Output Growth in the Baltic Countries in 1913–1938: New Estimates

Abstract

Assessing the economic performance success of restorations by OIST, it is possible to draw on the MPD (2020) data for the 1973–2018 period. However, MPD does not provide data for the years 1913–1938. This does not mean that there is no output data on the Baltic countries during this period. During the interwar years, Estonian and Latvian economists calculated national income for many years, almost completely covering the second decade of independence. There are three such estimates (for 1924, 1938, 1939) for Lithuania as well. However, they are not strictly comparable due to the lack of standard calculation methodology. Some were used in the first attempt at a cross-time comparable measurement of total output of the Baltic countries in 1913–1928, which was published in the third (and last) edition of the survey The Economic Forces of the World, released by Dresdner Bank (1930), and remains a broadly used source in historical statistics. In this publication, estimates of national income in national currencies were converted to German Reichsmark (RM) at gold parity.

Assessing the economic performance success of restorations by OIST, it is possible to draw on the MPD (2020) data for the 1973–2018 period. However, MPD does not provide data for the years 1913–1938. This does not mean that there is no output data on the Baltic countries during this period. During the interwar years, Estonian and Latvian economists calculated national income for many years, almost completely covering the second decade of independence. There are three such estimates (for 1924, 1938, 1939) for Lithuania as well. However, they are not strictly comparable due to the lack of standard calculation methodology. Some were used in the first attempt at a cross-time comparable measurement of total output of the Baltic countries in 1913–1928, which was published in the third (and last) edition of the survey The Economic Forces of the World, released by Dresdner Bank (1930), and remains a broadly used source in historical statistics. In this publication, estimates of national income in national currencies were converted to German Reichsmark (RM) at gold parity.

Colin Clark is deservedly celebrated for starting a new era in cross-country and cross-time comparison of total and per capita output 80 years ago (in Clark 1938), when he published his landmark comparison of the national income of 28 countries on a purchasing power parity (PPP) basis. Clark used 1929 as his benchmark year and the United States as the numeraire country. Therefore, following contemporary usage, we will designate his monetary measurement unit (which he called the ‘international unit’ or IU) as ‘1929 international $’ (int$) or ‘1929 international Clark $’ (C$ 1929). In Clark’s famous ranking, the Baltic countries share company with poorer countries, and Lithuania with 80 C$ 1929 per capita closes the list (Table 8.1).

Table 8.1 National income per capita (in int$ 1929) of 27 countries in 1925–1934. Data source: Clark (1938: 66). Clark himself provides only the national income value per head of occupied population. National income per capita calculated by the author, dividing Clark’s 1925–1934 national income annual mean totals by populations as of 1930, provided by Maddison (2008). For reader’s convenience, the names of Baltic countries are in bold (also in other tables in this chapter)

Clark’s estimates refer to a 1925–1934 national income mean value and do not help us compare the growth performance of the included countries. Joan R. Roses and Nikolaus Wolf (2010: 187–190) published estimates of the GDP (in GK$ 1990) for 1922, 1929 and 1938 for many European countries in The Cambridge Economic History of Modern Europe. Table 8.2 provides a ranking of most European countries by GDPpc according to Maddison (2008), expanded by estimates from Roses and Wolf (2010: 190) for Estonia and Latvia in 1922–1938 and Poland in 1922.¹ They provide a brighter picture of the international standing of Estonia and Latvia after the first decade of independence (in 1929) than Clark did in 1938 (to recall, his estimate refers only to a 1925–1934 annual mean).

Table 8.2 European country ranking by GDPpc in 1922–1938 according to Roses and Wolf (2010: 190). GDPpc values (except for Estonia, Latvia and Poland) are from Maddison (2008)

According to Clark’s ranking (see Table 8.1), these countries markedly outperformed only the South-Eastern (Balkan) European countries, standing nearly at the same level with Poland and other South European countries (except for Greece). According to the Roses and Wolf (2010) ranking, the distance separating Estonia and Latvia from South-Eastern Europe was greater, with their GDPpc in GK$ 1990 outpacing that of Bulgaria and Romania two and more times. They outperformed Central Europe (Hungary and Poland) and South European countries, except for Italy, and were on an equal level with Ireland.

Importantly, Estonia and Latvia in their view displayed strong positive growth dynamics not only in comparison with the 1922 level but also in terms of relative ascendancy or improved ranking position; Latvia rose from 18th position in 1922 to 15th in 1929 and to the tenth by 1938, trailing just behind Norway, France and Sweden. Estonia fell from 13th position to 14th in 1922–1929 and then rose to 11th place by 1938. Estonia and Latvia (except for 1922) were also richer than Finland, while Clark (1938) presented the opposite view.

For Estonia, Roses and Wolf (2010) provide a brighter picture than the estimates by Estonian historian Jaak Valge (2003). In 1932, Juhan Janusson published calculations of the Estonian national income at current prices in the national currency (Estonian crown; EEK) and its structure in 1928–1930 (Janusson 1932). He later revised and extended this pioneering work, covering the 1929–1936 period (Janusson 1937). Arvo Horm (1940) extended this time series even further by 2 more years (1937–1938). After a long break, Jaak Valge (2003) extended the time series covered in interwar publications backwards, including also the 1923–1928 period.

Janusson’s calculation is grounded on the assumption that only agriculture, industries, crafts and real estate create added value. Transforming Janusson’s calculation of national income in 1929–1938 into that of GDP for the same period, Valge also included the service sector. Valge converted estimates in EEK into GK$ 1990, so uniquely for Estonia we have output estimates comparable with those in the MPD and those provided by Roses and Wolf (2010). Thus, for the year 1929, estimates by Roses and Wolf are greater than Valge’s figures by 28.4% and by 37.4% for the year 1938 (Table 8.3).

