Inequality, the emergence of a political idea in the 20th century

Concevoir les inégalités dans le monde : l'émergence d'une idée politique au xxe siècle
Article Index
Measuring living standards internationally
Initial assessments of infant mortality in Africa
Initial comparisons of national income per capita…

Controversial indicators of standards of living lie at the heart of inequality issues. A history of international comparisons of living conditions shows that these indicators definitely help change how the world is seen and whether human rights are recognized. By putting everyone into a space of equivalence that enables quantification, statistics serve as new tools for generating discussion about the reconfiguration of power relationships.

In historical terms, the measurement and understanding of inequalities in the world are relatively recent concerns. During the first part of the twentieth century, when most of the countries that would later become known as "developing countries" were undergoing colonization, there was no question of comparing the living standards of European populations, or those of European origin, with the situation of African or Asian populations, nor of reducing the gap between one and the other. At the time, a barrier of incomparability existed between colonized peoples and those who were not, which, depending on the various colonial empires, was expressed in a legal form, but also on the basis of differences in race and civilisation: the view was held that "whites", "blacks" and "yellows" could not have the same needs or social aspirations. Without doubt, the development of international statistical data has, since the nineteenth century, played a role in the beginnings of a hierarchal definition of living standards across the world. But this remains "geographically limited to civilised countries" (Horvath, 1972): in other words, concerning mainly European and North American countries which, due to their assumed degree of "civilization", share the same referential space. The emergence of comparative data on living standards for societies that underwent industrialization in the nineteenth century is also closely linked to the development of social legislation intended to "govern misery" (Procacci, 1993), something which did not apply in the colonies.

The aim of this article is to put into perspective the conditions under which comparable standards of living in the world emerged in the mid-twentieth century, focusing particularly on the contradictory role that may have been played by the data on one side, and the paradigm of race on the other. More specifically, this article focuses on the first ever collection of data on living conditions in French colonial Africa. How has this knowledge been materially developed? To which questions, both internal and external to the empire, was it addressed? What were the different political uses of such data? Through the analysis of these questions, a number of contemporary issues will be raised regarding the measurement of inequality in the world, issues which echo controversies that appeared some sixty years ago.

The first evaluations of inequality in industrial societies

Since the mid-nineteenth century, a number of industrial societies have collected statistical information to measure and compare the living conditions of populations in terms of collective standards. Although most of this data primarily resulted from the initiatives of individual scholars, industrialists and doctors, it was correlated with new social legislation that was being established in certain countries. In England, where the adoption of the Poor Law Amendment Act in 1834 led to multiple controversies over the government's management of poverty, several major surveys were thus made in the second half of that century to estimate the numbers of poor people living in certain cities. The most important of these surveys were conducted successively by Charles Booth and Benjamin Rowntree who are now considered as the main precursors of the "poverty line" concept: from the perspective of their research, poverty became a category that was defined by a quantitative threshold, allowing it to be compared from one place to another (Carré, Révauger, 1995). At around the same time, similar surveys on the working classes were being conducted in France, while from 1841 onwards the first social laws were passed, which were expected to reduce the hardships endured by the workers. One of the stated objectives of these surveys was to allow "a comparison between places and people" in the words of Frederic Le Play, who was the author of an important series of monographic studies conducted in the second half of the century on workers" conditions (Le Play, 1879).

From such research the concept of social inequality emerged, in the sense that it is currently perceived (Desrosieres, 2008), although at the time these social inequalities mainly remained thought of at the level of a single country, where the same assistance and management mechanisms applied. Admittedly, there were some examples of cross-national comparisons in this area, particularly the work of Le Play (1879) and Rowntree (1910), but also within the framework of the first international statistical congresses that, in the second half of the nineteenth century, gathered specialists in demography, health and work (Brian, 1989). Unemployment figures in particular - which, from the 1880s, gradually replaced poverty as a category of public action in industrial societies (Topalov, 1994) - led to international comparisons of the actual situation itself, but also of the policies that were put in place to deal with the problem. But such comparative approaches to the understanding of living conditions remained rare until the early twentieth century.

