The Broader Picture
The Broader Picture
Global Determinants of Female Agency and Economic Growth
Abstract and Keywords
This chapter introduces the concept of female agency and sets up the framework for the rest of the book. The authors explain how age at marriage can be an indicator of the degree of power a woman has in a relationship, as well as her role in society. This chapter argues that the European Marriage Pattern (free choice of marriage partner) played a fundamental role in the economic development of Western Europe, leading to the Industrial Revolution and a higher standard of living. The authors explore the contemporary situation, presenting correlations between marriage age and composite measures of gender equality, as well as zooming out to the global level to discuss differences in how women have fared in terms of human capital formation, access to the capital market, and participation in the labor market.
If we are going to see real development in the world, then our best investment is WOMEN
Describing Female Agency
Economic development is often considered to be a male affair, engineered by male entrepreneurs, male politicians, and male managers. Economic historians trying to analyze long-term economic change have generally also concentrated on men as the main carriers of the processes they try to understand. In this book we make the point that women matter, and in our specific case, that women matter for understanding economic change in preindustrial Europe, as well as for economic development generally, both in the past and today. One of the unfortunate consequences of the increasing specialization among academics is that the study of women’s history has become a separate specialism, part of gender studies, and therefore has become isolated, to some extent, from the mainstream of economic history. Much of the work of gender historians is focused on calculating women’s “invisible” contribution to the economy, via their activities in the household economy. Regardless of the necessity of this particular quantitative exercise, it does not help us unravel the mechanisms that contribute to or discourage female empowerment more broadly construed. Moreover, the processes we describe in this book go back so far in time that putting figures on them is hard enough, let alone identifying the degree to which women have contributed to those developments. As a result, economic historians have not devoted much time to integrating new ideas and approaches that concentrate on the role of (p.2) women in their analyses of long-term change.1 We think, however, that based on insights from development economics and Unified Growth Theory, there are reasons to pay much more attention to the role played by women as engines of economic change.
Women represent roughly half the world’s population and are therefore a significant economic resource; however, throughout history and around the world, the fairer sex has been discriminated against and frequently left without a voice, on both political and domestic fronts.2 This goes hand in hand with the fact that male dominance has been a near universal phenomenon, although the degree to which it is practiced has varied.3 Intrinsically, this is unfair from a social justice, human rights perspective, and improving the lot of such a large share of the population should be a goal in its own right. However, equality between men and women has more wide-ranging effects than those experienced by women alone. This implies that there is also an instrumental importance to focusing on achieving gender equality.
The case for the instrumental role of women in economic development has repeatedly been made convincingly by development economists. The seminal paper on this dates back to 1992, from the then chief economist of the World Bank, Lawrence Summers (1992), who gave a concise overview of the reasons why development requires “investing in all the people.” He concluded that “education of girls may be the highest return investment available in the developing world.”4 He gave a number of reasons for this. First, he pointed out that women are overrepresented among the poor, and that, in particular in South Asia, there exists a “more general pattern of female deprivation” resulting in the infamous phenomenon of the millions of “missing women” noted by Amartya Sen.5 This is the result of the fact that fewer girls are born—or are registered at birth—than would be expected on the basis of the normal ratio between girls and boys. They are missing from the birth records because they have been aborted or die shortly after birth (due to infanticide or neglect).
Such practices are usually related to the patriarchal traditions of these societies, where girls leave the household of their parents after marriage, and therefore do not take care of them during old age. Boys, on the other hand, continue to live with their parents. Their wives move to live with the groom’s family and are then supposed to care for their in-laws. This gender bias leads to a vicious cycle (p.3) of “underinvestment in girls.” Summers noticed that “increasing educational opportunities for girls offers the best chance of breaking the vicious cycle. . . . Among other things, women spend more of their income on children than their husbands do, and educated women are more likely to seek medical care and improve sanitation practices.”6 And perhaps most importantly: “Educated women choose to have fewer children and can provide more for those they do have.”
