Changing Morals, Changing Attitudes: Educational ...
Changing Morals, Changing Attitudes: Educational development in the Netherlands Indies and Belgian Congo, ca. 1880-1960
Ewout Frankema Utrecht University
Paper prepared for the workshop
Colonial Extraction in the Netherlands Indies and Belgian Congo: Institutions, institutional change and long term consequences
3-4 December, Utrecht
Work in progress
Abstract This paper focuses on the comparative nature and pace of educational development in the Netherlands Indies and Belgian Congo after the adoption of the ethical policy by the Dutch government (1901) and the annexation of Leopold s Free State by the Belgian state (1908). The organization of the colonial education system differed fundamentally in both colonies. In the Netherlands Indies the colonial state was the main driving force behind the spread of mass education. In Belgian Congo the education system was developed by Christian missionaries, and in particular Belgian Catholic missionaries. This paper explores the implications of these organizational differences for the development of mass education, the allocation of educational funds and speculates about the long term consequences for post-colonial development.
I would like to thank Bas van Leeuwen for supplying me with the underlying data of his statistical appendices on education in the Netherlands Indies, see Bas van Leeuwen (2007). 1
1. Introduction
The development and diffusion of formal primary education was a key feature of the civilization mission envisaged by Western colonial powers in Africa and Asia in the late 19th to mid 20th century. Against the background of the excessive exploitation and coercion of indigenous labour under the cultivation system (het cultuurstelsel) in the Netherlands Indies (ca. 1830-1870) and the domanial regime (het domaniale stelsel) in the Congo Free State (1891-1908) such paternalistic ideals got a sharp political and moral edge. This paper studies the comparative nature and pace of educational development in both colonies after the major political reforms that intended to end illegitimate exploitation and to enhance the moral elevation and material living conditions of the colonial population. While the intentions of both colonial administrations in the Netherlands Indies and Belgian Congo were more or less comparable, school enrolment rates at the end of the colonial era varied considerably, albeit in a different way than one might have expected: despite the fact that the Dutch colonial administration invested much more in education, and had done so for a much longer time as well, primary school enrolment rates in the Congo were approximately twice as high at the eve of World War II. Differences in enrolment rates are meaningful because scholars have stressed the persistent developmental effects of colonial educational legacies. In sub-Saharan Africa, for example, colonial primary school enrolment rates have been found to positively relate to post-colonial literacy rates, governance quality and per capita GDP (Brown, 2000; Lloyd et al., 2000; Cogneau, 2003; Bolt and Bezemer, 2009). In East and Southeast Asia the contribution of formal education to the post-war growth acceleration is widely acknowledged (Worldbank, 1993; Birdsall et al., 1997). Despite the critical notes of some scholars (Booth, 2003), there is consensus that education played an important role in Asian miracle growth and that colonial policies have laid the foundations of the post-colonial education system, for the better or worse (Booth, 2007). In their own region the colonial educational legacies in the Netherlands Indies and Belgian Congo stand out differently. The Dutch have often been accused of neglecting their responsibility to educate the indigenous population (de inlanders) (van der Veur, 1969, p. 17).1 Only in the 1940s did the expansion of school enrolment levels gain momentum, when the influence of the Dutch was already greatly reduced (van Leeuwen, 2007). The education record of Belgian Congo, on the other hand, looks really impressive in comparison to its African neighbors. Also in comparison with the average British African colony, which were 1
We will refer to the anachronistic, but more neutral term Indonesians henceforth. 2
regarded as the more advanced, the Belgians seem to have operated quite effectively in the Congo, developing a system of mass education with high enrolment rates and limited public resources. Indeed, up to 1940 the Dutch colonial government devoted a higher share of its total expenditure budget to education and spent much more in absolute terms. Public investments in Indonesia education had much more time to accumulate and mature, simply because the Dutch had established state control over large parts of the Indonesian archipelago since the early 19th century, whereas the Belgian government only accepted formal responsibility in 1908, when King Leopold II was forced to give up his stakes in the Congo. In the remainder of this paper it will be argued that this puzzle can only be understood when looking at two critical factors. Firstly, the relative importance of private and public education: Christian missionary societies, and especially the Catholic denominations, were the driving force behind the spread of mass education in the Congo, but played only a minor role in the Islam oriented Netherlands Indies. In the Netherlands Indies the colonial state was the main actor in both developing and financing mass education. Second, the role of what I would like to call the compartmentalisation of education for separate target groups, which was a defining characteristic of the Dutch education system, but of subordinate importance in Belgian Congo. At the end of the paper I will offer some speculative thoughts about the implications of the colonial educational legacy for long run social and economic development in both areas.
2. The comparative development of primary education
Before focusing on a direct comparison of school enrolment rates between the Netherlands Indies and Belgian Congo we compare both countries to their closest neighbors in respectively Southeast Asia and Central Africa. We will confine ourselves to primary school enrolment rates, because these constituted between 95 and 99% of total enrolments in all the countries under discussion prior to 1950. Table 1 shows the Congolese primary school enrolment rates for the age group 5-14 in 1950, together with the figures for its neighbors and the unweighted averages of the British, French and Portuguese African empires as a whole. The table shows that roughly one-third of the specified age group in Belgian Congo was involved in at least the most basic type (i.e. the lower grades) of elementary education. With this figure Belgian Congo ranked among a select 3
group of mainly British African colonies such as Zambia, Zimbabwe, Nyasaland, Mauritius and Lesotho, with gross enrolment rates (5-14) surpassing 30% before 1950. Such rates were higher than the British African average (24,2%) and far higher than the French (9,4%) or Portuguese African (8,5%) average. Enrolment rates in Congo s eight neighbor countries underline an optimistic interpretation of the Belgian colonial educational system.2
Table 1: Gross primary school enrolment rates (age 5-14) in Belgian Congo, its direct neighbours and British, French and Portuguese Africa, ca. 1950 GER age 5-14
GER age 5-14
Belgian Congo Congo Brazaville (Fr.) Central African Rep. (Fr.) Sudan (Br.) Uganda (Br.)
33 24 16 6 25
Tanzania (Br.) Zambia (Br.) Angola (P.) British Africa average French Africa average
26 35 1 24,2 9,4
Ruanda-Burundi (Be.)
11
Portuguese Africa average
8,5
Sources: Frankema (2010); UNESCO (1965), Statistical Yearbook 1964, Paris Notes: Averages are unweighted, including 16 British colonies, 16 French and 4 Portuguese. For some countries (Zambia, CAR) we used the more familiar post-colonial instead of the colonial names (Northern Rhodesia, Oubangi-Chari).
The Netherlands Indies offer a different picture. Table 2 replicates the data from Furnivall s (1943) classic comparative study Educational progress in Southeast Asia. In the final years of effective Dutch rule in the Indies, i.e. the late 1930s, gross primary school enrolment rates were considerably lower than in neighboring countries such as British Malaya, Japanese ruled Formosa (Taiwan) or the US administered Philippines. In Thailand, the only independent country in this group, the number of people enrolled in primary school as a proportion of the total population was about three times as high as in the Netherlands Indies. Only in Burma and French Indo-China enrolment rates were more or less similar. That the share of unrecognized schools, which mainly consisted of Islamic (boarding) schools (madrasahs, pesantrens), was relatively high in the Netherlands Indies (an estimated 16,4% in 1938) does not alter this conclusion. Transforming Furnivall s estimates into gross enrolment rates of the age group 5-14 (third column of table 2), enrolment rates appear a little higher than in Central Africa, but with a similar large degree of intra-regional variation.3
2
Note that the gross enrolment rate of the official primary school age group (6-11) was even considerably higher (i.e. 55%) because it reduces the denominator with 40%. 3 We assume that the 5-14 age group constitutes 25% of the total population. This is the average of the 22-27% range for pre-war developing countries suggested by Benavot and Riddle (1988, p. 199). 4
Table 2: Gross enrolment rates of total population in Southeast Asia, 1936-39 Recognized primary schools
All schools
Recognized primary schools GER age 5-14
Netherlands Indies Malaya (Br.) Burma (Br.) Thailand Indo-China (Fr.) Taiwan (Jap.)
3,4 6 3,9 9,7 2,1 11,2
4 7,8 5,5 10,7 2,5 11,4
13,6 24 15,6 38,8 8,4 44,8
Philippines (US)
10,8
11,5
43,2
Sources: (Furnivall, 1943, pp. 111-2); (Benavot and Riddle, 1988) Notes: All schools, include the unrecognized schools as well as a virtually negligible share of students enrolled in secondary and tertiary schools.