Between the two world wars, the measurement and international comparisons of living conditions became more formalized due to the influence of the League of Nations (LON) and the International Labour Office (ILO). At the disposal of these institutions was quantitative data that governments were collecting on an increasingly routine basis for political purposes, although the production of statistics in this manner did not necessarily obey the same methodological framework. One of the LON's objectives was to work towards a standardization that would enable the comparison between circumstances in different nations and also the different ways in which the public authorities were managing the issues. This was particularly important in the 1930s as the economic crisis had resulted in the deterioration of living conditions throughout most industrialized countries. In 1932, the ILO thus published a "study of international comparisons of costs of living" that was based on a series of surveys involving working people from 15 cities in Europe and the United States. In 1935, the LON Health Committee launched a programme to assess levels of nutrition in the world by adopting universal standards on calorie and nutritional requirements, and guidelines to harmonize the methods of investigation. In 1938, the ILO carried out another study on "the living standards of workers" in which, for the first time by an international organization, some "generally accepted objective standards" of living were proposed (Bonnecase, 2011).

Such activities of measurement and comparison, while enabling a figure to be put on inequality in the world, remained however largely confined to a white and civilised world: colonized populations were generally excluded from this common reference point. Significantly, discussions at the ILO and LON during the inter-war period, when touching on Africa, did not concern living conditions and social legislation, but rather so called specifically African problems, especially slavery and forced labour (Cooper, 2004), the persistence of which was an ongoing concern in international fora. However, when the issue under consideration was health, nutritional conditions or mortality trends, the debate tended to focus mainly on industrialized countries. Indeed, colonial administrations had very little quantitative data on these issues. While the improvement of living conditions of colonized peoples was a key aspect of the colonial justification for conquest, it was not considered that this required measurement in order to be demonstrated: as evidence of such improvement, reports of clinics, agricultural development work and schools counted more than statistics. Certainly, colonial administrators produced an abundance of statistics within the framework of the routine management of their territories, but they related more to the colonial work itself, rather than to the lives of colonized peoples. What tended to be measured were things like the number of people treated or vaccinated in the clinics, the number of tonnes of grain produced in a region, the number of children enrolled in schools, i.e. the performance of the different colonial services. However, such data ultimately said very little about the way people lived in colonies: no administrator sought to specify to what extent colonized populations were likely to eat a poorer diet or to die younger than European populations. Inequality between these two population groups was not measured until the early 1940s, simply because it was considered that they did not belong to the same world.

Mobilization and internationalization of knowledge

Following the Second World War, the colonial empires were embroiled in a crisis of legitimacy led by heterogeneous actors, including new United Nations (UN) bodies and mobilized populations within colonies. French and British Africa had already witnessed great strikes in the late 1930s. In the British Empire, these strikes contributed to the adoption of the Colonial Development and Welfare Act of 1940, which represented a real break in colonial policy: the improvement of living conditions in the colonies, which until then was based primarily on local budgets, became the object of metropolitan public investment. In the French Empire, a similar break occurred in 1946 with the establishment of investment funds for economic and social development, while the colonies were shaken by a new phase of protests (Cooper, 2004).

This combination of institutional reforms and social mobilizations fundamentally altered the understanding of living conditions within imperial territories. On the one hand, new laws and policies to establish the transfer of funds from metropoles to colonies, encouraged administrations and the public to put the socio-economic disparities between the two population groups under further scrutiny; while on the other hand, colonial populations that had been mobilized in the fight to improve their living conditions eventually began to compare their own situation with those living in the metropolitan state. For example, railway workers in French West Africa, who were among the originators of the major strikes that caused alarm in the post-war colonial federation, requested that their working conditions and wages were determined according to the same rules as applied throughout the entire Empire. It was ultimately colonial differentialism, in terms of a growing awareness of living conditions, that was at the heart of the conflict.

The then newly formed UN organization became one of the main initiators of external pressure, on top of the internal social protests in imperial territories. In the post-war period, the UN quickly adopted an anti-colonial stance under the influence of the USSR and the United States, countries that officially opposed colonization, and also that of former colonies which had gained independence. While most criticism focused on the legal status of colonized peoples and their exclusion from the political sphere, the issue of living conditions was also taken on board. According to the charter adopted in January 1945 by the UN, Member States which "assume responsibility for the administration of territories whose peoples do not administrate themselves yet [...] accept the obligation to promote as much as possible their prosperity". They commit themselves "to this end to communicate regularly [...] statistical information and others of a technical nature relating to economic, social and educational conditions in the territories for which they are respectively responsible" (United Nations Charter, Article 73 e). According to this clause, itself marked by an "ideology of development" that grew following the War (Rist, 1996), colonial powers were obliged to provide statistical proof of their willingness to improve the social and economic situation of colonized populations.

