1On 17 April 1610, King Philip III gave the order to expel from Aragon and Catalonia the only Moriscos [1] remaining in Spain. [2] On 29 May, the proclamation was read in Zaragoza, then later in all other villages of the kingdom. The expulsion of most Aragonese Moriscos began on 3 July and was steadily pursued throughout July and August. The now empty villages presented economic opportunities for population growth. From that time onwards, the rulers of those villages—who were generally from noble families—began to slowly and painstakingly repopulate the communities with traditional Christians [3] (Colás, 2010). This exceptional historical circumstance is ripe for analysing the relationships between different demographic variables, especially that of population pressure. [4]

2Malthus’s ‘checks’, whereby population growth is kept in line with the food supply, may help in understanding this pressure. ‘Positive checks’ involuntarily reduce a population through epidemics, starvation, or war. ‘Preventive checks’ imply a degree of fertility control, perhaps by delaying age at marriage or maintaining a higher rate of singlehood. Following the expulsion of the Moriscos, a situation of low population pressure arose, which begs the following questions. Would Malthusian preventive checks have diminished in the localities concerned? Would neighbouring localities with the same socio-economic characteristics exhibit different demographic behaviours?

3Historically, population pressure has played a central role in demographic behaviour, especially regarding age at marriage, which indirectly influences fertility. Hajnal’s (1965) Western European marriage pattern, corresponding to the period from the 18th century to the first half of the 20th century, describes how high population pressure is controlled by maintaining an older average age at first marriage than in previous periods or in other regions of the world, and a larger percentage of people remaining unmarried. The same pattern was common in the Iberian Peninsula, although to a lesser extent than in other European countries (Cachinero, 1982). When the age at marriage increased, financial capacity and the share of unmarried people also increased, while fertility decreased (Wrigley and Schofield, 1981).

4This article examines this historical context to grasp the role played by population pressure in population growth. Using demographic microdata, we thus seek to understand why repopulated and non-repopulated Aragonese localities experienced different population growth rates after the expulsion of the Moriscos. [5] Our initial hypothesis is that the reproductive behaviour in villages differed due to changes in population pressure, and therefore fertility increased in localities experiencing low population pressure. This difference is assumed to result from a relaxation in Malthusian preventive checks. We expect that greater availability of land per person in localities with low population pressure will lead to higher proportions of children choosing to remain in their birthplace, marrying at younger ages, having more children of their own, and accumulating more wealth in adulthood than in areas with high population pressure.

5Focusing on the mechanisms of Malthusian preventive checks and how they were affected, our analysis draws on the abundant literature regarding the function and modifications of such checks in various pre-industrial societies, especially in England and France (Cummins, 2003; Weisdorf and Sharp, 2009; Cinnirella et al., 2017). While these societies placed greater importance on preventive checks for controlling population growth, this case study presents the opposite scenario: a society that had traditionally experienced high demographic pressure before becoming suddenly and unexpectedly subjected to low population pressure.

6Our research relies on the family reconstitution method and microdata to connect individuals and families in eight Aragonese municipalities. Five of these villages were inhabited by Moriscos and repopulated after their expulsion, while the other three had long been populated by traditional Christians. In addition, we employ a little-used source to approximate poverty levels relating to relaxed Malthusian preventive checks, namely collections of wills regarding the funerals of the rural popular classes. To find out whether low population pressure was associated with high average levels of wealth, descriptive statistics are used to compare the repopulated and non-repopulated localities.

7Section I summarizes similar studies that have examined the effects of various shocks on Malthusian preventive checks. Section II discusses the data and methods. In Section III, we analyse the economic context of the localities under study. In Section IV, we study the evolution of the populations and their growth during the 17th and 18th centuries. Section V is twofold. First, we analyse the effects of low demographic pressure on fertility and nuptiality. Then, contextualized within Malthusian theory, we explore other wealth-related variables that hindered population growth in the repopulated localities. Finally, we present our conclusions on the causes underlying the different population growth patterns in these repopulated and non-repopulated localities.

