In defense of the domestic market. Japanese imports and textile entrepreneurs in Peru, 1929-19391

FRANCO LOBO COLLANTES

Doctoral candidate, El Colegio de México

[email protected]

Abstract. This article analyzes the impact of the financial crisis of 1929 and large Japanese textile imports on the Peruvian cotton textile sector in the 1930s, and how leading enterprises defended themselves. We start by describing the characteristics of the enterprises that led the textile industry before 1929. We then suggest that the 1929 crisis and the increase in Japanese imports led textile entrepreneurs to mobilize their capacity to influence the government in order to protect their companies, something that had not occurred in previous periods.

Keywords: Textile industry; Peru; imports; Japon; protectionism, global financial crisis, 1929.

Introduction

The enterprise-state relationship has been a source of controversy in Latin American economic history. Approaches that view this relationship positively—arguing that it promotes enterprise growth and, consequently, employment and innovation, successfully engineering institutional change—have been in the minority. In contrast, negative views of this relationship, which conceive it as excessive control of the government by private interests, have predominated. Nevertheless, besides these differing interpretations, the enterprise-state relationship can be clearly observed in the contemporary world. During the 19th and much of the 20th centuries, this relationship was conceived as more direct, since entrepreneurs themselves participated in politics as members of parliament, local officials, or heads of state in some cases. In the late 20th century, this trend diminished and the relationship became more indirect, through mechanisms such as campaign financing, and less visible.

Still, there are cases in which enterprises use more institutionalized mechanisms intended to change certain rules of the game in order to defend their interests. For Douglass North (1993), these modifications should be understood as institutional change; that is, a process that originates in the demands of “entrepreneurs” to change the existing institutional framework, in various spheres, within the context of the costs as they perceive them. The factors that incentivize change can be exogenous (changes in relative or preferential prices) or endogenous (processes of learning or acquisition of skills). Following North, changes in formal rules (laws and regulations that are part of the rules of the game within a government organization) are the consequence of the actions by agents in the political arena. In addition, the author suggests that the direction of this change is influenced by path dependence, a prior trajectory of externalities, learning, and historical developments that reinforce its path. In our case study in historical and entrepreneurial economics, we think that this is a key element in understanding the changes, continuities, and parameters that allows us to grasp the origin and implications of some of the economic measures taken in the course of Peruvian history. Despite the rich neoclassical literature on foreign trade and economic policy, our analysis of the sources consulted for this study leads us to the conclusion that neo-institutional theory provides a more coherent explanation of the subject matter discussed here.

Employing the approach described, the objective of this article is to analyze the actions undertaken by producers of cotton textiles in Peru in order to protect themselves during the 1930s, when Japanese imports entered the Peruvian market in large volumes to compete with locally produced goods. Using sources from a business association, published at that time, as well as legal norms promulgated, we reconstruct the measures designed and implemented to protect this industry, and their possible longer-term effects. We argue that the 1929 crash and the large volume of Japanese imports combined to negatively affect the leading cotton textile enterprises in Peru; that is, the foreign commercial houses. This situation prompted these enterprises to develop institutional mechanisms in an attempt to convince the government to establish import quotas to protect their factories, something that had not occurred in previous decades. However, for the commercial houses, the increase in exports during the second half of the 1920s renewed an interest in foreign trade that predated the crisis (path dependence), which reduced their capacity to lobby for protectionist industrial policies.

1.Entrepreneurs and the cotton textile industry before the 1929 crash

During the export boom in Peru, which lasted from the end of the 19th century to the world crisis in 1929, cotton was one of the country’s principal export products; indeed, by the time the crisis hit the Peruvian economy, it was the second-most exported product (Contreras, 2011). According to Martin Monsalve (2011), given the availability of this raw material, specialization in the textile industry was widespread. In a context of a small internal market and a scarcity of labor, only industries that had access to abundant sources of raw materials could prosper. By taking advantage of these resources as well as by lowering their costs, they were able to export their products. Amid these conditions, in addition to the protection provided by an increase in tariffs and the devaluation of silver coins, the cotton industry led a process of import substitution during the last decade of the 19th century and the beginning of the 20th (Thorp & Bertram, 2013). Thus, of the seven cotton textile factories in Peru in 1905, six were created between the late 1800s and 1903 (Monsalve, 2011).

There were three types of entrepreneurs that entered the cotton textile industry taking advantage of these conditions, illustrating the modes of investment in this sector from the end of the 19th century to the beginning of the 1930s. The first group was made up of local elites, who led the process of economic recovery and growth until the early 1900s (Thorp & Bertram, 2013). The Prado family stood out in this group, and Mariano Ignacio Prado y Ugarteche was in charge of running various family investments. The Prado family began to specialize in the textile industry with the establishment of the Fábrica Nacional de Tejidos de Lana (1892) and the San Jacinto cotton textile factory (1896). In addition, the family had various investments in the Santa Catalina textile and electrical complex. However, despite this specialization, the family’s investment portfolio was diverse, spanning electrical power distribution, railroads, real estate development, and other investments. In addition, investment in banking was key to the family’s economic activities: it was the main stockholder in the Banco Popular and Seguros La Popular, in addition to minority holdings in the Banco Italiano and the Banco de Perú y Londres (Quiroz, 1990). According to Alfonso Quiroz (1990), whose study provides detailed information on the investments of such families, local elites’ emphasis on diversification in the internal market is the consequence of what he calls the “demonstration effect”; that is, learning from and imitating the foreign companies that came to Peru, in order to compete with them and defend their own interests. Thus, as Quiroz indicates, the local elite shaped itself into “economic investment groups,” defined by their broad diversification, for which the banking sector was fundamental in terms of arranging inter-sectoral transfers so that investments could achieve greater economic efficiency. This is how the cotton textile industry was able to obtain the financing needed for expansion at a time when credit for the production sector was scarce due to the limited development of the capital market (Monsalve, 2011).