This political reconsideration of living conditions in the world in general, and in the colonies in particular, went together with the technical effort within international organizations to produce new evaluation standards. Each specialized institution exerted standardizing actions within its own field. FAO, for example, in 1946 launched the first major "world food survey" and three years later instituted a "Committee on Calorie Requirements", which reviewed the validity of the pre-war minimum food requirement standards that were set by the LON, but which had most often been applied only to populations in Europe and North America. The primary objective was to collect standardized food information in territories where it was clearly lacking: in retrospect it was realized the extent to which, when undernourishment and malnutrition was considered in the 1930s, "we were almost always thinking of Eastern and South-Eastern European countries, with the Bosphorus [appearing] as the limit of the intellectual horizon" (FAO, 1955). Subsequently, the mental picture of food inequality was expanded to include non-European populations.

More generally, it was the overall measurement of living conditions across the world that became the subject of debate within international organizations. The United Nations General Assembly (UNGA) thus adopted a first resolution in 1949 requesting the Economic and Social Council to assess the social situation in the world from "quantitative indices of satisfaction of needs whose existence is universally recognized". Three years later, it adopted a second resolution in which it called for the development of "adequate statistical methods and techniques so as best to facilitate the gathering and use of pertinent data in order to enable the Secretary-General to publish regular annual reports showing changes in absolute levels of living conditions in all countries" (Bonnecase, 2011). The intention was to bring all of the world's populations into the same field of reference, where differences were both comparable and measurable, so that it became no longer only living conditions that were under discussion, but also living standards.

This led to the formation of a committee of experts from the United Nations Economic and Social Council, which in the early 1950s was charged with considering "the definition and evaluation of standards of living in an international perspective". It included economists, demographers and other experts who, like Louis-Joseph Lebret in France, had previously worked on the theme of poverty in different countries and was by then dedicated to the study of inequalities in living standards internationally. Several controversies afflicted the work of this committee. The main one concerned the relevance of a single index for the standard of living, since "the use of a uniform and universal system of values" could contribute to "penalising international technical progress in the West". Infant mortality rate and life expectancy were discussed at this time, although it was said that "longevity does not necessarily indicate a degree of sanitary quality". However, the main statistic highlighted was the national income per capita: it appeared as a "fairly comprehensive indicator of the determinants of living conditions as a whole", even though it may generate "rather false conclusions on the differences between living standards in different regions of the world", since "the non-market goods and services are partly excluded from the measurement" (Economic and Social Council of the United nations, 1954).

These controversies are of interest as they foreshadow more recent debates, particularly on the dominance of productive indicators in the measurement of well-being, but also on the relevance of universal indices applied to different populations, regardless of how these populations may experience their own living conditions (Destremau, Salama, 2002): such controversies, far from appearing in the last twenty years, were present from the outset in the history of the international measurement of living standards.

Compiling statistics

The international demand for statistics was locally translated into a proliferation of surveys on living conditions in the colonies, which was sometimes met with resistance from colonial administrators who were initially responsible for such investigations and often reluctant to internationalize knowledge. Taking the food situation in French West Africa as an example, the first ever research conducted here into nutrition was entrusted to a new agency, known as the Anthropological Mission, largely comprising military doctors and pharmacists. In the second half of the 1940s, the Mission carried out an extensive survey in several of the federation's colonies, which involved the observation of family meals and the weighing of food that went into the composition of these meals. For the first time the French government was seeking to quantify, according to clearly defined methodological procedures, the average calorie and nutritional intake in African towns and villages. However, underlying this innovative research was the racial assumption that Africans and Europeans did not necessarily have the same physiological needs: the fact that the people surveyed did not have a more malnourished bodily appearance, given their calculated food intake, led to the assumption that "the body of black people does not operate like [that of the whites]" (Pales, 1954). Investigators then tested this hypothesis for a number of nutrients before dismissing it, on a case by case basis, and concluded the uniqueness of the human physiology.

