0.48* (0.27) −0.66*** (0.02) −0.66*** (0.02) −0.55*** (0.02) 1.26*** (0.17) −0.66*** (0.02) −0.66*** (0.01) −0.55*** (0.01) Commodity-period fixed effects NO YES YES YES NO YES YES YES Variety-quarter fixed effects NO YES YES YES NO YES YES YES Interval 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 N 891 891 891 693 891 891 891 693 R2 0.00 1.00 1.00 1.00 Pseudo-R2 0.00 0.56 0.56 0.56 Estimation method OLS OLS OLS OLS PPML PPML PPML PPML . (1) . (2) . (3) . (4) . (5) . (6) . (7) . (8) . Poster −0.05 (0.03) 0.03 (0.03) 0.03 (0.03) −0.12 (0.35) 0.01 (0.02) 0.01 (0.02) Poster (four periods) 0.00 (0.03) −0.01 (0.01) Constant 0.48* (0.27) −0.66*** (0.02) −0.66*** (0.02) −0.55*** (0.02) 1.26*** (0.17) −0.66*** (0.02) −0.66*** (0.01) −0.55*** (0.01) Commodity-period fixed effects NO YES YES YES NO YES YES YES Variety-quarter fixed effects NO YES YES YES NO YES YES YES Interval 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 N 891 891 891 693 891 891 891 693 R2 0.00 1.00 1.00 1.00 Pseudo-R2 0.00 0.56 0.56 0.56 Estimation method OLS OLS OLS OLS PPML PPML PPML PPML Notes: * indicates statistical significance at the 10% level and *** at the 1% level. Robust standard errors have been clustered at the variety level and are reported in parentheses. The dependent variable is the average unit value of Britain’s variety imports. A variety is a bilaterally-defined commodity. The dependent variable is expressed in natural logarithms in cols. 1–4 and in levels in cols. 5–8. Poster is a binary variable taking a value of 1 beginning in the period after the variety-specific poster was displayed, and continuing for the duration of the interval. Poster (four periods) is a binary variable taking a value of 1 for only four consecutive periods, beginning in the period after the variety-specific poster was displayed. Open in new tab Appendix Table 2 Britain’s variety import prices, 1923Q2–1929Q4 . (1) . (2) . (3) . (4) . (5) . (6) . (7) . (8) . Poster −0.05 (0.03) 0.03 (0.03) 0.03 (0.03) −0.12 (0.35) 0.01 (0.02) 0.01 (0.02) Poster (four periods) 0.00 (0.03) −0.01 (0.01) Constant 0.48* (0.27) −0.66*** (0.02) −0.66*** (0.02) −0.55*** (0.02) 1.26*** (0.17) −0.66*** (0.02) −0.66*** (0.01) −0.55*** (0.01) Commodity-period fixed effects NO YES YES YES NO YES YES YES Variety-quarter fixed effects NO YES YES YES NO YES YES YES Interval 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 N 891 891 891 693 891 891 891 693 R2 0.00 1.00 1.00 1.00 Pseudo-R2 0.00 0.56 0.56 0.56 Estimation method OLS OLS OLS OLS PPML PPML PPML PPML . (1) . (2) . (3) . (4) . (5) . (6) . (7) . (8) . Poster −0.05 (0.03) 0.03 (0.03) 0.03 (0.03) −0.12 (0.35) 0.01 (0.02) 0.01 (0.02) Poster (four periods) 0.00 (0.03) −0.01 (0.01) Constant 0.48* (0.27) −0.66*** (0.02) −0.66*** (0.02) −0.55*** (0.02) 1.26*** (0.17) −0.66*** (0.02) −0.66*** (0.01) −0.55*** (0.01) Commodity-period fixed effects NO YES YES YES NO YES YES YES Variety-quarter fixed effects NO YES YES YES NO YES YES YES Interval 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1929Q4 1923Q2–1928Q2 N 891 891 891 693 891 891 891 693 R2 0.00 1.00 1.00 1.00 Pseudo-R2 0.00 0.56 0.56 0.56 Estimation method OLS OLS OLS OLS PPML PPML PPML PPML Notes: * indicates statistical significance at the 10% level and *** at the 1% level. Robust standard errors have been clustered at the variety level and are reported in parentheses. The dependent variable is the average unit value of Britain’s variety imports. A variety is a bilaterally-defined commodity. The dependent variable is expressed in natural logarithms in cols. 1–4 and in levels in cols. 5–8. Poster is a binary variable taking a value of 1 beginning in the period after the variety-specific poster was displayed, and continuing for the duration of the interval. Poster (four periods) is a binary variable taking a value of 