Home Contact us Site Map

Cette page existe uniquement en anglais

Home > Activities > Committees > Seminar Report

Conference on Demography and Poverty

Florence, Italy, 2-4 March 1995
Organised by IUSSP, uNICEF International Child Development Centre (ICDC) and the university of Florence


The IUSSP Seminar on Demography and Poverty, co-sponsored by IUSSP, uNICEF International Child Development Centre (ICDC), and the university of Florence took place in Florence, Italy, on March 2-4, 1995. Fittingly, the seminar was held at the home of the ICDC, the Spedale Degli Innocenti, built during the Renaissance by the master architect Filippo Brunelleschi as an institution for the care of orphaned and abandoned children.

The purpose of the meeting was to shed new light on an issue that is central to economic demography: the relationship between demographic behaviour and poverty, a topic that has a tendency to be dominated more by ideology and preconception than by empirical evidence. The seminar demonstrated the usefulness of greater collaborative research in this area, in terms of defining and measuring the various dimensions of poverty, sharing data and techniques for analysis, and developing appropriate criteria and methods for measuring the impact of poverty-reduction programmes on demographic behaviour. The seminar was divided into six sessions:

Session 1: introduction

In the first session of the seminar the definition and quantitative measurement of poverty were examined and a basic framework for studying the relationship between population and poverty was laid out in order to provide useful background to the remainder of the seminar. There are three distinct approaches to examining the population-poverty link: a descriptive approach (best suited to answer questions about the number and characteristics of the poor); a macro-level approach (best suited to address issues surrounding the relationship between population growth (best suited to more in-depth attempts to understand how demographic phenomena and behaviours affect the ability of individuals, families or groups to escape poverty) (Livi Bacci).

Being poor is, at first glance, a relatively simple concept; but poverty assessments are typically subjective and clouded in conceptual and methodological uncertainties. If one uses a measure based purely on income, where should the poverty line be drawn? Should other dimensions of poverty be considered? There are many pitfalls to using cross-sectional data on income to measure poverty and well-being (Anand and Morduch). using data from Bangladesh, Anand and Morduch illustrated how relying solely on current income as a measure of poverty leads to the counter-intuitive conclusion that the poor impoverish themselves further by having a large number of children. Furthermore, poverty is not necessarily a chronic condition, and individuals and families may be able to move into or out of poverty over their life-cycle, often as a result of demographic events.

Session 2: population change and poverty : long-term issues

The second session consisted of four papers that examined, to varying degrees, the magnitude and nature of the long-term relationship between population change and poverty. While the alleviation of poverty and inequality is generally considered a high priority in the development process, there is considerably more debate about whether population growth contributes to, hinders, or has little net effect on this laudable goal. As several participants observed, sorting out the nature of the interaction between population and poverty is extremely complex, and much of the research conducted on this topic to date has not succeeded in identifying a set of causal connections between demographic and economic changes, making definitive conclusions difficult (Lipton). Furthermore, when examining relationships at this level of generality, it is important to consider trends broken down by race as well as identifying certain vulnerable groups, such as children (Bairoch).

In an attempt to inject fresh insight into this complex morass, authors in this session used aggregate-level data to examine historical and contemporary inter-relationships between population growth, poverty, and the environment. Lipton constructed a macro-level data set of contemporary indicators of poverty and key demographic variables to explore the contemporary nature of the correlation between poverty and demography. As Lipton points out, the period from 1955-1980 has witnessed both unprecedented disimpoverishment and rapid population growth that has led to a fair amount of revisionism on the part of analysts, particularly with regard to generalisations about the adverse effects of population growth that once characterised the population debate. Lipton finds that the data support the rationale-choice model of human behaviour but there is still a role for government intervention due to market failure and market imperfections, and problems of externalities to childbearing and coordination failure.

using data from three critical periods in demographic history, Bairoch tested the neo-Malthusian hypothesis that rapid population growth is accompanied by declining real wages and increasing inequality. Bairoch noted that events around the 14th Century (when there was a strong decline in the total population) and the beginning of the Industrial Revolution (when population growth was very rapid) match fairly closely a priori expectations. However, the relationship breaks down when one examines the events that took place in the late 1970s - when population growth in developing countries fell below 0.7 percent for the first time in 200 years. In this latter period, there was a large increase in poverty and inequality in the uS and among OECD countries.