The second group of entrepreneurs was made up of immigrants, among whom the Italians stand out. In 1908, the city of Lima had a population of 172,000, and more than half of its 6,000 European inhabitants came from Italy. In terms of their importance to industry, Italians owned or had holdings alongside Peruvians in 86 industrial companies producing consumer or intermediate goods, including, in three cases, major stakes in cotton textile factories (Durand, 2004). The most representative case of this type of entrepreneur was Gio Batta Isola Canessa, a native of Genoa who was a merchant marine captain and owner of sailing ships. He transported coal from Liverpool to Callao, where he discovered great business opportunities (Durand, 2004). In 1883, he arrived in Callao with Giacomo Gerbolini, with whom he founded a company specializing in the coal trade. The success of the enterprise was so great that he was able to expand into the internal market, taking advantage of the aforementioned conditions. He established the San Jacinto cotton textile factory, as the majority owner, in partnership with the Prado family. In addition, he became one of the main stockholders of the Banco Italiano and of Seguros Italia, in addition to holdings in companies of various types (Portocarrero, 2014). According to Francisco Durand (2004), the Italian colony’s positioning strategy in the industrial market had three key elements. First, the most important entrepreneurs had prior experience in industrial and commercial fields, or had received specialized education. Second, they took advantage of this knowledge to identify market niches that were little-explored by local businessmen, and positioned themselves in new spaces in the urban market. Third, their sources of financing represented a competitive advantage; while most did not arrive with a fortune in hand, the small businesses they established or the savings they amassed from skilled jobs availed them of the resources necessary to set up small factories. Nevertheless, to sustain these factories over time and to expand, the founding of the Banco Italiano was important since it created sufficient funds to support and consolidate these first industrial initiatives.

The success of investments in the cotton textile sector attracted a third group of entrepreneurs—foreign commercial houses—who by the end of the 1920s, were the largest cotton textile producers in Peru. Foremost among these was the W. R. Grace & Co. from the United States, followed by the British Duncan Fox & Co. Both were characterized by their great capacity for diversification in the most profitable sectors of the economy. W. R. Grace & Co. started to amass its fortune with the acquisition of the Cartavio hacienda in 1882, but did not begin producing cotton for export until 1891. During the same decade, it acquired the Vitarte cotton textile factory and later, in 1903, it inaugurated the Inca textile factory. Similarly, the Duncan Fox & Co. commercial house started its expansion by dominating the cotton export market on the northern coast of Peru, especially in the region of Piura, and later acquired the El Progreso cotton textile factory. In 1918, it established a similar factory, La Union (Durand, 2004; Miller, 2011, 2015). According to Martín Monsalve (2011), the methods used by the commercial houses to control textile enterprises were common at the time: they offered to take care of both the local sales and the import needs of the textile firm; later, when some kind of financial problem occurred, the commercial house bought part of the textile firm, and little by little took control of it. Monsalve explains that W. R. Grace & Co. employed this strategy to acquire the La Victoria cotton textile factory from the Prado family, after the commercial house first took charge of its administration.

The expansion of these commercial houses brought about a “denationalization” of the cotton textile industry: ownership of mills in this sector by national enterprises went down from 71% in 1910 to 44% in 1918, and finally, to 24% in 1931 (Thorp & Bertram, 2013). According to Monsalve (2011), the foreign commercial houses had two competitive advantages. The first was their distribution chain: they took advantage of their experience in the wool and cotton trade, which allowed them to keep in touch with wholesalers and retailers from whom they obtained information about the tastes and preferences of consumers. The second advantage was access to credit, as discussed above. Strong competition created the need to constantly invest large amounts of capital, in turn necessitating financial backing or access to credit. For example, by 1918, the capital invested in cotton textiles was double that invested in wool textiles, and by 1933, it was triple that.

In conclusion, after analyzing the different actors that invested in cotton textiles, we found that their success depended on the following factors: access to raw materials, acquisition of specialized knowledge, access to information about the market, and, most importantly, access to credit under conditions in which an underdeveloped capital market limited the financing of production. Thus, by the end of the 1920s and the beginning of the 1930s, this industrial sector was controlled by enterprises that had gained access to all these requirements; namely, the foreign commercial houses W. R. Grace & Co. and Duncan Fox & Co.. This happened at precisely a time when the textile industry was growing more rapidly than most other sectors of the Peruvian economy. In 1929, textiles accounted for 28% of total industrial production (Seminario, 2015), and cotton textile production accounted for 30% of all the industrial jobs in Peru (Thorp & Bertram, 2013).

2.The 1929 crisis in Peru

Thorp and Bertram (2013) define the impact of the crisis in Peru as profound but brief, given that the fall in exports was reversed rapidly. This initial decline was due to a significant reduction in demand for raw materials from industrialized countries, where the crisis originated. According to estimates by Bruno Seminario (2015), the peak of the crisis was 1932, when GDP fell 19% below 1929 levels, while per capita GDP fell by 23%. These steep declines were the consequence of steep falls in exports, which 33% declined during this period. However, by the mid-1930s, these indicators had returned to pre-crisis levels: GDP in 1934, per capita GDP in 1936, and exports in 1935. In the specific case of cotton, the value of exports dropped precipitously—by 42% in 1931—but then rapidly rallied to pre-crisis levels by 1933. Thanks to this swift recovery (and the sharp drop in oil exports), cotton became Peru’s primary export product starting in 1933 (Contreras, 2011).

