The Paradox of Black Incomes in Puerto Rico in the Early Decades of U.S. Colonialism

Abstract

This paper examines racial income inequality in early twentieth-century Puerto Rico. It finds, surprisingly, that Black men had an income advantage relative to White and Mulatto men in 1910–1920. The effect of race on income in Puerto Rico was smaller than that of other covariates such as urban status, sex, and literacy. A comparison with the state of Louisiana and with the United States as a whole in the same Census years shows that Puerto Rico was exceptional by U.S. standards, displaying much lower levels of racial inequality. Most of the income advantage Black men had can be attributed to the fact that they were more urban than Mulatto or White men, but part of this surprising advantage can be attributed to the existence in the countryside of a layer of skilled Black workers. Overall, Black men had equal or slightly higher occupational scores than Whites. The coexistence of slavery with other forms of coerced labor affecting individuals of all races in the nineteenth century, as well as the emergence of a stratum of Black skilled workers which survived into the twentieth century and thrived economically when the sugar industry experienced an explosive boom after 1898, is at the root of Black income equalization in the Puerto Rican countryside and in the island as a whole during the early twentieth century.

The past several decades have seen an expansion of interest in racial inequality in Latin America, especially for the purpose of making comparisons with the United States. ¹ The most salient work has focused on Brazil, and has found that levels of socioeconomic inequality along racial lines in that country often match or even surpass those found in the United States. ² Puerto Rico should serve as another comparative case to inform our understanding of race and racial inequality in the Americas. The island has a significant population identifying as Afrodescendant, with 17.5 percent of the island’s population identifying as “Black” alone or in combination with another race in the most recent, 2020 census. ³ And, conveniently for comparisons with the mainland United States, Puerto Rico counts demographic trends in its population using the same census categories.

Yet studies on racial inequality in Puerto Rico, particularly regarding the twentieth and twenty-first centuries, are remarkably scarce. Most existing scholarship on race and racism in the island lacks concrete measures of inequality. In an attempt to address this dearth of scholarship and advance the study of the Puerto Rican case in comparative research on racial inequality in the Americas, this article examines racial income inequality in early twentieth-century Puerto Rico, at a time when the island’s society was transitioning from the period of Spanish colonialism (1508–1898) to a new colonial regime under the United States (1898–present). It makes use of 12 percent samples of the U.S. Census conducted on the island in 1910 and 1920, stored in the Integrated Public Use Microdata Series (IPUMS), which contains a variable called “occupational score,” a proxy for income. ⁴ Analysis of the data reveals an interesting paradox: in 1910 and 1920, when the population was classified by the U.S. Census into “Blacks,” “Mulattos” and “Whites,” Black men had on average higher occupational scores than men of the latter two racial categories. ⁵

A brief review of the literature on racial inequality in Puerto Rico is followed by measurement of inequality in the early twentieth century in comparison to other covariates such as sex, literacy, and urban status. Inequality on the island is juxtaposed to inequality in the state of Louisiana and in the United States as a whole, utilizing equivalent data from the U.S. Census. A final section examines the historical roots of the relatively low levels of racial inequality in Puerto Rico in the early twentieth century when compared to the United States.

The “Silence” on Racial Inequality in Modern Puerto Rico

Racial inequality in Puerto Rico has received relatively limited attention. Historians have paid more attention than sociologists, and most of their work focuses on the eighteenth and nineteenth centuries, not on the twentieth century or contemporary conditions. In the 1980s, Francisco Scarano demonstrated that productivity per-slave in the sugar plantations of Ponce, Puerto Rico’s principal plantation district in the nineteenth century, was higher than in Louisiana or anywhere else in the Caribbean, an indication of the intensity of exploitation of slave labor in Puerto Rico. In the process, Scarano blew up myths about the supposedly benign nature of slavery on the island. ⁶ Jorge Chinea has studied the immigration of skilled Afrodescendant laborers from the Eastern Caribbean to Puerto Rico in the period 1800–1850. ⁷ Luis Figueroa has examined in detail the contracts into which libertos, or freedmen, entered after abolition in 1873, the labor demands of these freedmen, and their opportunities for skill acquisition. ⁸ More recently, David Stark has offered a detailed account of family formation among slaves in Puerto Rico’s eighteenth-century hato, or cattle-ranching economy, showing that family formation among slaves was more prevalent and stable in the eighteenth century than in the nineteenth, when the condition of slaves deteriorated as a result of the emergence of a classic plantation economy. ⁹ In a masterful study based on more than a hundred legal cases surrounding marriages, María del Carmen Baerga illustrates a complex pattern of intertwined racial, social, and economic mobility in nineteenth-century Puerto Rico and demonstrates that the meanings of the word “race” on the island at the time were very distant from contemporary conceptions. ¹⁰

If on the Spanish colonial period there is a rich body of literature on racial inequality and the political economy of race in Puerto Rico, with respect to the U.S. colonial period there emerges a separation in scholarship between work that focuses on race and work that focuses on inequality and political economy. In the 1930s, important works were produced in Puerto Rico criticizing the plantation economy and sugar monoculture, ¹¹ income inequality produced by the plantation economy, ¹² and rural poverty and disease. ¹³ But for reasons that probably reflect the relatively large size of a White, impoverished rural proletariat, and the lack of salience of race as a category to many of these rural proletarians, ¹⁴ these studies emphasized class and gave little attention to race.

