The decline of infant and child mortality among Spanish Gitanos or Calé (1871−2005): A microdemographic study in Andalusia
Juan F. Gamella1
Elisa Martín Carrasco-Muñoz2
Abstract
BACKGROUND
Most Romani groups in Europe have experienced a decline in childhood mortality
during recent decades. These crucial transformations are rarely addressed in research or
public policy.
OBJECTIVE
This paper analyzes the timing and structure of the decline of childhood mortality
among the Gitano people of Spain.
METHODS
The paper is based on the family and genealogy reconstitution of the Gitano population
of 22 contiguous localities in Southern Spain. Registry data from over 19,100 people
and 3,501 reconstituted families was included in a dense genealogical grid ranging over
150 years. From this database we produced annual time series of infant and child
mortality and of the registered causes of death from 1871 to 2005.
RESULTS
The analyzed data shows a rapid decline in infant and child mortality from about 1949
to 1970. The onset of the definitive decline occurred in the late 1940s and early 1950s.
Child mortality was higher in the pre-transitional period and started to decline earlier,
although it took longer to converge with majority rates. The mortality transition in the
Gitano minority paralleled that of the dominant majority, but with important delays and
higher mortality rates. The causes of death show the deprivation suffered by Gitano
people.
CONCLUSIONS
The childhood mortality decline facilitated the most important changes experienced
recently by the Gitano minority, including its fertility transition and the transformation
of Gitanos’ gender and family systems.
CONTRIBUTION
This is one of the first historical reconstructions of the mortality transition of a Romani
population.
1. Introduction
The Gitanos or Calé of Spain are an ethnic minority related to the other Romani groups
in Europe and America. Notwithstanding their common remote origin, Romani groups
have adapted to the surrounding societies in all the regions where they have lived and
present considerable sociocultural heterogeneity (Matras 2015; Piasere 2004; Fraser
1992). The Gitanos seem to descend from the first migratory waves of Romani groups
into Western Europe, which were documented in the 15th century (Pym 2007; Leblon
1985). Their customs and life patterns are the product of a long coexistence with local
Spanish populations, often marked by persecution, forced assimilation, and
discrimination, but also by cooperation, hybridization, and by their creative
appropriation of majority customs (Gómez Alfaro 1993, 1999; Leblon 2003; Gamella,
Gómez Alfaro, and Pérez 2014).
In the democratic and decentralized regime developed after the end of Franco’s
dictatorship, Gitanos have gained access to free and universal health care, public
education, pensions, and housing benefits. This has induced a remarkable process of
social integration and cultural convergence with their Payo (non-Gitano) neighbors.
Nevertheless, most Gitano groups have preserved a vibrant sense of themselves as a
distinct cultural group, and have developed new forms of reaffirmation, resistance, and
mobilization in cultural, religious, and political realms (San Román 1997; Gay Blasco
1999; Cantón et al. 2004; Mirga 2014; Gamella, Fernández, and Adiego 2015).
Arguably, the most far-reaching transformation experienced by the Gitano people
of Spain in the 20th century was the dramatic fall in their infant and child mortality
patterns. Almost all newborn Gitano children survive today, whereas in 1950 about 200
per 1,000 died before their fifth birthday.3
The change is even more pronounced compared to previous decades: in the mid-1920s the risk of a Gitano child dying in
childhood was about sixty times greater than in the early years of the 21st century (see
Tables 1 and 2 below, and Gamella, Martín, and Quesada 20144
). The improvements in
child survival induced an unprecedented population growth that multiplied the size of
the Calé population. In turn, this demographic expansion led to an intense Gitano
migration from Andalusia and Extremadura to other more prosperous and industrialized
regions of Spain and, later, to other European countries. These movements and
resettlements altered the geographic distribution of Gitanos and therefore affected the
identity of the new generations of Calé. Moreover, the drop in childhood mortality
facilitated a decline in fertility that became generalized from the late 1980s. The
intentional control and reduction of fertility is completing the distinct demographic
transition of this minority. This process has enormous consequences for the entire
Gitano community, including consequences that are rarely considered in reviews of the
demographic transition (Lee and Reher 2011): convergence with the majority
population and an increase in intermarriage.
Ultimately, as the burden of reproduction falls disproportionately on women,
widespread child survival has had crucial long-term effects on the lives of Gitano
women and on the gender arrangements within the group.