In Latin America in general, changes in relative prices and a more active government role in the recovery led to greater economic diversification, which prepared the ground for the regional state-led industrialization project that CEPAL promoted after WWII (Bértola & Ocampo, 2013). Nevertheless, in the view of Thorp and Bertram (2013), the rapid recuperation of Peruvian exports together, with the reduction of public investment (due to the loan obtained by president Augusto Leguía in 1920), limited opportunities for initiatives to search for an alternative model. Thus, while consumer goods as a percentage of total imports fell from 30% in 1928 to 24% in 1935, they stabilized from the middle of the decade, dropping by a single percentage point further by 1939 to 23%. However, according to the data provided by the same authors for cotton textiles, the fall in their share of total imports continued throughout the decade: dropping from 11% in 1928 to 9% in 1935 and, finally, to 5% in 1939. This fall in imports of consumer goods was partially offset by domestic production; in the specific case of cotton textiles, production went from 57% of total consumption in 1928 to 68% in 1932 (Thorp & Bertram, 2013). One analyst at the time described the situation of the textile industry in the 1930s as follows:

The factories are as follows: “Inca,” “ El Progreso,” “Vitarte,” “La Victoria,” “San Jacinto,” “La Unión” and “La Bellota,” in Lima; “La Industrial,” in Arequipa; “Huáscar,” in Cuzco; “Malatesta,” in Ica; and “Manufactura Italiana,” in Sullana [...]. Together they represent more than S/.40,000,000 in capital. They provide jobs for more than 4,000 workers, with an average hourly income of S/.0.51 (fifty-one cents), equivalent to 0.37 Swiss francs, in which those of Italy and Poland are calculated [...]. It has around 4,000 looms and 100,000 spindles [...]. Its annual productive capacity can be estimated as 800,000 yard pieces each, with a value of 1,000,000 soles [...].2 (Desmaison, 1935, pp. 85-86)

This short period of import substitution, in addition to the collapse of international commerce, fostered a more positive attitude toward industry among the commercial houses that dominated textile production. In the period before the crisis, the aforementioned three types of textile industry investors had certain limitations when it came to promoting industrial growth or defending it by way of assuring protectionist policies. According to Alfonso Quiroz (1990), in order to defend its interests and attempt to displace foreign capital in the more profitable areas of the economy, the local elite strengthened their presence in financial entities in order to control intersectoral capital flows. However, to accomplish this, they had to sustain an appreciated currency in order to prevent capital flight. This was achieved by promoting the gold standard, a measure that disincentivized investment in industry given the loss of the effective protection that had been acquired through devaluation in the late 19th and early 20th centuries. For this reason, the author concludes that one of the costs of the local elite’s efforts to achieve autonomy vis-à-vis foreign capital was the limitation of diversification into industry. In the case of Italian immigrants, Francisco Durand (2004) finds they had certain limitations when it came to influencing the elite and the government. From a social perspective, he notes that the Italians, while finding themselves above Peruvian indigenous people and Asian immigrants, were on the lowest rung of the European immigrants, and were thus less integrated into the elite than others such as the English. In Durand’s view, even though these industrialists were generally seen as marginal, the fact that they were white and European, as well as their last names and level of education, gave them access to the upper middle class, which allowed them upward mobility albeit short of entry into the elite.

Finally, in the case of the commercial houses that dominated cotton textile production by the 1920s, Martín Monsalve (2011) argues that it was impossible to expect them to pressure bankers or congress to make changes to foreign exchange or tariff policies in favor of their industry. This is because these companies had also invested in importing manufactured products and exporting raw materials, and therefore such policies would be more prejudicial than beneficial. According to Monsalve, for these companies, investment in cotton textile factories was not part of a process of vertical integration but simply diversification of investments in order to control the internal and export markets. However, with the crisis of 1929, this logic was altered, with a reduction in foreign trade and an increase in textile production through import substitution. In this new context, the commercial houses saw an improvement in their prospects of convincing the government to defend their interests against any threat. While no evidence has been found of lobbying by the entrepreneurial sector as a collective, in the case of the cotton textile industry the threat came from Asia, with the so-called “Japanese avalanche.”

3.Importation of Japanese textiles

The importation of Japanese cotton in the 1930s was achieved due to the Japanese community’s establishment in the cotton trade starting in the previous decade. The arrival of the first Japanese immigrants in the late 19th century was the result of migratory policies designed by the Japanese Empire in response to demographic pressures stemming from the large and inactive rural population precipitated by the ongoing industrial boom. Through an agreement with the Peruvian government, 18,000 Japanese immigrants (15,800 men, 2,100 women, and a few more than 200 children) arrived in the country between 1899 and 1924, transported by Japanese shipping companies. During the five years after this so-called “contract immigration” came to an end, 7,000 more Japanese came to Peru to join their families (Lausent-Herrera, 1991). In 1910, 1,000 Japanese lived in Lima, the port of Callao, and surrounding areas; by 1920, the community had grown to 4,622, displacing the Chinese as the largest Asian group; finally, in 1936, the Japanese community numbered 22,000 immigrants (Lausent-Herrera, 1991).

Until 1920, the majority of Japanese immigrants came to work as peons on cotton haciendas on the Peruvian coast, later accumulating a little capital and becoming landowners or tenants of larger properties. This transformation would not have been possible without the capital provided by the tanamoshi, a system of mutual aid established by Japanese merchants to help one another. This was an independent source of financing that gave the community an important competitive advantage (Lausent-Herrera, 1991).