Despite these prejudices, research did however reveal the relative importance of nutritional problems in African populations, which was not obvious at the time. According to the results, the average calorie intake over one year was sufficient on average, while showing large variations depending on the timing and on the location studied: the strong decrease in the level of consumption during the "hunger gap", although a widely known phenomenon, appeared in retrospect to have been underestimated. The same applied to qualitative deficiencies, which appeared relatively high in terms of what was considered by FAO nutritionists to be normal requirements. The administration, however, did not readily accept these findings as evidence, particularly due to the small number of surveys on which they were based, which called into question their validity in terms of the whole of West Africa. Furthermore, the researchers knew that the people they observed strongly distrusted them, meaning that, in the eyes of the former, there was an additional parameter that significantly distorted the value of the results: for example, when the calculated intake appeared too low, the studied populations were sometimes suspected of concealing a proportion of their nutrient intake; or, conversely, when the intake was higher than expected, investigators suspected individuals of making exaggerated claims about their wealth (Bonnecase, 2009).

In the 1950s, this process of quantifying calorie intakes was continued, but under the auspices of new players: the colonial administrators, whom had previously held a quasi-monopoly on research in the colonies, began to be gradually replaced by professional nutritionists, sociologists and economists. New organizations acquired a central role in the compilation of data on living standards, especially ORSTOM (Office of Overseas Scientific and Technical Research), which was created in 1944, and the INSEE (National Institute of Statistical and Economic Studies), established in 1946. Methods of investigation were also changing, particularly through the adoption of the sampling technique. Thus, in 1957, the first food surveys based on probability samples were conducted in francophone West Africa. The same type of research was carried out in British colonies where Nutrition Committees were established after the War. By the late 1950s, Africa was still not yet considered as the main continent of hunger, in comparison to perceptions thereafter. However, compared to the situation ten years earlier, there was increasing recognition - backed up by data - of the level of undernourishment and malnutrition experienced by a number of African populations (Bonnecase, 2011).

A similar trend was emerging in the assessment of health levels, while demographics and infant mortality rates in particular were regarded by international organizations as the key indicators. In the immediate post-war period, colonial administrations had only a very small amount of synthetic data on the issue: most of the demographic knowledge available was based on administrative censuses that were regularly organized in various districts for tax purposes. Besides the fact that these figures did not paint a remotely accurate picture of reality, because many people developed strategies for concealing information from the census enumerators, they had not previously been regarded as of particular significance in terms of living conditions. If administrators were keen to take up the fight against "the depopulation of Africa" in the inter-war period, known in the French Empire as the "decline of the black race" (Conklin, 1997), it was mainly because it was seen as an obstacle to the development of the continent,.

During the 1950s, demographers in Africa instigated the first large population studies. In 1950, the French Imperial government ordered a review of its territories, not only to investigate whether population growth was compatible with an "industrial take-off", but also to assess "the elevation of the health status and the improvement of the way of life" (Bonnecase, 2011). Some demographic sample surveys were organized by INSEE agents which, for the first time, enabled the establishment of synthetic data at the level of entire colonies. Certainly, even in the eyes of their creators, these surveys produced results that appeared quite random, a fact that was due to the material difficulties encountered during the investigations, as well as the people's persistent distrust of enumeration efforts. However, these figures, especially the ones on infant mortality, enabled a hierarchy of health to be drawn up in which African territories now had their place. These data, in the absence of alternatives, continued to be regarded as definitive by international bodies until the first general population censuses on the continent in the 1970s.

National income was a third fundamental research area that contributed to the ranking of colonial territories on a scale of living conditions in the post-war period. On this aspect, a lack of knowledge was not a colonial specificity: even in industrial countries it was not until the 1930s or 1940s that the public authorities made regular records of national income while also developing their economic policies in the wake of Keynesianism (Vanoli, 2002). In the United States, the Senate called for the establishment of "national income estimates" for the first time in 1934, in a context of new investment policies that had been set in place to deal with the crisis; in Great Britain, such calculations became systematic from 1941 for the purpose of rationalising the war effort; in France, this trend towards modern national accounts emerged slightly later with the creation in 1945 of the Commissariat du Plan and the INSEE in 1946 (Desrosieres, 2003). The earliest calculations of national income in the colonies was closely linked to the launch of proactive investment, the formulation of which, if not the practice, required new macroeconomic data. However, the compilation of such data also reflected the will to evaluate the effects of policies "on raising living standards, [...] by generating from all existing statistical data a single figure for each country and for each year, a figure which should indicate what is commonly called, in European countries, the national income" (Ministère de la France d'Outre mer, 1955).