I – The adaptability of Malthusian preventive checks

8Populations have demonstrated remarkable adaptability, in response to new social and economic changes as well as to varying population pressures. Jedwab et al. (2019) analysed the evolution of European city populations following the Black Death pandemic of the 14th century and found that cities recovered their populations over the following 200 years, although the suitability of land and historical trade networks determined their recovery. Due to the lack of microdata, it is uncertain whether migration from rural areas or changes in demographic behaviour fuelled the urban repopulation.

9European colonial expansion is another example. With conquest came changes in demographic behaviour in contexts of low population pressure. Charbonneau’s well-known analyses (1975, 1984, 1999) of 16th-century French Canada found differences between individuals who moved there and those who remained in France. Female colonists were 3 years younger at first marriage than the women in France, while natural fertility was higher among the new Canadians, with an average of 6.9 children compared to 6.4. Settlers’ life expectancy and remarriage rates were also higher. These results are consistent with Malthus’s and Adam Smith’s examinations of North American British colonies, which demonstrate, overall, that different population pressures lead to different demographic behaviours.

10No large bibliography exists for studying the effects of repopulation on demographic behaviour following Spain’s expulsion of the Moriscos, especially none containing microdata. However, aggregate data are available that confirm different population growth rates during the 17th century. For example, varying population growth rates during repopulation were found in the Valencian region (from which Moriscos were also expelled) according to the type of locality. From these results, Ardit (1993) proposed distinct reproductive behaviour patterns among repopulated versus non-repopulated localities in the region, where Morisco villages were repopulated mainly by young couples who were relatively absent from the non-repopulated villages, which thus experienced an ageing population and decreased fertility. Other studies of Valencia by Bernat and Badenes (1994) confirmed the greater relative growth in repopulated localities with similar explanations for the differences, as did Ginés (1998), who used different population data to find that repopulated municipalities had a 0.42% annual growth rate between 1609 and 1652, versus −0.24% for the entire regional population of Valencia. Ginés explained these differences as follows: ‘[The repopulated areas] are recipients of a young population of reproductive age, and the latter [non-repopulated areas] are donor populations where the older people would remain, with consequent ageing and slowing of growth’ (p. 177). Based on differences in the age compositions of the non-repopulated and repopulated municipalities, this argument may have been valid for the first decades following the expulsion, but it is less so for later decades, when population growth would have depended not only on these villages’ rates of emigration and immigration but also on those of mortality, marriage, and births. Our analysis considers all these variables.

11While the research cited above indicates that population growth rates likely differed between repopulated and non-repopulated localities, at least during the first decades following the expulsion, questions nevertheless remain regarding explanations for these long-term differences. This article uses microdata covering over 2 centuries to delve into these unresolved questions. Our analysis is conducted in the framework of Malthusian population theories on pre-industrial Europe, which have been analysed in diverse contexts. In their book on the historical English population, Wrigley and Schofield state that marriage in that society tended to be postponed or avoided altogether (Wrigley and Schofield, 1981). In their parish-level analysis of the post-1540 period until the advent of the industrial revolution, Kelly and Ó Gráda (2012) found evidence for the effects of Malthusian preventive checks on marriages and births (Schofield, 1985), and while Nicolini (2007) found that preventive checks in England disappeared before 1740, not all authors agree with such an early termination. Cinnirella et al. (2017) showed that, in the period 1540–1850, families adjusted their fertility according to economic conditions and the stock of dependent children in their homes. Malthusian preventive checks therefore played an important role in preindustrial England. Given that rural Spain was similarly and typically preindustrial (Reher and Sanz-Gimeno, 2000), this article intends to identify the role of preventive checks when low population pressure emerged following the expulsion of the Moriscos.

