Published on Development Impact


This page in:
Development Impact logo

Family is arguably the oldest social institution. Families not only have an evolutionary function, but they also serve as economic units enabling their members to pool resources, share consumption, engage in joint production, and insure each other, among other things. Neglecting to incorporate the family into our analytical frameworks and data collection can result in, at best, an incomplete understanding of economic behaviors and outcomes, and at worst, erroneous conclusions.

So, how do we define Family?

Studies of family structure in developed societies have historically focused on the idea of the nuclear family, i.e., a monogamous married couple with their dependent children (or perhaps a single mother with children). This is an extremely restrictive definition and is not representative of the global variation in traditional and modern family structures. Natalie Bau and Raquel Fernandez adopt the following definition (based upon Brown et al, 2020) in a recent Handbook chapter on culture and the family: “A family can be defined as the smallest group of individuals who see themselves as connected to one another... Families tend to reside together and share economic opportunities and other rights and responsibilities.” Moreover, “the function of families is to fulfill basic human needs such as providing for children, defining parental roles, regulating sexuality, and passing property and knowledge between generations.”

As you might notice, this definition of family overlaps with that of a household. Often the terms “household” and “family” are used inter-changeably, yet it is important to recognize that they are not equivalent. In household surveys, we typically define a household as a group of individuals who co-reside and share a common kitchen (or cooking pot), regardless of blood or marital ties (as an aside: here is a fun thread by Sonalde Desai on the practical definitional dilemmas one encounters in the field during survey data collection). Clearly, this leaves out family members who do not cohabit. As well, individuals may co-reside with non-family members.

There are many different types of families.

Given the emphasis on the nuclear family structure, it is unsurprising that the theoretical and empirical literature in family economics has predominantly examined marital and parent-child relationships, with family formation coinciding with marriage or cohabitation. However, the majority of the world population does not live in a nuclear family arrangement, with the share of individuals living in an extended family being as high as 45% in the Asia-Pacific region. In fact, historically, the nuclear family has not been the predominant ancestral family type in much of Africa or South Asia (Bau and Fernandez, 2022).

Similarly, polygyny (i.e., the practice of a man having multiple wives simultaneously) has been highly prevalent historically, and remains quite common even today, in Sub-Saharan Africa. There is also substantial temporal and geographical variation in other practices that are correlated with family or household formation and co-residence patterns. For instance, in societies practicing patrilocality, sons typically reside with and provide elderly support to their parents after marriage. In contrast, in matrilocal households, parents tend to co-reside with their married daughters. It is evident that these familial arrangements have implications for, among other things, differential parental investments in sons and daughters (Bau, 2021).

Even if we ignore the historical variation in family structures, the concept of the “modern family” has also undergone significant evolution in recent times.  This evolution is marked by the emergence of alternative family arrangements, spurred by trends such as declining fertility rates, delayed marriage, increasing divorce rates, remarriages, and the legalization of same-sex marriage.

In fact, as David points out in this recent post, family structure may itself be endogenous as the very concept of the family can change with interventions and economic decisions (such as migration).

It is important for economists and policymakers to take into account family structure.

Ignoring the underlying family structure may lead to a flawed understanding of why identical policies have different effects in different contexts, and policymakers may end up with poorly designed policies that may be ineffective or have unintended consequences. This is nicely illustrated by Rossi (2018) who shows that it is crucial to take into account the family structure to analyze why the fertility transition in sub-Saharan Africa has been relatively slow. Fertility choices in countries like Senegal are made in the context of polygamous households where women’s status and their access to resources are determined by their relative number of children. Rossi argues that polygamy leads to competition between co-wives for more children and undermines fertility transition. If we ignore the family structure, we will fail to understand why supply-side family planning interventions that improve access to modern contraception may be ineffective in spurring fertility decline.

Additionally, a focus on couples in isolation overlooks the significant influence wielded by family members other than the married couple in various household decisions in non-polygamous societies. For instance, in South Asia, mothers-in-law play a predominant role in women’s family-planning and reproductive decision-making (Anukriti et al 2023, 2022, 2020). In fact, as the likely matriarch of the household, a woman’s mother-in-law may play an even stronger role than her husband, especially during the early years of an arranged marriage. Limiting our analysis of fertility decisions to bargaining only between the wife and the husband leads to an incomplete understanding of the intrahousehold decision-making process. Moreover, well-intended interventions that exclusively target women, or married couples, risk being ineffective if they overlook the status and authority of mothers-in-law within the household, especially when preferences and incentives between women and their mothers-in-law are misaligned.

The nuclear-family bias also influences what type of data is collected. For instance, the Demographic Health Surveys measure women’s intrahousehold decision-making power solely in relation to their husbands. Similarly, data regarding the preferred number of sons and daughters is typically gathered exclusively from husbands and wives.

Figuring out the ties that bind.

The two examples above are drawn from very different contexts, but the lesson is generalizable. Studying socioeconomic behaviors and outcomes of individuals may require understanding household and family dynamics well-beyond the concept of a nuclear family or married couples. We need more nuanced conceptual frameworks, better data, and more research that can explain the diverse array of family configurations observed worldwide. In a subsequent blog, I hope to share some ideas on how we can attempt to do so. 

S Anukriti

Senior Economist, Development Research Group

Join the Conversation

The content of this field is kept private and will not be shown publicly
Remaining characters: 1000