Conversely, studies of race in twentieth- and twenty-first-century Puerto Rico mostly ignore political economy. A classic reference point is the study by Tomás Blanco, which claimed in the late 1930s that racial prejudice in Puerto Rico was relatively benign. ¹⁵ In a revision to the original article published in 1942 as a booklet, Blanco put forward the notion that, compared with the violence of racial segregation and discrimination in the United States, “our prejudice is innocent child’s play.” ¹⁶ Blanco’s study on “prejudice” was limited to impressionistic data, and did not contain any measurements of socio-economic inequality. It has elicited multiple eloquent criticisms, but these are typically also circumscribed to notions of “prejudice” and lack empirical research about socioeconomic differences by race. ¹⁷

A mid-twentieth-century surge of interest in Puerto Rico as a “social laboratory,” ¹⁸ meanwhile, produced some studies of race and racism on the island. These works range from statements about the inexistence of racial categories in Puerto Rico to more sober assessments about the role of race in Puerto Rican society. Some authors argued, along the typical apologetic axis of some mestizaje theories, that “racial discrimination becomes virtually impossible, since a majority of the people are of mixed Indian, white, and Negro blood,” while at the same time arguing for the predominance of class over race: “Dividing Puerto Ricans into socio-economic classes is perhaps simpler than separating them according to race.” ¹⁹ Some saw a racial system in which “anybody with a drop of white blood is a white man.” ²⁰ Some viewed racial discrimination as a middle- and upper-class phenomenon exclusively: “Apart from small groups of the middle and upper class any ordinary gathering of Puerto Ricans represents a striking and unmistakable example of the complete acceptance of social intermingling of people of different color and racial characteristics.” ²¹ Sidney Mintz argued that an individual’s color may vary according to their socioeconomic status in a system in which racial criteria could not be completely disentangled from socioeconomic status criteria. ²² While Mintz circumscribed his findings to the Puerto Rican rural working class, other studies examining upper segments of society found very distinct patterns of racial discrimination. ²³ Eric Williams, the paramount historian of the Caribbean, argued similarly that the “absence of legal discrimination against the Negro arises from the fact that racial differences are subordinate to those of class.… Thus it is that, by virtue of the absence of legal discrimination, the high degree of social mobility, the emphasis on class and the political equality that prevails, unity among Negroes on the race question does not exist in Puerto Rico.” ²⁴ While some of these studies pointed to socio-economic differences organized by race, for example the observation that “colored workers” were overrepresented in the population of urban slums, ²⁵ none of the studies quoted above contained systematic quantitative assessments of racial inequality in income or living conditions.

Some recent works have focused specifically on the issue of racial categorization, but without empirical assessments of levels of inequality. ²⁶ Most studies of race in Puerto Rico under U.S. rule have focused on the ideological or discursive level, examining issues of cultural expression, prejudice or discrimination but largely ignoring measurement of concrete social and economic inequality in areas such as income, home ownership, educational achievement, health outcomes, rates of incarceration, or other measures of the wellbeing of Afrodescendants relative to Whites. ²⁷ The recently published memoirs of the First Congress of Afrodescendance (afrodescendencia) in Puerto Rico are a telling example of this lack of measurement of social inequality. Among thirty-three papers published, covering cultural issues, music, poetry, literature, self-organization, and multiple other topics, not a single article contains empirical research on income, housing conditions, health, education, or other indexes of well-being among Afrodescendants when compared to Whites or indeed any other racial group. ²⁸

In short, research on race in twentieth and twenty-first-century Puerto Rico is characterized by a remarkable lack of attention to concrete measures of living standards and socioeconomic inequality. The following sections analyze patterns of socioeconomic inequality along racial lines in the early twentieth century in Puerto Rico and seek to explain these patterns historically.

The Paradox of Higher Black Incomes in Puerto Rico in 1910: Occupational Scores and Occupational Structure

In order to adequately understand racial inequality in early twentieth-century Puerto Rico, it is useful not only to compare racial inequality to other axes of inequality within the island such as urban status, gender, or literacy, but also to make external comparisons to the mainland United States. Because the Puerto Rico data is from the U.S. Census, the same variables appear in the microdata samples for the United States in 1910 and 1920. Louisiana probably offers the best grounds for comparison with Puerto Rico among all states of the Union. As shown in Table 1, in 1910 Louisiana was highly rural (69 percent) like Puerto Rico (71 percent). It had Black and Mulatto (“creole”) populations, and a significant sugar industry, also like Puerto Rico. ²⁹ Although Louisiana’s literacy rate in 1910, at 67 percent, was higher than Puerto Rico’s (27 percent), it was the third-lowest rate of any state in 1910.

Table 1. General Demographic Indicators: United States, Louisiana, and Puerto Rico, 1910–1920

In our comparative analysis of racial inequality in early twentieth-century Puerto Rico, we rely on 12 percent samples of the U.S. Census conducted on the island in 1910 and 1920, stored in the Integrated Public Use Microdata Series (IPUMS), which contains a variable called “occupational score,” a proxy for income that sorts workers according to the average income in their occupation. OCCSCORE is flawed in that it assigns all members of an occupation the same score, such that using it to calculate overall racial differentials in earnings tilts the scale towards racial equality. Indeed, there is evidence that the variable OCCSCORE biases racial differences towards zero when compared to true earnings. ³⁰ However, using U.S. data to compare OCCSCORE to true earnings, Martin Saavedra and Tate Twinam find that OCCSCORE yields a clear gap in earnings by race and the sign is always in the same direction as true earnings. That is, Whites always earn more than Blacks, and the pattern in the size of the gap in OCCSCORE closely tracks the pattern of true earnings. These results support the use of occupational scores to assess racial inequality when other data is not available, as is the case here, as long as we acknowledge OCCSCORE’s limitations.

Table 2 displays average occupational scores in relative terms, with scores of different race/gender intersectional groups expressed as a percentage of the average score of a White male, in Puerto Rico, Louisiana, and the United States as a whole. ³¹ To understand the Puerto Rican figures, it is necessary to first be aware of the urban/rural divide in Puerto Rican society at the beginning of the twentieth century, illustrated in comparative terms in Table 3. More than two-thirds of the population (71 percent) was rural in 1910. There were stark differences in income between town and country, with the average rural income being only 49 percent of urban income. Within this context of sharp differentiation of income by urban/rural status, the racial distribution of the population along the urban/rural divide mattered immensely. Even though within the cities the incomes of Blacks were lower than those of Whites, on the island as a whole Black incomes were higher, for two reasons: (1) Blacks were the most urban group of the population—45 percent of Blacks lived in cities, compared to just 24 percent of Whites; and (2) in the countryside, Black, White, and Mulatto men had more or less equal average incomes. ³²

Table 2. Relative Incomes of Men and Women by Race in Puerto Rico, Louisiana, and the United States, 1910–1920

Table 3. Urbanization Rate by Racial Group, 1910–1920

Table 4 reports the relative occupational scores for different categories of the population in Puerto Rico, Louisiana, and the United States according to urban status, gender, and race, for people ages sixteen to sixty-five who were in the labor force. The income of a White, rural male worker is set at one hundred to express all incomes in relative terms in each location. Rural, White, male workers comprised the modal category within the labor force. In Puerto Rico, the incomes of urban Black males are considerably higher (179 percent) than the income of rural White males. Likewise, the urban-rural disparity being large, White urban women had slightly higher incomes than White rural males (104 percent). Rural Black males had an average income that was practically equal to that of rural White males (99 percent). This is in sharp contrast to the United States, where income differentials between men of different races were much larger.