Summers’s pioneering paper was symbolic of the changing views on the role of women in economic development that occurred during the 1980s and ’90s. It has led to plenty of new research on this issue, demonstrating the key role played by women—and by the education of women.7 This culminated in the World Bank’s World Development Report of 2012, which focuses on gender equality and development. The report presents three main arguments for why gender equality matters for economic development. First, as just mentioned, gender equality matters in its own right. Following Amartya Sen, it argues that development can be seen as a process of expanding freedoms for all people. From that perspective, achieving gender equality is a goal in itself: “just as development means less income poverty or better access to justice, it should also mean fewer gaps in well-being between males and females.”8 The second argument is that it is “smart economics” to improve the position of women: in most developing societies (and in many developed ones, as well) women’s skills and talents are underused and misallocated because they face discrimination in markets and institutions that prevent them from completing their education, entering certain occupations, and earning the same wages as men. The economic (and social) losses are huge, and eliminating the barriers that keep them from realizing their capabilities will have strong, positive effects on development.9 Third, women’s endowments, agency (the ability to make meaningful decisions about their life course), and opportunities shape those of the next generation. Women have a large—and usually decisive—say in decisions about reproduction. They play a key role in deciding how many children are “wanted” and how these are “spaced” (of course, given the constraints of nature);10 as we will see, it is their position and human capital that in most societies will determine the level of fertility and hence population growth. Moreover, women play a key role in raising the next generation in their transfer of values and “human capital” to their children. Improvements in women’s education and health have strong positive impacts on these and other outcomes for their children.
(p.4) Summing up, women in practically all societies play a crucial role in steering economic and demographic change, or have the potential to do so; they usually do this from a relatively subordinate position, in that they generally do not own the big companies or dominate politics, but if they did, they might be even more influential than those who currently do.
With this book we intend to give further support to the theory that the position of women in society matters for economic progress, presented on the basis of evidence from long-term European history. The historical situation of Europe can be seen as a case study to determine which parameters really matter for economic growth and female empowerment, allowing for future comparison to other societies. The degree to which women are allowed to play an influential role in the economy and society varies greatly from one society to another. This is largely determined by the power balances within households, between men and women, and between generations. These power balances are embedded in complex ideologies about the desired role of men and women in society, and are influenced by economic opportunities and incentives for both sexes. Access to wage labor, for example, may have a strong emancipatory effect and could strengthen women’s position. Even culture—and the cultural ideas about the roles of men and women in family and household—is malleable and subject to fundamental change. But the current world economy is characterized by strong differences in the distribution of power between men and women, which has important consequences for the pace and character of economic change in various parts of the world.11
In this introduction we will not as yet address the big question of what the causes of inequalities are between men and women in different societies,12 but will first briefly sketch the global differences in these power balances between men and women, and demonstrate how these can be applied to evaluate societies in the past. Next, we turn to the question of their consequences, and focus on the role of reproduction and human capital for economic change.
Marriage Ages and the Global Dimensions of Gender Inequality
How can we study female agency and gender inequality that existed in the past? Today, institutions such as the United Nations and the World Economic Forum collect datasets covering diverse dimensions of gender inequality, but for the (p.5) more distant past such data are often unavailable. In this book, we have taken a different approach by focusing on marriage—and in particular, on the age at which women and men marry—as reflecting the power distribution between women and men, as well as that between parents and children. This is not self-explanatory, so in this section we sketch the reasoning behind our use of this indicator.
Marriage is a topic that has been analyzed by a wide range of scholars, from economists, to anthropologists, to sociologists.13 The different ways that marriages are organized around the world reflect the differences in male–female unions. Marriage plays an essential role in family formation, and as such is a key determinant of investments in the next generation. The norms surrounding marriage form part of a wider set of family values that govern the transfer of property, power hierarchies, and the responsibilities of relatives. Together, the various societally prescribed values to do with marriage make up a marriage system.
The age of marriage of women is closely linked to the marriage system: when marriage is arranged by the parents (or the broader family), girls are usually married off at a very early age. Societies in which arranged marriage is the norm are, for all sorts of reasons, fundamentally different from societies in which the younger generation gets to choose his or her own partner. The transition from the one system to the other is one of the most fascinating and important changes in our history. In Western Europe, this transition occurred very early on (or alternatively, parts of this region never developed a fully patriarchal system); in some areas, it was already underway in the Middle Ages and resulted in the emergence of the European Marriage Pattern, or EMP (as identified by Hajnal in 1965).14 Why this was the case and the consequences for societal and economic development belong, in our view, to the big questions of European history that we address in chapter 2. It resulted in a situation in which men and women marry relatively late (often older than 24), with a very small age gap between spouses (one to three years, on average). Normally, marriages would not be arranged; individuals sought their own partners, and this took time. In addition, it was only possible to begin this search for a partner at a certain age. Moreover, the “search process” in the marriage market was targeted at more or less equal partners, resulting in the small age gap.