II – Area and data

12The study area comprises eight rural Aragonese villages in north-eastern Spain. Five of these villages were almost entirely (over 95%) populated by Moriscos until 1610, namely Alfamén, Botorrita, Mezalocha, Mozota, and Muel. Three were populated by traditional Christians: Jaulín, Longares, and Tosos. All these villages are situated close together in the same central area of Aragon (Figure 1). For example, the Morisco village of Botorrita lies only about 7 km from the traditionally Christian village of Jaulín. As most Aragonese Moriscos resided around the Ebro River and its tributaries, the study area is a tributary called the Huerva River, which is not unlike other nearby Morisco areas. Altogether, the five repopulated localities in this study represented 6.2% of localities inhabited almost entirely by Aragonese Moriscos and 6.8% of the Morisco population, thus making them slightly more populated than the average village. [6] These localities were chosen for analysis because of their central geographical position, while organizationally they were no different from other areas. All these localities have ample parish archives, although problems with conservation occasionally arise. [7] Nevertheless, the availability of the archives in this valley favours it over others.

Figure 1

Study area

Study area

13Before the demographic transition, most of the population was characterized by natural fertility (married couples having children with little to no fertility control [Henry, 1961]). Natural fertility refers here exclusively to high fertility and no fertility control within marriage. The children in these villages were born from the time the mother got married until she reached menopause, resulting in many births and a high child-mortality rate that balanced family size. However, the numerous births and increasing surviving children led to population growth that continually increased the demographic pressure.

14Information on individuals in the study area was obtained from the parish archives, which stores records of baptisms, confirmations, marriages, and deaths. We connected and analysed them using the standardized family reconstitution method (Fleury and Henry, 1956) and manually endeavoured to minimize errors while linking the records. All birth certificates were classified by families according to the names and surnames of the parents. Only two families shared the same parents’ precise combination of names and surnames in the same locality and time, and were thus excluded from the analysis.

15The data were linked with the marriage, death, and confirmation registers to identify all family members and dates. Individual birth data was matched with that person’s other events in the parish registers. Thus, for each individual who was born in one of the study localities and remained there all their life, we could determine their dates of birth, marriage, and death, along with details on their parents and children. The information from Morisco villages records births, marriages, and deaths from 1610 or earlier—except for Botorrita and Alfamén, whose archives began, respectively, in 1640 and 1716. [8] The archives in the traditionally Christian villages begin in the second half of the 16th century. All this information is compiled in the Alfamén and Middle Huerva Database.

16Due to under-reporting of childhood mortality, the data for the 16th century and most of the 17th century present the same problems found in almost all Catholic archives. [9] Under-reporting also exists for certain years in the Jaulín and Muel archives, with the former missing many births, marriages, and deaths in 1662–1679, while the latter under-reported in 1661–1679 and 1693–1695. Thus, the analyses excluded potentially affected individuals. For example, Jaulín families that had children during the 1660s were excluded from the marital fertility analysis because it was impossible to guarantee complete birth information on the entire family, as some children may have been born during the under-reported period. While the main problem associated with under-reporting is reduced sample size, we have no reason to believe that demographic behaviours in localities with under-reported periods should differ either from nearby localities or from previous and subsequent periods (no selection bias). Despite these problems, the quality of the analysed records was generally high, as they were controlled and supervised by archbishopric representatives. All individuals born or married in the area were assigned at least one family identification number, but most adults were given two: one for the family they were born into and another for the family they married (those who married more than once were assigned more family identification numbers). Only single immigrants were not assigned to any family. The period 1610–1800 included 4,134 families, of which 3,144 had at least one child. The remaining cases concern people who married in the study area but left or died before having their first child. Of the 3,144 families with children, 972 were considered complete families because information was available about the wedding and the children, and both spouses lived at least 45 years.

