Literature Review and Contributions to the Literature
The contributions to the literature from this paper could be multi-fold. First of all, this paper con-tributes to better understanding the functional intricacies of extractive colonial institutions, and in this particular case, the French colonial military conscription system in West Africa.
The general literature on colonial institutions has started with an exclusive focus on the long-term developmental impacts of various institutions put in place during colonial times, often at a rather grand and macro cross-country level (Engerman & Sokoloff, 1997; Acemoglu et al., 2001, 2002, 2005), where the proclaimed causal inference might be a bit strenuous to grab given other deep underlying factors at play. The more recent literature has attempted to identify causal developmental impacts of institutions at the more refined local level within a specific geographical region where the deep cultural, geographical or other long-lasting determinants seem to be more or less held constant, such as the focus on French West Africa (Huillery, 2009), the British Raj of India (Banerjee & Iyer, 2005), Spanish exploitative mining regimes in current-day Peru (Dell, 2010) or simply a specific type of “local institution” of very extractive nature and put in place for a rather short period of time during colonial years, such as colonial concessions (Lowes & Montero, 2016), the native chief rule instituted by the British in Sierra Leone (Acemoglu et al., 2014) and colonial opppression episodes in Madagascar (Garcia Ponce & Wantchekon, 2011).
Indeed, the historical and contemporary relevance on the inquiry for long-term developmental outcomes is undoubtedly valid, but such long-run repercussion analyses, nonetheless require very careful close-up scrutiny at the potential diverging or reversing historical impacts of the institutional layout under study in different important historical junctures throughout space and time.
As such, before jumping towards a grand conclusion on how colonial institutions significantly affected the living conditions of the indigenous in contemporary times, rather first and foremost, it might be equally important to comprehend how these institutions most effectively affected the organization of daily lives back during the colonial years and if such impacts during the colonial years served as stimulants or hindrance for further development bifurcations for the regions mostly affected by such ruling. Such inquiries have largely escaped the attention of most economic historians of development until very recently, in further trying to understand the role played by forced labor institutions during colonial times on colonial expenditure capacities, Van Waijenburg (2018) found an enormous contribution of the forced labor « corvée » system to colonial fiscal capacities in French African colonies. Similarly, by exploiting colonial data on the extensive extraction of convict labor for public works projects in British Nigeria, Archibong and Obikili (2019) identified that the share of convict labor made up of a substantial share of the total public works expenditures at the time.
In this sense, this paper also tries to evaluate, in the first place, the historical relevance of the military conscription system in French West Africa, on how it might have served to influence the living patterns of the indigenous Africans or the administrative structures of the French colonial authorities back then, and such research can be further seen as an important building block for the subsequent and more extensive analysis of the medium- or long-term repercussions of colonial military conscription in Africa.
Additionally, this paper also speaks to the vast literature on exploring the administrative complex-ities of the French direct colonial rule. The current literature started with a focus on the differential long-term developmental impacts between the direct French colonial rule and the indirect British colo-nial rule, starting with the seminal research from La Porta et al. (1997) and Porta et al. (1998) where the authors evaluated the long-term investment repercussions of French-civil-law versus British-common-law systems, to the more recent focus on ethnicity identification (Ali et al., 2018) or ethnic conflict consequences (Wucherpfennig et al., 2016) due to the differential inclusion versus exclusion ethnic politics during colonial times. However, starting from very early on, the literature has already em-phasized the great within-heterogeneities in colonial rule in French and British colonies respectively (Crowder, 1964; Berry, 1992). More recently, Frankema and Van Waijenburg (2014) explored how the colonial authorities in Africa adopted varying ruling strategies to cope with heterogeneous local ruling difficulties due to local factors such as geography, population density, etc. For instance, they explored how the German adopted both direct and indirect rules in Namibia and how the Portuguese colo-nizers applied both direct rules and concession policies to private companies depending on various geographical regions in Mozambique in the 19th century. The same has been observed by Cogneau et al. (2018) where they emphasized the second French colonial empire’s adaptive ruling capacities towards differing socioeconomic and historical contexts.
For colonial rule in French West Africa, a lot of the past research in the economics literaure has focused on the rather « omnipotent » role played by the local district-level administrators in influencing investments in public goods such as education and health expenditures that displayed striking long-term persistence (Huillery, 2009; Cai, 2015; Dray, 2016). For the analyses conducted in this paper, instead of excessively focusing on the seemingly decisive role of district administrators, which may be subject to debate, I tried to look more into how the interaction between French colonial officials of different ranks took place via the military conscription system. Such research contributes to a potential better understanding of certain mal-functioning and inefficiencies associated with the strignent top-down French colonial rules in FWA.