The Japanese success story started in the 1920s. They achieved urban hegemony among stores selling household items, and as barbers and hairdressers. They also excelled in the provision of coal, in bakeries, and in bazaars, where they sold textiles from their home country.3 In addition to urban commerce, the Japanese community acquired a major presence in cotton production in the 1930s. In 1934, they controlled more than 15% of that market in Peru, and as much as 50% of the cotton harvest in the valleys near Lima, where they were most prominent (Industria Peruana, 1934a). For Isabelle Lausent-Herrera (1991), the Japanese strategy consisted of intensive use of publicity, generating special offers and stock clearances. This price reduction policy pushed the competition to do the same, but the low-cost financial support of the tanamoshi gave the Japanese the ability to maintain low prices and gain advantage over other merchants who only had access to credit at very high interest rates. In addition to the distribution chain and access to credit at low interest, another advantage of Japanese textile imports lay in the design of the products, given that “[...] even when they are less durable (because, especially, of a raw material that is less strong and select), they have, on the other hand, an attractive design” (Desmaison, 1935, p. xxx).

Because of the success of Japanese businesses there, in 1936-1937 Peru accounted for the fifth-highest volume of remittances sent to this Asian country, only surpassed by the United States, Manchukuo, Brazil, and Canada (Lausent-Herrera, 1991). In addition, the trade between the two countries grew exponentially during the 1930s, though it should be noted that Peru operated at a deficit. Isabelle Lausent-Herrera (1991) provides evidence of this for the period 1935-1939, but we think it is more illustrative to begin observing this phenomenon before the impact of the 1929 crisis, and examine how it grew during the years of the depression. Although the prices are in current soles and the increases may seem excessive, Table 1 demonstrates the constant deficits run by Peru, and that imports of textiles and other items from Japan were constantly higher than exports of cotton from Peru.

There are several developments worth noting. First, the large increase in the deficit starting in 1933 coincides, as we will see later, with actions taken by local producers to defend themselves against the arrival of Japanese imports. Second, the insignificance of these imports for Peruvian trade as a whole, since in 1934 trade with Japan represented only 3.2% of the total. Thus, the problem that developed and the actions taken should be limited to the negative effects on the cotton textile industry, especially since an large percentage of Japanese imports were cotton textiles, as Table 2 illustrates.

Table 1
Trade balance between Peru and Japan, 1929-1934 (in thousands of current soles)

Year	Exports	Imports	Deficit	% of total trade
1929	99.5	2,356.1	2,256.4	0.5
1930	2.7	2,361.7	2,358.9	0.6
1931	21.4	1,654.9	1,633.4	0.5
1932	37.3	1,259.7	1,222.3	0.5
1933	999.9	5,277.3	4,278.1	1.7
1934	5,084.5	10,225.6	5,414.1	3.2

Source: compiled by author from data in Desmaison (1935).

Table 2
Imports of Japanese cotton textiles by type, 1929-1934
(in thousands of current soles)

Year	Yarn	Fabric	Finished textile products	Total (a)	% of Japanese cotton textiles in total imports from Japan
1929	1.3	481	600.5	1,083	45.9
1930	4.3	593	622.2	1,219.6	51.6
1931	2.1	398.5	361.9	762.6	46.1
1932	4.4	311.1	371.8	687.4	54.5
1933	5.7	685.9	1,054.2	1,745.8	33.1
1934	37.2	3,532.2	2,864.1	6,433	62.9

Source: compiled by author from data in Desmaison (1935).

Yarn imports from Japan were very small, but fabric and clothing imports from that country expanded almost in tandem, adversely affecting Peruvian producers of fabrics and finished textile products. In addition, textiles accounted for a significant share of total Japanese imports, ranging between 45% and 55% from 1929 to 1932, falling sharply to 33% in 1932, and soaring to 62% in 1934 to the consternation of local producers. This development is correlated with the fall in the share of locally produced cotton textiles in total consumption, which went from 68% in 1932 to 55% in 1934 (Thorp & Bertram, 2013). This is also evident in the reduction in orders by local cotton textiles enterprises, which fell by 17% between the first quarter of 1932 and the same period in 1935 (Desmaison, 1935).

As we can see from this analysis, there was evident damage to the Peruvian cotton textile industry, and leading entrepreneurs began to express concerns that resonated with the government.

4.First reactions

In 1934, when the toll taken by Japanese imports on local industry was already evident, the Sociedad Nacional de Industrias (SNI), the association that represented the industrial sector, published a detailed article giving its perspective on the situation. Under the provocative title “Japanese Infiltration,” the text refers to the harm caused by the arrival of these imports and the total opposition to their sale on the local market. The SNI highlighted the problems that these imports created for the economy, and the danger posed by Japan’s competence:

Peru has no problem which is more serious, more pressing, or which has a more profound effect on its constitution, its development or its economic independence than Japanese infiltration, cleverly prepared by one of the most farsighted nations on earth, methodically dosed so as to avoid brusque or disruptive resistance, and benefiting from the lack of awareness of the danger that, for a long time, private interests and public powers betrayed. (Industria Peruana, 1934a, p. 405)

This warning was a reaction to the prominence that this highly active community acquired in Lima, the Japanese merchants competing with Peruvian bodega owners, as well as the increasing number of Japanese peons on large cotton-producing landholdings, having triggered strong feelings of repudiation (Lausent-Herrera, 1991). The SNI’s strong opposition to proposals to establish a Japanese bank in Lima should be understood in this context:

A Japanese bank would protect its clientele, [the] Japanese, from restrictions on the amount of available credit; it would also mean that the people it protected would be free of all dependence on the economic organization of the country, and it would be equivalent, with a plaza under siege, to establishing an arsenal to supply the occupiers. (Industria Peruana, 1934b)

Finally, the SNI stressed the importance of relationships with European immigrants for the development of Peru, which could not be compared with the damaging relationship with Japan:

In other times, Peru has experienced the hegemony of various European influences. Until the end of the 19th century, the commercial hegemony of the English, without displacements or parasitical absorptions, promoted national industry, commerce and agriculture, so much so that one could say that its progress was almost exclusively due to support from British banking and commerce. We have also experienced Yanqui hegemony, which was at once a creator and exploiter of riches that were previously unproductive or only partially exploited. Before Japanese immigration, the Italian colony in Peru was the most numerous and economically strong; but we think it makes no sense to compare the immediate assimilation to the point of indistinguishability, in all aspects of national life, between Peruvians and Italians permanently settled in our country, and Japanese impermeability, the fruit of unbridgeable determinism. (Industria Peruana, 1934b)

The “impermeability” referred to in this quote, a characteristic of the secretiveness of Japanese society, was also rejected by the rest of the population. The assumption behind the request for Japanese workers to come to Peru was that they would return to their country with a little capital, or that, in any case, they would be integrated into the rural world of Peru. However, the fact that they stayed in the capital and were resistant to Peruvian influences, in addition to their attachment to Japan, resulted in their greater marginalization. According to Isabelle Lausent-Herrera (1991), the economic monopolization and the secretiveness of the Japanese community led to racist feelings against this Asian group of the type that already existed against the Chinese community. This repudiation manifested itself in a climate of violence in those Lima neighborhoods where large numbers of Japanese lived. In addition, these developments were not limited to popular sectors. According Tirso Molinari (2006), who analyzes the Unión Revolucionaria, a party that was hegemonic during the 1930s, the government itself propagated racist and xenophobic attitudes toward Asian migrants.

This factor could be important in understanding the government’s rapid acceptance of requests from local cotton textile entrepreneurs, given that it provided legitimacy to the project of protecting the internal market.

5.From words to actions: enacting institutional change

As mentioned in the first section, the foreign commercial houses, W. R. Grace & Co. and Duncan Fox & Co., dominated Peruvian cotton textile production; thus, they were the first to be affected by Japanese imports. It was the British commercial house, Duncan Fox—owner of the El Progreso and La Union factories—that led the defense against Japanese competition.4 Its leadership was not accidental, since it was one of the commercial houses that was least affected by the 1929 crisis due to its high level of economic diversification (Miller, 2015). The method that this enterprise chose to initiate its efforts was to channel the problem through the Sociedad Nacional de Industrias, which had already spoken out about the damage caused to local industry. A key actor in this process was George Bertie, Duncan Fox’s representative in the SNI and president of its Textile Committee, since he was already the visible link between the company and the SNI, and the main proponent of the former’s requests for the defense of its interests.

Bertie took two types of actions. First, he asked for the abrogation of the existing commercial treaty with Japan and, second, requested an increase in tariffs on imported cotton. The first had to do with the Treaty of Friendship, Commerce, and Navigation with the Japanese Empire, signed in 1924 and ratified in 1930. On August 24, 1934, Duncan Fox, through the SNI and represented by George Bertie, officially asked the Peruvian government for the treaty to be annulled. Although there is no available information about the negotiations that followed, the government apparently agreed, given that in October of the same year, the Ministry of Foreign Relations announced the abrogation of the treaty—the only stipulation being that it would cease to be in force one year from then, that is, in October 1935 (Desmaison, 1935; Industria Peruana, 1934c).

At the same time, work continued on the second objective; and in July 1934 George Bertie, acting through the SNI, sent a second request to the government:

Since this is about the principal national manufacturing industry, which processes raw materials produced in the country and provides jobs for many thousands of employees and workers, I have not hesitated in contacting your office to request protectionist measures that effectively counter the invasion of competitors, [a process] which is supported by the extremely low wages in Japan, the depreciation of the currency of that country and the decisive cooperation that is provided by your government [...]. The most immediate measure would be to increase tariffs, raising in adequate proportion headings nos. 25, 26, 30, 32, 39, 40, 51 and 52, which are the ones that are at the lowest limits in relation to the quality and quantity of the articles included. (Industria Peruana, 1934d)

Since this was a request that could have great implications, to the extent that changes in tariffs could affect other interests, the government set up a commission to evaluate the matter within a month of receiving the request. This commission was made up of representatives of the relevant governmental bodies, such as the Ministry of Finance, the Ministry of Development, and the customs authority, as well as the business associations of those affected, including the Cámara de Comercio de Lima and the SNI. The fact that there was no representative to defend Japanese interests, and that the representative of the SNI was George Bertie himself, speaks to the intentions of the commission (Desmaison, 1935).

Finally, on September 18 of the same year, the commission prepared a report that was favorable to the industrialists’ request. A new tariff was prepared through the recently created Consultative Body on Tariffs (Cuerpo Consultivo de Aranceles). The commission had agreed with the rates proposed for the section on cotton—the only change made was to also apply these rates to printed textiles. However, the difficulties that arose did not pertain to the measures themselves but rather to the time it took to implement them (Industria Peruana, 1934e). Because the first measure, the abrogation of the commercial treaty with Japan, would take a year to go into effect, the commission hoped that the pertinent tariffs could be applied without waiting for the approval of a complete set of tariff rates. This, however, was unacceptable: not because of a lack of will by the government but because of international factors. During those same months, commercial agreements were being negotiated with various countries including Great Britain, Chile, France, and others, in the context of the bilateralism that predominated in the 1930s. Thus, any unexpected modification of tariffs could jeopardize these negotiations (Desmaison, 1935).