The example of the EMP illustrates how marriage ages of men and women, and the difference between those ages, can be an indicator of different balances of power—balances between the sexes and between parents and their children. In a system of arranged marriages, there is a tendency to marry girls off at very (p.6) early ages—sometimes as young or younger than 12. The influence the bride has on this decision is minimal, if not nonexistent; she has, in other words, (almost) no “agency.” The rationale underpinning this system is explained in a rather whimsical fashion in the following answer by the regent of Banten (on Java) to the question of why girls in the region married at such a young age:
If you allow a girl to grow up freely until she has reached the age when her passions are aroused, you run the risk . . . of her choosing a person with whom she will never enjoy a happy marriage. You can prevent this by compelling her to marry another man, but as they say in North Banten: “[W]hen the heart is baked, don’t try to knead it.” . . . [Y]ou increase the chance of destroying marital fidelity. . . . It’s a completely different story for girls between the ages of seven and ten, who have yet to experience the feeling of love.15
The point is that, to a certain degree, the balance of power between (young) women and their parents—their families—can be derived from the marriage age of women. Marriage at an older age can have far-ranging implications, as Claudia Goldin sketches in her 2006 Richard Ely lecture. Women who marry later have a chance to form an adult identity that may include a career or profession, while those who marry very early will likely see themselves as secondary workers, optimizing time allocations to accommodate their husbands’ labor-market decisions.16
The relationship between husband and wife within a marriage is also, to a certain degree, reflected in the age difference between the partners. In Antiquity, men would often marry only after the age of 30 and wed much younger women (12–17 years old), which suggests a substantial degree of inequality.17 Results from present-day Bangladesh suggest that smaller spousal age gaps are associated with better status for brides, measured by socioeconomic factors.18 Contemporary work has also shown that the differences in status that result from a wide age gap between husband and wife can result in increased exposure to domestic violence and a power dynamic that undermines wives’ agency.19 Casterline et al. find that when it comes to the determinants of spousal age gaps across countries, explanations based on shortages of women in the marriage market do not suffice.20 Rather, they conclude that spousal age gaps seem to be (p.7) a reflection of preferences for a given age gap, which they suggest is in turn determined by family organization and the position of women. The EMP, which evolved during the Middle Ages, was characterized by a relatively small spousal age gap, and as already explained, by a relatively late age at marriage not only for men (comparable to the age in Antiquity) but also for women. Especially, this last aspect is notable: societies with high marriage ages for men are widespread, but societies with high marriage ages for women are much more unusual.
Studying marriage ages, therefore, gives us two measures of the degree of “agency” women have: their absolute age at marriage (the higher, the more “agency”) and the spousal age gap (the lower, the stronger the position of women in the household). The combination of these parameters brings together two domains—freedom of choice in whom to marry and equality within marriage—based on hypotheses that we developed earlier regarding the evolution of the EMP.21 Here, we wish to explore the validity of the hypothesized link between marriage patterns and female agency by turning to contemporary indicators of the position of women.
In the past decade, international organizations, such as the World Bank, have developed indices to evaluate the position of women on a worldwide scale and to enable comparisons internationally. The question that we ask ourselves is whether there is currently a relationship between the position of women in these different parts of the world and their age at marriage. Do women in countries where they have a better position also marry later than those in countries where they cannot decide on their future, where no investment is made in their education, and their health is poor? Can we, in other words, use data on marriage age in the past to gain insight into the “agency” of women in the past?
In order to compare measures of marriage patterns with more elaborate measures of the position of women, we use the Gender Inequality Index, or GII.22 This index first appeared in the UN Development Program’s 2010 Human Development Report, calculated on the basis of data for 2008.23 The GII examines three different aspects (or dimensions) of gender inequality: participation in the labor market, “empowerment,” and reproductive health. It gives an indication of the level of development as it relates to the inequality between men and women, measured along these three specified dimensions. It varies from 0 to 1, where 0 means no inequality between men and women and 1 means total (p.8) inequality. The GII report for 2010 demonstrates that the gender aspect with the greatest impact on human development is reproductive health, as measured by maternal mortality during childbirth and the adolescent fertility rate.