Furthermore, this paper also contributes to the rather scarce literature on various informal in-surance mechanisms exploited by indigenous people in historical or colonial settings. Blouin (2016) explored how the arrival of the Belgian colonizers destroyed the precolonial insurance institution ex-istent between the Hutus and Tutsis in Rwanda and the newly institutionalized forced labor system during the colonial times might have ultimately undermined the inter-ethnic trust cemented over the years via the informal insurance device. In the same vein, it could be equally conceived that similar ancient insurance structures might have been wiped away along the conquest of West Africa by the French Army. While on the other hand, extreme economic adversity and the lack of guarantee for sub-sistence survival might have further propelled the indigenous people to exploit seemingly “extractive colonial institutions” to their own advantage as a potential shelter and insurance of last resort. Such an exploration is particularly interesting because there’s rather scant research on how historically colo-nial or other oppressed subjects “subjectively” opted for such extractive institutions for the benefits of themselves.
The term Tirailleurs Sénégalais (in English “Senegalese Riflemen”) was first coined by the then-General Governor of French West Africa Louis Faidherbe in 1857, to refer to the first permanent unit of black Senegalese soldiers recruited by him into the French Army. Ever since then, this term has been generalized and adopted as an all-encompassing appellation for the section of the colonial French Army composed of indigenous black African recruits, either they were conscripted from French West Africa or French Equatorial Africa.
In actuality, the history of recruiting indigenous Africans into the French Army preceded this first creation of a permanent unit by Governor Faidherbe. As early as the 1820s, a small Bataillon d’Afrique of a few hundred recruits had already been formed in Senegal during the initial years of the post-Napoleonic era (de Boisboissel, 1931). Throughout the 19th century, the recruitment of African soldiers more or less followed a “Slaves into Soldiers” principle, sustained by the scandalous Rachat system, where the French Army repurchased slaves from the slave-owning masters at discounted rates (Zuccarelli, 1962), or its variants in the post-slavery era, such as payment of enlistment bonuses to the previous slave masters (Renault, 1972) or incorporation of prisoners-of-war and porters, etc.
Such a slave-mercenary-based conquest army was partially transformed into a more regular insti-tution with incorporation of soldiers from all walks of life after the institutionalization of the civilian rule of French West Africa in 1905. It was also the time when this entire system of Tirailleurs Sénégalais witnessed its largest spike in terms of recruits ever since its creation. At the time, the French colo-nial authorities quickly realized that they needed more military men to guard the newly-conquered territory than to conquer them in the first place (M. J. Echenberg, 1991, p. 25). As such, a partial conscription law was put in place in 1912 and subsequently this law successfully incorporated an as-tounding figure of 170,000 black indigenous Africans to have served in the French Army during the First World War.
The First World War fundamentally changed the entire institution of Tirailleurs Sénégalais forever. Although there was a lot of post-war debate on how the recruitment system should be sustained or reformed, ultimately the French colonial authorities decided to formerly institute the wartime recruit-ment system by introducing a universal peacetime conscription system in French West Africa in 1919. Henceforth, such conscription patterns as stipulated by the universal law sustained until the breakout of Second World War, when the war-time special conscription scheme predominated again. After the Second World War, the purpose of the conscription system was fundamentally altered as it was pro-fessionalized, with a significant increase in fitness standards and incorporation of soldiers from local military schools.
The ultimate decision to maintain a conscription army during the peacetime years was one that was going to most profoundly influence the living patterns of the indigenous as well as military colonial rule in the region. Although the two World Wars seemingly levied an enormous number of indigenous Africans in sheer numbers, it needs to be taken into account that the peacetime conscription system continuously lasted for a period of more than two decades and its ultimate total levy on the most-able local young males far exceeded the toll exacted by the two World Wars. This would also be the sample period of interest for the key analyses conducted in this paper, during the period of which the functional complexities inside the French colonial administrative hierarchy with regard to military conscription, and the indigenous people’s behavioral responses towards this seemingly “extractive system” would be duly explored.
The administrative organization of French West Africa (FWA)
Before moving forward to the detailed depiction of the peacetime conscription system, it’s important to bear in mind the overall administrative structure of the entire FWA, which is the focus for the empirical analysis of this paper.
As can be seen from Figure 1, French West Africa (FWA) was a federation of eight former French colonies, namely Senegal, Cote d’Ivoire, Guinea, Niger, French Sudan (current-day Mali), Dahomey (current-day Benin) and Upper Volta (current-day Burkina Faso). It straddles across a vast and diverse territory with more than 4,800,000 square kilometers in size (roughly twice the size of modern-day Algeria, or eight times the size of metropolitan France) with extremely low population densities, as it’s been estimated that in 1910 the region was inhabited by only around 12 million people.