Figure 1 places the estimated primary school enrolment rates in both countries in a temporal perspective, showing that enrolment rates in Belgian Congo were substantially higher than in the Netherlands Indies between 1930 and 1950. Before we further analyze this graph, however, we need to say a little bit more about the definition of primary education. The timeseries for the Netherlands Indies are taken from van Leeuwen (2007, pp. 264-6). These include the enrolment of Indonesians as well as Europeans (mainly Dutch) and other Asians (mainly Chinese), but they exclude a considerable number of Indonesian attending the unrecognized (and non-subsidized) Islam schools. Since the colonial administration only had a vague idea about the real size of private religious education, the official statistics tend to underreport the number of unrecognized schools, if they provide them at all. There may be good reasons to exclude the unrecognized schools from a long run timeseries of enrolment rates as van Leeuwen suggests, because the religious curriculum was not designed with the purpose to enhance secular knowledge or practical skills. On the other hand, in so far Koran study enhanced the spread of literacy (and there is evidence that it did) the economic side-effects of religious education cannot be neglected (Reid, 1988, p. 218; van Leeuwen, 2007, p. 49). Nevertheless, the role of Islam schools gradually declined as the colonial state initiated the development of a mass education system. When we combine the official enrolment estimates of the unrecognized schools reported in the Koloniale Jaarcijfers of the 1880s (ca. 250.000-300.000) to Furnivall s estimate in 1938 (ca. 450.000), the relative share of the traditional Islamic schools would have gone down from ca. 75% to 16% of total primary enrolment levels. The dotted line in figure 1 reflects this trend. The relative decline must have after the adoption of the Ethical Policy (de Ethische Politiek) during the first decade of the 20th century. As of 1907 the colonial administration 5
started to support the establishment of village schools, the so-called desa schools, as a basic and cheap indigenous alternative for the more exclusive European schools. In the desa schools students were taught in the vernacular, instead of Dutch, which was the norm in European education (Brugmans, 1938, pp. 302-317). These village schools were the major driver behind increasing enrolment rates between 1907 and 1940.
Figure 1: Gross enrolment rates (age 6-11) in Belgian Congo and the Netherlands Indies, 1880-2000 160 140 120 100 80 60 40 20 0 1880
1890
1900
1910
Indonesia
1920
1930
1940
1950
1960
1970
Indonesia, incl. non-recognized schools
1980
1990
2000
Congo
Sources: For Indonesia (van Leeuwen, 2007); Jaarcijfers voor het Koninkrijk der Nederlanden (Kolonieën) 1880-192x and Statistisch Jaaroverzicht van Nederlandsch-Indië 192x-1939; (Furnivall, 1943); UNESCO, Statistical Yearbooks, various issues 1964-1999; For Congo (Depaepe and van Rompaey, 1995, P. 247) and Annuaire Statistique de la Belgique et du Congo Belge 1911-1960
But figure 1 also indicates that even when we take the unrecognized schools into account, gross primary school enrolment rates probably never surpassed 20% before 1940. The big push to mass education occurred after the retreat of the Dutch: enrolment rates jumped from 20 to 80% of the school age population in less than two decades between 1945 and 1962. Around 1980 the Suharto administration had reached the historical milestone of universal primary education. In the Congo Free State (1885-1908) and the first decades of formal Belgian rule in the Congo (1908-1960) education statistics were never systematically collected. Only during the period 1930-1960 the Belgian Annuaire Statistique provides annual information on the
6
number and type of schools and the number of students enrolled. Schooling was the exclusive domain of Protestant and Catholic missionary societies until World War II and remained highly dependent on the missions until the end of the colonial era. In theory, all missionary societies were free to settle and develop activities in the Congo as stipulated in the international agreements following the Berlin conference. In practice, the treatment of various missionary societies was uneven. Protestant missions did not receive any financial support before 1948, while Catholic missionaries were increasingly supported with state subsidies and could also rely on the less visible advantages of close administrative contacts. The fact that Protestant missionaries in the Congo were nearly all foreigners , mainly from Britain, Germany and the US did not make their position stronger. Another problem in estimating gross enrolment rates in Belgian Congo concerns the lack of reliable population data. Colonial population censuses were complicated by the sheer impossible task to survey an immense territory with hardly any physical infrastructure outside the main waterways. The devastating pandemic of sleeping sickness during the first decade of the 20th century and the death toll as a result destructive rubber policies during, roughly, the second half of Leopold s reign, make it extremely difficult to estimate the size of the population at any given point in time before 1950. The short-cut we have used here is to backward extrapolate the post-1950 population series with a fixed annual growth rate of 1,5%. We don t think this produces a very reliable time series, but we do believe it is better than taking the prewar census estimates, which almost certainly underestimate the population. Before the 1930s the enrolment levels are thus based on a few scattered observations. Referring to the work of Liesenborghs (1939), Depaepe and van Rompaey (1995, p. 38 & 247) estimate the number of students in 1908 at ca. 46.000, of which ca. 27.000 were enrolled in Protestant and 19.000 in Catholic mission schools. After 1908 the numbers increase rapidly to approximately 100.000 in 1913, 150.000 in 1921 and 350.000 in 1929. As Liesenborghs already noted, this was a spectacular rise (1939, p. 484). These figures include both the recognized schools, which were liable to state subsidies and the unrecognized schools which were entirely operated on private funds. As educational expenditure budgets increased, especially after the mid-1920s, Catholic mission schools were increasingly recognized by the colonial administration. The substantial minority of Protestant mission schools started to receive subsidies in the late 1940s.
7
3. The impressive rise of enrolment rates in Belgian Congo
Why did enrolment rates rise so rapidly in Belgian Congo after 1908? For an answer to this question it is important to note that the colonial government of Belgian Congo only became actively engaged in the spread of mass education after the Second World War. Virtually all education was missionary education and the Catholic mission schools maintained their dominant position until the end of colonial rule, despite the development of public lay education (neutral onderwijs) and the incorporation of Protestant missions in the colonial education subsidy program after 1948 (Depaepe and van Rompaey, 1995, pp. xx). The formal transition from the Congo Free State to Belgian Congo in 1908 produced a more stable and secure political-institutional context, which set the doors wide open for a rapid expansion of missionary fieldwork in the Congo. The timing and nature of this development conformed to a more general pattern in colonial Africa, where missionary activities preceded the establishment of formal European rule, but only really expanded thereafter.4 A second factor, which was also not unique for the Congo, was the rapid improvement in tropical medicine, especially in the field of malaria eradication. Despite the widespread use of quinine in the late 19th century, working and living in the Congo remained an extremely risky undertaking for Europeans. The lethal disease environment was the major reason why Belgian politicians, including king Leopold himself, refused to actually visit the dependencies in Africa. As scientists discovered the true source of malaria (the mosquitoes, instead of the bad airs mal aria of swamps) in the early 20th century, major steps in malaria prevention (rather than curing) were taken, which also freed the way for further missionary expansion. What was really unique in the Belgian Congo was the comparatively high proportion of foreigners residing in the colony. Richens has shown that in a sample of 33 African colonies Belgian Congo counted by far the highest number of white administrators in the late 1930s: 728 against an African colonial average of 94! Of course, the Congo was a vast area containing one of the largest populations in the region, but this does not explain the entire gap. In British Nigeria, which had roughly twice as many inhabitants, the number of administrators was less than half (353). Indeed, the number of administrators per capita was comparatively high in Belgian Congo (Richens, 2009, pp. 21 & 64-5). This was also the case 4
Frankema has estimated that enrolment levels in British Africa underwent a tenfold increase, from ca. 100.000 to over 1.000.000 between 1900 and 1938. In East and Central Africa the acceleration in educational expansion started right after the formal British occupation of new territories (Frankema, 2010).
8
for the number of missionaries. Oliver has estimated that the total number of missionaries (Catholic and Protestant) in the whole of East Africa never exceeded 3.500 (Oliver, 1962, pp. 231-245). Table 3 shows that the numbers in Belgian Congo were not only higher already before 1940, but also continued to increase thereafter.
Table 3: Absolute numbers and indices of missionary presence and students enrolled in Belgian Congo, 1908-1957 (1938 = 100) 1908
1929
1938
1950
1957
Missionaries
500
2.500*
3.732
5.336
7.205
Index
13
67
100
143
193
46.000
350.000
562.851
970.372
1.718.931
8
62
100
172
305
School enrolment Index
Sources: 1908 from (Stengers and Vansina, 1985); 1929 (Depaepe and van Rompaey, 1995); 1938-1957 Annuaire Statistique de la Belgique et du Congo Belge 1911-1960. Notes: 1929 number of missionaries is guesstimate based on 1931 figures from Annuaire Statistique de la Belgique et du Congo Belge 1911-1960.
Table 3 shows that primary school enrolment levels in Belgian Congo were positively correlated with the number of Christian missionaries. This was not a linear relationship, as the number of foreign missionaries increased with a factor 15, while enrolments increased with a factor 37 between 1908 and 1957. Indeed, the expansion of missionary education also entailed the Africanization of missionary field work (Oliver, 1962; Frankema, 2010). In 1958, for instance, there were no less than 6.934 schools run by Protestant missions in the Congo, which were managed by ca. 1.550 Protestant missionaries. The far majority of teachers and school heads were Congolese. In other words, when it comes to explaining the spectacular rise in enrolment rates in Belgian Congo, we not only have to understand the comparatively high rate of missionary activity in the Congo, but also the apparent enthusiasm with which the Congolese embraced Christian education. The answer to the first question is partly related to the fact that the Congo was a Belgian colony: together with the mandate over Ruanda-Urundi granted by the UN after World War I, the Congo was Belgium s only overseas possession. Belgian citizens with colonial ambitions thus concentrated their efforts almost exclusively on the Congo. In the geographically much larger and far more dispersed British and French empires, British and French citizens diversified their attention. This small colonizer argument of course also applied to some extent to the Netherlands Indies, but with the crucial difference that in terms 9
of what we could call the population support ratio , that is the ratio of colonial versus metropolitan population, the Dutch were administering a much larger colony. Table 4 compares the population support ratio of Britain, France, Belgium and the Netherlands, showing that the difference was around a factor 5.