The other two papers in this session attempted to unravel some of the complex interrelationships between population, poverty and the environment. Despite the interest in and the importance of this topic there is a relatively small body of carefully designed research in this area and the exact nature of the interaction remains quite elusive. The poor are often blamed for a large share of environmental damage but this characterisation is unfair because, individually, the poor have the least amount of power to damage the environment. In fact, in urban areas, the poor are avid recyclers so without their efforts the environmental degradation might be much worse (Satterthwaite). Furthermore, there is a common misconception that cities with the poorest levels of air quality, water pollution, and the highest build-up of hazardous and toxic wastes have the highest levels of environment-related diseases. This is not always the case, however, because the prevalence of many environment-related diseases is strongly determined by access to piped water, and decent sanitation and health care which are not necessarily strongly correlated with such things as air quality, water pollution, or hazardous/toxic wastes (Satterthwaite).

Encompassing over 3,200 square kilometers, India is the sixth largest country in the world in terms of surface area, and the second largest in terms of population size. It is also one of the world’s poorest countries with an estimated GNP per capita of $470 in 1992 (World Development Report, 1994); consequently, much can be learned about the relationship between population, poverty and the environment from a closer examination of the Indian evidence. In the 1970s the Ford Foundation commissioned an 8-volume report on the relationship between population growth, poverty and the environment in India entitled, the Second India Study. In the final paper in this session Repetto returned to the Second India Study to examine to what extent the predictions offered in the original magnum opus were borne out in reality. Repetto concluded that '... the evidence that has emerged in India over the 20 years since the Second India Study was carried out suggests that a rapid demographic transition, poverty alleviation and development can occur simultaneously even in poor and populous countries. But unsustainable spirals of population growth, environmental degradation and impoverishment are also possible. Population growth such as India’s makes it harder to reduce underemployment and poverty, raise educational levels and environmental quality, or provide adequate infrastructure and basic services, but it doesn’t make these tasks impossible. The key determinant of success seem to lie in the social framework, broadly conceived to include the system of norms, economic arrangements, policies and institutions of governance' (Repetto).

Session 3: demographic behaviour and poverty

Papers in the third session dealt with micro-level relationships between demographic behaviour and poverty. Three of the authors concentrated on the possibility of causation running from demography variables to poverty outcomes (Gage, David, Lloyd), and one author concentrated on the direction of the causal chain running from poverty variables to demographic outcomes (Garcia).

Two papers in this session dealt with the poverty consequences of family building patterns. David explored the evidence that high fertility and short spacing of births have negative consequences for the life-chances of family members, especially children. These negative outcomes include higher morbidity and mortality, and worse nutrition. However, her review highlighted the unresolved problem of unobserved heterogeneity in many of the models that have produced much of the evidence to date and underscored the need to collect better data on the socioeconomic status, living condition and other aspects of social functioning of families - data that are related not only to current household status but also to circumstances in the early stages of family formation and in the family background of parents (David).

For a variety of reasons, teenage mothers tend to be disproportionately represented among the poor and among those dependent on PUBLIC assistance. Gage’s paper addressed the social implications of adolescent pregnancy and childbearing, a topic of intense debate and controversy in recent years. Gage reviewed the evidence on the potential negative consequences of adolescent fertility including, inter alia, lower educational attainment, poorer employment opportunities, lower social mobility and poorer wages. Gage also reviewed some of the medical consequences of teenage pregnancy for both mothers and children, including unsafe, sometimes fatal, abortions in many countries where abortion is illegal or restricted. Drawing mostly from the Sub-Saharan African and North American experience, Gage pointed out that the social context of adolescent fertility varies widely across countries and continents, and that these diverse cultural and social circumstances strongly influence our perceptions of the consequences of early childbearing, a theme that was expounded in a recent report from a uS National Academy of Sciences’ panel on which she served.