Table 4. Relative Income by Intersectional Category, 1910-1920 (White Rural Men in Each Place = 100)

* People in the labor force ages sixteen to sixty-five.

Simply put, in 1910 there was more inequality in income based on race in Louisiana and the United States than in Puerto Rico. These observations are consistent with Ayala-McCormick’s findings for the twenty-first century that racial inequality in incomes is much less severe in Puerto Rico than in the United States or in other Latin American societies, such as Brazil, which tends to be used as the paradigmatic “Latin American racial system.”

Returning to the question of imperfect income proxies, does OCCSCORE bias racial inequality toward zero in Puerto Rico? Histograms of the distribution of occupational scores in Puerto Rico and Louisiana clearly demonstrate very different racial occupational distributions. In the Puerto Rican case, the proportion of Whites in the bar representing the second-lowest quantile is higher than the proportion of Blacks. In the middle of the distribution, between OCCSCORES of 20 and 40, the proportion of Blacks is higher than that of Whites in Puerto Rico. The situation is very different in Louisiana, where in the left-hand side of the distribution, representing the lowest scores, the proportion of Blacks surpasses that of Whites, and conversely, on the right side of the distribution, representing the higher scores, the proportion of Whites clearly surpasses that of Blacks. It should be noted, however, that the almost total exclusion of Blacks from the very highest occupational scores is common to Puerto Rico and Louisiana.

The differences in occupational scores highlighted above reflect differences in racial occupational distributions. To begin with, Whites in Louisiana display a more diverse occupational structure than either Blacks in Louisiana or Blacks and Whites in Puerto Rico. As illustrated in Table 5, a full 35 percent of White workers in Louisiana had occupations other than the ten top occupations, whereas for Blacks in Louisiana, Blacks in Puerto Rico, and Whites in Puerto Rico, the top ten occupations encompassed 90 percent, 83 percent, and 86 percent of workers, respectively. In Louisiana, farmers represented 27 percent of Black and 23 percent of White occupations. It should be noted that this category included owners and tenants, and probably included a large number of sharecroppers. However, Puerto Rico had a much smaller percentage of farmers—9 percent of Whites and 3 percent of Blacks—and a much larger proportion of landless laborers, with the category of farm workers accounting for 53 percent of Black workers and 55 percent of White workers.

Table 5. Percent of the Labor Force Engaged in the Top 10 Occupations

Note: nec = not elsewhere classified

A striking difference is the relatively small number of White rural proletarians in Louisiana (7 percent), whereas in Puerto Rico the proportion of dispossessed rural workers was actually slightly larger among Whites (55 percent) than among Blacks (53 percent). Another notable difference is the existence of a layer of Black skilled workers in urban contexts in Puerto Rico: for example, carpenters and Blacksmiths were among the top 10 occupations for Blacks in Puerto Rico, but not among Whites. In short, Puerto Rican Whites were relatively worse off than Louisiana Whites, and Puerto Rican Blacks were relatively better off occupationally than Louisiana Blacks. Overall, the most significant differences between the distribution of occupations in Puerto Rico and Louisiana was the existence of a large White rural proletariat in Puerto Rico and its absence in Louisiana, on the one hand, and the existence of a stratum of Black skilled workers in Puerto Rico and its absence in Louisiana, on the other.

Finally, a linear regression of income (OCCSCORE) on demographic indicators provides evidence that the effect of race on income was small after controlling for other covariates in Puerto Rico, both in relation to other axes of inequality within the island and in relation to the United States (Table 6). ³³ The analysis shows that binary variables for rural status, being female, and being illiterate all had greater negative effects on income than being Black in Puerto Rico (model 1). The negative effect of rural status on income score was more than three times greater than the coefficient for “Black” and almost six times greater than the coefficient for “Mulatto.” Being female as opposed to male also has a negative effect that was more than two and a half times greater than that of being Black relative to the category White. Illiteracy had a similar effect on income, with a negative coefficient two and a half times greater than that of “Black.” Curiously, nativity is also a stronger predictor of income, with native-born Puerto Ricans earning much less than non-natives. ³⁴ Finally, married and older persons tend to earn more, but the effect of age tends to decrease over time. Both variables have a smaller effect on income than race.

Table 6. Fixed Effects Regression Models for Occupational Score, People Ages Sixteen to Sixty-Five in the Labor Force (1910–1920)

Robust standard errors in parentheses; *** p<0.001, ** p<0.01, * p<0.05, + p<0.1

The negative coefficient expressing the disadvantage of Black workers was almost three times greater in Louisiana (Model 2) or the United States (Model 3) than in Puerto Rico. Thus, not only was the negative effect of being Black on income relatively small when compared to other axes of inequality within Puerto Rico, it was also much smaller than the same effect in Louisiana and the United States. A question for future research is why the disparity in female income by race is larger than for male income by race in Puerto Rico.

While much of the explanation for the relatively high income of Black men relative to White men resides in the fact that the Black population was more urban than the White population, in the countryside in Puerto Rico Black men earned on average essentially the same incomes as White men in 1910–1920 (99 percent). This is contrary to what one would expect in a society which had plantation slavery for centuries. How can this be explained?