Figure 1.1 shows the correlation between the GII and female singulate mean age at marriage (SMAM), and figure 1.2 shows the correlation between the GII and the Girlpower index (the latter calculated by subtracting the spousal age gap from the female SMAM). These correlations are both remarkably high, even though marriage age is not part of the construction of the GII, and therefore (p.9) cannot be an explanation for the high level of correlation. Algeria has a higher female SMAM than would be expected on the basis of its GII score. The most extreme outlier is Jamaica; Jamaica was ranked eighty-ninth in the world when the GII is used but ranked first in the world when female SMAM is used.
The high average marriage age of Jamaican women is difficult to explain. It could be related to the phenomenon of “consensual unions,” a practice common in Jamaica. In such a system of consensual union, couples live in an unwedded state until they have achieved economic stability, after which they marry. Keith F. Otterbein, in an article on Caribbean family organization, describes the dominant role of women in such a system. Men have, according to a number of authors, a low economic position, which means they must delay marriage and often do not play a significant role in the raising of children.24
Empowerment and Economic Development: Reproduction and Human Capital
We approach the topic of female empowerment by studying changing marriage practices, not because we want to suggest that the place of women is in the household; quite the contrary, we think it is a fruitful way to explore the changing role and position of women in general. One of the vital elements in understanding the link between marriage/the household and economic performance is the recognition of human capital. A body of literature suggests that the reproductive activities of women can greatly affect economic development (see Dilli 2017 for an overview). Human capital is at the core of the knowledge economies that have developed over the past centuries in the West, and it is thus of utmost importance that we measure the human capital of the women involved and understand how their empowerment affects the next generation’s access to education. The relationship between gender (in)equality and human capital is, then, just another part of the whole cycle from reproduction to production that we study here, but in this section we focus on the role of human capital.
We start by briefly touching upon some debates about these issues so as to provide a theoretical grounding for the historical analysis that will follow. One of the main themes of that analysis derives from the economic analysis of households and demographic behavior, and it concerns the tradeoff between the quantity and the quality of offspring. Given the time constraints in household production—one can do only one thing at a time, and each activity therefore has its opportunity costs—the “production” of children is a costly affair. (p.10) A rather crude distinction that can be made (which is also used by biologists to study different reproductive patterns of behavior) is that between a strategy that concentrates on the quantity of offspring (thus does not invest much in the raising of the individual child) and a strategy that aims to optimize the quality of offspring. In the latter case, only a few children are born, but they receive a lot of attention from their parents. Becker famously argued that economic development requires such a switch from a quantity to a quality strategy.25 This has since become one of the core ideas of Unified Growth Theory, which tries to explain how an economy can move from a Malthusian equilibrium to modern economic growth.
In a Malthusian economy, people have as many offspring as possible and do not invest in their human capital. Economic growth will, in such a situation, induce more population growth, with the end result being that a population is not able to achieve a level of welfare above subsistence. Such improvement above subsistence is only possible when the quality strategy becomes more important: people then have fewer children and invest more in them.26 Put differently, economic development requires higher levels of human capital formation; according to Unified Growth Theory, the increase in skills, knowledge, and ideas is the most important source of long-term growth. And this requires investment by parents in the futures of their children. Summing up, there is a demographic cause of the transition from Malthusian stagnation to modern economic growth, which results in strong improvements in general welfare.
Women normally spend more time having (and raising) children than men. Therefore, it is the opportunity costs for women—and not those of men—that really matter here and that will have a large effect on the decision-making process that accompanies reproductive choices. The hypothesis that has received the most attention in this respect is that the human capital of women, which is obviously positively related to their earning power, is negatively related to fertility. This gendered version of the quantity–quality tradeoff has been most clearly developed by Diebolt and Perrin, who also tested the idea for France in 1851.27 Figure 1.3 gives an overview of this relationship as found in data for developing countries.