In a different paper in this session, Lloyd reviewed the complex relationships between household structure and function on the one hand, and the well-being or poverty of individuals in a household on the other. As Lloyd pointed out, there are several important dimensions of household economics that affect the relationship between household structure and poverty including: (1) the determinants of household formation and affiliation; (2) the existence of economic links between households; and (3) the distribution of resources within households. Lloyd found considerable variation in individual household affiliation over time, and large differences in resource flows between and within households. Consequently, she concluded that there is nothing inherent in a household’s structure that either predicts poverty or, alternatively, promises resource adequacy and that, in the future, poverty should be assessed directly at the individual level, rather than the household level.

The final paper in the session focused on changing patterns of urban poverty in Mexico and Latin America and families’ responses to the decline in living conditions. Garcia documented the significant deterioration in economic conditions in Latin America in the 1980s that resulted in more households being classified as below the poverty line in the mid-1980s than was the case in the 1970s. This reversal in the continent’s economic fortunes has led to changes in family dynamics that may either alleviate or accentuate the conditions of poverty. The most common family survival strategies in Latin America include, inter alia, increased labour force participation of family members (particularly women), an increase in the mean number of jobs held per family member, modified patterns of family consumption, internal and international migration of family members, and establishment or reactivation of networks of assistance between neighbours and relatives. At the same time, the decline in living standards has also been associated with growing domestic violence - a phenomenon that Garcia argued is more widespread than is often realized - and increasing family instability.

Session 4: mortality and poverty

The fourth session consisted of three papers that concentrated on ways in which poverty can influence demographic outcomes. Studies of the relationship between poverty and demography have focused traditionally on the relationship in the other direction: namely that population growth may cause poverty and analyses of the impact of poverty on demographic variables are rarer. As noted above, one of the major problems with this type of analysis is trying to identify specific causal mechanisms, which are extremely complex and usually require access to accurate time-series data.

One example of the complexities involved in jointly modelling demographic and economic behaviour is the relationship between household mortality and poverty. Most studies that focus on the effect of poverty on mortality concentrate on attempting to better understand how household income or maternal and paternal education affect the probability of a child’s survival. Few studies take a household level approach and study how household well-being affects the probability of death in the family. The latter approach is more complicated because of the direct feedback following the death of an adult in the family: i.e. the loss of an adult member becomes a cause of poverty rather than an outcome of poverty, particularly if he or she is the main wage earner in the household. In this case, the death of an adult family member simultaneously increases the dependency ratio and decreases the household’s per capita income. D’Souza’s paper focused on one particular aspect of this problem: namely the economic impact of the death of a husband on a widow. Based on preliminary analysis of a recent poverty study conducted by the author in various parts of Kinshasa, Zaire, D’Souza reported that the death of a husband can often lead a widow into poverty and force children in the household to abruptly abandon their formal education.

A second paper in this session discussed the increasingly important links between AIDS and poverty (Basu). Many factors are responsible for the rapid spread of the AIDS epidemic. Although AIDS appeared to have started out as a disease of affluence, it is becoming increasingly clear that AIDS has become an outcome of conditions that constitute various dimensions of poverty - economic hardship, sexual powerlessness, poor health services and perhaps biological vulnerability. Basu’s paper points out how the poor are disadvantaged with respect to both the risk of acquiring the infection and the risk of mitigating the household impact of the disease. Several, and in some cases overlapping, categories of the poor were identified as being particularly vulnerable on both these counts - geographic areas, migrants, women, children and drug-users. Basu also pointed out the weakness of many countries’ curative health services and argued that an inadequate supply of clean syringes and unsreened blood products in these countries result in a certain number of avoidable new infections every year.

As was noted above, poverty does not necessarily have to be a permanent state. The final paper in this session considered the potential consequences of acute short-term poverty, specifically famine, on long-term endemic poverty. Osmani reviewed several Malthusian and neo-Malthusian hypotheses and concluded that the short run demographic and non-demographic effects of famine may have some irreversible consequences for the individual, the family and the society, especially among the poorest segments of society who cannot easily rebuild their assets following a famine because they have limited ability to save or borrow money. Other irreversibilities pertain to the long-term physical and mental development of individuals as a function of childhood malnutrition, the structure of the family and the structure of the moral economy.