Table 4: The metropolitan-colonial population-support ratio in the British, French, Dutch and Belgian overseas empires, ca. 1938
United Kingdom France Netherlands Belgium
Metropolitan population 1938
Colonial population 1938
Population support ratio C/M
47,5 42 8,7 8,4
358,1 65,2 68,3 13,9
7,54 1,55 7,85 1,65
Source: Statistical Abstract of the British Colonies 1938-1940; Annuaire Statistique de la France; Maddison 2009.
Clearly, there were other factors playing a role as well. After all, the differences in the population support ratio between France and Belgium were negligible. One of these factors has to do with metropolitan policy considerations. From the very inception of the Congo Free State king Leopold encouraged Catholic missionary activity as part of his strategy to physically occupy the Congo basin with minimal public funds. Missionary education was needed to train Congoleze for the army (La Force Publique) and the administration, but they brought in their own funds for their evangelical labour. The initial key role of the missions in the colonial administration created path dependency. The French were more reluctant to give so much power to missionary societies, especially against the background of the heated debates about the separation between state and church during the late 19th century (Cowan et al., 1965; Subramanian, 1979). Yet, by refusing missionary help in developing colonial education, the colonial government deliberately constrained its reach. Another part of the explanation for low enrolment rates in French Africa, in comparison to Belgian Congo, relates to the fact that the French had occupied most of the Islamic areas, particularly in West Africa. Similar to Lugard s policies in British Northern Nigeria, the French did not support Christian missionary activity in the Islamic heartlands in fear of burdening their relationship with the Muslim population and their representatives. Education thus became the exclusive domain of the neutral colonial state from the onset. The state capacity to spread mass education was limited, however, not in the least place because of the small numbers of colonial administrators. In French colonies where Islam was less 10
influential, such as Madagascar or French Congo, Christian missionary activity was less problematic and far more widespread and school enrolment rates in these colonies were considerably higher (Frankema, 2010). The situation in French West Africa was comparable with the Netherlands Indies in the sense that the state took the initiative to spread education and strictly monitored, if not actively discouraged, Christian missionary activity. Of course, Islamic education provided an alternative to public lay education or missionary schooling, but the number of Islamic schools remained limited and did not develop rapidly over time. Public education in the metropolitan language (French or Dutch), rather than vernacular education or traditional Islamic education, offered the real opportunities for social mobility.
4. Educational finance and government expenditure, ca. 1880-1960
While private missionary education remained the norm in Belgian Congo after 1908, public education dominated in the Netherlands Indies and the colonial administration only further enlarged its control over the education system after 1907. These different patterns of educational development in the first half of the 20th century are also reflected in the sources and allocation of educational finances. Figure 2 shows the percentage share of the regular government expenditure budget devoted to education in the Netherlands Indies and Belgian Congo between 1880 and 1960. The graph shows that government spending on education in the Netherlands Indies had a longer history and that, until 1940, the share in total expenditure was higher than in Belgian Congo. The rapid catch-up in educational expenditure in Belgian Congo between 1908 and 1922 is a little misleading because the size of the Congo government budget remained small. Figure 3 shows educational expenses per head of the population converted to current US $ using the official exchange rate. This is a rather crude measure, because exchange rates can deviate substantially from purchasing power parities, but it suffices to make the point that absolute expenditure levels in Belgian Congo remained modest (below 10 dollar cents per capita) and that the Dutch in the 1930s still spent three to four times as much on education as the Belgian colonial government.
11
Figure 2: The percentage share of regular government expenditure devoted to education in the Netherlands Indies and Belgian Congo, 1880-1960
0,20
0,15
0,10
0,05
0,00 1880
1890
1900
1910
1920
1930
1940
1950
1960
Sources: For Indonesia (van Leeuwen, 2007); For Congo Annuaire Statistique de la Belgique et du Congo Belge 1911-1960.
Figure 3: Government expenditure on education per capita in the Netherlands Indies and Belgian Congo, 1880-1960 (current US $, y-axis log scale) 10,0 Indonesia
Congo
1,0
0,1
0,0 1880
1890
1900
1910
1920
1930
1940
1950
1960
Sources: For Indonesia (van Leeuwen, 2007), van Leeuwen has included expenditure on the state level and provincial level, as published in the reports of the Dutch Indies Education Committee; For Congo Annuaire Statistique de la Belgique et du Congo Belge 1911-1960. Notes: Dutch florins and Belgian Francs converted to US Dollars using official exchange rates. 12
Figure 4: Government expenditure on education per student enrolled in the Netherlands Indies and Belgian Congo, 1890-1960 (current US $, y-axis log scale) 100 Indonesia
Congo
10
1
0 1890
1900
1910
1920
1930
1940
1950
1960
Sources: see figures 2 and 3.
Because enrolment rates in the Netherlands Indies were so much lower, the differences in government expenses per student enrolled, as shown in figure 4 (y-axis log scale!), even fluctuated around a rate of 1:10 between 1910 and 1940. The horizontal trend in both countries indicates that the increasing size of the education budget was mainly consumed by expanding school enrolment rates, rather than an autonomous rise in the available budget per student enrolled. In fact, given the substantial amount of interwar inflation, the nominal persistence of education expenses suggests a long run decline in real terms. Although the average sum of ca. 10 US $ per student in the Netherlands Indies was quite respectable, we have to keep in mind that the distribution around the average was highly unequal. In 1929, for instance, about half of the total education budget was spent on European education, which was reserved for only about 10% of the total enrolled (van Leeuwen, 2007). The ethnic distribution of public resources between the Europeans on the one hand, and the Indonesians and Chinese on the other hand, became more egalitarian over time however. In the late 19th century European education was almost exclusively restricted to Dutch citizens, but by the late 1930s ca. 75% of the students enrolled were of Indonesian or Chinese origin (van Leeuwen, 2007). Needless to say, these were mainly children from higher social ranks.
13
The political choices regarding the allocation of government expenses also had important implications for the mission schools in the Netherlands Indies. In the last quarter of the 19th century around a quarter of the recognized students on Java and Madura went to a Christian mission school (zendingschool) and this share was even much higher in the outer areas (Hartgerink, 1942, p. 39). Yet, the financial resources of the missions were limited and the colonial administration refused financial support. The government preferred to annex mission schools which were unable to stay in business, rather than subsidize them. In the first decade of the 20th century the colonial government deliberately transferred part of its educational responsibility to the missions in some of the outer areas, such as the Moluccas, under the promise of state subsidies (Hartgerink, 1942, pp. 141-149). Hence, the missions were used to tie the loose ends of the public educational system together. In Belgian Congo the missionary societies were not only carrying the colonial education system, they were also its primary financiers. The rise in government expenditures during the second half of the 1920s, which also occurred in British Africa in response to the influential Phelps Stokes report (1922), only sufficed to compensate for the increasing costs of mission school expansion. As stated above, only Catholic schools could hope to receive some additional funding, but even among the latter the government funds were distributed rather selectively. The central schools in the major population centers received most of the money, while the village schools in the immense hinterland received next to nothing. The reason was that the central schools had to be better equipped because of their importance for the colonial order as such. These schools trained the clerks and typewriters necessary for administrative tasks, the medical assistants, technicians and agricultural experts necessary to develop the economy and also the catechists and soldiers necessary to evangelize and secure order (Depaepe and van Rompaey, 1995, pp. 63-69). The financial involvement of the Belgian Catholic missions in the Congo played a crucial role in the 1925 convention with the Belgian state. The catholic schools would receive more financial support in exchange for extra efforts to spread mass education across the colony. The contract was not signed until 1929, however, because the missions rejected the proposal of a state inspection that would monitor and guarantee the quality of colonial education. The fact that the Catholic missions eventually won this battle tells a lot about the balance of power: the Catholics exerted a de facto monopoly on the development of colonial education which they held on to until 1948 (Depaepe and van Rompaey, 1995, pp. 60-63). This does not mean that the existing financial problems were resolved or even relieved. The number of schools and students expanded at a much faster rate than the home 14
country funds of the missions. Local communities thus had to make substantial contributions in kind, which could include all sorts of materials, food, clothing or in the supply of unpaid labour on the mission fields or in construction work. Native teachers were paid a fraction of the European teacher salary, which also made it much cheaper to expand the networks of schools around a particular mission station.
Insert table of missionary funds in Belgian Congo. Compare with Indonesia.