Table 8.3 National income and GDP of Estonia in 1923–1938. Data sources: Valge (2003: 2211, 2719, 2721, 2276). GDP per capita in EEK calculated from total output and data on Estonia’s population in Valge (2003: 2721)

Roses and Wolf (2010) provide no estimates for Lithuania. On this country, only ‘guestimates’ by Lithuanian historian Gediminas Vaskela (2014: 116) for 1938 (1800–1900 GK$ 1990) and calculations by Adomas Klimantas and Aras Žirgulis (2020: 249) for 1937 (2064 GK$ 1990) are available. No MPD comparable output estimates are available for Latvia, except for Vaskela’s (2014: 113) guestimate that, in 1938, Latvia’s GDP could have been approximately 2800 GK$ 1990.

Roses and Wolf also do not provide estimates for 1913. However, inclusion of this year is necessary to provide a realistic picture of the economic progress achieved by the Baltic countries after their separation from Russia. The real measure of economic progress in the interwar independence period is how much their economies grew after recovering from the devastation of the war in comparison to the highest level achieved under foreign rule on the eve of World War I. Using 1922 as the base year, we cannot avoid an overestimation of their achievements because, at this time, recovery growth was still incomplete.²

New measurements are also needed because the available estimates became outdated following new releases of the MPD in 2013 and 2018. In the MPD 2018 release, following the practices of the Penn World Tables (PWT), MPD 2018 provided two estimates for each year: real GDP per capita at constant national prices, obtained from national accounts data (RGDPNApc), which is commended for cross-country comparisons of growth, and current year GDP per capita (CGDPpc), preferred for cross-country comparisons of economic standards of living (Bolt et al. 2018; Feenstra et al. 2015).

Following this approach, we (Norkus and Markevičiūtė 2021) derived the estimates of the CGDPpc and RGDPNApc in 1913, 1922, 1929 and 1938 for all three Baltic countries. To make them comparable with Roses and Wolf’s (2010) results, we selected 1922, 1929 and 1938. Like Roses and Wolf (2010), we used the indirect or econometric method, instead of the direct approach, including construction of historical national accounts in the system of national accounts (SNA). Admittedly, the direct method is the ‘king’s road’ in historical national income estimations. However, only using the indirect approach, it is possible to cover years with scarce data for construction of national accounts. In particular, this applies to 1913. The SNA framework is crafted to monitor economic activities in nation states. However, the territories of Estonia, Latvia and Lithuania were not even separate administrative units by 1913. The territory of Estonia was divided in two (Estland and Livland) and that of Latvia into three (Kurland, Livland and Vitebsk) provinces (governorates) of tsarist Russia. The territory of Lithuania also includes part of the German empire (the Memel/Klaipėda area), in addition to parts of the Kaunas, Vilnius and Suvalkai provinces of the Russian Empire.

We used a method that seems to be becoming the standard indirect approach in similar primary data situations. This method was pioneered by Robert C. Allen (2000) and then used by Malanima (2010), Arroyo Abad et al. (2012), Alvarez-Nogal and Prados de la Escosura (2013), Ridolfi (2016), Malinowski (2016) and Malinowski and Van Zanden (2017). This method involves the use of a model wherein GDP per capita is estimated from its values in benchmark countries and from data about real wages, food self-sufficiency and agricultural employment, using urban population data as a proxy. Importantly, this method was used in constructing the most recent releases of MPD (2018, 2020) by Bolt, Inklaar, de Jong and van Zanden to fix many data gaps. This point may provide some reason to claim that our estimates are better in terms of methodological compatibility with the recent MPD versions than the Roses and Wolf (2010) results.

In earlier research, our method was applied mainly to pre-industrial economies. However, under closer consideration, many of them were early industrialising economies (in fact, Malanima (2010) closes with the year 1913). There is no clear line separating ‘pre-industrial’ from ‘industrialising’ countries, and the Baltic countries were just late industrialisers. As our data (see the next section) shows, in 1913 Lithuania was a completely pre-industrial economy, and Latvia was apparently crossing the threshold separating pre-industrial from industrial economies, while Estonia was a case in between. What was peculiar about them (and should make them interesting cases to the broader research community interested in economic growth and comparative development) is the reversal of the master industrialisation process and re-agrarianisation in the interwar period.

The Groningen Growth and Development Centre (GDDC) experts used this method to calculate the output values for most African countries (since 1950) in MPD 2018 and MPD 2020. Primary data availability for the Baltic countries in 1913 and 1922 are not much better than for African countries in the 1950s, but we could use the GDP data of 13 European countries as benchmarks. However, GDDC experts used only one country (South African Republic). In addition, we empirically calibrated the model by replacing the assumption of complete food self-sufficiency with empirically estimated values and measured labour productivity in industry and agriculture in the Baltic countries for the relevant years. Therefore, we believe that our results for the Baltic countries in 1913–1938 are no less reliable than GDPpc values in MPD 2018 and MPD 2020, derived using the same method.

However, briefly after the ultimate publication of our estimates (in Norkus and Markevičiūtė 2021), the GDDC experts did reverse to the older approach used by Maddison, providing only one GDPpc estimate (conceptually closer to RGDPNApc) but changing the base year. Table 8.4 contains a recalculation of the GDPpc for the Baltic countries from their values for 13 European countries in the MPD 2020 in 2011 GK$.³ The equivalents of these figures in 1990 GK$ are available in Norkus and Markevičiūtė (2021, Table 19).