Historical Sources of Occupational Diversity

The striking difference detailed above in levels of socioeconomic inequality by race between Puerto Rico and the United States is part of a larger difference between the history and legacy of slavery in the Iberio-American world and slavery the United States. To begin with, in the late eighteenth and nineteenth centuries there were large populations of free people of color in the Spanish colonies and in Brazil, a situation that is in sharp contrast with that of the United States, where African ascendancy and slave status were much more closely associated. In most regions of Latin America, the vast majority of Afrodescendants were free in the year 1800. The ratio of the free to the enslaved population among Afrodescendants before independence from Spain were as follows in the colonial territories which correspond to the following modern countries: 62 to 1 in Mexico, 17.5 to 1 in Paraguay, 9.25 to 1 in Panama, 5.6 to 1 in Ecuador, and 4.0 to 1 in Colombia. Only in Brazil and Cuba, societies which actually experienced expansions in slave imports and in the number of enslaved persons in the nineteenth century, were there more slaves than free Afrodescendants in the year 1800. The free to enslaved ratio was .81 to 1 in Brazil and .52 to 1 in Cuba. ³⁵ Even then, in Brazil the demographic balance shifted dramatically in favor of free persons as the nineteenth century advanced, so that by 1872, a population of 4,245,428 free Afrodescendants constituted 42.8 percent of the total population, outnumbering the 1,510,806 slaves, who made up 15.2 percent of the total population. This means that in Brazil, the country which received by far the largest number of slaves during the epoch of the slave trade, there were almost three free Afrodescendants for every slave in 1872. In the United States, by contrast, the slave population comprised 18 and 13 percent of the total population in 1790 and 1860, respectively, but free people of color were a mere 1.5 percent of the population in both years; there were more than ten slaves for every free Afrodescendant both in the late eighteenth century and on the eve of the U.S. Civil War. ³⁶

The size of the free populations of color made them central to the functioning of Latin American economies and provided opportunities for the acquisition of skills and assets that could not be acquired or transferred so easily among slaves. The existence of large free populations of color in colonial times distinguishes Latin America from the United States. In terms of the size of the free population of color, the United States is clearly an outlier in the Americas.

In addition to the minute size of the free population of Afrodescendants, other markers of the exceptional racial regime in the United States were the formal bans on the exercise of trades, crafts, or professions, and limitations placed on the geographical mobility of people of African descent. Restrictions were much more extensive in the United States than in Latin America generally. In the United States, free people of color experienced extraordinary barriers to geographical mobility: many states demanded the departure of newly freed slaves from their territories (Virginia, 1805–1806; North Carolina, 1830; Mississippi, 1829; and Tennessee, 1831). Iowa, in 1859, even ordered the expulsion of all free people of color. Most of the South and some northern states forbade the immigration of Afrodescendants from other states (Virginia, 1793; South Carolina, 1800; Maryland, 1806; Delaware, 1807; Georgia, 1818; Mississippi, 1822; North Carolina, 1826; Tennessee, 1831; Oregon, 1858; Kentucky, 1852; and Indiana, 1852). Some states even banned the return of any free Blacks born in that state if they had left for any reason; for example, Georgia in 1835. Some southern states banned free Blacks from accessing public education or preaching a religion (Georgia, 1829; Virginia, 1830–1831; North Carolina, 1830, and Missouri, 1847). Throughout the nineteenth century, the professional mobility of free people of color was increasingly limited, with restrictions on the economic activities they could develop. In 1860, South Carolina went so far as to require free Blacks to wear badges engraved with their names, occupation, and a registration number. ³⁷

The size and centrality of the free people of color in Latin America stands in sharp contrast to the pattern prevalent in the United States. Puerto Rico was no exception to the general Latin American configuration. In 1830, 50.1 percent of the population was White, 39.3 percent were free people of color, and 10.6 percent were slaves. ³⁸ The existence of a large layer of free Afrodescendants was not merely a matter of numbers. The size of the free population of color also entailed occupational diversity and transfers of skills and capital across generations, making Afrodescendants central to the functioning of urban economies and providing a certain amount of social mobility. This was in sharp contrast to the U.S. experience, not only because local Afrodescendants who were free exercised many crafts in Puerto Rico, but also because free immigrants of color from the non-Hispanic Caribbean nourished the ranks of skilled workers in Puerto Rico throughout the nineteenth century. ³⁹

The restrictions on occupational and geographical mobility placed on Afrodescendants in the United States, and conversely, the lack of comparable restrictions in Latin America, must surely have affected the urban/rural distribution of populations. In 1910 in Puerto Rico, Blacks were more urban than Mulattos, and Mulattos in turn were more urban than Whites. In Louisiana, and in the United States as a whole in the same year, the reverse was true; Blacks were more rural than Mulattos, and Mulattos were more rural than Whites (see Table 3). The much larger urbanization rate among Blacks in Puerto Rico when compared to the United States must have been in part a product of the fact that on the island the free population of color was larger than the slave population throughout the entire nineteenth century and enjoyed geographical mobility without impediments to settlement in urban locations.

The acknowledgement of differences between slave and post-slave regimes by geographic region forms part of a new re-evaluation of the Tannenbaum thesis, in which historians are now recognizing that the differences in legal regimes of slavery between the United States and the Iberian-American colonies had long-term consequences because the institution of coartación in the latter, under which many slaves had the right to purchase their own freedom, led to manumissions which, in the long run, produced large populations of free people of color, in sharp contrast to the reduced free population of color in the United States. ⁴⁰

However, in some of the societies with large numbers of free people of color, Black/White inequality was very high (Brazil), while in others it was not (Puerto Rico), a conundrum that requires further research. Judging from our snapshot look at Puerto Rico in the early years of the U.S. colonial period, the island seems to be among the societies with large populations of free people of color, but even within this group, it appears to have relatively low racial inequality, which coincides with recent findings about Puerto Rico in the twenty-first century. ⁴¹ What accounts for this seeming anomaly?