5. The quality of education
It has been argued that the rather flexible private education system in Belgian Congo has positively affected the quantity of education as measured in terms of enrolment rates. But how did affect the comparative quality of education? Ideally, we would like to assess the quality of primary education on the basis of such key skills as reading, writing and calculus acquired by children that have just finished primary school. These skills are a function of a numbe rfo factors such as teacher quality, the actual years spent in school, teaching methods and practices, and the actual time devoted to training key competencies, rather than focusing on religious instruction for instance. Given the lack of observable results educational quality is very difficult to measure, however, in particular for earlier periods. The historiographical literature offers some hypotheses, however. First, in both countries public schools richer were clearly richer than private schools. Public schools could thus hire more and/or better teachers, reduce class size and buy more advanced teaching equipment. In the Netherlands Indies the differences in educational quality between the Islam schools, mission schools and desa-schools on the one hand, and the official European schools (gouvernement scholen) appeared obvious to the informed observer. European schools received more money per student and pupil-teacher ratios were considerably lower: in 1932-33 the European schools had 31,6 pupils per teacher against 49,4 in the desa schools and 50,7 in the Dutch-Chinese schools. For Belgian Congo similar evaluations exist regarding the distinction between the rural mission schools and the urban central schools. The latter were also benefitting from more state support. Yet, in both colonies rural schools or village schools provided the backbone of what we would call mass education. A possible reason to believe that the desa-schools in Indonesia 15
had more to offer than the average rural school in the Congo relates to the secular nature of the curriculum in the Indonesian schools and the strong emphasis on religious and moral instruction in the Congolese Catholic mission schools. In the desa-schools reading, writing and calculus formed the core subject. The curriculum was standardized (at least in theory) and the quality of the education was monitored by a state inspection. The primary objective of the missions in Congo was to spread Christianity. Education was only conceived as a powerful means to pursue this goal. The missionaries were also cautious not to overeducate native converts (Depaepe and van Rompaey, 1995). Rural schools in the Congo would therefore focus on religious instruction, agricultural training and the improvement of moral discipline. Girls education focused on housekeeping skills, child care, hygiene and the importance of such Christian values as monogamy. Yet, if we shift our perspective from the supply to the demand side, it seems that the arguments turn around. The Dutch government failed to convince the Indonesian of the value added of the vernacular schools, whereas the applications for Dutch language schools were far higher than the available seats. The attitude of the Indonesians with respect to village schools only changed fundamentally after the collapse of Dutch rule in the 1940s. Apparently, parents did not believe that it was worth the investment to send their children to school in a society where social mobility depended so much on the personal connections with the Dutch colonial elite. In the Congo Free State some children had been forced to attend the mission school in the early years, as they had been released from Arab slave traders or kidnapped to serve in the colonial army (van Reybrouck, 2010). Yet, missionaries could use much more soft pressure to convince parents to send their children to school, precisely because they interfered in a much broader spectrum of daily affairs than the state ever could. The local mission school was really embedded in the power structures of the connected communities. It is therefore difficult to argue that the demand for education among the Congolese in the countryside was larger than among the Indonesians, but refusing the opportunity to attend school definitely had major negative consequences in the Congo in terms of social exclusion.
Potential quality indicators (Yet to be elaborated) Pupil-teacher ratio s 1965: does not reveal significant difference. Grade repetition and drop-out rates 1970: does not reveal significant difference. Post-1950 literacy rates: is in line with school enrolment ratios, i.e. higher in the Congo. Student test scores in Indonesia in the 1990s: very poor in international perspective. 16
6. Long run implications of colonial educational development
The long run implications of the different systems of colonial education in the Netherlands Indies and Belgian Congo are important to consider, not only because the comparison is instructive for long run economic development in both countries after independence, but also because it may be instructive for the diverging paths of economic development of Africa and Southeast Asia at large. Indeed, missionary societies have been the major driver of formal primary education across Africa, whereas in Southeast Asia colonial states often played a more decisive role.
Speculations (Yet to be elaborated) 1. A possible advantage of a public education system is that the state has gained experience with the coordination of education and has reserved a permanent budget for educational development. 2. Higher rates of government revenue per capita in Indonesia have benefitted post-war educational expansion, these advantages may relate to the longer period of colonial rule and the systematic extraction of revenues during the culture system. 3. Education in both colonies became the only way to social mobility in the official ranks of the colonial state. Since the European or official schools were located in the major cities, the limited access to higher social circles aggravated rural-urban inequality. This became problematic for both post-colonial governments. 4. The high rates of illiteracy in Indonesia reach far into the 20th century and are a direct consequence of the limited supply of education and the compartmentalization. 5. Compared to its African neighbor states the Congo still seems to benefit from the colonial education legacy, for instance through comparatively high literacy rates anno 2000. 6. Rural education collapses in Belgian Congo during the decolonization, because it suddenly lost its top management. Such problems of continuation did not occur in post-colonial Indonesia.
7. Conclusion Yet to be written
17
References
Benavot, A., Riddle, P., 1988. The expansion of primary education, 1870-1940: trends and issues Sociology of Education 61, 191-210. Birdsall, N., Ross, D., Sabot, R., 1997. Education, Growth and Inequality. in: Birdsall, N. and Jasperson, F. (Eds.), Pathways to Growth. Comparing East Asia and Latin America, Inter-American Development Bank, Johns Hopkins University Press, Washington D.C., pp: 93-130. Bolt, J., Bezemer, D., 2009. Understanding Long-Run African Growth: Colonial Institutions or Colonial Education? Journal of Development Studies 45, 24-54. Booth, A., 2007. Colonial Legacies. Economic and Social Development in East and Southeast Asia. University of Hawai'i Press, Honolulu Booth, A., E., 2003. Education, Equality and Economic Development in Asian-Pacific Economies. in: Andersson, M. and Gunnarsson, C. (Eds.), Development and Structural Change in Asia-Pacific, RoutledgeCurzon, London, pp: 148-169. Brown, D. S., 2000. Democracy, colonization, and human capital in sub-Saharan Africa. Studies in Comparative International Development 35, 20-40. Brugmans, I. J., 1938. Geschiedenis van het Onderwijs in Nederlandsch-Indië. Wolters, Groningen Cogneau, D., 2003. Colonisation, School and Development in Africa. An empirical analysis. DIAL Document de Travail DT/2003/01, DIAL Cowan, L. G., O'Connel, J., Scanlon, D. G., 1965. Education and nation building. Frederick A. Praeger, New York Depaepe, M., van Rompaey, L., 1995. In het teken van de bevoogding. De educatieve actie in Belgisch-Kongo (1908-1960). Garant, Leuven Frankema, E. H. P., 2010. How Missions Made the Difference: On the Origins of Formal Schooling in sub-Saharan Africa. Working paper, Utrecht University Furnivall, J. S., 1943. Educational Progress in Southeast Asia. Institute of Pacific Relations, New York Hartgerink, H. J. H., 1942. De Staten-Generaal en het Volksonderwijs in Nederlandsch-Indië, 1848-1918. Wolters, Groningen Liesenborghs, O., 1939. Het Belgisch Koloniaal Onderwijswezen. Vlaamsch Opvoedkundig Tijdschrift 21, 448-490. 18
Lloyd, C. B., Kaufman, C. E., Hewett, P., 2000. The spread of primary schooling in sub-Saharan Africa: implications for fertility change. Population and Development Review 26, 483-515. Oliver, R., 1962. The Missionary Factor in East Africa. 2nd Longmans, Green and co ltd, London Reid, A., 1988. Southeast Asia in the Age of Commerce 1450-1680: The Lands below the Winds. Yale University Press, New Haven, CT Richens, P., 2009. The Economic Legacies of the 'Thin White Line': Indirect Rule and the Comparative Development of Sub-Saharan Africa. Working Paper No. 131/09, London School of Economics and Political Science Stengers, J., Vansina, J., 1985. King Leopold's Congo, 1886-1908. in: Oliver, R. and Sanderson, G. N. (Eds.), The Cambridge History of Africa. Volume 6, from 1870 to 1905, Cambridge University Press, Cambridge, pp: 315-358. Subramanian, V., 1979. Consequences of Christian Missionary Education. Third World Quarterly 1, 129-131. UNESCO, 1965. Statistical Yearbook 1964. UNESCO, Paris van der Veur, P. W., 1969. Education and Social Change in Colonial Indonesia. Progress and Procrastination in Education in Indonesia prior to World War II. CIS Papers in International Studies, Southeast Asia Series no. 12, Ohio University City. van Leeuwen, B., 2007. Human capital and economic growth in India, Indonesia and Japan. A quantitative analysis, 1890-2000. PhD thesis, Utrecht University, Utrecht van Reybrouck, D., 2010. Congo. Een Geschiedenis. De Bezige Bij, Amsterdam Worldbank, 1993. The East Asian Miracle. Economic Growth and Public Policy. The World Bank, Washington, D.C.
19
This document was created with Win2PDF available at http://www.win2pdf.com. The unregistered version of Win2PDF is for evaluation or non-commercial use only. This page will not be added after purchasing Win2PDF.