Table 8.4 GDPpc (in 2011 int$) of Baltic and selected comparator countries in 1913–2018. Data sources: MPD (2020); author’s own calculations (credits to J. Markevičiūtė for assistance); Valge’s (2003: 2726) alternative figure for Estonia in 1938

Our GDP per capita figures for Estonia in 1938 (3796 int$ 2011) may be an underestimate. It corresponds to 2377 GK$ 1990 (see Norkus and Markevičiūtė 2021, Table 19) and thus makes up 86.6% of Valge’s estimate for 1938. For 1929, our estimate (2015 GK$ 1990) makes up 92.3% of Valge’s (2003) fig. (2182 GK$ 1990). Both of Valge’s estimates are within 95% confidence intervals for our estimates (1522.37 at the lower end and 2667.64 at the upper end for 1929; 1848.19; 3058.31 GK$ 1990) for 1938 and do not disagree dramatically with ours. However, Valge’s estimates may nonetheless be more reliable because they are grounded in historical national accounts (Janusson 1932, 1937; Horm 1940). Therefore, the increase of Estonia’s output by 25.8% in 1929–1938, as implied by Valge’s estimates, may be closer to the truth than the 17.4% growth implied by our own estimates. Applying Valge’s growth rate in 1929–1938 to our estimate for 1929, we receive 4068 int$ 2011 as an alternative estimate for 1938.

Our estimate for Lithuania in 1938 (3619 int$ 2011, which is equivalent to 2267 GK$ 1990) is higher than that of Klimantas and Žirgulis (2020) for 1937 (2064 GK$ 1990). However, the last estimate is deeply within the boundaries of the 95% confidence interval of our estimate (1761.3 lower, 2917.2 upper bound). Even more importantly, it is known that 1936–1938 was a time of very rapid growth in Lithuania’s economy (Dargis 1978). Even if we do not accept Dargis’ (1978: 93) judgement that Lithuania’s national income increased in 1938 by 18.23% in comparison with the previous year, the 10% annual growth estimate is firmly validated by data about the increase in Lithuania’s budget income and foreign trade in 1937–1938. Assuming this kind of increase makes our estimate for 1938 nearly coincide with Klimantas and Žirgulis’ (2020) result.

The main aim of the economic output measurement at the PPP is to facilitate cross-national comparison of economic wellbeing and growth performance. The real meaning of our estimates and their plausibility transpires when we use them to establish the position of the Baltic countries in the European or worldwide country ranking and to compare our ranking with the received rankings. Therefore, we included into rankings not only our benchmark countries but also all European countries with relevant GDPpc data available in the MPD 2013 and 2018, some ‘Western offshoot’ countries (Australia, Canada, the United States) and Japan.

To facilitate the appraisal of the reliability of our findings, we also included some Latin American countries because some scholars have argued that interwar Latin American together with East European countries belonged to the same (peripheric) structural level of the world economy (e.g. Aldcroft 2006; Berend 1998, 2006; Janos 2000; Szlajfer 1990; Turnock 1997, 2006). To provide a long-term perspective, the table includes output data for the 1989–2018 period, which will be used when applying the OOST in the next chapter. However, many of the former communist countries (first of all, new states) are absent from the table because MPD 2020 only provides data for the last decades.

Our estimates break new ground by establishing the standing of the Baltic countries in the international context as of 1913. They disclose that the GDPpc level of Latvia surpassed the all-Russia mean by 70% and approached the levels of Norway and Ireland in 1913. Estonia had near parity with Finland, while Lithuania was markedly (17.6%) above the all-Russia mean. However, both countries lagged behind Latvia, which surpassed all the East European countries. None of the Baltic countries were poor by American standards (with a GDPpc less than 20% of that in the United States according to the AST), although Lithuania was not far away from the poverty threshold in 1913–1929, which was less than 2023 int$ 2011 in 1913, 2002 in 1922 and 2391 in 1929.

Next, our findings do not support the picture of Estonia and Latvia as star growth performers in interwar Europe, emerging from Roses and Wolf’s (2010) estimates (see Fig. 8.1). They also contradict their claim that in 1929–1938, growth performance of Estonia and Latvia was stronger than Finland’s. The growth of the Baltic countries was relatively weak for the interwar period. By relative weakness, we mean the lack of advancement of the Baltic countries in the ranking order as of 1938 in comparison with their international standing in 1913.

Fig. 8.1

Changes of GDPpc (in 1990 GK$) of the Baltic countries in 1913–1938. Author’s own production. Data sources: Roses and Wolf (2010: 190); author’s own calculations (credits to J. Markevičiūtė for assistance)

After the loss of the Russian market (see below), all three Baltic countries pursued the ‘Denmarkisation’ of their economies during the independence period to become rich food-exporting countries (Kõll 1994). While they possibly could succeed at this endeavour in the longer term, by 1938, the success of this economic policy was still limited. If Denmark’s food production exceeded domestic food consumption by up to three times, in the Baltic countries, the ratio of total food production to total domestic food consumption (food sufficiency) did not surpass 1.2 times by 1938 (see Norkus and Markevičiūtė 2021, Appendix 11).

For Estonia and Latvia, the overall pattern is one of long recovery after the destruction brought on by World War I, accomplished by the end of the first decade of independence. The next (and last) decade was one of considerable growth, which, however, did not lead much above (more than 20%) the 1913 RGDPNApc level. The output per capita increase (up to 40% above the 1913 level) by 1937–1938 was much more pronounced in Lithuania but was limited to the second decade of independence. Although all three Baltic states pursued policies of ‘Denmarkisation’, the lag separating them from Denmark and the other Nordic countries increased during the interwar independence period, and we did not find a conspicuous improvement of the place of Latvia in the international ranking by 1938, as Roses and Wolf claim (cp. Table 8.2). Whether ‘Denmarkisation’ would have succeeded in the longer term can only be speculated upon.

The data in Table 8.4 display a weak but discernible pattern of regional convergence. In 1913, the GDPpc of Estonia and Lithuania were 87% and 69% of the CGDPpc of Latvia, respectively. In 1938, these figures were 83% and 80%. If Lithuania was still rather far behind Latvia, it was catching up with Estonia, with Lithuania’s GDPpc growing most strongly (3.65% annually, with 2.07% for Latvia and 1.80–2.58% for Estonia) in 1929–1938 after Lithuania’s stagnation during the first decade, which also needs an explanation.