The first national incomes of colonial territories were obtained from 1948 onwards in British Africa, and from 1951 in French Africa: some statisticians, generally from the Central Statistical Office or the INSEE, sought to apply an accounting system in overseas territories that had already been applied in the metropoles. However, they remained extremely cautious about the value of their results, in part because the validity of the statistical production on which their calculations were based remained questionable, due to the practical conditions in which they were developed by the colonial administrations, and secondly because it was difficult to define "the boundaries of production" (Jerven 2009), i.e. to agree on which activity sectors - especially the non-monetized sectors of the domestic economy - should be included in the calculations. The meaning accorded to the averages obtained was also a subject of controversy: while figures on national income per capita were intended to reflect the standard of living of colonial populations, the interpretation and "international comparisons that could arise" remained highly objectionable to statisticians, because of the strong heterogeneity of calculation conditions from one country to another (Courcier, Dubois, Fabre, 1958). Harmonization efforts were undertaken in the 1950s under the auspices of the UN (United Nations statistical office, 1957), which at the beginning of the next decade gave rise to the establishment of "national accounts manuals adapted to developing countries" (Courcier, Le Hegarat, 1963). But many differences remained between countries with regard to the classification of economic agents, the monetary transcription system of non-market activities, the manner in which they were or were not included in the calculations or, more fundamentally, empirical data upon which the evaluation of production was based (Courcier, Ady, 1961).

The circumstances under which the statistics on national incomes developed did not however prevent them from being used in ways that their designers had not intended, and once established it wasn't long before international fora appropriated the data. Statistics on hunger or infant mortality were used to formally place colonized people into an international hierarchy of living standards, placing them on the same continuum as non-colonized populations. Viewed through the prism of "development", these scales indicated the degree to which each territory had "advanced". Although this placed them in a position of relative inferiority, it appeared that they were now being promised the same social and economic future as developed countries: the gap between colonized and non-colonized populations became measurable, which simultaneously gave rise to the belief that it would be possible for them to catch up.

Discussing inequalities in an imperial context

The significance of such statistics, beyond the technical conditions in which they were developed, therefore became eminently political. Not only did the data serve to highlight the existence of food, health and socio-economic issues in the territories concerned, but it also contributed to the emergence of a new cognitive framework from which the world's inequalities could be expressed, thereby providing a basis for new demands for equality.

These demands first manifested through international activism, and especially through the notion of "Third Worldism" which developed during the 1950s, a notion which takes its name from the works of Alfred Sauvy, a French economist and demographer who coined the term Third World in 1952, who referred to this Third World as "exploited, scorned like the Third Estate", and which "also wants to be something" (Sauvy, 1952). This ideology, in addition to the various economic theories and political struggles that structured it, was initially fuelled by data that showed the inequalities between the "third world" and "developed worlds". In the symbolic area of food, the 1950s were the years that saw the emergence of the concept of "world hunger", which implied the comparison of levels of food and nutrition throughout the world. At this juncture it is worth noting the story of Josué de Castro, a Brazilian geographer who was the first pamphleteer of books on this subject: he conducted research into the food shortages that afflicted his own country in the 1940s, before establishing in the 1950s a global "geopolitics of hunger" largely informed by statistical data (Castro, 1951), which served as a reference for many international campaigns over the next decade.

In the 1950s, knowledge on living standards also provided new opportunities for, sometimes antagonistic, discussion for the political actors in the colonial empires. From the point of view of metropolitan governments, they served as a means of legitimization while new policies of investment were implemented in the colonies. For example, it became possible through the calculation of data on national income to quantify the "overall progress" delivered by colonisation, whatever the technical quality of the figures. In the French Empire, the strong growth of African economies during the 1950s was thus quickly indexed by the colonial authorities, both as evidence that "thanks to French efforts, living standards had risen in colonies" and as an encouragement to "pursue development policies" that had been undertaken 10 years previously. Admittedly, the per capita income in African territories is often "characteristic of underdeveloped countries". But other nations, among the most opposed to colonialism in international fora, are "equally underdeveloped" and even "in a more serious manner" (AOF General Government, 1953).