In figure 1.3, the average years of education for women are graphed against their “demand for children” (as captured by fertility). These data are for 80 countries (from Algeria to Zimbabwe), for which, for five-year periods between 1950/54 and 2000/04, estimates of the educational attainment of women (and men) are available. The data are then plotted against estimates (p.11) of fertility for the same countries and the same years.28 The strong correlation between the two variables, which is also confirmed by econometric research,29 is clearly evident. The levels of educational attainment vary from close to zero to ten years of schooling; women who are not schooled at all have 5.5 to 8 children, whereas highly educated women have fewer children (2.1 is about the level at which a population reproduces itself, a threshold that will be attained with more than six years of schooling).30 Different countries have different trajectories within this fertility/human capital space, as figure 1.4 shows, where the developments in the Philippines (with a very high level of fertility) are contrasted with those of India (with a relatively low level of fertility) and Kenya (which falls in between).
Figures 1.3 and 1.4 present the classic switch from a situation with low levels of human capital formation and high levels of fertility (the bottom right quadrants of the figures) to one with high levels of human capital formation and low levels of fertility (the upper left quadrants). The point is that the degree to which women have acquired human capital will determine the degree to which they prefer quality of offspring over quantity. Once the process is set in motion—and women begin investing more in their children—there will be (p.12) the tendency for every generation to further stress quality over quantity of offspring. Most countries therefore move slowly along the curve shown in figures 1.3 and 1.4.
Econometric research has also demonstrated that men’s increased human capital does not have the same effect on fertility. On the contrary, there are strong indications that when men earn more and have more human capital, they tend to demand more children.31 It is the education of women that drives the downward movement. One way to demonstrate this is to show the effect of difference in educational attainment between men and women has on fertility (figure 1.5). It appears that when the gender gap is very large (increasing to 100% difference in education), fertility is also high; an equal distribution of education between men and women seems to be associated with, on average, lower demand for children.
Why is it that well-educated women have fewer, more educated children? A number of mechanisms seem to be at play. As mentioned, economists who theorize about the family and fertility have argued that opportunity costs are involved. When women have higher human capital, their levels of productivity are higher and their wages are higher. In such a situation, it is more “expensive” for women to raise children in comparison to working. The crucial mechanism here is a changing “demand for children,” but there are other links as well. It is often suggested that education enhances the degree to which women are open to new ideas, including ideas about birth control.32 There are direct links between (p.13) education and reproduction and age at marriage, which brings us back to the point made at the beginning of this introduction.
Going to school usually conflicts with marrying at an early age—in particular in societies where marriage is tightly linked to reproduction (bearing children as early as possible). A low age at marriage is often a feature of societies in which arranged marriage dominates, in which parents directly control the marriage decisions of their children. Controlling the marriage decision of your children is easier when (in particular) daughters have not acquired too much human capital, which corresponds with a relatively young age. Older children are usually less inclined to obey their parents.33 So, in a system of arranged marriages, it may not be in the interests of the parents to send their girls to school. Thus, the gap in human capital between men and women may be much larger in societies where arranged marriages dominate, than in systems of consensual marriage. And a low age at marriage will normally lead to high levels of fertility (during the complete life cycle of the women). As Mason has demonstrated, there are many such links between women’s autonomy (relative to men) and female age at first marriage.34
Female agency, then, not only is linked to human capital but there is also an obvious connection with capital markets. Ever since the microcredit system (p.14) was developed as a way for the uneducated to rise out of poverty, women’s access to financial capital markets has received much attention. The focus of many modern initiatives in this field (such as the Grameen Bank) has been to provide small loans to groups of women that would enable them to set up their own business (see the discussion in chapter 4). Notwithstanding the critique the microcredit system has recently received, connecting women to financial markets is in our view an essential part of the story of female empowerment and economic growth. The most recent, very thoughtful assessment of the microfinance movement by Banarjee and Duflo also concluded that, despite certain limitations (not all men and women have the potential to develop the right entrepreneurial features, and it is unclear how one moves from microfinance to forming an employment-creating firm), it is an impressive movement that is reaching the poor.35 As we will demonstrate later in the book, the early modern form of economic growth in Western Europe, which led to the Industrial Revolution, could rely upon a well-developed capital market with remarkably low interest rates and a variety of financial instruments used intensively not only by the wealthy classes. Among the participants in those capital markets were a fair number of women who demonstrated an “investment behavior” that was little different from that of their male counterparts.