Explanation of these patterns in data should start with the reporting of the policies of the Russian Minister of Finance (in 1892–1903), Sergei Witte, who increased protective tariffs to defend Russia’s fledgling industries from the competition of foreign companies. Another aim was to attract foreign direct investment (FDI), as after tariffs were increased, Western companies exporting to Russia could profit from its huge market only by establishing their subsidiaries beyond the customs wall. Latvia (especially Riga) was the most attractive place for such FDI because of its geographic location and qualified workforce (Karnups 2013).

Together with the Estland province, the Kurland and Livland provinces were known as the Baltic provinces of the Russian Empire that enjoyed considerable autonomy, which included a legislation different from the core of Russia (Plakans 1995). However, only members of the small but economically and socially dominant German minority (primarily landlord nobles) had participation rights in the self-governmental bodies, while the indigenous Latvian population remained nearly completely excluded until World War I.

However, it shared the Lutheran Protestant religion with the Baltic Germans, becoming inculcated by this powerful socialisation agency exhibiting many common features of mentality, summarily described by sociologists as the Protestant ethic (Norkus 2022, 2012: 222–241). One important feature was high literacy rates. Due to the zeal of Lutheran ministers to enable their parishioners to fulfil their religious duty of reading the Holy Scripture, compulsory elementary education in Latvian was introduced in the eighteenth century. As a result, according to the 1897 Russian population census, the Estland, Livland and Kurland provinces had the highest literacy rates in tsarist Russia (Raun 2017).

They included the territories of both Estonia and Latvia. However, Latvia had an important advantage over Estonia (and Finland)—the ice-free ports of Riga, Liepaja (Libava) and Ventspils (Windau). After completion of the Rīga-Orel, Moscow-Rīga, Rīga-Pskov-Rybinsk (with Russia), Liepāja-Romny (with Ukraine) and Moscow-Ventspils railway links, this advantage offset by far the advantages of greater closeness to Saint Petersburg enjoyed by Estonia and Finland. In 1913, 28.2% of total exports and 20.6% of total imports of the Russian Empire went through Latvian ports (Skujeneeks 1927: 677).

However, Latvia did not become simply a Russian transit area (Krūmiņš 2017). Somewhat paradoxically, the Russian economic policies of the early twentieth century that were used to industrialise the latter by attracting FDI to Russia, while defending its huge market with protective tariffs, were most successful not in inland Russia but in Latvia, transforming it (and to a lesser extent also Estonia) into the first largely industrialised region in the Russian Empire. Due to the advantages of its geographical location, an established reputation as the hub of Baltic trade and the social capital of Riga’s (mainly German) economic elite, Latvia’s ice-free Baltic seaports were extremely successful in attracting FDI investments to establish many large and technologically advanced production sites for the inland Russia market. Considering this industrial production as Latvia’s export, this country can be considered the first case of export-oriented, FDI-driven industrialisation, promoted by World Bank and IMF experts as the ‘king’s road’ for achieving growth since the late twentieth century.

Another Latvian advantage was Riga’s long history of trading with the rest of Europe, creating a strong entrepreneurial (mainly German) elite with easy access to commercial credit and good knowledge of market opportunities in both Russia and Europe. ‘Foreign capital did select Latvia as its operational centre in the Russian Empire because of its geographic proximity, congenial environment, and availability of skilled middlemen and consultants among local entrepreneurs’ (Netesin 1980: 237). This selection helped Latvia to become the third largest (after Saint Petersburg and Moscow) industrial centre of imperial Russia. Its specialisation was industrial production for the Russian market using imported raw materials and semi-manufactured goods that were not or could not be produced in Russia itself (Netesin 1980; Krūmiņš and Šiliņš 2017; Krastiņš 2018: 17–73). Contemporaries perceived it as the ‘Belgium of the East’ (Hiden and Salmon 1991: 76–77). Concerning the size of the disparity between Latvia and the all-Russia per capita output mean, the data about cross-regional differences in per capita industry and trade turnover in Russia may provide an important clue; in Kurland and Livland, the values of this indicator were two times above the all-Russia mean (Netesin 1980: 55).

In 1913, 63% of Latvia’s total industry output was exported to Russian regions, 11% was exported to foreign countries and only 26% was sold in Latvia (Krastiņš 2018: 72). In our time, this kind of industrial output marketing structure is displayed by the most successful newly industrialising countries, such as South Korea or Taiwan (e.g. see Amsden 1989; Birdsall et al. 1993). Expanding this analogy, we should note that a significant part of Latvia’s export industrial production was relatively technologically advanced by this time. Riga was where the first Russian planes and cars were designed and produced, together with a large proportion of railway carriages and other railway equipment (Krūmiņš and Šiliņš 2017: 43–44). The largest factory in Riga, Provodnik, a Russian–French joint venture, offers a good representation of the profile of pre-World War I Latvia’s ‘entrepôt industry’. In 1913, it employed 13,500 workers and produced 55 million roubles worth of production, 95% of which was exported to inland Russia and abroad (Karnups 2013: 26).

A pertinent reservation is that our findings reflect conditions in mainland Latvia, encompassing Kurland and the southern part of Livland, which together with the lands of contemporary Estonia in the seventeenth to early eighteenth centuries were ruled by Sweden and were then annexed by Russia in 1721. They preserved broad internal autonomy, used by a German minority to perpetuate its social privileges and foster cultural connections with German-speaking lands in Central Europe, which contributed to the appeal of Riga as an operational centre for international business interested in the Russian market. However, around 30% of the Latvian population lived in Eastern Latvia (Latgale), which had been under Polish–Lithuanian rule since the sixteenth century and was annexed by Russia in 1772. By 1913, its level of socio-economic development was closer to Eastern Lithuania. No such marked cross-regional disparities existed in Estonia, but they were characteristic of Lithuania, where Eastern Lithuania and the larger part of Western Lithuania (Samogitia) were less socio-economically advanced than its central and south-western areas.