The Rural Paradox

The income advantage of Black males in Puerto Rico in the first years of U.S. colonial rule has an historical explanation. It is, of course, counter to intuition in a society of the Americas conquered by Europeans who subjugated the native population and then proceeded to import enslaved Africans to work on plantations. An important dimension of the relative equality in incomes by race is that, unlike the United States, where extra-economic coercion was principally although not exclusively a racialized phenomenon concentrated on the exploited slave and sharecropping Black population, in Puerto Rico the landowning classes spent the better part of the nineteenth century trying to coerce the nominally free peasantry ⁴²—whether White, Mulatto, or Black—into forced labor.

The expansion of sugarcane plantations and slavery in Puerto Rico during the first half of the nineteenth century displaced the coastal peasantry upland and inland to the mountains of the Cordillera Central, where they sought to reproduce the independent subsistence lifestyle that they had enjoyed in the eighteenth century. ⁴³ In the second half of the nineteenth century, the subsistence farmers of the highlands were subjected to extra-economic coercion as the coffee export economy expanded. Beginning with the Reglamento de Jornaleros of 1849, the colonial state imposed on the free but title-less rural peasants labor obligations and a régimen de la libreta, or passbook system, that forced them to work for the larger landowners. This libreta system was abolished shortly after slavery, in 1873. ⁴⁴

To retain a rural labor force, landowners in nineteenth-century Puerto Rico developed schemes in which they granted land in usufruct, as well as some wages, to rural workers in exchange for work at harvest time. The relationship was known as agrego and the workers involved as agregados. Puerto Rico’s rural peasant-proletarians were stratified by race and skill in a complex continuum that defies easy characterization. ⁴⁵ In the century’s second half, coastal sugar plantations began to rely increasingly on the labor of agregados and on jornaleros, or day laborers coerced by the libreta system, to complement the enslaved labor force.

Governor Juan de La Pezuela’s introduction of a passbook system in his infamous Reglamento de Jornaleros of 1849 was designed to coerce subsistence farmers into performing servile work in sugar plantations and coffee haciendas, in favor of the landowning elites. Previous laws against vagrancy had tried to criminalize and coerce peasants into working for large landowners, ⁴⁶ but de la Pezuela’s Reglamento marked a qualitative leap in coercion. ⁴⁷ Coercion applied to those without property titles, including those considered White. Under the libreta regime, subsistence peasants without titles to land were forced to register as employees of landowners and planters. Failure to prove useful employment could result in imprisonment. After slavery was abolished in 1873, the formerly enslaved entered into an “apprenticeship” system that lasted three years. The passbook regime ended a few months after slavery was abolished. Other measures were initiated to try to maintain extra-economic coercion of labor, notably Governor José Laureano Sanz’s “Proyecto de Reglamento de Jornaleros (15 de abril, 1874),” ⁴⁸ but recent scholarship has questioned whether these were effective. ⁴⁹ Even during the epoch of slavery, other forms of coerced labor had played an important role in the production of agricultural commodities for export. The importance of non-slave coerced White labor during the regime of slavery was such that anthropologist Sidney Mintz, who studied slavery, the transition to free labor, and twentieth-century plantation life in various Caribbean societies, referred to Puerto Rico’s nineteenth-century sugar estates not as slave plantations but as “slave-and-agregado plantations.” ⁵⁰

In the last three decades of the nineteenth century, Puerto Rican peasants experienced a deterioration of the usufruct rights they had enjoyed as agregados on the farms of titled landowners, and they gradually became proletarianized as coffee exports exploded and reached a peak in the year 1896. ⁵¹ The emerging rural working class was thus composed of peasants of all races who were subjected to various forms of extra-economic coercion, and to these were added the libertos, freed slaves thrown into an emerging market for labor power after emancipation was proclaimed by the Spanish colonial state in 1873–1876. ⁵²

The existence of generalized labor coercion had two effects: a compression of inequality downward among all rural workers—Black, White, and Mulatto—and the emergence of a stratum of skilled Black workers who were well-positioned to benefit from the boom in the sugar industry in the twentieth century’s early decades. The second effect was a product of the preference of the planters for training slaves rather than libreta workers in the sugar industry under the slave regime because the latter, while forced to show to the authorities that they were employed by some landowner, were not tied to any particular employer. As Mintz explains,

Even more important for the evolution of Puerto Rican race relations than the absence of a bipartite “racial” division is the fact that the ex-slaves were generally more skilled at sugar-producing techniques of all kinds than were the descendants of freemen, whatever their racial background. The association of “race” and technical skill was thus an important aspect of Puerto Rican labor history. This association is doubtless rooted in the traditional distribution of tasks on slave-run haciendas before emancipation; slaves, rather than agregados, had usually been given the mill jobs and artisans’ responsibilities because there was no question as to the regularity or dependability of their labor. ⁵³

The preference for the employment of slaves in skilled plantation work carried over into the post-emancipation world. As a British consular official remarked shortly after the abolition of slavery, “In the process of sugar making, the more skilled liberto is generally employed within the boiling house while the free laborer does the rougher task of cutting and carrying the cane.” ⁵⁴ The issue of dependability was central, since in the first three years after abolition former slaves were bound to employers in an “apprenticeship” system: “The ‘liberto’ is appreciated not only on account of his superior ability for the work, but also because he is available at all times, and cannot leave his employment during a busy season.” ⁵⁵ The freedmen were “working side by side with the white native and the British black,” ⁵⁶ and furthermore, according to historian Andrés Ramos Mattei, the skills of some of these libertos placed them in a superior position in the hierarchy of the labor regime in many haciendas. ⁵⁷

The sugar industry and sugar-producing zones experienced a boom in the early twentieth century due to the inclusion of Puerto Rico within the tariff wall of the United States after 1902, while the highland coffee economy began a simultaneous process of contraction. Relative to Spanish colonial times, during which Puerto Rican coffee was protected in the Spanish and Cuban markets, Puerto Rican coffee enjoyed no tariff protection within the U.S. imperial system because there were no continental producers of coffee demanding tariff protection from Congress. Sugar was protected because the long-established U.S. sugarcane industry and a rising sugar beet industry after 1890 successfully demanded it. As the Puerto Rican coffee sector declined while the sugar sector expanded rapidly, an exodus of dispossessed peasants from the coffee highlands to the coastal zones ensued. The sugar boom of the first half of the nineteenth century displaced peasants from the coast to the highlands as sugar plantations manned by enslaved workers expanded; the sugar boom of the first half of the twentieth century led dispossessed peasants from the highlands to migrate from the ruined coffee region back to the coast. ⁵⁸