Faced with these uncertainties, an alternative solution was sought. On November 29, 1934, George Bertie, in his role as president of Textile Committee, convened a meeting of the SNI to which he invited fabric producers, representatives of the most important manufacturing industries, and some intellectuals specialized in the sector. Bertie described the objective of the meeting in this way:

[...] presenting for the consideration of those in attendance a measure that, at their discretion and with urgency, should be proposed to the Supreme Government, in defense of our textile industry which is seriously affected by Japanese competition. (Industria Peruana, 1934f)

The proposal involved recommending to the government a system of import quotas that was not only beneficial to local textile producers, but also to the overall trade balance and to improving the relationship with the country’s most important trade partners. Alejandro Desmaison, who describes this process in great detail, summarizes the gist of the proposal as follows:

Each country that sells cotton products to Peru will only be able to do so in a quantity proportional to the cotton (raw material) that they bought from us during the penultimate year. The choice of this year was made on the basis of the date of publication of the “Anuario del Comercio Exterior del Perú.” Reasonable compensation is sought: those who sell us cotton fabrics have to buy raw materials from us. (Desmaison, 1935, p. 92)

A fundamental part of the argument presented by Duncan Fox’s representative regarding a favorable trade balance was that Japanese textiles were not only displacing local producers, but had begun to gain ground on English textiles, which supplied a different market than local producers, and for this reason their coexistence was not conflictive. This issue became especially serious in the context of bilateral relations, since after the Ottawa Conference (British Empire Economic Conference), Great Britain negotiated trade agreements with other countries based on “I buy from those who buy from me.” Given that Great Britain accounted for 27% of Peruvian trade in 1934, nine times more than Japan, adopting this measure was given priority (Desmaison, 1935). In addition, according to Bill Albert (1982), 1934–1935 bilateral negotiations for a commercial agreement between Great Britain and Peru discussed the quota of sugar that Peru could export to the British market so long as British textiles did not lose market share because of Japanese expansion; this indicates that the quota system was widely accepted and promoted by British merchants. According to Albert, the proposal for a system of quotas emerged from these negotiations; English representatives in Lima had recommended that Duncan Fox & Co. put forward the system as its own proposal rather than as a recommendation of these representatives. However, we think that this cannot explain the active participation of the commercial house in the previous actions, during which bilateral conversations were not discussed. In addition, as Rory Miller (2015) indicates, Duncan Fox had no interest in sugar and thus the level of its involvement in the commercial agreement with Great Britain was not a priority. Thus, in our opinion, the process of negotiating a trade agreement created a climate that favored quotas, but these did not specifically result from this context.

This argument includes an additional element: since Duncan Fox was an English commercial house, and given how important cotton and sugar exports to Great Britain and importation of English products was to them, import quotas benefited them in all possible ways.

As a result of the consensus reached within the industry association, George Bertie, through the SNI, sent the proposal to the government. It reacted positively and on February 28, 1935, created a commission to analyze the request. Just like the commission established to analyze the tariffs, this one was made up of various representatives of the parties involved and, in the case of the SNI, by Augusto Maurer, the president of the association, demonstrating its level of commitment to the matter (Desmaison, 1935).

The commission worked during March and April to prepare their final report to the government, which, given that the congress had been closed down, issued a decree on May 10, 1935, establishing a regime of import quotas on manufactured cotton products. In the justifications included in the decree, the emphasis is not only on the economic effects of the crisis of local industry, but also on the social effects:

[...] the acute crisis resulting from importation on a vast scale of foreign fabrics [...] invading the national market [and] displacing their Peruvian counterparts [...]. That, if these importations continue to increase this industry would be seriously damaged, since it would result in the complete paralysis of locally-owned factories and the unemployment of numerous personnel they employ, seriously affecting the national economy. (Industria Peruana, 1935a)

The quotas were established on the basis of the conditions that existed in 1929; it was argued that this was a normal year, before the international crisis started affecting economic performance (Industria Peruana, 1935a). However, we agree with Isabelle Lausent-Herrera (1991) that this year was chosen in order to reduce trade with Japan and to benefit Europeans.

The quota system is illustrated in Table 3. We can see that the quota assigned to Japan was among the highest, although below those assigned to Great Britain, the United States, and Italy, providing evidence of the importance of Peruvian cotton exports to the Asian country in 1929, in keeping with the prominence of the Japanese colony on cotton plantations. Also evident is the predominance of Great Britain, with a quota that was much higher than the rest of the countries, and almost double that of the United States. Finally, it should be noted that there was little trade in cotton with other Latin American countries, which—in comparison with Europe and Asia—accounted for an insignificant percentage of this trade.

Table 3
System of quotas for the importation of cotton manufactured products, by countries in 1934

Country	Gross kilograms	Country	Gross kilograms	Country	Gross kilograms
Great Britain	845,387	Belgium	154,545	Brazil	3,590
United States	476,804	Spain	68,472	Chile	241
Italy	448,190	France	51,471	Panama	158
Japan	204,238	Hong Kong	20,390	China	66
Germany	176,680	The Netherlands	7,974	Colombia	3

Source: Industria Peruana (1935a).

The decree establishing the quotas stipulated that they would be in effect for six months, and charged the General Superintendency of Customs (Superintendencia General de Aduanas) with supervising the implementation of the quota system. To achieve this, new decrees created a register of current accounts by countries and by importers in order to control manufacturers. In addition, Peruvian consuls abroad opened special legers to keep track of consular invoices intended to document manufactured cotton products (Industrias Peruana, 1935b). Finally, the decree covering the system of quotas prohibited price increases by national fabric producers while the regime was in place, under the penalty of its abolishment (Industria Peruana, 1935a).

