The Persistent Legacy of the Fallen Empires. Assessing the Effects of Poland's Historical Partitions on Contemporary Social Norms Regarding Education

Given the prominent role of education for both individual well-being and the socioeconomic development of regions and countries, there is a need for understanding what makes an education system successful and why educational achievements differ across territories. This fundamental issue has been repeatedly tackled from various academic perspectives — inter alia by economists, sociologists, educators, and anthropologists. One promising approach is to look at regional differences in educational performance from the perspective of cultural norms. Socioeconomic phenomena are path dependent — and not only due to cumulative causative mechanisms, as described by Myrdal (1957), but also as a consequence of institutional persistence. Much empirical research shows that institutional arrangements are very abiding and continue to affect the functioning of societies long after the disappearance of the formal authorities that had imposed them.

The goal of this paper is to better understand the importance of the long-erased borders between the partitioning empires that passed through today's Poland on today's regional differences in educational achievements. As I will argue, there are reasons to believe that differing cultural norms established back in the 19th and early 20th centuries still affect the performance of students within those old borders.

The existence of educational discrepancies between the territories formerly under Austrian, Prussian, and Russian rule has been proven in earlier studies by Herbst and Rivkin (2013) and Bukowski (2018). I do not intend to question those analyses but rather to further investigate the institutional channels through which the bygone borders may still affect the quality of education. Although the effect of Poland's partitions on student achievement has already been measured, the attempts to explain this phenomenon were rather indirect and sometimes speculative, inviting further research.

The remaining part of this paper is organised as follows: Section 2 reviews the literature on how former institutions may affect different aspects of socioeconomic development, education included. Section 3 introduces the historical background of the research, explaining the link between the history of Poland in the period 1795–1918 and the institutional diversity of today's Polish regions. This section also reviews some earlier studies on the effect of the partitions on educational quality in Poland. Section 4 describes our identification strategy, including the methodological approach and data used in the analysis. Section 5 presents the results of estimations, and Section 6 is devoted to conclusions.

Cultural differences certainly exist between countries, but they are also observed within contemporary states. Even regions that have been functioning for many years under a uniform administrative and political system may still exhibit, due to their historical experience, distinct institutional and cultural features. This shows that socioeconomic development is path dependent. History matters, and we cannot understand present-day choices without reconstructing the evolution of institutions (North, 1995).

No doubt the most recognised empirical work on the impact of historical norms on contemporary social attitudes is that of Putnam, Leonardi, and Nanetti (1993), who demonstrate the link between the medieval civic tradition and the quality of social capital in modern Italian regions. Recently some of Putnam's findings were confirmed in a more formal study by Guiso, Sapienza, and Zingales (2016), which shows that Italian cities that experienced self-government in the Middle Ages indeed have a higher level of civic capital today than similar cities that did not.

In their seminal work, Acemoglu, Johnson, and Robinson (2005) reach back to the 15th century to discuss the rise of the European superpowers and to understand the reasons for the divergent development of European countries in the subsequent centuries. The authors argue that the rise of Western Europe from that century is largely due to growth in countries with access to the Atlantic Ocean and to substantial trade with the New World, Africa, and Asia via the Atlantic. However the long-term economic performance of those empires turns out to depend on the strength of domestic institutions predating the era of Atlantic trade. Where early political institutions placed significant checks on the monarchy, the growth of Atlantic trade strengthened merchant groups by constraining the power of the monarchy and helped merchants obtain changes in institutions to protect property rights. These changes were central to subsequent economic growth.

Another important aspect in the literature on the persistence of socioeconomic differences between countries is devoted to the long-term effects of colonial institutions, such as legal systems, proprietary rights in land, or forced-labour systems, on the colonised countries (Banerjee & Iyer, 2005; Dell, 2010; La Porta et al., 1997, 1998; Nunn, 2009).

It seems obvious that institutional outcomes are transmitted between generations through both formal and informal norms. Norms may persist locally even if formal institutions in a territory have changed (due to shifted borders or political change). They may also move along with migrating peoples. Simpser (2013) compares individuals who share an institutional environment but whose ancestors may have originated in different countries. This analysis shows that a proxy measure of past overall attitudes towards corruption in the country of ancestry explains substantial variation in attitudes towards corruption across the individuals in the sample, furnishing strong evidence of intergenerational transmission. One may expect that a similar mechanism is relevant for understanding the educational attainment of immigrant children.