III – Economic context of the studied localities

17The Aragonese villages populated by Moriscos and all other municipalities in this study area were traditionally manorial and accustomed to paying higher taxes than freeholds [10] (Colás, 1993). They earned their incomes primarily through raising sheep or other forms of agriculture, such as growing cereal crops, fruits, and vegetables on the banks of the Huerva River. The quality of the land was similar; however, lands without access to water—such as Alfamén, Jaulín, and parts of Muel—were largely devoted to grazing and, because they were considered relatively unattractive, they were thus repopulated slowly. Muel ceramics were recognized for their quality before the expulsion, while the formerly Morisco glass kiln in Alfamén was appropriated by the new settlers. Some inhabitants also worked in trade and transporting goods. The available 18th-century data includes the census of Aranda (1768) and the census of Floridablanca (1785–1787), by which we find an economic and social distribution that was typical of small agrarian localities and similar in both repopulated and non-repopulated localities. We assume that, in the first years after repopulation, the settlers had access to land ownership but faced relatively high taxes that in the previous period had been covered by greater numbers of people. Although the origins of some new settlers remain unknown, a portion of them clearly hailed from nearby locations within a radius of less than 100 km and usually less than 50 km. [11] The documentation (mainly wills, when available) does not indicate if they owned properties in their localities of origin. This scenario is congruent with the results found in other Aragonese and Valencian cases (Ciscar, 1993; Ainaga and Ainaga, 1996). Most settlers married, but their family sizes are unknown. They were all traditional Christians, and no evidence indicates that their demographic behaviour in their localities of origin differed from that in any other nearby traditionally Christian village.

IV – Evolution of the population

18No modern periodic census had been conducted on Spain’s population until well into the 19th century. Although neighbourhood lists and censuses for tax or military purposes were compiled, they tended to underestimate the true residential population (Salas, 2008) and are thus not useful for this study. We calculated the residential population based on the information available in the Alfamén and Middle Huerva Database, considering individuals to be village residents from the first to the last time their names appear in the parish records (including the birthdates of their children) and using 1 July of each year as a reference. This method means that a man born in January 1656 and deceased in January 1698 was counted as an inhabitant from 1656 to 1697 because he was not alive on 1 July 1698. If the first registered date of a woman born outside the area was her marriage in April 1690 and she died in November 1708, she was counted as an inhabitant from 1690 to 1708—even though she could have arrived earlier to work as a servant. This methodology tends to underestimate the residential population, as an individual would likely reside in a village before their first event was recorded and be unlikely to leave immediately after their last recorded event. Above all, an underestimation problem occurs for native individuals who migrated while single because they would only be counted as residents on their dates of birth and confirmation, despite residing in the village until their outmigration date.

19To determine whether this study’s estimates approximate reality, we compared our results with resident lists from the Status Animarum, which began in the study area in 1747. This parish family book includes all village residents aged roughly 7 and over. [12] Although it contains 3,925 individuals for the year 1750, our data analysis resulted in an estimate of 3,564 individuals, thus leading to a 9.32% underestimation of the true recorded population. For 1760, 1770, and 1780, the underestimations were, respectively, 2.47%, 1.82%, and 6.50%. Our estimates thus approximated the true population. For some repopulated localities with no data available before the parish registers began, we assume continuous population growth from zero inhabitants in 1610 to the first available value. For example, if the first available population data of a repopulated village is for the year 1650 and this locality had 80 inhabitants, we calculate a two-inhabitant-per-year increase in the village population from 1610 to 1650 (a span of 40 years).

20No resident lists exist for the repopulated localities before the expulsion, although the number of inhabitants can be approximated from some sources. Table 1 presents estimates of the numbers of Moriscos residing in the studied villages before the expulsion. These are based on the Marquis of Aytona’s census conducted before the expulsion to determine the Morisco population of Aragon. According to this study’s estimates, the table indicates the years when the repopulated villages returned to their pre-expulsion population levels and when they reached 25%, 50%, and 75% of the pre-1610 numbers of inhabitants. The study area clearly experienced slow repopulation, as Alfamén and Mozota did not reach their pre-expulsion population levels until the last quarter of the 19th century and Muel not until the beginning of the 20th. Alfamén faced particular challenges, requiring more than 125 years to reach 25% of the pre-expulsion level. In contrast, Botorrita recovered its population level in just over a century and Mezalocha in about 170 years. The localities that offered less land would repopulate the slowest, such as Alfamén and Muel, which rented much of their lands to shepherds.