The reconstruction of the developments in this section clearly shows how, through the maneuvers of George Bertie and utilization of the industrial association, Duncan Fox & Co. was able to defend its interests through government measures that originated as a result of its own initiatives. This constituted a change in the institutional framework that resulted from external factors—unexpected competition that altered the equilibrium existing in Peru until the mid-1930s, creating incentives to deploy the negotiating capacity of the affected industries. These types of victories could have created an national industrial project promoted by the private sector. However, changes in the external market created a different attitude toward industrial development due to path dependence.

6.Unexpected flexibility: path dependence

After observing how Duncan Fox & Co. worked with great persistence to implement measures that would benefit it in different ways, even though not all of its attempts were successful, one would have thought that these measures would last in the medium-term. However, this turned out not to be the case. When the six-month quota regime expired at the end of November, it was prolonged through the end of December while a new decree was prepared; this was promulgated on December 30 and included significant changes (Industria Peruana, 1936a).

First, it eliminated the system of quotas for countries and focused exclusively on Japanese imports. For the Sociedad Nacional de Industrias, the only factor that justified the system of quotas for all countries was the existence of the Peruvian-Japanese commercial treaty; however, after this expired, in October 1935, it was possible to focus directly on Japan. In addition, as the SNI commented (and as noted above), the Peruvian textile industry was never affected by the textile industries of the United States and Europe, and thus there was no reason to maintain a system of quotas for all countries when the situation had changed:

It is necessary to clarify, once more, that the local cotton textile industry was never affected, nor is it now, by imports from Europe and the United States. Rather, these have provided stimulus and incentives to achieve the improvements that we observe today with uttermost complaisance. (Industria Peruana, 1936b)

Nevertheless, we think that the second change was more significant. The new quota set exclusively for imported Japanese textiles was 612,714 kilograms, which was 300% greater than the previous quota. According to Isabelle Lausent-Herrera (1991), the way in which Japan defended its interests in Peru when faced with the first quota regime was to ensure that 70% of textiles sent to Peru went into Japanese hands, and the rest to Peruvian merchants. Lausent-Herrara argues that this favored the Japanese colony because it allowed it to continue setting low sale prices, which continued to negatively affect local producers. It was because of this situation that the quota was increased under the second system. Without rejecting this argument, we believe that this flexibility in relation to Japanese imports was related to the new attitude adopted by Duncan Fox in the context of the new international conditions. By 1936-1937, the value of cotton and sugar exports had increased considerably to reach their highest level of that decade (Contreras, 2014). This growth prompted Duncan Fox to pay more attention to the external market than to the internal market, as had been the case in the decades before the crisis of 1929, and so the company’s loss of interest in local production was conspicuous. In mid-1937, another import quota on Japanese textiles was imposed, this time at 850,000 kilograms, which once again increased Peru’s capacity to import from this Asian country (Industria Peruana, 1936c). In addition, after an energetic campaign through its publication Industria Peruana, from 1936 onward the SNI stopped publishing articles and commentaries on this subject. Information on the new 1937 quota was published in a small note given secondary importance. In addition, articles by George Bertie and the type of activism in which he engaged—frequently published between 1934 and 1935—ended in 1936.

In summary and based on the information presented in this section, the logic that Monsalve (2011) describes as operating in previous decades—in which the commercial houses invested in industry more to diversify their actions as a way of controlling the internal market than to achieve vertical integration—continued to operate in the 1930s, which greatly reduced the scope of import quotas and cannot be conceived of as an industrial policy. The path dependence of commercial houses led to a return to their previous strategies in relation to foreign commerce, which had been very profitable, and therefore industrial protectionism became a secondary concern.

7.Conclusions

The case of cotton textile entrepreneurs and the mechanisms they employed to defend themselves from foreign competition demonstrates how they were able to include a relationship with the government among their strategies. This sheds new light on the state-entrepreneur relationship and on the degree to which private logics can be imposed on the public sphere. From this perspective, we propose some conclusions based on the case discussed here, employing a neo-institutional perspective.

First, during the first three decades of the 20th century in Peru, entrepreneurs from the local elite, the immigrant population, and foreign commercial houses used various business strategies that—by the 1920s—ended up positioning commercial houses as the dominant entrepreneurial group in the cotton industry. However, this was conditioned by their secondary role in investments in this type of enterprise, given—as their very name indicates—the commercial houses focused primarily on the export and import of a range of products. This limited their leadership role among the dominant groups in the sector, something that we think is key to understanding later decisions that were influenced by path dependence.

Second, the impact of the crisis of 1929, despite its short duration, opened a small window to promote activities linked to the internal market, since the fall in foreign trade led the commercial houses to assume a greater interest in activities such as industry in order to offset their losses in other endeavors. However, as Douglass North (1993) points out, although a variety of factors can incentivize institutional change, such change is conditioned by costs. Therefore, if alterations resulting from external disruptions do not offset the cost of lobbying, institutional change will not take place, even when the situation of entrepreneurs is not optimal. In this sense, the upheavals in Peru resulting from the 1929 crisis were insufficient to lead to the kind of entrepreneurial lobbying that would be required to promote industrialization.

Third, while the 1929 crisis created an incentive for commercial houses to increase their investments in the internal market, the sharp increase in Japanese imports was what led to the decision to defend their share of the internal market through institutional change. Although trade with Japan represented only an insignificant percentage of total Peruvian foreign trade, the majority of imports from that country were cotton textiles, which, given their low prices and the distribution chains created by the immigrant community, posed a significant threat to local production controlled by the commercial houses.

Fourth, the rejection by local industrialists of Japanese products had a racial component that reflected the attitudes of the Peruvian population toward this immigrant community. Both the increase in the Japanese colony’s participation in the urban and rural economies, as well as this community’s insularity and lack of integration, led to non-acceptance by various sectors of society. This was a key variable in legitimating the demands of local producers which allowed for consensus to be reached more easily.