Conversely, from the perspective of trade unions and the mobilized political parties of the colonies, the new data on living standards supported a denunciation of poverty, a state in which many people within the empire continued to live in, even calling into question the very idea of "colonial achievement". In the French Empire, newly elected African Parliamentarians thus became able to highlight alarming statistics to the National Assembly, such as "dietary intakes that were often less than 2000 calories" and "rates of infant mortality between 25% to 50%, compared to 4% to 7% in European countries", denouncing the insufficient efforts to improve living conditions. In the colonies, trade unions, equipped with supporting figures, became able to demonstrate the contrast between the "imperialist bourgeoisie, living in luxury," and "the poverty experienced by the vast majority of colonial populations" (Bonnecase, 2011). Within colonies, the new statistical knowledge even enabled the quantification of differences that had never previously been measured: in French West Africa for example, national income data were used to calculate that "living conditions" in the federation were on average 40 times higher "among whites" than "among blacks", finding that "modern sector revenues accrue mainly to the former" while income from "mixed and traditional sectors" were more for the latter (Capet and Fabre, 1957). Ultimately, a new expression of justice had emerged within the imperial territories: while colonial knowledge on the social situation in colonies had long been based on the idea of a natural difference between the colonized and those who were not, it became possible with the introduction of statistics to imagine that the former could become equal to the latter, not only in terms of political rights and legal status, but also in terms of material living conditions.


In The statistical argument, Alain Desrosières explains that "applying and building a space of equivalence enabling quantification, and thus measurement, has both political and technical results. It is political in that it changes the world: comparing the aristocracy with commoners led to the night of the 4th August [abolition of feudalism during the French revolution]; comparing blacks with whites called for the abolition of slavery; comparing women with men called for a real universal suffrage that included women" (Desrosieres, 2008). Statistical information, such as that described in this article, has undoubtedly helped change the way in which people see the world, if not actually directly changing it. We should not, however, credit such data with too much importance nor view them as causal agents in the history of political change: ultimately, these statistics only served as a new tool to engender discussion on the reconfiguration of power relations, but they have not in their own right changed the nature of inequality.

Finally, how can such a story enlighten us in terms of the modern controversies on the measurement of inequality in the world? The first lesson is that arguments over the choice of a "good indicator" - and over the place granted to productive criteria in particular - have not appeared in the 1990s with the advent of the Human Development Index, or in the 2000s with discussions on the "Measurement of Economic and Social Performance" (Fitoussi, Sen, Stiglitz, 2009): they were already part of the international debate on the definition and evaluation of living standards in the 1950s. The second lesson is that these first statistics on the world's living standards were used without consideration for the material conditions under which they were developed, which again foreshadows more recent realities: while contemporary statistics on poverty can sometimes resemble a "great swindle" (Lorraine Data, 2009) in both the North and the South, this does not prevent them from indicating realities and being used to either manage or refute them. A third lesson is that the first international standards of living statistics have contributed to the building of a common space of possibilities, and this may nuance the normative and somewhat oppressive dimension that they can also carry. The emergence of post-war "development" has often been seen as a device of power, the imposition of a social and economic trajectory onto large populations that was mainly informed by Western history (Escobar, 1995). However, historically this knowledge was also considered by some of these populations, that had experienced a form of government based on race, as a means to express a fundamental claim: to be considered, if not measured, in the same way as everyone else.

Measuring living standards internationally

The use of living standards indicators to enable comparisons between all of the world's countries is a recent approach. The first trials of these indicators were conducted in the mid-1950s, showing that the objective to take multiple facets of life into account is not new, but the earliest attempts were hampered by the absence of measurable factors or statistical systems.
Show Media

Initial assessments of infant mortality in Africa

Initial demographic sample surveys, aiming in particular to measure infant mortality, have been conducted in sub-Saharan Africa since the 1950s. The results, despite uncertainties, highlight significant differences across the world, with the infant mortality risk being three to ten times greater in Africa than in Europe.
Show Media

Initial comparisons of national income per capita in developing countries

Income measurement and its comparison have given a new prominence to inequalities in the world, while also reinforcing the idea that the need to catch up with developed countries should be at the heart of any development policy.
Show Media