Table 1

Pre-expulsion households and inhabitants in 1610 and population recovery over time by localities

Locality Number of households Inhabitants Year in which x% of the pre-expulsion population was recovered 25% 50% 75% 100% Alfamén 124 558 1737 1771 1802 1875 Botorrita 38 171 1656 1680 1722 1734 Mezalocha 98 441 1632 1673 1716 1783 Mozota 59 266 1625 1703 1733 1876 Muel 306 1,377 1681 1743 1764 1907

Pre-expulsion households and inhabitants in 1610 and population recovery over time by localities

Note: The population was calculated at a conservative ratio of 4.5 inhabitants per household.

21The Morisco expulsion reduced the Aragonese population by 15%–20%, and the pace of each village’s repopulation depended on the conditions imposed by the lords of the manors (Ainaga, 1989; Ainaga and Ainaga, 1996; Lloret, 2002; Colás, 2010). In most places, the conditions of the houses and lands were better for new arrivals than what the Moriscos had experienced, although each village’s natural wealth played a decisive role and was not always distributed equally (Lacarra, 1979). For example, families linked to the marquis’ administrators and local prominent men formed a local oligarchy in some Valencian localities by obtaining the best and largest properties (La Parra, 1990). Colás (2010) states that the prospect of migration was unattractive because new land owners were subjected to high taxes and tithes on most agrarian production and because they were prohibited from leaving. This context could have generated a selection of settlers by attracting only the poorest people. Indeed, the demand for land was much lower than the supply. In Muel, for example, approximately 306 families were expelled from the village, but only 16 came to repopulate it in 1611 (Burillo, 1997); and although the king granted Spanish nobility the right to force settlers to remain in the villages, the reality was different (La Parra, 1990), as no information is available for six of the 16 Muel settlers just 1 year after repopulation. Fifty years later, 11 of the 16 surnames had disappeared from this village, indicating that settlers left if they found better elsewhere.

22Figure 2 illustrates the population growth in repopulated and non-repopulated localities from 1611–1800. The average annual growth rate in the repopulated areas was 11.3% for 1611–1649, 1.0% for 1650–1699, 1.1% for 1700–1749, and 0.6% for 1750–1799, while the respective rates in non-repopulated localities were 0.9%, 0.7%, 0.7%, and 0.4%. Therefore, the repopulated localities grew slightly faster throughout the whole period and saw intense growth only in the first years (1611–1625). Some authors indicate a population plateau in Aragon throughout the 17th century (e.g. Salas, 1989), although we found no such plateau in this study area.

Figure 2

Total population and average annual rate of population growth from 1611 to 1799, by type of village

Total population and average annual rate of population growth from 1611 to 1799, by type of village

V – Malthusian preventive checks: demographic and economic perspectives

23This section discusses the microdata to compare the repopulated and non-populated areas’ demographic behaviour and economic situations. Our analysis focuses on native individuals for whom dates of birth are available. [13] New settlers (‘immigrants’) are introduced for comparison in only the last part of this section (on the economic variables). First, we excluded immigrants from the analysis due to the scarcity of their pre-arrival information. In addition, the native individuals were few in the first decades of repopulation, when the villages were smaller, which limits the relevance of a comparison. We thus focused our analysis on 1650 onwards, not because of different patterns in the previous period, but because the scarcity of data for some variables could severely affect the results. Complementary immigrant data (such as registers with age at marriage for some decades) ensured that no important differences existed between natives and immigrants. Therefore, we do not believe our selection criteria bias the sample. All restrictions in this analysis are adapted to the period and data in accordance with the standardized criteria used in historical demography, and they were selected to compare similar populations. For example, our analysis of marital fertility exclusively included complete families and married individuals reaching 45 years old. Therefore, knowing the ages at marriage for this period required knowing the birthdates and wedding dates. This implies that we can only know the age at marriage of native individuals who, as we will discuss later (in relation to Figure 7), were between 50% and 95% of the adult population in these localities in the period 1650–1800. However, we have no evidence to suggest that this criterion biases the results. These data restrictions resulted in the available sample size changing in each analysis. In some cases, the size of the sample might affect the results but not the long-term trends.