Fifth, the measures were proposed by the British company Duncan Fox & Co., one of the two commercial houses with the largest share of local cotton textile production. Through the efforts of George Bertie, Duncan Fox’s representative in the Sociedad Nacional de Industrias, this association of industrialists presented a proposal to the government that led to the establishment of a system of quotas to limit Japanese imports and benefited those from Europe, especially from Great Britain, consolidating that country’s interests in local industry and foreign trade.

Sixth, and finally, the flexibility of Duncan Fox and the Sociedad Nacional de Industrias (SNI) toward Japanese imports so soon after the implementation of import quotas is explained by the revival of foreign trade—especially in agricultural exports, in which Duncan Fox had investments. Thus, the company turned its attention to these types of activities and de-emphasized protection of the local cotton industry. This reversal is explained by path dependence since the commercial house’s main connection with the Peruvian economy was through foreign trade, because they were well-acquainted with this sector and had comparative advantages within it. Thus investment in industry was a secondary component in their strategies.

As a result of these conclusions, we think that it is crucial for historiography to pay more attention to the entrepreneur-state relationship from a perspective that articulates the economy with politics (such as the neo-institutional approach), by carrying out archival work that makes it possible to reconstitute processes and thereby reach conclusions that enable the analysis of this link and or better understanding of its efficiency. In this way, we can better understand the influence of such links in the development of Latin American economies, the character of governance in Latin America, and the process of legitimation of elites.

References

Albert, B. (1982). Sugar and Anglo-Peruvian Trade Negotiations in the 1930s. Journal of Latin American Studies, 14(1), 121-142. Cambridge: Cambridge University Press.

Bértola, L., & Ocampo, J. (2013). El desarrollo económico de América Latina desde la independencia. Mexico City: Fondo de Cultura Económica.

Contreras, C. (Ed.). (2011). Compendio de historia económica del Perú (vol. IV: Economía de la primera centuria independiente: anexos). Lima: Banco Central de la Reserva del Perú, Instituto de Estudios Peruanos.

Contreras, C. (Ed.). (2014). Compendio de historia económica del Perú (vol. V: La economía peruana entre la gran depresión y el reformismo militar). Lima: Banco Central de Reserva del Perú, Instituto de Estudios Peruanos.

Desmaison, A. (1935). El sistema de contingentes en el comercio internacional. Lima.

Durand, F. (2004). El poder incierto. Trayectoria económica y política del empresario peruano. Lima: Congreso de la República del Perú.

Industria Peruana. (1934a). La infiltración japonesa. Industria Peruana, p. 405. Sociedad Nacional de Industrias.

Industria Peruana. (1934b). Banco japonés. Industria Peruana, p. 408. Sociedad Nacional de Industrias.

Industria Peruana. (1934c). Tratado de Amistad, Comercio y Navegación. Industria Peruana, p. 454. Sociedad Nacional de Industrias.

Industria Peruana. (1934d). Petición del comité textil. Industria Peruana, pp. 495-6. Sociedad Nacional de Industrias.

Industria Peruana. (1934e). Cuerpo Consultivo de Aranceles. Industria Peruana, p. 510. Sociedad Nacional de Industrias.

Industria Peruana. (1934f). Comité textil. Industria Peruana, p. 550. Sociedad Nacional de Industrias.

Industria Peruana. (1935a). Decreto de 10 de mayo de 1935. Industria Peruana, p. 188.

Industria Peruana. (1935b). Superintendencia General de Aduanas. Industria Peruana, p. 235. Sociedad Nacional de Industrias.

Industria Peruana. (1936a). Cuotas de importación. Industria Peruana, p. 1. Sociedad Nacional de Industrias.

Industria Peruana. (1936b). La industria textil. Industria Peruana, p. 7. Sociedad Nacional de Industrias.

Industria Peruana. (1936c). Cuotas de importación. Industria Peruana, p. 42. Sociedad Nacional de Industrias.

Lausent-Herrera, I. (1991). Pasado y presente de la comunidad japonesa en el Perú. Lima: Instituto Francés de Estudios Andinos.

Miller, R. (2011). Empresas británicas. Economía y política en el Perú, 1850-1914. Lima: Banco Central de Reserva del Perú, Instituto de Estudios Peruanos.

Miller, R. (2015). The British Commercial Houses in Peru and Chile between the Two World Wars: Success and Failure. Estudios de Economía, 42(2), 93-119. Santiago de Chile: Universidad de Chile.

Molinari, T. (2006), El fascismo en el Perú: La Unión Revolucionaria, 1931-1936. Lima: Universidad Nacional Mayor de San Marcos, Fondo Editorial de la Facultad de Ciencias Sociales.

Monsalve, M. (2011). Industria y mercado interno, 1821-1930. In C. Contreras (Ed.), Compendio de historia económica del Perú (tomo V: Economía de la primera centuria independiente). Lima: Banco Central de Reserva del Perú, Instituto de Estudios Peruanos.

North, D. (1993). Instituciones, cambio institucional y desempeño económico. Mexico City: Fondo de Cultura Económica.

Portocarrero, F. (2014). Grandes fortunas en el Perú: 1916-1960. Lima: Universidad del Pacífico.

Quiroz, A. (1990). Banqueros en conflicto: estructura financiera y económica peruana, 1884-1930. Lima: Centro de Investigaciones de la Universidad del Pacífico.

Seminario, B. (2015). La economía peruana en la era moderna. Lima: Universidad del Pacífico.

Thorp, R., & Bertram, G. (2013). Perú, 1890-1977: crecimiento y políticas en una economía abierta. Lima: Universidad del Pacífico.

---