The discussion of the pre-World War I Latvian phenomenon of the ‘entrepôt industry’ also provides a clue of where to look for the causes of Latvia’s relative decline during the interwar period, meaning both the lack of advancement (pace Roses and Wolf 2010) in the international output per capita ranking and the decrease in the disparity between Latvia and Lithuania. The operation of this industry was interrupted immediately after the outbreak of World War I, stopping the import of necessary components (Aizsilnieks 1968: 30–32; Netesin 1980: 170–173). With the German military approaching Riga in 1915, by 1917 nearly all Latvian industrial equipment was evacuated to inland Russia. According to the regulations of the peace treaty between the newly established independent Republic of Latvia and Soviet Russia on 11 August 1920, it should have been returned, but only 1% actually came back (Krastiņš 2018: 81).

Among the Baltic countries, Latvia’s territory was the most severely devastated by military activity. With the exception of the West Estonian archipelago, which was seized by a German fleet in a landing operation in October 1917, Estonia remained nearly unscathed by military action. Lithuania’s territory was an area of military activity for a rather brief time (summer–early autumn 1915), with the German military ousting Russian troops in mobile warfare. Military activity then moved to Latvia, where some 24% of its territory became the location of positional warfare for more than 2 years. On 7% of its territory, more than 50% of buildings were destroyed, and on 17%, the losses amounted to 10–50%.

While these landscapes resembled those in Eastern France, which were locations of positional warfare on the German Western Front, their economic rehabilitation took much longer. Differently from France, the newly established independent state of Latvia did not receive reparations from Germany, and there were no parts of its territory unscathed by military activity that could provide an economic basis for recovery of the devastated lands. During the independence war of 1918–1920, most of its territory was devastated once again by the Red Army. That army was repulsed with the help of German volunteers, who mutinied in autumn 1919, plundering everything that remained up for grabs before their final departure to Germany. Latvia emerged from the independence war heavily indebted, and as a young state with an insecure future, it could not borrow cheaply on international credit markets.

The Russian market became closed to industrial production of the Baltic countries due to the autarkic economic policies of the Soviet government, implemented using the state monopoly of foreign trade. This behaviour was first experienced by its neighbour Estonia because it was the first to make peace with the Bolsheviks, and the greater part of its industry was not evacuated. In 1920–1922, the Estonian government allocated generous credits to revive enterprises that were in operation before World War I for Russian markets, in the expectation that they would shortly open again. Those hopes were dashed, plunging the country into a severe financial and economic crisis in 1924 (Valge 2006). Another painful experience was that the industrial products of the Baltic countries could not compete in Western markets, also being out-competed in the markets of newly established East European states, which were difficult to access because of the economic nationalist policies of their governments (Kofman 1997 (1992)). If in 1914 only 26% of Latvian industrial production was locally consumed, by 1927 it was able to export only 11.7% of its production (Karnups 2017: 116).

The majority of Latvia’s industrial work force, which in 1915–1917 was evacuated to inland Russia together with industrial equipment, did come back, serving as the most important asset for the partial restoration of industry, working for the internal market (Krastiņš 2018: 79–86). However, briefly before the outbreak of the world economic crisis of 1929–1933, the gross output value of Latvian industry (at gold parity) was only 47% of its 1913 scope (Buševics 1930: 129). Urbanisation did not recover to the 1914 level even by 1939. The excess labour force moved back to agriculture, which was possible due to agrarian reforms that were implemented in all three Baltic countries beginning in the early 1920s.

They targeted large landlords, who in 1919 owned 48.12% of total land in Latvia, 46.93% in Estonia and 26.20% in Lithuania (Vaskela 1998: 57). The aims of this reform were mainly political and social. The political purpose was to undermine the economic power basis of the socially dominant national minorities (German in Estonia and Latvia and Polish in Lithuania). The social purpose was to provide for the rural and part of the urban working class, making them small landowning farmers, immune to radical leftist (communist) agitation.

While agrarian reforms did succeed in their political and social aims, making partial de-industrialisation of the Baltic countries relatively painless, their economic consequences were controversial. In contrast to the warnings made by critics of agrarian reform (e.g. see Sering 1921), there was no decrease in the total agricultural output. However, most of the output produced on these small new farms was for the consumption of farmers and their families, while agricultural production on the large estates and farms of the pre-World War I era was commercially oriented. In a relatively short time, small farmers of the Baltic countries discovered their comparative advantages in animal husbandry (Kõll 1994). Its products (butter and bacon) became the main export articles of all three Baltic countries during the second decade of independence. During the first decade, they were able to maintain their balances of payments by exporting flax, wood and wood products, which also remained the second most important (after butter) export commodity of Latvia during the second decade of independence (Krastiņš 2018: 99).

Therefore, the interwar Baltic countries were economically successful in the sense of what success means for minimally developed peripheral countries—surviving in the world market as exporters of a few basic commodities (Williamson 2011). However, this form of success also means belonging to low-income countries, because such production creates relatively low added value. These structural characteristics of their economies, which are shared by all three Baltic countries (with Estonia and Latvia becoming more similar to Lithuania during the interwar years), explain their relatively low GDP per capita levels in 1922–1938 and their regional convergence pattern disclosed by our analysis.

Assessing the reliability of our findings about Lithuania’s GDPpc levels as of 1913 (see Table 8.4), it is important to consider that the Russian government perceived Lithuania primarily as an area of military activity in the coming war with Germany, consciously underinvesting into its infrastructure (Meškauskienė 1963). Due to restricted local economic opportunities, Lithuania became an area of intense emigration, some 25% of its population (600,000) leaving the country in 1868–1914 (Truska 1961: 79). An important centre of Lithuanian immigration was its neighbour Riga, where Lithuanians were the second largest (after Latvians) national group among industry workers in Riga. According to data about the ethnic composition of employees working for the 259 largest industrial enterprises of Riga in 1908, ethnic Latvians made up 48.83%, Lithuanians 27.47%, Russians 14.68% and Germans 5.96% (Netesin 1958: 18).