1 – The roles of singlehood and age at marriage

24At this point in the analysis, our objective is to examine whether the reduced population pressure in repopulated localities could have relaxed the Malthusian preventive checks. Figure 3 confirms that increasing population pressure favoured increased age at marriage, which is consistent with Western Europe. The mean age at first marriage of women and men in the repopulated villages was 3 years younger than in the traditionally Christian villages throughout the 17th century. Both locality types tended to converge over the long term due to a gradual age increase in repopulated villages and an age decrease in non-repopulated villages. Valero (1984) placed the average female age at marriage between 22.5 and 23.5 years for nearby Zaragoza town residents during the 17th and 18th centuries; for men it was about 25.5 years. These data are similar to those obtained for the non-repopulated villages. In short, the repopulated localities presented an exceptional situation.

Figure 3

Mean age at first marriage for native women and men by year of birth and type of village

Mean age at first marriage for native women and men by year of birth and type of village

Note: Eleven-year moving average. Marriages at over age 40 are not included.

25In a natural fertility context, the Malthusian theory posits that a lower age among women at first marriage leads to greater fertility. Figure 4 shows the marital fertility in first marriages between native spouses who both lived to be at least 45 years old. Marital fertility was generally greater in the repopulated localities, which partly explains the high population growth. As a consequence of converging age at marriage, there is once again a tendency towards convergence (with the small sample size causing variability), which becomes complete by the end of the 18th century. Both types of localities display similar behaviours in terms of age at the last child’s birth and in birth intervals (Appendix Figure A.1 and Table A.1). Therefore, stopping and spacing strategies for fertility control (e.g. Van Bavel, 2004) do not decisively explain these fertility differences. In the Valencian case, Casey (1979) did not detect higher fertility among married people in the repopulated villages, although his sample size of only 12 marriages was small.

26Another preventive check identified by Malthusian theories is the elevated celibacy rate associated with societies experiencing high demographic pressure, a hypothesis confirmed by the results for this study area (Figure 5). The repopulated villages display a low rate of celibacy at age 45 (consistently under 5% of the population) compared to non-repopulated villages but with a slight tendency towards convergence during the last part of the 18th century. Livi-Bacci (1968) calculated that, for Aragon in 1787, 9.9% of men and 4.9% of women remained unmarried until they were 40 to 50 years old, data that are close to those of the non-repopulated localities. While outmigration could affect these results, we have no reason (nor proof) to think that a higher proportion of single people moved from repopulated than from non-repopulated villages.

Figure 4

Mean marital fertility by wife’s year of birth and type of village

Mean marital fertility by wife’s year of birth and type of village

Note: Fertility calculated for families in which both spouses survived beyond age 45.

Figure 5

Percentage of unmarried men and women by birth year and type of village

Percentage of unmarried men and women by birth year and type of village

27The Morisco expulsion arguably entailed the relaxation of Malthusian preventive checks. The repopulated villages grew faster than those that were not. However, why did the repopulated localities fail to grow faster in accordance with the strong relaxation of checks? Additional factors likely affected these areas’ population growth. Because the Malthusian theory predicts that preventive checks are relaxed as a consequence of an improved economic situation in areas with low demographic pressure, we next analyse the economic differences between locations and how they affected population growth.

2 – The role of economic constraints

28We first analysed the data on people who were born and remained in a single locality, a practice known as ‘sedentism’. We assumed that natives would have had fewer incentives to migrate from repopulated villages that offered more and better economic opportunities. Figure 6 shows the percentage of individuals for whom evidence indicates they remained in their village of birth beyond age 20 and died there. The results show that individuals in non-repopulated villages were initially more likely to remain in their birthplaces at even double the proportion of individuals in the repopulated villages until the second part of the 18th century, when they attained similar levels. The repopulated villages could not retain most of their young inhabitants.