Colonial French West Africa (Afrique Occidentale Française) as a federation was officially created in 1895 and the civilian government was established in 1904. Although the major military conquests were over by this time, civilian administration only gradually became effective across the entire FWA from the early 1900s till 1920s, with it taking the longest for geographically remote districts such as those of Mauritania, etc (Huillery, 2009). This also mirrored the period of time when the Tirailleurs Sénégalais witnessed its first major rise in recruitment, mainly for consolidating the colonial rule over the conquered regions and for soldier supply of the upcoming First World War. Hence on the whole, the period between 1904 and 1920 could be more or less regarded as an “adjustment period” of FWA where the colonial ruling both in terms of military prestige and civilian administration gradually established a firm basis for the forthcoming peacetime era (1920-1938). The structure of the colonial administration of FWA was argued to be stringently pyramidal: At the head of the federation we have the general governor (gouverneur général), who is followed by the lieutenant governors at the colony level (lieutenant-gouverneur) and ultimately, the administrators (administrateur or commandant de cercle) were the rulers of the respective districts within each given colony.
Hence, on the paper, FWA seemed to be an officially and strictly centralized regime. Yet there has been ample anecdotal historical evidence suggesting that the district administrators possessed the real powers in deciding the arrangements of the major daily activities at the local level. Some histo-rians even went so far as to claim that the administrators were the “real rulers of the French Empire” (Delavignette, 1939) by being “omnipotent and omnipresent” (El Mechat, 2009). Indeed, the district ad-ministrators were in charge of a series important tasks such as taking charge of tax collection, assisting in military conscription work, overseeing local public investment projects in terms of infrastructure construction, health and education expenditures, etc. It’s been argued that it was potentially the large geographical distance in the immense FWA region as well as the lack of communication means at the time that contributed to the rather omnipotent local power capacities of the administrators (Cohen et al., 1973).
The stringent “top-down” approach of peacetime conscription
Hence, it is against this backdrop of the colonial FWA administrative system that the peacetime con-scription system was carried out. Unsurprisingly, on the surface, the peacetime military conscription system was reputed to be a strictly “top-down” operational process. To start with, every autumn, the minister of defense in Paris would indicate the number of Tirailleurs Sénégalais required to be recruited for the entirety of French West Africa. Subsequently, this number was sent to the cabinet of French government for approval. Once this annual quota was accepted, it purportedly became “immutable” and was sent to the General Governor (head of all French West African colonies) in Dakar, Senegal for actual execution of the conscription. The General Governor would then distribute the quotas to the eight colonies following a fixed formula developed over the years with some regard to the population distribution across the colonies. Henceforth, the Lieutenant-Governor (head of a particular colony) would further divide the received colony-level quota into district-specific conscription targets and sent them to district administrators for the ultimate and actual fulfilment of military conscription at this most dis-aggregate administrative level, which is also the level of analysis for this paper.
Around the month of December or January in the following year, after having received the fixed recruitment quota, the district administrators would start to make an enumeration list of indigenous young males, who were regarded as potential recruits for the district in that given year. This process was usually carried out with the assistance of the local chiefs (who might have more knowledge of the distribution of population at the local level) or with available population census lists.
Subsequently, the actual military conscription took place in the form of a mobile drafting com-mission board (committee) that moved from one district to another to carry out the recruitment pro-cedures. Such a drafting commission board was usually composed of one French army officer (the chairman of the committee), the district administrator (second to the army officer), two clerks (one from the military and the other from the civilian administration) and one military physician (in charge of medical examination for the potential recruits). The enumerated soldiers would then be called upon in their residential villages or towns to travel to the drafting centers in their own district for medical examination. It was at this point in time that a certain proportion of the enumerated individuals would choose to escape from the conscription calls and some even went so far as to flee away from their residential districts to evade the search from the military authorities.
The individuals who were “present” would then receive medical examinations to be determined whether they were fit for the military or not. The standards of such medical examinations could be left to much speculation, but what needs to be emphasized is that at the end of the day, the number of “fit” out of all the present individuals is at best a “deemed fit” measure, not necessarily reflecting the true health conditions of the potential recruits. As such, the “deemed fit” measure could be largely subject to endogenous manipulation of the drafting board officers given potential changes in the recruitment process that might require them to do so. Subsequently, among the soldiers who were deemed fit, the drafting committee first asked who would “volunteer” to join the army and these volunteers were then directly recruited as “first portion” of the army. Among the rest, according to the official target (quota) for this year’s recruits, the committee would perform a lottery to recruit a portion of the remaining soldiers, who would join the previous volunteered soldiers and the two together would constitute the overall “first-portion” soldiers. Ultimately, the remaining fit soldiers who were left out of the lottery would form the “second portion”, also known as “military reservists”, who might be activated in times of urgent needs, such as the two World Wars, etc.