Two remarkable findings about Lithuania’s growth performance in 1913–1929 need further explanation. First, Lithuania’s economy experienced an early recovery to prewar levels that was accomplished by 1922. Second, there was protracted stagnation after this recovery, lasting until 1929 (see Fig. 8.1). Concerning the first point, it should be taken into consideration that because of the absence of notable industry before World War I, Lithuania did not suffer from de-industrialisation. It was an underdeveloped agrarian country before World War I and remained such in the interwar period, exchanging its peripheral status in the Russian Empire for an analogous position in the capitalist world economy, with a strong economic dependence on Germany. During the German occupation in 1915–1918, the German administration overtaxed Lithuanian farmers, commandeering horses and livestock. However, these losses could be promptly recouped after the war.

On account of its strategic calculation to establish a client buffer state on the border with Eastern Prussia, the German government provided the Lithuanian government with a 100 million RM loan early in 1919. In exchange, Lithuania remained in monetary union with Germany until 2 October 1922. The loan enabled Lithuania’s government to pay its bills in the difficult year of 1919 without overtaxing its population. Importantly, this debt did not need to be repaid because it was nearly extinguished by German hyperinflation by 1923 (Daščioras 1971, 1986). Before its outbreak in 1922, the monetary union with Germany enabled Lithuanian farmers to profit from a high demand for food in postwar Germany and satisfied Lithuania’s demand for investment goods, necessary for postwar reconstruction at beneficial prices, providing Lithuania with very favourable terms of trade in 1919–1922.

As in Latvia, military activity in Lithuania damaged its infrastructure. However, by 1918 it had already been reconstructed by the German administration itself, which needed roads to connect Germany with troops on the front line. As the Russian administration avoided investing in Lithuania’s infrastructure before World War I for the same reason, the German military was forced to do so itself. Thus, during the German occupation in 1915–1918, 458 km of railways (265 km wide-gauge and 193 km narrow-gauge) was constructed (Šalčius 1998: 226; Lastas 1981: 70–71) together with telegraph and telephone lines. This work amounts to more than the total railway construction during the independence period of 1918–1920 (cp. Matyckas 1968: 241–242; Meškauskas 1992: 199–201). Therefore, and perhaps uniquely, Lithuania after World War I had a much better railway and road network than before the war.

The next favourable factor contributing to the rapid postwar recovery of Lithuania was the huge Lithuanian diaspora. The main reason for emigration before World War I was limited economic opportunities in Lithuania. Most of the 600,000 persons who emigrated in 1868–1914 left for the United States, where in 1899–1914 alone US immigration authorities counted 252,594 persons of Lithuanian ethnicity (Eidintas 2005: 61). Interrupted by World War I, the flow of remittances flooded the country starting in 1919, and some 20,000 emigrants returned to their homeland, looking for investment opportunities for their savings. With such favourable circumstances for rapid recovery, how then can we explain the prolonged economic stagnation in the following years? Did stagnation exist at all?

First, after termination of the monetary union with Germany, high inflation was replaced by deflation, plunging Lithuania into a prolonged economic crisis, culminating in 1925–1926 (Daščioras 1975: 56; Dirvelė 1961). The crisis manifested itself as a contraction of the amount of money in circulation, exorbitantly expensive credit, bankruptcy of many newly founded businesses (many of them founded by returned American Lithuanians) and a decrease of Lithuania’s exports in 1924, followed by stagnation until 1928 (Vaskela 2011). Because imports did decrease less than exports, Lithuania’s gold and hard currency reserves were depleting. Only the continuing inflow of emigrant remittances from the United States saved the newly introduced Lithuanian currency from devaluation.

The main cause of the economic crisis was the sudden loss of Germany as Lithuania’s only food export market at this time. After the introduction of the Lithuanian national currency, the litas (LTL), Germany was not able to pay for Lithuanian food imports until 1924, while after the end of the German mark’s hyperinflation, Germany imposed heavy food import duties and introduced sanitary border controls to protect its own farmers and landlords in neighbouring East Prussia (Daščioras 1975, 1983, 1986). It took Lithuania several years and considerable investment to develop food industry capacities to produce milk and meat products that could penetrate the demanding, faraway British market and successfully compete with established insiders there (see Glemža 1972). Until then, Lithuania could pay for imports only from revenue earned from exporting flax and forest products and from emigrant remittances.

Many researchers consider the economic crisis in 1925–1927 an important contributing cause of the authoritarian coup in Lithuania on 17 December 1926, which ousted the left-of-centre government that had governed Lithuania since June 1926 (e.g. see Butkus 2007; Eidintas 2015; Lopata 1998; Norkus 2016). While the electorate’s discontent with economic difficulties contributed to the defeat of the incumbent Lithuanian Christian Democratic Party, the victorious coalition of the Peasant Popular Union and Social Democratic Party of Lithuania was not able to balance the budget without cutting defence expenditures and discharging a large part of the military. Protecting their jobs, angered officers plotted a coup and then gave over rule to a civilian government under the Lithuanian Nationalist (tautininkai) party.

However, the above are only conjunctural or proximate causes explaining the delay of the sustained growth of independent Lithuania’s economy beyond its 1913 level. There were also deeper structural obstacles. The most important of these was partition of Lithuania’s territory in the wake of the conflict with Poland over Lithuania’s historical capital Vilnius (see Senn 1966; Briedis 2012). Map 8.1 provides Lithuania’s official map, which until 1938 decorated the classrooms of all Lithuanian schools and the rooms of its government offices. According to this map, around 40% of its claimed 88,111 km² population was ‘under foreign occupation’ (Pakštas 1968: 461).