Another factor concerns female participation in the labor market. That is, the stronger the position of women in society, the more they are able to work outside the household—and vice versa: the more they participate in the labor market, the stronger their position in society. This again has a substantial effect on the age at marriage and on the opportunity costs women bear during marriage. In short, access to labor markets is a strategic link, connecting women’s autonomy, level of fertility, and human capital formation—a link that constitutes the focus of this book.
Why This Book?
Nowadays, states implement policies to increase women’s access to education, labor markets, and capital markets. Such top-down policies were absent in Europe’s past. Still, women participated in labor markets to a remarkable degree, and they gradually built up human capital as a result. As we will explain in this book, they also had a role in capital market activities. Both these developments were, we believe, linked to the rise of a particular form of marriage pattern of which we still see the effects in Europe. The European Marriage Pattern is (p.15) characterized by high numbers of single men and women, high marriage ages among men and women, and neolocality, meaning the couple forms a new nuclear household and does not reside with the parents of either the bride or the groom. We argue that the development of the EMP originated in the late Middle Ages and that it was one of the main causes of the gradual move away from a Malthusian state and toward an economy that was able to generate long-term economic growth. This story of female empowerment, therefore, is crucial for understanding the specific developmental path of the Western European economy, which resulted in the Industrial Revolution of the late eighteenth century.
The book relies largely on the work we have undertaken in the past few years exploring various aspects of the topics mentioned here. Much of this work has been published in other versions, but has not been given the global perspective we have sketched here, with clearer identification of the links between various societal processes and the mechanisms governing those processes. We start by bringing our ideas on female empowerment and economic development together in an explanatory framework based on the emergence of the European Marriage Pattern. In chapter 2, we argue that the EMP has played a fundamental role in Western Europe’s economic development, and we link this development to other aspects that have been discussed in this chapter. As we will explain, the EMP emerged in northwestern Europe in the late medieval period as a result of teachings of the Catholic Church on marriage based on consensus, the rise of labor markets, and specific institutions concerning property transfer between generations that enhanced wage labor for women. These influences resulted in a demographic regime embedded in a highly commercial environment, in which households interacted frequently with labor, capital, and commodity markets. We also discuss possible long-term consequences of these developments for human capital formation and institution building.
In the following chapters, then, we describe the effects of the EMP on the three areas we have mentioned here: human capital formation (chapter 3), access to the capital market (chapter 4), and participation in the labor market (chapter 5), thereby further testing and developing the central hypotheses as to the EMP’s long-term consequences. Dealing with these factors demanded that we develop different methods and use a multitude of sources. In chapter 3 on human capital, for example, we apply a fairly new approach to tracing human capital formation during the pre-statistical era. Instead of using the standard parameter of literacy to trace the development of human capital, we focus on numeracy, which can be retrieved via the age-heaping method. As we demonstrate, this approach yields an entirely different picture of the education of women, whether obtained via the normal schooling system or by a system of learning-by-doing via their activities in the economy. We present new evidence for the development of age heaping of men and women in Flanders and the Netherlands (p.16) in the late medieval and early modern period so as to test ideas based on previous research on the human capital formation of women and men in the North Sea area. We demonstrate that already in the sixteenth century, age heaping was gradually disappearing in large parts of the Low Countries, that (as we expected) there was almost no gender gap (women even outperformed men at times), and that differences in numeracy between town and country were quite small. This probably points to a very early rise of numeracy (or at least a “number sense”) in this region, linked to demographic change and commercial development in both town and country.
In chapter 4, we discuss female (and male) participation in the capital market, principally in regard to the functioning of the private capital market in Holland during the late medieval period. First, we argue that, in the absence of banks’ and state agencies’ involvement in the supply of credit, entrepreneurial access to credit was determined by two interrelated factors. The first factor was the quality of property-rights protection and the extent to which property could be used as collateral. The second factor was the rate of interest for borrowing money at the time, as well as occurrences of such borrowing compared with the interest rates on risk-free investments.
For our case study, which examines the small town of Edam and its hinterland, De Zeevang, in the fifteenth and sixteenth centuries, we demonstrate that properties were used as collateral on a large scale, and that interest rates on both small and large loans were relatively low (about 6%). This discovery is essential to understanding what made capital so accessible in early modern Europe, while it also links nicely to current forms of microcredit in which women play a central role. As a result of the well-developed capital market in our study area, many households (whether headed by men or women) had financial assets and/or debts, and the degree of financial sophistication was relatively high. Then, in the second part of chapter 4 we demonstrate that women were very active on this capital market and that their choice of financial instruments for investment was not much different from that of male investors.