Session 5: coping with poverty: demographic outcomes

Session 5 focused on the demographic effects of coping mechanisms by which individuals, households and communities handle the problem of chronic or transient poverty and vulnerability. The rapid transition of Eastern Europe has been accompanied by a sudden fall in industrial output and rapid impoverishment of large sections of the population and has provided demographers with a natural experiment to observe the magnitude of demographic responses to rapid and severe impoverishment. Cornia and Paniccia document how between 1989-1994 the transition in Eastern Europe has triggered a large demographic response: marriage rates have fallen by between 13 and 40 percent and birth rates have fallen by up to 50 percent. Death rates have risen significantly except in Central Europe. By 1993, as a result of these demographic changes, population size was declining, often sharply, in practically all countries of the region. In Russia - where male-life expectancy at birth had fallen to the same level as Pakistan - population size is envisioned to contract substantially by the end of the century (Cornia and Paniccia).

Another common coping response to poverty is migration. Although this also occurred in Eastern Europe, it was not the focus of the paper by Cornia and Paniccia. However, using information from surveys conducted in six sub-Saharan African countries between 1986/7-1992, Wery explored the complex relationship between poverty and migration. unlike the uSSR, this is an area with chronic long-term economic problems and where many governments have been forced to implement stabilisation and adjustment policies under the watchful eye of the International Monetary Fund. Wery identified two types of adaptation: through out-migration and through changes in labour force participation. In turn, these changes affect the rates and timing of demographic events or strategies. For example, Wery documented how out-migration can affect, inter alia, the production and consumption patterns in the household, the labour force participation of other household members, the dependency ratio and the local price of dowry. In turn, these changes can drive up the age at first marriage and the timing of childbearing.

A wealth of evidence supports the assertion that hunger and malnutrition remain the most devastating problems facing the majority of the world’s poor, and the last paper in this session reviewed the nature of the link between poverty and nutritional security (Ferro-Luzzi and Branca). Twenty percent of the population of developing countries does not receive enough food and have to adapt to energy levels that are insufficient to maintain their health. 786 million people do not consume enough food daily to undertake even light physical labour (Ferro-Luzzi and Branca). Poverty can lead to nutritional insecurity, which is turn can trigger a number of adaptive biological responses, both voluntary and involuntary, including weight loss, reduced physical activity and increased metabolic efficiency. At the same time, food deprivation can cause temporary or irreversible deterioration of the productive capacity in man, creating a feedback loop to more permanent poverty - an argument that will be familiar to students of the theory of efficiency wages as an explanation of wage-rigidity and long-term unemployment in urban labour markets in developing countries.

Session 6: roundtable discussion

The final session of the seminar consisted of a roundtable discussion on future research and data priorities for expanding our understanding of the complex interrelationships between demographic and poverty variables. Many of the participants argued that there was an urgent need for more interdisciplinary and multi-dimensional work, similar to that generated at this seminar. There was also an impassioned plea for future work to account explicitly for the relationship between women’s position and various demographic and economic outcomes.

There was general agreement about the need for more high-quality micro-level case studies that are firmly grounded in theoretical concepts, i.e. those that set out to specify and test particular causal mechanisms between poverty, however defined, and demographic outcomes. Doubtless, this will require both new and creative thinking on the part of researchers and more and better data, particularly data that can match demographic events not only to current household status, but also to circumstances in the early stages of family formation and in the family background of parents.

Finally, it should be noted that the seminar coincided with the first Innocenti lecture, delivered by Professor Amartya Sen on the second day of the seminar. This new lecture series is being sponsored by the Istituto degli Innocenti and aims at improving our understanding of some of the most pressing problems facing children, at raising collective awareness of the magnitude of these problems, and at making concrete suggestions about the actions needed for their solutions.

Barney Cohen

IUSSP, 3-5 rue Nicolas, F-75980 Paris cedex 20, France
Tel  +33 1 56 06 21 73 - Fax +33 1 56 06 22 04 - contact us