1.1 The death of Gitano children
For centuries the recurrent death of children was a common experience in the homes of
Gitano people. This tragic reality often emerges both in the discourse of Calé women
themselves (Gamella 2000, 2011) and in any cursory view of the corresponding civil
and parish records. Infant and child mortality was also common among the majority
population in Spain, but the higher fertility rates of Gitano women resulted in more
early deaths per mother. In our records there are many instances of Gitano women that
suffered the loss of a very high number of children.
For instance, Salvadora B. was a Gitano woman born in the city of Guadix in
1852. In our review of the civil registries of Guadix and neighboring towns we found
records of 11 children born to Salvadora in the 21-year period between 1871 and 1892.
Eight of these children died before their third birthday. Surprisingly, when Salvadora’s
husband, Juan, died at 48, his death certificate stated that he “was single, although for many years had lived as a married couple with Salvadora B. … [the] union of which
resulted in three surviving children”.5
In the winter of 1924 Salvadora died from
exposure. She was 72 and lived in a cave in miserable conditions. The civil and parish
records contain traces of the dramatic life of this woman, whose common law marriage
was not officially recognized and who was listed as castellana nueva (the official
euphemism for Gitanos) in several entries. Therefore, the registered side of Salvadora’s
biography reflects the bureaucratic ideology that stressed the separate ethnic identity of
Gitanos while at the same time ignoring their own cultural definitions of marriage and
relatedness. In the following decades Salvadora’s three surviving children registered 28
children themselves, of whom 15 died in childhood or adolescence. Hence, of
Salvadora’s 39 children and grandchildren, only 41% lived long enough to marry and
reproduce. Nevertheless, in our genealogical reconstitution we found 576 direct
descendants of Salvadora, most of them alive today. Hence, the lives of Salvadora and
her descendants show how the loss of many children was a common experience for
Gitano women until fairly recently, but also how successful their reproductive strategies
have been despite this loss.
Even if it goes unnoticed by the world at large, the death of a child usually deals a
terrible emotional and physical blow to the mourning parents. Today many Gitano
women still cry when they remember the death of a child that occurred decades ago,
and they still clearly remember the circumstances, symptoms, and events as they
unfolded to their tragic end. Many women can also tell the stories of the deaths of their
infant siblings based on their own or their relatives’ recollections. These narratives offer
important sources of either confirmation or rebuttal of the archival data we gathered.
Given the omnipresence of this crucial issue in the consciousness of Gitano women, it
is surprising it has not received more scholarly attention and scrutiny.
1.2 Objectives
In this paper we will assess the structural dynamics of the decline in infant (under 1
year of age) and child (under 5 years of age) mortality of the Gitano people of 22
contiguous communities in the province of Granada in Southern Spain. We will use
mostly records kept by the Civil Registry since its inception in 1871, but our analysis
will concentrate on the period beginning in 1920 when the available data becomes more
reliable.
This paper will provide a model of the infant and early childhood mortality
transition in the Gitano community, including the timing of the onset of definitive decline, the intensity of change, the main phases of the process, and the relative growth
in neonatal deaths. We will also compare our results to data reflecting both the entire
province and the Spanish population at large.
Our model also includes data on the causes of child deaths as they were recorded
in the researched archives, and tracks the changes in these diagnoses over time. The
causes of death offer key insights into the underlying social determinants that affected
the survival of children from the most underprivileged families.
This study aims to situate the mortality transition of Spanish Gitanos in its social,
political, and epidemiological context, and hence facilitate the analysis of the main
“factors responsible for this ‘secular’ and seemingly irreversible decline” (Corsini and
Viazzo 1997: xiii).
1.3 A gap in Romani studies
This study contributes to filling a gap in Romani Studies. The dominant representations
of Romani groups have ignored the evolution of childhood mortality, as well as other
key demographic transformations. This is especially remarkable considering that
demographic differences are among the most salient aspects of the ethno-cultural
contrast between Romani populations and mainstream majorities everywhere.
Particularly, in the best ethnographic monographs available, little or no attention is
paid to the issue of the death of children, even when the experience, celebration, and
commemoration of death is a crucial topic in these works (Sutherland 1975; Okely
1983; Williams 1993; Stewart 1997; Engebrigtsen 2007).
Nevertheless, the decline in infant and child mortality also seems to have occurred
in other countries of Central and Eastern Europe, such as Romania, Hungary, Slovakia,
and Bulgaria, which have large and varied Romani populations (see Ladányi and
Szlényi 2006; Scheffel 2005; Kohler and Preston 2011; Burlea 2012). However, we do
not know much about it, or how these processes have diverged from those occurring
among neighboring majority populations.