1 for only four consecutive periods, beginning in the period after the variety-specific poster was displayed. Open in new tab Appendix Table 3 Event study, 1923Q2–1929Q4 Dependent variable: Britain’s variety import volume . Price −2.11*** (0.73) i = −9 0.32 (0.38) i = −8 −0.25 (0.37) i = −7 −0.62 (0.68) i = −6 0.22 (0.19) i = −5 0.39 (0.36) i = −4 −0.44 (0.28) i = −3 0.19 (0.22) i = −2 0.19 (0.18) i = −1 0.01 (0.18) i = 1 (period of poster release) −0.01 (0.18) i = 2 0.07 (0.24) i = 3 −0.05 (0.37) i = 4 −0.10 (0.29) i = 5 0.15 (0.21) i = 6 0.11 (0.22) i = 7 0.00 (0.32) i = 8 −0.14 (0.31) i = 9 0.38 (0.25) Constant 13.85*** (0.54) Commodity-period fixed effects YES Variety-quarter fixed effects YES N 891 R2 0.89 Estimation method OLS Dependent variable: Britain’s variety import volume . Price −2.11*** (0.73) i = −9 0.32 (0.38) i = −8 −0.25 (0.37) i = −7 −0.62 (0.68) i = −6 0.22 (0.19) i = −5 0.39 (0.36) i = −4 −0.44 (0.28) i = −3 0.19 (0.22) i = −2 0.19 (0.18) i = −1 0.01 (0.18) i = 1 (period of poster release) −0.01 (0.18) i = 2 0.07 (0.24) i = 3 −0.05 (0.37) i = 4 −0.10 (0.29) i = 5 0.15 (0.21) i = 6 0.11 (0.22) i = 7 0.00 (0.32) i = 8 −0.14 (0.31) i = 9 0.38 (0.25) Constant 13.85*** (0.54) Commodity-period fixed effects YES Variety-quarter fixed effects YES N 891 R2 0.89 Estimation method OLS Notes: *** indicates statistical significance at the 1% level. Robust standard errors have been clustered at the variety level and are reported in parentheses. The dependent variable is the natural logarithm of the volume of Britain’s variety imports. For the definition of the i variables, see the text. Open in new tab Appendix Table 3 Event study, 1923Q2–1929Q4 Dependent variable: Britain’s variety import volume . Price −2.11*** (0.73) i = −9 0.32 (0.38) i = −8 −0.25 (0.37) i = −7 −0.62 (0.68) i = −6 0.22 (0.19) i = −5 0.39 (0.36) i = −4 −0.44 (0.28) i = −3 0.19 (0.22) i = −2 0.19 (0.18) i = −1 0.01 (0.18) i = 1 (period of poster release) −0.01 (0.18) i = 2 0.07 (0.24) i = 3 −0.05 (0.37) i = 4 −0.10 (0.29) i = 5 0.15 (0.21) i = 6 0.11 (0.22) i = 7 0.00 (0.32) i = 8 −0.14 (0.31) i = 9 0.38 (0.25) Constant 13.85*** (0.54) Commodity-period fixed effects YES Variety-quarter fixed effects YES N 891 R2 0.89 Estimation method OLS Dependent variable: Britain’s variety import volume . Price −2.11*** (0.73) i = −9 0.32 (0.38) i = −8 −0.25 (0.37) i = −7 −0.62 (0.68) i = −6 0.22 (0.19) i = −5 0.39 (0.36) i = −4 −0.44 (0.28) i = −3 0.19 (0.22) i = −2 0.19 (0.18) i = −1 0.01 (0.18) i = 1 (period of poster release) −0.01 (0.18) i = 2 0.07 (0.24) i = 3 −0.05 (0.37) i = 4 −0.10 (0.29) i = 5 0.15 (0.21) i = 6 0.11 (0.22) i = 7 0.00 (0.32) i = 8 −0.14 (0.31) i = 9 0.38 (0.25) Constant 13.85*** (0.54) Commodity-period fixed effects YES Variety-quarter fixed effects YES N 891 R2 0.89 Estimation method OLS Notes: *** indicates statistical significance at the 1% level. Robust standard errors have been clustered at the variety level and are reported in parentheses. The dependent variable is the natural logarithm of the volume of Britain’s variety imports. For the definition of the i variables, see the text. Open in new tab © The Author(s) 2021. Published by Oxford University Press on behalf of the European Historical Economics Society. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. © The Author(s) 2021. Published by Oxford University Press on behalf of the European Historical Economics Society. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
TI - Britain’s Empire Marketing Board and the failure of soft trade policy, 1926–33
JF - European Review of Economic History
DO - 10.1093/ereh/heab005
DA - 2021-11-02
UR - https://www.deepdyve.com/lp/oxford-university-press/britain-s-empire-marketing-board-and-the-failure-of-soft-trade-policy-wcmDV2phms
SP - 780
EP - 805
VL - 25
IS - 4
DP - DeepDyve
ER -