Chapter 5 focuses on major changes in the position of women participating in the labor market between about 1400 and 1800. We look in particular at wages, and we document that women’s wages in the North Sea area were almost on a par with those of men during the late Middle Ages, but then fell to much lower levels during the seventeenth and eighteenth centuries (a result that has recently been confirmed by Humphries and Weisdorf36). We try to offer an explanation of this remarkable increase in the wage gap and also speculate about the long-term demographic consequences of this change.
(p.17) The next two chapters focus on what we have called the “dark side” of the EMP: the nuclear hardship hypothesis, as formulated by Laslett, that notwithstanding its potential benefits for economic growth, this particular marriage pattern resulted in relatively small households that were more vulnerable to exogenous shocks.37 We first develop the idea more fully, then argue that Laslett indeed had a point: the degree to which the “resulting” nuclear household could spread risks is limited, not just owing to the small size of families but also to the so-called double squeeze, whereby the older marriage ages of both parents and their offspring mean that the period during which parents and children require help and support in their lives occurs simultaneously (chapter 6). Hence, we would expect a negative impact on the lives—and life expectancies—of people living in early modern EMP areas. However, evidence on life expectancy and the occurrence and effect of famines in late medieval and early modern Europe appear to be inconsistent with this hypothesis: there is evidence that longevity was actually higher in the North Sea area than outside it, and that this region managed to suppress and/or cushion the negative effects of harvest failures more successfully than the rest of Europe.
Addressing this paradox is the subject of the following chapter, in which we look at the “institutional effect” of the EMP. In chapter 7, we concentrate on two examples of institutionalized forms of “solutions” that developed in EMP societies to avoid the hardship Laslett was referring to. We first take a fresh look at servanthood as part of a “commercial household” economy, which may have added flexibility to the EMP household. Whereas servanthood, in the literature, is often approached from the supply side—stressing the large number of single men and women in early modern societies—we view the large presence of servants in EMP areas from the demand side. That is, having a servant in the house could be a solution to the difficulties created by changes in household demographics.
Another phenomenon that may help explain part of the aforementioned paradox is discussed in the second part of chapter 7, which focuses on the Flemish beguinages, a movement of single women who chose to live in communities of equal-minded single women—communities that were in appearance similar to religious convents but in practice functioned quite differently. In their “radicalness,” beguinages perhaps illustrate the basic features of the EMP (and the cultural space in which the EMP emerged) even better than the marriage pattern itself. On the basis of a large dataset of these beguinages, we explore the emergence and persistence of this phenomenon in the Low Countries between 1300 and 1900. The explanations offered by the literature (gender imbalances in the (p.18) cities and lack of access to convents) are discussed, and an alternative interpretation stressing the voluntary nature of the movement, based on the desire of the women to remain single without entering a strict religious community such as a convent, is advanced.
In the concluding chapter, we enter the debate on how to measure the EMP—via the age of marriage of women or via the underlying institutions—and how to link the EMP with economic development in Europe between 1300 and 1800. We respond to criticism by two scholars, Tracy Dennison and Sheilagh Ogilvie, who have argued that a clear link between the EMP and economic growth cannot be established.38 To make our point, we broaden the scope of the book by including the whole of Eurasia in our analysis. Finally, we formally test the central hypothesis of this book—that the marriage system of Western Europe, which facilitated a relatively high level of female agency—enhanced economic growth in early modern Europe.
Methodology and Sources
The book is based on a number of papers written by the authors, often together with authors from other disciplines. Annemarie Bouman, an anthropologist, contributed substantially to the analysis of the nuclear hardship hypothesis in chapters 6 and 7; economist and economic historian Alexandra de Pleijt conducted much of the formal testing of the link between the EMP and economic growth presented in chapter 8, and the medievalist Jaco Zuijderduijn (and research assistant Heleen Kole) helped to digitize and analyze the late Medieval sources from the Edam region that form the core of chapter 4. This illustrates the interdisciplinary approach of this project, bringing together ideas and data from economic and demographic history, anthropology, economics, and development studies.