Map 8.1

Official map of Lithuania’s borders in 1923–1938 (in Lithuanian and French). Source: Finansų ministerija. Centralinis statistikos biuras 1928: 5. https://archive.org/details/lietuva-skaitmenimis-1918-1928-1928. Accessed 10 January 2023). Public Domain Mark 1.0 The eastern and southern territories are marked as the ‘Polish occupied part of Lithuania’. The claimed state borders are rimmed in orange. The de facto border with Poland is rimmed in dark green and is designated as an ‘occupation line’

In this map, the eastern and southern borders of Lithuania are marked according to the Soviet–Lithuanian Peace Treaty of 12 July 1920 (see also Chap. 5). They were not recognised by Poland’s government, which after the ultimate victory in the Soviet–Polish war of 1919–1921 established its control over Grodno and Vilnius with adjacent territories. Poland’s main argument was the ethnic composition of the population in the contested territories, the Poles being a majority in larger cities and their surrounding territories. Along with the historical argument that Vilnius was the historical capital of the Grand Duchy of Lithuania, Lithuanians argued in terms of economic geography. Drawing on rich evidence, they contended that independent Lithuania is not viable economically without Vilnius as the economic centre of gravitation of its claimed territory. This claim is the subject matter of two studies published by Lithuanian experts (Klimas 2018 (1917); Rimka 1922 (1918)), which remain useful as sources for historical statistics.

Polish authors (e.g. Studnicki 1922; Gorzuchowski 1928; Wielhorski 1928) did not contradict Lithuanian arguments about the economic indispensability of Vilnius for the Lithuanian state. Quite the opposite, they provided additional evidence in its support. They did however draw very different political conclusions from this argument: Lithuania can survive economically only by reconciling with Poland and entering with it into an economic and political union, which may mean confederation, federation or just autonomy inside Greater Poland. Observing Lithuania’s economic difficulties after the dissolution of the German–Lithuanian monetary union in 1922 discussed above, they interpreted them only as confirmation of the thesis that a truncated ‘Litwa Kowienska’⁴ is not economically viable. They predicted that it would collapse economically in the near future, at which point the last large Lithuanian forests would be cut down and exported to earn hard currency to pay for Lithuania’s imports (e.g. Gorzuchowski 1928: 22–23). Wielhorski (1928: 205) even argued that ‘generally, cultural, economic life and wealth of Lithuania’s population decreased over the last 25 years, becoming even more primitive in comparison with their condition by the end of the last century’.

Lithuania considered itself at war with Poland until 1928. Until 1938, there was no traffic or direct trade between Poland and Lithuania, which meant total economic separation of eastern Lithuania from the interwar independent Republic of Lithuania. Arguably, the postwar recovery of Estonia and Tallinn would have been much longer and more difficult if, because of some political reasons, Tallinn and Riga would have been excluded from their political borders. Despite the structural damage to Lithuania’s economy imposed by the loss of eastern Lithuania, doomsday forecasts of its grim economic future were belied by its finding a place in the international division of labour, allowing for sustainable growth and the upgrading of its economic structure during the second decade of Lithuania’s independence.

An important contribution to Lithuania’s economic achievements during this period was made by the economic integration of the Klaipėda (Memel) region. While it constituted only 4.3% of interwar Lithuania’s actual territory (55,670 km²), and its population (141,650 according to a census dated 20 January 1925) was 6.4% of Lithuania’s total population in this year (Pakštas 1968), its economic significance exceeded these figures because it was the only port in Lithuania, and it was ice-free. Inhabited by a mixed German–Lithuanian population, it was detached from Germany by the Versailles peace treaty, and in 1920–1923, it was under the administration of France. In 1923, Allied powers recognised Vilnius as legitimately being a part of Poland, compensating Lithuania by transferring the Klaipėda (Memel) region under its sovereignty, which was granted broad territorial autonomy and self-government rights.

However, while Klaipėda was a large asset for independent Lithuania’s long-term economic development, its economic integration was another challenge for Lithuania’s still very weak economy in the 1920s. Although output per capita in the Klaipėda region exceeded that in mainland Lithuania by up to 50% (Vaskela 2014: 89), the taxation rights of Lithuania’s central government were limited by the Klaipėda Convention, signed by Lithuania in 1924 with the United Kingdom, France, Italy and Japan, strongly protecting the fiscal rights of local government (Kalijarvi 1937; Plieg 1962; Žostautaitė 1992).

Instead of receiving transfers from the Klaipėda region to its central budget, Lithuania’s government had to heavily invest in the reconstruction of the Klaipėda port (Gaigalaitė 1979: 56) and in connecting it to mainland Lithuania’s transport network. Until 1932, when construction of the Šiauliai–Klaipėda railway branch was concluded, Klaipėda could be reached from Kaunas by railway only via a bypass crossing the territory of Latvia and could not compete without government subsidies, with Königsberg, Danzig and Liepāja as loading and landing ports for Lithuania’s exports (Lastas 1981: 72).

Before World War I, industry in Klaipėda specialised in processing timber shipped down the Nemunas River from eastern Lithuania and Belarus, which then was mainly sold on the German market (Pfeiffer 1918; Meškauskas 1992: 19–20; Willoweit 1969, 1: 406–412; Žukas 2010). As this traffic ceased after the partition of Lithuania’s historical territory between Poland and independent Lithuania, Klaipėda’s large timber industry could not contribute to Lithuania’s economy. Out of the 42 sawmills operating in the Klaipėda region in 1913, only two worked in 1926. Therefore, the export of timber and timber products from Klaipėda Port made up only 5% of its total exports in 1912 (Gaigalaitė 1979: 56).