Ewout Frankema Utrecht University
Paper prepared for the workshop
Colonial Extraction in the Netherlands Indies and Belgian Congo: Institutions, institutional change and long term consequences
3-4 December, Utrecht
Work in progress
Abstract This paper focuses on the comparative nature and pace of educational development in the Netherlands Indies and Belgian Congo after the adoption of the ethical policy by the Dutch government (1901) and the annexation of Leopold s Free State by the Belgian state (1908). The organization of the colonial education system differed fundamentally in both colonies. In the Netherlands Indies the colonial state was the main driving force behind the spread of mass education. In Belgian Congo the education system was developed by Christian missionaries, and in particular Belgian Catholic missionaries. This paper explores the implications of these organizational differences for the development of mass education, the allocation of educational funds and speculates about the long term consequences for post-colonial development.
I would like to thank Bas van Leeuwen for supplying me with the underlying data of his statistical appendices on education in the Netherlands Indies, see Bas van Leeuwen (2007). 1
1. Introduction
The development and diffusion of formal primary education was a key feature of the civilization mission envisaged by Western colonial powers in Africa and Asia in the late 19th to mid 20th century. Against the background of the excessive exploitation and coercion of indigenous labour under the cultivation system (het cultuurstelsel) in the Netherlands Indies (ca. 1830-1870) and the domanial regime (het domaniale stelsel) in the Congo Free State (1891-1908) such paternalistic ideals got a sharp political and moral edge. This paper studies the comparative nature and pace of educational development in both colonies after the major political reforms that intended to end illegitimate exploitation and to enhance the moral elevation and material living conditions of the colonial population. While the intentions of both colonial administrations in the Netherlands Indies and Belgian Congo were more or less comparable, school enrolment rates at the end of the colonial era varied considerably, albeit in a different way than one might have expected: despite the fact that the Dutch colonial administration invested much more in education, and had done so for a much longer time as well, primary school enrolment rates in the Congo were approximately twice as high at the eve of World War II. Differences in enrolment rates are meaningful because scholars have stressed the persistent developmental effects of colonial educational legacies. In sub-Saharan Africa, for example, colonial primary school enrolment rates have been found to positively relate to post-colonial literacy rates, governance quality and per capita GDP (Brown, 2000; Lloyd et al., 2000; Cogneau, 2003; Bolt and Bezemer, 2009). In East and Southeast Asia the contribution of formal education to the post-war growth acceleration is widely acknowledged (Worldbank, 1993; Birdsall et al., 1997). Despite the critical notes of some scholars (Booth, 2003), there is consensus that education played an important role in Asian miracle growth and that colonial policies have laid the foundations of the post-colonial education system, for the better or worse (Booth, 2007). In their own region the colonial educational legacies in the Netherlands Indies and Belgian Congo stand out differently. The Dutch have often been accused of neglecting their responsibility to educate the indigenous population (de inlanders) (van der Veur, 1969, p. 17).1 Only in the 1940s did the expansion of school enrolment levels gain momentum, when the influence of the Dutch was already greatly reduced (van Leeuwen, 2007). The education record of Belgian Congo, on the other hand, looks really impressive in comparison to its African neighbors. Also in comparison with the average British African colony, which were 1
We will refer to the anachronistic, but more neutral term Indonesians henceforth. 2
regarded as the more advanced, the Belgians seem to have operated quite effectively in the Congo, developing a system of mass education with high enrolment rates and limited public resources. Indeed, up to 1940 the Dutch colonial government devoted a higher share of its total expenditure budget to education and spent much more in absolute terms. Public investments in Indonesia education had much more time to accumulate and mature, simply because the Dutch had established state control over large parts of the Indonesian archipelago since the early 19th century, whereas the Belgian government only accepted formal responsibility in 1908, when King Leopold II was forced to give up his stakes in the Congo. In the remainder of this paper it will be argued that this puzzle can only be understood when looking at two critical factors. Firstly, the relative importance of private and public education: Christian missionary societies, and especially the Catholic denominations, were the driving force behind the spread of mass education in the Congo, but played only a minor role in the Islam oriented Netherlands Indies. In the Netherlands Indies the colonial state was the main actor in both developing and financing mass education. Second, the role of what I would like to call the compartmentalisation of education for separate target groups, which was a defining characteristic of the Dutch education system, but of subordinate importance in Belgian Congo. At the end of the paper I will offer some speculative thoughts about the implications of the colonial educational legacy for long run social and economic development in both areas.
2. The comparative development of primary education
Before focusing on a direct comparison of school enrolment rates between the Netherlands Indies and Belgian Congo we compare both countries to their closest neighbors in respectively Southeast Asia and Central Africa. We will confine ourselves to primary school enrolment rates, because these constituted between 95 and 99% of total enrolments in all the countries under discussion prior to 1950. Table 1 shows the Congolese primary school enrolment rates for the age group 5-14 in 1950, together with the figures for its neighbors and the unweighted averages of the British, French and Portuguese African empires as a whole. The table shows that roughly one-third of the specified age group in Belgian Congo was involved in at least the most basic type (i.e. the lower grades) of elementary education. With this figure Belgian Congo ranked among a select 3
group of mainly British African colonies such as Zambia, Zimbabwe, Nyasaland, Mauritius and Lesotho, with gross enrolment rates (5-14) surpassing 30% before 1950. Such rates were higher than the British African average (24,2%) and far higher than the French (9,4%) or Portuguese African (8,5%) average. Enrolment rates in Congo s eight neighbor countries underline an optimistic interpretation of the Belgian colonial educational system.2
Table 1: Gross primary school enrolment rates (age 5-14) in Belgian Congo, its direct neighbours and British, French and Portuguese Africa, ca. 1950 GER age 5-14
GER age 5-14
Belgian Congo Congo Brazaville (Fr.) Central African Rep. (Fr.) Sudan (Br.) Uganda (Br.)
33 24 16 6 25
Tanzania (Br.) Zambia (Br.) Angola (P.) British Africa average French Africa average
26 35 1 24,2 9,4
Ruanda-Burundi (Be.)
11
Portuguese Africa average
8,5
Sources: Frankema (2010); UNESCO (1965), Statistical Yearbook 1964, Paris Notes: Averages are unweighted, including 16 British colonies, 16 French and 4 Portuguese. For some countries (Zambia, CAR) we used the more familiar post-colonial instead of the colonial names (Northern Rhodesia, Oubangi-Chari).
The Netherlands Indies offer a different picture. Table 2 replicates the data from Furnivall s (1943) classic comparative study Educational progress in Southeast Asia. In the final years of effective Dutch rule in the Indies, i.e. the late 1930s, gross primary school enrolment rates were considerably lower than in neighboring countries such as British Malaya, Japanese ruled Formosa (Taiwan) or the US administered Philippines. In Thailand, the only independent country in this group, the number of people enrolled in primary school as a proportion of the total population was about three times as high as in the Netherlands Indies. Only in Burma and French Indo-China enrolment rates were more or less similar. That the share of unrecognized schools, which mainly consisted of Islamic (boarding) schools (madrasahs, pesantrens), was relatively high in the Netherlands Indies (an estimated 16,4% in 1938) does not alter this conclusion. Transforming Furnivall s estimates into gross enrolment rates of the age group 5-14 (third column of table 2), enrolment rates appear a little higher than in Central Africa, but with a similar large degree of intra-regional variation.3
2
Note that the gross enrolment rate of the official primary school age group (6-11) was even considerably higher (i.e. 55%) because it reduces the denominator with 40%. 3 We assume that the 5-14 age group constitutes 25% of the total population. This is the average of the 22-27% range for pre-war developing countries suggested by Benavot and Riddle (1988, p. 199). 4
Table 2: Gross enrolment rates of total population in Southeast Asia, 1936-39 Recognized primary schools
All schools
Recognized primary schools GER age 5-14
Netherlands Indies Malaya (Br.) Burma (Br.) Thailand Indo-China (Fr.) Taiwan (Jap.)
3,4 6 3,9 9,7 2,1 11,2
4 7,8 5,5 10,7 2,5 11,4
13,6 24 15,6 38,8 8,4 44,8
Philippines (US)
10,8
11,5
43,2
Sources: (Furnivall, 1943, pp. 111-2); (Benavot and Riddle, 1988) Notes: All schools, include the unrecognized schools as well as a virtually negligible share of students enrolled in secondary and tertiary schools.