A few years earlier Fernández and Fogli (2009) had examined the work and fertility behaviour of second-generation American women. Cultural heritage was proxied with past female labour-force participation and total fertility rates from the woman's country of ancestry. The authors show that the cultural proxies have significant explanatory power even after controlling for education and spousal characteristics.

As with many other aspects of socioeconomic life, education has been studied in the context of institutional differences emerging as a consequence of the colonial division of the world. Bertocchi and Canova (2002) demonstrate that, compared to other colonised nations, former British and French colonies tended to invest more in education after gaining independence. Huillery (2009) argues that the impact of the institutional approach in the early colonial era on subsequent socioeconomic development can be observed even within a single former colonial empire. As she shows, differences in early investments in human capital and infrastructure between the districts in former French West Africa are still reflected in the current performance of these districts with respect to education.

The projection of historical borders on the maps of educational quality is also observed in Europe and North America. In Italy, a country that was unified in the mid-19th century, academic achievement is much higher in the northern region, in line with the differences in per capita income (Lynn, 2010). The United States also exhibits significant inequality in achievement between the Northern and Southern states, again showing that students in richer regions tend to have higher educational achievements, controlling for family variables (Parcel & Dufur, 2009). The relationship between economic development and educational performance may be considered in the context of the 19th century's formation of public education systems, which happened earlier in the industrialised countries compared to lagging agricultural areas. According to some interpretations (Galor & Moav, 2006), physical capital accumulation in the process of industrialisation enhanced the importance of human capital in production and generated incentives for capitalists to support the provision of public education for the masses, triggering the demise of the existing class structure.

In light of the economic literature on the institutional determinants of various socioeconomic phenomena, the most likely channel through which culture affects academic performance is motivation to study, which can be related to student identity, social pressure, and the perceived importance of education in realising life projects. In other words, ingrained norms and social expectations affect student effort at school. The prevalence of certain norms and attitudes in a given territory may be very persistent, as they are continuously transmitted between parents and their offspring. As demonstrated by Bisin and Verider (2001), parents are motivated to transmit their preferences to children because of a form of paternalistic altruism (“imperfect empathy”). At the same time, the parents’ involvement in their children's education depends on the characteristics of the neighbourhood. The complementarity of parental effort and the social norms prevailing in the neighbourhood, demonstrated by Patacchini and Zenou (2011) on the basis of British data, may explain why the geography of academic achievements and educational attitudes is so persistent.

Sociological research reveals that social divisions and students’ identities can be a dominant influence on scholastic achievement (Akerlof & Kranton, 2002; Coleman, 1960). An education system will fail if its expectations of students are inconsistent with the value system and personal goals of students themselves.

Although it seems that identity and social norms play an important role in determining student motivation to perform at school, cultural factors may also affect motivation to study by shaping the pecuniary and nonpecuniary returns of schooling. Chauvel (1999) notes that there are three large models for the social structure in Europe: the German, the Romance, and that of the less developed Catholic countries. The German model is characterised by limited stratification and low value placed on formal education. In turn, the Catholic model involves a strong link between education and socioeconomic status.

Poland is a valuable case for research on the impact of history on socioeconomic development. The first reason for this is its turbulent history. Between 1772 and 1795, following a series of military defeats, Poland was divided among its three neighbouring powers: Russia, Prussia, and Austria. This division took place in three stages and eventually involved the entirety of Polish territory. Up until the reunification of Poland in 1918, the three regions were exposed to very different political and administrative cultures and experienced very different rates of economic growth, with the territories under Russian rule being generally less advanced economically and lagging in terms of the development of modern social and political structures.

Indeed, research shows that territorial differences referring to many socioeconomic phenomena in today's Poland clearly reflect the historical partitions. Zarycki (1997, 2000) demonstrates that the historical borders are reflected in the political choices of Polish voters. Voters in the post-Prussian territory, and generally in the west of Poland, tend to be more pro-European and liberal compared with the post-Russian region, which in turn is more conservative. Grosfeld and Zuravskaya (2015) elaborate more on this subject, arguing that the political east-west split in Poland refers in reality to voters’ attitudes towards the communist past (western regions are more anticommunist), while the liberal-conservative axis runs rather from the north to the south (from the Congress Kingdom to Galicia).

Congress Kingdom or the Kingdom of Poland is the name for Poland's territories under Russian rule used since the Vienna Congress (1815).

Becker et al. (2016) exploit the fact that in several Eastern European countries, communities on both sides of the long-gone Habsburg border have been sharing common formal institutions for a century now, ever since the demise of the Austro-Hungarian empire. Identifying individuals living within a restricted swath on either side of the former border, they find that historical Habsburg affiliation relates to increased current trust and reduced corruption in courts and police.