In this paper, I do not attempt to answer the overarching question on what fundamentally motivated the indigenous Africans to join the French military, or to dissect the intricate interplay of various factions of power inside the colonial administration to further determine the observed conscription patterns. Instead, I will try to answer a few relevant questions along similar research agenda and within the available data structure at hand as described below.
Maneuver of conscription from the colonial authorities
The maneuver of military conscription from the colonial authorities’ perspective could be regarded as a simple three-dimension (non-) cooperative game among three principal agents: the local African chiefs, the French military officers and the French civilian district administrators.
As previously mentioned, the civilian district administrators often required (crucial) assistance from the local chiefs in drafting up the enumeration list with “recruitable” and “identifiable” in-digenous Africans. The chiefs hence purportedly faced a dilemma between serving for the French authorities (in the case of which they would usually be provided with financial incentives, such as salaries for being hired to perform certain tasks) or playing favorites for their own people at the local level, where they might also want to keep the most-able young males as labor force for themselves.
On the other hand, even within the French colonial authorities themselves, there could be compet-ing interests between the civilian administrators and the military officers regarding the conscription procedures. The district administrators might themselves have preferred to keep the best prime-age labor for local exploitation either in large-scale local cash crop plantations or other construction of public infrastructure projects, in the objective of principally contributing to local welfare. In the same vein, they could also have preferred to keep the best labor locally in order to appease potential tension with the local chiefs.
Such competing interests are made more likely to result in a strategic maneuver of the recruitment procedure, further given the fact that on the one hand, the “mobile” nature of the drafting board required the military officers to travel across several districts to conduct the reruitment procedures at a rather high frequency. This renders it rather impractical for them to obtain thorough knowledge of the recruitment landscape at the local level. Such a reality concerning the actual conscription procedure was vividly documented by M. J. Echenberg (1991, p. 251):
“The army officer, the physician and one clerk would have to visit perhaps ten to fifteen cercles in the colony when recruitment took place. Only the civilian component of the drafting board had some continuity and knowledge of the local demography, etc.”
While on the other hand, nonetheless, the district administrators and the local chiefs possessed more specific knowledge of the recruitment capacities of their own regions and hence had more room for manipulating, if not the “quantity”, the “quality” of the potential recruits at the local level than an officer who came in only for the week to conduct the draft board hearings.
Therefore, although the conscription procedure was on paper supposed to be a strictly “top-down” procedure, as the most conventional kind of direct-ruling French colonial institution was purported to be, there are ample rationales to believe that there are more administrative and functional intricacies within these seemingly pyramidal colonial institutions.
Although the data available wouldn’t allow me to capture the exact magnitude of the respective roles played by either the military or civilian officials (or local chiefs) on influencing the conscription outcomes such as volunteering or present rates, or how the competition for prime-age male labor exactly played out between the three parties, one plausible arena where such functional intricacies could be explored is how the lower-ranked colonial officials were adjusting or reacting towards a change in the recruitment target as stipulated by the colony-level administrative officers.
Table of contents :
2 Literature Review and Contributions to the Literature
3.1 Historical background
3.2 The administrative organization of French West Africa (FWA)
3.3 The stringent“top-down”approach of peace time conscription
4 Theoretical Hypotheses
4.1 Maneuver of conscription from the colonial authorities
4.2 Attitudes toward smilitary conscription from the indigenous
5 Data and Empirical Strategies
5.1 Conscription data
5.2 Majorout come variables of interest
5.3 Specifications for the impact of recruitment targets
5.4 Specifications for the impact of weather shocks
5.5 Weather data
5.6 Construction of other variables of interest
5.6.1 Construction of other weather-related variables
5.6.2 « Bartik-like »fill-in method formissing values oftargets
5.6.3 Construction of control variables
6 Summary Statistics
6.1 Main variables analyses
6.2 Correlation analyses
7.1 Variance Decomposition
7.2 Results of theim pacts of recruitment target changes
7.2.1 Direct impact son the recruitment procedures
7.2.2 Heterogeneous impacts given labor constraint and target pressure
7.3 Results of the impacts of weather shocks on military participation
Main Figures and Tables