References to infant and child mortality are scattered throughout publications
concerning health status and access to health care, or surveys on living conditions and
official demographic data (Cook et al. 2013; Ringold, Orentstein, and Wilkens 2005;
Kalibová 1993, 2000; Costarelli 1993). Most of the studies are cross-sectional. For
instance, some studies compare mortality rates across a particular country and
recurrently find much higher rates of infant mortality in the regions where Roma
populations are concentrated (Rychtaříková and Dzúrová 1992: 630) and in urban
neighborhoods with a high proportion of Roma residents (Rosicova et al. 2011:
526−528). Recently, Kohler and Preston (2011) made an important analysis of differential mortality patterns among religious and ethnic groups in Bulgaria, using
nominal data in 1990s’ censuses. However, their analyses “were restricted to the
noninstitutionalized adult population aged 20 and over” and thus they “avoided
potential problems in measuring mortality of children in the census-based data set”
(Kohler and Preston 2011: 93).
In the rapidly increasing literature on Romani groups, very few publications are
devoted to childhood mortality, much less from a historical perspective. Childhood
mortality is rarely treated as a key variable in other epidemiological, social, and
economic transformations. Even in important papers on the anthropological
demography of Romani groups there are no references to infant mortality (Durst 2010,
2002). One important exception can be found in the work of Ladányi and Szlényi, who
studied the transformation of a village in northeastern Hungary from a multiethnic
peasant village into a segregated Roma ghetto, and were able to collect demographic
data of considerable quality and detail for 1857 to 2000. They also were able to follow
the long-term evolution of infant mortality and its major shifts, concluding “the
dramatic decline in infant mortality between 1951−1988 was one of the most
spectacular achievements of socialist policy. The success of these policies places the
blame for high rates of Gypsy infant mortality squarely on the shoulders of pre-war
public health authorities” (Ladányi and Szlényi 2006: 67). This study proves that local
registry data often includes Romani people and can provide extraordinary results if
studied patiently and in an integrated form. However, historical studies such as this are
rare.
In sum, the literature on the demographic history of Romani peoples is very
limited in scale and content and demographic concepts and models have been ignored
by most historical or cultural studies of these groups. Their potential, however, is
obvious, both in terms of theory construction and in the analysis and design of public
policy.
2. Methods and data sources
The study of the history of infant and child mortality presents severe technical
difficulties and data problems even for larger and better-known populations (Corsini
and Viazzo 1997; Schofield, Reher, and Bideau 1991). Regarding Spanish Gitanos,
some historical developments compound the task. First, a Royal Order in 1783
prohibited references to the ethnic identity of Gitano people in Spanish public records.
This interdiction was followed, albeit irregularly,6
and for the last two centuries there is
no aggregated official data that can be used for the demographic study of this minority.
ggregated official data that can be used for the demographic study of this minority.
Secondly, the relatively small Gitano population has been dispersed in numerous
localities in all regions where they have lived. For instance, in 1785 the last available
census of the whole Gitano population gathered data on about 12,500 persons living in
over 650 localities in almost all Spanish regions (Gamella, Gómez Alfaro, and Pérez
2014). Today the roughly half a million Gitanos live in more than 1,000 villages, towns,
and cities all over Spain (Fundación Secretariado Gitano 2008).
Thirdly, most experts have assumed that Gitano families did not register the births
or deaths of their dear ones until very recently (see, for instance, San Román 1997;
Ramírez Heredia 2005). This assumption was in line with the popular misrepresentation
of Romani groups as essential nomads; that is, people who maintained weak and
uncertain links with their places of birth and residence. Contrary to this assumption, we
have found that Gitano families have been registering their births, deaths, and official
marriages in the parish registers of Andalusia since the beginning of the 18th century,
and often before.7
This practice was reinforced by the establishment of the Civil
Registry in 1871, and became commonplace in the 20th century. Some underregistration, however, was common in the first decades of the functioning of the Civil
Registry. Under-registration of infant deaths among the Gitano people may have been
important at least until the second decade of the 20th century. On the other hand,
common law marriages by ‘Gitano law’ were obviously not registered until they were
sanctioned by Catholic and civil authorities, a process that increased during the 20th
Century (Martín and Gamella 2005).
In a previous paper we have described the methodological and technical strategies
followed in our genealogical and familial reconstruction. Readers are referred to that
paper (Gamella, Martín, and Quesada 2014). Below we summarize its main points.
2.1 Gitano identity and identification processes
....