Some of the existing research on the persistent trace of the fallen empires on Poland's territory refers directly to education and student achievements. All of these studies followed the introduction of the countrywide standardised tests in Poland (2002) and used the test scores as the main endogenous variable. Nonetheless, these studies are quite differentiated in their approach and methods. Early works, ones that exploited the first editions of the school tests, rely mostly on descriptive and cartographic analyses (Herczyński & Herbst, 2002; Śleszyński, 2003). Later attempts employ cross-section regressions (Czapiewski & Śleszyński, 2007; Herbst, 2006), fixed effects models (Herbst & Rivkin, 2013), and Geographical Regression Discontinuity Design (Bukowski, 2018). All of the existing analyses were carried out on the data aggregated to the municipal or even higher level.

Summarising the existing research on the geography of educational quality in Poland, most authors observe that, even when controlling for different educational inputs and local characteristics, Galicia (the partition once under Austrian rule) outperforms other regions of Poland in terms of academic achievements, as measured by the standardised school tests (Bukowski, 2018; Herbst, 2006; Herbst & Rivkin, 2013; Herczyński & Herbst, 2002). Some studies also acknowledge that the western territories of Poland, formerly under Prussian rule, perform surprisingly poorly, taking into account their dominant role in the domestic economy (Herbst & Rivkin, 2013).

The conclusions from the literature review are consistent with the pattern in Figure 1 showing the municipal averages of ninth-grade students’ test scores in mathematics (2015). The achievements in southeastern Poland, which belonged to the Austro-Hungarian empire until 1918, are high compared to other regions of Poland. The average scores by the historical partitions are given in Table 1, again showing that students in the ex-Austrian territory outperform those living in other locations. The gap between the mean test scores in mathematics achieved in the Austrian and Prussian partitions accounts for a standard deviation of about 0.25, which is more than the post-Austrian territory's advantage over the Russian partition (0.13) and less than the gap between the Austrian partition and the territories acquired by Poland after World War II (0.34). The poor performance of students in the new (post-German) territories in the west and north of Poland is noted in most of the existing studies. In this context, a recent work by Becker et al. (2018) has yielded findings which at first glance may look surprising — namely, in showing that descendants of the forced migrants (who account for almost 30% of the population in the acquired areas) tend to have on average one extra year of schooling and are driven by a higher propensity to finish secondary and higher education. However, as shown by the authors, this advantage is likely to be offset by the very low educational attainment of the autochthonous population, and therefore the findings of Becker et al. are not in contradiction with the earlier studies on the geography of educational achievements in Poland.

Figure 1

Although the higher quality of education in the former Austrian partition is commonly recognised, things are worse in terms of understanding the mechanisms behind this phenomenon. The channels of the persistence of regional differences in educational achievement in Poland are a challenging and unsolved research puzzle. Herbst (2006) considered regional differences in returns to education as a promising subject for further research. Herbst and Rivkin (2013) elaborated on this channel, arguing that investment in education (and eventual migration) is generally more profitable in the eastern regions of Poland, which contributes to higher academic achievement, and this contrasts with the low economic performance of this area.

More recently, Bukowski (2018) has attempted to explain the long-lasting regional differences in educational quality in Poland via the concept of identity and student motivation. Bukowski argues that the interaction between institutional and individual identity might be crucial for the formation of human capital. In the Austrian educational system, the positive framing of Polish identity (including the use of Polish as the language of instruction) might have created a positive social norm regarding education among the Poles. Conversely, in the Prussian and Russian systems, Polish identity had a negative framing, and it thus might have led to either a neutral or negative social norm.

The degree of cultural autonomy, as a factor shaping a positive attitude towards education in 19th-century Galicia, is emphasised in the recent article by Herbst and Kaliszewska (2017). The authors also elaborate on other institutional arrangements distinguishing the former Austrian partition from the territories ruled by the other imperial powers: the high social status of teachers, the important role of the school as a means of social advancement, and much higher scholarisation rates in rural areas (especially compared to territories under Russian rule).

Although the works quoted above generally support the view that the persistent discrepancy in educational achievements between the partitions is caused by different social norms about education, the attempts to verify this claim amount to either qualitative insights or indirect quantitative analyses based on circumstantial evidence (Bukowski, 2018; Herbst & Kaliszewska, 2017; Herbst & Rivkin, 2013).