These “whiter” highlanders were at a disadvantage relative to the established coastal workers who possessed skills that were marketable in the mushrooming new sugar mills owned by expanding U.S corporations and local capitalists. As Mintz explains, “Because of the practice of training slaves as hacienda technicians while agregados were given the less specialized jobs, some of the Negro people of the coast were economically more secure, better educated, and more fully adapted to a wage-earning way of life than were the highland newcomers.” ⁵⁹ Ex-slaves were also well-represented among rural trades such as the paleros, whose skills in constructing and maintaining irrigation and drainage systems to match the slope of the land and the water table were absolutely necessary to keep the cane from rotting. ⁶⁰

Table 7 gives the results of a regression of occupational score on race and other demographic variables among rural males in the labor force. In the Puerto Rico sample, one can see that the coefficient is negative and significant for “Mulatto,” but insignificant for “Black” (Model 1). This shows that in the Puerto Rican countryside, the effect of race on income is indistinguishable between Blacks and Whites, but Mulattoes are at a disadvantage relative to Whites. Meanwhile, in both Louisiana and the United States, rural Blacks and Mulattoes are both disadvantaged, and Blacks in both Louisiana and the United States have slightly lower incomes than Mulattoes relative to Whites (Models 2 and 3).

Table 7. Fixed Effects Regression Models for Occupational Score, Rural Males Ages Sixteen to Sixty-Five in the Labor Force (1910–1920)

Robust standard errors in parentheses; *** p<0.001, ** p<0.01, * p<0.05, + p<0.1

When the coffee industry experienced difficulties in the early twentieth century and rural workers from the economically ruined coffee-producing highlands descended to the coast seeking employment in the sugar-producing municipalities, they entered a labor market in which many of the skilled jobs were in the hands of descendants of slaves. This led to the emergence of patterns of “race relations” that were different from what one might expect in a colonial society in which the Spanish had established racial hierarchies. According to Mintz, “the white coastal agregados who had worked alongside the Negro people of the coast before the American occupation, helped to cement social relations between white highland newcomers and the Negro people on the haciendas. No distinctions were made in the giving out of jobs in the field phase of the industry; as had always been the case, Black and white worked together, often with the Negro the teacher, the highland white the willing apprentice.” ⁶¹

The existence of a stratum of Black skilled workers affected patterns of intermarriage: “marriages between coastal people of differing appearance had been common since the eighteenth century, and the highland newcomers mixed freely. As one old ex-slave remarked, ‘In affairs of the heart, no one gives orders.’” ⁶² In the sugar zones, skilled men with year-round work in the mills were preferred potential marriage mates over cane cutters exposed to the five-month dead season of cane agriculture, during which there was no employment. Elsewhere, Mintz argues that in the rural, sugarcane-growing community he observed in 1948–1949 and in 1953, “marriages between persons of dramatically different phenotypes are common; physical type probably does enter into the choice of a mate, but its priority as a consideration seems to be very low.” ⁶³

In addition to their skills related to the sugar industry, Blacks were highly represented among other skilled trades. They carried their skills to the towns into which ex-slaves moved after emancipation, so that a good share of the skilled workers in urban regions were “colored” (Black or Mulatto) when the U.S. occupation began in 1898. According to the Census carried out by the U.S. War Department in 1899, whereas “colored” people represented 38 percent of the total population of the island, they were 64 percent of Blacksmiths, 52 percent of carpenters, 77 percent of masons, 44 percent of gold and silver workers, 37 percent of printers and lithographers, 58 percent of shoemakers, 42 percent of tinsmiths, and 58 percent of machinists, and were also overrepresented in several other skilled trades. ⁶⁴ U.S. Commissioner Henry K. Carroll reported in 1899 on meeting the artisans of San Juan as follows: “The artisans are better educated, have better food, and wear better clothes. As their work is chiefly in the cities, it is a necessity for them to be suitably dressed. At the invitation of the commissioner, the artisans of San Juan, who are organized into a dozen or more gremios or unions, came to his headquarters one evening and were examined. There were eleven of them, representing painters, tinsmiths, silversmiths, bookbinders, cigar makers, printers, masons, carpenters, bakers, shoemakers, and boatmen. Nine of the eleven were colored men, who seem to monopolize the trades, at least in the capital. All except one wrote his name and occupation in the stenographer’s notebook. They were neatly dressed, well-appearing, intelligent men. Each spoke of his own trade.” ⁶⁵

The historical formation of the rural working class of Puerto Rico, with roots in the coercion of the libreta system and slavery, differed dramatically from that of the United States, in which extra-economic coercion was limited to Black workers once the last vestiges of indentured servitude disappeared in the 1840s. Thus, one cannot make assumptions about racial stratification in the Puerto Rican countryside based on the U.S. experience, as if that particular combination of extra-economic coercion attached strictly to race held ground universally. In Puerto Rico, sugar plantations especially, but also coffee haciendas in the epoch of slavery, had been characterized by a mixed labor force of agregados, jornaleros coerced by the libreta system, free West Indian workers, and slaves, organized in a continuum but with a history of common struggle across race lines against the extra-economic coercion imposed by landowners and enforced by the colonial state’s infamous Guardia Civil (military police). ⁶⁶

The legacy of this common history of extra-economic coercion has been variously characterized. Recent scholarship argues, “The fact that forced labor in nineteenth-century Puerto Rico cast a net that transcended racial boundaries, as well as the salience of a divide between foreign landed and merchant elites and marginal creole landowners, regardless of race, are important factors to consider in analyzing the role of race in Puerto Rican society over the longue durée.” ⁶⁷ Mintz said as much in 1951: “The epic of forced, non-slave, white labour in Puerto Rico […] suggests that the history of labour can be fully interpreted only when slavery, forced labour, and all other means for relating labour to the instruments of production are seen in relation to one another in any historical period. In this light, studies of slavery and forced labour which emphasize racial differences or moral considerations may be made amenable to considerable reinterpretation.” ⁶⁸

In the twentieth century, the abrupt changes brought about by the transition from colonialism under Spain to colonialism under the United States reconfigured economic relations and created vast market opportunities for producers of sugar and tobacco, but not for producers of coffee. ⁶⁹ In this reconfiguration, the principal booming sector was the sugar industry and wages there were higher than in the coffee- and tobacco-producing regions. The Black population of Puerto Rico was concentrated precisely in the coastal districts that experienced a sugar boom in the first two decades of the twentieth century.