Figure 1 places the estimated primary school enrolment rates in both countries in a temporal perspective, showing that enrolment rates in Belgian Congo were substantially higher than in the Netherlands Indies between 1930 and 1950. Before we further analyze this graph, however, we need to say a little bit more about the definition of primary education. The timeseries for the Netherlands Indies are taken from van Leeuwen (2007, pp. 264-6). These include the enrolment of Indonesians as well as Europeans (mainly Dutch) and other Asians (mainly Chinese), but they exclude a considerable number of Indonesian attending the unrecognized (and non-subsidized) Islam schools. Since the colonial administration only had a vague idea about the real size of private religious education, the official statistics tend to underreport the number of unrecognized schools, if they provide them at all. There may be good reasons to exclude the unrecognized schools from a long run timeseries of enrolment rates as van Leeuwen suggests, because the religious curriculum was not designed with the purpose to enhance secular knowledge or practical skills. On the other hand, in so far Koran study enhanced the spread of literacy (and there is evidence that it did) the economic side-effects of religious education cannot be neglected (Reid, 1988, p. 218; van Leeuwen, 2007, p. 49). Nevertheless, the role of Islam schools gradually declined as the colonial state initiated the development of a mass education system. When we combine the official enrolment estimates of the unrecognized schools reported in the Koloniale Jaarcijfers of the 1880s (ca. 250.000-300.000) to Furnivall s estimate in 1938 (ca. 450.000), the relative share of the traditional Islamic schools would have gone down from ca. 75% to 16% of total primary enrolment levels. The dotted line in figure 1 reflects this trend. The relative decline must have after the adoption of the Ethical Policy (de Ethische Politiek) during the first decade of the 20th century. As of 1907 the colonial administration 5
started to support the establishment of village schools, the so-called desa schools, as a basic and cheap indigenous alternative for the more exclusive European schools. In the desa schools students were taught in the vernacular, instead of Dutch, which was the norm in European education (Brugmans, 1938, pp. 302-317). These village schools were the major driver behind increasing enrolment rates between 1907 and 1940.
Figure 1: Gross enrolment rates (age 6-11) in Belgian Congo and the Netherlands Indies, 1880-2000 160 140 120 100 80 60 40 20 0 1880
1890
1900
1910
Indonesia
1920
1930
1940
1950
1960
1970
Indonesia, incl. non-recognized schools
1980
1990
2000
Congo
Sources: For Indonesia (van Leeuwen, 2007); Jaarcijfers voor het Koninkrijk der Nederlanden (Kolonieën) 1880-192x and Statistisch Jaaroverzicht van Nederlandsch-Indië 192x-1939; (Furnivall, 1943); UNESCO, Statistical Yearbooks, various issues 1964-1999; For Congo (Depaepe and van Rompaey, 1995, P. 247) and Annuaire Statistique de la Belgique et du Congo Belge 1911-1960
But figure 1 also indicates that even when we take the unrecognized schools into account, gross primary school enrolment rates probably never surpassed 20% before 1940. The big push to mass education occurred after the retreat of the Dutch: enrolment rates jumped from 20 to 80% of the school age population in less than two decades between 1945 and 1962. Around 1980 the Suharto administration had reached the historical milestone of universal primary education. In the Congo Free State (1885-1908) and the first decades of formal Belgian rule in the Congo (1908-1960) education statistics were never systematically collected. Only during the period 1930-1960 the Belgian Annuaire Statistique provides annual information on the
6
number and type of schools and the number of students enrolled. Schooling was the exclusive domain of Protestant and Catholic missionary societies until World War II and remained highly dependent on the missions until the end of the colonial era. In theory, all missionary societies were free to settle and develop activities in the Congo as stipulated in the international agreements following the Berlin conference. In practice, the treatment of various missionary societies was uneven. Protestant missions did not receive any financial support before 1948, while Catholic missionaries were increasingly supported with state subsidies and could also rely on the less visible advantages of close administrative contacts. The fact that Protestant missionaries in the Congo were nearly all foreigners , mainly from Britain, Germany and the US did not make their position stronger. Another problem in estimating gross enrolment rates in Belgian Congo concerns the lack of reliable population data. Colonial population censuses were complicated by the sheer impossible task to survey an immense territory with hardly any physical infrastructure outside the main waterways. The devastating pandemic of sleeping sickness during the first decade of the 20th century and the death toll as a result destructive rubber policies during, roughly, the second half of Leopold s reign, make it extremely difficult to estimate the size of the population at any given point in time before 1950. The short-cut we have used here is to backward extrapolate the post-1950 population series with a fixed annual growth rate of 1,5%. We don t think this produces a very reliable time series, but we do believe it is better than taking the prewar census estimates, which almost certainly underestimate the population. Before the 1930s the enrolment levels are thus based on a few scattered observations. Referring to the work of Liesenborghs (1939), Depaepe and van Rompaey (1995, p. 38 & 247) estimate the number of students in 1908 at ca. 46.000, of which ca. 27.000 were enrolled in Protestant and 19.000 in Catholic mission schools. After 1908 the numbers increase rapidly to approximately 100.000 in 1913, 150.000 in 1921 and 350.000 in 1929. As Liesenborghs already noted, this was a spectacular rise (1939, p. 484). These figures include both the recognized schools, which were liable to state subsidies and the unrecognized schools which were entirely operated on private funds. As educational expenditure budgets increased, especially after the mid-1920s, Catholic mission schools were increasingly recognized by the colonial administration. The substantial minority of Protestant mission schools started to receive subsidies in the late 1940s.
7
3. The impressive rise of enrolment rates in Belgian Congo
Why did enrolment rates rise so rapidly in Belgian Congo after 1908? For an answer to this question it is important to note that the colonial government of Belgian Congo only became actively engaged in the spread of mass education after the Second World War. Virtually all education was missionary education and the Catholic mission schools maintained their dominant position until the end of colonial rule, despite the development of public lay education (neutral onderwijs) and the incorporation of Protestant missions in the colonial education subsidy program after 1948 (Depaepe and van Rompaey, 1995, pp. xx). The formal transition from the Congo Free State to Belgian Congo in 1908 produced a more stable and secure political-institutional context, which set the doors wide open for a rapid expansion of missionary fieldwork in the Congo. The timing and nature of this development conformed to a more general pattern in colonial Africa, where missionary activities preceded the establishment of formal European rule, but only really expanded thereafter.4 A second factor, which was also not unique for the Congo, was the rapid improvement in tropical medicine, especially in the field of malaria eradication. Despite the widespread use of quinine in the late 19th century, working and living in the Congo remained an extremely risky undertaking for Europeans. The lethal disease environment was the major reason why Belgian politicians, including king Leopold himself, refused to actually visit the dependencies in Africa. As scientists discovered the true source of malaria (the mosquitoes, instead of the bad airs mal aria of swamps) in the early 20th century, major steps in malaria prevention (rather than curing) were taken, which also freed the way for further missionary expansion. What was really unique in the Belgian Congo was the comparatively high proportion of foreigners residing in the colony. Richens has shown that in a sample of 33 African colonies Belgian Congo counted by far the highest number of white administrators in the late 1930s: 728 against an African colonial average of 94! Of course, the Congo was a vast area containing one of the largest populations in the region, but this does not explain the entire gap. In British Nigeria, which had roughly twice as many inhabitants, the number of administrators was less than half (353). Indeed, the number of administrators per capita was comparatively high in Belgian Congo (Richens, 2009, pp. 21 & 64-5). This was also the case 4
Frankema has estimated that enrolment levels in British Africa underwent a tenfold increase, from ca. 100.000 to over 1.000.000 between 1900 and 1938. In East and Central Africa the acceleration in educational expansion started right after the formal British occupation of new territories (Frankema, 2010).
8
for the number of missionaries. Oliver has estimated that the total number of missionaries (Catholic and Protestant) in the whole of East Africa never exceeded 3.500 (Oliver, 1962, pp. 231-245). Table 3 shows that the numbers in Belgian Congo were not only higher already before 1940, but also continued to increase thereafter.
Table 3: Absolute numbers and indices of missionary presence and students enrolled in Belgian Congo, 1908-1957 (1938 = 100) 1908
1929
1938
1950
1957
Missionaries
500
2.500*
3.732
5.336
7.205
Index
13
67
100
143
193
46.000
350.000
562.851
970.372
1.718.931
8
62
100
172
305
School enrolment Index
Sources: 1908 from (Stengers and Vansina, 1985); 1929 (Depaepe and van Rompaey, 1995); 1938-1957 Annuaire Statistique de la Belgique et du Congo Belge 1911-1960. Notes: 1929 number of missionaries is guesstimate based on 1931 figures from Annuaire Statistique de la Belgique et du Congo Belge 1911-1960.
Table 3 shows that primary school enrolment levels in Belgian Congo were positively correlated with the number of Christian missionaries. This was not a linear relationship, as the number of foreign missionaries increased with a factor 15, while enrolments increased with a factor 37 between 1908 and 1957. Indeed, the expansion of missionary education also entailed the Africanization of missionary field work (Oliver, 1962; Frankema, 2010). In 1958, for instance, there were no less than 6.934 schools run by Protestant missions in the Congo, which were managed by ca. 1.550 Protestant missionaries. The far majority of teachers and school heads were Congolese. In other words, when it comes to explaining the spectacular rise in enrolment rates in Belgian Congo, we not only have to understand the comparatively high rate of missionary activity in the Congo, but also the apparent enthusiasm with which the Congolese embraced Christian education. The answer to the first question is partly related to the fact that the Congo was a Belgian colony: together with the mandate over Ruanda-Urundi granted by the UN after World War I, the Congo was Belgium s only overseas possession. Belgian citizens with colonial ambitions thus concentrated their efforts almost exclusively on the Congo. In the geographically much larger and far more dispersed British and French empires, British and French citizens diversified their attention. This small colonizer argument of course also applied to some extent to the Netherlands Indies, but with the crucial difference that in terms 9
of what we could call the population support ratio , that is the ratio of colonial versus metropolitan population, the Dutch were administering a much larger colony. Table 4 compares the population support ratio of Britain, France, Belgium and the Netherlands, showing that the difference was around a factor 5.