In later years, Klaipėda’s timber industry gradually recovered, the number of sawmills increasing to six in 1930 and then to 18 in 1937 (Willoweit 1969, 2: 640–641). This increase became possible after the Soviet Union decided to supply it (partly for political reasons) with timber at preferential prices since 1931, while the Lithuanian government provided exclusive lower railway fares for this deal (Mel’tyukhov 2015: 388–389; 408–409). Overall, it took several years for Klaipėda’s regional economy to exchange its role as a timber-processing industry centre in the advanced industrial economy of Germany for the role of independent Lithuania’s ‘little Riga’ (Willoweit 1969, 2: 632–652).

By identifying Klaipėda as Lithuania’s ‘little Riga’, we pinpoint the structural and functional similarity between the position of the region in Lithuania’s agrarian economy in the 1930s and that of Riga (and other Baltic ports) in the economy of the Russian Empire before World War I. In addition to serving (since 1932) as Lithuania’s only export and import port, Klaipėda also hosted newly emerged industrial sectors, producing consumption goods (textiles, rubber products and fertilisers) for Lithuania’s market using raw materials imported from abroad and profiting from its protectionist economic policies in the 1930s (Meškauskas 1992: 161–162; Walter 1939: 68–82). However, in this comparison, the adjective ‘little’ should of course be emphasised because of differences in scale and the brief duration of Klaipėda’s new prosperity. Nazi Germany re-annexed the city in 1939, while the next year the Soviet Union did the same with Lithuania itself.

According to textbook wisdom in classical development economics, with Estonian economist Ragnar Nurkse (1907–1959) as one of its founding fathers (see Kattel et al. 2009), economic modernisation involves moving the labour force from less-productive employment in agriculture to industry, provided the rise in the productivity of agriculture is sufficient to maintain an urban population and exports (see Kuznets 1966, 1971; Ros 2000: 69–133; Shirmai 2015: 73–130). This movement happened in Latvia and to a lesser extent in Estonia in the late nineteenth and early twentieth centuries.

Uniquely, some features of the precocious industrialisation of Latvia anticipated the export-driven industrialisation of the newly industrialising countries in the twentieth century (Kiely 1998; 97–114; McCombie and Thirlwall 1994; Weiss 2011). They include FDI as the main source of investments (instead of internal accumulation, advocated by classical development economics) and specialisation in the production of relatively technologically advanced products, using the comparative advantage of an abundant supply of skilled labour.

The case of Latvia’s (and to a lesser extent Estonia’s) early rise and subsequent decline is not unique in the turbulent economic history of the late nineteenth–early twentieth centuries. Latvia shares its record of early enrichment and ensuing decline with Argentina, Uruguay and Chile, which by 1914 had GDPpc at PPP levels as high or even higher than those of Scandinavian countries (Williamson 2011: 146–168; Hofman 2000: 7–29). In the case of the precocious relative richness of some Latin American countries, the long-term unsustainability of their affluence may be explained by insufficient diversification of their economies, making them all too dependent on the price shocks in demand for their basic export commodities. However, changes in the economic fortunes of the Baltic countries can only be explained by taking into account factors and forces that transcend the boundaries of narrowly conceived economic history.

The main factor was the imperial expansion of Russia. Since the seventeenth century, the ‘great strategy’ of this expansion (LeDonne 2004; Lieven 2000) was the conquest of territories presently enclosed within the borders of Estonia and Northern Latvia to gain access to the Baltic Sea. This aim was achieved by the Russian victory in the Northern War of 1699–1725. The grip of Russia over the eastern shores of the Baltic was expanded and consolidated by the partitions of the Polish–Lithuanian Commonwealth, when by 1795 Russia had annexed the remaining parts of Latvia (Latgale and Kurland) as well as almost the whole territory of Lithuania.

Another factor is nationalism, which is a political movement claiming for territorial ethnocultural minorities the right to live in politically independent nation states (Smith 2010: 9, 25–30). Arising in the second half of the nineteenth century, Baltic national movements used the situation created by the defeat of the Russian Empire in WWI and its dissolution in 1917–1918, followed by the defeat of its vanquisher power Germany, to establish and consolidate independent nation states. During the interwar period, all three Baltic nations used the institutions of the modern state to foster the development of national ‘high cultures’ in the vernacular languages (Kasekamp 2010; Plakans 2011). As Russia embarked on building an autarkic centrally planned economy, there was no possibility for restoration of Latvia’s and Estonia’s entrepôt industrial economies.

Thus, the industrialisation of Estonia and Latvia did not survive beyond WWI. The price paid for the integration of newly established independent nation states into the world market was their partial de-industrialisation, causing them to more closely resemble underdeveloped agrarian countries from the eastern European periphery (including Lithuania). The rather slow growth of both countries in the interwar years simply reflects this structural transformation of their economies. The loss of the Russian market after WWI had a depressing effect on the economic growth of the Baltic countries not dissimilar to that which the dissolution of the Habsburg monarchy had on its constituent lands (e.g. see Berend 1998: 228–229; Pasvolsky 1928). ‘While the mechanisms are not clear, specialists point to the profound structural weaknesses in the region after imperial collapse and the rise of protectionism as new states instituted their own tariff regimes’ (Good 1996: 76).

The new development pathway of Baltic countries, inspired by the success story of Denmark, which did economically modernise by building export-oriented highly productive agrarian capitalism grounded in family farms, was violently interrupted by the Stalinist export of the Russian Revolution, making the Baltic states parts of the restored Russian Empire once again in 1940. Under Soviet occupation, the Russification policies of the Soviet authorities and mass immigration from other Soviet republics (Misiunas and Taagepera 1993 (1983): 214–218, 264–295) made the Baltic peoples worry about their survival in the Soviet Russian ‘melting pot’. Therefore, the interwar period remained in the historical memories of indigenous populations as a kind of ‘golden age’, although our findings do not support the narrative of this time as being an ‘economic miracle’, as is revealed in some émigré literature and a number of publications from the post-communist years (e.g. see the influential account by Laar 2002).