Puerto Rico: Percent of the Population that is Black, by Municipio (1910)

Puerto Rico: Percent of the Population that is Afrodescendant, by Municipio (1910)

The boom increased incomes for coastal rural workers relative to those of the tobacco and coffee districts. Local governments enjoyed an increase in tax receipts from the sugar industry, which helps to explain the large increase in educational investment in the early twentieth century biased toward the coastal sugar-producing regions, where the Black population was concentrated. ⁷⁰ Between 1898 and 1930, the White/non-White gap in literacy rates decreased from 11.63 percent for those educated under the Spanish regime to 3.27 percent for those educated under the U.S. regime after 1898. ⁷¹

Mintz’s ethnographic perceptions about race relations within the Puerto Rican working class can be checked against the quantitative data in the IPUMS database. The percentage of skilled workers seems to have been higher among Black rural workers than among White ones. In the cane-growing municipalities in 1910, 6.3 percent of rural Black workers were classified as “skilled and artisan” compared to 3.96 percent of Mulattos and 3.03 percent of Whites. Moreover, the average income of these Black workers (occupational score) was 112 percent that of skilled Whites, and the income ratio of skilled Mulattos to skilled Whites was 102 percent. Blacks were also proportionately more represented in other occupational classifications which indicate industrial work, with 6.54 percent of Black workers in the cane-growing municipalities compared to 5.02 percent of White workers being classified as “urban working class,” despite the fact that they worked in the countryside. This classification reflected employment in the sugar mills as opposed to rural employment in the fields as cane cutters. The higher concentration of Blacks in these relatively well-paid strata of the working class may explain why, in the countryside as a whole, Blacks had only slightly lower (1910, 98.71 percent) or even slightly higher (1920, 100.20 percent) occupational scores than Whites. ⁷²

Thus, the preference of planters for training slaves and libertos in the skilled jobs of the sugar industry in the last decades of the nineteenth century produced a stratum of skilled, specifically Black workers who transferred their skills intergenerationally and thus facilitated the emergence of a relatively well-paid, disproportionately Black segment of the working class in the sugar-producing municipalities in the early twentieth century. This particular configuration, so contrary to commonsense perceptions of the occupational status and income of Blacks relative to other groups, is in turn unintelligible without an examination of the specific combination of different forms of coerced labor in the nineteenth century.

Professional Inequality and Property Inequality

The socioeconomic profile of Black Puerto Ricans in the early twentieth century was not an entirely positive story. First of all, whereas Blacks were overrepresented in the skilled strata of the working class in the sugarcane municipalities, they were underrepresented among the category “farmer,” which indicates either property or management of a farm. In 1910, for example, 10.3 percent of rural White males were classified as farmers, whereas only 5.21 percent of Mulattos and 4.82 percent of Blacks were so classified. According to the agricultural census, the average size of a farm owned by Blacks or Mulattos was smaller than the average farm size of Whites. In 1910, Whites made up 66 percent of the population but owned or operated 76 percent of the farms. By contrast, 24 percent of the farms were owned or operated by “colored” (Black and Mulatto) farmers, who made up 34 percent of the population. The average farm operated by a White owner was 36 acres in size, whereas farms operated by “colored” owners averaged 15 acres. The average value of land and buildings of the former was $1,204, and of the latter, $315. ⁷³ It should be kept it mind nevertheless that the proportion of farmers of all races in Puerto Rico was rather small when compared to Louisiana, and that the dispossessed class of agricultural laborers or jornaleros encompassed the majority of the population, both Black and White.

To avoid underestimating the level of racial inequality in Puerto Rico, it should be emphasized that the relative equality of the occupational distribution was not characteristic of the property distribution. The existence of a very large segment of landless workers was the fundamental agrarian legacy of Spanish colonialism in Puerto Rico, ⁷⁴ where a process of dispossession in the last three decades of the nineteenth century in the predominantly White coffee highlands accelerated the process of proletarianization that had started earlier in the century in the coastal, sugar-growing municipalities. ⁷⁵ The phenomenal concentration of land in tracts of over 500 acres placed the overwhelming balance of agrarian resources in the lands of a small and powerful landowning oligarchy. This oligarchy was overwhelmingly White, although there were historically prominent Black families and individuals among the planters and in the upper echelons of the class structure. ⁷⁶ Nevertheless, this mostly White landowning elite dominated a multiracial dispossessed agrarian working class, and the percentage of dispossessed farm laborers among Whites was slightly higher than among Blacks, in sharp contrast to Louisiana, where the class of dispossessed rural Whites was much smaller.