Table 4: The metropolitan-colonial population-support ratio in the British, French, Dutch and Belgian overseas empires, ca. 1938
United Kingdom France Netherlands Belgium
Metropolitan population 1938
Colonial population 1938
Population support ratio C/M
47,5 42 8,7 8,4
358,1 65,2 68,3 13,9
7,54 1,55 7,85 1,65
Source: Statistical Abstract of the British Colonies 1938-1940; Annuaire Statistique de la France; Maddison 2009.
Clearly, there were other factors playing a role as well. After all, the differences in the population support ratio between France and Belgium were negligible. One of these factors has to do with metropolitan policy considerations. From the very inception of the Congo Free State king Leopold encouraged Catholic missionary activity as part of his strategy to physically occupy the Congo basin with minimal public funds. Missionary education was needed to train Congoleze for the army (La Force Publique) and the administration, but they brought in their own funds for their evangelical labour. The initial key role of the missions in the colonial administration created path dependency. The French were more reluctant to give so much power to missionary societies, especially against the background of the heated debates about the separation between state and church during the late 19th century (Cowan et al., 1965; Subramanian, 1979). Yet, by refusing missionary help in developing colonial education, the colonial government deliberately constrained its reach. Another part of the explanation for low enrolment rates in French Africa, in comparison to Belgian Congo, relates to the fact that the French had occupied most of the Islamic areas, particularly in West Africa. Similar to Lugard s policies in British Northern Nigeria, the French did not support Christian missionary activity in the Islamic heartlands in fear of burdening their relationship with the Muslim population and their representatives. Education thus became the exclusive domain of the neutral colonial state from the onset. The state capacity to spread mass education was limited, however, not in the least place because of the small numbers of colonial administrators. In French colonies where Islam was less 10
influential, such as Madagascar or French Congo, Christian missionary activity was less problematic and far more widespread and school enrolment rates in these colonies were considerably higher (Frankema, 2010). The situation in French West Africa was comparable with the Netherlands Indies in the sense that the state took the initiative to spread education and strictly monitored, if not actively discouraged, Christian missionary activity. Of course, Islamic education provided an alternative to public lay education or missionary schooling, but the number of Islamic schools remained limited and did not develop rapidly over time. Public education in the metropolitan language (French or Dutch), rather than vernacular education or traditional Islamic education, offered the real opportunities for social mobility.
4. Educational finance and government expenditure, ca. 1880-1960
While private missionary education remained the norm in Belgian Congo after 1908, public education dominated in the Netherlands Indies and the colonial administration only further enlarged its control over the education system after 1907. These different patterns of educational development in the first half of the 20th century are also reflected in the sources and allocation of educational finances. Figure 2 shows the percentage share of the regular government expenditure budget devoted to education in the Netherlands Indies and Belgian Congo between 1880 and 1960. The graph shows that government spending on education in the Netherlands Indies had a longer history and that, until 1940, the share in total expenditure was higher than in Belgian Congo. The rapid catch-up in educational expenditure in Belgian Congo between 1908 and 1922 is a little misleading because the size of the Congo government budget remained small. Figure 3 shows educational expenses per head of the population converted to current US $ using the official exchange rate. This is a rather crude measure, because exchange rates can deviate substantially from purchasing power parities, but it suffices to make the point that absolute expenditure levels in Belgian Congo remained modest (below 10 dollar cents per capita) and that the Dutch in the 1930s still spent three to four times as much on education as the Belgian colonial government.
11
Figure 2: The percentage share of regular government expenditure devoted to education in the Netherlands Indies and Belgian Congo, 1880-1960
0,20
0,15
0,10
0,05
0,00 1880
1890
1900
1910
1920
1930
1940
1950
1960
Sources: For Indonesia (van Leeuwen, 2007); For Congo Annuaire Statistique de la Belgique et du Congo Belge 1911-1960.
Figure 3: Government expenditure on education per capita in the Netherlands Indies and Belgian Congo, 1880-1960 (current US $, y-axis log scale) 10,0 Indonesia
Congo
1,0
0,1
0,0 1880
1890
1900
1910
1920
1930
1940
1950
1960
Sources: For Indonesia (van Leeuwen, 2007), van Leeuwen has included expenditure on the state level and provincial level, as published in the reports of the Dutch Indies Education Committee; For Congo Annuaire Statistique de la Belgique et du Congo Belge 1911-1960. Notes: Dutch florins and Belgian Francs converted to US Dollars using official exchange rates. 12
Figure 4: Government expenditure on education per student enrolled in the Netherlands Indies and Belgian Congo, 1890-1960 (current US $, y-axis log scale) 100 Indonesia
Congo
10
1
0 1890
1900
1910
1920
1930
1940
1950
1960
Sources: see figures 2 and 3.
Because enrolment rates in the Netherlands Indies were so much lower, the differences in government expenses per student enrolled, as shown in figure 4 (y-axis log scale!), even fluctuated around a rate of 1:10 between 1910 and 1940. The horizontal trend in both countries indicates that the increasing size of the education budget was mainly consumed by expanding school enrolment rates, rather than an autonomous rise in the available budget per student enrolled. In fact, given the substantial amount of interwar inflation, the nominal persistence of education expenses suggests a long run decline in real terms. Although the average sum of ca. 10 US $ per student in the Netherlands Indies was quite respectable, we have to keep in mind that the distribution around the average was highly unequal. In 1929, for instance, about half of the total education budget was spent on European education, which was reserved for only about 10% of the total enrolled (van Leeuwen, 2007). The ethnic distribution of public resources between the Europeans on the one hand, and the Indonesians and Chinese on the other hand, became more egalitarian over time however. In the late 19th century European education was almost exclusively restricted to Dutch citizens, but by the late 1930s ca. 75% of the students enrolled were of Indonesian or Chinese origin (van Leeuwen, 2007). Needless to say, these were mainly children from higher social ranks.
13
The political choices regarding the allocation of government expenses also had important implications for the mission schools in the Netherlands Indies. In the last quarter of the 19th century around a quarter of the recognized students on Java and Madura went to a Christian mission school (zendingschool) and this share was even much higher in the outer areas (Hartgerink, 1942, p. 39). Yet, the financial resources of the missions were limited and the colonial administration refused financial support. The government preferred to annex mission schools which were unable to stay in business, rather than subsidize them. In the first decade of the 20th century the colonial government deliberately transferred part of its educational responsibility to the missions in some of the outer areas, such as the Moluccas, under the promise of state subsidies (Hartgerink, 1942, pp. 141-149). Hence, the missions were used to tie the loose ends of the public educational system together. In Belgian Congo the missionary societies were not only carrying the colonial education system, they were also its primary financiers. The rise in government expenditures during the second half of the 1920s, which also occurred in British Africa in response to the influential Phelps Stokes report (1922), only sufficed to compensate for the increasing costs of mission school expansion. As stated above, only Catholic schools could hope to receive some additional funding, but even among the latter the government funds were distributed rather selectively. The central schools in the major population centers received most of the money, while the village schools in the immense hinterland received next to nothing. The reason was that the central schools had to be better equipped because of their importance for the colonial order as such. These schools trained the clerks and typewriters necessary for administrative tasks, the medical assistants, technicians and agricultural experts necessary to develop the economy and also the catechists and soldiers necessary to evangelize and secure order (Depaepe and van Rompaey, 1995, pp. 63-69). The financial involvement of the Belgian Catholic missions in the Congo played a crucial role in the 1925 convention with the Belgian state. The catholic schools would receive more financial support in exchange for extra efforts to spread mass education across the colony. The contract was not signed until 1929, however, because the missions rejected the proposal of a state inspection that would monitor and guarantee the quality of colonial education. The fact that the Catholic missions eventually won this battle tells a lot about the balance of power: the Catholics exerted a de facto monopoly on the development of colonial education which they held on to until 1948 (Depaepe and van Rompaey, 1995, pp. 60-63). This does not mean that the existing financial problems were resolved or even relieved. The number of schools and students expanded at a much faster rate than the home 14
country funds of the missions. Local communities thus had to make substantial contributions in kind, which could include all sorts of materials, food, clothing or in the supply of unpaid labour on the mission fields or in construction work. Native teachers were paid a fraction of the European teacher salary, which also made it much cheaper to expand the networks of schools around a particular mission station.
Insert table of missionary funds in Belgian Congo. Compare with Indonesia.