Figure 6

Percentage of women and men who died in their birth village by year of birth and type of village

Percentage of women and men who died in their birth village by year of birth and type of village

Note: Eleven-year moving average.

29The causes of these differences are possibly linked to a smaller marriage market or greater propensity for mobility in repopulated villages or to the economic context. In earlier research, we found that, at least during the 18th century, the poorest individuals had the weakest incentives to remain in their places of origin (Marco-Gracia, 2017). To test the hypothesis of the link between a worse economic situation and higher mobility, we analyse the poverty levels through an indirect source: wills or, more precisely, what we call here ‘funeral wills’.

30These wills were intended strictly for funeral expenses. Few wills in this area were notarized, and most were dictated to the parish priest when those who did not die suddenly received the extreme unction. Such wills did not indicate how to distribute the person’s wealth but only what to spend at their funeral and how, which the priest was required to record in death certificates reviewed by his superiors. Not everyone drew up a will, although the priest asked all dying people of sound mind to do so, as he crucially depended on the funeral bequest for a part of his earnings. Over 90% of adults made a funeral will, and if the priest found that the dying person had undeclared resources yet bequeathed insufficient funds for the funeral, he would try to increase the declared amount. Among those who left a will, we consider only people older than 20 who resided in the study area—in other words, those who, according to the death registers, did not die suddenly or suffer a mental illness. Our analysis further compared the numbers of people who were too poor to pay for their funeral expenses. We also compare the percentages of individuals who drew up a funeral will and of those who could not afford a funeral to determine whether the proportions of the latter differed by village type. Such information about extreme poverty levels increases our knowledge of the socio-economic situations in both types of localities, which allows us to look beyond the outmigration rate when analysing whether the repopulated localities offered more and better economic opportunities.

31Figure 7 compares these individuals based on sex, place of origin (natives or immigrants), and year of death. To analyse only individuals living in the study area (and not beggars or passers-by), the analysis included only immigrants with at least two children (born on different dates) to ensure that these individuals were settled in the area and had sufficient presence to participate in the local economy. The results reveal that residents in repopulated areas (natives and immigrants) were more likely to die in poverty. The non-repopulated localities were economically more favourable throughout the 17th century for both natives and immigrants of both sexes.

32Village wealth can also be explored using funeral costs. Beyond the basic expense of a funeral, many extras like additional priests or more masses enhanced the standing of the deceased and added to the final cost. On average for most of the period, people spent slightly more on their funerals in non-repopulated than repopulated localities (Appendix Table A.2), suggesting that the economic situation of the former was better than the latter. The lower rate of sedentism in the repopulated villages could be due to less favourable economic circumstances, despite relaxed Malthusian preventive checks.

3 – Village attractiveness

33Analysing the number of immigrants in the villages also helps to determine how attractive the localities were in terms of their potential for economic, marital, and social opportunities. In this way, we can ascertain, on the one hand, whether more settlers moved to villages with low population pressure than to non-repopulated villages and, on the other, if the trend changed during the 2 centuries under analysis. Figure 8 shows individuals who died in a village in which they were not born, giving an approximation of immigration. It reveals marked differences during the first decades, with a high percentage of immigrants in repopulated localities due to a lack of natives at the time and to incentives to settle there. From the middle of the 17th century, the percentages were still slightly higher in repopulated localities. However, given that the repopulated villages had fewer inhabitants, more immigrants arrived in non-repopulated localities. At the same time, the importance of immigrants to fertility rates differed according to locality and period. In repopulated localities, the percentages of children born to immigrant fathers were: 50.4% in 1650–1699, 34.3% in 1700–1749, and 25.7% in 1750–1799. For non-repopulated localities, these percentages were, respectively, 29.2%, 28.5%, and 29.0%. Similar results were obtained for mothers. These results confirm that, until the middle of the 18th century, the growth of the repopulated villages was highly influenced by those who arrived in the area. From the middle of the century onwards, both types of villages show similar values. The repopulated villages seem to have been more attractive in the first decades after the expulsion of the Moriscos, losing their potential to attract as their population size increased.