The most recent piece of such “patchwork” analysis is provided by Bukowski (2018), who demonstrates that the achievement gap between the Austrian and Russian partitions is higher with respect to sixth-grade test scores (which is a low-stakes test) than in the case of ninth-grade test scores (which are high stakes because they are used as an admission criterion to high school). According to that author, this shows that the high performance of students in Galicia is driven by long-standing social norms regarding education rather than by expected returns from education. To investigate these norms, Bukowski eventually uses data from three different surveys (of which two are large, general socioeconomic surveys) turning selected education-related questionnaire items into dependent variables and then regressing them on partition “dummies” and control variables. The surveys are not specifically designed to study the norms regarding education, and they do not allow us to deepen the analysis (as they typically include only two to three relevant questions). They are also based on different methodological approaches and have different targeting and sampling. For example, only one of the surveys is addressed to the parents of school-aged children, while the two others were conducted among the adult population.

The results are somewhat puzzling. They suggest that, compared to the other partitions, people from the former Austrian empire are more likely to declare that education is important in their life and more likely to support higher government spending on education. On the other hand, they are less likely to desire higher education for their children and less trustful of education as a factor in lifetime success. Bukowski convincingly argues that overall the results can be interpreted as evidence for a positive social norm regarding local educational institutions in the former Austrian territory.

Differences between the early educational systems in the three partitions, which may have resulted in different social norms regarding education, are well described in the works of Polish historians (see Dybiec et al., 2015; Kucha, 1982; Możdżeń, 2006; Szymański, 1983; Truchim, 1967–1968; Wroczyński, 1980; Żniński, 1914). In the Prussian partition, where compulsory schooling had been formally introduced by 1819 following the ordinance of Friedrich Wilhelm III, both the network of schools and the curricula were centrally designed. Not only were schools meant to raise the level of useful qualifications in society, but they were also considered an important measure to integrate the annexed territories into Prussia. Prussian schools, teaching exclusively in German, very quickly outnumbered the already existing Polish institutions, and even the latter became obliged to offer full instruction in German in addition to classes held in Polish. In time, particularly after Otto Bismarck became German chancellor, Germanisation pressure grew stronger. Between 1873 and 1887 the Prussian authorities gradually limited instruction in Polish, and in 1887 they withdrew the Polish language as a separate subject even in elementary schools. From 1887 on, secondary schools became obliged to use German as the language of instruction in religion classes, as well. In the 1880s the government also monopolised teacher nominations, something hitherto having usually been the task of local authorities.

In contrast to the Prussian partition, the Russian authorities never introduced compulsory schooling in the seized territories; this was instituted only after Poland had regained its independence in 1918. Although the government of the Russia-dependent Kingdom of Poland made some attempts to develop a school network, the outcomes were rather limited. According to Kmiecik (1969), in 1904 in the rural parts of the kingdom, there were only 32 children attending elementary school per 1,000 inhabitants — much fewer compared to the European part of “proper” Russia (42.9). Among the men recruited to the Russian army from the kingdom's territory in 1886, 83.1% were illiterate.

In the early 19th century, the schooling system in the Russian partition was underdeveloped compared to the other partitions, but it was largely autonomous. The situation changed dramatically after 1863, as a consequence of the failed national uprising (the January Uprising). Following the Jugenheim edicts signed by the Russian czar Alexander II, the Polish language was eliminated from schools, which henceforth were to use only Russian textbooks. Reading and writing were now taught using “civil script” (a form of the Cyrillic alphabet), and school instruction included Orthodox religious singing. These reforms naturally met with strong opposition in Poland, as the population was overwhelmingly Catholic, and it had always relied on the Latin alphabet.

Another important measure of Russification via the school system involved replacing Polish teachers with Russians. As a result of this policy, in 1903 Poles accounted for only 22% of all teachers in secondary schools for boys.

The Russian government's tightening of schooling policies contributed to the emergence of secret courses held in Polish. Among the territories of partitioned Poland, underground schooling was most developed in the Russian region, where it covered all levels of education, including seminars for teachers.

Compared to the Russian and Prussian territories, the political climate surrounding the school system in Galicia was very different. The law on compulsory schooling for children aged 6 to 14 was announced in 1869, 50 years after the corresponding regulation in Prussia and 50 years prior to the freshly independent Polish state's introduction of compulsory schooling in the former Russian partition. From the very beginning the Austrian law stated that public education was free and that it was provided in the children's mother tongue.

What distinguished teachers from Galicia from those in other partitions was not only the dominant share of Poles but also the large share of teachers of peasant origin and the minor participation of the gentry class. The teaching profession offered the opportunity for social advancement and guaranteed a path to prestige in the local community.