5. The quality of education
It has been argued that the rather flexible private education system in Belgian Congo has positively affected the quantity of education as measured in terms of enrolment rates. But how did affect the comparative quality of education? Ideally, we would like to assess the quality of primary education on the basis of such key skills as reading, writing and calculus acquired by children that have just finished primary school. These skills are a function of a numbe rfo factors such as teacher quality, the actual years spent in school, teaching methods and practices, and the actual time devoted to training key competencies, rather than focusing on religious instruction for instance. Given the lack of observable results educational quality is very difficult to measure, however, in particular for earlier periods. The historiographical literature offers some hypotheses, however. First, in both countries public schools richer were clearly richer than private schools. Public schools could thus hire more and/or better teachers, reduce class size and buy more advanced teaching equipment. In the Netherlands Indies the differences in educational quality between the Islam schools, mission schools and desa-schools on the one hand, and the official European schools (gouvernement scholen) appeared obvious to the informed observer. European schools received more money per student and pupil-teacher ratios were considerably lower: in 1932-33 the European schools had 31,6 pupils per teacher against 49,4 in the desa schools and 50,7 in the Dutch-Chinese schools. For Belgian Congo similar evaluations exist regarding the distinction between the rural mission schools and the urban central schools. The latter were also benefitting from more state support. Yet, in both colonies rural schools or village schools provided the backbone of what we would call mass education. A possible reason to believe that the desa-schools in Indonesia 15
had more to offer than the average rural school in the Congo relates to the secular nature of the curriculum in the Indonesian schools and the strong emphasis on religious and moral instruction in the Congolese Catholic mission schools. In the desa-schools reading, writing and calculus formed the core subject. The curriculum was standardized (at least in theory) and the quality of the education was monitored by a state inspection. The primary objective of the missions in Congo was to spread Christianity. Education was only conceived as a powerful means to pursue this goal. The missionaries were also cautious not to overeducate native converts (Depaepe and van Rompaey, 1995). Rural schools in the Congo would therefore focus on religious instruction, agricultural training and the improvement of moral discipline. Girls education focused on housekeeping skills, child care, hygiene and the importance of such Christian values as monogamy. Yet, if we shift our perspective from the supply to the demand side, it seems that the arguments turn around. The Dutch government failed to convince the Indonesian of the value added of the vernacular schools, whereas the applications for Dutch language schools were far higher than the available seats. The attitude of the Indonesians with respect to village schools only changed fundamentally after the collapse of Dutch rule in the 1940s. Apparently, parents did not believe that it was worth the investment to send their children to school in a society where social mobility depended so much on the personal connections with the Dutch colonial elite. In the Congo Free State some children had been forced to attend the mission school in the early years, as they had been released from Arab slave traders or kidnapped to serve in the colonial army (van Reybrouck, 2010). Yet, missionaries could use much more soft pressure to convince parents to send their children to school, precisely because they interfered in a much broader spectrum of daily affairs than the state ever could. The local mission school was really embedded in the power structures of the connected communities. It is therefore difficult to argue that the demand for education among the Congolese in the countryside was larger than among the Indonesians, but refusing the opportunity to attend school definitely had major negative consequences in the Congo in terms of social exclusion.
Potential quality indicators (Yet to be elaborated) Pupil-teacher ratio s 1965: does not reveal significant difference. Grade repetition and drop-out rates 1970: does not reveal significant difference. Post-1950 literacy rates: is in line with school enrolment ratios, i.e. higher in the Congo. Student test scores in Indonesia in the 1990s: very poor in international perspective. 16
6. Long run implications of colonial educational development
The long run implications of the different systems of colonial education in the Netherlands Indies and Belgian Congo are important to consider, not only because the comparison is instructive for long run economic development in both countries after independence, but also because it may be instructive for the diverging paths of economic development of Africa and Southeast Asia at large. Indeed, missionary societies have been the major driver of formal primary education across Africa, whereas in Southeast Asia colonial states often played a more decisive role.
Speculations (Yet to be elaborated) 1. A possible advantage of a public education system is that the state has gained experience with the coordination of education and has reserved a permanent budget for educational development. 2. Higher rates of government revenue per capita in Indonesia have benefitted post-war educational expansion, these advantages may relate to the longer period of colonial rule and the systematic extraction of revenues during the culture system. 3. Education in both colonies became the only way to social mobility in the official ranks of the colonial state. Since the European or official schools were located in the major cities, the limited access to higher social circles aggravated rural-urban inequality. This became problematic for both post-colonial governments. 4. The high rates of illiteracy in Indonesia reach far into the 20th century and are a direct consequence of the limited supply of education and the compartmentalization. 5. Compared to its African neighbor states the Congo still seems to benefit from the colonial education legacy, for instance through comparatively high literacy rates anno 2000. 6. Rural education collapses in Belgian Congo during the decolonization, because it suddenly lost its top management. Such problems of continuation did not occur in post-colonial Indonesia.
7. Conclusion Yet to be written
17
References
Benavot, A., Riddle, P., 1988. The expansion of primary education, 1870-1940: trends and issues Sociology of Education 61, 191-210. Birdsall, N., Ross, D., Sabot, R., 1997. Education, Growth and Inequality. in: Birdsall, N. and Jasperson, F. (Eds.), Pathways to Growth. Comparing East Asia and Latin America, Inter-American Development Bank, Johns Hopkins University Press, Washington D.C., pp: 93-130. Bolt, J., Bezemer, D., 2009. Understanding Long-Run African Growth: Colonial Institutions or Colonial Education? Journal of Development Studies 45, 24-54. Booth, A., 2007. Colonial Legacies. Economic and Social Development in East and Southeast Asia. University of Hawai'i Press, Honolulu Booth, A., E., 2003. Education, Equality and Economic Development in Asian-Pacific Economies. in: Andersson, M. and Gunnarsson, C. (Eds.), Development and Structural Change in Asia-Pacific, RoutledgeCurzon, London, pp: 148-169. Brown, D. S., 2000. Democracy, colonization, and human capital in sub-Saharan Africa. Studies in Comparative International Development 35, 20-40. Brugmans, I. J., 1938. Geschiedenis van het Onderwijs in Nederlandsch-Indië. Wolters, Groningen Cogneau, D., 2003. Colonisation, School and Development in Africa. An empirical analysis. DIAL Document de Travail DT/2003/01, DIAL Cowan, L. G., O'Connel, J., Scanlon, D. G., 1965. Education and nation building. Frederick A. Praeger, New York Depaepe, M., van Rompaey, L., 1995. In het teken van de bevoogding. De educatieve actie in Belgisch-Kongo (1908-1960). Garant, Leuven Frankema, E. H. P., 2010. How Missions Made the Difference: On the Origins of Formal Schooling in sub-Saharan Africa. Working paper, Utrecht University Furnivall, J. S., 1943. Educational Progress in Southeast Asia. Institute of Pacific Relations, New York Hartgerink, H. J. H., 1942. De Staten-Generaal en het Volksonderwijs in Nederlandsch-Indië, 1848-1918. Wolters, Groningen Liesenborghs, O., 1939. Het Belgisch Koloniaal Onderwijswezen. Vlaamsch Opvoedkundig Tijdschrift 21, 448-490. 18
Lloyd, C. B., Kaufman, C. E., Hewett, P., 2000. The spread of primary schooling in sub-Saharan Africa: implications for fertility change. Population and Development Review 26, 483-515. Oliver, R., 1962. The Missionary Factor in East Africa. 2nd Longmans, Green and co ltd, London Reid, A., 1988. Southeast Asia in the Age of Commerce 1450-1680: The Lands below the Winds. Yale University Press, New Haven, CT Richens, P., 2009. The Economic Legacies of the 'Thin White Line': Indirect Rule and the Comparative Development of Sub-Saharan Africa. Working Paper No. 131/09, London School of Economics and Political Science Stengers, J., Vansina, J., 1985. King Leopold's Congo, 1886-1908. in: Oliver, R. and Sanderson, G. N. (Eds.), The Cambridge History of Africa. Volume 6, from 1870 to 1905, Cambridge University Press, Cambridge, pp: 315-358. Subramanian, V., 1979. Consequences of Christian Missionary Education. Third World Quarterly 1, 129-131. UNESCO, 1965. Statistical Yearbook 1964. UNESCO, Paris van der Veur, P. W., 1969. Education and Social Change in Colonial Indonesia. Progress and Procrastination in Education in Indonesia prior to World War II. CIS Papers in International Studies, Southeast Asia Series no. 12, Ohio University City. van Leeuwen, B., 2007. Human capital and economic growth in India, Indonesia and Japan. A quantitative analysis, 1890-2000. PhD thesis, Utrecht University, Utrecht van Reybrouck, D., 2010. Congo. Een Geschiedenis. De Bezige Bij, Amsterdam Worldbank, 1993. The East Asian Miracle. Economic Growth and Public Policy. The World Bank, Washington, D.C.
19
This document was created with Win2PDF available at http://www.win2pdf.com. The unregistered version of Win2PDF is for evaluation or non-commercial use only. This page will not be added after purchasing Win2PDF.