Figure 7

Percentage of native and immigrant individuals too poor to make a will, by gender, year of death, and type of village

Percentage of native and immigrant individuals too poor to make a will, by gender, year of death, and type of village

Note: Only individuals over 20 years were included in the analysis. Immigrants with fewer than two children were excluded.

Figure 8

Percentage of men and women who died where they were not born, by year of death and type of village

Percentage of men and women who died where they were not born, by year of death and type of village

Conclusion

34The expulsion of the Aragonese Moriscos created an exceptional situation of low population pressure in empty localities. Our aim was to determine why the rate of population growth consistently remained higher in the repopulated municipalities than in the non-repopulated ones. We investigated whether Malthus’ theories on fertility and preventive checks applied in practice to this extraordinary context.

35On the one hand, low population pressure resulting from meagre repopulation may have favoured a decrease in age at marriage, thus leading to an increase in marital fertility. No evidence indicates the practices of spacing or stopping births as fertility control strategies. Furthermore, there was a low rate of celibacy. These circumstances likely resulted from a relaxation of Malthusian preventive checks. Thus, these variables favoured population growth in the repopulated villages.

36On the other hand, our analysis of this area shows no apparent relationship between a locality’s low demographic pressure and a good economic situation. The new natives born into the repopulated villages had less incentive to remain there due to the limited opportunities, perhaps because of the ‘extractive’ institutional framework (Acemoglu and Robinson, 2012). [14] To determine whether an economic reason can explain this behaviour, we used information from wills to discover if each inhabitant left sufficient funds to pay for their funeral. We found that, throughout the 17th century, a greater percentage of individuals residing in non-repopulated localities could pay for their funerals than those in repopulated villages. Our analysis also shows that average funeral costs were higher in non-repopulated localities. Thus, the repopulated localities could have been poorer, with too few opportunities for young people to remain.

37Extractive institutions may also have played an important role in the repopulation process by imposing high local taxes that could have discouraged settlement in the empty localities. Poor people probably embraced relocating to these villages to improve their initial situation, even if they were still worse off than their traditionally Christian neighbours. In contrast, economic circumstances may have favoured the youth outmigration from the repopulated villages, thereby conditioning population growth and reducing it to levels close to those in non-repopulated localities. Thus, the fewer economic opportunities could have relaxed population growth. Whatever the case, all the variables included in this study—both those that favoured population growth and those that did not—tended to converge over the following 2 centuries.

38Overall, these results support Malthus’ hypothesis on the negative relationship between low population pressure and preventive checks. Our initial hypothesis has been partly confirmed. The results show that young people in the repopulated villages married at a younger age and had higher marital fertility. However, the economic opportunities offered by these villages, where population pressure was low, were not sufficient to attract newcomers to settle. These localities did not retain the young population born in the area; neither did the residents accumulate more wealth than those in non-repopulated localities. The scenario after the expulsion of the Moriscos was not as beneficial for the repopulators as could be expected from the theoretical framework. Our conclusions must be corroborated by further studies of other regions while considering complementary variables such as socio-economic data (for example, using occupations as a socio-economic indicator). One could then compare the socio-economic structure of different types of localities and the demographic behaviour at the individual level according to family socio-economic status. For example, could the repopulated villages have presented a differentiated socio-economic structure from their neighbouring villages? Could the socioeconomic structure of a village have conditioned its population growth? Could the low age at marriage (and, therefore, the high marital fertility) be the consequence of a higher proportion of individuals from high socio-economic groups? Pursuing these questions will aid in understanding what happened in the long term in localities with differing population pressure.