But first and foremost, this is an economic view of a crucial development in European history: the genesis of the EMP in the Middle Ages and its implications for economic development in almost all its dimensions (labor, capital, human capital). The conceptual framework we have developed in chapter 2 to analyze these processes makes intensive use of data and sources related to the Low Countries and England, which together, we think, form the “core region” where the EMP emerged in the late Middle Ages. We not only describe it by using a number of indicators to picture these developments, such as marriage age and spousal age gap, but we also look into the (relatively) short-term and (p.19) (very) long-term effects of the EMP. Given the wide time span covered here, the reader will have to connect the many “dots” we are plotting, using a large variety of historical and present-day sources, with comparisons on different levels, from the very micro level of the individual to the broader macro view of global developments.
Moreover, we introduce and use a wide variety of historical sources to put some “flesh” on the bones of the theories we have set out, but we realize this, too, adds to the challenges we set before our readers. Our databases are based on many different types of sources, and some of them can be considered rather unusual in the field of economic history: from the quite common censuses and population registers, to age references on portraits (which have previously been ignored by historians and art historians alike) as useful sources for reconstructing age awareness and levels of numeracy; and from tediously collected references to women’s wages during periods when survival, more than bookkeeping, must have been in the minds of their employers, to data on literacy collected for a wide variety of countries around the world.
Our types of sources are commented upon in the chapters where they are used, so that their applicability for our study becomes clear. For practical reasons, we have mainly used sources related to the Low Countries, or the Netherlands and Belgium taken together, in those chapters where we test the hypotheses about the consequences and implications of the EMP. This is because the Low Countries were part of the core area of the EMP as it emerged in the late Middle Ages. This concentration on the Low Countries, and how it affects our data selection, is most evident in the chapter on numeracy, in relation to human capital formation. When using the age-heaping method to detect levels of numeracy, it’s important to use sources at the level of the individual person, not by household. This is because it is never certain whether the age recorded was reported directly by the individual, and in general the recording of ages is something not often found in historical sources. We did, however, find a substantial number of recorded ages for the Low Countries, and therefore we have used these records extensively.
To put this into perspective, for age-heaping data on other parts of Europe, we relied on existing studies and published sources. In contrast, in chapter 5, we paid a lot of attention to developments in the other part of the core area of the EMP—England—to sketch the long-term impact that EMP had on labor markets. The underlying idea is that the countries bordering the North Sea followed a more or less similar development path during the late medieval and early modern period; here, the EMP emerged in its clearest form and a process of long-term economic development began that would, by the late eighteenth century (in England), result in the Industrial Revolution. This “Little Divergence”—in a way, a spin-off of the “Great Divergence” debate, which focused on the separate development (p.20) paths of Western Europe versus Asia (China, in particular)39 —has now been established, based on quantitative research into real wages, GDP per capita, human capital, and political institutions.40 This book contributes to this debate by focusing on the micro dimension of human behavior, its institutions (such as the way in which marriage is organized), and its long-term implications.
We also present readers with a wide variety of methods used to analyze the long-term processes we study here. Which method is suitable depends on the research question we are answering in each chapter, on the causal relationships we are identifying, on the available materials, and on the level of detail needed. Our chosen methods are based on various statistical and econometric standards, but in principle these choices should not detract from our main results. Our biggest challenge was to sketch the important and influential mechanisms governing the behavior of both early modern and present-day Europeans in their daily choices. Many present-day institutions have their roots in changes that go back literally centuries, but we are often unaware how these changes are related to each other, let alone how they interact and may, in turn, lead to new mechanisms or undermine previously established norms.
(1) Important exceptions to this rule do exist, such as Jan de Vries’s (1994) work on the “Industrious Revolution.” However, recent syntheses such as Clark (2007b), Landes (1998), Pomeranz (2000), and Allen (2001) do not expand on this issue.
(22) In the working paper on which this section is based, we also compared marriage ages with the Global Gender Gap developed by the World Economic Forum, but the correlations with both the SMAM and the Girlpower Index were lower, albeit still positive; see Carmichael, De Moor, and van Zanden 2011a.
(23) In the tables, one can observe that the Netherlands scores highest on the GII.
(30) In fact, the elasticity of the curve presented in figure 1.3 is close to –1, meaning that one extra year of schooling implies one child less, a rule that also seems to apply to most of the countries when taken separately (see figure 1.4).