Finally, the existence of a skilled Black working class stands in sharp contrast to the almost total exclusion of Blacks and Mulattos from the professions. According to the first U.S. census of Puerto Rico, in 1900, 93 percent of dentists, 95 percent of engineers, 88 percent of teachers, 95 percent of physicians and surgeons, and 88 percent of journalists were White. ⁷⁷ This exclusion was part of the legacy of Spanish colonialism, where entry into universities or the exercise of certain professions required White status and were zealously guarded, in a contested process in which the reputations of individuals and families and their “purity of blood” were litigated in the courts. ⁷⁸ To be sure, the flexibility of the racial system and the hybridity of the population meant that many of the so-called “White” professionals actually had African ancestry. The overall pattern of exclusion, however, extended into the early twentieth century and was noticed by U.S. administrators. Among teachers, the clergy, and in the Guardia Civil, not only were Blacks and Mulattos almost totally absent, but peninsular Spaniards seem to have monopolized these positions even at the expense of local creole Whites. ⁷⁹

The large and occupationally diversified population of color in the eighteenth and nineteenth centuries fulfilled crucial functions at almost all levels of the economy and the labor force was relatively diversified racially in the middle range of the occupational distribution, but the top echelons of the colonial labor hierarchy were mostly closed to Afrodescendants.

Discussion

It has become commonplace in academic discourse to talk of “structural,” “systemic,” or “institutional” racism without specifying the politico-economic processes that generate racial inequality, or the mechanisms which reproduce it. Often, these terms invoke a sort of black box whose internal gears are invisible but which, it is presumed, will inevitably generate and reproduce racial inequality. While much scholarship suffers from this bias, there are strands of scholarship that use these terms and specify what they mean, such as in research on how redlining in the United States had immense consequences for the intergenerational transfer of assets, exacerbating wealth inequality by race. ⁸⁰

The example of early twentieth-century Puerto Rico raises some interesting questions, given the assumptions by some researchers that the “black box” uniformly produces equally egregious Black disadvantage everywhere. ⁸¹ If the functioning of the gears within the black box are assumed to represent White supremacy, the latter understood as an invariant structure throughout the Americas, how can one account for variation in levels of inequality or the Puerto Rican anomaly in Black incomes in the early twentieth century? We agree with Stuart Hall that lack of attention to comparative differences in inequality “is often little more than a gestural stance which persuades us to the misleading view that, because racism is everywhere a deeply anti-human and anti-social practice, that therefore it is everywhere the same—either in its forms, its relations to other structures and processes, or its effects.” […] “in the analysis of particular historical forms of racism, we would do well to operate at a more concrete, historicized level of abstraction (i.e., not racism in general but racisms).” ⁸²

Closer empirical attention to the existence of different strata of the population, their specific location in the political economy, the coexistence of diverse systems of extra-economic coercion affecting different ethnic or racial groups, is thus required. In Puerto Rico, the combination of a higher rate of urbanization among Blacks with the existence of a skilled Black stratum of the agrarian working class in the sugar districts produced an interesting, and somewhat surprising and counterintuitive pattern: on average, Blacks had higher incomes than either Whites or Mulattos in Puerto Rico as a whole in 1910–1920. Within cities, Blacks had lower incomes than Whites, but because the percentage of Blacks that were urban was much higher than that of Whites, and because urban income scores were twice those of the countryside, the island-wide average for Blacks was higher than for Whites. Urban status, in turn, reflects the historical reality that most people of African descent in Puerto Rico were free throughout the nineteenth century, even as the island experienced a plantation boom in 1800–1850. Freedom was combined with geographical mobility to produce higher urbanization rates for Blacks. Geographical and occupational mobility matter.

Attempts to explain racial inequality that invoke the existence of White supremacy in all societies of the Americas are therefore insufficient and inadequate, for such a putatively uniform force cannot produce variable degrees of inequality, or at least cannot explain them. There are, instead, large variations in levels of racial inequality across countries. Yet, Latin America tends to be treated as a monolith in U.S.-centered discussions of racial inequality, which overlook immense differences in levels of Black/White racial inequality within the region. Not only is there immense variation in levels of inequality within Latin America, ⁸³ but some Latin American societies, notably Brazil, display levels of income inequality between Whites and Blacks that are even larger than those of the United States. ⁸⁴

If anything, the case of Puerto Rico shows that attention to the nuances of the history and political economy of race in the Americas may turn up unexpected surprises and may open the door for further comparative study of racial inequality. Paramount in such a comparative enterprise seems to be the coexistence of slavery with other regimes of labor coercion, including indentured servitude, which in the English Caribbean lasted until 1918, ⁸⁵ and other forms of coerced labor comparable to Puerto Rico’s Reglamento de Jornaleros. The size of the free Afrodescendent population relative to the slave population is also important, because societies with large numbers of free people of color provided mechanisms for wealth and skill acquisition that could be transferred across generations, something which was more difficult to achieve for enslaved persons in the plantations everywhere.

In Puerto Rico, the colonial state’s attempts to provide labor for planters in the coastal sugar districts and coffee hacendados in the highlands cast a net of labor coercion that spanned a population larger than the strictly Afrodescendant, nominally free peasantry, and subjected “White” highland peasants in the coffee-producing region to distinct forms of coerced labor. ⁸⁶ At the same time, most Afrodescendants were free in the nineteenth century, but many experienced forms of extra-economic coercion other than slavery. “Free” and “White” peasants were subjected to forms of extra-economic coercion also. The boundaries between freedom and unfreedom, and between coerced and free labor, were very different from those of the U.S. South. A much higher percentage of the Afrodescendant population was free in the epoch of slavery, and there were higher degrees of unfreedom among White peasants (forms of coerced labor and extra-economic coercion). This influenced the type of organizing that emerged to fight inequality. Under the U.S. regime after 1898, workers in Puerto Rico organized a formidable interracial union federation, the Free Federation of Workers. Class organization as workers was much stronger than organization by race. This pattern of organizing has lasted to this day, and it has had a large impact on Black organizing itself. As Eric Williams pointed out in 1945, racial differences have tended to be subordinate to those of class. ⁸⁷

Variability in degrees of inequality among societies in Latin America and between Latin America and the United States, in turn, should help to explain variability in the social movements that struggle against inequality. The example of Puerto Rico early in the twentieth century points to the importance of the legacy of coexistence of slavery with other forms of extra-economic coercion affecting broad segments of the population, not just the Black population. The specificities of colonial legacies, labor regimes, forms of extra-economic coercion, and systems of racial stratification, should open new vistas into the diversity of forms and means of fighting against inequality that have characterized societies different from the United States.