Cultural and administrative autonomy, Polish instruction at all tiers of schooling, and relatively close relationships with teachers together created important, positive incentives for Polish students. Another crucial factor was the existence of universities in the region (Lviv University and the Jagiellonian University in Kraków), which allowed the most talented students to continue education after graduation from secondary school. At this time Galician universities were the only ones operating within the territory of partitioned Poland and teaching in Polish, as the University of Warsaw (established 1816) was closed by the Russian government after just 15 years of operation and reopened except for a short period between 1862 and 1869.

Overall, given the differences in the political, cultural, and administrative conditions in which the public education systems were formed within the divided Polish territory, the hypothesis of the divergence of education-related social norms that may persist today seems justified.

The goal of this work is to start where the previous quantitative research left off, that is, to directly address the questions concerning the norms through which history may affect present-day educational performance. The main hypothesis is that residents of the former Austrian partition have more positive norm regarding education compared to those living in other regions of Poland. A secondary hypothesis is that, due to different historical experience, they may also perceive education more in terms of a formative process and have less utilitarian expectations about it.

I use the results of the self-designed survey of parents having 15–16-year-old children, which aimed to assess the attitudes towards education, institutions, and regional historical legacy in the three historical partitions of Poland. The conceptual framework of the analysis and the survey itself are described in the following section.

In this work I have investigated the institutional legacy of the Austrian empire as the possible explanation for the comparatively higher academic achievements of students in the southeast of Poland. The persistence of former institutions and their impact on present-day socioeconomic life is well documented in interdisciplinary research the world over. However, we do not fully understand the channels through which the historical experience determines within a country the regional diversity that we observe today, and in particular the diversity of educational achievements.

In the first stage the RDD was used to verify the existence of educational discrepancies between the neighbouring territories that belonged to different countries in the period 1795–1918. The results (discussed in Appendix 1) confirmed the findings of the earlier studies showing that students from Galicia perform significantly better than those in the former Russian partition, while there is no significant difference between the former Prussian and Russian territories. My major goal was to directly address the questions concerning the institutional channel through which history may affect present-day educational performance. Based on suggestions from the interdisciplinary literature on the subject, I decided to investigate the social norms regarding education in different partitions as a possible explanation for the observed gap in achievements. I conducted a self-designed survey of parents having 15–16-year-old children, aiming to assess their attitudes towards education, public institutions, and regional historical legacy. Survey items were then used to classify parents according to their compliance with three different norms related to education. Next I performed logit estimations, and postestimation prediction of probabilities of compliance with particular norms towards education, as conditional on living in the former Austrian partition.

All three norms proved significantly different between the former Austrian partition and the remaining territories of Poland. The most pronounced difference was in trust in educational institutions. Parents in Galicia (the former Austrian territory) were about 10% more likely to express trust in schooling than did their counterparts on the other side of the former border. This result proved robust for the sensitivity tests. Adding variables depicting parents’ opinions on their child's school, their attitude towards the partitions as historical events, and their general openness (general trust question) did not depress the estimated impact of the Austrian legacy on trust in educational institutions.

Besides having more trust in schooling, parents in Galicia are also 7% more likely to care about the formative role of school and 4% less likely to recognise the monetary returns on education.

The results are very much in line with findings by Becker et al. (2016), who demonstrated that living in regions formerly under Habsburg rule increases the level of trust and reduces the perceived corruption in courts and police. In this context schools may be considered another type of public institution. The findings also support the earlier suggestions of Bukowski (2018) and Herbst and Kaliszewska (2017), who point to social norms and the degree of identification with public institutions as possible sources of higher educational performance in the former Austrian partition. However, different from what is suggested by these studies, a positive attitude towards education in Galicia is not primarily driven by personal experience, and it is not limited to the familiar, local school system, but it applies to education in general. Similarly, a higher trust in education is not associated with the promise of material success or social advancement but rather stems from the conviction that the school is important for personal development.

The research shows that the intergenerational transfer of long-inculcated norms may have a significant impact on territorial patterns of socioeconomic development, even if the administrative and political divisions that were at the foundation of these norms vanished a long time ago. This supports the findings of Bisin and Verdier (2001) and Patacchini and Zenou (2011) on how parental effort combined with local social norms contribute to the tenacity of regional differences in educational performance. With respect to educational policy, our results show that traditional measures of coping with inequalities, relying on redistribution of resources, providing extra teaching hours, or intensified evaluation of achievements may be ineffective if deeply rooted local norms regarding education are not taken into account and if policies fail to address multiple generations, and not just students at schools.